NCP Nutrition in Clinical Practice

Volume 33 • Number 4 • August 2018

MANAGING GI DISORDERS
Bacteria, Bones, and Stones: Managing
Complications of Short Bowel Syndrome

Dietary Fermentable Oligosaccharides, Disaccharides,

Monosaccharides, and Polyols (FODMAPs)
and Gastrointestinal Disease

Nutrition Management of Necrotizing Enterocolitis

Enteral Nutrition in the Management of Crohn’s

Disease: Reviewing Mechanisms of Actions
and Highlighting Potential Venues for Enhancing
the Efficacy

Enteral Nutrition for Pediatric Crohn’s Disease:

An Underutilized Therapy

Case Report of Wernicke’s Encephalopathy After

Sleeve Gastrectomy

Copper Deficiency Myelopathy After Upper

Gastrointestinal Surgery

Peer-reviewed, practical solutions in clinical nutrition wileyonlinelibrary.com/journal/ncp

ISSN: 0884-5336

NCP_33_4_cover000.indd 1 29/06/18 6:59 PM

 Nutrition in Clinical Practice
EDITOR-IN-CHIEF
Jeanette M. Hasse, PhD, RD, LD, FADA, CNSC
Dallas, TX
ASSOCIATE EDITORS
Roland N. Dickerson, PharmD, BCNSP Mary S. McCarthy, PhD, RN, CNSC, FAAN
Memphis, TN Tacoma, WA
Mary Marian, DCN, RDN, CSO, FAND Jayshil Patel, MD
Tucson, AZ Milwaukee, WI

CONTRIBUTING EDITOR
Ryan T. Hurt, MD, PhD
Rochester, MN

ASPEN STAFF
Colleen K. Harper, CAE Catherine Wattenberg Catherine J. Klein, PhD
Interim Chief Executive Officer Director of Publications Publications Manager

EDITORIAL BOARD
Teruyoshi Amagai, MD, PhD Ronald L. Koretz, MD Carol J. Rollins, MS, RD, CNSC,
Nishinomiya, Japan Granada Hills, CA PharmD, BCNSP
Tucson, AZ
Deborah A. Andris, MSN, APNP Debra S. Kovacevich, RN, MPH
Milwaukee, WI Ann Arbor, MI Mary K. Russell, MS, RDN, LDN
David A. August, MD, FACS, CNSP Kenneth A. Kudsk, MD Chicago, IL
New Brunswick, NJ Madison, WI
Denise Baird Schwartz, MS, RD,
Albert Barrocas, MD, FACS Laura E. Matarese, PhD, RD, LDN, CNSC, FADA, FAND, FASPEN
East Point, GA FADA, FASPEN Studio City, CA
David L. Burns, MD, FACG Greenville, NC
David S. Seres, MD, ScM, PNS
Burlington, MA Carol McGinnis, DNP, RN, CNS, CNSC New York, NY
Yvon Carpentier, MD Sioux Falls, SD
Brussels, Belgium Pierre Singer, MD
Remy Meier, MD Tel Aviv, Israel
Isabel Correia, MD, PhD Liestal, Switzerland
Minas Gerais, Brazil Annalynn Skipper, PhD, RD, FADA
Sarah J. Miller, PharmD, BCNSP
Chicago, IL
Gail Cresci, PhD, RD, LD, CNSC Missoula, MT
Cleveland, OH Krishnan Sriram, MD, FRCS(C),
Charles M. Mueller, PhD, RD, CNSC, CDN
Robert S. DeChicco, MS, RD, New York, NY FACS, FCCM
LD, CNSC Chicago, IL
Cleveland, OH Gerard E. Mullin, MD, CNSC
Baltimore, MD Dan Waitzberg, MD, PhD, FASPEN
Rose Ann DiMaria-Ghalili, PhD, RN, CNSC São Paulo, Brazil
Philadelphia, PA Lynne M. Murphy, MSN, RN
Washington, DC Jacqueline Wessel, MEd, RDN, CNSC, CSP,
Beverly J. Holcombe, PharmD, BCNSP,
Luis Nin, MD CLE, LD
FASHP
Montevideo, Uruguay Cincinnati, OH
Ocean Isle Beach, NC
Caroline M. Kiss, RD, DCN Kim Robien, PhD, RD, CSO Patricia H. Worthington, RN, MSN, CNSC
Basel, Switzerland Washington, DC Philadelphia, PA
NUTRITION IN CLINICAL PRACTICE (Print ISSN: 0884-5336; Online ISSN: 1941-2452) is published bimonthly in
February, April, June, August, October, and December on behalf of the American Society for Parenteral and Enteral Nutrition
(ASPEN) by Wiley Subscription Services, Inc., a Wiley Company, 111 River Street, Hoboken, NJ 07030-5774 USA. Periodicals
postage paid at Hoboken, NJ and additional mailing offices. Postmaster: Send all address changes to Nutrition in Clinical Practice,
John Wiley & Sons Inc., C/O The Sheridan Press, PO Box 465, Hanover, PA 17331 USA.
COPYRIGHT  C 2018 by ASPEN. All rights reserved. No part of this publication may be reproduced, stored or transmitted

in any form or by any means without the prior permission in writing from the copyright holder. Authorization to copy
items for internal and personal use is granted by the copyright holder for libraries and other users registered with their local
Reproduction Rights Organization (RRO), e.g. Copyright Clearance Center (CCC), 222 Rosewood Drive, Danvers, MA 01923,
USA (www.copyright.com), provided the appropriate fee is paid directly to the RRO. This consent does not extend to other kinds
of copying such as copying for general distribution, for advertising or promotional purposes, for republication, for creating new
collective works or for resale. Permissions for such reuse can be obtained using the RightsLink “Request Permissions” link on
Wiley Online Library. Special requests should be addressed to: permissions@wiley.com
INFORMATION FOR SUBSCRIBERS Nutrition in Clinical Practice is published in 6 issues per year. Institutional subscription
prices for 2018 are:
Print & Online: US$607 (US and Rest of World), €563 (Europe), £470 (UK).
Prices are exclusive of tax. Asia-Pacific GST, Canadian GST/HST and European VAT will be applied at the appropriate rates. For
more information on current tax rates, please go to www.wileyonlinelibrary.com/tax-vat. The price includes online access to the
current and all online backfiles to January 1st 2014, where available. For other pricing options, including access information and
terms and conditions, please visit www.wileyonlinelibrary.com/access.
DELIVERY TERMS AND LEGAL TITLE Where the subscription price includes print issues and delivery is to the recipient’s
address, delivery terms are Delivered at Place (DAP); the recipient is responsible for paying any import duty or taxes. Title to
all issues transfers Free of Board (FOB) our shipping point, freight prepaid. We will endeavour to fulfil claims for missing or
damaged copies within six months of publication, within our reasonable discretion and subject to availability.
BACK ISSUES Single issues from current and recent volumes are available at the current single issue price from cs-
journals@wiley.com.
DISCLAIMER The Publisher, ASPEN and Editors cannot be held responsible for errors or any consequences arising from the
use of information contained in this journal; the views and opinions expressed do not necessarily reflect those of the Publisher,
ASPEN and Editors, neither does the publication of advertisements constitute any endorsement by the Publisher, ASPEN and
Editors of the products advertised.
PUBLISHER Nutrition in Clinical Practice is published by Wiley Periodicals, Inc., 350 Main St., Malden, MA 02148. E-mail:
cs-journals@wiley.com
JOURNAL CUSTOMER SERVICES: For ordering information, claims and any enquiry concerning your journal subscription
please go to www.wileycustomerhelp.com/ask or contact your nearest office.
Americas: Email: cs-journals@wiley.com; Tel: +1 781 388 8598 or +1 800 835 6770 (toll free in the USA & Canada).
Europe, Middle East and Africa: Email: cs-journals@wiley.com; Tel: +44 (0) 1865 778315. Asia Pacific: Email: cs-
journals@wiley.com; Tel: +65 6511 8000.
Japan: For Japanese speaking support, Email: cs-japan@wiley.com.
Visit our Online Customer Help available in 7 languages at www.wileycustomerhelp.com/ask
PRODUCTION EDITOR email: NCPprod@wiley.com.
Wiley’s Corporate Citizenship initiative seeks to address the environmental, social, economic, and ethical challenges faced in our
business and which are important to our diverse stakeholder groups. Since launching the initiative, we have focused on sharing
our content with those in need, enhancing community philanthropy, reducing our carbon impact, creating global guidelines and
best practices for paper use, establishing a vendor code of ethics, and engaging our colleagues and other stakeholders in our
efforts. Follow our progress at www.wiley.com/go/citizenship
Wiley is a founding member of the UN-backed HINARI, AGORA, and OARE initiatives. They are now collectively known as
Research4Life, making online scientific content available free or at nominal cost to researchers in developing countries. Please
visit Wiley’s Content Access – Corporate Citizenship site: http://www.wiley.com/WileyCDA/Section/id-390082.html
American Society for Parenteral and Enteral Nutrition (ASPEN), 8401 Colesville Road, Suite 510, Silver Spring, MD 20910.
Printed in the USA by The Sheridan Group.
Nutrition in Clinical Practice accepts articles for Open Access publication. Please visit http://olabout.wiley.com/WileyCDA/
Section/id-828081.html for further information about OnlineOpen.

This journal is a member of the Committee on Publication Ethics (COPE) (http://publicationethics.org/about)

 Nutrition in Clinical Practice
Volume 33 Issue 4 August 2018

Editorial
Editor’s Note 452
Jeanette M. Hasse, PhD, RD, LD, FADA, CNSC

Invited Reviews
Bacteria, Bones, and Stones: Managing Complications of Short Bowel Syndrome 454
Erika Johnson, MS, RD, CSR, LD, CNSC; Long Vu, MD;
and Laura E. Matarese, PhD, RDN, LDN, CNSC, FADA, FASPEN, FAND
Dietary Fermentable Oligosaccharides, Disaccharides, Monosaccharides, and Polyols (FODMAPs) 468
and Gastrointestinal Disease
Nimish Vakil, AGAF, FACG, FASGE
Nutrition Management of Necrotizing Enterocolitis 476
Vikram J. Christian, MBBS; Elizabeth Polzin, MBA, RD, CD, CNSC; and Scott Welak, MD

Reviews
Enteral Nutrition in the Management of Crohn’s Disease: Reviewing Mechanisms of Actions and Highlighting 483
Potential Venues for Enhancing the Efficacy
Moftah H. Alhagamhmad, MBBCH, DCH UK, PhD
Enteral Nutrition for Pediatric Crohn’s Disease: An Underutilized Therapy 493
Ala K. Shaikhkhalil, MD; and Wallace Crandall, MD, MMM

Clinical Observations
Case Report of Wernicke’s Encephalopathy After Sleeve Gastrectomy 510
Leslie A. Hamilton, PharmD, BCPS, BCCCP; Sarah H. Darby, BS; Allan J. Hamilton, MD;
Matthew H. Wilkerson, MD; and Kabel A. Morgan, MD
Copper Deficiency Myelopathy After Upper Gastrointestinal Surgery 515
Dominic King, MBChB; Keith Siau, MBChB; Latha Senthil, MBBS; Katherine F. Kane, MD;
and Sheldon C. Cooper, MD

Clinical Research
Reduction of Parenteral Nutrition and Hydration Support and Safety With Long-Term Teduglutide Treatment in Patients 520
With Short Bowel Syndrome−Associated Intestinal Failure: STEPS-3 Study
Douglas L. Seidner, MD, AGAF, FACG, CNSC; Ken Fujioka, MD; Joseph I. Boullata, PharmD, RPh, BCNSP, FASPEN;
Kishore Iyer, MBBS, FRCS, FACS; Hak-Myung Lee, PhD; and Thomas R. Ziegler, MD
Survey of Nutrition Management Practices in Centers for Pediatric Intestinal Rehabilitation 528
Anita M. Nucci, PhD, RD, LD; Kipp Ellsworth, MS, RD, CSP, CNSC; Austin Michalski, RD, LD, CNSC;
Emily Nagel, MS, RD, CNSC; and Jackie Wessel, MEd, RD, CNSC, CSP, CLE; on behalf of the ASPEN Pediatric
Intestinal Failure Section
Soluble Fiber Use in Pediatric Short Bowel Syndrome: A Survey on Prevailing Practices 539
Meredith Linley Harvie, MD; Margaret Alyssa Tucker Norris, MS, RD, LDN, CLC;
and Wednesday Marie A. Sevilla, MD, MPH, CNSC
Use of a Gastroschisis Feeding Guideline to Improve Standardization of Care and Patient Outcomes at an Urban 545
Children’s Hospital
R. Colby Passaro, MPH; Kate B. Savoie, MD, MS; and Eunice Y. Huang, MD, MS, FACS, FAAP
Impact of Gastrostomy Feeding Tube Placement on the 1-Year Trajectory of Care in Patients After Stroke 553
Janina Wilmskoetter, PhD; Annie N. Simpson, PhD; Sarah L. Logan, PhD; Kit N. Simpson, DrPH;
and Heather S. Bonilha, PhD, CCC-SLP
Strategies for the Enhancement of Nutrition Practice in a New York State Level 1 Trauma Center: A Hospital’s Journey 567
Lisa Musillo, MS, RDN, CNSC; Laryssa Marie Grguric, RDN, CNSC; Edward Coffield, PhD; Frank Aversano, MD;
Jeremy Bosworth, MD; and Richard Batista, MD
Gastrostomy Tube Feeding in Children With Developmental or Acquired Disorders: A Longitudinal Comparison 576
on Healthcare Provision and Eating Outcomes 4 Years After Gastrostomy
Ellen Backman, MSc; Ann-Kristin Karlsson, PhD; and Lotta Sjögreen, PhD

Letter to the Editor

Four-Oil Lipid Emulsion (Smoflipid) as a Tool in Managing Parenteral Nutrition Shortages 584
Lisa Mostafavifar, PharmD, BCPS, BCNSP; and David C. Evans, MD, FACS

Cover Art Note

The cover art depicts the human gastrointestinal tract; this issue of NCP tackles the nutrition management of some difficult
gastrointestinal disorders.
Cover art credit: 
C Getty Images

The ASPEN Rhoads Research Foundation depends on gifts from caring individuals to continue its work. Honor the memory of
a loved one or celebrate a special occasion by giving a gift to the ASPEN Rhoads Research Foundation.

We have now made it easier to support the foundation by providing a secure online donation form on our website:
www.nutritioncare.org. In addition to the online donation form, you can also charge your gift to a Visa, Mastercard, or American
Express by phoning our office, or you can mail a personal check directly to the ASPEN Rhoads Research Foundation.

Editor’s Note
Serious digestive disorders and other conditions that im-
pede nutrient absorption often require intervention with
nutrition support to restore and optimize nutrition status.
This issue of NCP features articles that tackle the nutri-
tion management of some difficult gastrointestinal (GI)
disorders.
Patients who have had extensive bowel resection often
develop short bowel syndrome (SBS), accompanied by
disruptions in the production of bile acids and digestive
enzymes, and dilation and dysmotility of the small bowel.
In this issue, Johnson, Vu, and Matarese discuss how these
conditions increase the risk of small intestinal bacterial
overgrowth, further complicating nutrient absorption. The
authors also delve into how metabolic changes that occur in
SBS can lead to bone demineralization and nephrolithiasis.
Two studies in this issue focus on SBS-related intestinal
failure. Seidner and colleagues undertook the third (STEPS-
3) in a series of studies to monitor the safety and efficacy of
teduglutide as a therapy for a cohort of parenteral nutrition
(PN)-dependent patients. They observed that treatment for
up to 42 months was associated with sustained efficacy and
a reduction in the PN support needed to maintain fluid
balance or nutrition intake.
In a Survey of Nutrition Management Practices in Centers
for Pediatric Intestinal Rehabilitation, Nucci et al collected
information to describe nutrition strategies used by special-
ists in pediatric intestinal rehabilitation at 14 medical centers
to promote gut adaptation and manage complications that
have not been previously summarized. Although many
enteral nutrition (EN) and PN management strategies were
relatively consistent, there was variability in laboratory test
selection and monitoring frequency among practitioners.
The authors encourage collaboration among centers to
establish biochemical monitoring standards for pediatric
patients with intestinal failure. Also surveying pediatric
practitioners, Harvie et al collected information to describe
the use of fiber to treat chronic high stool output in pediatric
patients with SBS treated in the US and abroad. For prema-
ture infants, most respondents (53.4%) did not initiate fiber
until the patient was >40 weeks corrected age. Supplemental
fiber was added in the form of pectin (52.4%) or from pectin-
rich foods (green beans, sweet potatoes and bananas) and
was discontinued if abdominal distention (67%), increased
emesis (43%), or bloody stools (38%) were observed or
if stool output increased (37%) or significantly decreased
(33%). In a review of pediatric literature, Christian, Polzin,
and Welak discuss how EN factors into recovery from
Nutrition in Clinical Practice
Volume 33 Number 4
August 2018 452–453

C 2018 American Society for
Parenteral and Enteral Nutrition
DOI: 10.1002/ncp.10183
wileyonlinelibrary.com
necrotizing enterocolitis (NEC) among premature infants.
This paper summarizes recommendations for PN support
when NEC occurs, recommended EN after uncomplicated
NEC, and provides considerations for EN after bowel
resection. The readers are encouraged to listen to the NCP
podcast featuring a discussion of this paper with 2 of its
authors.
Symptoms of abdominal pain, distention, and bloating
in adults have been treated by modifying the content of
short-chain carbohydrates in the patient’s diet. Dr. Vakil
reviews the emerging literature on FODMAP, an acronym
for fermentable oligosaccharides, disaccharides, monosac-
charides and polyols, as a treatment for GI disorders
and raises awareness of a lack of information available
to identify low-FODMAP-containing US enteral formu-
las. Continuing education credits are available for this
article.
Crohn’s disease (CD) causes inflammation in various
segments of the digestive tract, which can be painful and ac-
companied by diarrhea. Over time, these symptoms lead to
poor intake, poor nutrient absorption, weight loss, and mal-
nutrition. Two papers in this issue address EN in the man-
agement of CD. Alhagamhmad reviewed the pathogenesis
of this disease and summarizes studies of the mechanisms
of action for how EN may suppress intestinal inflammation
and enhance mucosal healing, including modulation of
gut microbiota. Similarly, Drs. Shaikhkhalil and Crandall
reviewed the pediatric literature on the efficacy, advantages,
and methods of administration of EN in the treatment
of CD. In addition, they discussed barriers to use of EN
and suggested methods practitioners could undertake to
increase usage.
Nutrient deficiencies related to GI surgery are explored in
2 articles. Historically, Wernicke’s Encephalopathy (WE), a
condition of thiamine deficiency, has been most commonly
associated with chronic alcoholism. Recently, thiamine de-
ficiency has also emerged as a concern following bariatric
surgery. Hamilton and colleagues describe a case of WE that
developed in a young woman, 3 months after she underwent
sleeve gastrectomy. Further concerning was the differential
diagnosis for symptoms of acute ischemic stroke, which
improved with the administration of IV thiamin. The au-
thors suggest there may be a role for prophylactic thiamin
in the bariatric surgery patient population to prevent WE.
King et al present a case of copper deficiency myelopa-
thy, which developed in a chronically malnourished, 50-
year-old woman who had undergone upper GI surgery.

Magnetic resonance image with T2-weighted enhancement
revealed a subacute combined degeneration of the spinal
cord, which is characteristically associated with vitamin B12
deficiency. The patient was initially treated with copper-
containing PN followed by 3 monthly hydroxocobalamin
intramuscular injections and enteral supplements in the
form of copper amino acid chelate. To prevent deleterious
neurologic impairment, the authors suggest that follow
up monitoring of blood levels of copper after upper GI
surgery to be a prudent means of recognizing copper
deficiency.
453
These and other nutrition support articles in this issue
summarize current practice and raise awareness of emerging
research into how you might restore nutrients to overcome
the incidence of malnutrition in patients with severe GI
disorders.
Jeanette M. Hasse, PhD, RD, LD, FADA, CNSC
Editor-in-Chief, Nutrition in Clinical Practice
Invited Review

Nutrition in Clinical Practice

Volume 33 Number 4
Bacteria, Bones, and Stones: Managing Complications August 2018 454–466

C 2018 American Society for

of Short Bowel Syndrome Parenteral and Enteral Nutrition

DOI: 10.1002/ncp.10113
wileyonlinelibrary.com

Erika Johnson, MS, RD, CSR, LD, CNSC1 ; Long Vu, MD1 ;
and Laura E. Matarese, PhD, RDN, LDN, CNSC, FADA, FASPEN, FAND2

Abstract
Short bowel syndrome (SBS) occurs in patients who have had extensive resection. The primary physiologic consequence is
malabsorption, resulting in fluid and electrolyte abnormalities and malnutrition. Nutrient digestion, absorption, and assimilation
may also be diminished by disturbances in the production of bile acids and digestive enzymes. Small bowel dilation, dysmotility,
loss of ileocecal valve, and anatomical changes combined with acid suppression and antimotility drugs increase the risk of small
intestinal bacterial overgrowth, further contributing to malabsorption. Metabolic changes that occur in SBS due to loss of colonic
regulation of gastric and small bowel function can also lead to depletion of calcium, magnesium, and vitamin D, resulting in
demineralization of bone and the eventual development of bone disease. Persistent inflammation, steroid use, parenteral nutrition,
chronic metabolic acidosis, and renal insufficiency may exacerbate the problem and contribute to the development of osteoporosis.
Multiple factors increase the risk of nephrolithiasis in SBS. In the setting of fat malabsorption, increased free fatty acids are available
to bind to calcium, resulting in an increased concentration of unbound oxalate, which is readily absorbed across the colonic mucosa
where it travels to the kidney. In addition, there is an increase in colonic permeability to oxalate stemming from the effects of
unabsorbed bile salts. The risk of nephrolithiasis is compounded by volume depletion, metabolic acidosis, and hypomagnesemia,
resulting in a decrease in renal perfusion, urine output, pH, and citrate excretion. This review examines the causes and treatments
of small intestinal bacterial overgrowth, bone demineralization, and nephrolithiasis in SBS. (Nutr Clin Pract. 2018;33:454–466)

Keywords
short bowel syndrome; small intestinal bacterial overgrowth; metabolic bone disease; nephrolithiasis

Introduction increase and there is a gradual transition from aerobic to

Short bowel syndrome (SBS) is the most common form
of intestinal failure. It is a debilitating, complex disorder
with multiple complications, some of which are potentially
life threatening.1,2 SBS results from the loss of portions of
the intestines due to surgical resection or loss of intestinal
function.2,3 Regardless of the reasons for the resection, the
result is an inability to maintain fluid and nutrient balances
through the consumption of a normal diet.2 The result-
ing malabsorption can cause numerous catheter, liver and
biliary, metabolic, renal, bone mineralization, or financial
and quality-of-life complications.1,3,4 This review focuses
on bacterial overgrowth including d-lactic acidosis (DLA),
metabolic bone disease (MBD), and nephrolithiasis.
anaerobic organisms that occur in the more distal sections
of the bowel. Once across to the ileocecal valve, bacte-
rial counts increase from 107 –109 organisms/mL in the
terminal ileum to approximately 1010 –1012 organisms/mL
in the colon.5 In the proximal section of the small bowel
(SB), only a few acid-resistant bacteria, mainly Lactobacilli,
Streptococci, and Escherichia coli exist.6 In the jejunum and
From the 1 Center for Human Nutrition, Digestive Disease and
Surgery Institute, Cleveland Clinic, Cleveland, Ohio, USA; and
2 Department of Internal Medicine and Surgery, Brody School of
Medicine, East Carolina University, Greenville, North Carolina,
USA.
Financial disclosure: None declared.
Conflicts of interests: None declared.

Bacteria This article originally appeared online on June 21, 2018.

Microbiota Changes in SBS
There are changes to the intestinal microbiotia of pa-
tients with SBS when compared with that of healthy nor-
mal adults. In the nonresected intestine, bacterial counts
Corresponding Author:
Laura E. Matarese, PhD, RDN, LDN, CNSC, FADA, FASPEN,
FAND, Department of Internal Medicine and Surgery, Brody School
of Medicine, East Carolina University, 600 Moye Blvd., Vidant MA
342, Mail Stop 734, Greenville, NC 27834, USA.
Email: mataresel@ecu.edu
Johnson et al
colon, numerous and predominantly anaerobic microbial
organisms reside, such as Bacteroides, Bifidobacteria, and
Clostridia, with Lactobacilli making up <1%.7
In patients with SBS, the overall amount of bacteria
are found to be quantitatively similar, but the difference
is noticed in the variety of microbial species. The changes
observed in a resected intestine is that the bacterial diversity
is noticeably reduced, and there is a shift between dominant
and subdominant organisms. For example, after a small
bowel resection, the microbiota of fecal matter and the
remaining distal intestine becomes depleted in the variety
of Bacteroides and Clostridium species; Lactobacilli become
more predominant in the entire remaining intestine.7,8 An-
other observed change is the presence of some strains such
as Lactobacillus mucosae, which have never been accounted
for in the intestinal microbiota of healthy human adults,
but have been found in SBS patients.8 In the setting of
SBS, the microbial contents of the colon and the metabolic
capabilities of the microbiota have an influence on a variety
of intestinal functions and become an essential component
in the course of management.
Small Intestinal Bacterial Overgrowth
Pathophysiology. In SBS, anatomical changes, fast tran-
sit, malabsorption, and hyperacidity occur after massive
resection, and this creates an optimal condition for the
alteration of bacteria throughout the intestine. Small in-
testinal bacterial overgrowth (SIBO) may occur when there
is a disproportional or excess of a particular species of
bacteria in the proximal SB, this does not necessarily equate
to harmful pathological consequences. In fact, there are
some benefits to having microbiota, even in excess. Take,
for instance, the role of bacteria in the energy salvage
of a patient with SBS. In the presence of a colon in
continuity, undigested carbohydrates pass into the colon
and are fermented by colonic bacteria into short chain
fatty acids and other organic acids. These fatty acids are
absorbed by the colon and provide additional energy as well
as enhance absorption of sodium and water. In pathologic
cases of SIBO, elevated bacterial counts have been found in
the proximal SB, commonly with bacterial species including
Streptococci, Bacteroides, Escherichia, and Lactobacilli.5 In
SBS patients, the dominant bacteria becomes Lactobacilli,
which is rather harmless, but an invasion of the SB with
an excessive or imbalanced population does carry a risk of
developing pathological SIBO. This can cause inflammation
and exacerbate malabsorption.
With the initial aid of gastric acid, Lactobacilli survive
past the stomach and are present in the proximal SB. As a
result of the malabsorption of nutrients, carbohydrates are
more abundant for Lactobacilli to ferment into lactic acid
and further contribute to the acidic environment optimal
for the sustained survival of Lactobacilli. Even the increase
455
in production of lactate by the bacteria overwhelm the
buffering capacity of the exocrine secretions of the pan-
creas, resulting in a low pH throughout the entire resected
bowel.6
The loss of the ileocecal valve was once thought to
promote SIBO because it acted as a barrier to the reflux of
material from the colon into the SB. However, it has been
suggested that the ileocecal valve does not independently
affect the risk of SIBO.9 Instead, the increased risk for
SIBO is thought to be related to reduced ileal length,
impaired peristalsis, and gastrointestinal acidity.5 The use
of antimotility medications may also play a role.10,11
Clinical significance. SIBO may contribute to the onset of
malnutrition, mucosal injury, and onset of symptoms. One
of the most significant nutrition implications is the risk
of developing malnutrition associated with malabsorption.
This can occur when deconjugation of bile acids in the
proximal SB disrupt fat digestion and lead to the production
of lithocholic acid.12 Under normal conditions, conjugated
bile acid enter the gastrointestinal tract in the duodenum
and are important in the emulsification, digestion, and ab-
sorption of dietary fat that occur in the SB. Deconjugation
and subsequent dehydroxylation of the bile acids result in
the formation of the secondary bile acids deoxycholic acid
and lithocholic acid.13 Bacteria such as Lactobacilli are able
to deconjugate bile acids, and as previously stated, there
is an increased amount of Lactobacilli in the intestines of
patients with SBS. Therefore, conjugated bile acids entering
the SB will be rapidly deconjugated.6 As a result of the
deconjugation of bile acids by Lactobacilli, there is only a
small amount of conjugated bile acid available, leading to
insufficient concentrations for fat absorption.
Another major pathophysiologic consequence of SIBO
relates to the inflammatory epithelial changes that sub-
sequently occur in the mucosa. Lithocholic acid, which
is produced from deconjugation of bile acids, is toxic to
enterocytes. Mucosal injury may also result from bacterial
adherence, production of enterotoxins, and other metabolic
products produced by bacteria. Enterocyte injury leads to a
loss of activity of brush-border activity, villous atrophy, and
altered permeability, the latter predisposing to the devel-
opment of a protein-losing enteropathy, impaired immune
defense, and bacterial translocation. Biopsy specimens of
the SB may provide the best evidence of the harmful impact
of the bacteria present. Biopsy provides a visual of the
inflammatory changes, villus blunting, and the presence of
adherents or intracellular bacteria, and it can also support
the diagnosis of pathologic SIBO.5
The consequences of malnutrition and inflammation
caused by SIBO are largely responsible for the variety
of symptoms that occur. Common symptoms associated
with SIBO include gas, bloating, and abdominal discomfort
caused by the fermentation of undigested carbohydrates by
456
Table 1. Symptoms Associated With Small Intestinal
Bacterial Overgrowth.
r ppm, indicates SIBO.5,12,14 However, the test results will
Increase flatulence
r Bloating
r Abdominal discomfort
r Foul smelling stools
r Explosive diarrhea
r Steatorrhea
r Frothy, foamy stools
r Decrease in oral intake
r and false negative rates.
Micronutrient deficiencies (vitamins A, B1, B12, D,
and E and iron)
bacteria (Table 1). Other symptoms that occur as a result of
malabsorption include steatorrhea and fat-soluble vitamin
deficiencies. Vitamin B12 is another nutrient deficiency that
can occur because of the use of the vitamin by anaerobic
bacteria. This deficiency can result in megaloblastic anemia
or neurological symptoms. Secretory diarrhea may occur
when bile acid absorption is disrupted, which happens when
lactobacilli deconjugate the bile acid.13
Diagnosis. SIBO is defined as a bacterial colony count
>100,000 colony-forming units per millimeter (105
CFU/mL) from the direct measurement of jejunal fluid or
the presence of bacteria from colon or oropharynx in the
SB.14 Nonspecific symptoms may suggest the presence of
SIBO such as increase in watery or frothy stool output,
especially if associated with bloating, flatulence, and
abdominal pain. A physical exam can show abdominal
distention with an audible succussion splash, which can
indicate fluid and gas in the distended bowel loops.5 There
is, as of yet, no widespread agreement as to the optimal
test for SIBO, especially in the setting of SBS. The gold
standard, although rarely used in practice, for diagnosing
SIBO is a quantitative culture of aspirated SB fluid.5
However, this test is not without significant limitations.
The test is relatively expensive, as it involves endoscopy or
fluoroscopy to place the aspiration catheter; there is a risk
for contamination of the specimen during its transit into
the SB; and it requires a detailed bacteriologic analysis,
which is especially difficult with anaerobic organisms. It has
been reported that only about 40% of the total gut flora
can be identified using conventional culture methods.11
As a result of the invasiveness of the direct method of
measurement, several indirect tests have been developed to
test for SIBO. These tests are widely used as a surrogate
to direct aspiration. Hydrogen breath tests have become
the most commonly used because of their low cost and
relative simplicity. The hydrogen test uses a carbohydrate
as a substrate, such as glucose, lactulose, or xylose. How
the test works is that the excessive bacteria in the SB
will metabolize the substrate releasing hydrogen that is
subsequently absorbed and then released in expired air.
Nutrition in Clinical Practice 33(4)
An increase in hydrogen in the breath sample after oral
ingestion of the carbohydrate substrate, typically 10–20
be flawed in the setting of malabsorption because of the
delivery of undigested substrate to the colon, making it im-
possible to differentiate between SB and colonic production
of hydrogen. Despite their popular use, hydrogen breath
tests are not recommended for use in testing for SIBO in the
SBS population because of the number of high false positive
Because the majority of patients with SBS present with
multiple predisposing factors to SIBO, clinicians often forgo
the diagnostic testing and presume SIBO based on the
presence and strength of symptoms and will often treat
empirically.
DLA
Pathophysiology. DLA is a rare syndrome that occurs in
individuals with SBS and an intact colon. Alteration of
the colonic microbiota plays a major role in the production
of D-lactate. Development of this syndrome occurs when
there is an excess production of D-lactate that exceeds the
clearance of D-lactate and leads to an accumulation of this
substance.
SBS can cause an excess production of D-lactate due
to the malabsorption of carbohydrates. This occurs when a
large quantity of undigested carbohydrates reach the colon,
the bacteria in the colon ferment the undigested carbohy-
drates to SCFA, lactic acid and other organic acids. Due
to the larger than normal amount of carbohydrates that
reach the colon, there is a greater than normal production
of organic acids which lower the luminal pH of the colon.
The low pH alters the colonic microbiota and supports the
survival of acidophilic bacteria that favor production of
lactic acid. As populations of both D-lactic-acid producing
and L-lactic-acid producing bacteria increase in the colon,
the pH continues to decline, progressively promoting the
growth and survival of lactic acid producing bacteria,
including bacteria such as Lactabillus fermenti, Lactobillus
acidophilus, and Streptococcus, all of which produce D-
lactate. Eventually, the gut flora is predominated by lactic
acid producing bacteria, and in some cases primarily those
that produce D-lactate. Once the intestinal flora has a
high concentration of D-lactate producing bacteria, large
amounts of carbohydrate, particularly readily fermentable
simple sugars, reaching the colon will result in a high
production of D-lactate.15 Carbohydrate malabsorption
can alter gut microbiota, predisposing patients with SBS
to SIBO, and SIBO can cause malabsorption of carbo-
hydrates further supporting the development of D-lactate
production.
In addition to excess production of D-lactate, in the
setting of SBS, there is potential for impaired clearance of
Johnson et al
Table 2. Symptoms of D-Lactic Acidosis.13,15
r Encephalopathy
r recommended to wait for the results to confirm elevated D-
Slurred speech
r Gait disturbance
r Weakness
r Tachypnea
it. D-Lactate is metabolized to pyruvate through the enzyme
D-2-hydroxy acid dehydrogenase (D-2-HDH). The highest
concentration of D-2-HDH is found in the kidney and liver.
These 2 organs are often compromised in SBS patients,
potentially reducing D-2-HDH concentration and D-lactate
metabolism. Another potential for impaired D-lactate
metabolism, involves oxalate, which is a potent inhibitor of
D-2-HDH and is often elevated in the blood of patients
with SBS.15-18 The inhibition of D-2-HDH by oxalate
and the potential impact of oxalate upon renal function,
may reduce metabolism and impair urinary excretion of
D-lactate.
Clinical significance. A key neurological symptom experi-
enced in patients with DLA is encephalopathy and this may
pose a diagnostic challenge for clinicians since the presenta-
tion can be mistaken for another neurological condition.18
Other neurological symptoms include altered aggressive
behavior, slurred speech, and ataxia, with patients often ap-
pearing drunk. The actual etiology of neurologic symptoms
seen in DLA is somewhat unclear. It appears that acidosis
and elevation of blood D-lactate levels alone are not capable
of producing the observed neurological manifestations. The
neurological symptoms do not always correlate with the
level of D-lactate, and it is unsettled whether it is D-lactate
or a simultaneously produced metabolite that causes the
syndrome of intoxication19 (Table 2). From our clinical
experience, SIBO and DLA appears to affect their quality
of life and ability to wean parenteral nutrition (PN).
Diagnosis. A high level of suspicion is key in diagnosing
DLA in patients with SBS, an intact colon in continuity, that
present with classic neurological disturbances. The level of
suspicion should be raised if the patient recently received
a course of antibiotic therapy or started a probiotic, both
of which can alter the intestinal flora. A diet history will
frequently reveal ingestion of relatively large amounts of
carbohydrates in the form of simple sugars prior to the onset
of symptoms.
If a patient presents with the described clinical fea-
tures and is acidotic, a diagnosis can be confirmed with a
normal L-lactate level and in increase in urine or serum
D-lactate level > 3mmol/L.16 Conventional lab assays for
lactate usually only measure L-lactate, which is why the
test is usually normal in DLA. D-lactate measuring kits are
457
available but assays are sent out since so infrequently used
and it can take weeks for results to be available. It is not
lactate levels before treating. A thorough work up should
rule out other indications that might be implicated such as
diabetic ketoacidosis, hepatic encephalopathy, iron toxicity,
ingestion of toxic substances, or uremic acidosis in advance
chronic kidney disease.
It might be difficult to measure elevated D-lactate levels
when a patient is having neurological symptoms because the
symptoms are transient and the patient may not be available
to collect a urine or blood sample. Often times a diagnosis
is suspected during an office visit, after a patient reports the
symptoms during a discussion with a clinician.
Treatment for SIBO and DLA
Nutrition therapy. The diet for SBS remains an important
component for managing SIBO and DLA. The usual dietary
modifications to reduce stool output should be emphasized
with the patient, and maintaining adequate hydration is
important. Patients are still recommended to limit simple
sugars such as cakes, cookies, and candies. Restriction of
readily fermentable simple carbohydrates should combine
with a modest amount of complex carbohydrates with 4–6
small frequent meals to avoid exposure of the gut flora to
large boluses of undigested carbohydrates. Fats should not
be restricted, unless they provoke steatorrhea and watery
diarrhea. In which case, a moderate-fat diet should be
implemented and combined with medical therapies. For
patients with significant bile acid loss, from a terminal ileum
resection, and a colon in continuity, bile acid resins could ac-
company a moderate-fat diet. Pancreatic enzymes could be
combined with a moderate fat diet in patients with or with-
out a colon present. In extreme cases, SIBO may complicate
adequate oral intake and pathologic bacterial may cause
mucosal damage, so nutrition support may be required.
For patients who experience DLA, the recommendation
to restrict simple sugars is a preventive therapy. During
an acute episode of DLA, treatment involves withholding
enteral sources of carbohydrate and remaining nil per os or
nothing by mouth until severe neurologic symptoms resolve.
To correct acidosis, intravenous bicarbonate and fluid can
be administered, but lactated ringer’s solution should be
avoided because they contain both D-lactate and L-lactate.
If the patient is at risk or has severe malnutrition, with
an anticipated prolonged oral food restriction, then PN
support is warranted; carbohydrate provided parenterally
does not increase D-lactate levels.20,21 Once patients recover
from neurologic symptoms, PN should be transitioned to
enteral feedings.
Micronutrient deficiency and their associated symptoms
are common in patients with SBS, and supplementation will
be necessary. For reasons described previously, in the setting
458
of SIBO, the replacement of fat-soluble vitamins A, D, and
E are likely as well as vitamin B12 and iron. Because of the
possibility that a problem with pyruvate metabolism may
be involved in the development of neurologic symptoms in
DLA, thiamine status should be assessed and supplemented
if indicated.16 An oral calcium supplement is recommended
for patients with SBS with a colon in continuity, not to
correct deficiencies but, rather, as an oxalate binder to
reduce the risk of increased free oxalate in circulation.
Medical therapy. SIBO is implicated in most cases of DLA,
therefore the prevention of this overgrowth is important
in the long-term management and prevention of DLA.
Medications associated with or known to inhibit intestinal
motility or the inhibition of gastric acid secretion should
be eliminated or substituted when treating SIBO. This is
difficult in the SBS population because a majority of the
treatment involves antimotility, antidiarrheal, and antise-
cretory medications to manage malabsorption and diarrhea
and control gastric acid hypersecretion after a SB resection.
However, after resection, the remaining bowel progresses
through several phases of adaptation, and the remaining
bowel begins to increase absorptive function. During this
time, absorption improves and it may be necessary to
reassess the need for antimotility and antisecretory medi-
cations and discontinue use if possible because this could
reduce the risk of developing SIBO for some individuals.
When medication correction is not an option, microbial
modification is attempted with antibiotic and probiotic
therapies. Currently, there are very few well controlled
trials for the treatment of SIBO, so recommendations are
based on clinical experience, and they are purely empirical
approaches. This is done with caution because antibiotic
therapy can increase or decrease D-lactate production.
A suggestion is to start with a single 7–14 day course of
antibiotic, preferably one that is nonabsorbable and covers
both aerobic and anaerobic organisms. When patients ex-
perience a reoccurrence of symptoms, a repeat course of
antibiotics is needed. Rotating among different antibiotic
regimens is recommended to prevent the development of
resistance.
Because there is concern regarding antimicrobial resis-
tance, antibiotic allergic reactions, and Clostridium difficile
diarrhea associated with the prolong use of antibiotics, there
is growing interest in the use of probiotics. Despite current
popularity and a growing number of experimental studies
that have shown potential benefits to probiotics, the role of
these agents in the management of SIBO remains unclear.
There is considerable discrepancies when it comes to the
use of probiotics to enhance outcomes of managing SIBO
and DLA. Among clinicians, there is a common practice
to avoid Lactobacillus strains of probiotics in SBS with a
colon in continuity because of the belief that there is a risk
Nutrition in Clinical Practice 33(4)
of developing DLA. However, these claims are made in the
absence of quality scientific data.22
One of the goals of treating SIBO is to reduce the risk
of an overproduction of D-lactic acid, which in turn would
reduce the risk of developing an episode of DLA. This is
done by attempting to reduce the number of pathogenic
bacteria present and to reduce the count of D-lactic acid
producers. Although DLA could occur without being ex-
acerbated by a course of antibiotics or the initiation of
a probiotic, the use of these therapies has been linked to
DLA episodes. Although it is reasonable to treat SIBO and
DLA with antibiotic and probiotic therapies, the results
with chronic preventive therapy have not been consistent
and are unpredictable.
The following antibiotic therapies are used to treat SIBO
and acute flares of DLA: tremethoprim-sulfamethoxazole,
neomycin, ciprofloxacin, vancomycin, and metronidazole.
However, each of these antibiotics has also been associated
with the onset of DLA,15,23 so use with caution is advised
(Table 3).
Oh et al24 reported the beneficial effects with the antibi-
otic neomycin in treating DLA. In this case, the patient was
able to remain free of recurrent episodes for at least 1 year.
In a different report, neomycin was not effective in reducing
the clinical symptoms of DLA nor reducing Lactobacilli.6
Rifaximin is increasingly used in the treatment of SIBO
and has been shown in controlled studies to be a good choice
for management.11 In at least 1 trial, rifaximin showed a
higher SIBO decontamination rate than metronidazole.25
However, even rifaximin appears to come with a degree
of doubt because it has been discussed that lactic acid
bacteria can grow at high levels, even in the presence of high
concentration of intraluminal rifaximin.17
In 1 of the original cases of DLA reported in a hu-
man, a probiotic, specifically Lactobacillus acidophilus, was
started prior to the onset of the episode.24 Since that first
report was published, other reports have been published
implicating probiotics in cases of DLA and also being used
successfully in the prevention of recurrent episodes.26,27 In
fact, some newer studies are showing promising results of
a combination of antibiotics and probiotic therapies.28,29
There is also a growing number of studies that are pro-
viding preliminary evidence that support probiotics as an
effective prevention and treatment option for SIBO.10,27
With all that potential comes the need for well-designed,
randomized controlled trials testing for strain-specific ef-
fects that may support specific claims. In addition to
finding an effective therapeutic agent, it is also important
to determine the duration and frequency of treatment
administration.22
Surgical options. Among patients with SBS, medical and
antibiotic therapies may fail, indicating the need for surgical
strategies. Options include bowel tapering, lengthening, or
resection (Table 4).
Johnson et al
Table 3. Antibiotic Use With Small Intestinal Bacterial Overgrowth and D-Lactic Acidosis in Patients With Small Bowel
Syndrome.
General Guidelines Antibiotic
r No standard protocol
r May increase or decrease d-lactate Amoxicillin-clavulanate
r Use with caution Ciprofloxacin
r Alternate between 2 different Rifaximin
single antibiotic regimens Metronidaxole
Trimethoprim-sulfamethoxazole
Table 4. Treatment Options for Small Intestinal Bacterial
Overgrowth and D-Lactic Acidosis in Patients With Short
Bowel Syndrome.
Diet modification
r 4–6 small meals
r Simple sugar restriction
r Low to moderate fat
r Lactose free, as needed
r Replacement of micronutrient deficiencies
(fat-soluble vitamins, B1, B12, iron)
r Calcium supplement with colon in continuity
Probiotics and antibiotics
r Empirical trials
r Role of probiotics remain unproven
r Use cautiously in patients with a history of D-lactic
acidosis
Other medications
r Re-evaluate need for proton pump inhibitor 6
months postoperatively small bowel resection
r Re-evaluate need for antidiarrheal, antisecretory, and
antimotility agents from time to time
Surgical
r When no medical or dietary modification treatments
are effective
r Bowel tapering, lengthening, or colon removal
Bones
MBD
A spectrum of disease. MBD affects bone strength through
a spectrum of disease from compromised bone density
as seen in osteopenia and osteoporosis to defective bone
mineralization as seen in osteomalacia. Risk factors for
MBD include both nonmodifiable and modifiable charac-
teristics (Table 5).30 The poor bone mineralization seen in
osteomalacia is usually caused by vitamin D deficiency or
another pathologic process that ultimately results in vitamin
D deficiency such as phosphate wasting disorders, renal
losses, metabolic acidosis, and malabsorption.31
Bone physiology. Rather than forming a completely rigid
and unchanging structure, the bones of the skeletal system
459
Adult Dose
Single-course regimens (7–14 days)
500 mg 3x/day
500 mg 2x/day
550 mg 3x/day
500 mg 3x/day
1 double-strength tablet 2x/day
Table 5. Risk Factors for Metabolic Bone Disease.30
Nonmodifiable Modifiable
r Female sex r Dietary deficiencies
r Older age in vitamin D,
(particularly calcium, phosphorus
menopause) r Tobacco use
r Race (White and r Alcohol use
Asian) r Elevated body mass
r Family history of index
metabolic bone r Decreased physical
disease activity
constitute a dynamic organ that is constantly being broken
down and remade anew. Osteoclasts of the mononuclear
monocyte–macrophage precursor lineage break down and
resorb old bone, preventing long-term build-up of stress
damage. Meanwhile, osteoprogenitor cells develop the os-
teoblasts that generate new protein matrices that form a
framework for bone mineralization. This bone remodeling
takes place in both the dense outer cortical bone as well as
the spongy inner trabecular bone. The overall bone mass
and density are the major determinants of bone strength.32
Epidemiology. Decreased bone strength leads to an in-
creased likelihood of bone fractures, which are a major
cause of morbidity and mortality particularly in the elderly
population. The overall prevalence of osteoporosis in the
U.S. population aged older than 50 years is 10.3% for
10.2 million adults, and an additional 43.4 million adults
are thought to have osteopenia.33 These fragile patients
generate >1.5 million fractures annually.34 Patients with
hip fractures in particular have a 3-fold increase in mor-
tality risk linked to fracture-related complications such as
thromboembolism, infections, and heart failure.35 Those
patients that survive continue to have increased morbidity
with limitations in mobility, basic activities of daily living,
self-care, and quality of life as up to one-third require
placement in assisted living or nursing home support.34,36
The treatment of osteoporosis and recovery from fracture
also comprises a significant financial burden in the U.S.
elderly, estimated at 16 billion dollars in 2002.34
460
Diagnosis. Although conventional plain-film radiographs
can help detect complications of low bone mineral density
(BMD) such as fractures, it has poor sensitivity to detecting
early changes of MBD. The dual-energy X-ray absorpti-
metry scan has long been considered the gold standard
for diagnosing osteoporosis. It is precise, accurate, has low
radiation exposure, and is relatively inexpensive.8 The test
generates 2 measures, a t-score and z-score, for several
different bone sites. The t-score is the number of standard
deviations between the patient’s BMD and the BMD of a
healthy 30-year-old. The z-score is a similar but compares
the patient’s BMD to the BMD of people similar in age,
race, and sex. The t-score is the most relevant number
for osteoporosis, which is defined as a t-score ࣘ −2.5.
Osteopenia is diagnosed with a t-score between −1.0 and
−2.5. The t-scores >−1 are considered normal BMD.37,38
MBD in SBS
Prevalence. Although there are some congenital causes of
SBS, the majority of adult cases are caused by extensive
surgical resection for various diseases, including mesen-
teric infarction, Crohn’s disease, radiation enteritis, and
intestinal volvulus. The decreased anatomical length of
bowel creates a reduced intestinal absorptive surface area
that can lead to increased losses of fluids and electrolytes,
compensatory restriction of enteral nutrition to reduce
losses, increased gastrointestinal tract transit time, and the
development of SIBO.39 The failure of the intestines to
absorb an adequate amount of vitamin D and calcium
predisposes the SBS patient to osteoporosis, and that risk
is higher than for the general population. In a 2017 study,
Nygaard et al40 examined a population of 167 patients
with chronic intestinal failure, of whom 80% had SBS.
The prevalence of osteoporosis in these patients was 56.9%
compared with 24.2% in the control group for an odds ratio
of 4.2; a further 31.7% of patients had osteopenia.40 Almost
90% of these patients with SBS had MBD.
Risk factors. The vast majority of patients with SBS have
some form of MBD from osteopenia to osteoporosis. Mal-
absorption of vitamin D and calcium certainly is an integral
component, but other aspects of the disease play crucial
roles in the development of MBD. The short bowel length
decreases fluid absorption time, leading to chronic diarrhea
and loss of bicarbonate-rich fluid, and can cause a nonanion
gap metabolic acidosis. This situation can be compounded
further by the development of SIBO causing anion gap
DLA. In addition, renal calcium wasting can further ex-
acerbate calcium depletion via an extraintestinal route.41,42
The binding of calcium to fatty acids via saponification and
generating Ca-soap can inhibit calcium absorption as well.43
Other disease states requiring intestinal surgery that
leads to SBS can also be risk factors for MBD. Patients
Nutrition in Clinical Practice 33(4)
often undergo bowel resections with abdominal malignancy,
which can impact their ability to intake enough calcium
from anorexia and malnutrition as well as direct mal-
absorption. Other malignancies can cause hypogonadism
and paraneoplastic syndromes that can directly affect cal-
cium metabolism and worsen MBD.41 For patients with
inflammatory bowel disease, mucosal inflammation of the
intestines can lead to malabsorption, and the elevation of
cytokines is believed to stimulate osteoclast activity, causing
more bone turnover.42 The avoidance of trigger foods such
as dairy products could lead to vitamin D and calcium
deficiency as well. Furthermore, corticosteroid use has been
strongly linked to the development of osteoporosis in the
general population and certainly puts patients prescribed
with frequent steroid tapers for inflammatory bowel disease
flares at risk for MBD.41,44
Deficiencies. SBS can lead to deficiencies in protein, cal-
cium, magnesium, and vitamin D, which are all important
for optimal bone health.42 There is a high prevalence of
vitamin D deficiency as noted by Ellegård et al45 in a
Swedish cohort of 106 patients with intestinal insufficiency
or intestinal failure in which 67% of the patients were
deficient (defined as <50 nmol/L) and 88% were found to
have low bone density. These results are consistent with a
similar Danish study of 167 patients that found vitamin
D deficiency in approximately 50% of the patients and
osteoporosis in 56.9%.40 This relationship of vitamin D
deficiency and low BMD also holds in different populations
around the world as seen in studies from Canada and
China, and often the vitamin D levels remain low despite
supplementation.46,47
Deficiencies in other related vitamins and minerals such
as vitamin K, vitamin C, copper, and fluoride can play a role
in SBS leading to MBD. Vitamin K is typically synthesized
by normal gut flora that can be disrupted in SBS. Vitamin K
is more commonly known for its role in blood coagulation,
but the vitamin K-dependent proteins osteocalcin, matrix
Gla protein, and protein S are all involved in the process of
bone formation, so a deficiency in vitamin K can also impact
bone metabolism and decrease bone mineralization.42 In a
retrospective study of 189 Canadian patients on home PN
(HPN), a trend was noted for higher hip and lumbar spine
t-scores for patients who had vitamin K supplementation
when compared with those who had no supplementation.48
MBD With HPN
Prevalence. For patients with SBS, malabsorption and di-
arrhea often limit giving nutrition via enteral route. The use
of HPN is common and allows for the provision of calories,
vitamins, and other nutrients via intravenous route as a
bridge to intestinal surgery or transplant or even continued
indefinitely. A cross-sectional and retrospective multicenter
Johnson et al
study of 165 European patients on long-term HPN for a
duration >6 months found a high prevalence of reduced
BMD, with a t-score at any site <−1 in 84% and t-scores
ࣘ−2.5 in 41%.49 MBD is common among patients using
HPN and often may go underdiagnosed without adequate
screening in this patient population.
Risk factors. In addition to the numerous risk factors for
MBD from SBS, there are risk factors for MBD from PN
as well that affect calcium excretion. Adequate amounts of
calcium and phosphorus are required, although the levels
that can be provided in solution are limited by calcium-
phosphate precipitation.41 The proteins supplied by amino
acids constitute a titratable acid, and their low pH will
also increase urinary calcium excretion. Higher levels of
sodium will increase the renal glomerular filtration rate and
thus increase urinary calcium excretion, and higher levels
of dextrose will increase serum insulin concentrations and
increase calcium excretion as well.42 In the home setting,
patients often perform cycling of their HPN infusion,
reducing duration from 24 hours to >12 hours, which has
been shown to increase urinary calcium excretion. Although
the exact mechanism is unknown, it is likely related to
the increased rate of HPN infusion when the duration is
shortened.41
Other factors in PN that affect BMD include those that
adversely alter bone metabolism. Magnesium has an effect
on parathyroid hormone (PTH) secretion and renal activity
that plays an important role in calcium homeostasis.50
Heparin-associated osteoporosis can also occur, particu-
larly for those who require treatment for deep venous
thrombosis, which is a common complication with central
intravenous lines used for PN access.42 Aluminum is a
common element found in many components of PN, and
its toxicity can impair PTH secretion and decrease levels of
activated vitamin D.41
PN preparation guidelines. Given the risk factors for MBD
from PN, it is imperative that PN formula preparations are
adjusted to avoid worsening of MBD. Calcium gluconate
should be provided at 15 mEq/day and with phosphate in a
ratio of 1:2, for 15 mEq calcium to 30 mmol phosphorus
in the PN solution. Magnesium should be dosed around
15–20 mEq/day, depending on the volume of diarrhea.41
Amino acids should not exceed 1.5 g/kg/day for short
term and upon resolution of critical illness, should be
reduced to 0.8–1.0 g/kg/day to avoid metabolic acidosis and
hypercalciuria.42 Acetate levels in the PN solution are kept
to a goal serum bicarbonate in the mid-range of normal
to also avoid acidosis.41 Aluminum toxicity was a larger
problem in the 1980s with the use of casein hydrolysate as a
protein source, although some PN additives still have a small
amount of aluminum contamination.42 For all patients, the
461
guidelines recommend keeping aluminum to a minimum,
ideally <25 mcg/L.51
The European Society for Clinical Nutrition and
Metabolism guidelines also suggest adding vitamin D to
PN for patients with intestinal failure as so many of them
have MBD or are at high risk for MBD; the recommended
intravenous dose is 200 international units, which is typically
provided in the form of a multivitamin supplement.52
Additional supplementation of vitamin D is indicated in
patients with low 25-hydroxyvitamin D levels and high
PTH.45 Unfortunately, there are no intravenous formula-
tions of vitamin D available in the United States that can
be added to PN aside from that found in the multivitamin,
so additional supplementation is given by enteral route. It
should be noted that vitamin D is a negative acute phase
reactant and care should be taken when checking levels in
patients with acute inflammation.53 Interestingly, in patients
with low BMD, low PTH, and low 1,25-hydroxyvitamin D
levels, but a normal 25-hydroxvitamin D level, the vitamin
D supplement can be temporarily withdrawn with levels
rechecked at yearly intervals. The removal of vitamin D
from PN did not have adverse effects and is actually shown
to significantly increase bone mineral content.54
Long-term HPN. Despite the presence of risk factors for
MBD from PN, many studies have shown that long-term
PN does not exacerbate osteoporosis. A Danish study
followed 75 patients receiving HPN from 1995 to 2003.
The mean duration of HPN was 4 years, and patients
received an average of 4.4 dual-energy X-ray absorptimetry
scans during the study period. There was a moderate but
statistically significant 1% decrease in BMD, but this loss
was not significantly greater than that seen in age-matched
and sex-matched healthy controls.55 In some studies, there
was actually an increase in BMD, such as the multicenter
study by Pironi et al,56 which followed 65 patients on HPN
during an average of 18 months and found a statistically
significant increase in z-score at the lumbar spine. A French
prospective study followed 56 patients with intestinal failure
on HPN for 5 years of whom 67% had osteoporosis at
baseline and found an improvement in trabecular BMD,
particularly in patients whose onset of intestinal failure was
in adulthood.57
Management of MBD
Patient evaluation. Evaluation of MBD should incorporate
a thorough history and physical examination as well as
relevant blood work. Many patients with osteoporosis do
not have symptoms, but a history of bone fractures from
minor trauma or vertebral compression fractures is suspect.
Patients can complain of skeletal bone pain or proximal
muscle weakness. A family history of osteoporosis may also
be present. Social history including tobacco and alcohol
462
use must also be elicited. Secondary causes of osteoporo-
sis include Cushing’s syndrome, hyperparathyroidism, and
hypogonadism in men, so signs of these conditions should
be investigated on physical exam. A medication history is
essential as corticosteroids, estrogens, and loop diuretics can
have an impact in MBD.41
Laboratory work should consist of studies to support the
history and physical including basic metabolic panel, mag-
nesium, and phosphorus levels. PTH, thyroid-stimulating
hormone, and 25-hydroxyvitamin D levels can be checked,
particularly if patients have malabsorption. Commonly
the recommendation is for a 24-hour urine calcium to be
monitored every 6–12 months to check for excess urinary
calcium excretion.41 However, many impracticalities exist
in the home setting with adequate urine collection and re-
moving calcium supplements prior to collection. In addition
to biochemistry, European Society for Clinical Nutrition
and Metabolism guidelines recommend routine monitoring
of BMD with dual-energy X-ray absorptimetry starting at
baseline and then yearly intervals, although this scanning
frequency is not well supported by studies at this time.52
Patients should ideally be followed longitudinally using the
same scanner every time to increase precision and minimize
variation in measurement and interpretation.37
Nonpharmacologic treatment. It is important for patients
with MBD or at risk of MBD to continue to treat their
underlying disease, infections, and metabolic acidosis and
reduce chronic inflammation where possible. The treatment
of the secondary causes of osteoporosis involves diagnos-
ing and managing hyperthyroidism, hypogonadism, and
hyperparathyroidism.41 Healthy lifestyle changes should
also be counseled for patient to discontinue tobacco use,
minimize alcohol intake, and encourage regular physical
activity, particularly weight-bearing exercise. Not only do
these measures increase BMD but also decrease fall rates.31
Medications. There are several classes of medications that
are approved by the Food and Drug Administration for the
treatment of osteoporosis. These drugs include conjugated
estrogens, selective estrogen receptor modulators, bispho-
sphonates, calcitonin, and receptor activator of nuclear
factor-κB ligand (RANKL) inhibitors. The pharmacologic
mechanism of action is to inhibit osteoclasts and re-
duce bone resorption.41,58 Bisphosphonates are considered
the first-line therapy for osteoporosis in postmenopausal
women based on efficacy, safety, and cost. Unfortunately,
they are poorly absorbed from the gastrointestinal tract
and commonly cause gastrointestinal intolerance, which
makes them difficult to use in patients with SBS.59 However,
intravenous preparations exist and an early randomized
controlled trial of 20 patients taking clodronate found it
inhibited bone resorption and increased BMD. However,
the BMD increase was only statistically significant in the
Nutrition in Clinical Practice 33(4)
forearm, but there was a trend for increasing BMD in
other bone sites.60 A Canadian case series of 11 patients
also noted an improvement in hip t-score with the use of
intravenous palmidronate.61
For patients who are intolerant of bisphosphonates, an
alternative therapy is teriparatide, which is recombinant
PTH. This medication is given by subcutaneous route and
has a different mechanism of action; rather than inhibit
osteoclasts, it stimulates osteoblasts to increase bone
anabolism. There is a case report of a patient intolerant
of oral alendronate on teriparatide for 18 months whose
spinal t-score increased from −2.5 to 0.62 The patient’s
BMD normalized from the osteoporotic range.
Newer pharmacologic therapy includes the RANKL
inhibitor denosumab, so named because it targets the
RANKL, a cytokine that is responsible for osteoclast func-
tion and differentiation. By binding RANKL, denosumab
interferes with its ability to bind the receptor RANK,
shutting down osteoclasts and inhibiting bone resorption.63
Denosumab has been shown to reduce hip fractures in post-
menopausal women and has also been studied in patients on
HPN with intestinal failure. A total of 15 patients assessed
on denosumab after 1 year were found to have significant
improvement in t-scores and z-scores of the lumbar spine
and femur compared with no difference found for the 17
patients in the control group.58 The treatment was generally
well tolerated.
Nephrolithiasis (Stones)
Individuals with SBS are susceptible to numerous potential
complications. One of these is the development of renal
stones. This can result in frequent episodes of renal colic,
obstructive uropathy, and urinary tract infections. In some
cases, the deposition of stones can lead to the progressive
impairment of renal function.
Prevalence
The prevalence of symptomatic nephrolithiasis, renal
stones, ranges from 42% to 50% in patients with SBS
compared with 8.4% in the general population.64-66 The inci-
dence of stones for patients with a jejunostomy or ileostomy
ranges from 5% to 15%.67,68 If the colon is retained such
as those with a jejuno-colic anastomosis, they have a 25%
chance of developing symptomatic nephrolithiasis.69
Types, Causes, and Prevention
The composition of the stones will vary depending on the
type of surgical resection and several other predisposing
factors. Both uric acid and calcium oxalate renal stones are
common in patients with a jejunostomy or ileostomy68,70-73
and are thought to be related to chronic dehydration,
low urine volume, and reduced urine sodium.74-76 This
Johnson et al 463

Table 6. High Oxalate Foods.a,83-88

Food Type Examples

Fruits Blackberries, black raspberries, blueberries, red currants, dewberries, figs, grapes, gooseberries,
kiwi, lemon peel, lime peel, orange peel, red raspberries, rhubarb, strawberries, tangerines,
any juice made from above fruits
Vegetables Beans (green, wax, dried), beets (tops, roots, greens), celery, chives, collards, dandelion,
eggplant, escarole, kale, leeks, mustard greens, okra, parsley, parsnips, peppers (green),
pokeweed, rutabagas, sorrel, spinach, summer squash, sweet potatoes, Swiss chard, tomato
soup, vegetable soup, watercress, yams
Meat substitutes, beans, nuts, Almonds, baked beans canned in tomato sauce, cashews, green beans (waxed and dried),
seeds peanut butter, peanuts, pecans, sesame seeds, sunflower seeds, tofu (soybean curd), walnuts
Beverages Any juice made from high oxalate fruits, draft beer, chocolate (plain), chocolate milk, cocoa,
coffee powder (instant), Ovaltine, tea (brewed)
Starches Fig Newtons, fruit cake, graham crackers, grits, white corn, kamut, marmalade, soybean
crackers, wheat germ
Condiments Ground cinnamon, raw parsley, pepper, >1 tsp/day, ginger, soy sauce
a Partial listing.

Table 7. Methods to Decrease Risk of Stone Formation in Short Bowel Syndrome.

Methods Rationale

Adequate fluid and salt intake
Low oxalate diet
Low-fat diet
Increase foods with vitamin C
or citrate supplements
Calcium supplementation
Magnesium supplements
Probiotics
Cholestyramine
In short bowel syndrome, urine volume and urine sodium are reduced due to malabsorption.
Increase fluid and salt intake to increase urine output to a goal of 2 liters per day prevent
stone formation
Reduce oxalate intake for those with a colon to reduce oxalate excretion
Reduce steatorrhea that leads to binding of intraluminal calcium by fatty acids, which leads to
nonavailability of calcium to bind oxalate, which leads to increased absorption of dietary
oxalate
Vitamin C converted to citrate which prevents oxalate and calcium crystal formation
Bind oxalate in the gut
Hypomagnesemia leads to reduced citrate excretion and urine pH
Oxalate degrading
Bind oxalate in the colon

highlights the importance of adequate fluid and salt intake
to increase urine output and prevent stone formation. Urine
volume should be increased as much as possible so that
the urine output exceeds 2 liters per day.68 This can be
challenging in patients with SBS and may require the use
of supplemental intravenous fluids or even PN. In addition,
patients with the largest resections or the most extensive
disease have the greatest degree of hyperoxaluria and are
more likely to form renal stones.77,78 Individuals with an
ileal pouch anal anastomosis have a slightly smaller risk
for renal stones, presumably due to better fluid and salt
balance.79 Calcium oxalate stone formation results from
hypocitraturia and hypomagnesuria; uric acid stones are
generally related to low urine volume and low pH.66 SBS
patients with hyperglycemia due to diabetes or poorly
managed PN are at risk for developing stones.80 Diets high
in citrate-containing fruits and vegetables such as citrus
fruits, tomatoes, and Jerusalem artichokes reduce oxalate
stone formation by competing for calcium binding in the
urine.81 Hypomagnesemia leads to reduced citrate excretion
and urine pH. Thus, it is important to maintain adequate
magnesium levels.
Patients with SBS with the colon or portions of the colon
in circuit are more likely to experience stone formation;
these are generally calcium oxalate in nature. Normally,
calcium binds to oxalate in the intestine allowing it to
be excreted in the stool. In SBS, unabsorbed fat bind to
calcium. This leaves oxalate free to be absorbed by the
colon. More oxalate is absorbed, resulting in a higher risk
of developing kidney stones.82 These may occur as early as
2 months after surgical resection or as late as 67 months
following an operation.64 Consumption of an oxalate-free
diet will reduce urinary oxalate excretion.77 Oxalate is preva-
lent in many foods, making the diet difficult to comply with
for some patients (Table 6).83-88 Alternatively, increasing
dietary calcium intake in ranges of 600–2000 mg/day with
464
meals has been shown to decrease urinary oxalate excretion
by competing with luminal oxalate and limiting intestinal
absorption.89-91
Patients with SBS often have excess fecal fat due to
steatorrhea. Unabsorbed fatty acids compete with oxalate
for calcium resulting in greater oxalate absorption.77 A
low-fat diet should be used to reduce the fecal output that
leads to binding of intraluminal calcium, which in turn
results in nonavailability of calcium to bind oxalate. The
situation is often further complicated by the presence of
unabsorbed bile salts as well as long-chain fatty acids.74
Cholestyramine has been shown to reduce oxaluria by
chelating bile salts and preventing stone formation.79 How-
ever, cholestyramine also binds and impedes the absorption
of other nutrients including fat-soluble vitamins and may
even worsen diarrhea in some patients.
Certain vitamins have been implicated in the devel-
opment and treatment of oxalate stones. Pyridoxine and
thiamine deficiency has been associated with enhanced
metabolic generation of oxalate or enhanced ileal absorp-
tion. As a result, pyridoxine in doses of 150 mg per day
has been used to reduce endogenous oxalate production.92
Thiamine deficiency has been shown to cause hyperoxaluria
in animal models.93
Oxalate has been shown to be degraded by bacteria
in the colon.94,95 As a result, there has been interest in
using probiotics to change the composition of the gut
microbiota. Certain probiotics have been used to reduce
oxalate absorption.96 The best probiotic and specific dosage
to use still needs to be determined. The oxalate-degrading
activity of different probiotics varies, thus more research
is indicated to determine the exact strain and dose.97 Pre-
vention and treatment of both uric acid and oxalate stones
is based on the gastrointestinal anatomy and may require
a multipronged approach to include dietary interventions,
medications, and fluid management (Table 7).
Summary
SBS can result in significant acute and chronic complica-
tions resulting from changes in gastrointestinal anatomy
and the results of malabsorption. These complications
can have a negative impact on the patient’s quality of
life. SIBO, bone demineralization, and nephrolithiasis are
common and debilitating in these patients. Comprehensive
monitoring and treatment can minimize these complications
and the progression of further disease.
Statement of Authorship
E. Johnson, L. Vu, and L. E. Matarese contributed to con-
ception/design of the article; E. Johnson, L. E. Vu, and L. E.
Matarese contributed to acquisition, analysis, or interpretation
of the data; E. Johnson, L. Vu, and L. E. Matarese drafted the
Nutrition in Clinical Practice 33(4)
manuscript; E. Johnson, L. Vu, and L. E. Matarese critically
revised the manuscript; and E. Johnson, L. Vu, and L. E.
Matarese agree to be fully accountable for ensuring the integrity
and accuracy of the work. All authors read and approved the
final manuscript.
References
1. Hofstetter S, Stern L, Willet J. Key issues in addressing the clinical and
humanistic burden of short bowel syndrome in the US. Curr Med Res
Opin. 2013;29(5):495-504.
2. O’Keefe SJ, Buchman AL, Fishbein TM, Jeejeebhoy KN, Jeppesen PB,
Shaffer J. Clin Gastroenterol Hepatol. 2006;4(1):6-10.
3. Nightingale JMD. The short bowel. In: Nightingale JMD, ed. Intestinal
Failure. London, England: Greenwich Medical Media; 2001:177-198.
4. Buchman AL, Scolapio J, Fryer J. AGA technical review on short
bowel syndrome and intestinal transplantation. Gastroenterology.
2003;124(4):1111-1134.
5. Dibaise JK, Young RJ, Vanderhoof JA. Enteric microbial flora,
bacterial overgrowth, and short-bowel syndrome. Clin Gastroenterol
Hepatol. 2006;4(1):11-20.
6. Bongaerts GPA, Tolboom JJ, Naber AHJ, et al. Role of bacteria in the
pathogenesis of short bowel syndrome-associated d-lactic academia.
Microb Pathog. 1997;22(5):285-293.
7. Kaneko T, Bando Y, Kurihara H, Satomi K, Nonoyama K, Matsuura
N. Fecal microflora in a patient with short-bowel syndrome an identi-
fication of dominant lactobacilli. J Clin Microbiol. 1997;35(12):3181-
3185.
8. Joly F, Mayeur C, Bruneau A, et al. Drastic changes in fecal and
mucosa-associated microbiota in adult patients with short bowel syn-
drome. Biochimie. 2010;92(7):753-761.
9. Matarese L, O’Keefe SJ, Kandil H, Bond G, Costa G, Abu-Elmagd K.
Short bowel syndrome: clinical guidelines for nutrition management.
Nutr Clin Pract. 2005;20(5):493-5020.
10. Cresci G, Bawden E. Gut microbiome: what we do and don’t know.
Nutr Clin Pract. 2015;30(6):734-746.
11. Quigley E, Abu-Shanab A. Small intestinal bacterial overgrowth. Infect
Dis Clin N Am. 2010;24(4):943-959.
12. Quigley EMM, Quera R. Small intestinal bacterial overgrowth: roles
of antibiotics, prebiotics, and probiotics. Gastroenterology. 2006;130(2
Suppl 1):S78-S90.
13. Bongaerts GPA, Severijnen R, Tangerman A, Verrips A, Tolboom JJM.
Bile acid deconjugation by lactobacilli and its effects on patients with a
short small bowel. J Gastroenterol. 2000;35(11):801-804.
14. Schiller L. Evaluation of small bowel bacterial overgrowth. Curr
Gastroenterol Rep. 2007;9(5):373-377.
15. Peterson C. D-lactic acidosis. Nutr Clin Pract. 2005;20(6):634-645.
16. Uribarri J, Oh M, Carroll H. D-lactic acidosis: a review of clinical
presentation, biochemical features, and pathophysiologic mechanisms.
Medicine. 1998;77(2):73-82.
17. White L. D-lactic acidosis: more prevalent than we think. Practical
gastroenterology. https://med.virginia.edu/ginutrition/articles-from-
practical-gastroenterology/. Accessed March 8, 2018.
18. Kowlgi NG, Chhabra L. D-lactic acidosis: an underrecognized
complication of short bowel syndrome. Gastroenterol Res Pract.
2015;2015:476215.
19. Goulet O, Joly F. The pathological microbiota: intestinal microbiota in
short bowel syndrome. Gastroenterol Clinical Bio. 2010;34(1):S37-S43.
20. Ramakrishnan T, Stokes P. Beneficial effects of fasting and low
carbohydrate diet in D-lactic acidosis associated with short bowel
syndrome. JPEN J Parenter Enteral Nutr. 1985;9(3):361-363.
21. Bongaerts GPA, Severijnen R. Arguments for a lower carbohydrate–
higher fat diet in patients with a short small bowel. Med Hypothesis.
2006;67(2):280-280.
Johnson et al
22. Vitetta L, Coulson S, Thomsen M, Nguyen T, Hall S. Probiotics, D-
lactic acidosis, oxidative stress and strain specificity. Gut Microbes.
2017;8(4):311-322.
23. Attar A, Flourie B, Rambaud JC, et al. Antibiotic efficacy in small
intestinal bacterial overgrowth–related chronic diarrhea: a crossover,
randomized trial. Am Gastro Assoc. 1999;117(4):794-797.
24. Oh MS, Phelps KR, Traube M, Barbosa-Saldivar J, Boxhill C, Carroll
HJ. D-lactic acidosis in a man with the short bowel syndrome. N Engl
J Med. 1979;301(5):249-52.
25. Lauritano EC, Gabrielli M, Scarpellini E, et al. Antibiotic therapy in
small intestinal bacterial overgrowth: rifaximin versus metronidazole.
Eur Rev Med Pharmacol. 2009;13(2):111-116.
26. Munakata S, Arakawa C, Kohira R, Fujita Y, Fuchigami T,
Mugishima H. A case of D-lactic acid encephalopathy associated with
use of probiotics. Brain Develop. 2010;32(8):691-694.
27. Vanderoof Y, Murray K. Treatment strategies for small bowel bac-
terial overgrowth in short bowel syndrome. J Med Gastro Nutrition.
1998;27(2):155-160.
28. Gigante A, Sardo L, Gasperini ML. D-lactic acidosis 25 years
after bariatric surgery duet to salmonella enteritidis. Nutrition.
2012;28(1):108-111.
29. Takahashi K, Terashima H, Kohno K, Ohkohchi N. A stand-alone
symbiotic treatment for the prevention of D-lactic acidosis in short
bowel syndrome. Int Surg. 2013;98(2):110-113.
30. NIH Consensus Development Panel on Osteoporosis Prevention, Di-
agnosis, and Therapy. Osteoporosis prevention, diagnosis, and therapy.
J Am Med Assoc. 2001;285(6):785-795. https://doi.org/10.1001/jama.
285.6.785
31. Goltzman D. Functions of vitamin D in bone. Histochem Cell Biol.
2018;149(4):305-312. https://doi.org/10.1007/s00418-018-1648-y
32. Clarke B. Normal bone anatomy and physiology. Clin J Am
Soc Nephrol. 2008;3(suppl 3):S131-S139. https://doi.org/10.2215/CJN.
04151206
33. Wright NC, Looker AC, Saag KG, et al. The recent prevalence of
osteoporosis and low bone mass in the United States based on bone
mineral density at the femoral neck or lumbar spine. J Bone Miner Res.
2014;29(11):2520-2526. https://doi.org/10.1002/jbmr.2269
34. Blume SW, Curtis JR. Medical costs of osteoporosis in the el-
derly Medicare population. Osteoporos Int. 2011;22(6):1835-1844.
https://doi.org/10.1007/s00198-010-1419-7
35. Panula J, Pihlajamäki H, Mattila VM, et al. Mortality and
cause of death in hip fracture patients aged 65 or older—a
population-based study. BMC Musculoskelet Disord. 2011;12(1):105.
https://doi.org/10.1186/1471-2474-12-105
36. Dyer SM, Crotty M, Fairhall N, et al. A critical review of the
long-term disability outcomes following hip fracture. BMC Geriatr.
2016;16(1):158. https://doi.org/10.1186/s12877-016-0332-0
37. Lewiecki EM, Lane NE. Common mistakes in the clinical use of bone
mineral density testing. Nat Clin Pract Rheumatol. 2008;4(12):667-674.
https://doi.org/10.1038/ncprheum0928
38. Nayak S, Edwards DL, Saleh AA, Greenspan SL. Systematic review
and meta-analysis of the performance of clinical risk assessment instru-
ments for screening for osteoporosis or low bone density. Osteoporos
Int. 2015;26(5):1543-1554. https://doi.org/10.1007/s00198-015-3025-1
39. Pironi L, Arends J, Baxter J, et al. ESPEN endorsed recommendations.
Definition and classification of intestinal failure in adults. Clin Nutr.
2015;34(2):171-180. https://doi.org/10.1016/j.clnu.2014.08.017
40. Nygaard L, Skallerup A, Olesen SS, et al. Osteoporosis in patients with
intestinal insufficiency and intestinal failure: prevalence and clinical
risk factors [published online ahead of print August 2017]. Clin Nutr.
https://doi.org/10.1016/j.clnu.2017.07.018
41. Seidner DL. Parenteral nutrition-associated metabolic bone dis-
ease. JPEN J Parenter Enteral Nutr. 2002;26(5 suppl):S37-S42.
https://doi.org/10.1177/014860710202600511
465
42. Ferrone M, Geraci M. A review of the relationship between parenteral
nutrition and metabolic bone disease. Nutr Clin Pract. 2007;22(3):329-
339. https://doi.org/10.1177/0115426507022003329
43. Gacs G, Barltrop D. Significance of Ca-soap formation for calcium
absorption in the rat. Gut. 1977;18(1):64-68. https://doi.org/10.1136/
gut.18.1.64
44. Abitbol V, Roux C, Chaussade S, et al. Metabolic bone assess-
ment in patients with inflammatory bowel disease. Gastroenterology.
1995;108(2):417-422. https://doi.org/10.1016/0016-5085(95)90068-3
45. Ellegård L, Kurlberg G, Bosaeus I. High prevalence of vitamin D
deficiency and osteoporosis in out-patients with intestinal failure. Clin
Nutr. 2013;32(6):983-987. https://doi.org/10.1016/j.clnu.2013.02.005
46. Fan S, Ni X, Wang J, et al. High prevalence of suboptimal vitamin
D status and bone loss in adult short bowel syndrome even after
weaning off parenteral nutrition. Nutr Clin Pract. 2017;32(2):258-265.
https://doi.org/10.1177/0884533616665784
47. Napartivaumnuay N, Gramlich L. The prevalence of vitamin D
insufficiency and deficiency and their relationship with bone min-
eral density and fracture risk in adults receiving long-term home
parenteral nutrition. Nutrients. 2017;9(5):481. https://doi.org/10.3390/
nu9050481
48. Aljarallah B, Fernandes G, Jeejeebhoy KN, et al. The Canadian
Home Total Parenteral Nutrition (HTPN) registry: vitamin K
supplementation and bone mineral density. JPEN J Parenter
Enteral Nutr. 2012;36(4):415-420. https://doi.org/10.1177/
0148607111431983
49. Pironi L, Morselli Labate AM, et al. Prevalence of bone disease in
patients on home parenteral nutrition. Clin Nutr. 2002;21(4):289-296.
https://doi.org/10.1054/clnu.2002.0548
50. Rodriguez-Ortiz ME, Canalejo A, Herencia C, et al. Magnesium
modulates parathyroid hormone secretion and upregulates parathyroid
receptor expression at moderately low calcium concentration. Nephrol
Dial Transplant. 2014;29(2):282-289. https://doi.org/10.1093/ndt/gft400
51. Klein GL. Aluminum in parenteral solutions revisited—again. Am J
Clin Nutr. 1995;61(3):449-456. https://doi.org/10.1093/ajcn/61.3.449
52. Pironi L, Arends J, Bozzetti F, et al. ESPEN guidelines on
chronic intestinal failure in adults. Clin Nutr. 2016;35(2):247-307.
https://doi.org/10.1016/j.clnu.2016.01.020
53. Waldron JL, Ashby HL, Cornes MP, et al. Vitamin D: a
negative acute phase reactant. J Clin Pathol. 2013;66(7):620-622.
https://doi.org/10.1136/jclinpath-2012-201301
54. Verhage AH, Cheong WK, Allard JP, Jeejeebhoy KN. Increase
in lumbar spine bone mineral content in patients on long-term
parenteral nutrition without vitamin D supplementation. JPEN
J Parenter Enteral Nutr. 1995;19(6):431-436. https://doi.org/10.
1177/0148607195019006431
55. Haderslev K, Tjellesen L, Haderslev P, Staun M. Assessment of the
longitudinal changes in bone mineral density in patients receiving home
parenteral nutrition. JPEN J Parenter Enteral Nutr. 2004;28(5):289-
294. https://doi.org/10.1177/0148607104028005289
56. Pironi L, Tjellesen l, De Francesco A, et al. Bone mineral density in
patients on home parenteral nutrition: a follow-up study. Clin Nutr.
2004;23(6):1288-1302. https://doi.org/10.1016/S0261-5614(04)00061-5
57. Cohen-Solal M, Baudoin C, Joly F, et al. Osteoporosis in patients
on long-term home parenteral nutrition: a longitudinal study.
J Bone Miner Res. 2003;18(11):1989-1994. https://doi.org/10.
1359/jbmr.2003.18.11.1989
58. Szczepanek K, Pedziwiatr M, Klek S. Denosumab improves bone
mineral density in patients with intestinal failure receiving home
parenteral nutrition: results from a randomized, controlled clinical trial
[published online ahead of print March 2017]. JPEN J Parenter Enteral
Nutr. https://doi.org/10.1177/0148607117695247
59. Lin JH. Bisphosphonates: a review of their pharmacokinetic properties.
Bone. 1996;18(2):75-85. https://doi.org/10.1016/8756-3282(95)00445-9
466
60. Haderslev K, Tjellesen L, Sorensen H, Staun M. Effect of cyclical
intravenous clodronate therapy on bone mineral density and
markers of bone turnover in patients receiving home parenteral
nutrition. Nutr Clin Pract. 2003;18(2):176-176. https://doi.org/
10.1177/0115426503018002176
61. Raman M, Aghdassi E, Baun M, et al. Metabolic bone disease
in patients receiving home parenteral nutrition: a Canadian study
and review. JPEN J Parenter Enteral Nutr. 2006;30(6):492-496.
https://doi.org/10.1177/0148607106030006492
62. Pazianas M, Compher C, Schiavone-Gatto P, Kinosian BP. In-
testinal failure-associated metabolic bone diseases and response to
teriparatide. Nutr Clin Pract. 2006;21(6):605-609. https://doi.org/
10.1177/0115426506021006605
63. Cummings SR, Martin JS, McClung MR, et al. Denosumab for preven-
tion of fractures in postmenopausal women with osteoporosis. N Engl
J Med. 2009;361(8):756-765. https://doi.org/10.1056/NEJMoa0809493
64. Nightingale JM, Lennard-Jones JE, Gertner DJ, et al. Colonic preser-
vation reduces need for parenteral therapy, increases incidence of renal
stones, but does not change high prevalence of gall stones in patients
with a short bowel. Gut. 1992;33:1493-1497.
65. Scales CD Jr, Smith AC, Hanley JM, Saigal CS. Prevalence of kidney
stones in the United States. Eur Urol. 2012;62(1):160-165.
66. Teichman JM. Clinical practice. Acute renal colic from ureteral calcu-
lus. N Engl J Med. 2004;350(7):684.
67. Clarke AM, McKenzie RG. Ileostomy and the risk of urinary uric acid
stones. Lancet. 1969;2(7617):395-397.
68. Tomson CRV. Nephrocalcinosis and nephrolithiasis. In: Nightingale
JMD, ed. Intestinal Failure. London, England: Greenwich Medical
Media Limited; 2001:227-242.
69. Nightingale JMD. Hepatobiliary, renal and bone complications of
intestinal failure. Best Pract Res Clin Gastroenterol. 2003:17(6):907-
929.
70. Deren JJ, Porush JG, Levitt MF, Khilnani MT. Nephrolithiasis as a
complication of ulcerative colitis and regional enteritis. Ann Inter Med.
1962;56(6):843-853.
71. Maratka Z, Nedbal J. Urolithiasis as a complication of the surgical
treatment of ulcerative colitis. Gut. 1964;5:214-217.
72. Bennett RC, Jepson RP. Uric acid stone formation following ileostomy.
ANZ J Surg. 1966;36(2):153-158.
73. Kennedy HJ, Lee ECG, Claridge G, Truelove SC. The health of
subjects living with a permanent ileostomy. Q J Med. 1982;51(203):341-
357.
74. Gallagher ND, Harrison DD, Skyring AP. Fluid and electrolyte distur-
bances in patients with long established ileostomies. Gut. 1962;3(3):219-
223.
75. Clarke AM, Chirnside A. Chronic dehydration and sodium depletion
in patients with established ileostomies. Lancet. 1967;2(7519):740-
743.
76. Kennedy HJ, Al-Dujaili EAS, Edwards CRW, Truelove SC. Water
and electrolyte balance in subjects with a permanent ileostomy. Gut.
1983;24(8):702-705.
77. Chadwick VS, Modha K, Dowling RH. Mechanism for hyperoxaluria
in patients with ileal dysfunction. N Engl J Med. 1973;289(4):172-176.
78. Dobbins JW, Binder HJ. Effect of bile salts and fatty acids on the
colonic absorption of oxalate. Gastroenterology. 1976;70(6):1096-1100.
Nutrition in Clinical Practice 33(4)
79. Smith LH, Fromm H, Hofmann AF. Acquired hyperoxaluria,
nephrolithiasis and intestinal disease. Description of a syndrome. N
Engl J Med. 1972;286(26):1371-1375.
80. Weinberg AE, Patel CJ, Chertow GM, Leppert JT. Diabetic severity
and risk of kidney stone disease. Eur Urol. 2014;65(1):242-247. https://
doi.org/10.1016/j.eururo.2013.03.026
81. Goldfarb DS, Asplin JR. Effect of grapefruit juice on urinary
lithogenicity. J Urol. 2001;166(1):263-267.
82. Hylander E, Jarnum S. Jensen HJ, Thale M. Enteric hyperoxaluria:
dependence on small intestinal resection, colectomy, and steator-
rhoea in chronic inflammatory bowel disease. Scand J Gastroenterol.
1978;13(5):577-588.
83. Ney DM, Hofmann AF, Fischer C, Stubblefield N. The Low Oxy-
late Diet Book. San Diego, CA: University of California Press;
1981.
84. Massey LK, Roman-Smith H, Sutton RA. Effect of dietary oxalate and
calcium on urinary oxalate and risk of formation of calcium oxalate
kidney stones. J Am Diet Assoc. 1993;93(8):901-906.
85. Holmes RP, Kennedy M. Estimation of the oxalate content of foods
and daily oxalate intake. Kidney Int. 2000;57(2000):1662-1667.
86. Zarembski PM, Hodgkinson A. The oxalic acid content of English
diets. Br J Nutr. 1962;16:627-634.
87. Kasidas GP, Rose GA. Oxalate contents of some common foods; deter-
mination of by an enzymatic method. J Hum Nutr. 1980;34(4):255-266.
88. Siener R, Honow R, Voss S, Seidler A, Hesse A. Oxalate content of
cereals and cereal products. J Agric Food Chem. 2006;54(8):3008.
89. Lemann J, Pleuss JA, Worcester EM, Hornick L, Schrab D, Hoffmann
RG. Urinary oxalate excretion increases with body size and decreases
with increasing dietary calcium intake among healthy adults. Kidney
Int. 1996;49(1):200-208.
90. Barilla DE, Notz C, Kennedy D, Pak CYC. Renal oxalate excretion
following oral oxalate loads in patients with ileal disease and with renal
and absorptive hypercalciurias: effect of calcium and magnesium. Am
J Med. 1978;64(4):579-585.
91. Hylander E, Jarnum S, Nelsen K. Calcium treatment of enteric
hyperoxaluria after JIB for morbid obesity. Scand J Gastoenterol.
1980;15(3):349-352.
92. Lyon ES, Borden TA, Ellis JE, Vermeulen CW. Calcium oxalate
lithiasis produced by pyridoxine deficiency and inhibition by high
magnesium diets. Invest Urol. 1966;4(2):133-142.
93. Sidhu H, Gupta R, Thind SK, Nath R. Oxalate metabolism in
thiamine-deficient rats. Ann Nutr Metab. 1987;31(6):354-361.
94. Argenzio RA, Liacoz JA, Allison MJ. Role of intestinal oxalate
degrading bacteria in the pathogenesis of enteric hyperoxaluria. Gas-
troenterology. 1985;88:1309.
95. Allison MJ, Cook HM, Milne DB, Gallagher S, Clayman RV. Ox-
alate degradation by gastrointestinal bacteria from humans. J Nutr.
1986;116(3):455-460.
96. Okombo J, Liebman M. Probiotic-induced reduction of gastrointesti-
nal oxalate absorption in healthy subjects. Urol Res. 2010;38(3):169-
178.
97. Mogna L, Pane M, Nicola S, Raiteri E. Screening of different pro-
biotic strains for their in vitro ability to metabolize oxalates: any
prospective use in humans? J Clin Gastroenterol. 2014;48(Suppl 1):
S91-S95.
Invited Review

Nutrition in Clinical Practice

Volume 33 Number 4
Dietary Fermentable Oligosaccharides, Disaccharides, August 2018 468–475

C 2018 American Society for

Monosaccharides, and Polyols (FODMAPs) and Parenteral and Enteral Nutrition

DOI: 10.1002/ncp.10108
Gastrointestinal Disease wileyonlinelibrary.com

Nimish Vakil, AGAF, FACG, FASGE1,2

Abstract
FODMAP is an acronym for fermentable oligosaccharides, disaccharides, monosaccharides, and polyols. Dietary modification of
FODMAPs has been shown to have significant effects on the physiology of the gastrointestinal tract and improves
symptoms of abdominal pain, distention, and bloating in patients with irritable bowel syndrome. Structured withdrawal and
reintroduction of FODMAPs supervised by a dietitian is the optimal practice for dietary FODMAP modification in irritable bowel
syndrome. FODMAPs are present in enteral feeding formulas and may have a role in diarrhea and bloating in tube-fed patients.
Emerging areas of research include the effects of dietary modification of FODMAPs on the microbiome, micronutrient absorption,
and caloric intake. FODMAP dietary modification is an emerging area in other gastrointestinal disorders and is of relevance to all
practicing dietitians. (Nutr Clin Pract. 2018;33:468–475)

Keywords
disaccharides; enteral nutrition; FODMAP; irritable bowel syndrome; inflammatory bowel diseases; microbiota; monosacccha-
rides; oligosaccharides; polysaccharides; sugar alcohols

Introduction glucosidases break down the dextrin and maltose further

FODMAP is an acronym for fermentable oligosaccharides,
disaccharides, monosaccharides and polyols. FODMAPs
are short-chain carbohydrates that have the following 3
characteristics: they are poorly absorbed in the small intes-
tine, they are fermentable, and they are osmotically active.
The FODMAP content of foods may be found in Table 1.
Carbohydrates and Their Absorption
Carbohydrates consist of sugars, including monosaccha-
rides and disaccharides; polyols, which are sugar alco-
hols such as sorbitol and xylitol; and oligosaccharides,
which are carbohydrates containing 3–10 monosaccharides
found in plants such as onions. Polysaccharides are large
chains of monosaccharides. Polysaccharides may be de-
signed for energy storage, such as glycogen or starch or
serve a structural function such as cellulose or chitins. A
fructan is a polysaccharide made up of fructose. Fructo-
oligosaccharides are fructans with a short chain length. A
detailed description of carbohydrates and their absorption
may be found elsewhere.1
The digestion of carbohydrates begins with the break-
down of complex carbohydrates by salivary and pancreatic
amylase into simpler sugars such as dextrin and maltose.
Enzymes in the brush border of the small intestine called
and with other disaccharidases (lactase and sucrase) break
sugars down further into monosaccharide units. Lactose
is a disaccharide composed of glucose and galactose and
is broken down into its component monosaccharides by
lactase. In the absence of the lactase enzyme, lactose is
carried to the distal small bowel and colon where it is
used by bacteria and can cause symptoms of bloating and
diarrhea. Glucose, galactose, and fructose are absorbed
through the epithelium and are transported by the portal
vein to the liver. At low concentrations, glucose is absorbed
using a sodium-dependent active transporter. At higher
concentrations, a second facilitated transporter becomes
active. Galactose is absorbed using the same transporters.
Fructose is an important monosaccharide that is present
in a free from in the diet, as a subunit of a disaccharide
From the 1 University of Wisconsin School of Medicine and Public
Health, Madison, Wisconsin, USA; and 2 Aurora Health Care,
Summit, Wisconsin, USA.
Financial disclosure: None declared.
Conflicts of interests: None declared.
This article originally appeared online on June 5, 2018.
Corresponding Author:
Nimish Vakil, MD, Aurora Summit Medical Center, 36500 Aurora
Drive, Summit, WI 53066.
Email: nvakil@wisc.edu
Vakil 469

Table 1. High FODMAP Foods by Component.

Fructans and Galacto

Fructose Lactose oligo-saccharides Polyols

Fruits: apples, cherries, figs,
mango, pears, watermelon
Vegetables: artichokes,
asparagus, dried tomatoes
and sugar snap peas
Sweeteners: honey, high
fructose corn syrup, agave
nectar
Alcohol: sherry, port, rum
sweet wines
Cheese: soft cheeses (ricotta,
cottage, mozzarella)
Milk: sheep, cow, goat and
buffalo
Milk products: yogurt,
ice-cream, custard,
condensed and evaporated
milk
Legumes: beans (black, fava,
kidney, navy), soy (beans,
flour, milk)
Nuts and grains: wheat, rye,
barley, pistachios and
cashews, Almonds (>10)
Vegetables: garlic, leek,
scallion, onion or garlic
powder
Fruit: grapefruit, watermelon. Vegetables: cauliflower, celery,
Plums, prunes, peaches, figs,
dates, currants and banana
Other: teas (oolong,
chamomile and fennel),
inulin (also appears as
chicory root extract), carob
Sweetening agents: sorbitol,
mannitol, xylitol, isomal
Fruits: nectarines, peaches,
pears, blackberries, cherries,
plums, prunes, watermelon,
lychee
Sweetners: isomalt, sorbitol,
mannitol, xylitol, maltitol,
lactitol
snow peas, sweet potato
Other: hard candies, toffee,
jams and preserves, chewing
gum, chocolates, protein
powders, baked goods, cough
drops and throat lozenges

FODMAP, fermentable oligosaccharides, disaccharides, monosaccharides, and polyols.

and as a component of the polysaccharide fructan. Fructose
consumption has increased in the U.S. diet in recent years.
The prevalence of high-fructose corn syrup used as a
sweetener in various foods increased from 16% in 1977 to
42% in 1998 and has remained stable at that rate.2 Fructose
is absorbed by 2 transporters GLUT-2 and GLUT-5, which
are located in the brush border epithelium of the small
intestine. Absorption through the GLUT-2 pathway is facil-
itated by ingestion of glucose, but the GLUT-5 pathway is
independent of glucose coadministration.3 Approximately
half the population cannot tolerate 25 grams of fructose.4
With the average consumption of fructose ranging from 11–
54 grams a day, the increasing use of fructose as a sweetener
exposes many patients to high doses of fructose.
An exception to the absorptive mechanisms described
previously is the handling of dietary fiber. Fiber is a complex
polysaccharide that cannot be absorbed and passes through
the gastrointestinal (GI) tract into the stool. Colonic bacte-
ria may metabolize some parts of dietary fiber.
Physiologic Effects of FODMAPs
Effect on intestinal water. The effect of high FODMAP
diets has been studied in humans with an ileostomy by ex-
amining the fluid effluent from the ileostomy. Other studies
have measured intestinal water by using magnetic resonance
imaging to quantify intestinal water.5,6 High FODMAP
diets increase intestinal water. This may contribute to pain
because of intestinal distention in patients with visceral hy-
persensitivity and diarrhea as a result of the increased water
content of luminal content.7 Poorly absorbed carbohydrates
such as the FODMAPs increase intestinal water by their
osmotic activity.8
Effects on colonic gas. Passage of undigested or unab-
sorbed short-chain carbohydrates into the colon provides
a substrate for colonic bacteria to produce hydrogen and
methane. Studies in healthy individuals and patients with
irritable bowel syndrome using magnetic resonance imag-
ing show a significant increase in colonic gas with high
FODMAP diets. In patients with irritable bowel syndrome,
the increased gas is associated with the development of
abdominal symptoms.9,10 Studies using magnetic resonance
imaging show that polysaccharides such as inulin commonly
derived from plant products such as chicory root increase
colonic gas in both healthy individuals and patients with
irritable bowel syndrome and are associated with symptoms
in patients with irritable bowel syndrome.5 A double-blind,
placebo-controlled trial of the administration of a fructan
(maltodextrin) to children with irritable bowel syndrome
showed a significant worsening of bloating, flatulence, and
abdominal pain symptoms with fructan exposure.11
Effects on short-chain fatty acids and visceral hypersensitivity.
In addition to the production of methane and hydrogen,
the fermentation of short-chain carbohydrates in the colon
results in the production of short-chain fatty acids. An
important feature of functional gastrointestinal disorders
such as irritable bowel syndrome is the heightened sensi-
tivity of these patients to gut stimulation or distention, a
470
condition termed visceral hypersensitivity. Short-chain fatty
acids in luminal contents have been shown create visceral
hypersensitivity in animal studies.12,13 Low FODMAP diets
are associated with a decrease in liposaccharides (a complex
of fat and polysaccharides) in the luminal content of the
colon.11 Intracolonic administration of fecal supernatant
from patients with irritable bowel syndrome induces visceral
hypersensitivity in rats.11 Administration of a lipopolysac-
charide antagonist blocked the increase and fecal super-
natant from healthy individuals, and patients with irritable
bowel syndrome on a low FODMAP diet did not have
an effect on visceral hypersensitivity.11 Low FODMAP
diets may therefore alter visceral sensitivity by changing
the composition of luminal contents (decreasing luminal
short-chain fatty acids and liposaccharides). Short-chain
fatty acids have a beneficial effect on epithelial function,
and therefore concerns remain about the prolonged use
of low FODMAP diets on colonic health and the risk of
colon cancer.8 There are a number of limitations with the
studies that have been performed to date, including small
sample sizes, conflicting results, and difficulties with the
measurement of short-chain fatty acids in the ascending
colon.8
Effect on the microbiome. The gastrointestinal tract harbors
a large number of organisms in an ecosystem that is
unique for the individual and relatively stable over time.14
This ecosystem consists of luminal bacteria and bacteria
that are adherent to the mucosal surface. The microbiota
communicates with wall of the organ and can influence
function. The 2 main families or phyla of organisms are
Firmicutes and Bateroidetes that together make up 90%
of the microbiome.15 Alterations in the microbiome could
change the cross-talk between the bacteria and intestine,
altering function and creating symptoms. Diets that differ
in FODMAP content have been shown to have an effect on
the gut microbiome of the host.16 In patients with irritable
bowel syndrome, symptoms improved on a low FODMAP
diet, but concentrations of bifidobacter (a beneficial or-
ganism in the colon) decreased.17,18 A probiotic coadmin-
istered with the low FODMAP diet restored bifidobacter
concentrations.15 A small study showed that a 3-week low
FODMAP diet when compared with a high FODMAP diet
resulted in a higher abundance of Adlercreutzia, a genus
that uses hydrogen.17 Investigation into the effects on the
FODMAP diet on the microbiome are in their infancy.
Effect on the metabolome. The metabolome refers to the
complete set of small-molecule chemicals found within a
biological sample. Within the human body there is an en-
dogenous metabolome and also a food-related metabolome,
which is of growing interest. The food metabolome is
defined as the part of the human metabolome directly
derived from the digestion and biotransformation of foods
Nutrition in Clinical Practice 33(4)
and their constituents. In a randomized, controlled, single-
blind trial, McIntosh et al19 evaluated the effect of low and
high FODMAP diets on the food-related metabolome by
measuring histamine, p Hydroxybenzoic acid, and azelaic
acid in the urine of patients with irritable bowel syndrome.
In patients on a high FODMAP diet, urinary levels of
histamine increased significantly, whereas histamine levels
dropped in patients randomized to the low FODMAP
diet. Histamine is an endogenous amine that is thought
to play a role in irritable bowel syndrome and studies
blocking histamine have shown a benefit in patients with
this disorder.20 A number of other agents may be involved
in the genesis of symptoms in irritable bowel syndrome, but
additional research is necessary in this area.
Disease States
Irritable Bowel Syndrome
Irritable bowel syndrome is a common and debilitating
medical condition that is associated with abdominal pain,
bloating distention, and changes in bowel habit. No single
curative treatment exists, and patients often report an asso-
ciation of symptoms with certain foods. A detailed review
of the criteria for a diagnosis may be found elsewhere.21
Short-term studies (4–6 weeks) on FODMAP restriction in
irritable bowel syndrome. Of the patients in various studies,
50%–80% report an improvement of their symptoms on
a low FODMAP diet. A recent meta-analysis evaluated 6
randomized controlled trials and 16 nonrandomized tri-
als and reported substantial improvements in abdominal
pain, bloating, and overall symptoms with odds ratios
ranging from 1.75–1.81.22 Dietary intervention with low
FODMAP dietary education when compared with sham
(placebo) dietary intervention has shown a substantial
benefit for dietary advice given by a dietitian when com-
pared with placebo with regard to irritable bowel syndrome
symptoms.15
Long-term results of FODMAP restriction on irritable bowel
syndrome. The evidence for long-term efficacy of a low
FODMAP diet in irritable bowel syndrome is limited in
quality and in the number of studies available. A dietary
strategy of reducing fructose and fructans was effective in
alleviating symptoms in 74% of patients at 14 months.23
The study was small, and recall bias limits the validity of
the observed benefit. A randomized trial comparing gut-
directed hypnotherapy with a low FODMAP diet evaluated
patients at 6 months and reported that 82% of patients in the
diet therapy group had improved symptoms when compared
with their baseline. Hypnotherapy achieved similar results,
but the numbers of patients in each group were small (n =
25).24 A retrospective study of patients with irritable bowel
Vakil
syndrome followed for 16 months on a low FODMAP diet
suggested that one third of patients continued to adhere
to the diet, whereas 84% of patients used a modified low
FODMAP diet. Of the patients, 54% used the diet on and
off based on symptom severity; 54% of patients reported
a partial improvement in symptoms and 32% reported
complete resolution.25 The British Dietetic Association rec-
ommends the low FODMAP diet as a second-line strategy
with dietary advice to be provided by a dietitian.26
Nonceliac gluten sensitivity. Nonceliac gluten sensitivity
is a disorder characterized by abdominal symptoms that
improve after gluten withdrawal in the absence of celiac
disease. This is one of the fastest growing segments of the
food industry in North America, and the sale of gluten-
free foods is expected to reach $7.5 billion in 2020.27 In a
double-blind crossover trial of patients with irritable bowel
syndrome and nonceliac gluten sensitivity, Biesiekierski et
al28 administered a low FODMAP diet for 2 weeks followed
by high-gluten (16 g gluten/day), low-gluten (2 g gluten/day
and 14 g whey protein/day), or control (16 g whey pro-
tein/day) diets for 1 week. In all patients, symptoms im-
proved significantly on a low FODMAP diet and worsened
equally when gluten or whey protein was added to the
diet, suggesting the lack of a specific effect for gluten in
aggravating symptoms. Wheat contains several potential
symptom inducers including gluten, fructans, and soluble
proteins. Wheat-containing foods are a major source of
fructans in the U.S. diet. To determine the effects of gluten
compared with fructans, Skodje et al29 studied patients
who did not have celiac disease by diagnostic tests and
who were on a self-imposed, gluten-free diet for relief of
abdominal symptoms.29 Participants were randomized in a
blinded manner to diets containing gluten (5.7 g), fructans
(2.1 g), or placebo concealed in muesli bars for 7 days.
After a washout period, patients were re-randomized until
every patient received each of the dietary interventions.
Symptoms were recorded using validated questionnaires.
Symptoms of bloating and overall symptoms of irritable
bowel syndrome were substantially worse in patients given
fructans, but there was no difference in symptoms between
those given placebo or gluten-enriched bars. These data
suggest that fructans may be the cause of symptoms in many
patients who report gluten sensitivity in the absence of celiac
disease. From a clinical standpoint, it is reasonable to treat
patients who present with gluten sensitivity and who do not
have celiac disease with the structured dietary approach for
the low FODMAP diet that is described later because their
symptoms may be related to fructans in wheat and they may
be sensitive to other FODMAPs as well.
Inflammatory bowel disease. Inflammatory bowel disease
(Crohn’s disease and ulcerative colitis) is known to be
affected by changes in the fecal microbiome, and some pa-
471
tients with these disorders report aggravation of symptoms
with certain foods. There are limited data on the FODMAP
diet in inflammatory bowel disease. In a small study of
patients with quiescent Crohn’s disease, a low FODMAP
diet was associated with changes in the fecal microbiome
when compared with a normal Australian diet.30 Symp-
toms of irritable bowel syndrome coexist in patients with
inflammatory bowel disease. A recent meta-analysis found
that patients with inflammatory bowel disease who were
still symptomatic despite adequate control of their disease
by objective tests (markers of inflammation or endoscopic
evidence of active disease) had significant improvement in
abdominal symptoms with a low FODMAP diet.30,31
Predicting a Response to the FODMAP Diet
Because of the cost and inconvenience associated with
the low FODMAP diet, a number of strategies are being
evaluated to determine if a response to the diet can be
predicted by diagnostic tests. A recent study suggests that
fecal bacterial profiles using a commercially available assay
may predict a response to the FODMAP diet in irritable
bowel syndrome. Bacterial abundance was higher in nonre-
sponders to the low FODMAP diet when compared with
responders.32,33 In a small recent study, the measurement
of volatile organic compounds in stool using a low-cost
assay predicted a response to a low FODMAP diet in 100%
of patients.34 Fructose and lactose breath testing are not
predictive of a response to a low FODMAP diet.35
Instructing Patients on the FODMAP Diet
The management of a FODMAP diet in clinical practice has
been proposed as consisting of the following 4 stages (Figure
1): (a) initial dietitian visit for FODMAP restriction, (b)
4–6 week dietitian visit for FODMAP reintroduction and
(c) elective third visit for FODMAP personalization, and
(d) long-term follow-up of irritable bowel syndrome and
nutrition status.36 During the first visit, there is a detailed
dietary assessment including an evaluation of dietary pref-
erences and foods known to cause symptoms. A nutrition
assessment is also useful as some patients may lose weight
on the FODMAP diet and the long-term nutrition conse-
quences await definition. A description of the FODMAP
diet, the association of FODMAPs with symptoms, and
the results of FODMAP withdrawal are provided along
with counseling on FODMAP restriction. Table 1 provides
a list of common foods containing FODMAPs. The initial
phase of management consists of removal of all the po-
tential sources of FODMAPs to reduce their concentration
in the diet below the threshold of symptom generation.
It is important to emphasize to the patient that the diet
is not permanent but compliance with the initial phase
is important to determine if the patient is going to be
a responder to dietary intervention. The importance of
472
Figure 1. Timeline for fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (FODMAP) intervention.
fructans in wheat (as opposed to gluten sensitivity) should
be explained. Some patients need a discussion of other
diets that they may have tried with varying benefits. A
number of diets are described in books and on the internet,
including the specific carbohydrate diet, the paleo diet, and
the Candida diet. Some of these diets restrict FODMAPs
and can result in symptom improvement, but the scientific
basis for these diets is inadequate and there is no structured
mechanism to reintroduce foods into the diet. Our approach
to the first visit for dietary intervention has evolved from
giving the patient a handout (which was ineffective) to a
structured hour-long visit when the patient meets with the
dietitian and the gastroenterologist. The existing therapeu-
tic relationship with the physician is an important part of
the management of irritable bowel syndrome and can be
used to reassure the patient that they can follow the diet and
to enhance compliance with diet therapy.37,38 Patients fre-
quently ask questions regarding the use of alcohol. Alcohol
can cause gastrointestinal symptoms by itself. Beer is low
in FODMAPs despite its origins in barley and rye because
the fermentation process destroys many fructans. Rum and
fortified wines such as sherry and port contain high amounts
of fructose and are forbidden in the initial phase. Dry wines
(defined as <4 g of sugar per liter) are safe during the
initial phase and include varietals such as sauvignon blanc,
albarino, and chardonnay in the recommended dose of 150
mL. Whiskey, vodka, and gin are low FODMAP alcohols
in a dose of 30 mL per day. The major risk of FODMAP
exposure comes from the mixers used with these drinks,
such as tonic water and soda. Exercise enthusiasts need to
be made aware of chicory root in energy bars, which often
cause distress during exercise because of their high fructan
content (Table 1). The duration of the initial restrictive
phase is in a state of evolution because of the effects of
the low FODMAP diet on the microbiota. Randomized
controlled trials have generally restricted FODMAPs for
6 weeks or longer, but recent descriptions of a reduction
in bifidobacter in the colon have prompted many groups
to reevaluate the duration of complete FODMAP with-
Nutrition in Clinical Practice 33(4)
drawal. A 4-week period of complete FODMAP withdrawal
is considered adequate.35 Supplementing with a probiotic
containing bifidobacter has been shown to be effective in
replenishing the microbiome and should be discussed with
the patient.
Structured Reintroduction
There is substantial evidence from randomized controlled
trials for the FODMAP withdrawal phase, but little struc-
tured evidence on the best method to reintroduce foods or
the order in which they should be reintroduced. Individual
centers have developed their own strategies for the struc-
tured addition of foods back to the diet. Our technique
is illustrated in Table 2, and the instructions provided
to the patient are summarized in Figure 2. We typically
schedule a 30-minute dietitian visit to explain the strategy of
gradual dose escalation of foods shown in Table 2. A daily
symptom chart is provided to the patient, and symptoms
are recorded every day. For example, with lactose addition,
a half cup is ingested on Monday, 1 cup on Tuesday, and
1½ cups on Wednesday. If the patient develops symptoms
on Monday, the experiment is stopped for that week and
lactose is listed as a substance the patient cannot tolerate.
If the patient is able to tolerate lactose, the experiment is
ended on Wednesday. From Thursday–Sunday, the patient
returns to the low FODMAP diet that was prescribed for
the preceding 4 weeks. On Monday, the next agent on the
list is tried using the same protocol. At the end of the
structured reintroduction, the patient will know the foods
that cannot be tolerated even in low doses. Other foods may
be tolerated but in limited amounts, and some foods may
generate no symptoms at all. Based on these responses, a
diet for long-term use that avoids some foods, limits others,
and permits others is crafted for the patient. Symptoms
are recorded every day, and it is immediately apparent to
the patient whether lactose can be tolerated and at what
dose. The strategy is repeated in subsequent weeks until
complete. A recent study showed that patients who were
Vakil 473

Table 2. Structured FODMAP Reintroduction Plan.

Challenge week What to eat on challenge days

Week 1: lactose ½ cup milk or ¾ cup plain yogurt (without sweeteners or other FODMAPs)
Week 2: fructose ½ mango or 1–2 tablespoons of honey
Week 3: polyols-sorbitol Blackberries 3–10 or ¼ avocado
Week 4: polyols-mannitol 1/3 cup cauliflower or ¾ cup sweet potato
Week 5: fructans-wheat 2 slices of whole wheat bread or 1 cup cooked pasta
Week 6: fructans-onion 1 tablespoon diced onion
Week 7: fructans-garlic 1 clove of garlic
Week 8: galacto oligo-saccharides ½ cup kidney beans, black beans, or thawed peas or 15–25 almonds

FODMAP, fermentable oligosaccharides, disaccharides, monosaccharides, and polyols.

unaffected, but calcium intake decreased and overall caloric

• Test only one FODMAP group at a time
• Continue avoiding other FODMAPs
• Test with a food that only contains that one FODMAP
• Eat the prescribed amount of the test food (don’t over
do it)
• Try to test on 3 days during the week allowing a rest
day between challenges
• Use the same test food on each of the 3 days but
increase the dose by 50% (e.g. go from ½ cup to one
cup to 1½ cup)
• Track symptoms every day (not just the challenge
days)
intake declined in some patients.14,40 In our experience, some
patients lose a small amount of weight in the initial 4–
6 weeks of dietary FODMAP restriction. Some patients
choose to follow a very restricted diet, and future research
will help determine if micronutrient deficiencies occur when
these diets are administered long term. Monitoring for
micronutrient deficiencies should be individualized based
on the degree of dietary restriction.

FODMAPs in Enteral Feeding

Figure 2. Principles of food reintroduction for patients.
FODMAP, fermentable oligosaccharides, disaccharides,
monosaccharides, and polyols.
initially treated with a low FODMAP diet and were then
allowed to liberalize their diet using a structured approach
were able to maintain the symptom improvement seen in the
highly restrictive phase and were able to increase the amount
of fiber in the diet. Interestingly, the microbiome was not
affected in these patients.39
Patient Support
Some patients welcome support tools and applications that
help them to choose low FODMAP foods. Table 3 lists re-
sources that our patients have found useful in helping them
navigate the FODMAP diet. They include an app that can
be used on mobile devices to identify the FODMAP content
of foods, charts that identify low and high FODMAP-
containing foods, recipes, and support from others on the
same journey.
Long-Term Monitoring of Nutrition Status
The effects the modified personalized FODMAP diet on
the gut microbiome are being actively studied as is the
benefit of adding probiotics. There is little information from
carefully conducted trials regarding the long-term nutrition
impact of the FODMAP diet. In one study, iron intake was
FODMAPs may be important in enteral feeding. Diarrhea
is a commonly encountered problem with enteral feeding
in clinical practice. The FODMAP content of the enteral
feeding solution may play a role in the genesis of symptoms.
A retrospective study of patients with diarrhea caused by
enteral feeding in Australia suggested that a longer hospital
stay and a longer duration of enteral feeding were risk
factors for diarrhea. Formula FODMAP levels in that
study ranged from 10.6–36.5 g/day. A low FODMAP
enteral formula was associated with a 5-fold reduction in
diarrhea rates.41 In a randomized, controlled trial of a low,
moderate, and high FODMAP enteral feeds in Korea, Yoon
et al42 demonstrated a significant reduction in diarrhea
and improvement in nutrition parameters and clinical
outcome in patients randomized to the low FODMAP
enteral feeding formula. Quantifying FODMAPs in enteral
formulas has proven difficult because of the interference
with in vitro assays of fructans and raffinose caused by
the maltodextrin content of the formula.43 At the present
time, there is no comprehensive list of low FODMAP-
containing enteral formulas in the United States, and the
FODMAP content of Australian enteral feeding formulas
described by Halmos et al41 may not apply to North
America. Some enteral feeding formulas contain fructose
and others contain inulin, and these are likely to have a
high FODMAP content. Controversy persists regarding the
use of fiber supplements in enteral tube feeding, and this
debate has recently been summarized elsewhere.44 It should
be noted that many enteral feeding formulas supplemented
474 Nutrition in Clinical Practice 33(4)

Table 3. Resources for Patients Being Started on the FODMAP Diet.

Resource Cost Link Comment

Monash University Mobile App
Low FODMAP recipes & support
group
Low FODMAP foods
Reintroducing FODMAPs book
A little bit yummy
FODMAP cookbooks
Low and high FODMAP checklists Free
$12.99 Apple and android app stores Links to an extensive database
of FODMAP containing
foods
Free https://www.facebook.com/groups/7435 Recipes and support group
58315679493/
Variable https://www.fodyfoods.com Sites for low FODMAP foods
http://trueselffoods.com including energy bars
$17.99 https://reintroducingfodmaps.com/ A book on FODMAP
reintroduction of patients
who need additional
information
Free https://app.alittlebityummy.com/meal-plan/ Multiple FODMAP resources
Variable Online book stores FODMAP Cookbooks written
by registered dietitians
http://www.katescarlata.com/low Checklist for low and high
fodmapdietchecklists/ FODMAP containing foods

FODMAP, fermentable oligosaccharides, disaccharides, monosaccharides, and polyols.

with fiber have another FODMAP added to the solution References

(fructo-oligosaccharides).43 Further work is urgently
needed in this area.
Future Directions With FODMAP and Dietary
Therapy
Future research is directed at test strategies to determine
who will benefit from the low FODMAP diet. We need more
data on the long-term outcome of treatment with the low
FODMAP diet and its effects on the microbiome and the
nutrition status of the individual. An important area of
research is in food science. The development of enzymes that
break down fructans in food could help enormous numbers
of people who currently purchase gluten-free products at
substantial cost.
Acknowledgments
I am grateful to Amanda Motl, RD, for the care of our
patients and for the information on the reintroduction phase
of FODMAP dietary therapy.
Statement of Authorship
N. Vakil contributed to conception/design of the research; N.
Vakil contributed to acquisition, analysis, or interpretation
of the data; N. Vakil drafted the manuscript; Nimish Vakil
critically revised the manuscript; and N. Vakil agree to be fully
accountable for ensuring the integrity and accuracy of the
work.
Supplementary Information
Additional supporting information may be found online in the
Supporting Information section at the end of the article.
1. Kelm N, Levin R, Havel P. Carbohydrates in Modern Nutrition in
Health and Diseases. 11th ed. Philadelphia: Lippincott Williams &
Wilkins; 2014:36-57.
2. Marriott B, Cole N, Lee N. National estimates of dietary fructose
intake increase from 1977 to 2004 in the United States. J Nutr.
2009;139(6):1228S-1235S.
3. Noelting J, DiBaise J. Mechanisms of fructose absorption. Clin Transl
Gastroenterol. 2015;6(11):e120.
4. Gibson PR, Newnham E, Barrett JS, Shepherd SJ, Muir JG. Review ar-
ticle: fructose malabsorption and the bigger picture. Aliment Pharmacol
Ther. 2007;25(4):349-363.
5. Barrett JS, Gearry RB, Muir JG, et al. Dietary poorly absorbed, short-
chain carbohydrates increase delivery of water and fermentable sub-
strates to the proximal colon. Aliment Pharmacol Ther. 2010;31(8):874-
882.
6. Marciani L, Cox EF, Hoad CL, et al. Postprandial changes in small
bowel water content in healthy subjects and patients with irritable bowel
syndrome. Gastroenterology. 2010;138(2):469-477.
7. Major G, Pritchard S, Murray K, et al. Colon hypersensitivity to dis-
tension, rather than excessive gas production, produces carbohydrate-
related symptoms in individuals with irritable bowel syndrome. Gas-
troenterology. 2017;152(1):124-133.
8. Shepherd S, Lomer M, Gibson P. Short-chain carbohydrates and func-
tional gastrointestinal disorders. Am J Gastroenterol. 2013;108(5):707-
717.
9. Ong DK, Mitchell SB, Barrett JS, et al. Manipulation of dietary short
chain carbohydrates alters the pattern of gas production and genesis
of symptoms in irritable bowel syndrome. J Gastroenterol Hepatol.
2010;25(8):1366-1373.
10. Staudacher HM, Whelan K. The low FODMAP diet: recent advances
in understanding its mechanisms and efficacy in irritable bowel syn-
drome. Gut. 2017;66(8):1517-1527.
11. Chumpitazi BP, McMeans AR, Vaughan A, et al. Fructans exacerbate
symptoms in a subset of children with irritable bowel syndrome. Clin
Gastroenterol Hepatol. 2018;16(2):219-225.
12. Bourdu S, Dapoigny M, Chapuy E, et al. Rectal instillation of butyrate
provides a novel clinically relevant model of noninflammatory colonic
hypersensitivity in rats. Gastroenterology. 2005;128(7):1996-2008.
Vakil
13. Zhou SY, Gillilland M 3rd, Wu X, et al. FODMAP diet mod-
ulates visceral nociception by lipopolysaccharide-mediated intestinal
inflammation and barrier dysfunction. J Clin Invest. 2018;128(1):267-
280.
14. Gut Microbiota for Health. http://www.gutmicrobiotaforhealth.com/
en/about-gut-microbiota-info/. Accessed March 24, 2018.
15. Staudacher H, Whelan K. Altered gastrointestinal microbiota in ir-
ritable bowel syndrome and itsmodification by diet: probiotics, pre-
biotics and the low FODMAP diet. Proc Nutr Soc 2016;75(3):306-
318.
16. Halmos EP, Christophersen CT, Bird AR, Shepherd SJ, Gibson PR,
Muir JG. Diets that differ in their FODMAP content alter the colonic
luminal microenvironment. Gut. 2015;64(1):93-100.
17. Staudacher HM, Lomer MCE, Farquharson FM, et al. A diet
low in FODMAPs reduces symptoms in patients with irritable
bowel syndrome and a probiotic restores bifidobacterium species:
a randomized controlled trial. Gastroenterology. 2017;153(4):936-
947.
18. Staudacher HM, Lomer MC, Anderson JL, et al. Fermentable car-
bohydrate restriction reduces luminal bifidobacteria and gastrointesti-
nal symptoms in patients with irritable bowel syndrome. J Nutr.
2012;142(4):1510-1518.
19. McIntosh K, Reed DE, Schneider T, et al. FODMAPs alter
symptoms and the metabolome of patients with irritable bowel
syndrome: a randomised controlled trial. Gut. 2017;66(7):1241-
1251.
20. Fabisiak A, Włodarczyk J, Fabisiak N, Storr M, Fichna J. Targeting
histamine receptors in irritable bowel syndrome: a critical appraisal.
J Neurogastroenterol Motil. 2017;23(3):341-348.
21. Lacy B, Mearin F, Chang L, et al. Bowel disorders. Gastroenterology.
2016;150(6):1393-1407.
22. Marsh A, Eslick EM, Eslick GD. Does a diet low in FODMAPs
reduce symptoms associated with functional gastrointestinal disorders?
A comprehensive systematic review and meta-analysis. Eur J Nutr.
2016;55(3):897-906.
23. Shepherd SJ, Gibson PR. Fructose malabsorption and symptoms of
irritable bowel syndrome: guidelines for effective dietary management.
J Am Diet Assoc. 2006;106(10):1631-1639.
24. Peters SL, Yao CK, Philpott H, et al. Randomised clinical trial: the
efficacy of gut directed hypnotherapy is similar to that of the low
FODMAP diet for the treatment of irritable bowel syndrome. Aliment
Pharmacol Ther. 2016;44(5):447-459.
25. Maagaard L, Ankersen DV, Végh Z, et al. Follow-up of patients with
functional bowel symptoms treated with a low FODMAP diet. World
J Gastroenterol. 2016;22(15):4009-4019.
26. McKenzie YA, Bowyer RK, Leach H, et al. British Dietetic Association
systematic review and evidence-based practice guidelines for the dietary
management of irritable bowel syndrome in adults (2016 update).
J Hum Nutr Diet. 2016;29(5):549-575.
27. Gluten-Free Products Market Analysis by Product (Bakery, Dairy
Alternatives, Desserts & Ice-Creams, Prepared Foods, Pasta & Rice),
By Distribution (Grocery Stores, Mass Merchandiser, Club Stores),
and Segment Forecasts, 2018–2025. https://www.grandviewresearch.
com/industry-analysis/gluten-free-products-market. Accessed March
3, 2018.
28. Biesiekierski JR, Peters SL, Newnham ED, Rosella O, Muir JG, Gib-
son PR. No effects of gluten in patients with self-reported non-celiac
gluten sensitivity after dietary reduction of fermentable, poorly ab-
sorbed, short-chain carbohydrates. Gastroenterology. 2013;145(2):320-
328.
29. Skodje GI, Sarna VK, Minelle IH, et al. Fructan, rather than gluten,
induces symptoms in patients with self-reported non-celiac gluten
sensitivity [published online ahead of print November 2, 2017]. Gas-
troenterology. https://doi.org/10.1053/j.gastro.2017.10.040
475
30. Halmos EP, Christophersen CT, Bird AR, Shepherd SJ, Muir JG,
Gibson PR. Consistent prebiotic effect on gut microbiota with altered
FODMAP intake in patients with Crohn’s disease: a randomised,
controlled cross-over trial of well-defined diets. Clin Transl Gastroen-
terol. 2016;7:e164. https://doi.org/10.1038/ctg.2016.22
31. Zhan YL, Zhan YA, Dai SX. Is a low FODMAP diet beneficial
for patients with inflammatory bowel disease? A meta-analysis and
systematic review [published online ahead of print May 24, 2017]. Clin
Nutr. https://doi.org/10.1016/j.clnu.2017.05.019
32. Bennet SMP, Böhn L, Störsrud S, et al. Multivariate modelling of fae-
cal bacterial profiles of patients with irritable bowel syndrome predicts
responsiveness to a diet low in FODMAPs [published online ahead of
print April 17, 2017]. Gut. https://doi.org/10.1136/gutjnl-2016-313128
33. Valeur J, Småstuen MC, Knudsen T, Lied GA, Røseth AG. Exploring
gut microbiota composition as an indicator of clinical response to
dietary FODMAP restriction in patients with irritable bowel syndrome
[published online ahead of print January 4, 2018]. Dig Dis Sci.
https://doi.org/10.1007/s10620-017-4893-3
34. Rossi M, Aggio R, Staudacher HM, et al. Volatile organic
compounds in feces associate with response to dietary intervention
in patients with irritable bowel syndrome [published online
ahead of print October 7, 2017]. Clin Gastroenterol Hepatol.
https://doi.org/10.1016/j.cgh.2017.09.055
35. Wilder-Smith CH, Olesen SS, Materna A, Drewes AM. Predictors
of response to a low-FODMAP diet in patients with functional
gastrointestinal disorders and lactose or fructose intolerance. Aliment
Pharmacol Ther. 2017;45(8):1094-1106.
36. Whelan K, Martin LD, Staudacher HM, Lomer MCE. The low
FODMAP diet in the management of irritable bowel syndrome: an
evidence-based review of FODMAP restriction, reintroduction and
personalisation in clinical practice [published online ahead of print
January 15, 2018]. J Hum Nutr Diet. https://doi.org/10.1111/jhn.12530
37. Halpert A, Godena E. Irritable bowel syndrome patients’ perspectives
on their relationships with healthcare providers. Scand J Gastroenterol.
2011;46(7-8):823-830.
38. Hulme K, Chilcot J, Smith MA. Doctor-patient relationship and
quality of life in irritable bowel syndrome: an exploratory study of the
potential mediating role of illness perceptions and acceptance. Psychol
Health Med. 2017;1-11. https://doi.org/10.1017/S0033291717003324.
[Epub ahead of print]
39. Harvie RM, Chisholm AW, Bisanz JE, et al. Long-term irritable
bowel syndrome symptom control with reintroduction of selected
FODMAPs. World J Gastroenterol. 2017;23(25):4632-4643.
40. Böhn L, Störsrud S, Liljebo T, et al. Diet low in FODMAPs
reduces symptoms of irritable bowel syndrome as well as tradi-
tional dietary advice: a randomized controlled trial. Gastroenterology.
2015;149(6):1399-1407.e2
41. Halmos EP, Muir JG, Barrett JS, Deng M, Shepherd SJ, Gibson PR.
Diarrhoea during enteral nutrition is predicted by the poorly absorbed
short-chain carbohydrate (FODMAP) content of the formula. Aliment
Pharmacol Ther. 2010;32(7):925-933.
42. Yoon SR, Lee JH, Lee JH, et al. Low-FODMAP formula improves
diarrhea and nutritional status in hospitalized patients receiving enteral
nutrition: a randomized, multicenter, double-blind clinical trial. Nutr J.
2015;14:116.
43. Halmos EP, Bogatyrev A, Ly E, Liels KL, Muir JG, Gibson PR.
Challenges of quantifying FODMAPs in enteral nutrition formulas:
evaluation of artifacts and solutions. JPEN J Parenter Enteral Nutr.
2017;41(8):1262-1271.
44. Tarleton S, Kraft C, DiBaise J. Fiber-enriched enteral formulae:
advantageous or adding fuel to the fire? Pract Gastroenterol. https://
med.virginia.edu/ginutrition/wp-content/uploads/sites/.../Parrish_Dec
_13.pdf accessed 2/4/2018
Invited Review

Nutrition in Clinical Practice

Volume 33 Number 4
Nutrition Management of Necrotizing Enterocolitis August 2018 476–482

C 2018 American Society for

Parenteral and Enteral Nutrition

DOI: 10.1002/ncp.10115
wileyonlinelibrary.com

Vikram J. Christian, MBBS1 ; Elizabeth Polzin, MBA, RD, CD, CNSC2 ;

and Scott Welak, MD3

Abstract
Necrotizing enterocolitis (NEC) is one of the most significant causes of morbidity and mortality among premature infants. The
exact cause is considered multifactorial and related to gastrointestinal immaturity, inflammation and enteral feeding. The role of
nutrition is vitally important in NEC. The main modifiable risk factor is the introduction and advancement of enteral feedings.
After an infant has recovered from NEC, enteral feeds should be cautiously resumed to prevent injury from prolonged use of
parenteral nutrition. The logistics of how, when, and what to feed are somewhat unclear and often depend on the severity of the
disease. For patients with an enterostomy, refeeding the distal intestine with the small-intestinal ostomy output may improve bowel
growth and prevent long-term complications. (Nutr Clin Pract. 2018;33:476–482)

Keywords
enteral nutriton; necrotizing enterocolitis; parenteral nutrition; premature infant

Introduction and Background Nutrition Before NEC

Necrotizing enterocolitis (NEC) is one of the most signif-
icant causes of morbidity and mortality among extremely
premature infants.1 Although the disease may affect infants
of any gestational age (0.2/1000 live births), the disease
occurs in 10% of infants born <1.5 kg.1 Mortality occurs
in 3% of all cases of NEC, but increases to 30% among
those who require surgical intervention.2 Survivors of NEC
are at risk for many long-term morbidities, most notably
gastrointestinal complications such as short gut syndrome
and dependence on parenteral nutrition (PN) for survival.
In addition, these infants face an increased risk for neurode-
velopmental impairment and chronic lung disease.2
NEC is a disease that is almost exclusively found in prema-
ture infants. The gastrointestinal tract develops throughout
pregnancy, both in tissue growth and maturity. When an
infant is born prematurely, that development is altered. Be-
cause of the complications with prolonged PN use, enteral
nutrition (EN) is required during the ex utero maturation of
the intestine.
From the 1 Division of Gastroenterology, Department of Pediatrics,
Medical College of Wisconsin, Milwaukee, Wisconsin, USA;
2 Clinical Nutrition Department, Children’s Hospital of Wisconsin,

Despite decades of basic scientific and clinical research,
the pathophysiology of NEC is still incompletely under-
stood. The exact cause is considered multifactorial and is
related to gastrointestinal immaturity, inflammation and
enteral feeding. Treatments such as ventilator support,
maintenance of hemodynamic stability, and intravenous
fluid resuscitation are not specific to treatment of NEC, but
are aimed at reducing its effects.
The role of nutrition is vitally important in NEC. The
main modifiable risk factor is the introduction and advance-
ment of enteral feedings. After recovery from NEC, many
infants have complications related to nutrition, including
malabsorption, surgical resection of the small intestine, and
special nutrition needs. This article reviews the relationship
between NEC and nutrition, including the effects of feeding
before the disease and changes that occur after infants
recover from the disease.
Milwaukee, Wisconsin, USA; and 3 Division of Neonatology,
Department of Pediatrics, Medical College of Wisconsin, Milwaukee,
Wisconsin, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on June 25, 2018.
Podcast available
Listen to a discussion of this manuscript with NCP Editor-in-Chief
Jeanette M. Hasse, PhD, RD, LD, FADA, CNSC, and authors
Vikram J. Christian, MBBS, and Scott Welak, MD. This and other
NCP podcasts are available at: https://onlinelibrary.wiley.com/
page/journal/19412452/homepage/podcasts
Corresponding Author:
Scott Welak, MD, Neonatology, Medical College of Wisconsin, 8701
Watertown Plank Road, Milwaukee, WI 53226, USA.
Email: swelak@mcw.edu
Christian et al
It is believed that NEC occurs from a combination of GI
immaturity, infection, inflammation, and enteral feeding.3
Of those risk factors, only enteral feeding is modifiable. In-
deed, 90% of NEC occurs after infants have received enteral
feedings. There are methods of enteral feeding that have
been shown to reduce the risk for NEC. One of the most
important strategies is an early introduction to enteral feeds
following a protocol of steady advancement of feedings.
Most clinicians will start infants on feeds within the first
few days of life and introduce feeds at 10–30 mL/kg/d.4
Subsequent advancements are in 10–30 mL/kg/d intervals
until the infant achieves his or her goal feeding volume.
When an infant demonstrates an intolerance to feeding
advancements, enteral feedings are often held at the current
volume or discontinued.
The most proven intervention to reduce the risk for
NEC is a diet of exclusively human expressed breast
milk (EBM).5-8 The protective factors of EBM are not
completely understood. EBM has a unique composition
compared with cow’s milk formula, including a different
casein/whey ratio that allows for easier digestion. EBM
also contains many potentially protective immunological
factors that have been thought to protect infants against
NEC. These factors include secretory immunoglobulin A,
lysozyme, and lactoferrin, which participate in host defense;
growth factors such as epidermal growth factors, insulin,
and insulin-like growth factor; anti-inflammatory cytokines
such as interleukin-10; and several antioxidant enzymes to
prevent free radical damage.9
Although many mothers are able and willing to provide
EBM for their infants, some are unable to due to several
factors including medical conditions, maternal medication
use, and inadequate free time. One study showed that only
30% of mothers were able to provide their own EBM to
their premature infant throughout the entire NICU course.7
Due to the benefits of human milk for the premature
infants, the use of donor human breast milk (DHBM)
has increased in recent years.10 Studies have shown that
the use of DHBM reduces the rates of NEC compared
with cow’s milk formula.5,11,12 One study demonstrated that
infants who received DHBM as their sole source of EN
had a 79% reduced risk for NEC, and that the number to
treat to prevent NEC with DHBM is 18.5.13 In addition
to preventing NEC, DHBM has been shown to have long-
term benefits, such as an improved lipoprotein profile in
adolescents.
Despite substantial research, few interventions have
shown promise in reducing the risk for NEC in preterm
infants. Many studies have shown improvement in the
prevention of NEC in the laboratory setting, but have not
translated into the clinical setting by significantly reducing
the incidence or severity of the disease. The most promising
development in the last decade has been use of probiotics.
Studies have shown that preterm neonates have a signifi-
477
cantly different intestinal bacteria composition than term
infants. These differences include a reduced diversity and
number of bacteria, and higher proportion of pathogenic
bacteria. Both laboratory and clinical studies have shown
that the supplementation of probiotics reduces the incidence
of NEC.14
The mechanism of protection for probiotics is still being
studied, but there is strong evidence that it affects the
signaling of Toll-Like Receptor-4 (TLR4).15 TLR4 is found
in many different tissues but is known to be abundant in
the small-intestinal epithelium during fetal development,
especially during the period premature infants are most at
risk for NEC. Laboratory studies have shown that TLR4
activation increases intestinal injury and inflammation in
animal models of NEC.16 The administration of probiotics
has been shown to reduce TLR4 activation and prevent
subsequent inflammation and tissue injury.17
Although there is significant evidence showing beneficial
effects of probiotics, their use for the prevention of NEC
in neonatal intensive care units (ICUs) is still limited.
Even though probiotics are available over the counter in
the United States as a dietary supplement, using them to
prevent or treat an illness requires additional regulation.
Currently, the U.S. Food and Drug Administration has
not approved any probiotic for use in premature infants
in the prevention of NEC. There are obvious concerns
that a provider may have regarding the administration of
a bacteria and the subsequent risk of that species causing
bacteremia. The preterm neonatal intestine is known to
have poor barrier function, and introducing bacteria into
the intestinal lumen could lead to a bacterial translocation.
There have been individual case reports of culture-positive
bacteremia in patients receiving lactobacillus. However,
in the larger meta-analysis, there have been no increases
in culture-proven sepsis.18 Additional studies are needed
to determine the proper method of probiotic adminis-
tration, including the strain, dose, and timing of the
probiotic.
PN During NEC
When an infant is diagnosed with NEC, all enteral feedings
are halted, typically for 7–10 days. Although there is little
evidence for the length of holding enteral feedings, most
clinicians use this time frame as a standard of practice.
The time until enteral feeds can be resumed depends on
the severity and extent of the disease. During this time, the
infant must receive all hydration and nutrition parenterally.
Due to the need for PN as nutrition support for at
least 7 days, central access is typically necessary. Infants
who develop extensive disease often have multiorgan com-
promise and may require mechanical ventilation. Infants
with significant disease often experience cardiovascular
compromise and hypotension, requiring frequent volume
478
resuscitation and inotropic support secondary to poor vas-
cular tone and excessive capillary leak. Edema is very com-
mon and will cause an increase in total fluid requirements.
It is not uncommon to see infants receiving fluids in excess
of 180–200 mL/kg/d with severe NEC.
Because of excessive fluid administration, multiorgan
failure, and subsequent renal insufficiency, infants with
NEC are at risk for electrolyte abnormalities. Excessive
capillary leak leads to edema of the intestine and soft tissues,
and hyponatremia can subsequently occur. Although total
body sodium may be relatively normal, their intravascular
sodium levels are low, and infants with severe NEC often
require significant sodium administration. Clinicians should
be aware of increasing requirements of sodium during this
time.
Acidosis can also be a significant sequela of NEC. In
severe cases of NEC, intestinal tissue becomes insufficiently
perfused and may become necrotic. Lactic acidosis can be
overwhelming, further worsening vascular tone and cardiac
output. It is common for infants to require acetate in their
PN to partially correct acidosis. The amount of acetate
delivered depends on the amount of sodium and potassium
in the PN. However, the main method for improving acidosis
is to stop tissue injury and necrosis.
If the infant has severe intestinal injury, the risk for
hyperkalemia increases. In most cases, hyperkalemia is not
common. However, in patients with rapidly progressing
NEC, the risk increases. There is a subset of NEC patients
with Thomsen-Freidenrich (T) cryptantigen activation,
which causes a rapid hemolysis and progressive intestinal
injury. These patients have a very quick decline and are more
likely to require surgery for their disease.19 The mortality
risk is also increased. Given that infants with severe NEC
are often hypotensive and have poor renal perfusion, the
excretion of potassium can be compromised. Therefore,
providers should be aware of the risk for hyperkalemia and
monitor electrolytes and urine output.
Patients with NEC are at risk for hypertriglyceridemia.
Poor hepatic perfusion leading to poor fat metabolism can
occur in severe NEC. Administration of lipid in the PN can
cause elevated triglyceride levels. Although some lipids are
needed to prevent essential fatty acid deficiency, care should
be used when determining lipid dosage.
Once patients have adequately recovered from NEC (at
least 7 days), clinicians will resume enteral feedings. PN
is continued until enteral feeds have reached the goal. In
severe cases of NEC, especially when intestinal resection
is required, some infants may take weeks to recover and
may not tolerate full enteral feedings for several months.
Prolonged PN use, defined as PN use for >21 days, may be
required. Providers should be aware that prolonged PN use
is a significant risk factor for cholestasis and PN-associated
liver disease. Some modifications can be made with long-
term PN use to reduce the risk for liver injury, such as not
Nutrition in Clinical Practice 33(4)
overfeeding, cycling lipid and PN infusions, or use of lipid
emulsion that includes fish oil.
Management of EN After Uncomplicated NEC
Once an infant recovers from NEC, reintroduction of EN is
considered. The importance of enteral feeding to intestinal
growth and maturation is well established. Enteral feeding
prevents gut atrophy, enhances mucosal adaptation, and
stimulates intestinal motility and growth.20 Mucosal villous
atrophy occurs within days in the absence of EN.4 This fact
highlights the importance of EN and the need to restart
feeds when clinically appropriate.
There is a lack of consensus on when EN can be reintro-
duced post-NEC. Frantz21 in 1975 reported recurrence of
NEC in several patients who had enteral feedings reestab-
lished <10 days after NEC. Recurrence of NEC is the
primary concern of resuming enteral feedings. Dimmitt22
recommends bowel rest for 7–14 days, depending on the
severity of the NEC episode. Most clinicians wait at least
1 week before resuming enteral feeds, reaching the goal of
complete enteral feeds 7–10 days later.20
Bohnhorst et al23 studied early vs late reintroduction
of enteral feeding after the diagnosis of NEC. Three con-
secutive days of lack of portal venous gas on abdominal
ultrasounds was considered a prerequisite to initiation of
enteral feeds. The median time of initiation of feeds in the
early introduction group was 4 days after onset of NEC
for medically treated NEC and 7 days for surgically treated
NEC. The median time of initiation of feeds in the late in-
troduction group was 10 days after onset of NEC. Late rein-
troduction was associated with delayed completion of EN.
Early introduction was associated with decreased duration
of central venous access, less catheter-related septicemia,
and less time to hospital discharge when compared with
the late introduction group. The authors concluded that
early feeding after NEC is feasible and associated with fewer
complications.23 The authors used a method of determining
readiness for feeds (abdominal ultrasound) that is not
universally followed. In 1993, Stringer et al24 described
12 neonates with recurrent NEC. The authors found no
consistent association between recurrent NEC and type or
timing of enteral feeds.24
Given the lack of further studies supporting early
reintroduction of feeds, the practice of waiting at least
7–14 days post-NEC may still be considered standard of
care. The decision to introduce feeds may be based on
clinical indicators and abdominal evaluations.20 Nonnu-
tritive sucking (pacifier usage), which has been shown
to increase mesenteric blood flow25 and trophic feeds
(<20 mL/kg/d), may be started initially to prepare the gut
for advancement of feeds. We recommend waiting at least 7
days before initiation of trophic feeds. Further advancement
Christian et al
of EN can be done when the following criteria have been
met:
• Adequate hemodynamics off vasoactive agents
• Reassuring abdominal examination
• Stable ventilation requirements
• Minimal to no electrolyte abnormalities
• Discontinuation of empiric antibiotics
• Reassuring abdominal radiograph
Breast milk is the best EN for premature infants and
is also the preferred source of EN as infants recover from
NEC.20,26 Breast milk contains growth hormones that pro-
mote adaptation of the gut and lactase that aids in digestion
of lactose.26 However, because of maternal indications or
parental preference, a cow’s milk formula may be used.
In 2017, Embleton and Zalewski20 systematically reviewed
the literature for data on reintroduction of feeds post-
NEC. No relevant trials focusing on the choice of enteral
feeding were identified at the time. Perks and Abad-Jorge26
recommended initiation of cow’s milk preterm formulas
with a higher caloric density because of the decreased
lactose and increased medium-chain triglycerides (MCT)
and long-chain triglycerides (LCT) in these products. We
recommend usage of these formulas for preterm infants.
Standard term formula may be used in term infants after
NEC.
Cow’s milk protein allergy has been implicated in the
pathogenesis of NEC. Gordon and colleagues27 in 2013
reclassified NEC based on risk factors that predisposed
an infant to development of NEC. In doing so he de-
fined “cow’s milk associated NEC” as a subset of NEC.27
Chuang et al28 studied serum and intestinal markers of
allergic-related inflammatory markers in patients with NEC
and compared them with specimens from infants with-
out inflammatory disease. They concluded that infants
with NEC demonstrated significant cow’s milk protein
sensitization, as evidenced by markers in the peripheral
blood. Minimal mucosal activation was also noted in some
cases with NEC.28 Faber et al29 reported a case of severe
cow’s milk allergic colitis resembling necrotizing colitis in
a premature infant. Similar cases of NEC attributed to
cow’s milk allergy have been reported.30,31 These studies
suggest that cow’s milk allergy should be considered as
a cause in cases of recurrent NEC or the occurrence of
NEC in the absence of typical risk factors. This also has
implications in selecting the enteral feed post-NEC. We
recommend usage of an extensively hydrolyzed formula
(EHF) in these specific circumstances to prevent recurrence
of NEC in an infant who is likely allergic to cow’s milk
protein.
Elemental formulas or amino acid–based formulas, in
contrast with EHFs, consist of predominantly free amino
acids as the source of nitrogen. They are therefore con-
479
sidered to be hypoallergenic and are expected to be well
tolerated in an infant with cow’s milk protein allergy. No
literature was found to support usage of elemental for-
mulas in a post-NEC infant. Moreover, because elemental
formulas are likely more expensive than EHFs, we recom-
mend usage of EHF in a post-NEC infant when cow’s
milk protein allergy is presumed contributory or causative.
Despite lack of supporting data, there are specific situations
in which elemental formulas may be preferred over EHF.
These situations include the development/persistence of
symptoms resembling cow’s milk protein allergy and/or
the development of recurrent NEC in an infant who is
fed EHF.
Reasonable advancement of feeds has been shown to
decrease the incidence of NEC.26 While reintroducing EN,
following this practice may prevent recurrence of NEC.
Perks and Abad-Jorge26 recommended the bolus method of
feeding, advancing at a rate of 10–35 mL/kg/d in cases of
medical NEC. We recommend starting bolus feeds initially,
to provide 20 mL/kg/d, gradually advancing at the rate
of 10–20 mL/kg/d, as tolerated. Feeds may be provided
by mouth or by nasogastric/gastrostomy tube. If bolus
feeds are not tolerated, feeds may be run continuously
via nasogastric/gastrostomy tube. As enteral feeds are
advanced, PN should be gradually weaned, until eventually
discontinued.
Management of EN After Bowel Resection
Reintroduction of feeds after surgical NEC is dependent
on length of remaining bowel and patient’s clinical status.
Indications to start and advance feeds are the same as those
of uncomplicated NEC.
Human breast milk remains the feed of choice in this
population as well. Lapillonne et al32 conducted a na-
tionwide cross-sectional questionnaire-based study among
French neonatal ICUs to describe the frequency and reasons
for using EHF. The authors found that among 1969 hos-
pitalized infants, 12% were receiving EHF. Eleven percent
of these EHF prescriptions were due to previous NEC.
According to the survey, the main reasons for using an
EHF when feeding post-NEC neonates were the absence of
human milk (75%) and the need for surgical management of
NEC (17%). Because these formulas do not typically con-
tain lactose and may contain significant amounts of MCTs,
the authors postulate that they may be better absorbed by
an infant recovering from NEC.32 This advantage afforded
to infants recovering from NEC may be unnecessary. This
was determined by Schaart et al33 in 2007. In this study, the
protein content of small-intestinal enterostomy effluent was
quantified in neonates who underwent bowel resection for
different reasons (NEC, atresia, perforation). The authors
concluded that the protein absorptive capacity of neonates
after bowel resection is intact (70%–90% of total enteral
480 Nutrition in Clinical Practice 33(4)

Table 1. Summary of Recommendations.

Nutrition Before NEC

1. For stable ELBW infants, enteral feeding should be started within the first 48 hours of life.
2. Every effort should be made to provide ELBW infants with their own mother’s EBM.
3. If the mother does not provide EBM, donor human breast milk should be used.
4. Although there are promising studies showing a reduced rate of NEC by using probiotics, clinical trials are still
needed to determine which strain, dose, and timing of probiotics is best used for ELBW infants.

PN During NEC
1. Strongly consider central venous access at the diagnosis of NEC.
2. Provide adequate nutrition support through PN.
3. Anticipate infants having an increased need for fluid requirements secondary to capillary leak.
4. Closely monitor serum sodium and appropriately correct hyponatremia.
5. Anticipate metabolic acidosis and give adequate amounts of acetate.

EN After Uncomplicated NEC

1. Trophic feeds (<20 mL/kg/d) may be started via nasogastric/gastrostomy tube after at least 7 days of bowel rest.
2. EN may be advanced based on clinical indicators:
r Adequate hemodynamics off vasoactive agents
r Reassuring abdominal examination
r Stable ventilation requirements
r Minimal to no electrolyte abnormalities
r Discontinuation of empiric antibiotics
r Reassuring abdominal radiograph
3. Human breast milk (mother or donor-sourced) is the preferred feed.
4. If breast milk is unavailable, preterm formula with higher MCT and LCT should be fed to preterm infants.
5. Standard term formula may be fed to term infants.
6. Extensively hydrolyzed formulas may be considered if cow’s milk protein allergy is suspected (in patients with
recurrent NEC or in the absence of typical risk factors).
7. Bolus feeds may be started by mouth or by nasogastric/gastrostomy tube to provide 20 mL/kg/d.
8. Volume of feeds should be advanced gradually at a rate of 10–20 mL/kg/d.

EN After Bowel Resection

1. Indications to start trophic feeds and advance EN are the same as those of uncomplicated NEC.
2. Human breast milk (mother or donor-sourced) is the preferred feed.
3. Semi-elemental or amino acid–based formula may be considered if the patient is intolerant to conventional
preterm/term formulas.
4. Continuous feeds may be started by nasogastric/gastrostomy tube to provide 20 mL/kg/d.
5. Volume of feeds should be advanced gradually at a rate of 10–20 mL/kg/d.

EBM, human expressed breast milk; ELBW, extremely low birth weight; EN, enteral nutrition; LCT, long-chain triglycerides; MCT,
medium-chain triglycerides; NEC, necrotizing enterocolitis; PN, parenteral nutrition.

protein intake was absorbed).33 Perks and Abad-Jorge26
recommended the use of breast milk or semi-elemental or
amino acid–based formula initially. We recommend human
breast milk, when available, or semi-elemental or amino
acid–based formulas should the patient be intolerant to
conventional preterm/term formulas.
Reintroduction of feeds should be a gradual process.
Trophic feeds (<20 mL/kg/d) may be started when the
patient is deemed ready to start feeds. Perks and Abad-
Jorge26 recommended continuous feeds delivered on a
pump over 24 hours, followed by gradually compressing the
continuous feeds as tolerated to a bolus-feeding schedule.
The authors also recommended advancing feeds at a rate
of 10–35 mL/kg/d.26 Continuous feeding is thought to be
nonphysiological, because it does not allow for the normal
pattern of gastrointestinal hormone release observed with
bolus feedings. Despite this, we recommend starting with
continuous feeds via nasogastric/gastrostomy tube to
provide 20 mL/kg/d, gradually advancing at the rate of
10–20 mL/kg/d. As the infant approaches the goal volume
of feeds, PN may be gradually weaned and eventually
discontinued.
Accretion of micronutrient stores occurs in the third
trimester.34 Preterm infants are therefore at risk for
Christian et al
micronutrient deficiency and require supplementation.
Preterm infants recovering from NEC are thought to have
increased needs for minerals, vitamins, electrolytes, zinc, and
iron.20 There is a higher risk for deficiency in infants who
have had a bowel resection. This deficiency would depend on
the length and portion of bowel resected, presence and loca-
tion of ostomy, and ostomy output. Post-NEC infants may
also require additional supplementation of levo-carnitine if
there is prolonged use of PN and/or significant ostomy/stool
losses.26
Refeeding of Ostomy Output
In the surgical infant with NEC where bowel resection
is necessary, an enterostomy and mucous fistula may be
created.35 In an infant with an enterostomy, enteral feeds are
digested by the proximal intestine, and the effluent empties
into an ostomy bag. This fluid is theoretically similar to
that which would enter the large intestine, where nutrients
and fluid would be absorbed by the colon. Therefore,
patients with an enterostomy are not able to fully benefit
from enteral feedings. However, some patients may benefit
from refeeding the ostomy effluent. Refeeding refers to
the ostomy output being collected and then refed through
the distal mucous fistula. This practice has been shown
to decrease fluid loss and to improve electrolyte balance,
bowel adaptation, and growth. The practice of refeeding
of proximal stoma effluent through a distal mucous fistula
uses the absorptive surface of the distal bowel for nutrient
absorption, stimulates mucosal growth and intestinal adap-
tation, and prevents atrophy of the distal bowel.36-38 Infants
who undergo mucous fistula refeeding have a shorter time
from initiation of enteral feeds to attainment of goal feeds
and PN discontinuation in the interoperative period.39
The practice of refeeding the output to the mucous fistula
is thought to improve bowel adaptation and decrease time
to full enteral feeds, but it does come with some risk.40
The safety of refeeding has posed concern for risk for
perforation, GI bleed, sepsis, and mortality.35 Refeeding
is thought to be safe and beneficial when administered
in an ICU using a protocol that specifies the equipment
and techniques used for mucous fistula cannulation and
refeeding parameters.39 Gause et al39 described a refeeding
protocol using a 6.0 Fr Foley catheter placed by a surgeon
while monitoring for signs of vascular compromise. The
authors’ protocol consisted of refeeding the succus entericus
collected from the proximal enterostomy and instilling this
into the mucous fistula via an electronic feeding pump at a
rate of 1:1.39 The authors recommended replacement of the
catheter in case of dislodgement in a patient who had been
tolerating refeeding. Forming an institution-specific proto-
col for mucous fistula refeeding is a necessary starting point
before initiation of refeeding in a neonatal unit. There are
no studies comparing refeeding protocols. More research is
481
needed to identify other risks and benefits of this practice
and before a refeeding protocol can be recommended to best
care for these patients.
Conclusion
NEC is a complex disease with significant mortality and
morbidity. Those who survive the disease are still at risk
for complications. Resuming enteral feeds can be difficult,
with many short- and long-term complications. Enteral
feeds should be cautiously resumed once the infant has
recovered, to prevent long-term injury from prolonged PN
use. The logistics of how and what to feed are somewhat
unclear, and often depends on the severity of the disease.
For patients with an enterostomy, refeeding the distal
intestine may improve bowel growth and prevent long-term
complications. Recommendations have been summarized
in Table 1. Future studies should be aimed at clarifying the
best modalities in enteral feeding after NEC, and hopefully
provide improved methods of caring for infants who survive
this terrible disease.
Statement of Authorship
V. J. Christian, E. Polzin, and S. Welak contributed to the
conception and design of the manuscript. V. J. Christian, E.
Polzin, and S. Welak contributed to acquisition, analysis, or
interpretation of the data. V. J. Christian, E. Polzin, and S.
Welak drafted the manuscript. V. J. Christian, E. Polzin, and
S. Welak critically revised the manuscript. V. J. Christian,
E. Polzin, and S. Welak agree to be fully accountable for
ensuring the integrity and accuracy of the work, and all
authors read and approved the final manuscript.
References
1. Abdullah F, Zhang Y, Camp M, et al. Necrotizing enterocolitis
in 20,822 infants: analysis of medical and surgical treatments. Clin
Pediatr. 2010;49(2):166-171.
2. Schnabl KL, Van Aerde JE, Thomson AB, Clandinin MT. Necrotizing
enterocolitis: a multifactorial disease with no cure. World J Gastroen-
terol. 2008;14(14):2142-2161.
3. Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med.
2011;364(3):255-264.
4. Kliegman RM, Walker WA, Yolken RH. Necrotizing enterocolitis:
research agenda for a disease of unknown etiology and pathogenesis.
Pediatr Res. 1993;34(6):701-708.
5. Sullivan S, Schanler RJ, Kim JH, et al. An exclusively human milk-
based diet is associated with a lower rate of necrotizing enterocolitis
than a diet of human milk and bovine milk-based products. J Pediatr.
2010;156(4):562-567.e561.
6. Quigley M, McGuire W. Formula versus donor breast milk for feeding
preterm or low birth weight infants. Cochrane Database Syst Rev.
2014(4):CD002971.
7. Schanler RJ, Lau C, Hurst NM, Smith EO. Randomized trial of
donor human milk versus preterm formula as substitutes for mothers’
own milk in the feeding of extremely premature infants. Pediatrics.
2005;116(2):400-406.
482
8. Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis.
Lancet. 1990;336(8730):1519-1523.
9. Hanson L, Silfverdal SA, Stromback L, et al. The immunological
role of breast feeding. Pediatr Allergy Immunol. 2001;12(suppl 14):
15-19.
10. McGuire W, Anthony MY. Donor human milk versus formula for pre-
venting necrotising enterocolitis in preterm infants: systematic review.
Arch Dis Child Fetal Neonatal Ed. 2003;88(1):F11-F14.
11. Cristofalo EA, Schanler RJ, Blanco CL, et al. Randomized
trial of exclusive human milk versus preterm formula diets in
extremely premature infants. J Pediatr. 2013;163(6):1592-1595.
e1591.
12. Kantorowska A, Wei JC, Cohen RS, Lawrence RA, Gould JB, Lee HC.
Impact of donor milk availability on breast milk use and necrotizing
enterocolitis rates. Pediatrics. 2016;137(3):e20153123.
13. Boyd CA, Quigley MA, Brocklehurst P. Donor breast milk versus in-
fant formula for preterm infants: systematic review and meta-analysis.
Arch Dis Child Fetal Neonatal Ed. 2007;92(3):F169-F175.
14. AlFaleh K, Anabrees J. Probiotics for prevention of necrotiz-
ing enterocolitis in preterm infants. Cochrane Database Syst Rev.
2014(4):CD005496.
15. Leaphart CL, Cavallo J, Gribar SC, et al. A critical role for TLR4 in
the pathogenesis of necrotizing enterocolitis by modulating intestinal
injury and repair. J Immunol. 2007;179(7):4808-4820.
16. Gribar SC, Anand RJ, Sodhi CP, Hackam DJ. The role of epithelial
Toll-like receptor signaling in the pathogenesis of intestinal inflamma-
tion. J Leukoc Biol. 2008;83(3):493-498.
17. Patel RM, Underwood MA. Probiotics and necrotizing enterocolitis.
Semin Pediatr Surg. 2018;27(1):39-46.
18. Sawh SC, Deshpande S, Jansen S, Reynaert CJ, Jones PM. Prevention
of necrotizing enterocolitis with probiotics: a systematic review and
meta-analysis. PeerJ. 2016;4:e2429.
19. Hall N, Ong EG, Ade-Ajayi N, et al. T cryptantigen activation is
associated with advanced necrotizing enterocolitis. J Pediatr Surg.
2002;37(5):791-793.
20. Embleton ND, Zalewski SP. How to feed a baby recovering from
necrotising enterocolitis when maternal milk is not available. Arch Dis
Child Fetal Neonatal Ed. 2017;102(6):F543-F546.
21. Frantz ID 3rd, L’Heureux P, Engel RR, Hunt CE. Necrotizing
enterocolitis. J Pediatr. 1975;86(2):259-263.
22. Dimmitt RA, Moss RL. Clinical management of necrotizing entero-
colitis. NeoReviews. 2001;2(5):110-117.
23. Bohnhorst B, Muller S, Dordelmann M, Peter CS, Petersen C, Poets
CF. Early feeding after necrotizing enterocolitis in preterm infants.
J Pediatr. 2003;143(4):484-487.
24. Stringer MD, Brereton RJ, Drake DP, Kiely EM, Capps SN, Spitz
L. Recurrent necrotizing enterocolitis. J Pediatr Surg. 1993;28(8):
979-981.
Nutrition in Clinical Practice 33(4)
25. Huang CF, Tsai MC, Chu CH, Lee MY, Shih YL, Cheng SN. The
influence of pacifier sucking on mesenteric blood flow in infants. Clin
Pediatr. 2003;42(6):543-546.
26. Perks P, Abad-Jorge A. Nutritional management of the infant with
necrotizing enterocolitis. Pract Gastroenterol. 2008;32(2):46-60.
27. Gordon P, Christensen R, Weitkamp JH, Maheshwari A. Mapping the
new world of necrotizing enterocolitis (NEC): review and opinion. EJ
Neonatol Res. 2012;2(4):145-172.
28. Chuang SL, Hayes PJ, Ogundipe E, Haddad M, MacDonald TT, Fell
JM. Cow’s milk protein-specific T-helper type I/II cytokine responses
in infants with necrotizing enterocolitis. Pediatr Allergy Immunol.
2009;20(1):45-52.
29. Faber MR, Rieu P, Semmekrot BA, Van Krieken JH, Tolboom JJ,
Draaisma JM. Allergic colitis presenting within the first hours of
premature life. Acta Paediatr. 2005;94(10):1514-1515.
30. de Peyer E, Walker-Smith J. Cow’s milk intolerance presenting as
necrotizing enterocolitis. Helv Paediatr Acta. 1978;32(6):509-515.
31. Walther FJ, Kootstra G. Necrotizing enterocolitis as a result of cow’s
milk allergy? Z Kinderchir. 1983;38(2):110-111.
32. Lapillonne A, Matar M, Adleff A, Chbihi M, Kermorvant-Duchemin
E, Campeotto F. Use of extensively hydrolysed formula for refeed-
ing neonates postnecrotising enterocolitis: a nationwide survey-based,
cross-sectional study. BMJ Open. 2016;6(7):e008613.
33. Schaart MW, de Bruijn AC, Tibboel D, Renes IB, van Goudoever JB.
Dietary protein absorption of the small intestine in human neonates.
JPEN J Parenter Enteral Nutr. 2007;31(6):482-486.
34. Shaikhkhalil AK, Curtiss J, Puthoff TD, Valentine CJ. Enteral zinc
supplementation and growth in extremely-low-birth-weight infants
with chronic lung disease. J Pediatr Gastroenterol Nutr. 2014;58(2):183-
187.
35. Haddock CA, Stanger JD, Albersheim SG, Casey LM, Butterworth
SA. Mucous fistula refeeding in neonates with enterostomies. J Pediatr
Surg. 2015;50(5):779-782.
36. Koike Y, Uchida K, Nagano Y, et al. Enteral refeeding is useful for
promoting growth in neonates with enterostomy before stoma closure.
J. Pediatr Surg. 2016;51(3):390-394.
37. Richardson L, Banerjee S, Rabe H. What is the evidence on the practice
of mucous fistula refeeding in neonates with short bowel syndrome?
J Pediatr Gastroenterol Nutr. 2006;43(2):267-270.
38. Al-Harbi K, Walton JM, Gardner V, Chessell L, Fitzgerald PG.
Mucous fistula refeeding in neonates with short bowel syndrome.
J Pediatr Surg. 1999;34(7):1100-1103.
39. Gause CD, Hayashi M, Haney C, et al. Mucous fistula refeeding
decreases parenteral nutrition exposure in postsurgical premature
neonates. J Pediatr Surg. 2016;51(11):1759-1765.
40. Lau EC, Fung AC, Wong KK, Tam PK. Beneficial effects of mucous
fistula refeeding in necrotizing enterocolitis neonates with enteros-
tomies. J Pediatr Surg. 2016;51(12):1914-1916.
Review

Nutrition in Clinical Practice

Volume 33 Number 4
Enteral Nutrition in the Management of Crohn’s Disease: August 2018 483–492

C 2018 American Society for

Reviewing Mechanisms of Actions and Highlighting Potential Parenteral and Enteral Nutrition
DOI: 10.1002/ncp.10004
Venues for Enhancing the Efficacy wileyonlinelibrary.com

Moftah H. Alhagamhmad, MBBCH, DCH UK, PhD

Abstract
Crohn’s disease (CD) is a chronic condition that affects the gut and has adverse effects on growth and development. There is a
global increase in the incidence and prevalence rates, and several factors are believed to contribute to this rise, including dietary
habits. In contrast, the use of enteral nutrition (EN) as an exclusive source of nutrition is increasingly becoming the preferred
induction treatment of pediatric CD patients in part to address the nutrition complications. However, EN therapy is considered
less effective in adults with CD. A better understanding of the molecular mechanisms of enteral therapy will help improve the
clinical management of CD. It is increasingly becoming evident that the therapeutic utility of EN is in part due to the reversal of
the microbial changes and the direct immunomodulatory effects. Moreover, there is a potential tendency for enhancing the efficacy
of EN therapy by improving the palatability of the given formulas and, more important, by magnifying the anti-inflammatory
properties. Recent observations have shown that the immunomodulatory effects of EN are mediated at least in part by blocking
nuclear factor-κB. Furthermore, it is likely that several ingredients of EN contribute to this activity, in particular glutamine and
arginine amino acids. In addition, manipulating the composition of EN therapy by altering concentrations of the key ingredients
is found to have the potential for more efficient therapy. In this review, the underlying mechanisms of EN actions will be discussed
further with a focus on the potential methods for enhancing the efficacy. (Nutr Clin Pract. 2018;33:483–492)

Keywords
Crohn’s disease; enteral nutrition; nutrition therapy; inflammation; immunomodulation

Crohn’s disease (CD) is an entity of inflammatory bowel
diseases (IBD) and can affect any part of the gastrointestinal
tract.1 It is a chronic, lifelong disorder that can potentially
compromise the quality of life and is characterized by
alternating periods of remission and relapse, imposing a
considerable negative impact on all aspects of life.2 In
addition to the chronic and current incurable nature of CD,
there is an increase in the global incidence and prevalence
rates.3 Recent reports show that CD is no longer confined,
as previously known, to the Western countries; rather, it
is increasingly becoming more prevalent in other parts of
the world with a concomitant rise in CD diagnoses among
children.4 Changes in dietary habits along with the other
environmental triggers are believed to account for the recent
changes in the patterns of CD epidemiology.4
Further challenges involved in the management of CD
include the accompanying nutrition complications, includ-
ing impairments in growth and development.5 It has been
reported that around 50% of CD patients may be under-
weight and stunted at the time of diagnosis.6 Furthermore,
the compromised nutrition status in these children may fur-
ther deteriorate with the use of immunosuppressive drugs.7
Thus, it appears that in addition to the induction of remis-
sion, addressing the nutrition complications is a vital goal
in CD management.8 It is noteworthy that enteral nutrition
(EN) with a polymeric formula (PF) and/or elemental
formula given as a sole nutrition source is increasingly be-
coming the first choice as an induction treatment of young
patients with active CD.9 In addition to its advantage in
terms of correcting the nutrition imbalances, enteral diet in
the form of exclusive enteral nutrition (EEN) is considered
a safe therapy with a therapeutic efficacy incomparable with
that of corticosteroids in inducing remissions and, more
From the Faculty of Medicine, University of Benghazi (Al-Arab
Medical University), Benghazi, Libya.
Financial disclosure: None declared.
Conflict of interest: None declared.
Received for publication April 21, 2017; accepted for publication
October 10, 2017.
This article originally appeared online on January 11, 2018.
Corresponding Author:
Moftah Alhagamhmad, MBBCH, DCH UK, PhD, Benghazi
University (Al-Arab Medical University), Medical School, P.O. Box
18251, Hawari Road, Benghazi, Libya.
Email: sufrani82@yahoo.com
484
Figure 1. The changes accompanying Crohn’s disease, including gut microbial changes and loss of integrity of intestinal epithelial
barrier with subsequent overactivation of mucosal immune responses.
important, superior efficacy in enhancing mucosal healing.6
This review discusses the underlying mechanisms of action
of enteral therapy and highlights the potential means for
enhancing the efficacy.
Pathogenesis of Crohn’s Disease
For a better understanding of the mode of action of EN and
its role in suppressing intestinal inflammation, the basic con-
cepts of CD aetiopathogenesis are briefly described in this
section (Figure 1). The pathogenesis is considered multifac-
torial with genetic, microbial, immunologic, environmental,
and other adjuvant factors playing a role.10 Altered immune
regulation is one of the suggested underlying mechanisms
of CD, leading to the ongoing inflammatory processes with
subsequent tissue destruction.11 An exaggerated immune
response is characterized by a rise in the proinflammatory
cytokines levels along with a drop in the anti-inflammatory
mediators.12 Gene expression of the proinflammatory me-
diators is modulated through activation of the complex
intracellular pathways, most notably nuclear factor (NF)-
κB13,14 and mitogen protein kinase cascades.15
Furthermore, it has been established that changes in the
gut microbiota also play a crucial role in the pathogenesis
and development of CD.16 Dysbiosis, or bacterial imbal-
ance, is a leading factor that disrupts the host–immune
system cross-talk with subsequent perturbation of mucosal
homeostasis.17 Moreover, an increase in the intestinal per-
meability is an additional contributory factor involved in the
disruption of the host–microbiome harmony.18 Certainly,
Nutrition in Clinical Practice 33(4)
a loss of function of the intestinal epithelial barrier al-
lows trafficking various agents, including luminal bacteria,
initiating uncontrolled compensatory immune responses
with concomitant overexpressions and the release of the
immunoregulatory cytokines.19
Mechanisms of Action
Although there is a considerable gap in our understanding
about the exact mechanisms of action of EN, there is
substantial evidence indicating that EN exerts therapeutic
benefits by interfering with the integral components of IBD
pathology.20 Immunomodulation, profound changes in the
gut microbiota, and enhanced mucosal healing of the leaky
gut appear to be the key factors beyond the attenuation of
inflammation and induction of clinical remissions following
treatment with EN21 (Figure 2).
Anti-Inflammatory Property
The initial observation on the anti-inflammatory effects
of EN was made on a group of CD patients with resis-
tant disease who had undergone remission while receiv-
ing elemental formula as preoperative nutrition support.22
Thereafter, several trials provided stronger evidence of the
immunomodulatory effects of EN.23 Certainly, in numerous
clinical trials involving CD patients, the EEN regimen
exerted a remarkable immunosuppressive effect, resulting
in inhibition of cytokines’ production with a concomitant
fall in the mucosal levels.24-26 In a trial involving 21 patients
with active CD, treatment with 8 weeks of EEN using PF
Alhagamhmad
Figure 2. Mode of action of enteral therapy in attenuating intestinal inflammation by reversing the pathological changes
characterizing Crohn’s disease. NF-κB, nuclear factor-kappa B; Iκκ, Inhibitor of kappa-B kinase; IκB, inhibitor of kappa-B; P,
phosphorylation.
resulted in a significant drop in the cytokine expression in
the colonic and ileal biopsies.27 Similar observations were
made in another trial that used elemental formula as an
induction therapy in 28 patients with active CD.25 The
mucosal concentrations of the key inflammatory mediators
in the collected biopsies of CD patients were not compa-
rable with those in the biopsies of matched controls.25 It is
notable that in another trial on 12 CD patients undergoing
a 6-week EEN protocol, the anti-inflammatory effects of
the given EN therapy proceeded its nutrition values.28 The
disease activity score and raised inflammatory markers were
corrected in the initial days of treatment, whereas changes
in the nutrition parameters, including weight scores and
subcutaneous fat thickness, were only visible by the second
and third weeks of treatment initiation.28
EN can alter the exaggerated immune responses and
thereby suppress inflammation. However, the exact mecha-
nisms of its interaction with the mucosal immune system to
attenuate intestinal inflammation remain unclear. De Jong
et al29 investigated the effect on 3 different immortalized
intestinal epithelial cells cultured as an in vitro model
of intestinal inflammation and showed that the therapy
exerted a direct suppressing effect. The treatment with PF
blocked tumor necrosis factor (TNF)-α induced interleukin
(IL)-8 responses in the stimulated cells. The block in IL-8
production was consistent with a delay in the degradation of
the inhibitory κ-B subunit, the major regulatory molecule of
NF-κB transcription factor. Subsequently, using the same
model, Nahidi et al30 further examined the interactions of
PF with the NF-κB pathway. PF modulated the expressions
485
of several genes linked to the NF-κB pathway, indicating
that EN could exert its effect by altering the activity of the
intracellular signaling pathways.30 However, this study was
unable to determine the determinants of PF activity and
the exact nature of the interaction with the NF-κB pathway.
Thereafter, additional in vitro experiments were conducted
to further explore the previous observations.31 Glutamine,
arginine, vitamin D3 , and α linolenic acid, present in PF,
were identified as potential immunoactive nutrients, and
their effects were studied on the stimulated epithelial cells.31
The authors noted that at concentrations equivalent to those
used in PF, glutamine, arginine, and vitamin D3 (at final con-
centrations of 12.7 mM, 1.8 mM, and 3.8 nM, respectively,
in the culture media) can suppress inflammation, indicating
that these components are likely at least in part to contribute
to the immunomodulatory effects observed following EEN
treatment.31 However, the observed effect of PF with its
whole components was superior to that of combined glu-
tamine, arginine, and vitamin D3 in blocking cytokine pro-
duction from the stimulated cells.31 This indicates that there
are other currently unidentified ingredients of PF that might
be contributing to its activity.32 The authors also showed
that the glutamine and arginine ingredients of PF can block
phosphorylation, a prerequisite process for the activation
of the key active components of NF-κB and P38 mitogen
protein kinase cascades,31 thereby blocking the major source
of proinflammatory cytokine production. Glutamine and
arginine can interfere with phosphorylation directly through
potential competitive adenosine 5’-triphosphate binding
in the signaling components and indirectly by enhancing
486 Nutrition in Clinical Practice 33(4)

Table 1. Studies on the Role of EN in Suppressing Inflammation.

Date, first author, reference, journal

name Study description and main findings

1973 Voitk et al22 Observational study: First report on the immunosuppressive effects of EN made on a
Archives of Surgery group of IBD patients waiting for surgery while receiving elemental formula as
preoperative nutrition support.
1994 Beattie et al26 Descriptive study: Effect of PF on 7 children with CD was evaluated. All children
Alimentary Pharmacology & experienced reduction in the disease activity and 6 of the total patients showed a
Therapeutics reduction in mucosal cytokines levels.
2000 Fell et al27 Prospective cohort study: Efficacy of EN in inducing remission was investigated on
Alimentary Pharmacology & 21 pediatric CD patients. Remission rate was 79% following 8-week regimen of PF
Therapeutics together with a significant reduction in mRNA expression of colonic and ileal
proinflammatory cytokines.
2007 de Jong et al29 In vitro model of IBD utilized intestinal epithelial cells cultured in lab: Treatment
Digestive Diseases and Sciences with PF suppressed TNF-α induced IL-8 productions from the stimulated cells.
2015 Nahidi et al30 In vitro model of IBD as above: Altered immune responses following the treatment
Genes & Nutrition with PF were shown to be mediated through modulation of the NF-κB pathway.
2016 Alhagamhmad et al31 In vitro model of IBD as above: Glutamine, arginine, and vitamin D3 part of PF
Journal of Parenteral and Enteral prompted a reduction in IL-8 production from the activated cells (potential key
Nutrition components of PF). Glutamine and arginine exerted their effects through
interfering with phosphorylation of NF-κB and P38 MAPK pathways. Amino
acids enriched PF prompted superior immunomodulating effects over standard PF.

CD, Crohn’s disease; EN, enteral nutrition; IBD, inflammatory bowel disease; IL, interleukin; MAPK, mitogen protein kinase cascades; NF,
nuclear factor; PF, polymeric formula; TNF, tumor necrosis factor.

production of intermediary active metabolites such as nitric
oxide that is involved in nitrosylation, blocking the tran-
scription molecules.31 Collectively, the therapeutic utility
of EN appears to be mediated at least in part by the
immunomodulatory effects exerted through the deactiva-
tion of the major intracellular signaling pathways NF-κB
and P38 mitogen protein kinase cascades. Furthermore, it
is likely that several individual components contribute to
this property, most notably glutamine and arginine amino
acids. Key studies on the role of EN in attenuating immune
responses are summarized in Table 1.
Enhancing Mucosal Healing
Disruption of the integrity of the intestinal epithelial bar-
rier, a prominent feature of CD, accounts for trafficking
of the luminal bacteria and other antigenic agents, re-
sulting in disruption of the gut microbial homoeostasis
and in turn stimulating the mucosal immune system.33
Thus, promoting mucosal healing has become a key goal
in IBD management.34 It is noteworthy that numerous
clinical studies have shown that enteral diet, administered
as EEN, is capable of enhancing mucosal healing in treated
CD patients.35,36 Furthermore, EN is superior to corticos-
teroids in terms of induction of higher mucosal remission
rates.37 In a trial involving 47 young CD patients, the
investigators noted that around two-thirds of the children
receiving EEN achieved mucosal healing compared with
<50% of those who received corticosteroids.38 The authors
also observed that 7 patients of the EEN group achieved
complete mucosal healing vs none in the corticosteroids
group.38
However, the exact mechanism by which EN restores
the epithelial barrier function has been poorly investigated
and remains unclear. It is well accepted that alterations
in the intestinal epithelial barrier function are caused by
an overproduction of the proinflammatory mediators such
as TNF-α.39,40 TNF-α enhances intestinal epithelial per-
meability by modulating the protein expression of myosin
II regulatory light-chain kinase (MLCK).41 Nahidi et al42
investigated the effects of PF on intestinal tight junction
barrier function in the presence of TNF-α in an in vitro
model using Caco-2 epithelial monolayers. Monolayer per-
meability was studied, and MLCK gene expression was eval-
uated. In these experiments, TNF-α exposure increased the
monolayer permeability, upregulated MLCK expression,
and reduced tight junction integrity.42 It is interesting to
note that these changes were reversed completely with the
use of PF in association with the inhibition of MLCK.42
These observations were further supported by studies on
a mouse model of colitis, characterized by increased gut
inflammatory markers and reduced gut barrier function.43
EEN treatment, given as PF, maintained the gut barrier
integrity and function, reversed the inflammatory changes,
and ameliorated colitis. The authors also found that these
changes were correlated with the down-regulation of the
gene expression of proinflammatory cytokines involved in
the disruption of tight junction proteins.43
Alhagamhmad 487

Table 2. Studies on the Role of EN in Mucosal Healing.

Date, first author, reference, journal

name Study description and main findings

2006 Borrelli et al37 Randomized controlled trial: A group of 37 children with active CD who were
Clinical Gastroenterology and randomized to receive either EN in THE form of PF or corticosteroids and healing
Hepatology of gut mucosa was assessed. PF showed superior efficacy over corticosteroids in
healing of gut inflammatory lesions.
2012 Nahidi et al42 In vitro model using Caco-2 epithelial monolayers: Effect of EN upon TNF-α
Journal of Gastroenterology induced disrupted intestinal tight junction function was assessed. PF enhanced the
integrity and reduced permeability in conjunction with reduced expression of
myosin II regulatory light-chain kinase protein expression.
2013 Nahidi et al43 Mouse model of colitis: Gut barrier function was assessed following treatments with
Biomed Research International EEN and corticosteroids. PF was superior to corticosteroids, given as
hydrocortisone, in maintaining epithelial barrier integrity and reversing
accompanied inflammatory changes.
2014 Grover et al35 Prospective open label study: Efficacy of EN in inducing endoscopic remission was
Journal of Gastroenterology investigated. Treatment with EEN was effective in inducing mucosal and
transmural remissions and reduced endoscopic relapses at 1 year.

CD, Crohn’s disease; EEN, exclusive enteral nutrition; EN, enteral nutrition; PF, polymeric formula; TNF, tumor necrosis factor.

The TNF-α induced increase in the MLCK gene activity
is mediated by activation of NF-κB and vice versa.40,44
Although it was not directly investigated in the previous
in vitro study,42 the downregulation of TNF-α-induced
MLCK protein expression by PF treatment is a potential
indicator of the inhibition of NF-κB activity by EEN
treatment. Overall, it appears that the anti-inflammatory
effects and enhanced mucosal healings observed after EN
treatment are mediated by attenuating production of the
proinflammatory mediators via blocking of the major in-
tracellular signaling pathways. Major studies that have
explored the role of EN in mucosal healing are summarized
in Table 2.
Gut Microbiota Modification
Additional factors that contribute to the therapeutic ben-
efits of EN in CD management are mediated through
modifications in the gut microbiota.45 Using the tempera-
ture gradient gel electrophoresis technique, the microflora
in the stool samples of CD patients who received an
enteral diet for 8 weeks and those of comparatively healthy
controls were investigated.46 Major modifications in the
microflora of the collected samples were observed following
the EEN treatment.46 Significant alterations were also noted
in the fecal bacterial composition following EN treatment
in another study that used an alternative technique (de-
naturing gradient gel electrophoresis).47 Furthermore, in
a recent study that used whole-genome high-throughput
sequencing techniques, the microbial diversity in EEN-
treated CD patients was significantly lower than that in the
matched controls.48 The authors were also able to identify
several bacterial species within the Firimictus family that
were correlated with disease activity during and after EEN
therapy.48
How precisely enteral therapy modulates changes in the
gut microbiota, however, remains unknown. It has been
proposed that enteral diets have prebiotic properties that
enable the modification of microbiota.46 One study has
also proposed limiting nutrient access during the EEN
regimen to starve the bacteria residing in the colon.49
According to an interesting finding reported by a recent
observational study, EN can alter the metabolism of colonic
gut microbiota.50 Significant reductions in the fecal concen-
trations of microbial metabolites (short-chain fatty acids, 1-
propanol, 1- butanol, and esters of short-chain fatty acids)
were noted following 2 weeks of EN treatment in CD
patients. These changes were consistent with improvements
in the patients’ symptoms along with a fall in the levels of
their measured inflammatory markers.50 Key studies that
explored the influence of EN on microbiota and dysbiosis
are summarized in Table 3.
An additional mechanism of action of enteral therapy
involves resting the bowel51 and excluding offensive dietary
factors involved in exacerbations of immune responses,
thereby interrupting the so-called “bacterial penetration
cycle.”52 Previous findings provided evidence that there are
a variety of diets and food preservatives that can compro-
mise the integrity of the intestinal epithelial barrier.53,54
In the “bacterial penetration cycle” hypothesis, it has been
proposed that the loss of intestinal epithelial barrier func-
tion, secondary to consumed dietary factors, allows the
adherence and trafficking of bacteria and other antigenic
substances, triggering the mucosal immune system.55 This
in turn results in exaggerated immune responses and sub-
sequent tissue destruction with further loss in the integrity
488 Nutrition in Clinical Practice 33(4)

Table 3. Studies on the Role of EN in Microbiota Modifications.

Date, first author, reference, journal

name Study description and main findings

2005 Lionetti et al46 Prospective study: Fecal microflora were collected from 9 children and adults with
Journal of Parenteral and Enteral active CD treated with PF and 5 healthy children and assessed with 16S rRNA
Nutrition gene and temperature gradient gel electrophoresis. The microflora were modified in
all treated CD patients.
2008 Leach et al47 Prospective study: Fecal samples of 6 CD patients and matched healthy controls were
Alimentary Pharmacology & collected before, during, and after EN treatment, and the diversity was evaluated
Therapeutics using denaturing gradient gel electrophoresis. Significant changes in the
composition of bacteria was noted with EN treatment and persisted for 4 months.
2015 Kaakoush et al48 Prospective study: Fecal samples collected from 5 CD patients treated with EN and
Clinical and Translational matched healthy controls were assessed using 16S rRNA gene and whole-genome
Gastroenterology high throughput sequencing. Microbial diversity in CD patients was significantly
lower than that in matched controls and the number of operational taxonomic
units, reflecting microbial diversity and decreased substantially upon starting EN
treatment and correlated with a drop in disease activity. A number of bacterial
species within the Firimictus family was found correlating with disease activity.
2016 Walton et al50 Observational study: Fecal samples collected from CD patients treated with EN and
European Journal of Clinical Nutrition healthy controls were studied by gas chromatography and mass spectrometry for
investigating effects of enteral therapy on gut microbiota metabolism. A reduction
in the fecal concentrations of microbial metabolites, including SCAFs, 1-propanol,
1-butanol, and esters of SCAFs, was noted following the treatment with EN.

CD, Crohn’s disease; EN, enteral nutrition; PF, polymeric formula; SCAFs, short-chain fatty acids.

of mucosal barrier, resulting in a continuous cycle of diet-
induced damage of the mucosal barrier, bacterial penetra-
tion, and triggered mucosal immune responses.52 Therefore,
the underlying mechanism by which EEN therapy induces
therapeutic benefits for the patients might at least in part be
associated to an interruption of this vicious cycle of diet-
facilitated bacterial penetration.
EEN as a Treatment Option for CD
Induction and Maintenance of Remission
Several clinical trials involving pediatric CD patients treated
with EEN or steroids have shown that EEN is as efficient
as corticosteroids in inducing remission.35,56,57 However,
the observed supported growth and enhanced development
following EN treatment with no reported adverse effects
as those with corticosteroid use indicate superiority of
EEN over conventional medications as a first-line therapy
for treating children with active CD.58 In a meta-analysis
involving 7 randomized controlled trials of pediatric CD
patients, Dziechciarz et al59 also reported that the efficacy
of EN was incomparable with that of steroids in inducing
remissions. Consistently, as per the 2014 consensus guide-
lines of the European Crohn’s and Colitis Organization
and the European Society of Pediatric Gastroenterology,
Hepatology and Nutrition for the medical management
of pediatric CD, an enteral diet given as EEN is the
preferred induction therapy for children with active CD.60
Furthermore, there is an emerging role for enteral therapy
in maintaining surgically and medically induced remissions
in CD patients.61,62
However, EEN is believed to be less effective in adult
CD patients than in children with CD.63,64 This is partially
attributed to poor compliance in adults.65 Thus, current
clinical practice does not support the use of EN therapy
as an induction treatment for adult CD patients, except
in those willing to undergo this therapy.66 An additional
drawback for EN is the long duration of therapy that can
be extended for several weeks as an induction of remission
and even longer if used as a maintenance therapy.67,68
Exclusive or Partial Nutrition
The enteral diet can be administered as EEN or in con-
junction with a normal diet as partial enteral nutrition
(PEN).69 However, PEN in combination with a free diet
appears to be less effective in inducing remissions than as
when given exclusively.70 Sigall-Boneh et al71 investigated
the effect of PEN together with a structured CD exclusion
diet and showed favorable outcomes. Nevertheless, PEN has
not been established as an induction treatment60 ; however,
it may have a role as a maintenance therapy.72
Disease Location and Treatment Duration
Disease location is perceived as a potential factor in
determining the efficacy of EEN for inducing remissions in
CD patients.73 EEN treatment has been shown to induce
remissions at higher rates in patients with small bowel
Alhagamhmad
involvement than in those with isolated colonic
involvement.74 In contrast, in a different trial, the
disease location did not influence efficacy.57 Regarding
the treatment duration, as an induction treatment, enteral
therapy is usually administered during a period of up to
8 weeks in pediatric patients with active CD.60 However,
an optimal treatment duration for enteral therapy, if
used as maintenance therapy, remains unknown and is
influenced by patient compliance.75 EN was administered
as a nocturnal supplement for 1 year after the induction
of remissions using EEN.76 In contrast, the follow-up
duration was longer (mean ࣘ 2.6 years) in patients who had
undergone surgical interventions for inducing remission.77
Discussing Potential Venues for Enhancing
the Efficacy
Improving Compliance
Despite the current limitations that discourage wide utiliza-
tion of enteral diet, improving the palatability might be a
realistic option to improve compliance. Guo et al78 investi-
gated the efficacy of EEN using more palatable formulas
in adult patients with active CD and assessed its effects.
Remission rate was 85% in the treated patients following a
4-week PF regimen.78 Furthermore, there was a significant
improvement in the quality-of-life score, social function,
and emotional status after the treatment. In addition,
around 61% of the recruited participants expressed their
willingness to undergo treatment again to induce remission
if their disease relapsed.78 Thus, these initial observations
indicate that enhancing compliance through the use of
more palatable formulas might be a viable option for better
utilization of the therapy; however, further clinical studies
are warranted to validate these preliminary observations.
Combined Therapy
Commonly, the enteral therapy regimen for treating CD
patients involves EEN6 or, as recently proposed, PEN is
given with a CD exclusion diet.71 However, there is a trend
of using enteral diet as a concomitant therapy given with
the other immunosuppressive medications used in IBD
management,79 which might provide an additional potential
setting for better utilization of EN therapy in the man-
agement of CD.80 Previous research enrolled 498 patients
with CD patients and investigated the effects of combined
EEN with immunosuppressive drugs.81 The recruited pa-
tients were randomly grouped into 1 of the following 3
groups: those who were undergoing preoperative immuno-
suppressive treatments alone, those undergoing preoper-
ative immunosuppressive treatments in combination with
EEN treatment, and those who received none of these
treatments. This study reported better outcomes in patients
who underwent the combination treatment. The rates of
489
stoma creation, postoperative complications, and urgent
need of surgical intervention operation were significantly
lower in patients who received the combination treatment
than in those in the other groups. The authors also noted
that the combination treatment group exhibited in addition
an extended preoperative drug-free interval.81 Furthermore,
favorable outcomes were also observed in a recent meta-
analysis that included 4 studies that investigated the efficacy
of infliximab as a monotherapy for CD patients or in
combination with an enteral diet.82 Concomitant therapy
was more effective for both, inducing and maintaining
clinical remission than biological therapy alone.82 However,
it should be noted that in these few trials, EEN was not
used as a monotherapy; therefore, whether enteral diet
works better as part of a concomitant therapy or when
administered solely warrants further investigations.
Magnifying Immunomodulatory Effect
Enhancing the anti-inflammatory activity of EN by manip-
ulating the concentrations of its active components is an
additional, promising option for enhancing the efficacy.83
Recent studies that used colonic epithelial cells cultured in
the laboratory as an in vitro model of IBD have shown that
the glutamine and arginine components of a polymeric diet,
at concentrations used in the commercial formula (PF), are
capable of suppressing the TNF-α-induced IL-8 production
from the stimulated cultured cells.31,84 Furthermore, these
constituents exhibited a dose-dependent anti-inflammatory
effect mediated through the synergistic interactions against
the I-κB kinase activity of the NF-κB pathway, indi-
cating the feasibility of using higher concentrations of
these active nutrients.31,84 Interestingly, manipulating the
glutamine and arginine components of the standard PF
enhanced the anti-inflammatory properties.31,84 Certainly,
PF enriched with glutamine and arginine amino acids (at
final concentrations of 50 mM and 20 mM, respectively,
in the culture media) was superior to the standard PF in
terms of amelioration of TNF-α-induced IL-8 response
in the stimulated intestinal epithelial cells.31 Interestingly,
the addition of curcumin, an additional immunoactive
nutrient, to glutamine-enriched and arginine-enriched PF
further enhanced the anti-inflammatory activity in these
experiments.83 Moreover, these initial observations were
investigated further in an in vivo mouse model of coli-
tis developed by exposing C57BL/6 mice to dextran sul-
fate sodium to induce intestinal inflammation.85 Amino
acid–enriched PF substantially prevented dextran sulfate
sodium–induced weight loss, reduced myeloperoxidase ac-
tivity, suppressed the expression of proinflammatory cy-
tokines, and partially ameliorated local colonic tissue injury
with no reported adverse effects. Although the treatment
with the enriched formula did not show measurable superi-
ority over the treatment with the standard PF, the reversal of
490
inflammatory changes was similar to that in standard PF but
with 15% lower total given volume.85 Thereafter, additional
experiments involving colonic biopsies collected from 10
children with active CD and matched controls investigated
further the efficacy of the enriched formula in attenuating
intestinal inflammations.85 The new formula enriched with
amino acids showed superiority over the standard PF in
suppressing inflammation in the cultured biopsies with no
adverse effects on tissue viability.85 These findings suggested
that altering the composition of the enteral diet might
provide a more effective and safe therapy with enhanced
immune effects.
Collectively, despite the proven efficacy of the currently
used enteral diet for the treatment of CD in clinical
practice, this therapy has limitations in ameliorating the
exaggerated immune responses and thereby in attenuating
intestinal inflammation. Nevertheless, the recent preclinical
studies have shown that manipulating concentrations of the
glutamine and arginine components of the polymeric diet
magnifies the immunomodulatory effects, which might have
the potential to enhance the therapeutic utilities of EEN
in treating patients with CD. However, it should be noted
that the translation into a real therapy in actual clinical
practice remains ambiguous. Indeed, a glutamine-fortified
polymeric diet did not show any measurable advantages over
the standard formula in a trial involving patients with CD.86
Furthermore, it has been claimed that enteral glutamine
and arginine supplementations did not provide significant
clinical benefits over the standard enteral formulas in tri-
als involving intensive care unit and burn patients.87,88 In
contrast, in several other trials, enteral diets enriched with
glutamine and/or arginine showed several positive clinical
outcomes in critically ill and surgical patients (acceler-
ated wound healings, reductions in nosocomial infection-
associated complications, and hospital mortality rate).89-91
In addition, considering the recent findings of enhanced
activity of the polymeric diet following optimization of
the amino acids component, enriched enteral therapy ap-
pears to possess superior therapeutic utility in treating
IBD patients over the standard formulations. This potential
superiority of enriched diets, however, might not be evident
if the key ingredients are not supplemented in the optimal
active concentrations. Thus, further studies are required to
validate these preliminary observations.
Summary
Enteral diet therapy plays an important role in the man-
agement of CD and can offer several therapeutic bene-
fits. Attenuating exaggerated immune responses, enhanc-
ing mucosal healing and counteracting dysbiosis appear
to be the underlying mechanisms of actions of EN. The
immunomodulatory effects of EN are attributed at least in
part to blocking the major intracellular signaling pathways,
Nutrition in Clinical Practice 33(4)
including NF-κB transcription factor. Furthermore, it is
likely that several components of EN, including glutamine
and arginine amino acids account for that activity. Al-
tering the composition of the enteral diet seems effective
in enhancing the anti-inflammatory property and thereby
providing the opportunity for better utilization in clinical
practice.
Acknowledgments
This review article is part of fulfillments for a clinical master’s
degree in pediatrics.
Statement of Authorship
M.H. Alhagamhmad contributed to the conception and de-
sign of the research; M.H. Alhagamhmad contributed to the
acquisition, interpretation and analysis of the data; M.H.
Alhagamhmad drafted and critically revised the manuscript.
M.H. Alhagamhmad agrees to be fully accountable for en-
suring the integrity and accuracy of the work, and read and
approved the final manuscript.
References
1. Hanauer SB. Inflammatory bowel disease: epidemiology, pathogenesis,
and therapeutic opportunities. Inflamm Bowel Dis. 2006;12:S3-S9.
2. Beasley G. P-203 disease activity, depression, and health related quality
of life in adolescents with inflammatory bowel disease attending a
support group. Inflamm Bowel Dis. 2016;22:S70.
3. Molodecky NA, Soon S, Rabi DM, et al. Increasing incidence and
prevalence of the inflammatory bowel diseases with time, based on
systematic review. Gastroenterology. 2012;142:46-54.
4. Alhagamhmad MH, Leach ST, Lemberg DA, et al. Changing patterns
in the epidemiology of Crohn disease. J Gastroenterol Hepatol Res.
2015;4:1805-1809.
5. Limketkai BN, Parian AM, Shah ND, et al. Short bowel syndrome and
intestinal failure in Crohn’s disease. Inflamm Bowel Dis. 2016;22:1209-
1218.
6. Alhagamhmad MH, Day AS, Lemberg DA, et al. An update of the
role of nutritional therapy in the management of Crohn’s disease. J
Gastroenterol. 2012;47:872-882.
7. Day A, Whitten K, Sidler M, et al. Systematic review: nutritional
therapy in paediatric Crohn’s disease. Aliment Pharmacol Therap.
2008;27:293-307.
8. Ezri J, Marques-Vidal P, Nydegger A. Impact of disease and treatments
on growth and puberty of pediatric patients with inflammatory bowel
disease. Digestion. 2012;85:308-319.
9. Day AS, Whitten KE, Lemberg DA, et al. Exclusive enteral feeding
as primary therapy for Crohn’s disease in Australian children and
adolescents: a feasible and effective approach. J Gastroenterol Hepatol.
2006;21:1609-1614.
10. Cashman KD, Shanahan F. Is nutrition an aetiological factor for
inflammatory bowel disease. Eur J Gastroenterol Hepatol. 2003;15:607-
613.
11. Hume G, Radford-Smith GL. The pathogenesis of Crohn s disease in
the 21st century. Pathology. 2002;34:561-567.
12. Rogler G, Andus T. Cytokines in inflammatory bowel disease. World J
Surg. 1998;22:382-389.
13. Schottelius A, Baldwin A Jr. A role for transcription factor NF-kB in
intestinal inflammation. Inter J Colorec Dis. 1999;14:18-28.
14. Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis.
Gastroenterology. 1998;115:182-205.
Alhagamhmad
15. Waetzig GH, Seegert D, Rosenstiel P, et al. p38 mitogen-activated pro-
tein kinase is activated and linked to TNF- signaling in inflammatory
bowel disease. J Immunol. 2002;168:5342-5351.
16. Pagliari D, Piccirillo CA, Larbi A, et al. The interactions between
innate immunity and microbiota in gastrointestinal diseases. J Immunol
Res. 2015;2015:1-3.
17. Alhagamhmad MH, Day AS, Lemberg DA, et al. An overview of the
bacterial contribution to Crohn disease pathogenesis. J Med Microbiol.
2016;65:1049-1059.
18. Goll R, van Beelen Granlund A. Intestinal barrier homeostasis in
inflammatory bowel disease. Scand J Gastroenterol. 2015;50:3-12.
19. Pastorelli L, De Salvo C, Mercado JR, et al. Central role of the gut
epithelial barrier in the pathogenesis of chronic intestinal inflamma-
tion: lessons learned from animal models and human genetics. Front
Immunol. 2013;4:1-12.
20. Tan X, Mao J, Tang H, et al. Mechanisms underlying clinical efficacy
of enteral nutrition in inflammatory bowel disease. Intern J Clini Expe
Med. 2017;10: 2026-2035.
21. Day AS, Lopez RN. Exclusive enteral nutrition in children with Crohn’s
disease. World J Gastroenterol.2015;21:6809-6816.
22. Voitk AJ, Echave V, Feller JH, et al. Experience with elemental diet in
the treatment of inflammatory bowel disease: is this primary therapy.
Arch Surg. 1973;107:329-333.
23. Heuschkel R. Enteral nutrition in Crohn disease: more than just
calories. J Paediatr Gastroenterol Nutr 2004; 38: 239-241.
24. Yamamoto T, Nakahigashi M, Saniabadi AR, et al. Impacts of long-
term enteral nutrition on clinical and endoscopic disease activities and
mucosal cytokines during remission in patients with Crohn’s disease: a
prospective study. Inflamma Bowel Dis. 2007;13:1493-1501.
25. Yamamoto T, Nakahigashi M, Umegae S, et al. Impact of elemental
diet on mucosal inflammation in patients with active Crohn’s disease:
cytokine production and endoscopic and histological findings. Inflamm
Bowel Dis. 2005;11:580-588.
26. Beattie R, Schiffrin E, Donnet-Hughes A, et al. Polgmeric nutrition
as the primary therapy in children with small bowel Crohn’s disease.
Aliment Pharmacol Therap. 1994;8:609-615.
27. Fell J, Paintin M, Arnaud-Battandier F, et al. Mucosal healing and a
fall in mucosal pro-inflammatory cytokine mRNA induced by a specific
oral polymeric diet in paediatric Crohn’s disease. Aliment Pharmacol
Therap. 2000;14:281-290.
28. Bannerjee K, Camacho-Hübner C, Babinska K, et al. Anti-
inflammatory and growth-stimulating effects precede nutritional
restitution during enteral feeding in Crohn disease. J Pediatr
Gastroenterol Nutr. 2004;38:270-275.
29. de Jong NS, Leach ST, Day AS. Polymeric formula has direct anti-
inflammatory effects on enterocytes in an in vitromodel of intestinal
inflammation. Dig Dis Sci. 2007;52:2029-2036.
30. Nahidi L, Corley SM, Wilkins MR, et al. The major pathway by which
polymeric formula reduces inflammation in intestinal epithelial cells: a
microarray-based analysis. Genes Nutr. 2015;10:1-9.
31. Alhagamhmad MH, Day AS, Lemberg DA, et al. Exploring and
enhancing the anti-inflammatory properties of polymeric formula.
JPEN J Parenter Enteral Nutr. 2016;41:436-445.
32. Meister D, Bode J, Shand A, et al. Anti-inflammatory effects of enteral
diet components on Crohn’s disease-affected tissues in vitro. Dig Liver
Dis. 2002;34:430-438.
33. McGuckin MA, Eri R, Simms LA, et al. Intestinal barrier dysfunction
in inflammatory bowel diseases. Inflamma Bowel Dis. 2009;15:100-113.
34. Shah S, Colombel JF, Sands B, et al. Systematic review with meta-
analysis: mucosal healing is associated with improved long-term out-
comes in Crohn’s disease. Aliment Pharmacol Therap. 2016;43:317-333.
35. Grover Z, Muir R, Lewindon P. Exclusive enteral nutrition induces
early clinical, mucosal and transmural remission in paediatric Crohn’s
disease. J Gastroenterol. 2014;49:638-645.
491
36. Grover Z, Burgess C, Muir R, et al. Early mucosal healing with
Exclusive Enteral Nutrition is associated with improved outcomes in
newly diagnosed children with luminal Crohn’s disease. J Crohn’s and
Colitis. 2016;10:1159-1164.
37. Borrelli O, Cordischi L, Cirulli M, et al. Polymeric diet alone versus
corticosteroids in the treatment of active pediatric Crohn’s disease:
a randomized controlled open-label trial. Clin Gastroenterol Hepatol.
2006;4:744-753.
38. Berni Canani R, Terrin G, Borrelli O, et al. Short-and long-term ther-
apeutic efficacy of nutritional therapy and corticosteroids in paediatric
Crohn’s disease. Dig Liver Dis. 2006;38:381-387.
39. Wang F, Graham WV, Wang Y, et al. Interferon-γ and tumor necrosis
factor-α synergize to induce intestinal epithelial barrier dysfunction
by up-regulating myosin light chain kinase expression. Am J Pathol.
2005;166:409-419.
40. Al-Sadi R, Boivin M, Ma T. Mechanism of cytokine modulation of
epithelial tight junction barrier. Front Biosci. 2009;14:2765-2778.
41. Ma TY, Boivin MA, Ye D, et al. Mechanism of TNF-α modulation of
Caco-2 intestinal epithelial tight junction barrier: role of myosin light-
chain kinase protein expression. Am J Physiol Gastro Liver Physiol.
2005;288:G422-G430.
42. Nahidi L, Day AS, Lemberg DA, et al. Differential effects of nutritional
and non-nutritional therapies on intestinal barrier function in an in
vitro model. J Gastroenterol. 2012;47:107-117.
43. Nahidi L, Leach ST, Mitchell HM, et al. Inflammatory bowel disease
therapies and gut function in a colitis mouse model. BioMed Res Intern.
2013;2013:1-15.
44. Ye D, Ma I, Ma TY. Molecular mechanism of tumor necrosis factor-α
modulation of intestinal epithelial tight junction barrier. Am J Physiol-
Gastrointest Liver Physiol. 2006;290:G496-G504.
45. Lionetti P, Callegari M, Cavicchi M, et al. P0127 PP Enteral nutrition-
induced remission is associated with profound modificatioin of in-
testinal microflora in Crohn’s disease. J Pediatr Gastroenterol Nutr.
2004;39:S106.
46. Lionetti P, Callegari ML, Ferrari S, et al. Enteral nutrition and
microflora in pediatric Crohn’s disease. JPEN J Parenter Enteral Nutr.
2005;29:S173-S178.
47. Leach S, Mitchell H, Eng W, et al. Sustained modulation of intestinal
bacteria by exclusive enteral nutrition used to treat children with
Crohn’s disease. Aliment Pharmacol Therap. 2008;28:724-733.
48. Kaakoush NO, Day AS, Leach ST, et al. Effect of exclusive enteral
nutrition on the microbiota of children with newly diagnosed Crohn’s
disease. Clin Trans Gastroenterol. 2015;6:1-11.
49. Nahidi L, Day AS, Lemberg DA, et al. Paediatric inflammatory
bowel disease: a mechanistic approach to investigate exclusive enteral
nutrition treatment. Scientifica. 2014;2014:1-9.
50. Walton C, Montoya MPB, Fowler DP, et al. Enteral feeding reduces
metabolic activity of the intestinal microbiome in Crohn’s disease: an
observational study. Eur J Clin Nutr. 2016;70:1052-1056.
51. Greenberg G, Fleming C, Jeejeebhoy K, et al. Controlled trial of bowel
rest and nutritional support in the management of Crohn’s disease. Gut.
1988;29:1309-1315.
52. Levine A, Wine E. Effects of enteral nutrition on Crohn’s disease:
clues to the impact of diet on disease pathogenesis. Inflamm Bowel Dis.
2013;19:1322-1329.
53. Shen W, Gaskins HR, McIntosh MK. Influence of dietary fat on
intestinal microbes, inflammation, barrier function and metabolic
outcomes. J Nutr Biochem. 2014;25:270-280.
54. Lerner A, Matthias T. Changes in intestinal tight junction permeability
associated with industrial food additives explain the rising incidence of
autoimmune disease. Autoimmun Rev. 2015;14:479-489.
55. Barnich N, Denizot J, Darfeuille-Michaud A. E. coli-mediated gut in-
flammation in genetically predisposed Crohn’s disease patients. Pathol
Biol. 2013;61:e65-e69.
492
56. Critch J, Day AS, Otley A, et al. Use of enteral nutrition for the
control of intestinal inflammation in pediatric Crohn disease. J Pediatr
Gastroenterolo Nutr. 2012;54:298-305.
57. Buchanan E, Gaunt W, Cardigan T, et al. The use of exclusive
enteral nutrition for induction of remission in children with Crohn’s
disease demonstrates that disease phenotype does not influence clinical
remission. Aliment Pharmacol Therap. 2009;30:501-507.
58. Heuschkel RB, Menache CC, Megerian JT, et al. Enteral nutrition and
corticosteroids in the treatment of acute Crohn’s disease in children. J
Pediatr Gastroenterol Nutr. 2000;31:8-15.
59. Dziechciarz P, Horvath A, Shamir R, et al. Meta-analysis: enteral
nutrition in active Crohn’s disease in children. Aliment Pharmacol
Therap. 2007;26:795-806.
60. Ruemmele F, Veres G, Kolho K-L, et al. Consensus guidelines of
ECCO/ESPGHAN on the medical management of pediatric Crohn’s
disease. J Crohn’s Colitis. 2014;8:1179-1207.
61. El-Matary W, Otley A, Critch J, et al. Enteral feeding therapy for
maintaining remission in Crohn’s disease: a systematic review. JPEN
J Parenter Enteral Nutr. 2015;41: 550-561.
62. Yamamoto T, Nakahigashi M, Umegae S, et al. Impact of long-term
enteral nutrition on clinical and endoscopic recurrence after resection
for Crohn’s disease: a prospective, non-randomized, parallel, controlled
study. Aliment Pharmacol Therap. 2007;25:67-72.
63. Hartman C, Eliakim R, Shamir R. Nutritional status and nutri-
tional therapy in inflammatory bowel diseases. World J Gastroenterol.
2009;15:2570-2578.
64. Zachos M, Tondeur M, Griffiths A. Enteral nutritional therapy for
induction of remission in Crohn’s disease. Cochrane Database Syst Rev.
2007;1:1-50.
65. Kansal S, Wagner J, Kirkwood C, et al. Enteral nutrition in Crohn’s
disease: an underused therapy. Gastroenterol Res Pract. 2013;2013:1-
11.
66. Wall CL, Day AS, Gearry RB. Use of exclusive enteral nutrition
in adults with Crohn’s disease: a review. World J Gastroenterol.
2013;19:7652-7660.
67. Whitten KE, Rogers P, OoI CY, Day AS. International survey of
enteral nutrition protocols used in children with Crohn’s disease. J Dig
Dis. 2012;13:107-112.
68. Gråfors JM, Casswall TH. Exclusive enteral nutrition in the treatment
of children with Crohn’s disease in Sweden: a questionnaire survey.
Acta Paediatr. 2011;100:1018-1022.
69. Kang Y, Kim S, Kim SY, et al. Effect of short-term partial enteral
nutrition on the treatment of younger patients with severe Crohn’s
disease. Gut Liver. 2015;9:87-93.
70. Johnson T, Macdonald S, Hill SM, et al. Treatment of active Crohn’s
disease in children using partial enteral nutrition with liquid formula: a
randomised controlled trial. Gut. 2006;55:356-361.
71. Sigall-Boneh R, Pfeffer-Gik T, Segal I, et al. Partial enteral nutrition
with a Crohn’s disease exclusion diet is effective for induction of
remission in children and young adults with Crohn’s disease. Inflamm
Bowel Dis. 2014;20:1353-1360.
72. Schulman JM, Pritzker L, Shaoul R. Maintenance of remission with
partial enteral nutrition therapy in pediatric Crohn’s disease: a retro-
spective study. Can J Gastroenterol Hepatol. 2017;2017:1-7.
73. de Bie C, Kindermann A, Escher J. Use of exclusive enteral nutrition
in paediatric Crohn’s disease in The Netherlands. J Crohn’s Colitis.
2013;7:263-270.
74. Afzal NA, Davies S, Paintin M, et al. Colonic Crohn’s disease in
children does not respond well to treatment with enteral nutrition if
the ileum is not involved. Dig Dis Sci. 2005;50:1471-1475.
Nutrition in Clinical Practice 33(4)
75. Yamamoto T, Nakahigashi M, Umegae S, et al. Enteral nutrition for
the maintenance of remission in Crohn’s disease: a systematic review.
Eur J Gastroenterol Hepatol. 2010;22:1-8.
76. Duncan H, Buchanan E, Cardigan T, et al. A retrospective study
showing maintenance treatment options for paediatric CD in the first
year following diagnosis after induction of remission with EEN: sup-
plemental enteral nutrition is better than nothing. BMC Gastroenterol.
2014;14:1-7.
77. Esaki M, Matsumoto T, Hizawa K, et al. Preventive effect of nutri-
tional therapy against postoperative recurrence of Crohn disease, with
reference to findings determined by intra-operative enteroscopy. Scan J
Gastroenterol. 2005;40:1431-1437.
78. Guo Z, Wu R, Zhu W, et al. Effect of exclusive enteral nutrition on
health-related quality of life for adults with active Crohn’s disease. Nutr
Clin Pract. 2013;28:499-505.
79. Wu X-L, Chen R-P, Tao L-P, et al. Infliximab combined with enteral
nutrition for managing Crohn’s disease complicated with intestinal
fistulas. Gastroenterol Res Pract. 2016;2016:1-6.
80. Triantafillidis JK, Vagianos C, Papalois AE. The role of enteral
nutrition in patients with inflammatory bowel disease: current aspects.
BioMed Res Intern. 2015;2015:1-12.
81. Li Y, Zuo L, Zhu W, et al. Role of exclusive enteral nutrition in the
preoperative optimization of patients with Crohn’s disease following
immunosuppressive therapy. Medicine. 2015;94:1-9.
82. Nguyen DL, Palmer LB, Nguyen ET, et al. Specialized enteral nutrition
therapy in Crohn’s disease patients on maintenance infliximab therapy:
a meta-analysis. Therap Advan Gastroenterol. 2015;8:168-175.
83. Alhagamhmad M, Leach S, Day AS, et al. Mo1232 development of
an enhanced nutritional therapy for the treatment of Crohn’s disease.
Gastroenterol. 2015;148:S-644.
84. Alhagamhmad M, Leach S, Day AS, et al. Mo1231 attenuation
of intestinal inflammation by co-supplementation of glutamine and
arginine. Gastroenterol. 2015;148:S-644.
85. Alhagamhmad MH, Lemberg DA, Day AS, et al. Advancing nutri-
tional therapy: a novel polymeric formulation attenuates intestinal in-
flammation in a murine colitis model and suppresses pro-inflammatory
cytokine production in ex-vivo cultured inflamed colonic biopsies. Clin
Nutr. 2017;36:497-505.
86. Akobeng AK, Miller V, Stanton J, et al. Double-blind randomized
controlled trial of glutamine-enriched polymeric diet in the treatment
of active Crohn’s disease. J Pediatr Gastroenterol Nutr. 2000;30:78-84.
87. Peterik A, Milbrandt EB, Darby JM. Immunonutrition in critical
illness: still fishing for the truth. Crit Care. 2009;13:1-3.
88. Van Zanten AR, Dhaliwal R, Garrel D, et al. Enteral glutamine
supplementation in critically ill patients: a systematic review and meta-
analysis. Crit Care. 2015;19:1-16.
89. Garrel D, Patenaude J, Nedelec B, et al. Decreased mortality and
infectious morbidity in adult burn patients given enteral glutamine
supplements: a prospective, controlled, randomized clinical trial. Crit
Care Med. 2003;31:2444-2449.
90. Lorenz KJ, Schallert R, Daniel V. Immunonutrition–the influence of
early postoperative glutamine supplementation in enteral/parenteral
nutrition on immune response, wound healing and length of hospital
stay in multiple trauma patients and patients after extensive surgery.
GMS Interdiscip Plast Reconstr Surg DGPW. 2015;4:1-10.
91. Caparrós T, Lopez J, Grau T. Early enteral nutrition in critically ill
patients with a high-protein diet enriched with arginine, fiber, and
antioxidants compared with a standard high-protein diet. The effect
on nosocomial infections and outcome. JPEN J Parenter Enteral Nutr.
2001;25:299-309.
Clinical Observations

Nutrition in Clinical Practice

Volume 33 Number 4
Case Report of Wernicke’s Encephalopathy After August 2018 510–514

C 2017 American Society for

Sleeve Gastrectomy Parenteral and Enteral Nutrition

DOI: 10.1177/0884533617722758
wileyonlinelibrary.com

Leslie A. Hamilton, PharmD, BCPS, BCCCP1 ; Sarah H. Darby, BS1 ;

Allan J. Hamilton, MD2 ; Matthew H. Wilkerson, MD3 ; and Kabel A. Morgan, MD3

Abstract
Background: We report a case of a patient who was 3 months post–sleeve gastrectomy and presented with acute stroke symptoms
ultimately due to Wernicke’s encephalopathy (WE) after bariatric surgery. A 20-year-old white female presented to an outside
hospital 3 months after sleeve gastrectomy complaining of nausea and vomiting. She initially underwent a cholecystectomy and later
became less responsive and required intubation. Magnetic resonance imaging changes, presumed to be an acute stroke, prompted
her transfer to our facility. Intravenous (IV) thiamin was administered, and the patient’s symptoms improved over the course of her
hospital stay. Results: Thiamin levels were markedly low, and the patient rapidly improved with the administration of IV thiamin.
The patient was discharged to inpatient rehabilitation. Conclusion: Bariatric surgery is a less common cause of WE but can lead
to acute WE due to malabsorption of thiamin. In patients undergoing bariatric surgery, clinicians should be vigilant about the
potential for WE to occur. In addition, based on history, WE should be considered in the differential diagnosis for symptoms of
acute ischemic stroke. (Nutr Clin Pract. 2018;33:510–514)

Keywords
bariatric surgery; sleeve gastrectomy; thiamin; thiamin deficiency; Wernicke encephalopathy

Historically associated with alcoholism, Wernicke’s en-
cephalopathy (WE) is an acute neurologic development due
to thiamin deficiency, with a classic triad of symptoms
consisting of ophthalmologic dysfunction and nystagmus,
ataxia, and mental status changes.1 The human body does
not produce any or store large amounts of thiamin, thus
requiring its consumption through the diet.2 Thiamin serves
as a cofactor for numerous pathways of carbohydrate
metabolism, which when altered leads to mitochondrial
damage and potential necrosis, apoptosis, and oxidative
stress.3
Because thiamin is obtained through the diet, deficiency
due to malnutrition is more likely in developing nations,
where proper nutrition may be more difficult to obtain.
In the Western world, where access to food may be less
difficult to obtain, WE is typically connected to alcoholism.
However, the number of nonalcoholic WE cases is likely to
increase due to malabsorption as an adverse result of the
growing number of bariatric operations being performed.
With more than one-third of the U.S. adult population
currently categorized as obese, it is not surprising to see
that the number of bariatric operations has increased from
approximately 16,000 per year in the early 1990s to almost
180,000 in 2012.4-6 Considering that these numbers repre-
sent only 1% of surgery-eligible patients, they are sure to
continue to grow and require clinicians to be on the lookout
for complications related to malabsorption postsurgery, be-
cause inadequate absorption of necessary vitamins and min-
erals can lead to preventable and dangerous deficiencies.7
Here we describe a case of a young woman presenting
with abnormal neurologic findings 3 months after sleeve
gastrectomy and later diagnosed with WE. A literature
review of PubMed with the search terms “Wernicke’s en-
cephalopathy,” “gastric sleeve,” and “bariatric surgery” and
of Medline with the terms “bariatric,” “bariatric surgery,”
“encephalopathy,” “surgery,” “Wernicke,” and “Wernicke’s
encephalopathy” returned a substantial list of articles in
From the 1 Department of Clinical Pharmacy, College of Pharmacy,
University of Tennessee Health Science Center, Knoxville, Tennessee,
USA; 2 Department of Anesthesiology and Pain Management, UT
Southwestern, Dallas, Texas, USA; and 3 Department of
Anesthesiology, Graduate School of Medicine, University of
Tennessee Medical Center, Knoxville, Tennessee, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on September 27, 2017.
Corresponding Author:
Leslie A. Hamilton, PharmD, BCPS, BCCCP, Department of Clinical
Pharmacy, College of Pharmacy, University of Tennessee Health
Science Center, 1924 Alcoa Highway, Box 117, Knoxville, TN 37909,
USA.
Email: lhamilt4@uthsc.edu
Hamilton et al 511

which a relative few were case reports of WE occurring after

bariatric surgery. The unifying link among the case reports is
that no patient example presented in quite the same manner.

Case Report
The patient is a 20-year-old white woman who was approx-
imately 3 months post–sleeve gastrectomy and presented to
an outside hospital in late August 2015 with 6 weeks of nau-
sea and vomiting due to gallstone pancreatitis. During those
6 weeks, she presented to an outside hospital 3 times and
reported a subjective history of subsisting on rehydration
solution secondary to this nausea and vomiting, with an 83-
lb weight loss in the 3 months since the procedure. During
her last presentation, she presented with hypotension to
the outside hospital’s emergency department, and elevated
lipase and amylase were found on laboratory tests. She was
admitted to its intensive care unit for volume resuscita-
tion, which was completed with lactated Ringer solution.
Her pancreatic enzymes were elevated due to suspected
gallstones, and she had a laparoscopic cholecystectomy
performed on outside hospital day 8 (HD 8). She was noted Figure 1. Magnetic resonance imaging: T1 weighted
to be neurologically intact after the procedure. Social history periaqueductal.
was negative for alcohol and tobacco use.
On the morning of outside HD 9, staff found that she
was markedly less responsive, although she was able to move
all extremities to command. This neurologic change was
evaluated with magnetic resonance imaging (MRI), which
showed bilateral diffusion restriction within the posterior
thalamus (1.5 × 0.5 cm) that extended into the posterior
midbrain adjacent to the aqueduct of Sylvius. This was
accompanied by increased signal on T2 weighting within
these areas of diffusion restriction (Figures 1–4). These
findings were consistent with an acute infarct, with top-of-
the-basilar syndrome. She was transferred to our facility
for evaluation and treatment of a suspected acute stroke.
The patient was intubated, sedated, and paralyzed prior to
arrival at our facility, with outside records reporting that she
was moving all extremities prior to intubation. No mention
was made about ophthalmoplegia or nystagmus in the
records from the outside hospital. Computer tomography
angiography of the head at our facility showed patency
of the basilar artery, bilateral posterior communicating
arteries, and bilateral posterior cerebral arteries, without ev-
idence of basilar occlusion or clot at the artery of Percheron.
The patient’s body mass index was 48 kg/m2 on ad-
Figure 2. Magnetic resonance imaging: T1 weighted.
mission (height, 1.68 m; weight, 135 kg). Pertinent abnor-
mal laboratory results on admission include the following:
sodium, 129 mEq/dL (reference range, 136–147); chloride, international normalized ratio, 1.38 (0.8–1.2); thiamin,
93 mEq/dL (98–109); glucose, 137 mg/dL (70–99); phospho- 29 nMol/L (66.5–200); thyroid-stimulating hormone, 0.441
rus, 2.1 mg/dL (2.7–4.5); magnesium, 1.5 mg/dL (1.6–2.6); IU/mL (0.45–4.5); serum albumin level, 3 g/dL (3.5–5.2);
amylase, 269 U/L (5–125); lipase, 229 U/L (8–78); aspartate prealbumin, 13 mg/dL (16–38); total cholesterol, 79 mg/dL
aminotransferase, 87 U/L (5–34); alanine aminotransferase, (100–199); HDL, 11 mg/dL (>40); and LDL, 29 mg/dL
63 U/L (0–55); prothrombin time, 14.8 seconds (9.1–12); (<100). Of note, although her thiamin level was drawn on
512
Figure 3. Magnetic resonance imaging: T2 fluid-attenuated
inversion recovery, periaqueductal.
Figure 4. Magnetic resonance imaging: T2 fluid-attenuated
inversion recovery.
the day of admission to our institution, it took 3 days for
the results to return from an outside laboratory. A urine
toxicology screen was performed at the outside hospital
and was positive for acetaminophen, benzodiazepines, and
cannabinoids.
A magnetic resonance venogram was performed on HD
1 and did not show evidence of a dural venous thrombus. A
Nutrition in Clinical Practice 33(4)
transthoracic echocardiogram was also performed on HD
1 and showed a normal left ventricular ejection fraction,
normal left atrial size, and no evidence of aortic valve
septal defects. Bilateral upper and lower extremity venous
duplex scans were negative for deep venous thrombosis.
Family members at bedside on HD 2 reported seizure-like
activity, at which time she was loaded with fosphenytoin
and levetiracetam, with levetiracetam continued for main-
tenance therapy. The electroencephalogram appreciated no
epileptiform abnormalities.
Inpatient medications included aspirin (81 mg per tube
daily), ceftriaxone (1 g, intravenous [IV], daily for HDs
8–15), enoxaparin (40 mg, subcutaneous, twice daily), es-
omeprazole (40 mg per tube daily), folic acid (1 mg per
tube daily), insulin (sliding scale, subcutaneous, daily),
levetiracetam (1000 mg, IV, twice daily), polyethylene glycol
3350 (17 g per tube daily), and thiamin (500 mg, intravenous,
× 1 dose on HD 3, followed by 100 mg, IV, daily). On HD
4, cholecalciferol (800 IU per tube daily) and multivitamin
(5 mL per tube daily) were initiated, with cyanocobalamin
(1000 mcg per tube daily) and pyridoxine (50 mg per tube
daily) started on HD 6. Modafinil (200 mg by mouth daily)
was added on HD 10.
Electrolyte deficiencies were aggressively replaced, and
enteral nutrition (EN) was started slowly with consideration
of the risks of the possibility of refeeding syndrome. Her ini-
tial EN goals were 1700 kcal and 121 g of protein (1.8 g/kg
based on ideal body weight). EN was a polymeric 1 kcal/mL
of fiber-containing formula titrated to goal rate over 3 days.
Her initial sodium of 129 mEq/L was corrected to 136 mE/L
over 3 days. After extubation on HD 5, a mechanical soft
diet was started on HD 7 with oral nutrition supplements.
Follow-up MRI done on HD 7 showed fluid-attenuated
inversion recovery (FLAIR) hyperintensities in bilateral
medial thalami on postcontrast enhancement consistent
with WE. Physical examination prior to extubation showed
a disconjugate gaze. Physical examination after extubation
showed limited horizontal eye movements with nystagmus,
dysarthric speech without aphasia, bilateral dysmetria on
finger-nose-finger, and bilateral upper extremity ataxia. The
patient did complain of vision and hearing difficulty, but
she reported that this was present prior to admission. She
did have mild improvement with nystagmus and dysarthria
throughout the rest of her hospital course. Levetiracetam
was stopped on HD 7, as it was thought that there was no
evidence that she had experienced a seizure. She remained
in the hospital for 8 more days for rehabilitation with
physical and occupational therapy and awaited inpatient
rehabilitation placement. She was discharged on HD 15.
Discussion
Bariatric surgery includes Roux-en-Y gastric bypass, sleeve
gastrectomy, gastric band placement, and biliopancreatic
Hamilton et al
diversion, and each of these procedures carries with it
varying degrees of risk of later malabsorption.8 Roux-en-
Y gastric bypass is considered a malabsorptive interven-
tion because it decreases contact time between food and
biliary and pancreatic enzymes, thus limiting digestion and
absorption.9 Other procedures, such as the sleeve gastrec-
tomy, are considered restrictive and may lead clinicians to
inappropriately consider vitamin deficiencies as less of a
threat. Without proper intake of thiamin, patients are at
risk for developing WE, with reports showing that up to
30% of patients develop thiamin deficiency postsurgery.9
Our patient initially presented to an outside hospital that
was unfamiliar with bariatric surgery, which led to a delay
in treatment and subsequent worsening of symptoms over
the course of her stay. Another case that faced delays
due to misdiagnosis occurred in a patient after Roux-
en-Y bypass surgery10 ; however, our patient underwent
sleeve gastrectomy. This illustrates that practitioners cannot
assume that any 1 type of bariatric surgery is safe from
complications. Surgery not only induces structural changes
to the gastrointestinal tract, which decreases absorption
capabilities, but also leaves patients with a reduced appetite,
further complicating the risk of vitamin deficiencies.9 Our
patient reported nausea and vomiting, which is not unex-
pected given that vomiting is considered to be an important
predisposing risk factor for postbariatric WE with poor
thiamin intake.11 Our patient’s combination of vomiting
and relying on a rehydration solution for nutrition made her
extremely susceptible to WE-type symptoms.
While vomiting and poor oral intake are contributing
factors to the development of WE, patients present with
a variety of signs and symptoms once they report to a
facility for medical attention. One case reports a 24-year-old
man presenting 7 months postsurgery with complaints of
difficulty walking, diplopia, and dysarthria. The patient had
normal MRI results.12 Another case describes a 30-year-old
woman whose nonadherence with vitamin supplementation
led to the development of diplopia and oscillopsia 2 months
after sleeve gastrectomy.13 Others report that a 38-year-old
woman complained of weakness, nausea, and vomiting 1
week postsurgery. The following week she presented again
with mental status changes, hypokinesis, diplopia, and an
inability to speak logically, though her MRI results were
normal.14 In another report, a 27-year-old woman was
diagnosed with WE 3 weeks after her surgery with mild
mental status changes. She denied vomiting, and computed
tomography scan and MRI results were both normal. Her
symptoms resolved approximately 4 hours after admis-
sion and administration of thiamin.15 Two months after
gastric banding, a 40-year-old woman experienced vom-
iting, strabismus, urinary incontinence, and apathy. MRI
showed high-intensity lesions in the dorsomedial thalami,
hypothalami, mammillary bodies, mesencephalic tectum,
and periaqueductal region.1 A 31-year-old man presented
513
with nystagmus, ataxia, confusion, and ophthalmoplegia 2
months after his sleeve gastrectomy. His MRI also showed
hyperintense signals on T2 and FLAIR image in thalamus,
periaqueductal area, and mammillary bodies.16
The documented case reports discussed here vary greatly
in terms of how and when patients present after bariatric
surgery. Postoperative complications have been shown to
have an early or late onset and vary among symptoms such
as vomiting, ataxia, eye movement disorders, dysarthria,
confusion, and even urinary incontinence. There is debate
over the number of patients who actually present with the
classic triad of WE symptoms, with numbers ranging from
16% to 40%, meaning that the trio of symptoms cannot
be relied on for diagnosis.1,11,17 A systematic review con-
ducted by Kröll and colleagues found 13 well-documented
cases of WE after sleeve gastrectomy. Of these cases, 8
patients reported experiencing vomiting postsurgery, and
the length of time until WE after surgery varied from 2
weeks to 7 months. Other symptoms experienced by patients
included peripheral polyneuropathy, muscle weakness, and
dysarthria, but these symptoms were not consistent among
patients.11
Computed tomography scan and MRI findings varied
as well, while some patients did not undergo scans at all.
With a variety of vague symptoms and an incomplete
patient history, the diagnosis of WE may be overlooked
until MRI is performed. Expected MRI findings for WE
include hyperintense signals on T2-weighted and FLAIR
images in the periaqueductal regions and bilateral medial
thalami.18 Mammillary body and cerebellar vermis atrophy
may also be seen but is unlikely in malnutrition patients
without alcoholism, because this is often the first thiamin-
related deficiency causing neurologic dysfunction.17 Our
patient’s positive MRI findings, initially misinterpreted as
an acute stroke, were what prompted her transfer from an
outside hospital to our facility. MRI should be considered
for patients who report WE-like symptoms after bariatric
surgery; imaging is considered confirmatory due to 93%
specificity.15 However, image findings as described here may
appear in only 58% of patients and result in misleading MRI
results.15 In 1 literature review, MRI findings consistent with
WE were found in just 6 of the 10 patients tested.11 A similar
review states that 7 of 10 patients who underwent MRI
showed positive findings congruent with WE.19 Thiamin
levels should be evaluated to solidify diagnosis, and a posi-
tive patient response to thiamin therapy will also support a
diagnosis. Dosing of thiamin for WE is variable, and while
the recommendation is to acutely administer IV thiamin
(given erratic absorption in malnourished patients), there is
no standard dosing recommendation.20
Of note, our patient was only 20 years of age, which is
the youngest reported development of WE after bariatric
surgery. At this age, brain maturation may be incomplete,
and further research is necessary to determine if she may
514
be at increased risk of long-term complications. Although
the number of case reports in the literature is growing, it
is unlikely that many practitioners have witnessed this type
of case firsthand; thus, other practitioners play a crucial
role in suggesting differential diagnoses as well. Prophylactic
thiamin administration may also be suggested to avoid de-
ficiency complications if possible. Risk of thiamin toxicity
is minimal due to rapid renal elimination. Pharmacists can
also provide support with preoperative and postoperative
medication regimens.
Conclusions
This report describes a case of WE in a young patient with a
sleeve gastrectomy mimicking an acute stroke. Practitioners
should be vigilant in these patients for signs of thiamin and
other nutrient malabsorption and potential WE. There may
be a role of prophylactic thiamin in the prevention of WE
among bariatric surgery patients.
Statement of Authorship
L. A. Hamilton, M. H. Wilkerson, and K. A. Morgan equally
contributed to the conception and design of the manuscript;
A. J. Hamilton and S. H. Darby contributed to the acquisition
of the data. All contributed to the analysis and interpretation
of the data, drafted the manuscript, critically revised the
manuscript, agree to be fully accountable for ensuring the
integrity and accuracy of the work, and read and approved the
final manuscript.
References
1. Kühn AL, Hertel F, Boulanger T, Diederich NJ. Vitamin B1 in
the treatment of Wernicke’s encephalopathy due to hyperemesis after
gastroplasty. J Clin Neurosci. 2012;19:1303-1305.
2. Voet D, Voet JG, Pratt CW. Fundamentals of Biochemistry: Life at the
Molecular Level. 4th ed. Hoboken, NJ: John Wiley & Sons; 2012:495.
3. Martin PR, Singleton CK, Hiller-Sturmhöfel S. The role of thiamine
deficiency in alcoholic brain disease. Alcohol Res Health. 2003;27:134.
4. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of child-
hood and adult obesity in the United States, 2011-2012. JAMA.
2014;311:806-814.
Nutrition in Clinical Practice 33(4)
5. National Institute of Diabetes and Digestive and Kidney Diseases.
Longitudinal assessment of bariatric surgery. http://www.niddk.nih.
gov/health-information/health-topics/weight-control/Bariatric-
Surgery. Published January 2010. Accessed December 17, 2015.
6. American Society for Metabolic and Bariatric Surgery. New procedure
estimates for bariatric surgery: what the numbers reveal. http://connect.
asmbs.org/may-2014-bariatric-surgery-growth.html. Published May
2014. Accessed December 17, 2015.
7. Nguyen NT, Vu S, Kim E, et al. Trends in utilization of bariatric
surgery, 2009-2012. Surg Endosc. 2016;30:2723-2727.
8. Athanasiou A, Angelou A, Diamantis T. Wernicke’s encephalopathy
after sleeve gastrectomy: where do we stand today? A reappraisal. Surg
Obes Relat Dis. 2014;10:563.
9. Tack J, Deloose E. Complications of bariatric surgery: dumping
syndrome, reflux, and vitamin deficiencies. Best Pract Res Clin Gas-
troenterol. 2014;4:741-749.
10. Iannelli A, Addeo P, Novellas S, Gugenheim J. Wernicke’s en-
cephalopathy after laproscopic Roux-en-Y gastric bypass: a misdiag-
nosed complication. Obes Surg. 2010;20:1594-1596.
11. Kröll D, Laimer M, Borbély YM, Laederach K, Candinas K, Nett
PC. Wernicke encephalopathy: a future problem even after sleeve
gastrectomy? A systematic literature review [published online October
17, 2015]. Obes Surg.
12. Jeong HJ, Park JW, Kim YJ, et al. Wernicke’s encephalopathy after
sleeve gastrectomy for morbid obesity. Ann Rehabil Med. 2011;35:583-
586.
13. Jenkins PF. Wernicke encephalopathy. Am Orthopt J. 2015;65:104-108.
14. Makarewicz W, Kaska L, Kobiela J, et al. Wernicke’s syndrome after
sleeve gastrectomy. Obes Surg. 2007;17:704-706.
15. Saab R, El Khoury M, Farhat S. Wernicke’s encephalopathy three
weeks after sleeve gastrectomy. Surg Obes Relat Dis. 2014;10:992-994.
16. Landais A, Saint-Georges G. Wernicke’s encephalopathy following
sleeve gastrectomy for morbid obesity. Rev Méd Interne. 2014;35:760-
763.
17. Zuccoli G, Pipitone N. Neuroimaging findings in acute Wernicke’s en-
cephalopathy: review of the literature. Am J Roentgenol. 2009;192:501-
508.
18. Rathakrishnan R, Smith EW. MRI in non-alcoholic Wernicke’s en-
cephalopathy. J Clin Neurosci. 2008;15:478-480.
19. Milone M, Di Minno MND, Lupoli R, et al. Wernicke encephalopathy
in subjects undergoing restrictive weight loss surgery: a systematic
review of literature data. Eur Eat Disord Rev. 2014;22:223-229.
20. Day E, Bentham PW, Callaghan R, Kuruvilla T, George S. Thi-
amine for prevention and treatment of Wernicke-Korsakoff syn-
drome in people who abuse alcohol. Cochrane Database Syst Rev.
2013;(7):CD004033.
Clinical Observations

Nutrition in Clinical Practice

Volume 33 Number 4
Copper Deficiency Myelopathy After Upper August 2018 515–519

C 2017 American Society for

Gastrointestinal Surgery Parenteral and Enteral Nutrition

DOI: 10.1177/0884533617713955
wileyonlinelibrary.com

Dominic King, MBChB; Keith Siau, MBChB; Latha Senthil, MBBS;

Katherine F. Kane, MD; and Sheldon C. Cooper, MD

Abstract
A well-functioning alimentary canal is required for adequate nutrient absorption. Disruption to the upper gastrointestinal tract
through surgery can lead to micronutrient malnourishment. Copper deficiency has been noted in up to 10% of those undergoing
Roux-en-Y gastric bypass surgery, but sequalae are not frequently reported. The resultant deficiency states can have profound and
long-term consequences if not realized early and managed appropriately. Here we present a case of copper deficiency myelopathy,
a condition indistinguishable from subacute combined degeneration of the spinal cord, following upper gastrointestinal bypass
surgery for gastric ulceration, further complicated by inadequate nutrition. (Nutr Clin Pract. 2018;33:515–519)

Keywords
bariatric surgery; copper deficiency; gastric bypass; micronutrients; myelopathy; spinal cord diseases; vitamins

A variety of micronutrient deficiencies are well known to
cause disorders of the nervous system.1 The importance of
a functioning digestive tract is therefore cruicial. Disruption
to the gut, whether through surgical approach or disease
state, inevitably interferes with micronutrient absorption
and can have devastating sequalae. The prevalence of cop-
per deficiency following Roux-en-Y gastric bypass surgery
in 1 cross-sectional study was found to be 9.6%,2 and
in a series of obese women undergoing bariatric surgery
with Roux-en-Y gastric bypass, 15.4% were found to be
copper deficient.3 Asymptomatic copper deficiency is com-
mon in a disrupted upper gastrointestinal tract; though
rare, symptoms can also be seen.3,4 We present a case of
symptomatic copper deficiency in a patient following upper
gastrointestinal surgery. This produced a myelopathy and
clinical mimic of the subacute combined degeneration of
the spinal cord (SACDSC), characteristically associated
with vitamin B12 deficiency.
Case
History of Surgical Events and Nutrition State
Leading Up to Admission
A 50-year-old woman was referred to our tertiary center for
completion gastrectomy in January 2015, requiring preoper-
ative parenteral nutrition (PN) containing all essential trace
elements and vitamins (provided in Additrace, Solivito, and
Vitlipid) due to severe malnutrition.
The patient had nonhealing gastric ulcerations that
had been treated with laprascopic distal gastrectomy and
gastrojejunostomy in August 2010. The surgery had been
complicated by recurrent ulceration at the gastrojejunal
anastamosis and gastric outlet obstruction, requiring
further surgical intervention. She underwent subtotal
gastrectomy with Roux-en-Y gastric bypass in May 2012
with subsequent jejunostomy feeding. This was further
complicated by poor remnant emptying, and in June 2013, a
further laparotomy was performed where 55 cm of jejunum
was resected and a surgical jejunostomy was placed for
which the patient became dependent for feeding. Until refer-
ral to our hospital and the current admission, there had been
no PN. There were recurrent complications with enteral
feeding tubes (eg, blocked tube and feed leakage), leading to
malnutrition.
Ulcer biopsies ruled out malignancy, and investiga-
tion for gastrin-producing tumors and inflammatory bowel
disease was unremarkable. Investigations for rarer causes
From University Hospitals Birmingham, Birmingham, United
Kingdom.
Financial disclosure: None declared.
Conflicts of interest: Dr Sheldon Cooper has the following conflicts of
interest to declare: Covidien, Advisory Board and educational
sponsorship for an international meeting; Eli Lilly, Advisory Board;
Baxter/Willow, speakers’ fee.
This article originally appeared online on June 28, 2017.
Corresponding Author:
Sheldon C. Cooper, MD, Department of Gastroenterology, QEHB,
UHB NHS Foundation Trust, Edgbaston, Birmingham, West
Midlands B15 2GW, UK.
Email: sheldon.cooper@nhs.net
516
Figure 1. Magnetic resonance image with T2-weighted enhancement. Subacute combined degeneration of the spinal cord is seen,
with high signal changes in the dorsal columns, represented by the arrows.
of gastric ulcerations were also negative, and serum and
urinary nonsteroidal anti-inflammatory drugs were absent.
High-grade gastric outlet obstruction continued to be
an issue as the patient continued to eat and drink, and
repeat gastroscopy confirmed further ulceration at the anas-
tomosis. The patient suffered from complex chronic pain
and found relief in venting her jejunostomy. She developed
hyponatremia, hypoalbuminemia, and ongoing weight loss
with malnutrition, leading to the admission in 2015 to our
tertiary center.
Assessment, Examination, and Imaging
On initial assessment, the patient was found to be cachectic,
with her lowest weight being 38 kg with a body mass
index (BMI) of 13.15. She was wheelchair bound, later
becoming bed bound and dependent for activities of daily
living. A year history of progressive finger and toe paraes-
thesia had been noted and investigated at the original
referring hospital with nutritient testing of vitamin B12
and folate and by noncontrasted magnetic resonance image
(MRI) of the brain and spine but with no pathology
found.
Examination revealed normal tone with bilateral upper
limb and lower limb weakness that was more severe distally.
Nutrition in Clinical Practice 33(4)
In the upper limbs, power was Medical Research Council
(MRC) grade 4 of 5 (reduced power against resistance) in
finger and wrist extension and flexion and normal more
proximally. In the lower limbs, power was MRC grade 4
in extension and flexion of hip and knee joints bilaterally.
More distally, power was reduced further with MRC grade
3 (active movement against gravity, without resistance) in
dorsiflexion and plantarflexion. By February 2015, power
in the lower limbs had progressed to MRC grade 2 (active
movement but only if gravity eliminated) in hip and knee
flexion and extension and remained 3 to 4 at the ankle.
Plantars were up going, and reflexes were bilaterally brisk,
although ankle jerks were absent. Sensation was reduced to
pinpoint and light touch in a stocking-and-glove distribu-
tion to the midthighs and forearms, respectively. The first
symptoms of distal numbness had been reported in 2013,
but progression occurred during the 2015 admission to hos-
pital. Vibration sense and proprioception were reduced in
the lower limbs, and continence was maintained. Neurology
advice suggested a diagnosis of SACDSC with a peripheral
neuropathy, secondary to nutrition deficiency.
An MRI scan of the spine with T2 weighting found
characteristic findings of SACDSC with high signal changes
in the dorsal columns, represented by a charcteristic “inverse
V sign” (Figure 1).
King et al 517

Table 1. Relevant Blood Laboratory Investigations.
Laboratory Investigation
Hemoglobin, g/L
White cell count, ×109 /L
Platelets, ×109 /L
Mean corpuscular volume, fL
Mean corpuscular hemoglobin, pg
Mean corpuscular hemoglobin
concentration, g/L
Ferritin, μg/L
Vitamin B12, ng/L
Homocysteine, μmol/L
Copper, μmol/L
Selenium, μmol/L
Zinc, μmol/L
withdrawn after 1 year from commencement. Nutrition
status has since been maintained, both serologically (in-
Values (Reference Range)
cluding serum copper levels comfortably within the normal
113 (115–165) range) and in terms of a stable BMI. Neurologic assessment
8.7 (4–11) found little in the way of neurophysiologic or functional
517 (150–450) improvement; the patient is currently chair bound, having
89.6 (80–99) previously been bed bound. Spinal MRI following admis-
30 (27–33) sion, at 1 month and at 1 year, showed no significant changes
335 (315–365) to the high signal abnormalities seen in Figure 1. Her most
42 (10–320)
recent copper levels were normal at 16.2 μmol/L, and she
180 (>200) continued to receive enteral supplements in the form of
9.2 (6.7–15.2) copper amino acid chelate, 1 tablet per day (the compound is
5.9 (11–25) 20 mg, which is equivalent of 2 mg of elemental copper). In
0.46 (0.9–1.7) the longer term, Forceval has been sufficient to maintain her
8.8 (11–24) copper levels with cessation of specific copper supplements,
but she remains under close nutrition supervision.

Discussion and Literature Review

Results of Laboratory Investigations
The patient was noted to be mildly cytopenic (Table 1).
Vitamin B12 levels at this time were found to be mildly
low at 180 ng/L (reference, >200 ng/L) but had been within
the normal range and were thought to be insufficiently low
enough to cause SACDSC, following review by our hema-
tologists; furthermore, homocysteine levels were normal at
9.2 μmol/L (6.7–15.2 μmol/L).
Vitamins A, E, and B6 levels were normal; antitissue
transglutaminase antibodies and vasculitic screen yielded
negative results; and cerebrospinal fluid investigation re-
vealed normal values (white cell count, 1/cmm; glucose,
3.2 mmol/L; protein, 0.14 g/L; polymerase chain reaction,
negative). Nerve conduction studies demonstrated sen-
sory/motor axonal neuropathy. Copper levels were sent and
revealed a low level, 5.9 μmol/L (reference, 11–25 μmol/L).
Borderline-low zinc levels (8.8 μmol/L) and selenium de-
ficiency (0.46, μmol/L; Table 1) were corrected with PN.
The PN copper component consisted of copper chloride,
initially at 40 μmol per PN bag (24 hours), in the form
of 20 mL of Additrace, double the standard amounts in
24 hours of PN (10 mL or 1 vial), which was reduced to
maintenance of 1 vial (10 mL) of Additrace (in addition
to Solivito and Vitlipid), providing 20 μmol in all PN
thereafter. Inpatient PN was converted to home PN 5 nights
of the week some 4.5 months after commencement.
The predominant copper deficiency produced this in-
distinguishable syndrome of SACDSC. Copper-containing
PN and restoration of general nutrition status with a BMI
of 21.94 have led to some resolution of symptoms. Three
monthly hydroxocobalamin intramuscular injections (1 mg;
following loading doses of 1 mg on alternate days for 2
weeks) were also provided per British National Formulary
guidance. The patient is now soley fed via jejunostomy. PN
was weaned down to just 2 nights per week, then finally
Copper is absorbed from the duodenum and the stomach via
the Ctr1 apical membrane transporter. Following transport
to the liver hepatocytes, either it is utilized, or it binds
to caeruloplasmin (not measured in this case as usually
measured in diagnostic workup for copper overload) for
systemic circulation to other tissues. Copper is a vital
micronutrient required for essential enzymatic processes in
the human body, including neurotransmitter metabolism,
iron transport, oxidative phosphorylation, purine synthesis,
and connective tissue synthesis.5,6 Deficiency of ATPase
transporters from the gastrointestinal cells to the portal
circulation or the hepatocytes to bile, leads to the copper
deficiency and excess states of Menke’s disease (kinky hair
disease) and Wilson’s disease, respectively.7-9 Excess copper
will bind to metallothioneins to prevent toxic levels and
has a higher affinity for binding than does zinc. Zinc is
known to upregulate these proteins in the enteric mucosa
and is the mechanism by which copper is lost, leading to
the deficiency states seen in zinc excess.10,11 Bile secretion
achieves homeostasis, with much reabsorbed, but main
losses occur via feces and small amounts in the urine.6
Copper deficiency as a cause of cytopenia has long been
accepted,6,10 but it is only more recently that a myelo-
pathic process has been described.12 Copper is normally
seen in high levels within the central nervous system. It
is thought that the copper-deficient myelopathy, analogous
to the SACDSC seen in vitamin B12 deficiency, is related
to copper’s role in the methylation cycle.13 Methionine
synthase (a part of the cycle) may be the common disrupted
pathway of nerve protein sythesis in copper and vitamin
B12 deficiency. SACDSC is an insidious progressive combi-
nation of dorsal column signs manifesting in ataxia, with
combinations of upper and lower motor signs, classically
absent ankle jerks, and up-going plantar response,14 as seen
in this case. Vitamin B12 and copper deficiency may also
518
cause peripheral neuropathy, confirmed in this case, which
can further complicate diagnosis.
Although an underlying cause of copper deficiency
may be found—commonly, upper gastrointestinal surgery
(as contributed in this case), malabsorption, zinc excess
(sometimes as a result of denture fixative), copper chelation
through penacillamine—sometimes no cause can be found.
Plantone et al described 2 cases of women with gait ataxia
(1 mild and 1 moderate) aged 50 and 53 years with MRI
dorsal column findings similar to those seen in our case,
with low copper levels but with no cause for the copper
deficiency found. Interestingly, oral suplementation at 8
mg/d of copper led to significant clinical and radiologic
improvement.15 Kumar et al described 3 cases of copper
deficiency myelopathy; for the 1 man (72 years old, who
had undergone partial gastrectomy at age 26), MRI findings
showed hyperintense signal change on T2 weighting in
the dorsal columns. He had a vitamin B12 deficiency, but
symptoms of ascending paraesthesia, increased limb weak-
nes, and ataxia progressed, despite replacement. Copper
deficiency was eventually discovered and treated, though
clinical improvement was not commmented on. The other
2 cases were of women in their late 40s where the cause
of the copper deficiency was not clear. One patient had
MRI changes, the other did not, and both had ataxic gaits
and reduced vibration and proprioception sense. All 3 cases
had evidence of sensory/motor axonal neuropathy, not seen
in the Plantone cases but seen in our case. One woman
received parentral and then oral copper supplementation
and achieved slight clinical improvement; the other woman
had only oral supplementation (at 2 mg/d), did not achieve
normal copper levels, and her syptoms deteriorated.11
A review of 16 cases of Scottish patients with severe
copper deficiency by Gabreyes et al showed that replace-
ment of copper and removal of exccess zinc improved
blood cytopaenias, though neurologic improvement was
seen much less commonly. Causes of copper deficiency in
this retrospective study included zinc excess supplementa-
tion in the form of dental fixatives, for proximal bowel
resection, for Wilson’s disease, and from topical zinc use
in patients with burns. Two patients, whose cases were not
related to excess zinc, had celiac disease or penacillamine as
treatment for systemic sclerosis. Of the 16, 12 had neurologic
symptoms with gait disturbance, and 10 had paraesthesias.
Eight patients had spinal imaging, of which 6 had dorsal
column signs, and 2 of the 7 patients who had nerve
conduction studies had sensory/motor neuropathy. Only 3
of 12 patients had any improvement, and 4 progressed to
be wheelchair bound. All patients received 5 mg/d of oral
copper sulphate.10
Copper deficiency myelopathy has a female preponder-
ance, as demonstrated in Jaiser and Winston’s review of
55 case reports. The authors demonstrated an age range
peaking in the fifth and sixth decades, as seen in this case;
Nutrition in Clinical Practice 33(4)
47% of these cases had had upper gastrointestinal surgery.6
Kumar and colleagues’ review of Roux-en-Y gastric bypass
surgery as treatement for obesity examined symptomatic
and asymptomatic cases and described up to 10% of pa-
tients as having copper deficiency postprocedure; however,
having symptoms was rare, with only 34 cases after bariatric
surgery found in the literature.4
It is pertinent to mention that the neurologic conse-
quences of copper deficiency can occur many years after
surgery. Indeed, in 1 review, the time for neuorolgic onset
following gastric surgery was 5–26 years and 10–46 years
for bariatric and nonbariatric patients, respectively.6 In 2
case reports, men in their 70s with neurologic symptoms
secondary to copper deficiency, akin to the patient in this
case, had had gastric bypass surgery in their 20s and
30s.11,16 Copper deficiency leading to neurologic sequaelae is
being recognized more and more in the literature. Although
surgery for gastric ulceration has become less and less
common, there is increased bariatric surgery, which can
equally lead to malabsorption, and indeed many obese
patients may be copper deplete prior to surgery.17,18
The UK recommended dietry intake of copper is 1.2
mg/d, which should be easily achieved given the ubiquitous
nature of copper in many foodstuffs.19 The literature has
very different treatement suggestions, with Kumar and
colleagues’ review of bariatric surgery finding suggested
ranges after Roux-en-Y gastric bypass surgery from 0.45
to 3.0 mg/d (which includes values lower than the daily
recommendation).4 The suggestion by Jaiser and Winston is
that 8 mg/d orally is suitable to try to prevent neurologic re-
laspe/progression following deficiency-induced myelopathy;
this case series found treatment ranges from 2 mg escalating
to 4–6 mg and 8 mg, and in some cases, 9 mg/d were used.
The tolerable upper intake level is thought to be 10 mg/d.6
Optimal duration, dose, and form of copper replacement
are not clear from the literature; in this case, long-term
supplementation is likely to be required as the underlying
cause remains. The British Obesity and Metabolic Surgery
Society advises that copper be measured annually following
gastric bypass surgery as a standard, with more common
measurements if poor wound healing is seen. More common
measurements were felt unecessary, as routine supplemen-
tation at a minimum of 2 mg/d should be commenced
as recommended by American and British guidelines.4,20
In severe deficiency, parentrally administered copper at
2–4 mg/d for 6 days was advised.21
Conclusions
This case demonstrates the need for clinicians to be aware of
nutrition deficiencies and the importance of copper malab-
sorption as a cause of neurologic impairment. The literature
suggests that the neurologic deterioration secondary to cop-
per deficiency can only be stabilized or partially improved by
King et al
supplementaion; therefore, swift recognition and treatment
are paramount.6 This point is further emphasized as the
consequences of copper deficiency may not be seen until
many years after the causative insult (eg, gastric bypass
surgery).10,22
We suggest that in the context of a malabsorbative state,
in particular through upper gastrointestinal surgery com-
pounded by poor nutrition intake, vigilance for a clinical
picture of SACDSC is vital. If it is suspected, T2-weighted
MRI of the whole spine is the investigation of choice.
We also suggest that follow-up blood tests that include
copper levels are a prudent low-cost means of recognizing
copper deficiency before it has any deleterious effect, and
oral supplementation in concordance with recommended
guidelines can prevent deficiencies before they occur.
Statement of Authorship
D. King, K. Siau, and S. C. Cooper equally contributed to
the conception and design of the case report; all authors
contributed to the acquisition and analysis of the data; all
authors contributed to the interpretation of the data; and
D. King drafted the manuscript. All authors critically revised
the manuscript, read and approved the final manuscript, and
agree to be accountable for ensuring integrity and accuracy of
the work.
References
1. Lanska DJ. Chapter 30: historical aspects of the major neurological
vitamin deficiency disorders: the water-soluble B vitamins. Handb Clin
Neurol. 2010;95:445-476.
2. Gletsu-Miller N, Broderius M, Frediani JK, et al. Incidence and
prevalence of copper deficiency following Roux-en-Y gastric bypass
surgery. Int J Obes (Lond). 2012;36:328-335.
3. Ernst B, Thurnheer M, Schultes B. Copper deficiency after gastric
bypass surgery. Obesity (Silver Spring). 2009;17:1980-1981.
4. Kumar P, Hamza N, Madhok B, et al. Copper deficiency after
gastric bypass for morbid obesity: a systematic review. Obes Surg.
2016;26:1335-1342.
5. Rech M, To L, Tovbin A, Smoot T, Mlynarek M. Heavy metal in
the intensive care unit: a review of current literature on trace element
supplementation in critically ill patients. Nutr Clin Pract. 2014;29:78-
89.
6. Jaiser SR, Winston GP. Copper deficiency myelopathy. J Neurol.
2010;257:869-881.
519
7. Gollan JL, Deller DJ. Studies on the nature and excretion of biliary
copper in man. Clin Sci. 1973;44:9-15.
8. Gandhi R, Kakkar R, Rajan S, Bhangale R, Desai S. Menkes kinky
hair syndrome: a rare neurodegenerative disease. Case Rep Radiol.
2012;2012:684309.
9. Menkes JH, Alter M, Steigleder GK, Weakley DR, Sung JH. A sex-
linked recessive disorder with retardation of growth, peculiar hair, and
focal cerebral and cerebellar degeneration. Pediatrics. 1962;29:764-779.
10. Gabreyes AA, Abbasi HN, Forbes KP, McQuaker G, Duncan A,
Morrison I. Hypocupremia associated cytopenia and myelopathy: a
national retrospective review. Eur J Haematol. 2013;90:1-9.
11. Kumar N, Crum B, Petersen RC, Vernino SA, Ahlskog JE. Copper
deficiency myelopathy. Arch Neurol. 2004;61:762-766.
12. Schleper B, Stuerenburg HJ. Copper deficiency-associated myelopathy
in a 46-year-old woman. J Neurol. 2001;248:705-706.
13. Winston GP, Jaiser SR. Copper deficiency myelopathy and subacute
combined degeneration of the cord—why is the phenotype so similar?
Med Hypotheses. 2008;71:229-236.
14. Kumar P, Clark M. Kumar and Clark’s Clinical Medicine. New York,
NY: Elsevier; 2016.
15. Plantone D, Primiano G, Renna R, et al. Copper deficiency myelopa-
thy: a report of two cases. J Spinal Cord Med. 2015;38:559-562.
16. Chhetri SK, Mills RJ, Shaunak S, Emsley HC. Copper deficiency. BMJ.
2014;348:g3691.
17. de Luis DA, Pacheco D, Izaola O, Terroba MC, Cuellar L, Cabezas G.
Micronutrient status in morbidly obese women before bariatric surgery.
Surg Obes Relat Dis. 2013;9:323-327.
18. Papamargaritis D, Aasheim ET, Sampson B, le Roux CW. Copper,
selenium and zinc levels after bariatric surgery in patients recommended
to take multivitamin-mineral supplementation. J Trace Elem Med Biol.
2015;31:167-172.
19. Panel on Dietary Reference Values of the Committee on Medical
Aspects of Food Policy. Dietary reference values for food energy and
nutrients for the United Kingdom. Rep Health Soc Subj (Lond).
1991;41:1-210.
20. O’Kane M, Pinkney J, Aasheim ET, Barth JH, Batterham RL,
Welbourn R. BOMSS Guidelines on Perioperative and Postoperative
Biochemical Monitoring and Micronutrient Replacement for Patients
Undergoing Bariatric Surgery. London, UK: British Obesity and
Metabolic Surgery Society; 2014.
21. Mechanick JI, Youdim A, Jones DB, et al. Clinical practice guidelines
for the perioperative nutritional, metabolic, and nonsurgical support of
the bariatric surgery patient—2013 update: cosponsored by American
Association of Clinical Endocrinologists, The Obesity Society, and
American Society for Metabolic and Bariatric Surgery. Obesity (Silver
Spring). 2013;21:S1-S27.
22. Duddy J. Copper deficiency masquerading as subacute combined
degeneration of the cord and myelodysplastic syndrome. Adv Clin
Neurosci Rehabil. 2007;7:20-12.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Reduction of Parenteral Nutrition and Hydration Support and August 2018 520–527

C 2018 The Authors. Nutrition in

Safety With Long-Term Teduglutide Treatment in Patients Clinical Practice published by

Wiley Periodicals, Inc. on behalf
With Short Bowel Syndrome−Associated Intestinal Failure: of American Society for
Parenteral and Enteral Nutrition
STEPS-3 Study DOI: 10.1002/ncp.10092
wileyonlinelibrary.com

Douglas L. Seidner, MD, AGAF, FACG, CNSC1 ; Ken Fujioka, MD2 ;

Joseph I. Boullata, PharmD, RPh, BCNSP, FASPEN3 ; Kishore Iyer, MBBS, FRCS,
FACS4 ; Hak-Myung Lee, PhD5 ; and Thomas R. Ziegler, MD6

Abstract
Background: Patients with intestinal failure associated with short bowel syndrome (SBS–IF) require parenteral support (PS)
to maintain fluid balance or nutrition. Teduglutide (TED) reduced PS requirements in patients with SBS–IF in the random-
ized, placebo (PBO)-controlled STEPS study (NCT00798967) and its 2-year, open-label extension, STEPS-2 (NCT00930644).
Methods: STEPS-3 (NCT01560403), a 1-year, open-label extension study in patients with SBS–IF who completed STEPS-2, further
monitored the safety and efficacy of TED (0.05 mg/kg/day). Baseline was the start of TED treatment, in either STEPS or STEPS-2.
At the end of STEPS-3, patients treated with TED in both STEPS and STEPS-2 (TED–TED) received TED for ࣘ42 months, and
patients treated with TED only in STEPS-2 (no TED treatment [NT]/PBO–TED) received TED for ࣘ36 months. Results: Fourteen
patients enrolled (TED–TED, n = 5; NT/PBO–TED, n = 9) and 13 completed STEPS-3. At the last dosing visit, mean (SD) PS was
reduced from baseline by 9.8 (14.4 [50%]) and 3.9 (2.8 [48%]) L/week in TED–TED and NT/PBO–TED, respectively. Mean (SD)
PS infusions decreased by 3.0 (4.6) and 2.1 (2.2) days per week from baseline in TED–TED and NT/PBO–TED, respectively. Two
patients achieved PS independence; 2 additional patients who achieved independence in STEPS-2 maintained enteral autonomy
throughout STEPS-3. All patients reported ࣙ1 treatment-emergent adverse event (TEAE); 3 patients had TEAEs that were reported
as treatment related. No patient had a treatment-related treatment-emergent serious AE. Conclusions: Long-term TED treatment
yielded a safety profile consistent with previous studies, sustained efficacy, and a further decline in PS requirements. (Nutr Clin
Pract. 2018;33:520–527)

Keywords
glucagon-like peptide-2; intestinal failure; parenteral nutrition; short bowel syndrome; teduglutide

In adults, intestinal failure associated with short bowel
syndrome (SBS−IF) is an uncommon but life-altering con-
dition that occurs as a consequence of surgical resection
or disease with or without partial or complete colonic
resection, typically in patients with inflammatory bowel
disease, intestinal obstruction, ischemic events, fistula resec-
tion, or malignancy.1 As a result of reduction in residual
small-bowel length and absorptive function, patients with
SBS−IF experience generalized protein-energy malnutri-
tion, dehydration, and micronutrient depletion due to a loss
in intestinal absorptive capacity and often require long-term
parenteral support (PS) in the form of parenteral nutrition
and/or intravenous hydration, depending on the extent of
bowel resection and the presence of colon-in-continuity.2
Patients with SBS−IF require PS to maintain hydration and
nutrition status, including energy, protein sources (amino
acids), vitamins, trace elements, and electrolytes.1
Although PS provides vital fluid and nutrient support
for patients with SBS−IF, this therapeutic approach is
associated with potentially serious or life-threatening com-
plications (eg, risk for catheter-related infection and liver
failure).3 Patients with SBS−IF also have impaired health-
related quality of life characterized by feelings of social
isolation, a disruption of activities of daily living, and
gastrointestinal (GI) symptoms (eg, diarrhea, abdominal
pain).4-6
Teduglutide (TED) is a synthetic analog of endoge-
nous glucagon-like peptide-2, which is synthesized in the
ileum and colon of healthy individuals.7 Subcutaneous
TED administration promotes intestinal adaptation, en-
hances intestinal absorptive capacity, and enables reduc-
tion of PS in patients with SBS−IF.8 TED is approved
in the United States (adults) and Europe (adults and
children aged ࣙ1 year) for the treatment of patients
Seidner et al 521

who are PS dependent.9,10 In the pivotal, phase III,
24 week, multinational, placebo (PBO)-controlled STEPS
study (NCT00798967; EudraCT: 2008-006193-15), the pri-
mary study endpoint (ࣙ20% reduction in PS volume at
Week 20 that was maintained at Week 24) was met in
63% of adult patients treated with subcutaneous TED
0.05 mg/kg/day compared with 30% of patients treated
with PBO (P < 0.01).8 In the open-label, 2-year, multi-
national extension study (STEPS-2; NCT00930644; Eu-
draCT: 2009-011679-65), treatment with subcutaneous
TED 0.05 mg/kg/day for ࣘ24 months was well tolerated and
resulted in additional reductions in weekly PS volume and
number of days of PS infusions per week while maintain-
ing patients’ nutrition status.11 The objective of STEPS-3
(NCT01560403) was to further assess the long-term safety
and efficacy of TED in patients with SBS−IF who had
completed the STEPS-2 study.
Methods
Study Design and Population
STEPS-3 was an additional 1-year, open-label extension
of the STEPS-2 open-label TED extension study, designed
primarily to assess the long-term safety of TED in patients
with SBS−IF. In addition, efficacy outcome data using
measures from the original STEPS study were collected.
STEPS-3 was conducted in the United States, and 5 of the
7 STEPS-2 study sites participated (ie, of the 88 patients
who participated in STEPS-2, 24 patients came from the
United States [7 sites; 18 patients] and Canada [3 sites;
6 patients]); the majority were from Europe (7 countries;
64 patients). The study was conducted in accordance
with the Declaration of Helsinki and the International
Conference on Harmonisation and Good Clinical Practice,
and the study protocol was approved by institutional review
boards.
Eligible adult participants in STEPS-3 had completed
24 months of TED treatment in STEPS-2, regardless of
whether they had been weaned from PS, and they had
provided written informed consent before the initiation
of study-related procedures. All patients received TED
0.05 mg/kg/day by subcutaneous injection.
The STEPS-3 study population was stratified into
2 subgroups for the purpose of data analysis, depending
on length of TED treatment (Figure 1). The TED–TED
subgroup was composed of patients who received TED
in the initial PBO-controlled trial (STEPS) and in the
open-label STEPS-2 study. The no teduglutide treatment
(NT)/PBO–TED subgroup included patients who received
NT (entered STEPS-2 directly) or PBO in the initial PBO-
controlled trial (STEPS) and TED in STEPS-2. Compliance
with the dosing regimen, a prespecified endpoint, was
monitored by counting used and unused vials of study drug
on return patient visits and was considered to have been
achieved if ࣙ80% of planned doses were administered (ie,
[vials dispensed–vials returned intact]/days on treatment).
Baseline demographics, patient characteristics, and efficacy
parameters were based on the intent-to-treat (ITT) popula-
tion. Baseline for the parameters assessed in STEPS-3 was
defined as the start of TED treatment in the STEPS study
for patients in TED–TED or the start of TED treatment
in STEPS-2 for patients in the NT/PBO–TED subgroup
(Figure 1). As such, observations collected for the patients
who received PBO in the initial STEPS trial were not
included in the analysis of the NT/PBO–TED subgroup.
The safety analysis population included all patients who

From the 1 Vanderbilt University Medical Center, Nashville, TN, USA; 2 Scripps Clinic, La Jolla, CA, USA; 3 Hospital of the University of
Pennsylvania and Drexel University, Philadelphia, PA, USA; 4 Mount Sinai Medical Center, New York, NY, USA; 5 Shire Human Genetic
Therapies, Inc., Lexington, MA, USA; and 6 Emory University School of Medicine, Atlanta, GA, USA.
Financial disclosure: The clinical trial was funded by NPS Pharmaceuticals, Inc., Lexington, MA, USA. NPS Pharmaceuticals, Inc., is a wholly
owned indirect subsidiary of Shire. Editorial support, funded by Shire, was provided by Complete Healthcare Communications, LLC (West
Chester, PA, USA). The sponsor of the study participated in study design, data collection, data analysis, data interpretation, review, and approval
of the final report, and provided the study drug.
Conflicts of interest: D.L. Seidner received research support from and served as a study investigator, consultant, and advisory board member for
NPS Pharmaceuticals, Inc., and as a consultant for Shire; K. Fujioka served as a consultant and study investigator for NPS Pharmaceuticals, Inc.,
and as a speaker for Shire; J.I. Boullata served as a study investigator and advisory board member for NPS Pharmaceuticals, Inc.; K. Iyer served
as a consultant, study investigator, and advisory board member for and received research support from NPS Pharmaceuticals, Inc., and has served
as a consultant for Shire; H.-M. Lee is an employee of Shire; and T.R. Ziegler served as a study investigator for NPS Pharmaceuticals, Inc.
Received for publication August 18, 2017; accepted for publication March 5, 2018.
This article originally appeared online on May 15, 2018.
Corresponding Author:
Douglas L. Seidner, MD, AGAF, FACG, CNSC, Vanderbilt University Medical Center, 1211 21st Avenue South,
Ste 514 MAB, Nashville, TN 37232-2713, USA.
Email: douglas.seidner@vanderbilt.edu
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
522 Nutrition in Clinical Practice 33(4)

Data Analysis
Completed Completed
STEPS STEPS-2 STEPS-3 Descriptive statistics were used to summarize the baseline
27 global sites 25 global sites 5 U.S. sites
randomized, open-label open-label and demographic characteristics, efficacy endpoints, and
double-blind extension extension TEAEs; the study was not designed or sufficiently powered
TED 0.05 mg/kg/day TED 0.05 mg/kg/day to determine the statistical significance of safety or efficacy
24 weeks ≤24 months ≤12 months
endpoints.
TED
0.05 mg/kg/day TED–TED TED–TED Results
(n=43) (n=37) (n=5)
Of the 14 patients who completed treatment with TED
Baseline in STEPS-2 and enrolled in STEPS-3 (ITT population),
13 patients completed the study (ie, were receiving TED
PBO PBO–TED at the time the study concluded). One patient was lost
(n=43) (n=39)
to follow-up after being treated with TED for 8 months;
Baseline
NT/PBO–TED however, available data from this patient were included in
(n=9)
Direct the analysis. The confluence of the rolling start dates and
enrollment the study end date meant that all patients did not receive
NT–TED
into STEPS-2
Not treated
(n=12) 12 months of TED treatment. The mean (SD) duration
in STEPS* of exposure to TED during STEPS-3 was 38.9 (9.8) weeks
Baseline
for the overall population, 41.5 (8.4) weeks for NT/PBO–
TED, and 34.3 (11.3) weeks for TED–TED. Combined
with the TED treatment in the STEPS-2 study, the total
Figure 1. Flow diagram of patients across the STEPS studies.
NT/PBO–TED received NT or PBO in initial PBO-controlled TED exposure time was ࣘ36 months for NT/PBO–TED
trial (STEPS) and TED in STEPS-2. TED–TED received and ࣘ42 months for TED–TED. Of the 14 patients enrolled
TED in initial PBO-controlled trial (STEPS) and in STEPS-2. in STEPS-3, 8 (57%) received ࣙ80% of the planned dose:
*Patients who completed fluid optimization and stabilization 67% (n = 6/9) in the NT/PBO–TED and 40% (n = 2/5) in the
but were not randomized in STEPS because of full study TED–TED subgroups. Patient demographics and baseline
enrollment were eligible for direct enrollment into STEPS-2. characteristics are summarized in Table 1.
NT, no teduglutide treatment; PBO, placebo; TED,
teduglutide.
Efficacy
At baseline, the mean (SD) PS volume requirements were
received ࣙ1 dose of study drug in this extension study and
13.4 (11.6) and 10.5 (7.5) L/week for patients in the
was identical to the ITT population.
TED−TED and NT/PBO−TED subgroups, respectively.
From study baseline to the final dosing visit, mean (SD)
Efficacy and Safety Measures total PS volume was reduced during the STEPS-3 study
Study visits were conducted every 3 months. Prespecified period by 9.8 (14.4) and 3.9 (2.8) L/week for patients in
efficacy parameters included absolute and relative change the TED–TED and NT/PBO–TED subgroups, respectively.
from baseline in actual PS volume received, reduction in The mean (SD) percentage reduction from baseline in PS
days of PS infusions per week, and number of patients who volume was 49.7% (72.4%) for the TED−TED subgroup
achieved independence from PS in the STEPS-3 study. Pre- and 47.8% (42.9%) for the NT/PBO−TED subgroup. In ad-
specified safety parameters included duration of exposure dition, patients in the TED−TED subgroup compared with
to TED; incidence of treatment-emergent adverse events the NT/PBO−TED subgroup exhibited greater reductions
(TEAEs) and treatment-emergent serious adverse events in actual mean PS volume at all visits (Figure 2). In addition
(TESAEs); physical examinations; vital signs; electrocardio- to the mean volume reduction with time during STEPS-3
gram results; colonoscopy evaluations; clinical laboratory compared with baseline, patients reduced the frequency
testing (including serum chemistries for liver and kidney of required PS infusions. For patients in the TED–TED
biochemical values, pancreatic enzymes, and electrolytes); subgroup, the reduction from baseline in mean (SD) days
and assessment of TED-specific antibody formation. Body per week receiving PS at the last dosing study visit in
weight and body mass index (BMI) were determined as this extension study was 3.0 (4.6) days. Patients in the
post hoc analyses because of a data correction to 1 pa- NT/PBO–TED subgroup had a reduction of 2.1 (2.2) days
tient’s weight. Electrocardiograms and colonoscopies were per week receiving PS. Eight of 14 patients had a ࣙ1-day
performed at the first and final study visits; all other safety reduction in PS, and 6 of 14 patients had a ࣙ3-day
parameters were assessed at every study visit. reduction. At the completion of the STEP-3 study,
Seidner et al 523

Table 1. Demographics and Baseline Characteristics.

NT/PBO–TEDa TED–TEDb All Patients

Characteristic (n = 9) (n = 5) (n = 14)

Mean (SD) age,c years 55.9 (12.2) 55.8 (10.7) 55.9 (11.3)
Age range, n (%)
<45 years 2 (22) 1 (20) 3 (21)
45–<65 years 6 (67) 3 (60) 9 (64)
ࣙ65 years 1 (11) 1 (20) 2 (14)
Women, n (%) 6 (67) 4 (80) 10 (71)
Race, n (%)
White 7 (78) 5 (100) 12 (86)
Black 2 (22) 0 2 (14)
Ethnicity, n (%)
Not Hispanic or Latino 9 (100) 5 (100) 14 (100)
Mean (SD) body weight, kg 68.5 (14.2) 58.4 (14.2) 64.9 (14.5)
Mean (SD) BMI, kg/m2 24.4 (4.2) 21.8 (3.2) 23.5 (3.9)
Reason for resection, n (%)
Vascular disease 1 (11) 3 (60) 4 (29)
Crohn’s disease 2 (22) 0 2 (14)
Injury 1 (11) 1 (20) 2 (14)
Volvulus 1 (11) 0 1 (7)
Cancer 0 0 0
Other 4 (44) 1 (20) 5 (36)
Colon remaining, n (%) 7 (78) 3 (60) 10 (71)
Colon-in-continuity, n (%) 5 (56) 3 (60) 8 (57)
Mean (SD) percentage of colon 52.6 (39.9) 50.0 (0.0) 51.8 (32.6)
remainingd
Median (range) estimated remaining 55 (17–100) 76 (30–190) 66 (17–190)
small intestine,e cm
Stoma, n (%) 2 (22) 3 (60) 5 (36)
Mean (SD) time since start of PS 6.5 (9.1) 5.0 (3.5) 6.0 (7.4)
dependence, years

BMI, body mass index; NT, no teduglutide treatment; PBO, placebo; PS, parenteral support (parenteral nutrition and/or intravenous hydration);
TED, teduglutide.
a NT/PBO–TED received NT or PBO in initial PBO-controlled trial (STEPS) and TED in STEPS-2.
b TED–TED received TED in initial PBO-controlled trial (STEPS) and in STEPS-2.
c Age at informed consent in initial PBO-controlled trial (STEPS).
d Includes only patients with a colon (NT/PBO−TED, n = 7; TED−TED, n = 3; all patients, n = 10).
e Includes only patients with known residual small intestine length (NT/PBO−TED, n = 7; TED−TED, n = 5; all patients, n = 12).

4 patients were independent from PS. Two female patients
(aged 43 and 41 years) with no stoma, colon-in-continuity,
and baseline PS volumes of 4.5 and 4.7 L/week achieved
enteral autonomy after 126 and 130 weeks, respectively,
with TED treatment. The other 2 patients had achieved
independence from PS during the STEPS-2 study and
maintained long-term enteral autonomy in STEPS-3. These
2 patients were women (aged 61 and 60 years) with no stoma
and baseline PS volumes of 4.1 and 6.3 L/week, respectively.
Safety
TED-specific antibody-positive samples were reported in
3 of 14 patients (21%), all of whom were in the NT/PBO–
TED subgroup. In 1 of these patients, neutralizing anti-
bodies specific for TED were detected at the first visit in
STEPS-3, but the Month 3 visit sample was negative
for such antibodies. Subsequent sample results collected
at Months 6 and 9 did not detect neutralizing TED
antibodies.
TEAEs were reported by 100% of patients; most were
mild to moderate in severity and none led to study discon-
tinuation. There were no deaths during the study. There
were no reports of pancreatic-related events or GI steno-
sis/obstruction in this extension study. Among the overall
study population, the most common TEAEs were asthenic
conditions (an adverse event [AE] grouping consisting of
medically similar terms including asthenia, decreased ac-
tivity, fatigue, and lethargy) and diarrhea (Table 2). Five
patients (36%) reported ࣙ1 TESAE during the study.
None of the 14 TESAEs reported were considered by the
study investigators to be related to treatment with TED
(Table S1). There were no electrocardiogram abnormalities
reported as a TEAE or TESAE. There was 1 reported
524 Nutrition in Clinical Practice 33(4)

serious cardiac AE in a patient with bilateral pneumo-

18
Mean (SE) PS Volume, L/week

NT/PBO–TED nia and acute congestive heart failure; the events were
16 TED–TED
13.4 deemed by the investigator to be unrelated to treatment
14
with TED.
12
Colonoscopy procedures were scheduled at the first
10
10.5 7.9
STEPS-3 study baseline visit (ie, the end of STEPS-2)
8
6.7 6.2 and at the end of STEPS-3 study visit (ie, Month 12 or
6 6.3 6.0
6.0
early termination). Per the study protocol, polyps were
4
removed at the baseline study visit before the patient could
2
continue in the STEPS-3 extension study. Four of the
0
0. .0 10 patients with remnant colon had colonoscopies at both
–2
BL 3 9 2 the first visit and the end-of-study visit. Of these 4 pa-
Months tients, 1 patient had hyperplastic colon polyps at the first
study visit and colon polyps at the final study visit (no
NT/PBO–TED 9 9 8 7 4
TED–TED 5 5 5 1 1
additional histology reported), and 1 patient had a tubular
adenoma at the final study visit only (normal baseline
colonoscopy). Three of the 10 patients with remnant colon
Figure 2. Mean (SE) reduction in parenteral support.
had only the first visit colonoscopy; 1 patient had cecal
NT/PBO–TED received NT or PBO in initial PBO-controlled
trial (STEPS) and TED in STEPS-2. TED–TED received polyps with histology of tubular adenoma. The remaining
TED in initial PBO-controlled trial (STEPS) and in STEPS-2. 3 of 10 patients with remnant colon had only a single,
The table below the graph has the number of patients for each unscheduled visit colonoscopy; 1 of these patients was
study visit that corresponds to each time point. BL, baseline; initially evaluated for possible polyps, but follow-up his-
NT, no teduglutide treatment; PBO, placebo; PS, parenteral tology showed that no polyp was present. Collectively, of
support (parenteral nutrition and/or intravenous hydration); the 10 patients with remnant colon, 3 patients had ade-
TED, teduglutide.
noma or polyps based on the colonoscopy and/or histology
evaluations.
Patients maintained their nutrition status, as
demonstrated by the lack of substantial changes in
Table 2. Treatment-Emergent Adverse Events Reported in ࣙ2
Patients Sorted by Incidence in Overall Population. mean (SD) body weight (TED−TED, −3.0 [6.7] kg;
NT/PBO−TED, −0.9 [4.6] kg) or BMI (TED−TED,
Adverse Event NT/PBO–TEDa TED–TEDb −1.0 [2.4] kg/m2 ; NT/PBO−TED, −0.3 [1.6] kg/m2 ) at
Group or (n = 9) (n = 5) the last dosing visit compared with baseline (Table S2).
Preferred Term n (%), Events n (%), Events In addition, patients exhibited stable electrolyte levels
(ie, calcium, magnesium, and phosphate), as reflected in
Asthenic 2 (22), 2 1 (20), 1
conditions the absence of meaningful changes from baseline at the
Diarrhea 1 (11), 1 2 (40), 3 end of study visit. There were no clinically meaningful
Abdominal pain 2 (22), 2 0 changes in mean (SD) concentrations for blood urea
Benign neoplasms 2 (22), 2 0 nitrogen (TED−TED, 3.4 [15.7] mmol/L; NT/PBO−TED,
gastrointestinal 0.7 [2.3] mmol/L), serum creatinine (TED−TED, 23.0
Catheter sepsis 0 2 (40), 2 [61.5] μmol/L; NT/PBO−TED, −3.4 [12.7] μmol/L), serum
Cognition and 2 (22), 3 0
sodium (TED−TED, −1.6 [7.3] mmol/L; NT/PBO−TED,
attention
disorders and 0.6 [3.0] mmol/L), or urine sodium (TED−TED, 56.2
disturbances [43.2] mmol/L; NT/PBO−TED, −22.3 [64.8] mmol/L) in
Dyspnea 0 2 (40), 2 patients at the end of study visit compared with baseline.
Hypersensitivity 2 (22), 2 0 One patient in the NT/PBO–TED subgroup reported an
Viral infection 2 (22), 2 0 elevation in serum creatinine levels (126 μmol/L; upper
Weight decreased 0 2 (40), 3 limit of normal, 115 μmol/L) that returned to 116 μmol/L
NT, no teduglutide treatment; PBO, placebo; TED, teduglutide.
by the 6 month visit and remained within normal limits
a NT/PBO–TED received NT or PBO in initial PBO-controlled trial through the end of study visit; this event was considered
(STEPS) and TED in STEPS-2. to be unrelated to treatment by the site investigator. Stable
b TED–TED received TED in initial PBO-controlled trial (STEPS)
serum albumin levels were maintained in all patients,
and in STEPS-2.
and mean decreases in serum concentrations of alkaline
phosphatase (ALP), alanine aminotransferase (ALT), and
aspartate aminotransferase (AST) were reported in all
Seidner et al 525

Table 3. Change in Serum Albumin Level and Liver Enzymes From Baseline to Last Dosing Study Visit.

Liver Function Tests/Liver Enzymes NT/PBO–TEDa (n = 9) TED–TEDb (n = 5)

Serum albumin level (normal range, 32–50 g/L)

Baseline, g/L, mean (SD) 39.1 (3.2) 41.4 (4.3)
Change from baseline, g/L, mean (SD) 0.7 (3.2) −2.8 (3.6)
>10% increase from baseline at EOT, n (%) 1 (11) 0
>10% decrease from baseline at EOT, n (%) 1 (11) 1 (20)
Alkaline phosphatase (normal range, 37–147 U/L)
Baseline, U/L, mean (SD) 124.8 (50.1) 130.4 (50.8)
Change from baseline, U/L, mean (SD) −18.7 (25.7) −17.0 (39.9)
>10% increase from baseline at EOT, n (%) 0 1 (20)
>10% decrease from baseline at EOT, n (%) 4 (44) 2 (40)
ALT (normal range, 0–55 U/L)
Baseline, U/L, mean (SD) 34.7 (22.6) 24.8 (4.9)
Change from baseline, U/L, mean (SD) −4.0 (14.3) −2.6 (10.7)
>10% increase from baseline at EOT, n (%) 3 (33) 2 (40)
>10% decrease from baseline at EOT, n (%) 6 (67) 2 (40)
AST (normal range, 0–45 U/L)
Baseline, U/L, mean (SD) 31.0 (10.8) 25.4 (5.6)
Change from baseline, U/L, mean (SD) −2.2 (6.6) −1.4 (8.1)
>10% increase from baseline at EOT, n (%) 2 (22) 1 (20)
>10% decrease from baseline at EOT, n (%) 5 (56) 2 (40)
Bilirubin (normal range, 5.1–25.7 μmol/L)
Baseline, μmol/L, mean (SD) 8.7 (7.5) 8.5 (4.0)
Change from baseline, μmol/L, mean (SD) −2.5 (6.2) −3.4 (4.4)
>10% increase from baseline at EOT, n (%) 1 (11) 0
>10% decrease from baseline at EOT, n (%) 4 (44) 3 (60)
GGT (normal range, 0–30 U/L)
Baseline, U/L, mean (SD) 79.0 (56.1) 57.8 (82.4)
Change from baseline, U/L, mean (SD) −9.9 (26.7) −4.4 (11.7)
>10% increase from baseline at EOT, n (%) 3 (33) 1 (20)
>10% decrease from baseline at EOT, n (%) 5 (56) 3 (60)

ALT, alanine aminotransferase; AST, aspartate aminotransferase; EOT, end of treatment; GGT, γ -glutamyl transferase; NT, no teduglutide
treatment; PBO, placebo; TED, teduglutide.
a NT/PBO–TED received NT or PBO in initial PBO-controlled trial (STEPS) and TED in STEPS-2.
b TED–TED received TED in initial PBO-controlled trial (STEPS) and in STEPS-2.

but 2 patients at the end of study visit compared with Discussion

baseline (Table 3). Furthermore, for each liver enzyme
measured, more patients experienced a >10% decrease
between baseline and end of treatment than experienced a
>10% increase in these measures (Table 3).
Two patients had mild hepatobiliary-related laboratory
findings reported at baseline that continued throughout the
study. The increases in ALT and γ -glutamyl transferase
(GGT) reported in these 2 patients were not deemed to be
clinically significant and were not considered by the site
investigator to be related to treatment with TED. During
the study, increased serum lipase concentrations (>3 times
upper limit of normal) were reported in 3 patients, and
increased serum amylase concentrations (ࣙ350 U/L) were
reported in 1 patient, without reported clinical evidence of
pancreatitis. An additional patient had decreased amylase
levels (ࣘ15 U/L) postbaseline.
The findings from the open-label STEPS-3 extension study
provide general support for the safety, tolerability, and
clinical utility of TED in the treatment of adults with
SBS−IF for ࣙ36 months.
The safety data in patients with extended exposure to
TED indicated that the drug was well tolerated, and data
were in line with the expectations based on previous studies.
No indication of new adverse safety signals emerged in the
STEPS-3 long-term extension study. The development of
anti-TED-specific antibodies was reported in some patients,
but this finding had no apparent effect on the efficacy
of TED, as determined by PS adjustments during the
course of this study among these individuals. Although
the intestinotrophic properties of TED may account for
the favorable effects of intestinal adaptation observed in
526
patients with SBS,12 the development of GI polyps is a
previously identified safety concern.8,11 In this small open-
label, 1-year STEPS-3 study, 10 patients had remnant colon
and underwent colonoscopy procedures. In total, 3 patients
were reported to have polyps at either the first study visit
and/or the final study visit, 2 of whom had histologically
confirmed adenoma. In the preceding open-label, 2-year
STEPS-2 study, 9 patients had reported polyps with no
case of overt malignancy related to polyps reported, 5 of
whom had histologically confirmed adenoma.11,13 Collec-
tively, when the overlap in the STEPS-2 final visit and
the STEPS-3 baseline visit is considered, 10 patients were
reported to have polyps throughout the STEPS study series.
Together, these colonoscopy results support the safety-
related recommendation in the U.S. and European pre-
scribing information stating that patients receiving TED
have regular colonoscopy for signs of neoplastic growth.9,10
The ongoing global registry study for patients with SBS
(ClinicalTrials.gov identifier: NCT01990040) will provide
needed data regarding the incidence of colorectal polyps
in the SBS population overall, as well as among patients
treated with TED, during a follow-up period of at least
10 years. Any available results from routine colonoscopies
will be captured as part of the clinical outcomes data
collection registry, and may permit a more comprehensive
analysis of the potential risk of polyp formation associated
specifically with TED therapy.
Patients maintained and improved nutrition status in
the presence of PS volume reductions, as evidenced by the
reported vital signs and chemistry laboratory tests during
the course of the study. Although 2 patients experienced
episodes of decreased body weight that were reported as
TEAEs, these episodes appeared to be transient; mean body
weight and BMI remained relatively stable during the course
of the study. Furthermore, electrolyte levels remained steady
or improved during the course of the study, and mean
serum kidney biochemical values generally remained stable
between baseline and the end of the study. In addition, mean
serum albumin levels were relatively unchanged at the end
of the study compared with baseline, indicating that stable
nutrition status was maintained during the course of the
study.14,15 It is notable that stable or improved nutrition
status found in this 1-year extension study is similar to that
reported in the previous 2-year STEPS-2 study,11 further
supporting the conclusion that extended exposure to TED
is well tolerated.
Long-term PS is associated with increased risk of liver
damage, and advanced liver disease has been reported in
0%–50% of patients receiving chronic PS; the greatest risk
is in patients with ultrashort bowel (<50 cm).16,17 During
the preceding STEPS-2 study, serum concentrations of
liver enzymes either declined or remained stable during
2.0–2.5 years of TED treatment in the study group as a
Nutrition in Clinical Practice 33(4)
whole. In the subgroup of patients who experienced ࣙ50%
reductions in PS volume, liver enzymes decreased by a mean
of 6%–44% from baseline values.11 The improvement in
liver biochemical values seen in STEPS-2 continued in this
study, with mean decreases from baseline in ALP, ALT, AST,
GGT, and bilirubin levels observed at the end of treatment.
In addition, for each liver enzyme measured, more patients
experienced >10% decreases from baseline than experi-
enced >10% increases from baseline at end of treatment.
Together, the results of this study support previous findings
indicating that in patients receiving TED, chronic PS can
be reduced with corresponding improvements in liver
function.
TED has been shown to enhance intestinal absorption
in patients with SBS−IF.18 Therefore, it is important that
patients who achieve improvements in intestinal fluid ab-
sorption during treatment with TED also achieve balanced
reductions in the need for PS to minimize the risk for
fluid overload. In this study, long-term treatment with TED
for ࣘ42 months was associated with sustained efficacy and a
further decline in PS requirements in patients with SBS−IF.
It is also noteworthy that some patients who achieved
PS autonomy in STEPS-2 maintained such independence
with continuous long-term TED in STEPS-3. However, the
finding that other patients reached enteral autonomy for the
first time after approximately 2 years of treatment with TED
in STEPS-3 suggests that some patients may need longer
exposure to TED to achieve PS independence.
Limitations of the study include the open-label design
and lack of a control arm; additionally, our study popu-
lation was a small selected cohort that met inclusion re-
quirements for longer-term follow-up. This extension study,
undertaken at the end of the clinical development program
for drug approval, had a study design and a data analysis
plan that was descriptive and was not intended to be suf-
ficiently powered for statistical significance analysis. Even
with this limitation, this type of study design can provide
important information on outcomes, and as reported here
for the STEPS-3 data, observed results that were in line with
those of previous studies.8,11 These factors may limit the
applicability of the current findings to patients in the real-
world clinical setting with a broader range of disease and
clinical characteristics and require further study. Although
the ongoing global registry study will add valuable real-
world experience in broader populations of patients with
SBS, this small extension study provides additional new
data that long-term treatment with TED has a safety profile
similar to that reported in previous studies and supports
the overall conclusion that TED is well tolerated. However,
the ideal time for response was not determined and may,
in fact, be undeterminable because, as this study shows,
PS reductions can be further enhanced with continuous
treatment.
Seidner et al
Acknowledgments
The authors wish to thank all participating patients and their
families and the network of clinical investigators, study coor-
dinators, and operations staff at all participating centers for
their contributions to the entire STEPS clinical trial program.
The authors also thank Peter Nagy, MD, and Clément Olivier,
MD, of Shire International GmbH (Zug, Switzerland), for their
insights during the finalization of the manuscript.
Statement of Authorship
D. L. Seidner, K. Fujioka, J. I. Boullata, K. Iyer, and T. R.
Ziegler contributed to the conception and design of the re-
search; D. L. Seidner, K. Fujioka, J. I. Boullata, K. Iyer,
and T. R. Ziegler contributed to the acquisition, analysis,
and interpretation of the data; H.-M. Lee contributed to the
analysis and interpretation of data; D. L. Seidner, K. Fujioka,
J. I. Boullata, K. Iyer, H.-M. Lee, and T. R. Ziegler drafted the
manuscript; D. L. Seidner, K. Fujioka, J. I. Boullata, K. Iyer,
H.-M. Lee, and T. R. Ziegler critically revised the manuscript;
D. L. Seidner, K. Fujioka, J. I. Boullata, K. Iyer, H.-M. Lee,
and T. R. Ziegler agree to be fully accountable for ensuring the
integrity and accuracy of the work; and all authors read and
approved the final manuscript.
Supplementary Information
Additional supporting information may be found online in the
supporting information tab for this article.
References
1. Jeppesen PB. Spectrum of short bowel syndrome in adults: intestinal
insufficiency to intestinal failure. JPEN J Parenter Enteral Nutr.
2014;38:8S-13S.
2. Matarese LE. Nutrition and fluid optimization for patients with short
bowel syndrome. JPEN J Parenter Enteral Nutr. 2013;37:161-170.
3. Winkler MF, Smith CE. Clinical, social, and economic impacts of
home parenteral nutrition dependence in short bowel syndrome. JPEN
J Parenter Enteral Nutr. 2014;38:32S-37S.
527
4. Carlsson E, Bosaeus I, Nordgren S. Quality of life and concerns in
patients with short bowel syndrome. Clin Nutr. 2003;22:445-452.
5. Kelly DG, Tappenden KA, Winkler MF. Short bowel syndrome:
highlights of patient management, quality of life, and survival. JPEN
J Parenter Enteral Nutr. 2014;38:427-437.
6. Kalaitzakis E, Carlsson E, Josefsson A, Bosaeus I. Quality of life in
short-bowel syndrome: impact of fatigue and gastrointestinal symp-
toms. Scand J Gastroenterol. 2008;43:1057-1065.
7. Janssen P, Rotondo A, Mule F, Tack J. Review article: a comparison
of glucagon-like peptides 1 and 2. Aliment Pharmacol Ther. 2013;37:18-
36.
8. Jeppesen PB, Pertkiewicz M, Messing B, et al. Teduglutide reduces need
for parenteral support among patients with short bowel syndrome with
intestinal failure. Gastroenterology. 2012;143:1473-1481.
 R
9. GATTEX (teduglutide [rDNA origin]). Full prescribing information.
Shire-NPS Pharmaceuticals, Inc., Lexington, MA; 2016.

R
10. Revestive (teduglutide). EMA summary of product characteristics.
Shire Pharmaceuticals Ireland Limited, Dublin, Ireland; 2016.
11. Schwartz LK, O’Keefe SJ, Fujioka K, et al. Long-term teduglutide for
the treatment of patients with intestinal failure associated with short
bowel syndrome. Clin Transl Gastroenterol. 2016;7:e142.
12. Jeppesen PB, Sanguinetti EL, Buchman A, et al. Teduglutide (ALX-
0600), a dipeptidyl peptidase IV resistant glucagon-like peptide 2 ana-
logue, improves intestinal function in short bowel syndrome patients.
Gut. 2005;54:1224-1231.
13. Rudzki S, Armstrong D, Jeppesen PB, Forbes A, Lee H-M. Post hoc
analysis of polyps in 9 short bowel syndrome patients treated with
teduglutide. Clin Nutr. 2016;35(Suppl. 1):S183-S184.
14. Don BR, Kaysen G. Serum albumin: relationship to inflammation and
nutrition. Semin Dial. 2004;17:432-437.
15. Hankins J. The role of albumin in fluid and electrolyte balance. J Infus
Nurs. 2006;29:260-265.
16. Wiles A, Woodward JM. Recent advances in the management of
intestinal failure-associated liver disease. Curr Opin Clin Nutr Metab
Care. 2009;12:265-272.
17. Dibb M, Teubner A, Theis V, Shaffer J, Lal S. Review article: the
management of long-term parenteral nutrition. Aliment Pharmacol
Ther. 2013;37:587-603.
18. Jeppesen PB, Hartmann B, Thulesen J, et al. Glucagon-like peptide
2 improves nutrient absorption and nutritional status in short-bowel
patients with no colon. Gastroenterology. 2001;120:806-815.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Survey of Nutrition Management Practices in Centers August 2018 528–538

C 2017 American Society for

for Pediatric Intestinal Rehabilitation Parenteral and Enteral Nutrition

DOI: 10.1177/0884533617719670
wileyonlinelibrary.com

Anita M. Nucci, PhD, RD, LD1 ; Kipp Ellsworth, MS, RD, CSP, CNSC2 ;
Austin Michalski, RD, LD, CNSC3 ; Emily Nagel, MS, RD, CNSC4 ;
and Jackie Wessel, MEd, RD, CNSC, CSP, CLE5 ;
on behalf of the ASPEN Pediatric Intestinal Failure Section

Abstract
Background: Nutrition management of pediatric intestinal failure (IF) requires interdisciplinary coordination of parenteral
nutrition (PN) and enteral nutrition (EN) support. Nutrition strategies used by specialists in pediatric intestinal rehabilitation
to promote gut adaptation and manage complications have not been previously summarized. Methods: A practice survey was
distributed to members of the dietitian subgroup of the American Society for Parenteral and Enteral Nutrition Pediatric Intestinal
Failure Section. The survey included 24 open-ended questions related to PN and enteral feeding strategies, nutrition management
of PN-associated liver disease, and laboratory monitoring. Results: Dietitians from 14 centers completed the survey. Management
components for patients at risk for cholestasis were consistent and included fat minimization, trace element modification, avoiding
PN overfeeding, and providing EN. Parenteral amino acid solutions designed for infants/young children are used in patients <1 or
2 years of age. Trace minerals are dosed individually in 10 of 14 centers. Eleven centers prescribe a continuous infusion of breast
milk or elemental formula 1–2 weeks after resection while 3 centers determine the formula type by the extent of resection. Most
(86%) centers do not have a protocol for initiating oral/motor therapy. Laboratory panel composition varied widely by center. The
selection and frequency of use depended on clinical variables, including cholestatic status, exclusive vs partial PN dependence,
postrepletion verification vs routine monitoring, intestinal anatomy, and acuity of care. Conclusion: EN and PN management
strategies are relatively consistent among U.S. centers. Collaborative initiatives are necessary to define better practices and establish
laboratory monitoring guidelines. (Nutr Clin Pract. 2018;33:528–538)

Keywords
enteral nutrition; liver failure; parenteral nutrition; pediatrics; short bowel syndrome; surveys and questionnaires

Short bowel syndrome (SBS) may occur after massive
resection of the small bowel due to necrotizing enterocolitis
or conditions such as small bowel atresia, malrotation
with midgut volvulus, or gastroschisis and is the most
common cause of intestinal failure (IF) in children.1 IF
can also occur due to impaired intestinal absorption or
altered motility in children with microvillus atrophy or
intestinal pseudo-obstruction. Pediatric IF is a clinical
condition characterized by malabsorption, malnutrition,
and growth retardation.2-4 Management of children with
IF is complex but can be optimized by interdisciplinary
coordination of nutrition support, medical therapies, and
surgical interventions.5 Many children with IF experience
progressive adaptation of their remaining small bowel over
a few months to years with eventual independence from
parenteral nutrition (PN). However, hepatobiliary disease
is a chronic complication of long-term use of PN and is
a major contributor to the high morbidity and mortality
of children with IF.5,6 Nutrition management during the
period of intestinal adaptation includes maintenance of
fluid and electrolyte balance as well as enteral nutrition
(EN) and PN support.7 Pediatric nutrition specialists rely on
established nutrition guidelines and interventions intended
to promote adaptation of the remnant small bowel and facil-
itate weaning of PN and achievement of enteral autonomy
before complications occur.8
From 1 Georgia State University, Atlanta, Georgia, USA; 2 Children’s
Healthcare of Atlanta, Atlanta, Georgia, USA; 3 University of
Michigan, C.S. Mott Children’s Hospital, Grand Rapids, Michigan,
USA; 4 Helen DeVos Children’s Hospital, Grand Rapids, Michigan,
USA; and 5 Cincinnati Children’s Hospital Medical Center,
Cincinnati, Ohio, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on July 21, 2017.
Corresponding Author:
Anita M. Nucci, PhD, RD, LD, Department of Nutrition, Georgia
State University, PO Box 3005, Atlanta, GA 30302-3995, USA.
Email: anucci@gsu.edu
Nucci et al
Nutrition practice guidelines for the general manage-
ment of children receiving PN have existed for several
decades.9-11 In 2004, a 2-part review of intestinal reha-
bilitation (IR) and SBS management by DiBaise et al1
was published. Part I included an overview of the patho-
physiology of SBS and a discussion of luminal nutrients,
hormones, growth factors, and other substances that may
enhance adaptation. Part II provided nutrition management
guidelines for SBS with dietary guidelines by the presence or
absence of a colon, suggestions for micronutrient supple-
mentation, and a discussion of the clinical benefits of man-
agement by an interdisciplinary IR team.5 More recently,
detailed macronutrient and micronutrient recommenda-
tions for the management of children with IF have been
published. These reports include initial postoperative as well
as short-term and long-term PN and EN recommendations
and laboratory monitoring guidelines12,13 with an emphasis
on the prevention of nutrient deficiencies and complica-
tions, such as IF-associated liver disease (IFALD).14 In
addition, multiple cohort studies have reported benefits of
restricting intravenous (IV) soy-based fats as a means of pre-
venting IFALD,15 as a treatment for infants diagnosed with
IFALD,16 and after reversal of cholestasis using fish oil–
based IV emulsion.17 Although clinical nutrition guidelines
for the management of children with IF are available, most
research that has been conducted in this population has been
descriptive or semi-experimental in design. Clinical trials are
costly and often require multicenter collaboration to obtain
an adequate sample size. However, experimental studies
to establish evidence-based criteria for care are needed.12
No published review of current nutrition practice for this
population has been reported. The purpose of this study is
to (1) summarize the nutrition strategies used by physicians
and dietitians in pediatric IR to promote gut adaptation and
manage complications in children with IF and (2) compare
current nutrition practice with existing recommendations.
Methods
The Pediatric Intestinal Failure Section of the American
Society for Parenteral and Enteral Nutrition (ASPEN) was
initiated in 2014. The section includes physician, dietitian,
and pharmacy subgroups. The dietitian subgroup includes
members from 27 pediatric hospitals within the United
States and 4 international facilities. The short-term goals of
the dietitian subgroup are (1) to prepare nutrition education
materials for use by healthcare professionals and care-
giver(s) of children with IF and (2) to create a network of
practitioners with a specialty in pediatric IF for the purpose
of sharing resources and discussing nutrition practice issues.
A long-term goal of the subgroup is to develop nutrition
research questions and form multicenter research collabo-
rations. A qualitative nutrition practice survey developed by
the dietitian subgroup was electronically distributed to all
529
subgroup members between January and September 2015.
The survey included 23 primarily open-ended questions
related to IR center composition, PN and enteral feeding
strategies, nutrition management of IFALD, and laboratory
monitoring (Table 1). The study was determined to be
exempt from federal regulations by university institutional
review boards.
Results
IR Center Structure
Dietitians from 14 centers completed the survey. The cen-
ters were located in Atlanta, Georgia; Aurora, Colorado;
Baltimore, Maryland; Boston, Massachusetts; Cleveland,
Ohio; Cincinnati, Ohio; Columbus, Ohio; Grand Rapids,
Michigan; Indianapolis, Indiana; Little Rock, Arkansas;
New Haven, Connecticut; St Louis, Missouri; and Seattle,
Washington. Seven of the 14 centers (50%) have an or-
ganized interdisciplinary IR service that manages children
with IF upon discharge from the neonatal intensive care
unit (NICU). While in the NICU, children with IF are
managed by the NICU service with consultation from the
IR team. Centers without an organized IR team use both
the gastroenterology and pediatric surgery services for care
management.
PN Management
Parenteral amino acid solutions designed for infants and
young children are used in patients <1 year of age in all
but 1 center, which uses the product up to 2 years of age. A
standard parenteral amino acid solution is used thereafter.
Twelve of 14 centers (86%) include protein calories when
calculating total caloric intake. While excluding protein
calories when calculating total caloric intake has not
been a standard of care for many years, we assessed
whether dietitians working in IR centers still follow this
practice. The general practice of dosing trace elements and
cycling/tapering of PN solutions is shown in Table 2. Eleven
of the 14 centers (79%) reported monitoring serum levels
of trace elements as a result of reduced intake, specifically
copper, selenium, and zinc. Dietitians from 10 centers
reported that the decision to initiate iron supplementation
is dependent upon laboratory determination of deficiency.
The general consensus was that enteral iron should be
provided when there is a sufficient length of proximal small
bowel and tolerance to enteral feedings is established. IV
iron is selected in the case of an extremely short bowel
and feeding intolerance where PN is the primary source of
nutrition. PN support standards of care did not significantly
differ for children who receive long-term PN (>1 month)
at the centers surveyed. However, 4 center dietitians (29%)
noted that laboratory monitoring differed by the length
of time on PN with the inclusion of parameters for
530 Nutrition in Clinical Practice 33(4)

Table 1. Pediatric Intestinal Failure Nutrition Practice Survey Questions.

Number Question

1 What medical services/teams manage your patients with IF?

2 What is your standard protein source for PN?
3 Do you use the same protein source for all patients (yes/no)?
4 If you use the same protein source for all patients, under what circumstances would you use a specialized
parenteral amino acid solution?
5 Do you use a trace mineral package or dose minerals individually for patients on PN?
6 For patients with IF on PN, when would you provide supplemental iron?
7 If supplemental iron is provided, how is it given and under what circumstances would each method be chosen?
8 Do you have any protocols for patients receiving long-term (>1 month) PN without EN intake?
9 Do you include protein calories when calculating the caloric intake of patients on PN?
10 What is your practice regarding cycling PN?
11 What is your practice regarding tapering of PN?
12 If you are caring for a patient with IF who is at risk for cholestasis, what is your general philosophy for
nutrition management?
13 If you are caring for a patient with IF who has cholestasis, what is your standard practice for nutrition
management?
14 If you are caring for a patient with cholestasis, what is your institution policy with regard to the provision of
parenteral trace minerals?
15 How many days are your patients with IF typically ordered “nothing by mouth” status post resection?
16 What is your formula of choice for children with IF status postresection?
17 What is your typical mode of feeding after resection?
18 What oral supplements do you typically recommend for patients with IF?
19 Do you have a protocol for starting oral/motor therapy for oral feedings after resection?
20 How soon are oral feedings initiated in children with IF at your center?
21 Do you use urine sodium to determine whether a patient with IF requires sodium supplementation?
22 What laboratory tests do you routinely follow in patients with IF on PN? Include macronutrients and
micronutrients and interval for follow-up (eg, daily, weekly, monthly).
23 Which of the following parameters do you follow in patients receiving PN?
a. Carbohydrate infusion rate (select: mg/kg/min or mg/kg/h)
b. Serum triglyceride
c. Serum cholesterol
d. Triene/tetraene ratio
e. Nonprotein nitrogen/calorie ratio
f. Other:

EN, enteral nutrition; IF, intestinal failure; PN, parenteral nutrition.

trace elements and essential fatty acid deficiency to the
monitoring schedule for those on long-term PN.
Enteral/Oral Management
Eight of the center dietitians (57%) reported a length of
time that children with IF are typically unable to receive
enteral/oral feeding after initial surgery (range, 5 days to 2
weeks). The type of EN and oral supplementation provided
to children with SBS after resection varied between centers
(Table 3). Most centers (n = 13) prefer the administration of
continuous feeds via nasogastric tube (NG) or gastrostomy
tube (GT) postresection, with 2 also recommending small
oral feedings as tolerated. One center prefers to provide
oral feeding in the absence of feeding intolerance and if
the child is developmentally ready for oral intake. Vitamin,
trace element, and fiber supplements are the products most
commonly prescribed.
Most centers (n = 12) do not have a formal protocol
for the initiation of oral feedings after small bowel resec-
tion. Nine of these centers do consult with occupational
or speech therapy for the purpose of encouraging age-
appropriate complementary food intake when the child is
developmentally ready. Two centers reported that they use a
protocol for starting oral intake. Of these, 1 center begins
with nonnutritive oral feedings of human milk or infant
formula and advances to small-volume oral boluses when
trophic feedings are tolerated. This center also initiates solid
foods at 6 months of age beginning with vegetables followed
by bananas, grains, meats, and fruit. The second center
begins with 2 oral feedings per day under the supervision of
an occupational or speech therapist when enteral feedings
have reached 3 mL/h. The timing of initiation of oral feed-
ings varies by center. Most centers did not report an exact
time point for the initiation of feedings but all provided
qualifications for initiation of oral feedings (Table 3).
Nucci et al 531

Table 2. Practice Responses to Questions Related to Parenteral Nutrition Management.

Number of Centers
Survey Question Response Rate Practice Response per Response

Do you use a trace element 14 of 14 r Trace element package only 0

package or dose minerals r Individual dosing 9
individually for patients on r Both trace element package and individual 5
PN? dosing are used

What is your practice regarding 11 of 14 Cycling criteriaa

cycling of PN? r Medical stability 6
r Minimum 3 months corrected age 3
r Weight >5 kg 1
r Intake of >50 kcal/kg/d from EN 1
r Tolerance of continuous enteral feeding 1
14 of 14 Cycling strategy
r 1- to 2-hour increments 3
r 2- to 4-hour increments 5
r No more than 4 hours off PN (infants) 1
r No specific strategy reported 5

What is your practice regarding 14 of 14 Tapering strategy

tapering of PN? r minutes up and down
1
r 30 minutes up and 1 hour down
1
r 1 hour up and down
8
r 2 hours down only
1
r hours up and down
1
r No specific strategy reported
2

EN, enteral nutrition; PN, parenteral nutrition.

a Centers may have reported >1 response.

Prevention and Management of IFALD
The center dietitians were queried regarding general
philosophies toward the nutrition management of patients
at risk for IFALD. Limiting IV fat was a major component
of the nutrition management of children with IF at risk
for developing cholestasis at all centers (Table 4). Among
the centers employing some form of fat restriction, typical
dosing was limited to 1–2 g/kg/d. The second major theme
among the centers participating in the survey was providing
EN in the form of trophic feedings or aggressive advance-
ment of feedings. Monitoring parameters included glucose
infusion rate (GIR), triene/tetraene ratio, trace elements,
and weight gain and linear growth. The center dietitians
were also asked to describe their standard practice for the
nutrition care of a child with IF as well as IFALD. Fat
minimization or restriction is the most common element of
nutrition support practice among all centers included in the
survey (Table 4). Each center typically limits IV fat to 1
g/kg/d in patients with cholestasis. However, some centers
(n = 3) reduce fat dose to <1 g/kg/d either by providing
a restricted daily dose or providing 1 g/kg 2 or 3 days
per week. Four dietitians reported that trace elements were
dosed individually at their centers for children with IFALD.
Two centers provide a reduced amount of copper and 3
centers remove manganese entirely from the PN solution.
Laboratory/Biochemical Monitoring
An overview of the IF center dietitian practice survey
responses to questions related to laboratory/biochemical
monitoring in children with IF is shown in Table 5. Most
centers (93%) use urine sodium to determine whether a
patient requires sodium supplementation. Dietitians from
all centers listed a variety of panels, vitamins, trace ele-
ments, and miscellaneous laboratories whose selection and
frequency of use appeared to depend largely on a patient’s
clinical variables such as noncholestatic vs cholestatic status,
exclusive vs partial PN dependence, postrepletion verifica-
tion vs routine monitoring, intestinal anatomy, and acuity
of care (ie, NICU vs general floor vs outpatient). Dietitians
from all centers use the GIR (mg/kg/min) as a clinical
parameter for patients receiving PN.
Discussion
While the provision of PN has resulted in a reduced
mortality rate for infants and children with IF, serious and
532 Nutrition in Clinical Practice 33(4)

Table 3. Practice Responses to Questions Related to Enteral/Oral Nutrition Management.

Response Number of Centers

Survey Question Rate Practice Response per Response

What is your formula of choice 14 of 14 r Human milk or elemental infant formula 8

for children with intestinal r Amino acid–based infant formula 3
failure status postresection? r Selection based on the extent of small bowel resectiona 3

What oral supplements do you 14 of 14 r Vitamin and trace element supplements 13

typically recommend for r Carbohydrate/fat or fat modular 4
patients with intestinal r Fiber supplements
failure?b ◦ Green beans 2
◦ Pectin 3
◦ Other fiber supplements 3

How soon are oral feedings 14 of 14 r As soon as possible 4

initiated in children with r Depends on stool/ostomy output 2
intestinal failure at your r Varies individually 2
center? r Within 7–10 days 1
r Medically stable for 2 weeks on EN 1
r Age appropriate (>37 weeks) 1
r Developmentally appropriate; 2–3 days after EN initiated 1
r After EN is initiated 1
r After EN tolerated at 5–10 mL/h 1

EN, enteral nutrition.

a Human milk or an elemental infant formula with significant resection; polymeric infant formula with minimal resection.
b Centers may have reported >1 response.

potentially life-threatening complications may still occur as
a result of long-term use of PN. Growth failure continues
to be observed in a high percentage of children with IF,18
the risk of nutrient deficiencies has been shown to increase
during the transition from PN to EN,19-21 and multiple
years of PN therapy can result in the development of
IFALD.14 The nutrition care of children with IF should
be individualized based upon the intestinal anatomy and
medical condition of each patient. The EN and PN man-
agement strategies reported by the pediatric IF practitioners
who participated in this survey are relatively consistent.
Only half of the dietitians surveyed function as part of an
organized IR program. The strategies currently being used
are designed to support intestinal adaptation and reverse
or slow the progression of complications. However, clinical
and laboratory monitoring practices vary considerably be-
tween centers. Collaborative initiatives such as the Pediatric
Intestinal Failure Section of ASPEN are ideally suited to
define better practices and establish monitoring guidelines
through multicenter research efforts.
PN is an established treatment for children with IF
that must be carefully managed to prevent or delay the
development of PN-associated complications.22 Methods
currently recommended for the prevention of IFALD in-
clude avoiding overfeeding, providing a balanced distri-
bution of macronutrients in the PN solution, cycling the
PN infusion, and providing EN as tolerated.23 All of the
dietitians surveyed reported that cycling of the PN solution
is a treatment goal in their practices and that advancement
(reduction in PN infusion hours) is determined by blood
glucose tolerance. Providing a dextrose solution over fewer
hours may result in a GIR greater than what is currently
recommended for children without cholestasis (<12–14
mg/kg/min).23 Therefore, monitoring for hyperglycemia or
hypoglycemia is essential during the cycled infusion and
while the infusion is off. Abrupt cessation of PN has been
reported to be well tolerated in children >2 years of age.24
However, many children with IF are <1 year of age at the
time of diagnosis.25 Tapering of the dextrose solution over
30–60 minutes before discontinuation may prevent rebound
hypoglycemia and is a common practice reported by the
dietitians who completed our survey.
Most centers provide human milk or amino acid–based
infant formula as the primary source of nutrition for chil-
dren with SBS. This practice aligns with recommendations
from the current literature. In a retrospective medical record
review of 30 neonates with SBS, Andorsky et al26 reported
that the use of either human milk or an amino acid–based
formula was associated with a shorter duration of PN.
Human milk contains long-chain fatty acids, free amino
acids such as glutamine, and growth factors that paired
with immune-enhancing properties may assist in intestinal
adaptation. An amino acid–based formula may decrease
the duration of PN since it contains a high percentage of
Nucci et al 533

Table 4. Practice Responses to Questions Related to Prevention and Management of Intestinal Failure–Associated Liver Disease.

Number of Centers
Survey Question Response Rate Practice Response per Response

If you are caring for a patient 14 of 14 r IV fat restriction 12

with short bowel syndrome r Initiate/advance EN 9
who is at risk for cholestasis, r Cycle PN 4
what is your general r Avoid overfeeding 4
philosophy for nutrition r Dose trace elements individually 2
management?a r Wean PN 2
r Monitor serum levels of trace elements 2
r Provide parenteral protein at age-appropriate 1
level

If you are caring for a patient 14 of 14 r IV fat restriction 12

with short bowel syndrome r Initiate/advance enteral nutrition 9
who has cholestasis, what is r Cycle PN 6
your standard practice for r Dose trace elements individually 4
nutrition management? r Provide fish oil–based IV fat emulsion 4
r Monitor serum levels of trace elements 4
r Avoid overfeeding 2
r Monitor for essential fatty acid deficiency 2
(triene/tetraene ratio)
r Limit parenteral protein dose 2
r Do not exceed maximum recommended glu- 1
cose infusion rate
r Provide enteral formula with higher MCT oil 1
content

EN, enteral nutrition; IV, intravenous; MCT, medium-chain triglyceride; PN, parenteral nutrition.
a Centers may have reported >1 response.

long-chain fatty acids, which may promote mucosal
adaptation.27 Although long-chain fatty acids promote
adaptation and provide a source of essential fatty acids,
medium-chain triglycerides are advantageous because they
can be directly absorbed by enterocytes. Most amino
acid–based formulas contain a mixture of long-chain and
medium-chain triglycerides, which may explain why most
centers prefer these formulas if human milk is not available.
Some researchers argue that there is insufficient evidence to
support 1 type of formula over another for patients with
SBS. Ksiazyk et al28 conducted a prospective, randomized,
crossover, double-blind study of 10 children receiving infant
formula containing either hydrolyzed or intact protein.
Children were trialed on each formula for 30 days (crossover
on day 31), and nitrogen balance and lactulose/mannitol
excretion ratio (to measure intestinal permeability) were
evaluated on days 1, 31, and 61 of the study. The researchers
reported that the 2 types of formula did not yield a sig-
nificant difference in intestinal permeability, caloric/energy
intake, weight gain, or nitrogen balance. However, feeding
tolerance was not assessed, and all infants were receiving
30% of calorie intake from hydrolyzed enteral feeds before
beginning the study protocol. In a review of published data
relevant to feeding strategies in children with SBS, Olieman
et al29 recommended the use of human milk or standard
polymeric formula depending on the age of the child.
Most centers initiate continuous feedings postresection
since they are believed to be better tolerated than bolus
feeds and promote adaptation via continual saturation of
the intestinal lumen. However, limited research evidence
exists on the topic. A frequently cited study to support the
use of continuous feeds was conducted by Parker et al30
over 30 years ago. The researchers reported that continuous
feedings yielded a significant increase in body weight and
enteral balance (enteral retention of formula and nutrients)
compared with oral bolus feeds. Unfortunately, the sample
size was very small (n = 9), and only 2 of the infants had
SBS. A more recent randomized crossover study of nutrient
absorption by mode of nutrition therapy (tube feeding
alone, oral feeding alone, combination of tube and oral
feeding) was conducted in adults with SBS (n = 15). Results
revealed significantly greater absorption of protein, fats, and
energy when a tube feeding was used alone or in conjunction
with oral feeding compared with when oral feeding was used
alone.31 Although continuous feedings may be better tol-
erated and improve nutrient absorption, bolus feedings are
desirable as they mimic the typical infant eating pattern and
are therefore more physiological. Bolus feedings allow for
534 Nutrition in Clinical Practice 33(4)

Table 5. Practice Responses to Questions Related to Laboratory/Biochemical Monitoring.

Number of Centers
Survey Question Response Rate Practice Response per Response

Do you use urine sodium to determine 14 of 14 r Yes 10

whether a patient with short bowel r Yes, occasionally 2
syndrome requires sodium r Yes, as well as serum osmolarity 1
supplementation? r No 1

What laboratory tests do you routinely
follow in patients on PN with short
bowel syndrome or intestinal failure?
Include macronutrients and
micronutrients and interval for
follow-up (eg, daily, weekly,
monthly).a
13 of 14b Panels (daily to monthly): CBC, CMP, Varying responses
BMP by 13 centers
Vitamins (every 3–12 months): serum
retinol, retinol binding protein, RBC
folate, vitamin B-12, methylmalonic
acid, vitamin D, vitamin E
Trace elements (every 3–12 months):
serum copper, serum zinc, selenium
(RBC and serum), serum iron, total
iron binding capacity, ferritin
Other (every 3–12 months): magnesium,
phosphorous, serum triglycerides, C-
reactive protein, prothrombin time, in-
ternational normalized ratio, thyroid-
stimulating hormone, serum citrulline,
triene/tetraene ratio

Which of the following parameters do 13 of 14b r Glucose infusion rate 14

you monitor in patients receiving r Serum triglyceride 13
PN?a r Serum cholesterol 4
r Triene/tetraene ratio 12
A. Glucose infusion rate r 5
Nonprotein calorie/nitrogen ratio
B. Serum triglyceride
C. Serum cholesterol
D. Triene/tetraene ratio
E. Nonprotein calorie/nitrogen ratio

BMP, basic metabolic panel; CBC, complete blood count; CMP, comprehensive metabolic panel; PN, parenteral nutrition; RBC, red blood cell.
a Centers may have reported >1 response.
b Question was not answered by 1 center.

periods of fasting, which may prevent hyperinsulinemia32
and promote appropriate bacterial clearance.33
Vitamins and trace elements were the supplements most
often reported to be prescribed. Several dietitians also
reported use of a carbohydrate and fat modular product,
purees, and pectin/fiber. Vegetable oils such as canola,
safflower, or flaxseed can be added to tube feedings to
increase caloric intake or supplement essential fatty acids
without adding significant volume. Fiber is often added to
feedings to slow bowel transit time and reduce stool output,
which assists with nutrient absorption and subsequently
growth.34 Soluble fiber is beneficial because it is fermented
into short-chain fatty acids, which can be metabolized by
colonocytes for energy.35 In a 2010 practice paper published
in the Journal of the Academy of Nutrition and Dietetics,
the authors noted that dietary fiber should be used when
developmentally appropriate (age 4–6 months) in the pres-
ence of a colon.29 Pectin, a water soluble dietary fiber,
may slow the motility of gastric contents and increase the
contact of nutrients with the intestinal lumen. However,
the dose used (1%–3% solution) is based on a single case
series of 2 formerly premature infants who had small bowel
resections.36 Drenckpohl et al37 documented the benefits of
using stage 2 baby food green beans to infant feedings in a
case series. After the addition of green beans to formula (4-
oz jar of green beans per 8-oz formula), stool consistency
improved in all of the infants. The researchers concluded
that the mixture of 32% soluble and 68% insoluble dietary
fiber present in the green beans (2–3 g fiber per 4-oz jar)
assisted in improving stool consistency.
While most dietitians did not report using a protocol for
initiating oral/motor therapy, most centers do recommend
therapy when appropriate. Oral aversion is common in
patients who receive nutrition support for a prolonged
Nucci et al
period of time. Consequently, oral motor therapy and early
oral feedings are important to the rehabilitation process.
These therapies may also provide beneficial support for
families who become frustrated with lack of progress with
oral feeds.38 Although most centers did not report an exact
time for initiation of complementary oral feedings, some
use criteria to assess readiness or generally stated that they
would begin oral feeds as soon as possible. The initiation
of oral feeds as soon as developmentally and medically
appropriate is important to prevent oral aversion.38 The
first year of life is the most critical period for the de-
velopment of oral feeding skills.33 Oral feeding promotes
epidermal growth factor release from salivary glands and
increases gastrointestinal secretion of trophic factors.33
Small bolus feedings by mouth (equal to or less than the
volume tolerated continuously per hour) should be trialed
to encourage development of swallowing. Solid food should
be introduced at a developmentally appropriate age under
the guidance of a feeding therapist.12 Unfortunately, no
clinical studies have been conducted to identify the preferred
time of initiation of complementary solid foods or the
optimal oral diet for children with SBS.29 Current nutrition
recommendations are subdivided by colon status. A high-
carbohydrate (50%–60% of energy), lower fat (20%–30%
of energy) diet may be indicated for patients with a colon
while a higher fat (30%–40% of energy), lower carbohydrate
(40%–50% of energy) diet may be preferred when the colon
is absent.5 Complex carbohydrates may be better tolerated
than simple sugars, and patients with a colon may need to
avoid or restrict oxalate intake to decrease the risk of oxalate
renal stones.5
Several components of nutrition management for chil-
dren at risk for IFALD were consistent among all centers
and were similar if not the same as those principles used
for patients at risk for developing cholestasis. The literature
related to minimization of soy-based fat emulsion (ࣘ1
g/kg/d) is primarily observational. Although restriction of
IV fat emulsion has been shown to reduce total bilirubin
levels in children with IFALD,16 the effect of the accom-
panying caloric reduction on growth in the IF population
is unknown.14 Levit et al39 conducted a clinical trial that
examined the safety and efficacy in preventing IFALD using
a low-dose soy-based fat emulsion (1 g/kg/d) vs a control
dose (ࣈ3 g/kg/d) in preterm infants (n = 136) after 14
days of PN. The authors reported that after 28 days of
life, no difference in growth or reduction of cholestasis
was observed between the randomized groups. However,
this was not a surgical population requiring prolonged PN
support, and the determination of cholestasis (ࣙ15% of the
total bilirubin at 28 days of life) differed from the definition
used in other fat minimization studies (conjugated bilirubin
>2 mg/dL).15,16 Essential fatty acid deficiency was observed
in infants with IFALD who were placed on a soy-based fat
emulsion reduction protocol (1 g/kg/d twice weekly) for a
535
mean of 50 days (range, 11–712 days).16 Deficiencies were
mild and resolved within 1–2 months after the number of
days of fat infusion were increased. The authors noted
the importance of monitoring markers of essential fatty
acid deficiency during fat minimization and suggested that
absolute values of linoleic and α-linolenic acid be examined
vs the triene/tetraene ratio, which is frequently used.16 Given
the importance of dietary fat on infant brain development,
there is concern about the effect of fat minimization on
the long-term neurodevelopment of infants with IFALD
treated with low-dose IV fat. Few studies have assessed this
risk. However, preliminary findings at 1 institution showed
that infants treated with IV fat emulsion reduction had
mostly normal neurodevelopmental outcomes at ages rang-
ing from 2–5 years.40 Another institution recently published
data showing no adverse effect of low-dose IV soy-based
fat emulsion on neurodevelopment or growth at 2 years of
age in a cohort of 15 premature infants compared with 15
premature infants who received standard dosing.41
The use of ω-3 or fish oil–based fat emulsions is associ-
ated with resolution of biochemical markers of IFALD42-44
but lacks rigorous clinical studies in the pediatric IF pop-
ulation. Puder et al45 conducted a trial of fish oil–based
fat emulsion in 42 infants with SBS and cholestasis. Of
the 38 patients who survived or were not transplanted, 19
(50%) experienced reversal of cholestasis (direct bilirubin
ࣘ2 mg/dL). The fish oil–based fat emulsions are thought
to improve liver function due to the anti-inflammatory
properties of the ω-3 fatty acids and improved triglyceride
clearance.42 In addition, they lack the phytosterols of soy-
based fats, which are thought to be relatively hepatotoxic.42
However, there are reports of continued liver disease de-
spite resolution of cholestasis,46-48 and similar reductions
in cholestasis are achieved by restricting traditional soy-
based IV fats. Fish oil–based fat emulsions are not cur-
rently approved by the Food and Drug Administration
(FDA) for use in the United States but remain available
for use under research or compassionate use protocols.
Alternatively, Smoflipid (Fresenius Kabi, Bad Homburg,
Germany), an IV fat emulsion that contains soybean oil,
medium-chain triglycerides, olive oil, and fish oil, is FDA
approved for use in adults. Although the safety and efficacy
of using Smoflipid in pediatric patients have not yet been
established,49 randomized controlled trials have examined
Smoflipid vs soy-based fat emulsion with encouraging re-
sults, including decreased serum γ -glutamyl transferase50
and total bilirubin51 in the Smoflipid groups. In a small
retrospective review, conjugated bilirubin decreased within
2 weeks after a change to Smoflipid from a soy-based fat
emulsion in children with IFALD.52
Reducing or removing trace elements from PN solutions
is a proposed strategy to decrease the risk of IFALD.13 Since
copper is eliminated via biliary excretion, parenteral intake
of copper may need to be reduced when IFALD is present,
536
while increased provision of copper may be necessary in
patients with a jejunostomy or other biliary output.10,53 Of
note, copper deficiency has been reported with copper-free
PN.53 Recent studies examining the practice of reducing
or removing copper from PN solutions of cholestatic pa-
tients have shown that a standard intake of 20 mcg/kg/d
for infants may be safe, and it is suggested that copper
levels be assessed prior to reducing or removing copper
from PN solutions.54-56 Manganese supplementation may
not be necessary for children with IFALD as manganese
is a contaminant in PN solutions.38 Therefore, use of
individual trace element preparations vs a packaged trace
element preparation may be the best practice in patients
with IFALD.10,38 Careful monitoring of trace element status
when micronutrients are removed from PN is warranted.
Provision of EN in some form continues to be a sig-
nificant component of the nutrition support practice for
patients with IFALD. Expert opinion based on experience
and available research seems to focus on advancing enteral
feedings to promote intestinal adaptation and weaning PN
support while maintaining proportional growth. EN is be-
lieved to promote intestinal adaptation by stimulating mu-
cosal hyperplasia, gastrointestinal hormone secretion, and
pancreaticobiliary secretions.1 These mechanisms may also
protect against the development of IFALD by improving
biliary flow and decreasing the risk of bacterial overgrowth,
which may further inhibit bile secretions.23 Most experts
agree on starting EN as early as medically feasible and
advancing slowly while taking advantage of available strate-
gies as previously discussed, including continuous feeds,
refeeding of ostomy output into distal mucous fistula, and
fiber supplementation.13,22,38
Our survey revealed considerable variability in labo-
ratory test selection and monitoring frequency among
pediatric IF practitioners. Most dietitians responding to
the survey indicated the use of specific laboratory panels
at defined intervals to address the reported high preva-
lence of micronutrient deficiencies in patients with pedi-
atric IF.4,19,21,57 The laboratory monitoring practices of
most centers are generally reflective of the evidence-based
literature. Most centers routinely monitor urine sodium
due to the known association of growth failure with a
low urine sodium level, particularly in those patients with
jejunostomies, ileostomies, and colostomies.58 Many centers
have laboratory monitoring protocols, similar to published
recommendations from children’s hospitals with established
IR centers, to address the high risk of micronutrient de-
ficiency in their patient populations.12,57 Although infre-
quently mentioned by survey respondents, the monitoring
of iodine status in pediatric IF patients has garnered much
attention in the literature since the submission of survey
results, particularly in light of numerous studies demon-
strating the potential increased risk of iodine deficiency in
pediatric patients requiring long-term PN supplementation.
Nutrition in Clinical Practice 33(4)
Two case studies and 1 retrospective cohort analysis re-
vealed varied use of iodine biomarkers (thyroid-stimulating
hormone, T4, serum vs urine iodine) and varied frequency
of monitoring59-61 in investigating iodine deficiency in their
chronic PN pediatric patients. The high variability in labora-
tory test selection and monitoring frequency may be reflec-
tive of a lack of national guidelines. Our survey has revealed
the importance of continual collaboration among centers to
establish biochemical monitoring standards, particularly in
the absence of the large numbers of subjects needed for well-
executed clinical trials in this subspecialty area of practice.
Areas in which collaboration may be beneficial include the
following:
r Designating the relevant micronutrients for monitor-
ing and the most reliable indicators (eg, erythrocyte
vs plasma, serum vitamin B-12 or serum methyl-
malonic acid or both)
r Establishing the monitoring frequency and indi-
cations for any protocol variance (eg, intestinal
anatomy, liver function, PN dependent vs nondepen-
dent)
r Facilitating monitoring across the inpatient/home
health spectrum of care, particularly in regard to
the management of parenteral trace element product
shortages62
The design of the practice survey has the advantage of
permitting collection of data that might not be captured in
a closed-ended questionnaire. However, this type of study
survey also has several limitations. Questions are subject
to individual interpretation (eg, what supplements do you
typically recommend for patients with IF?), and specific
components may not have been reported due to the open-
ended format (eg, what laboratory tests do you routinely
follow in patients with IF on PN?).
Conclusion
EN and PN management strategies in the presence and
absence of cholestasis are relatively consistent among U.S.
practitioners caring for children with IF. However, labora-
tory monitoring practices vary widely among centers. Al-
though only half of the dietitians surveyed function as part
of an organized IR program, interdisciplinary management
and early referral of children dependent on PN to an IR pro-
gram can result in positive outcomes, including cessation of
PN support, accelerated growth, and improved survival.10,14
Multicenter clinical trials to identify best nutrition practices
and to establish laboratory monitoring guidelines across
the spectrum of care for children with IF would be opti-
mal. Collaboration among centers should be encouraged
considering the difficulty of conducting large-scale clinical
trials in this population. Collaborative initiatives such as the
Pediatric Intestinal Failure Section of ASPEN appear well
Nucci et al
positioned to pragmatically address the myriad variables
affecting micronutrient status in pediatric IF patients and
to issue evidence-based laboratory monitoring guidelines.
Moreover, the section can foster collaboration and take an
active role in conducting clinical research to determine best
practices for enhancing intestinal adaptation and preventing
or delaying nutrition support-related complications.
Statement of Authorship
A. M. Nucci contributed to the conception and design of the
research, as well as the acquisition, analysis, and interpretation
of the data. K. Ellsworth, A. Michalski, E. Nagel, and J.
Wessel equally contributed to the design of the research. K.
Ellsworth, A. Michalski, and E. Nagel contributed to the
analysis and interpretation of the data. J. Wessel contributed to
the interpretation of the data. A. M. Nucci, K. Ellsworth, A.
Michalski, and E. Nagel drafted the manuscript. All authors
critically revised the manuscript, gave final approval of the
manuscript, and agree to be fully accountable for ensuring the
integrity and accuracy of the work.
References
1. DiBaise JK, Young RJ, Vanderhoof JA. Intestinal rehabilitation and
the short bowel syndrome: part 1. Am J Gastroenterol. 2004;99:1386-
1395.
2. Warner BW, Ziegler MM. Management of the short bowel syndrome
in the pediatric population. Pediatr Clin North Am. 1993;40:1335-1350.
3. Barksdale EM, Stanford A. The surgical management of short bowel
syndrome. Curr Gastroenterol Rep. 2002;4:229-237.
4. Wu J, Tang Q, Feng Y, et al. Nutrition assessment in children with short
bowel syndrome weaned off parenteral nutrition: a long-term follow-up
study. J Pediatr Surg. 2007;42:1372-1376.
5. DiBaise JK, Young RJ, Vanderhoof JA. Intestinal rehabilitation and
the short bowel syndrome: Part 2. Am J Gastroenterol. 2004;99:1823-
1832.
6. Duro D, Kamin D, Duggan C. Overview of pediatric short bowel
syndrome. J Pediatr Gastroenterol Nutr. 2008;47:S33-S36.
7. Vanderhoof JA, Young RJ, Thompson JS. New and emerging thera-
pies for short bowel syndrome in children. Pediatr Drugs. 2003;5(8):
525-531.
8. Goulet O, Sauvat F. Short bowel syndrome and intestinal transplanta-
tion in children. Curr Opin Clin Nutr Metab Care. 2006;9:304-313.
9. Kerner JA, ed. Manual of Pediatric Parenteral Nutrition. New York,
NY: John Wiley; 1983.
10. Greene HL, Hambidge KM, Schanler R, Tsang RC. Guidelines for the
use of vitamins, trace elements, calcium, magnesium, and phosphorus
in infants and children receiving total parenteral nutrition: report of
the Subcommittee on Pediatric Parenteral Nutrient Requirements from
the Committee on Clinical Practice Issues of the American Society for
Clinical Nutrition. Am J Clin Nutr. 1988;48:1324-1342.
11. Corkins MR, ed. The ASPEN Pediatric Nutrition Support Core Cur-
riculum. 2nd ed. Silver Spring, MD: American Society for Parenteral
and Enteral Nutrition; 2015.
12. Cole R, Kocoshis SA. Nutrition management of infants with sur-
gical short bowel syndrome and intestinal failure. Nutr Clin Pract.
2013;28:421-428.
13. Mayer O, Kerner JA. Management of short bowel syndrome in
postoperative very low birth weight infants. Semin Fetal Neonatal Med.
2017;22:49-56.
537
14. Wales PW, Allen N, Worthington P, George D, Compher C, the
American Society for Parenteral and Enteral Nutrition, Teitelbaum
D. A.S.P.E.N. clinical guidelines: support of pediatric patients with
intestinal failure at risk of parenteral nutrition–associated liver disease.
JPEN J Parenter Enteral Nutr. 2014;38:538-557.
15. Sanchez SE, Braun LP, Mercer LD, Sherrill M, Stevens J, Javid PJ.
The effect of lipid restriction on the prevention of parenteral nutrition–
associated cholestasis in surgical infants. J Pediatr Surg. 2013;48:573-
578.
16. Cober MP, Killu G, Brattain A, Welch KB, Kunisaki SM, Teit-
elbaum DH. Intravenous fat emulsions reduction for patients with
parenteral nutrition–associated liver disease. J Pediatr. 2012;160:
421-427.
17. Khan FA, Fisher JG, Sparks EA, et al. Preservation of biochem-
ical liver function with low-dose soy-based lipids in children with
intestinal failure–associated liver disease. J Pediatr Gastroenterol Nutr.
2015;60:375-377.
18. Nucci A, Burns RC, Armah T, et al. Interdisciplinary management
of pediatric intestinal failure: a 10-year review of rehabilitation and
transplantation. J Gastrointest Surg. 2008;12:429-436.
19. Ubesie AC, Kocoshis SA, Mezoff AG, Henderson CJ, Helmrath MA,
Cole CR. Multiple micronutrient deficiencies among patients with
intestinal failure during and after transition to enteral nutrition. J
Pediatr. 2013;163:1692-1696.
20. Yang CJ, Duro D, Zurakowski D, Lee M, Jaksic T, Duggan C. High
prevalence of multiple deficiencies in children with intestinal failure: a
longitudinal study. J Pediatr. 2011;159:39-44.
21. Gonzalez HF, Perez NB, Malpeli A, Martinez MI, Buono BD, Viteri
FE. Nutrition and immunological status in long-term follow up of
children with short bowel syndrome. JPEN J Parenter Enteral Nutr.
2005;29:186-191.
22. Squires RH, Iyer KR. Intestinal failure. In: Corkins MR, ed. The
ASPEN Pediatric Nutrition Support Core Curriculum. 2nd ed. Silver
Spring, MD: American Society for Parenteral and Enteral Nutrition;
2015:427-440.
23. Slicker J, Vermilyea S. Pediatric parenteral nutrition: putting the
microscope on macronutrients and micronutrients. Nutr Clin Pract.
2009;24:481-486.
24. Werlin SL, Wyatt D, Camitta B. Effect of abrupt discontinuation
of high glucose infusion rates during parenteral nutrition. J Pediatr.
1994;124:441-444.
25. Squires RH, Duggan C, Teitelbaum DH, et al. Natural history of
pediatric intestinal failure: initial report from the pediatric intestinal
failure consortium. J Pediatr. 2012;161:723-728.
26. Andorsky D, Lund D, Lillehei C, et al. Nutritional and other postop-
erative management of neonates with short bowel syndrome correlates
with clinical outcomes. J Pediatr. 2001;139:27-33.
27. Vanderhoof J, Langnas A. Short-bowel syndrome in children and
adults. Gastroenterology. 1997;113:1767-1778.
28. Ksiazyk J, Piena M, Keirkus J, Lyszkowska M. Hydrolyzed versus
nonhydrolyzed protein diet in short bowel syndrome in children. J
Pediatr Gastroenterol Nutr. 2002;35:615-618.
29. Olieman J, Penning C, Ijsselstijn H. Enteral nutrition in children with
short-bowel syndrome: current evidence and recommendations for the
clinician. J Am Diet Assoc. 2010;110(3):420-426.
30. Parker P, Stroop S, Greene H. A controlled comparison of continuous
intermittent feeding in the treatment of infants with intestinal disease.
J Pediatr. 1981;99:360-364.
31. Joly F, Dray X, Corcos O, Barbot L, Kapel N, Messing B. Tube feed-
ing improves intestinal absorption in short bowel syndrome patients.
Gastroenterology. 2009;136:824-831.
32. Batra A, Beattie RM. Management of short bowel syndrome in infancy.
Early Hum Dev. 2013;89:899-904.
538
33. Goulet O, Olieman J, Ksiazyk J. Neonatal short bowel syndrome as
a model of intestinal failure: physiological background for enteral
feeding. Clin Nutr. 2013;32:162-171.
34. Batra A, Beattie RM. Management of short bowel syndrome in infancy.
Early Hum Dev. 2013;89:899-904.
35. Nordgaard I, Hasen B, Mortensen P. Importance of colonic support
for energy absorption as small-bowel failure proceeds. Am J Clin Nutr.
1996;65:222-231.
36. Hawkins R, Henry B, Gottschalk ME. Pectin supplemented enteral
feedings in the treatment of short bowel syndrome in two infants. J Am
Diet Assoc. 1995;95:A-28.
37. Drenckpohl D, Hocker J, Shareef M, Vegunta R, Colgan C. Adding
dietary green beans resolves the diarrhea associated with bowel surgery
in neonates: a case study. Nutr Clin Pract. 2005;20:674-677.
38. Wessel J, Kocoshis S. Nutritional management of infants with short
bowel syndrome. Semin Perinatol. 2007;31:104-111.
39. Levit OL, Calkins KL, Gibson C, et al. Low-dose intravenous soybean
oil emulsion for prevention of cholestasis in preterm neonates. JPEN J
Parenter Enteral Nutr. 2016;40:374-382.
40. Blackmer AB, Warschausky S, Siddiqui S, et al. Preliminary find-
ings of long-term neurodevelopment outcomes of infants treated
with intravenous fat emulsion reduction for the management of par-
enteral nutrition–associated cholestasis. JPEN J Parenter Enteral Nutr.
2015;39:34-46.
41. Ong ML, Purdy IB, Levit OL, et al. Two-year neurodevelopment
and growth outcomes for preterm neonates who received low-dose
intravenous soybean oil [published online October 21, 2016]. JPEN J
Parenter Enteral Nutr.
42. Gura K, Duggan C, Puder M, et al. Reversal of parenteral nutrition–
associated liver disease in two infants with short bowel syndrome using
parenteral fish oil: implications for future management. Pediatrics.
2006;118:e197-e201.
43. Gura K, Lee S, Valim C, et al. Safety and efficacy of a fish-oil-based
fat emulsion in the treatment of parenteral nutrition–associated liver
disease. Pediatrics. 2008;121:e678-e686.
44. Diamond IR, Sterescu A, Pencharz PB, et al. Changing the paradigm:
omegaven for the treatment of liver failure in pediatric short bowel
syndrome. J Pediatr Gastroenterol Nutr. 2009;48:209-215.
45. Puder M, Valim C, Meisel JA, et al. Parenteral fish oil improves
outcomes in patients with parenteral nutrition associated liver injury.
Ann Surg. 2009;250:395-402.
46. Soden JS, Lovell MA, Brown K, Partrick DA, Sokol RJ. Failure of
resolution of portal fibrosis during omega-3 fatty acid lipid emulsion
therapy in two patients with irreversible intestinal failure. J Pediatr.
2010;156:327-331.
47. Mercer DF, Hobson BD, Fischer RT, et al. Hepatic fibrosis persists and
progresses despite biochemical improvement in children treated with in-
travenous fish oil emulsion. J Pediatr Gastroenterol Nutr. 2013;56:364-
369.
48. Nandivada P, Chang MI, Potemkin AK, et al. The natural history
of cirrhosis from parenteral nutrition–associated liver disease after
resolution of cholestasis with parenteral fish oil therapy. Ann Surg.
2015;261:172-179.
Nutrition in Clinical Practice 33(4)
49. Food and Drug Administration. SMOFLIPID. https://www.acce
ssdata.fda.gov/drugsatfda_docs/label/2016/207648lbl.pdf. Published
May, 2016. Accessed January 27, 2017.
50. Tomsits E, Pataki M, Tölgyesi A, Fekete G, Rischak K, Szollár
L. Safety and efficacy of a lipid emulsion containing a mixture of
soybean oil, medium-chain triglycerides, olive oil, and fish oil: a
randomised, double-blind clinical trial in premature infants requir-
ing parenteral nutrition. J Pediatr Gastroenterol Nutr. 2010;51:514-
521.
51. Goulet O, Antébi H, Wolf C, et al. A new intravenous fat emulsion
containing soybean oil, medium-chain triglycerides, olive oil, and fish
oil: a single-center, double-blind randomized study on efficacy and
safety in pediatric patients receiving home parenteral nutrition. JPEN
J Parenter Enteral Nutr. 2010;34:485-495.
52. Attard MI, Patel N, Simpson J. Change from intralipid to SMOF lipid
is associated with improved liver function in infants with PN associated
liver disease. Arch Dis Child. 2012;97:A54-A55.
53. Hurwitz M, Garcia MG, Poole RL, Kerner JA. Copper deficiency
during parenteral nutrition: a report of four pediatric cases. Nutr Clin
Pract. 2004;19:305-308.
54. Frem J, Sarson Y, Sternberg T, Cole CR. Copper supplementation
of cholestatic infants. J Pediatr Gastroenterol Nutr. 2010;50:650-
654.
55. Corkins MR, Martin VA, Szeszycki EE. Copper levels in cholestatic
infants on parenteral nutrition. JPEN J Parenter Enteral Nutr.
2013;37:92-96.
56. Blackmer AB, Bailey E. Management of copper deficiency in
cholestatic infants: review of the literature and a case series. Nutr Clin
Pract. 2013;28:75-86.
57. Chi-fu JY, Duro D, Zurakowski D, Lee M, Jaksic T, Duggan C.
High prevalence of multiple micronutrient deficiencies in children
with intestinal failure: a longitudinal study. J Pediatr. 2011;159:
39-44.
58. Mansour F, Petersen D, De Coppi P, Eaton S. Effect of sodium
deficiency on growth of surgical infants: a retrospective observational
study. Pediatr Surg Int. 2014;30:1279-1284.
59. Mortensen M, Williamson N, David C, Hsu EK, Javid PJ, Horslen
S. Iodine deficiency in a parenteral nutrition–dependent adolescent
with intestinal pseudo-obstruction. JPEN J Parenter Enteral Nutr.
2016;40:730-733.
60. Golekoh MC, Cole CR, Jones NHY. Severe hypothyroidism from
iodine deficiency associated with parenteral nutrition. JPEN J Parenter
Enteral Nutr. 2016;40:1191-1193.
61. Santoro JD, Nespor C, Poole RL, Kerner JA. Iodine supplementation
for pediatric patients receiving long-term parenteral nutrition. Nutr
Clin Pract. 2016;31:245-249.
62. A.S.P.E.N. Clinical Practice Committee Shortage Subcommittee. Par-
enteral nutrition trace elements product shortage considerations. http://
www.nutritioncare.org/News/General_News/Parenteral_Nutrition_
Trace_Element_Product_Shortage_Considerations/. Updated July 26,
2016. Accessed January 13, 2017.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Soluble Fiber Use in Pediatric Short Bowel Syndrome: August 2018 539–544

C 2018 American Society for

A Survey on Prevailing Practices Parenteral and Enteral Nutrition

DOI: 10.1002/ncp.10089
wileyonlinelibrary.com

Meredith Linley Harvie, MD1 ; Margaret Alyssa Tucker Norris, MS, RD, LDN,
CLC2 ; and Wednesday Marie A. Sevilla, MD, MPH, CNSC3

Abstract
Background: In pediatric short bowel syndrome (SBS), adding fiber to enteral feedings is 1 treatment method to manage increased
stool output. However, there are no standardized recommendations on the use of fiber in this setting, including type, dosage,
titration strategies, etc. Objective: The aim of this study is to determine current prevailing practices on the use of fiber in
the treatment of chronic high stool output in the pediatric SBS population. Methods: An anonymous electronic survey with 13
questions was sent through health professional electronic mailing lists. The survey was completed by healthcare professionals
including physicians (primary care, subspecialists, and surgeons), nurse practitioners, and registered dietitians. Results: A total of
94 responses were received. The most common supplemental fiber used was pectin (62.8%). The 2 major factors considered when
initiating fiber therapy were consistency of stool (74.5%) and volume of stool output (85.1%). The major factor that determined
discontinuation of fiber was abdominal distention (67%). A majority of providers waited 2 weeks or less to see improvement
following fiber initiation before discontinuing it. Conclusions: The goal of the survey was to gather more information with regard
to fiber use in the management of SBS patients. The data collected can be used to provide future direction on determining best
practices for fiber use in SBS patients. (Nutr Clin Pract. 2018;33:539–544)

Keywords
child; dietary fiber; enteral nutrition; intestinal rehabilitation; short bowel syndrome

Introduction time. On the contrary, viscous-forming or gel-forming

The primary goal of management in pediatric short bowel
syndrome (SBS) is to reach enteral autonomy through
intestinal rehabilitation. Enteral autonomy is clinically de-
fined as the patient’s ability to achieve growth and main-
tain hydration with caloric intake solely administered to
the gastrointestinal tract without the aid of parenteral
nutrition. One of the barriers to enteral autonomy is the
quick transit time of stools, which can lead to poor
nutrient absorption, electrolyte abnormalities, and poor
growth.1,2 One accepted treatment in the management of
loose stools in pediatric SBS is addition of fiber into enteral
or oral formula because the standard formula contains little
to none.
There are 2 main types of fiber: insoluble and soluble.
Insoluble fiber does not dissolve in water and can have a
laxative effect, making stooling worse. Soluble fiber can
improve stool consistency and is therefore the most readily
used. Soluble fiber can be subdivided based on whether
it is viscous-forming or gel-forming, readily fermented,
or a combination of both (Table 1).3 Each type has
its own advantages and disadvantages. A non-viscous
fiber dissolves in water but does not increase viscosity
of chyme, having little to no effect on luminal transit
fiber dissolves in water, forms a viscous gel, and is therefore
able to slow the digestion of complex nutrients by increasing
intestinal transit time.3 Readily fermented fiber may provide
a “prebiotic” effect, leading to the formation of short-chain
fatty acids (SCFAs) and reabsorption of sodium and
water. Non-fermented fiber creates a “stool normalizing”
effect by staying gelled throughout large bowel transit, caus-
ing bulking of loose stools.3 However, it does not provide
From the 1 University of Tennessee Health Sciences Center,
Department of Pediatrics, Memphis, Tennessee, USA; 2 University of
Tennessee Le Bonheur Children’s Hospital, Clinical Nutrition
Department, Memphis, Tennessee, USA; and the 3 Children’s
Hospital of Pittsburgh, Division of Pediatric Gastroenterology,
Hepatology and Nutrition, Pittsburgh, Pennsylvania, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on May 16, 2018.
Corresponding Author:
Wednesday Marie A. Sevilla, MD, MPH, CNSC, Children’s Hospital
of Pittsburgh, Division of Pediatric Gastroenterology, Hepatology
and Nutrition, 4401 Penn Avenue, 6th Floor Faculty Pavilion,
Pittsburgh, PA 15224.
Email: wednesday.sevilla2@chp.edu
540 Nutrition in Clinical Practice 33(4)

Table 1. Classification of Soluble Fibers Commonly Used. Table 2. Regions Represented by Respondents.

Non-Viscous, Viscous, Viscous, Readily Number of

Readily Fermented Non-Fermented Fermented Region States Respondents (n = 94)

Wheat dextrin Methylcellulose Pectin West WA, OR, CA, NV, UT, n = 14
Inulin Psyllium Guar gum AZ, NM, CO, WY
Oligosaccharides β-glucan Midwest ND, SD, NE, KS, MN, n = 18
IA, MO, IL, WI, IN,
OH, MI
South TX, OK, AR, LA, MS, n = 39
the potential positive benefits achieved in the fermentation AL, GA, FL, SC, NC,
process. TN, KY, WV, VA, MD
While different studies have chosen viscous-forming or Northeast ME, NH, VT, NY, MA, n = 14
gel-forming soluble fiber, there are currently no standard- RI, CT, NJ, PA
Outside Australia, Canada, Oman, n=9
ized recommendations for the use of fiber in pediatric SBS.
the U.S. Mexico
There are no studies that examine and compare the effects
of different types of fiber, initiation and maintenance doses,
titration strategies, or the accuracy of monitoring for clini-
cal response. The aim of this study is to determine current
prevailing practices on the use of fiber in the treatment of
chronic high stool output in the pediatric SBS population.

Methods
Pediatric health professionals, in multiple disciplines man-
aging pediatric patients with SBS, were surveyed. This
population was chosen to gain knowledge on details of the
use of fiber, the goal being to provide future directions for
determining best practices and future research strategies. An
anonymous electronic survey with 13 questions was sent
through health professional electronic mailing lists (see
Appendix). The survey was completed by healthcare pro-
fessionals including physicians (primary care, subspecialists,
and surgeons), nurse practitioners, and registered dieti-
tians. The electronic survey was chosen to reach a wide
variety of healthcare professionals as well as multiple re-
gions across the United States and the globe. Survey results
Figure 1. Percentage of healthcare professionals who
were then summarized. Survey instrument development and responded to survey.
data collection were completed with a statistical program
(Qualtrics, Provo, UT). The data were analyzed through
analytic software (SPSS, IBM, Armonk, NY). The survey outside the United States (Table 2). The majority of respon-
study was approved by the University of Tennessee Health dents were dietitians. Other medical professionals included
Science Center Institutional Review Board. physicians, nurse practitioners/physician assistants, and sur-
geons (Figure 1). A large percentage, 69% of respondents,
Results practiced as part of an intestinal rehabilitation team or
program, most of whom (40.9%) reported an average annual
Respondents SBS patient census of 10–50 patients.
A total of 94 responses were received. There was a good
Type of Fiber
representation of health professionals from various regions
of the United States, with 9 respondents from other coun- Based on the survey, the most common supplemental fiber
tries including Australia, Canada, Mexico, and Oman. The used in the treatment of loose stools in pediatric short bowel
regions included South, Midwest, Northeast, and West. Of patients was pectin (62.8%). Pectin was the most commonly
the 94 responses received, 40% percent of the participants used form of fiber in the Southern and Midwestern states, as
were from the South, 20% from the Midwest, 15% from well as outside the United States. In the survey, an op-
the Northeast, 15% from the West, and 10% were from tion for additional comments was provided. Some responses
Harvie et al
Figure 2. Most common practices on the use of fiber in
pediatric SBS. SBS, short bowel syndrome.
included supplemental fiber that was administered in the
form of green beans, which contain pectin, prepared as baby
food. Other pectin-rich foods, such as sweet potatoes and
bananas, were also mentioned. A small proportion admin-
istered wheat dextrin (35.1%) and guar gum (14.9%). Wheat
dextrin was the most commonly used form of fiber in the
Western and Northeastern states.
Amount of Fiber
Respondents who used pectin (52.4%) reported the max-
imum percentage content of added pectin to feeds was
3%. About 6% of the respondents used a maximum of
4% pectin added to feeds. Regarding guar gum or wheat
dextrin, titration of supplementation was administered a
teaspoon (ࣈ2g; 20.2%); a tablespoon (ࣈ6g; 8.3%); and
a gram (9.5%) at a time. It is uncertain how the healthcare
provider determined their titrations.
Initiating and Discontinuing Fiber
In determining when to initiate fiber therapy, the 2 major
factors were consistency of stool (74.5%) and volume of
stool output (85.1%). For premature infants, majority of the
respondents (53.4%) did not initiate fiber until the patient
was >40 weeks corrected age. There was not a majority
consensus on length of time the healthcare provider waited
to see if there was an improvement with fiber therapy, but
most did not continue >2 weeks if there was no improve-
ment. The major factor that determined discontinuation of
fiber was abdominal distention (67%), in addition to in-
creased emesis (43%) and bloody stools (38%). Others had
to discontinue due to increased stool production (37%) or
decreased stool production (33%).
A summary of the most common practices on the use of
fiber in pediatric SBS is listed in Figure 2.
Discussion
A major complication of SBS is intestinal failure, which
is the inability to maintain fluid, electrolyte, and nutrient
balance despite adequate intake. For SBS patients, intestinal
541
failure occurs due to the limited surface area and altered
gut motility. Intestinal adaptation is key to achieving enteral
autonomy. For intestinal adaptation to occur, the gut must
be stimulated by enteral nutrition to enhance digestive
and absorptive capacity. By slowing down intestinal transit
time, this will presumably allow more time for absorption.
Therefore, medical management of increased stool out-
put will occasionally include the use of antimotility agents
such as diphenoxylate-atropine and/or loperamide. These
medications inhibit intestinal peristalsis by slowing tran-
sit time,4 but they are not without side effects. Fiber, while its
effects may not be as immediate, is used as an alternative or
adjunct to these medications. Some side effects of fiber in-
clude abdominal distention, bloating, and pain. Depending
on the type of fiber, these side effects can range in intensity.
Inulin, a low-digestible carbohydrate, when consumed in
large quantities is associated with increased gastrointesti-
nal intolerance. On the contrary, wheat dextrin is much
more tolerable in high quantities (30–45 g/day).5 These side
effects can sometimes be minimized by starting with a small
amount and increasing slowly. There are only few subsets of
people who cannot tolerate fiber at all.
The addition of dietary fiber to pediatric formula for
non-SBS has shown some potential benefits including de-
creasing stool output. Khoshoo et al6 performed a prospec-
tive, double-blind, randomized trial comparing tolerance of
a peptide-based formula to a peptide-based formula with
insoluble and soluble fiber in children. The trial showed
improvement of stool consistency in children receiving the
formula with fiber. There are case studies that report the
use of soluble fiber (pectin and guar gum) to minimize
chronic diarrhea in the pediatric patient.1,7-10 Alam et al8
performed a double-blind, randomized, controlled trial that
showed a statistically significant reduction in duration of
diarrhea in children who received supplementation with sol-
uble fiber in the acute setting.8 The aforementioned studies
were not conducted in children with SBS. It remains to
be examined through randomized controlled trials whether
fiber should be added to the diets of SBS patients. While
there are no clinical studies to show the safety and efficacy
of fiber use in children with SBS, many health professionals
will use fiber based on experience rather than clinical
evidence.11 Current recommendations from the American
Dietetic Association, according to Olieman et al, are to
introduce soluble fiber around 4 to 6 months of age in
the SBS patient with an intact colon. The paper mentions
pectin, green beans, and guar gum as sources of soluble
fiber used, but it did not give details to why those fibers
were chosen. It also mentions the limited evidence on the
effect of soluble fibers on enteral feeding tolerance, as
previously mentioned.12 The aim of our survey was to
gather information on the current practice of soluble fiber
use in treatment of increased stool output in the setting of
SBS. A majority of respondents (62.77%) selected pectin as
542
their source of soluble fiber. Added to this is the use of
pectin-rich foods such as green beans, sweet potatoes and
bananas. There is evidence of improved adaptation of the
small intestine and colon in animal models of SBS with the
addition of pectin to an elemental diet.11
SBS patients are known to have gut microbiota dys-
biosis secondary to small bowel resection and interrup-
tion of normal developing gut microflora. A majority
of patients are diagnosed clinically with small bacterial
overgrowth as an effect of dysbiosis. Treatment often re-
quires multiple courses of antibiotics, further disrupting
and altering their gut microbiome.13 While soluble fiber
has been postulated to increase intestinal transit time, more
recent studies have noted certain fibers to act as a prebi-
otic with the capacity to alter the gut microbiota. A pre-
biotic is fermented by intestinal microflora stimulating the
growth of more favorable intestinal bacteria. This increase
in bacterial mass leads to increase in fecal mass (stool
bulking). The soluble fibers that have prebiotic proper-
ties include wheat dextrin, pectin, gums, and inulin.5 The
fermentation of these fibers leads to production of SCFAs,
which are thought to be essential in the metabolic activity of
colonocytes.14 Piper et al13 performed a study which con-
cluded that the microbiome of healthy children contained
an abundance of anti-inflammatory commensal bacteria,
which produce SCFAs, compared with SBS patients. SCFAs
help the intestinal tract thrive and are considered to play
a key role in immune function. By lowering luminal and
fecal pH, SCFAs impede the growth of potentially harmful
bacteria, such as certain Clostridia species.5 Animal studies
show a positive correlation between increased SCFAs and
an increased immune response. Rats provided a diet with
fructooligosaccarides (soluble fiber) continuously for more
than a week had increased salivary IgA and SCFAs.15 Other
studies mirrored these findings as well.16 These results em-
phasize the potential benefits of providing fiber in pediatric
patients with intestinal failure.
From our results, it appears the majority of health
professionals surveyed (65%) will wait 2 weeks or less to
see improvement with fiber supplementation. According
to Limketkai et al, intestinal adaptation can occur during 1
to 2 years following bowel resection;4 therefore, the trial of
fiber may need to be longer before pursuing other interven-
tions. Nakao et al17 completed a study that showed adults
with chronic watery stools can achieve normal stooling pat-
terns after soluble fiber was given for ࣙ4 weeks. The study
also mentioned that after the administration of soluble
fiber, there was a slight increase in iron, copper, and zinc
levels compared with pretreatment values. It is possible that
the 2-week period of observation after commencement of
treatment is largely driven by the adverse effects of supple-
menting with fiber. These adverse effects include symptoms
of intolerance such as increased stool output, leading to
weight loss and the perception of treatment failure. While
Nutrition in Clinical Practice 33(4)
relatively healthy individuals (without intestinal pathology)
may see benefits of soluble fiber at a faster rate, given the
complexity of anatomy and microbiome of the SBS patient,
it is unknown how quickly they should respond to fiber
initiation.
As mentioned previously, there is no standardization of
how much fiber supplementation to use in SBS patients. As
we learn more about the microbiome of SBS patients, the
dosage may become more individualized. Although it is
known that fiber and prebiotics can affect the intestinal
microflora, the individual responses can vary. There are
some who respond well to certain fibers and some who do
not respond at all. This could be secondary to individual
genetics, unique microflora, inadequate dosages, and/or lack
of bacteria to ferment the specific fiber chosen.14 Two stud-
ies showed that at lower doses of fructooligosaccharides and
galactooligosaccarides (2.5 g/day), there was no significant
increase in Bifidobacterium spp (produces SCFAs as end
products from fermentation); however, when doses were
increased (10g/day) there was significant increase in Bifi-
dobacterium spp.18,19 Based on the data by region, there is
consistency among regions in terms of type of fiber used,
but no consensus across the United States. It is possible that
people in certain regions have more similar microbiomes
and, therefore, tolerate and/or benefit from certain types of
soluble fiber differently.
Limitations of Current Study
The goal of the survey was to gather more information
regarding fiber use in SBS patients. We did not explore the
rationale of choosing 1 type of fiber vs another. The survey
was mostly focused on when and why providers initiated
fiber therapy and which type was used. It would be helpful to
know why certain fiber was chosen vs another, for example,
availability, cost, or training experience. The survey did not
provide clarity on whether fiber was used in combination
with other therapies to help with significant stool output
or overall growth. The survey did not indicate whether
fiber therapy improved or resolved the patients’ loose stools.
While a variety of healthcare professionals responded to
the survey, the majority of respondents were dietitians and
included very few surgeons. Ideally, in future studies, there
should be more of a balance in healthcare professionals to
help formulate the best practice.
Future Research
More studies are needed to demonstrate the clinical effec-
tiveness of fiber administration in pediatric SBS. Studies
exploring the type, dose, and response to fiber therapy
should be completed. The preferred study would be a
double-blind, randomized, controlled trial. Animal models
can also be used in regard to effectiveness of fiber in
SBS subjects. The introduction of soluble fiber has shown
Harvie et al 543

to increase absorption of key elements such as calcium  Do not use guar gum/wheat dextrin (5)
and iron. Carvalho et al20 performed a controlled study  other (6)
in rats that showed improvement of iron deficiency on a
diet containing soluble fiber. Prebiotic fibers, as previously Q3 If you use pectin, what is the maximum percentage you
mentioned, have been shown to improve calcium absorption use?
in both humans and rats,21 including a study showing
increased calcium absorption in pubertal adolescents fol- ◦ 1% (1)
lowing soluble fiber diet.22 While these studies were per- ◦ 2% (2)
formed in subjects without intestinal pathology, it opens ◦ 3% (3)
doors for further research involving children with intestinal ◦ 4% (4)
failure. ◦ Do not use pectin (5)
Research involving the microbiome has become a recent ◦ other (6)
area of interest in gastroenterology. Collecting more data on
the microbiomes of SBS patients would help develop more Q4 For which patients do you provide soluble fiber? Please
targeted therapies for these patients. Most studies that were select all that apply.
reviewed did not specifically mention whether the subjects
had an intact colon. Correlations with remnant anatomy (ie,  Those who have a colon (1)
length of remaining small bowel and large bowel, absence  Those without a colon (2)
of ileocecal valve), type of enteral feed, and frequency are  Both A and B (3)
other complications of SBS that should be studied.  Do not use pectin or guar gum/wheat dextrin (4)

Statement of Authorship Q5 How do you decide when to start supplemental fiber?

Please select all that apply.
W. Sevilla contributed to the conception and design of the
research; A. Norris contributed to the design of the research;
 Volume of stool output (1)
W. Sevilla and A. Norris contributed to the acquisition and
analysis of the data; W. Sevilla and M. L. Harvie equally  Age (2)
contributed to the interpretation of the data; M. L. Harvie  Weight (3)
drafted and critically revised the manuscript; A. Norris drafted  Volume of feeds (4)
the manuscript; and W. Sevilla critically revised the manuscript.  Consistency of stool (5)
All authors agree to be accountable for all aspects of work  Do not use soluble fiber (6)
ensuring integrity and accuracy, and they approved the final  Other (7)
manuscript.
Q6 How early do you begin soluble fiber?

Appendix ◦ <34 weeks gestational age (1)

◦ Between 34 and 40 weeks gestational age (2)
Survey Questions and Answer Choices
◦ >40 weeks gestational age (3)
◦ Other (4)
Fiber Use in Pediatric Short Bowel Patients Questionnaire
◦ Do not use soluble fiber (5)
Q1 What do you use for supplemental fiber in the treatment
Q7 How do you administer fiber? Please select all that
of loose stools (ie, slowing transit time)? Please select all that
apply.
apply.

 Pectin (1)  Mixed with feeds (1)

 Guar gum (ie, Nutrisource) (2)  Seperate from feeds (2)
 Wheat dextrin (ie, Benefiber) (3)  Other (3)
 Other (4)
Q8 Why do you stop supplemental fiber?
Q2 If you use guar gum or wheat dextrin, how do you titrate
up/down? Please select all that apply.  Increased stool output (1)
 Decreased stool output (2)
 By 1 tablespoon (1)  Increased emesis (3)
 By 1 gram (2)  Abdominal distention (4)
 By 1 teaspoon (3)  Bloody stools (5)
 Do not titrate (4)  Other (6)
544 Nutrition in Clinical Practice 33(4)

Q9 How long do you wait to see improvements once 3. McRorie J. Evidence-based approach to fiber supplements and clini-
initiating soluble fiber? cally meaningful health benefits, part 1. Nutr Today 2015; 50(2):82-89.
4. Limketkai B, Parian A, Shah N, Colombel J. Short bowel syn-
◦ 2–5 days (1) drome and intestinal failure in Crohnʼs disease. Inflamm Bowel Dis
2016;22(5):1209-1218.
◦ 1–2 weeks (2)
5. Slavin J. Fiber and prebiotics: mechanisms and health benefits. Nutri-
◦ 2–4 weeks (3) ents 2013;5(4):1417-1435.
◦ >1 month (4) 6. Khoshoo V, Sun S, Storm H. Tolerance of an enteral formula with
◦ Other (5) insoluble and prebiotic fiber in children with compromised gastroin-
testinal function. J Am Diet Assoc 2010;110(11):1728-1733.
Q10 What region is your facility located in? 7. Rabbani G, Teka T, Saha S, et al. Green banana and pectin Improve
small intestinal permeability and reduce fluid loss in Bangladeshi
◦ West (WA, OR, CA, NV, UT, AZ, NM, CO, WY, MT, children with persistent diarrhea. Dig Dis Sci 2004;49(3):475-484.
8. Alam N, Meier R, Schneider H, et al. Partially hydrolyzed guar gum–
ID) (1)
supplemented oral rehydration solution in the treatment of acute
◦ Mid-West (ND, SD, NE, KS, MN, IA, MO, IL WI, diarrhea in children. J Pediatr Gastroenterol Nutr 2000;31(5):503-507.
IN, OH, MI) (2) 9. Becker B, Kuhn U, Hardewig-Budny B. Double-blind, randomized
◦ South (TX, OK, AR, LA, MS, AL, GA, FL, SC, NC, evaluation of clinical efficacy and tolerability of an apple pectin-
TN, KY, WV, VA, MD) (3) chamomile extract in children with unspecific diarrhea. Arzneimit-
telforschung 2006;56(06):387-393.
◦ North East (ME, NH, VT, NY, MA, RI, CT, NJ, PA)
10. Drenckpohl D, Hocker J, Shareef M, Vegunta R, Colgan, C. Adding
(4) dietary green beans resolves the diarrhea associated with bowel surgery
◦ Outside the U.S.; specify (5) in neonates: a case study. Nutr Clin Pract 2005;20(6):674-677.
11. Goulet O, Olieman J, Ksiazyk J, et al. Neonatal short bowel syndrome
Q11 What type of medical professional are you? as a model of intestinal failure: physiological background for enteral
feeding. Clin Nutr 2013;32(2):162-171.
◦ Physician (1) 12. Olieman J, Penning C, Ijsselstijn H, et al. Enteral nutrition in children
◦ Dietitian (2) with short-bowel syndrome: current evidence and recommendations for
the clinician. J Am Diet Assoc 2010;110:420-426.
◦ Surgeon (3)
13. Piper H, Fan D, Coughlin L, et al. Severe gut microbiota dysbiosis is
◦ Nurse practitioner/physician assistant (4) associated with poor growth in patients with short bowel syndrome.
◦ Other (5) JPEN J Parenter Enteral Nutr 2017;41(7):1202-1212.
14. Holscher H. Dietary fiber and prebiotics and the gastrointestinal
Q12 What is your average annual short bowel patient microbiota. Gut Microbes 2017;8(2):172-184.
census? 15. Yamamoto Y, Takahahi T, To M, et al. The salivary IgA flow rate Is
increased by high concentrations of short-chain fatty acids in the cecum
◦ <10 (1) of rats ingesting fructooligosaccharides. Nutrients 2016;8(8):500.
16. Kim M, Qie Y, Park J, Kim C. Gut microbial metabolites fuel host
◦ 10–50 (2)
antibody responses. Cell Host Microbe 2016;20(2):202-214.
◦ 50–100 (3) 17. Nakao M, Ogura Y, Satake S, et al. Usefulness of soluble dietary fiber
◦ >100 (4) for the treatment of diarrhea during enteral nutrition in elderly patients.
◦ Other (5) Nutrition 2002;18(1):35-39.
18. Bouhnik Y, Vahedi K, Achour L, et al. Short-chain fructo-
Q13 Does your facility have a gut/intestinal rehabilitation oligosaccharide administration dose-dependently increases fecal bifi-
dobacteria in healthy humans. J Nutr 1999;129(12):113–116.
team and/or program?
19. Davis L, Martı́nez I, Walter J, Goin C, Hutkins R. Barcoded py-
rosequencing reveals that consumption of galactooligosaccharides
◦ Yes (1)
results in a highly specific bifidogenic response in humans. PLoS One
◦ No (2) 2011;6(9):e25200.
20. Carvalho L, Brait D, Vaz M, et al. Partially hydrolyzed guar gum
increases ferroportin expression in the colon of anemic growing rats.
References Nutrients 2017;9(3):228.
1. Rabbani G, Teka T, Zaman B, Majid N, Khatun M, Fuchs, G. Clinical 21. Wallace T, Marzorati M, Spence L, Weaver C, Williamson P. New fron-
studies in persistent diarrhea: dietary management with green banana tiers in fibers: innovative and emerging research on the gut microbiome
or pectin in Bangladeshi children. Gastroenterology 2001;121(3):554- and bone health. J Am Coll Nutr 2017;36(3):218-222.
560. 22. Abrams S, Griffin I, Hawthorne K, et al. A combination of prebiotic
2. Squires R, Duggan C, Teitelbaum D, et al. Natural history of pediatric short- and long-chain inulin-type fructans enhances calcium absorp-
intestinal failure: initial report from the pediatric intestinal failure tion and bone mineralization in young adolescents. Am J Clin Nutr
consortium. J Pediatr 2012;161(4):723-728.e2. 2005;82(2): 471-476.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Use of a Gastroschisis Feeding Guideline to Improve August 2018 545–552

C 2018 American Society for

Standardization of Care and Patient Outcomes Parenteral and Enteral Nutrition

DOI: 10.1002/ncp.10083
at an Urban Children’s Hospital wileyonlinelibrary.com

R. Colby Passaro, MPH1 ; Kate B. Savoie, MD, MS1,2,3 ; and

Eunice Y. Huang, MD, MS, FACS, FAAP1,2,3

Abstract
Background: This study examined clinical outcomes associated with the use of a gastroschisis-specific (GS) feeding advancement
guideline. Methods: We performed a retrospective study of all simple gastroschisis babies (N = 65) treated between June 2009June
2015. We compared patients treated on a postintestinal surgery guideline using either a 1-day (1D) or 3-day (3D) feeding
advancement from August 2009–August 2013 with infants treated on a GS guideline from September 2013–June 2015. Results:
Patients in the 2 groups were similar in sex, race, gestational age, weight, and comorbidities. Median time to full enteral nutrition
(EN) was 11 days for the 1D group, 22 days for the 3D group, and 18 days for the GS group (P < .01). However, lengths of stay
and estimated weight gain per day were similar among the groups. A total of 3 infants (10%) in the 1D group developed necrotizing
enterocolitis compared with none in the 3D or GS groups. Control chart analysis showed reduced variation in median time to full
EN in the GS group when compared with the 1D and 3D groups. Guideline adherence was significantly better with the GS guideline
when compared with the 1D or 3D guidelines (94% vs 72% vs 90%; P < .01). Conclusion: A GS protocol yielded reduced variation
in median time to full EN, significant improvement in percent adherence to the guideline, and zero cases of necrotizing enterocolitis.
Weight gain and lengths of stay were not adversely affected by slower feeds. (Nutr Clin Pract. 2018;33:545–552)

Keywords
gastroschisis; infant; necrotizing enterocolitis; digestive system; surgical procedures; enteral nutrition; tube feeding; neonate; clinical
protocols

Gastroschisis is a congenital abdominal wall defect requir-
ing emergent surgical intervention and is on the rise in the
United States. According to the Centers for Disease Control
and Prevention, the prevalence of gastroschisis increased
from 2.3 per 10,000 live births to 4.4 per 10,000 live births
between 1995 and 2005.1 Almost 50% of babies born with
gastroschisis are born prematurely, which is a risk factor
for the development of necrotizing enterocolitis (NEC) and
longer lengths of hospital stay. Some of these children
require prolonged parenteral nutrition (PN), which can lead
to severe cholestasis.2,3 Although gastroschisis is typically
diagnosed via antenatal ultrasound, it is difficult to reliably
predict postnatal morbidity.4
The use of protocols to standardize care has been
shown to be beneficial for critically ill patients, con-
tributing to improved patient outcomes and provider sat-
isfaction in retrospective studies.5,6 Protocolization also
helps to identify provider noncompliance to best practices
and may help identify associated factors that contribute
to poor outcomes.7 Moreover, the implementation of
evidence-based nutrition management protocols has proven
to improve patient outcomes in critically ill patients.8-10
One key element associated with morbidity, mortality,
and health outcomes of gastroschisis patients is their nu-
trition management.11 For example, a recent study showed
that time from gastroschisis closure to first feed predicts
outcomes in these infants.12 Despite that, there are few
published guidelines prescribing exactly how and when
enteral feeding should be administered, and how it should
be advanced.13,14 Cholestatic jaundice and infection can
also be a problem with long-term use of PN; therefore, the
prompt advancement of enteral nutrition (EN) is preferred
when safe.15-19
Prior to 2009, there was no standard feeding advance-
ment regimen in our institution; the advancement for
postintestinal surgery (PIS) infants in the neonatal intensive
From the 1 College of Medicine, University of Tennessee Health
Science Center, Memphis, Tennessee, USA; 2 Division of Pediatric
Surgery, Le Bonheur Children’s Hospital, Memphis, Tennessee, USA;
and 3 Department of General Surgery, University of Tennessee Health
Science Center, Memphis, Tennessee, USA.
Conflicts of interests: None declared.
This article originally appeared online on April 4, 2018.
Corresponding Author:
Eunice Y. Huang, MD, MS, FACS, FAAP, 51 N. Dunlap St., Second
Floor, Memphis, TN 38105.
Email: ehuang@uthsc.edu
546
care unit (ICU) was determined by physician preferences
and individual infant needs. In an effort to standardize
and improve the quality of care we deliver to our PIS
patients, we instituted a feeding advancement guideline for
postsurgical patients as a part of a quality improvement
project in August 2009. The guidelines were the product
of collaboration between pediatric surgeons, neonatologists,
pediatric gastroenterologists, neonatal dietitians, clinical
pharmacists, and speech therapists. Feeding advancement
was based on residual small-bowel length after intestinal
surgery and weight at initiation of feeds. We assessed the
safety and efficacy of this guideline in a previous study,
finding less severe cholestasis and increased breast milk use
in the postimplementation group, and 67% adherence to
the guideline.20 However, during this evaluation, we also
noted that our simple gastroschisis patients were being
categorized into >1 feeding advancement category (either
1-day advancement or 3-day advancement) by the clinicians
using this protocol despite the fact that these patients did
not have any bowel resected. In addition, we observed a few
episodes of NEC in our 1-day advancement cohort.
As a result, we developed a gastroschisis-specific (GS)
feeding guideline. The purpose of this guideline, imple-
mented in September 2013, was to more accurately meet
the needs of the simple gastroschisis patient population and
provide quality care by standardizing management. These
infants had the full length of their small intestine, did not
have associated complex intestinal problems and, therefore,
underwent standard abdominal wall closure with or without
initial silo placement. Under this guideline, all postoperative
infants with simple gastroschisis and no contraindications
to enteral feeding are advanced every 48 hours until 50% of
goal feeds, then every 24 hours until goal feeds. Advance-
ment was designed to be every other day at the start of
enteral feeding because of possible early dysmotility from
inflammation of the previously exposed intestines.
The aim of this study was to retrospectively compare
gastroschisis infants cared for after implementation of the
PIS guideline to those cared for after implementation of the
GS guideline. Our primary aim was to determine whether
implementation of the newer guideline resulted in a change
in the length of time needed to achieve full enteral nutrition
(FEN) after surgical closure. Our secondary goals were to
determine if the change in guideline resulted in decreased
in-hospital complication rates, decreased variation in time
to full feeds, and/or improved guideline compliance.
Methods
Study Design
After receiving approval from the university institutional
review board, we identified all neonates who underwent
abdominal surgery for gastroschisis between June 2009 and
Nutrition in Clinical Practice 33(4)
June 2015 as categorized by the International Classification
of Disease ninth edition diagnosis code 756.73. Inclusion
criteria were any infants with a diagnosis of simple gas-
troschisis (N = 65), excluding infants with associated com-
plex intestinal problems or intestinal resections (n = 14).
Infants were also excluded if they did not reach goal feeds
or if they died prior to discharge from the hospital (n = 5).
Patients surgically treated at our neonatal ICU unit
were sampled in the following 2 cohorts: 1 treated between
August 1, 2009 and August 31, 2013, on a PIS feeding
guideline (n = 43) and another treated between September 1,
2013 and June 30, 2015, on a GS feeding guideline
(n = 17). Babies treated before September 1, 2013, had
enteral intake increases every 1 day (n = 26) or 3 days
(n = 4), or a combination of these (n = 13), depending
on bowel length and continuity (Table 1). Those infants
managed on a combination were assigned to a group based
on their starting guideline, yielding only a 1-day group
(n = 29) and a 3-day group (n = 14). Of note, although
no patients included in this study had any bowel resection,
14 infants were fed according to the 3-day PIS guideline
based on the clinical assessment of the patient care team,
including the surgeon and neonatologist. All patients
were assessed every 12 hours, or as clinically appropriate,
for signs of intolerance. The babies treated on or after
September 1, 2013, received enteral intake increases every
other day until 50% of target and then were advanced every
day until target (Table 2).
Feeding Advancement Strategy
PN was initiated immediately for postoperative infants.
Enteral feeding was initiated, orally or by nasogastric tube
(based on the infant’s neurodevelopmental and cardiopul-
monary status), when the infant had return of bowel
function (ie, stool or flatus). When feeding was tolerated,
the overall caloric intake from feeding was calculated, and
the difference between total caloric needs and those pro-
vided by the feeding was supplemented with PN. Clinicians
were encouraged to use expressed breast milk as the first
choice for feeding, followed by donor breast milk, then
age-appropriate formula. Fortification of breast milk was
initiated only after the infant was tolerating >50% of his
or her target feeding.
The GS guideline was developed in September 2013. In
this guideline, all postoperative infants with simple gas-
troschisis in continuity with the colon and no contraindica-
tions to enteral feeding were advanced every 48 hours until
50% of goal, then every 24 hours until goal feeds (Table 2).
Outcome Measures
The primary outcome for our study was time to FEN.
Secondary outcomes included time to initiation of feeds
after abdominal wall closure, length of hospital stay from
Passaro et al 547

Table 1. Postintestinal Surgery Enteral Feeding Guidelines.

Patient Type and 1 Day 3 Day

Feeding

Residual small >40% residual small bowel in continuity with 20%–40% residual small bowel in continuity
bowela colon with colon
Patient weight <1500 g ࣙ1500 g <1500 g ࣙ1500 g
Initial enteral 12 mL/kg/d 12–24 mL/kg/d 1 mL every 3 12 mL/kg/d
feedsb continuous continuous OR hours for 3 days continuous
feeds bolus feeds then 12 mL/kg/d OR
continuous intermittent
feeds continuous
feeds
Feeding increases Increase by 12 Increase by 12 Increase by 12 mL/kg/hr every 3 days
mL/kg/d every mL/kg/d every
48 hours 24 hours
Feeding Evaluate every 3 hours. If intolerance, hold Evaluate every 3 hours. If intolerance, hold
intolerancec feeds for 12 hours and reassess feeds for 12 hours and reassess

No infants included in this study were started on the 7-day guideline because they were not eligible for it based on the aforementioned criteria.
a Normal intestinal length: 140 cm for gestational age of 19–27 weeks, 220 cm for gestational age of 27–35 weeks, and 300 cm for gestational age

>35 weeks.22
b Contraindications to enteral feeding include treatment with indomethacin within 48 hours, cardiac conditions that compromise gastrointestinal

perfusion, polycythemia with a hematocrit >65 mg/dL, significant metabolic acidosis, decompensating respiratory status, hemodynamic
instability, exchange transfusion within 48 hours, abdominal signs and symptoms, severe asphyxia within 72 hours, significant apnea, cyanosis or
bradycardia associated with gavage feedings, diagnosis of stage I necrotizing enterocolitis within 10 days or stage II necrotizing enterocolitis
within 14 days, and prior return of bowel function postoperatively.
c Feeding intolerance was defined as a change in the character of gastric residuals, emesis, or ostomy output; hemorrhagic stools; or abdominal

distention with redness.

Table 2. Gastroschisis-Specific Enteral Feeding Guidelines. admission to discharge, number of patients with at least 1
episode of feeding intolerance, number of patients with at
Patient Type and Simple Gastroschisis
least 1 in-hospital complication, frequency of in-hospital
Feeding
complications, including infection, transfusion, additional
Eligible candidates Postoperative infants with simple surgical intervention, or NEC, highest total bilirubin, to-
gastroschisis in continuity with tal bilirubin at discharge, daily weight gain (grams), and
colon and no contraindications to percent adherence to guideline.
enteral feedinga
Initial enteral feeds 12 mL/kg/d continuous or bolus feeds
Feeding increases Increase by 12 mL/kg/d every
48 hours until 50% of goal enteral
volume; then increase in
Definitions
increments of 12 mL/kg/d every FEN was defined as the infant reaching goal kilocalories per
24 hours until goal kilogram per day (kcal/kg/day) feeding, as determined by
b
Feeding intolerance Evaluate every 3 hours. If intolerance,
the neonatologist and dietitian taking care of patients in the
hold feeds for 12 hours and reassess
neonatal ICU. Time to FEN was calculated from time of
a Contraindications to enteral feeding include treatment with start of enteral feeding after abdominal closure to time of
indomethacin within 48 hours, cardiac conditions that compromise reaching FEN, for example, when both target volume and
gastrointestinal perfusion, polycythemia with a hematocrit >65
mg/dL, significant metabolic acidosis, decompensating respiratory
target calories had been reached.
status, hemodynamic instability, exchange transfusion within 48 The percent of remaining small bowel for infants treated
hours, abdominal signs and symptoms, severe asphyxia within 72 on the PIS protocol was determined by dividing the remain-
hours, significant apnea, cyanosis, or bradycardia associated with ing small-bowel length as measured by the operative surgeon
gavage feedings, diagnosis of stage I necrotizing enterocolitis within
10 days or stage II necrotizing enterocolitis within 14 days, and prior
by the expected average neonatal small-bowel length based
return of bowel function postoperatively. on gestational age as determined by a previous publication
b Feeding intolerance was defined as a change in the character of
(19–27 weeks gestational age, 140 cm; 27–35 weeks, 220 cm;
gastric residuals, emesis, or ostomy output; hemorrhagic stools; or >35 weeks, 300 cm), and multiplying that factor by 100.20-22
abdominal distention with redness.
Length of stay was calculated as the number of days
the infant spent in the hospital from their arrival date to
548 Nutrition in Clinical Practice 33(4)

discharge after gastroschisis repair. No infants included in Table 3. Baseline Demographics: Median (Quartile 1,
the study were discharged before reaching full enteral feeds. Quartile 3) or Frequency (%).
NEC was defined as a diagnosis made by the neona-
Baseline 1-Day GL 3-Day GL GS GL
tologist or surgeon treating the infants if they had signs Demographics (n = 29) (n = 14) (n = 17) P Value
consistent with Bell’s stage II or higher based on overall
clinical picture, imaging, or operative findings. Male 14 (48) 10 (71) 9 (53) .35
Adherence to feeding guidelines was calculated by as- Race
sessing the percentage of days that the clinician used the Black or 8 (28) 5 (36) 2 (12) .62
recommended advancement or nonadvancement of feeding African
American
based on the infant’s daily clinical status. This information White 15 (52) 7 (50) 11 (65)
was reviewed in the daily progress notes in each infant’s Other, multi, 6 (20) 2 (14) 4 (24)
medical record, where the EN volume was recorded; if the or
volume delivered was in accordance with the documented unspecified
guideline, that day was considered adherent. If there were Gestational age 36 37.4 36.5 .05
clinical reasons that resulted in delayed advancement, this (34.6, 37.1) (36.6, 37.6) (35.2, 37.4)
was considered “adherence to the guideline” because both Birthweight, g 2310 (1919, 2556 (2200, 2650 (2400, .14
2789) 2900) 2990)
guidelines allowed for change based on clinical status of the Non-GI- 19 (66) 5 (36) 10 (59) .18
infants. For those infants fed with the PIS guideline, adher- related
ence was assessed according to the documented designation, comorbidities
regardless of whether the protocol selected was “correct”
based on the criteria for 1-day or 3-day advancement. GL, guideline; GS, gastroschisis-specific.

Statistical Analysis
Data were summarized using proportions for categorical
variables and median (interquartile ranges) for nonpara-
metric continuous variables. To test for differences in the
1-day, 3-day, and GS guidelines for categorical variables,
the χ 2 test or Fisher’s exact test were used, as appropriate.
The Kruskall-Wallis test was used to test for differences in
continuous variables between the 3 groups. The Wilcoxon
Mann–Whitney test was used to test for differences in
continuous variables between the PIS and GS groups. All
statistical analyses were performed using Stata SE 12.0
(StataCorp, College Station, TX).
To better visualize the results of our process change and
assess the common vs special cause variations, Shewhart
control charts were developed in concert with the hospital
Department of Quality Improvement and Risk to identify
common and special cause variation in time-to-goal enteral
feeds and percent adherence to the guidelines. Control
charts are a standard assessment for quality improvements
that allow simple detection of events that are indicative of
actual process change. Common cause variations are caused
by unknown factors resulting in a random distribution of
outputs around the mean. They are always present in the
process. This variation falls between the upper and lower
control limits on the chart. Special cause variations are
caused by factors that result in a nonrandom distribution
of output and can be potentially removed by improving the
process.23 All quality improvement analyses were performed
using Minitab 18 (Minitab Inc., State College, PA).
Results
Cohort Demographics
There were 60 simple gastroschisis infants included in the
analysis: 43 infants were treated according to the PIS
guideline (29 infants on 1-day advancement; 14 on 3-day
advancement), and 17 were treated on the GS guideline
(Table 3). The median gestational age was 36.5 weeks
(quartile 1, 35.2; quartile 3, 37.45); 1 (1.7%) infant was
ࣘ28 weeks gestation, 2 (3.3%) were 29–32 weeks gestation,
29 (48.3%) were 33–36 weeks gestation, and 28 (46.7%)
were ࣙ37 weeks gestation. The median birth weight was
2472.5 grams (quartile 1, 2061.5; quartile 3, 2855). The most
common comorbidities were patent foramen ovale, patent
ductus arteriosus, intraventricular hemorrhage, laryngoma-
lacia, meconium aspiration syndrome, respiratory distress
syndrome, hydronephrosis, bladder exstrophy, and anemia
of prematurity.
Analysis of Outcomes, Adverse Events,
and Guideline Compliance
There was a significant difference in the median time to
FEN between the 3 groups: 11 days for the 1-day guideline,
22 days for the 3-day guideline, and 18 days for the GS
guideline (P < .01). This difference in median time to FEN
was not significant when comparing the PIS protocol babies
(combined 1-day and 3-day guidelines) to the GS protocol
babies (14 vs 18; P = .38). Length of stay was also similar
between the 3 groups.
Passaro et al 549

Table 4. Outcomes: Median (Quartile 1, Quartile 3) or Frequency (%).

Outcomes 1-Day GL (n = 29) 3-Day GL (n = 14) GS GL (n = 17) P Value

Time to initiation of feeds after 13 (10, 18) 13 (9, 17) 10 (7, 14) .14
reanastomosis
Time to full feeds 11 (8, 19) 22 (14, 32) 18 (15, 23) <.01
Length of stay 37 (25, 52) 51 (36, 59) 40 (29, 56) .15
Patients with at least 1 episode 13 (45) 6 (43) 6 (35) .81
of feeding intolerancea
Number of patients with at 7 (24) 3 (21) 3 (17) .77
least 1 in-hospital
complication
Number of complications, total 9 (–) 3 (–) 3 (–)
Infection 4 (14) 2 (14) 3 (17) 1.00
Transfusion 2 (7) 0 (0) 0 (0) .72
Additional surgical 0 (0) 1 (7) 0 (0) .23
intervention
Necrotizing enterocolitis 3 (10) 0 (0) 0 (0) .33
Highest total bilirubin 2.7 (0.8, 3.9) 2.3 (1.3, 4.8) 3.2 (1.2, 4.9) .63
Total bilirubin at discharge 0.5 (0.4, 1.3) 0.75 (0.4, 2.2) 0.8 (0.4, 3.7) .63
Daily weight gain, g 17.4 (12.5, 21.4) 19.2 (15.3, 26.1) 19.7 (17.7, 22.9) .25
Percent adherence to guideline 72 (45, 89) 90 (79, 100) 94 (87, 100) <.01

Bold text = P value <0.05. GL, guideline; GS, gastroschisis-specific.

a Feeding intolerance = cessation of feeds for at least 24 hours secondary to aspiration, emesis, oral aversion, surgical interventions related to

feeding/gastrointestinal disturbance, diarrhea, abdominal distention, blood in the stool, ileus, necrotizing enterocolitis, infection, or increased
ostomy output.

The number of infants with feeding intolerance was not
significantly different among the 3 groups. The number of
infants with at least 1 in-hospital complication in the 1-day
group (7; 24%) was higher than that in the 3-day group (2;
21%) and GS group (3; 18%); yet this difference was not
statistically significant (P = .77). There were 3 cases of NEC
(10%) in the group advanced using the 1-day PIS guideline
(Table 4).
There was no significant difference in the level of highest
total bilirubin while in the hospital or total bilirubin at
discharge between the 3 groups. Speed of advancement of
feeds did not affect daily weight gain: 17.4 grams/day for the
1-day group, 19.2 grams/day for the 3-day group, and 19.7
grams/day for the GS group (P = .25).
Median percent adherence to the guideline was much
higher in the GS (94%) and 3-day advancement (90%)
groups than in the 1-day advancement group (72%; P < .01;
Table 4).
Quality Improvement Analysis
The analysis of time-to-goal feeds using Shewhart control
charts showed both decreased common and special cause
variation during the GS guideline time period as compared
with the PIS period, both in the 1-day and 3-day groups
(Figure 1). Similar findings were noted when percent
adherence was assessed (Figure 2). There were instances of
special cause variation (according to the Westgard rules) in
both the 1-day and 3-day groups in the form of points above
the upper control limit for time-to-goal enteral feeds and
points below the lower control limit for percent adherence
to the guidelines.24 There was no special cause variation in
the GS guideline group.
Discussion
Although the survival rate for gastroschisis is >90%,
survivors are frequently affected by varying degrees of
gastrointestinal dysfunction, which can lead to significant
morbidity and prolonged need for PN.25 Appropriate nu-
trition management is thus a key component of quality
hospital care and good health outcomes for these post-
surgical infants.26 However, enteral feeding practices for
infants with gastroschisis are highly variable, and there are
few published guidelines addressing how to advance feeds
following abdominal wall closure.12 Our results show that
implementing a simple GS feeding guideline can improve
the standardization of care by reducing variation in mean
time to FEN and increasing provider adherence to feeding
advancement protocol. Moreover, this guideline is safe and
does not significantly change mean time-to-goal enteral
feeds, length of stay, or daily weight gain.
Full enteral feeding at a median of 18 days on the
GS guideline is consistent with the current literature for
feeding advancement in infants with simple gastroschisis.27
Although the infants managed on the 1-day feeding
guideline reached FEN in a shorter period of time (11
days), we contend that this difference is minimally clinically
550
Figure 1. Three-group time to enteral feeds.
Figure 2. Three-group percent adherence to guideline.
significant, especially as it did not impact daily weight gain
or length of stay, which are critical indicators of neonatal
nutrition status and cost of care, respectively.
Consistent with previous studies, we found that although
increasing the rate of feeding advancement can result in
a shorter time to FEN in infants with gastroschisis, this
speed brings with it a risk of NEC.28 We designed our
GS guideline to limit this risk by advancing feeding every
other day for the first half of the feeding advancement
Nutrition in Clinical Practice 33(4)
process. This strategy limits the stress put on the postsurgi-
cal inflamed intestines; gastroschisis infants are at a higher
risk of feeding intolerance shortly after surgery and may
not tolerate advancement every day. Our results suggest
this strategy may be effective as we noted a downward
trend in incidence of feeding intolerance from the PIS
infants to the GS infants, and we did not observe any
cases of NEC in infants managed according to the GS
guideline.
Passaro et al
Our results contribute to the small but growing body of
literature showing that standardization of feeding guide-
lines on EN administration improves patient outcomes.26
Our GS guideline successfully reduced variation in practice,
yielding increased adherence to the feeding advancement
protocol by providers. This standardization of care resulted
in less variation in time to FEN for the infants managed on
this protocol, and therefore fewer isolated cases of infants
with very long hospital stays. In other words, although the
1-day guideline may have been very successful for some
infants, it was very unsuccessful for others, in contrast to the
GS guideline, which resulted in a slight increase in days to
FEN for the group as a whole, but fewer adverse outcomes
such as NEC. Our results thus show that it is possible to
provide more efficient care without compromising patient
outcomes with a GS guideline.
There are several limitations to this study. First, it is a
retrospective study subject to errors caused by inconsisten-
cies or omissions in medical record data. We addressed this
limitation by strictly defining the outcomes of interest at
the outset of the study and having 1 researcher perform
data extraction to maintain consistency of interpretation of
data. We also performed multiple internal audits of the data
following the first extraction. Second, this was a small study,
increasing the risk for type II errors. There may have been
differences in outcomes between the 2 populations that were
not identified as a result of our small sample size. This is a
common issue in research projects evaluating rare diseases;
however, we plan to continue monitoring these outcomes in
our expanding population in the coming years. Because the
safety of our patients is of the utmost importance, we chose
to evaluate the safety and efficacy of this guideline now
rather than wait several more years to ensure statistically
significant results. The use of the Shewhart control charts
also helped us note changes in variation over time in these
small populations. Finally, the neonates in the PIS protocol
were treated during a different time period than those on the
GS protocol, and healthcare is a rapidly advancing field, so
there may have been additional differences in management
that were not accounted for by our analysis, confounding
our results.
Our results show that standardization of care with a
simple GS feeding guideline can yield fewer adverse health
outcomes without significantly impacting time-to-goal en-
teral feeds, length of stay, or weight gain. Future quality
improvement projects with a larger sample size are needed to
accurately assess the statistical significance of the reduction
in undesired outcomes such as feeding intolerance, in-
hospital infections, and postsurgical enterocolitis. The next
step for our quality improvement project is an increase in the
rate of feeding advancement for bigger babies with simple
gastroschisis: 24 kcal/kg per advancement for infants >1500
grams. The safety and efficacy of this change in our feeding
guideline will be evaluated in future studies.
551
Acknowledgments
We would like to thank Donna Vickery, administrative director
in the Department for Quality Improvement and Risk, for her
help with control chart development. We would also like to
thank Sandra Grimes for her help as our research coordinator.
Statement of Authorship
R. C. Passaro, K. Savoie, and E. Huang equally contributed
to the conception and design of the research; R. C. Passaro
contributed to the acquisition and analysis of the data; R.
C. Passaro, K. Savoie, and E. Huang contributed to the
interpretation of the data; and R. C. Passaro, K. Savoie, and E.
Huang drafted the manuscript. R. C. Passaro, K. Savoie, and
E. Huang critically revised the manuscript; R. C. Passaro, K.
Savoie, and E. Huang agree to be fully accountable for ensuring
the integrity and accuracy of the work; and all authors read and
approved the final manuscript.
References
1. Kirby RS, Marshall J, Tanner JP, et al. Prevalence and correlates of
gastroschisis in 15 states, 1995 to 2005. Obstet Gynecol. 2013;122(2 pt
1):275-281.
2. Gupta V, Trivedi A, Walker K, Holland JA. Neurodevelopmental
outcome of infants with gastroschisis at one-year follow-up. J Neonatal
Surg. 2015;4(2):12.
3. Mills JA, Lin Y, Macnab YC, Skarsgard ED. Perinatal predictors of
outcome in gastroschisis. J Perinatol. 2010;30(12):809-813.
4. D’Antonio F, Virgone C, Rizzo G, et al. Prenatal risk factors and out-
comes in gastroschisis: a meta-analysis. Pediatrics. 2015;136(1):e159-
e169.
5. Chang SY, Sevransky J, Martin GS. Protocols in the management of
critical illness. Crit Care. 2012;16(2):306.
6. Wellington A, Perlman JM. Infant-driven feeding in premature infants:
a quality improvement project. Arch Dis Child Fetal Neonatal Ed.
2015;100(6):F495-F500.
7. Cotton BA, Dossett LA, Au BK, Nunez TC, Robertson AM, Young
PP. Room for (performance) improvement: provider-related factors
associated with poor outcomes in massive transfusion. J Trauma.
2009;67(5):1004-1012.
8. Barr J, Hecht M, Flavin KE, Khorana A, Gould MK. Outcomes in crit-
ically ill patients before and after the implementation of an evidence-
based nutritional management protocol. Chest. 2004;125(4):1446-1457.
9. Dervan N, Dowsett J, Gleeson E, Carr S, Corish C. Evaluation of over-
and underfeeding following the introduction of a protocol for weaning
from parenteral to enteral nutrition in the intensive care unit. Nutr Clin
Pract. 2012;27(6):781-787.
10. McCallie KR, Lee HC, Mayer O, Cohen RS, Hintz SR, Rhine WD.
Improved outcomes with a standardized feeding protocol for very low
birth weight infants. J Perinatol. 2011;31(suppl 1):S61-S67.
11. Walter-Nicolet E, Rousseau V, Kieffer F, et al. Neonatal outcome of
gastroschisis is mainly influenced by nutritional management. J Pediatr
Gastroenterol Nutr. 2009;48(5):612-617.
12. Aljahdali A, Mohajerani N, Skarsgard ED. Effect of timing of enteral
feeding on outcome in gastroschisis. J Pediatr Surg. 2013;48(5):971-976.
13. Olieman JF, Penning C, Ijsselstijn H, et al. Enteral nutrition in children
with short-bowel syndrome: current evidence and recommendations for
the clinician. J Am Diet Assoc. 2010;110(3):420-426.
14. Kohler JA Sr, Perkins AM, Bass WT. Human milk versus formula after
gastroschisis repair: effects on time to full feeds and time to discharge.
J Perinatol. 2013;33(8):627-630.
552
15. Sondheimer JM, Asturias E, Cadnapaphornchai M. Infection and
cholestasis in neonates with intestinal resection and long-term par-
enteral nutrition. J Pediatr Gastroenterol Nutr. 1998;27(2):131-137.
16. Spencer AU, Yu S, Tracy TF, et al. Parenteral nutrition-associated
cholestasis in neonates: multivariate analysis of the potential protective
effect of taurine. JPEN J Parenter Enteral Nutr. 2005;29(5):337-343;
discussion 43–44.
17. Mungnirandr A, Khorana J, Ruangtrakool R. Correlation between
duration of postoperative parenteral nutrition and incidence of post-
operative complication in gastroschisis patients. J Med Assoc Thai.
2010;93(4):443-448.
18. Wessel JJ, Kocoshis SA. Nutritional management of infants with short
bowel syndrome. Semin Perinatol. 2007;31(2):104-111.
19. Fallon EM, Mitchell PD, Potemkin AK, et al. Cholestasis and growth
in neonates with gastroschisis. J Pediatr Surg. 2012;47(8):1529-1536.
20. Savoie KB, Bachier-Rodriguez M, Jones TL, et al. Standardization of
feeding advancement after neonatal gastrointestinal surgery: does it
improve outcomes? Nutr Clin Pract. 2016;31(6):810-818.
21. Holland AJ, Walker K, Badawi N. Gastroschisis: an update. Pediatr
Surg Int. 2010;26(9):871-878.
Nutrition in Clinical Practice 33(4)
22. Touloukian RJ, Smith GJ. Normal intestinal length in preterm infants.
J Pediatr Surg. 1983;18(6):720-723.
23. Koutras MV, Bersimis S, Maravelakis PE. Statistical process control
using shewhart control charts with supplementary runs rules. Methodol
Comput Appl. 2007;9(2):207-224.
24. Westgard JO, Barry PL. Cost-Effective Quality Control: Managing the
Quality and Productivity of Analytical Processes. Washington, DC:
American Association for Clinical Chemistry Press; 1986.
25. Overcash RT, DeUgarte DA, Stephenson ML, et al. Factors associated
with gastroschisis outcomes. Obstet Gynecol. 2014;124(3):551-557.
26. Loomis T, Byham-Gray L, Ziegler J, Parrott JS. Impact of standard-
ized feeding guidelines on enteral nutrition administration, growth
outcomes, metabolic bone disease, and cholestasis in the NICU.
J Pediatr Gastroenterol Nutr. 2014;59(1):93-98.
27. Miranda da Silva Alves F, Miranda ME, de Aguiar MJ, Bouzada
Viana MC. Nutritional management and postoperative prognosis
of newborns submitted to primary surgical repair of gastroschisis.
J Pediatr (Rio J). 2016;92(3):268-275.
28. Hay WW Jr. Strategies for feeding the preterm infant. Neonatology.
2008;94(4):245-254.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Impact of Gastrostomy Feeding Tube Placement August 2018 553–566

C 2018 American Society for

on the 1-Year Trajectory of Care in Patients After Stroke Parenteral and Enteral Nutrition
DOI: 10.1002/ncp.10015
wileyonlinelibrary.com

Janina Wilmskoetter, PhD1 ; Annie N. Simpson, PhD2,3 ; Sarah L. Logan, PhD4 ;

Kit N. Simpson, DrPH2 ; and Heather S. Bonilha, PhD, CCC-SLP1,3

Abstract
Background: Percutaneous endoscopic gastrostomy (PEG) feeding tubes are commonly placed in acute stroke patients with a
need for enteral nutrition. However, PEG tubes are associated with medical complications and a decrease in quality of life. We
compared the 1-year care trajectory of stroke patients with and without PEG tube placement to enhance knowledge about the
long-term impact of PEG tube placement. Methods: We conducted a retrospective analysis of commercially insured stroke patients
included in the Truven Health MarketScan Research Databases of 2011. We analyzed their index hospital stay and conducted 1-
month, 3-months, 6-months, and 1-year follow-ups. We compared admissions to inpatient rehabilitation facilities, acute hospitals,
skilled nursing facilities, outpatient hospital visits, and home visits for stroke patients with and without PEG tube placement using
unadjusted and adjusted modelling. Results: Of the 8911 included stroke patients, 148 patients (1.7%) had a PEG tube placed during
their index hospital stay. After controlling for age, gender, stroke severity, comorbidities, and stroke type, PEG tube placement was
an independent predictor for admissions to inpatient rehabilitation facilities and skilled nursing facilities. Furthermore, PEG tube
placement was an independent predictor for all-cause, unplanned hospital readmissions in a multivariable logistic model (area under
the receiver operating characteristic curve was .84). Conclusion: Stroke patients who receive a PEG tube can expect a significantly
different care trajectory after being discharged from the acute hospital. Our findings can aide in predicting recovery and planning
resources and identifying gaps and points for improvement in stroke care for patients with PEG tube placement. (Nutr Clin Pract.
2018;33:553–566)

Keywords
stroke; deglutition; deglutition disorders; enteral nutrition; hospitalization; patient readmission

Background decision to place a PEG tube need to carefully weigh the

Stroke is a leading cause of death, severe long-term dis-
ability, and healthcare expenditures in the United States,1,2
with 795,000 people annually experiencing a new or recur-
rent stroke.1 Patients commonly depend on enteral and/or
parenteral nutrition because complications from the stroke
impact their ability to meet nutrition and hydration needs,
for example, as a consequence of dysphagia (swallowing
disorders). When there is an anticipated prolonged need
for enteral nutrition, percutaneous endoscopic gastrostomy
(PEG) tubes are placed in stroke patients,3 with about 6% of
all stroke patients receiving a PEG tube during their acute
inpatient hospital stay.4
Despite nutrition and hydration benefits, PEG tubes are
associated with short- and long-term complications, such as
skin/wound or abdominal wall infections, tube obstruction,
and gastric hemorrhage.5-7 Also, PEG tubes are commonly
associated with a decrease in quality of life because patients
and their families experience the limitations in the patients’
ability to eat as an emotional and social burden.3 Thus,
patients, families, and clinicians who are involved in the
potential risks and benefits. One of the biggest factors for
consideration is the prognosis of the individual patient.
For this purpose, understanding the trajectory of care for
From the 1 Department of Health Sciences and Research, College of
Health Professions, Medical University of South Carolina,
Charleston, South Carolina, USA; 2 Department of Healthcare
Leadership and Management, College of Health Professions, Medical
University of South Carolina, Charleston, South Carolina, USA;
3 Department of Otolaryngology–Head and Neck Surgery, Medical
University of South Carolina, Charleston, South Carolina, USA; and
4 Department of Neurosciences, College of Medicine, Medical
University of South Carolina, Charleston, South Carolina, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on February 3, 2018.
Corresponding Author:
Heather Shaw Bonilha, PhD, CCC-SLP, Department of Health
Sciences & Research, College of Health Professions, Medical
University of South Carolina, 77 President St. MSC 700, Charleston,
SC 29425, USA.
Email: bonilhah@musc.edu
554
stroke patients with and without a PEG tube is crucial for
clinicians to adequately counsel patients and their families
about what they can expect if or when a PEG tube is placed,
to allocate resources, and plan follow-up visits, including
reevaluations.
Once a PEG tube is placed, stroke patients can ex-
perience a variety of different healthcare trajectories and
health outcomes. Patients may show complete recovery of
their swallowing abilities with a successful removal of their
feeding tube, or patients may show little or no improvement
of their swallowing abilities with a sustained need of their
feeding tube, or patients may die as a consequence of feeding
tube/dysphagia-related or unrelated complications.8-10 The
proportion of patients with poststroke dysphagia who have
their PEG tube removed varies broadly between 16.3% and
75% and likely depends on a multitude of patient factors
and practice patterns.11 Despite PEG tube removal being
a primary goal of stroke rehabilitation,12,13 some stroke
patients with a PEG tube will never recover their swallowing
abilities and will be in the need of ongoing healthcare
support.
Our previous research has shown that stroke patients
who receive a PEG tube during their acute hospital stay are
twice as likely to be readmitted within 30–60 days to an acute
care hospital when compared with stroke patients who did
not receive a PEG tube.14 For some patients, the primary
medical reason of their hospital readmission was directly
linked to their PEG tube. Reducing hospital readmissions
is not only a main incentive of acute care hospitals who
are penalized for 30-day readmissions through the Medicare
Hospital Readmission Reduction Program15 but also in the
interest of patients because readmissions are a surrogate for
quality of care. Although several factors other than PEG
tube placement may contribute to hospital readmissions in
stroke patients,16 we found in a previous study that PEG
tube placement was an independent risk factor for hospital
readmissions in stroke patients.14
When we selectively investigated stroke patients with
PEG tube placement regarding their risk factors for read-
missions, we found that discharge locations play a role
in the likelihood that stroke patients with a PEG tube
will be readmitted. More than half (53.80%) of all stroke
patients with a PEG tube were discharged from the acute
hospital stay to a skilled nursing facility (SNF), and the vast
majority (71.2%) of all 30-day readmitted stroke patients
with a PEG tube were originally discharged to a SNF.
We confirmed discharge location as one of several risk
factors in a prediction model for 30-day readmissions for
stroke patients with PEG tube placement. Besides discharge
location, number of comorbidities, hospital length of stay
(LOS), and stroke type were predictive for 30-day readmis-
sions. To the best of our knowledge, this is the only study
that assessed hospital readmissions of stroke patients with
PEG tube placement and that identified opportunities to
Nutrition in Clinical Practice 33(4)
prevent future readmissions.14 Our findings emphasize that
stroke patients with a PEG tube are at risk for hospital
readmissions because of clinically significant postacute care
complications.
Based on the currently available evidence, much remains
unknown about the healthcare trajectory of stroke patients
with a PEG tube after being discharged from the acute care
hospital. For instance, we lack an understanding of the
impact of PEG tube placement on discharge destinations,
access to rehabilitation, and hospital readmissions that oc-
cur longer than 2 months after the acute stroke. Specifically,
access to inpatient rehabilitation facilities (IRFs) seems to
be a crucial catalyst for favorable poststroke outcomes.
The American Heart Association/American Stroke Asso-
ciation recommend in their recently published guidelines
that—if possible—stroke patients should receive treatment
in an IRF rather than an SNF because of the better
overall outcomes and higher rate of returning to live in
communities.17
Understanding the trajectory of the care of stroke
patients with PEG tubes is crucial for various reasons.
During the inpatient hospital stay, clinicians need to take
the prognosis of an individual patient into account to
make informed decisions that can have an impact on the
patient’s postacute trajectory (such as placement of a PEG
tube) and to provide recommendations for their postacute
care (such as discharge destinations, treatment planning,
resource allocation). In addition, clinicians need to inform
patients and their families about possible postacute care
trajectories to allow them to form realistic expectations and
enable informed decision making. Moreover, information
about the continuum of care, such as the occurrence of
complications and nonfavorable outcomes (eg, rehospital-
izations following preventable complications), may allow
caregivers and stakeholders to identify key aspects in the
care of these patients that warrant improvement.
Thus, the purpose of our study was to provide informa-
tion about the continuum of care of stroke patients with
PEG tube placement when compared with stroke patients
without PEG tube placement by evaluating their index
hospital stay and their 1-year care trajectory. We aimed to
build on our previous research on practice patterns and
hospital readmissions by expanding the follow-up time to
1 year, expanding the patient population, and testing our
30-day readmission model in a new patient population. We
had 3 main objectives in our study. First, we sought to
determine the demographic characteristics, medical charac-
teristics, and rate of PEG tube placements during the index
(acute stroke) hospital stay of commercially insured stroke
survivors with no history of prestroke dysphagia. Second,
we sought to conduct a 1-year follow-up of these patients to
determine their care trajectory by investigating admissions
to IRFs, hospital readmissions, and admission to other
facility encounters (SNF, outpatient hospital, and home
Wilmskoetter et al
visits). Third, we sought to test our previously developed
prediction model for 30-day readmissions for stroke patients
with PEG tube placement.14 For all our aims, we sought
to compare stroke survivors with PEG tube placement to
patients without PEG tube placement during their index
hospital stay to understand the specific characteristics of
patients with poststroke PEG tube placement. Importantly,
as a major improvement of our previous research studies
and most stroke outcome studies that are based on ad-
ministrative databases, we were able to control for stroke
severity by using a novel, validated, stroke severity index
recently developed by our group specifically for application
in administrative databases.18
Methods
We conducted a retrospective analysis of commercially
insured stroke patients included in the Truven Health
MarketScan Research Databases. These databases capture
person-specific clinical utilization, expenditures, and enroll-
ment across inpatient, outpatient, prescription drug, and
carveout services. We analyzed 3 years of the MarketScan
databases: 2010 was used as the baseline period, 2011
was used to identify the index stroke patient cohort, and
2012 was used as the follow-up period. We only included
patients with a primary diagnosis of stroke in 2011 that
was either represented by an International Classification
of Diseases, Ninth Revision, Clinical Modification (ICD-9-
CM) code of 434.xx for ischemic or 431.xx for hemorrhagic
stroke. PEG tube placement was identified through the
ICD-9-CM procedure code 43.11. The 1-year baseline (year
2010) before the index hospital stay was used to exclude
patients with a previous stroke, dysphagia diagnosis, feeding
tube placement, or speech and language pathology (SLP)
utilization. To ensure a 1-year baseline and a minimum
follow-up of 1 month for all patients, we excluded all stroke
patients who were not enrolled in their insurance at least 1
year prior and 1 month after their index hospitalization in
2011.
Inpatient records for the index hospital stay in 2011 of
stroke patients with and without PEG tube placement were
analyzed with univariate analyses for demographic (age,
gender) and clinical information (ischemic or hemorrhagic
stroke, comorbidities measured by the Charlson Comorbid-
ity Index [CCI],19,20 died during hospital stay, LOS, presence
of dysphagia, discharge destination). We also employed a
recently developed validated stroke severity measure for use
in archival data, the Stroke Administrative Severity Index
(SASI).18 SASI was developed using the National Institute
of Health Stroke Scale as a theoretical framework—a
commonly clinically used stroke severity scale—and uses
diagnoses and procedure codes at discharge from the acute
hospital stay. SASI predicts 30-day postdischarge mortality
or discharge to hospice and thus models severity of the
555
patient’s condition at discharge from the acute hospital stay.
SASI includes 7 binary indicator variables (aphasia, coma,
dysphagia and/or dysarthria, hemiplegia or monoplegia,
neglect, nutritional infusion, tracheostomy, and/or ventila-
tion) that can be used separately or as a total weighted
score to control for stroke severity in administrative
databases.
The characteristics of stroke patients with and without
PEG tube placement were compared using χ 2 tests and
Wilcoxon-rank sum tests. Nonparametric tests were ap-
plied because all variables were not normally distributed
when tested with the Shapiro-Wilk test and by visual
examination of distribution. To adjust for confounding
variables, multivariable logistic regression modeling was
performed to predict binary outcome variables, and gen-
eralized linear modeling with negative binominal distribu-
tions and a log link function was performed to predict
hospital LOS.
Time points of analysis were the index hospital stay, 1
month (30 days), 3 (90 days), 6 (180 days), and 12 months
(360 days) postdischarge. Only patients who were enrolled in
the commercial insurance at the time point of analysis were
included in the specific analysis. Because patients drop out
of the insurance, we used changing cohorts across all time
points.
For all 4 follow-up time points, we assessed the trajectory
of care by analyzing inpatient and facility records regarding
admission rates and diagnoses, time between facility stays,
LOS, and discharge destinations. Specifically, we investi-
gated (1) admissions to IRFs, (2) hospital readmissions, and
(3) SNF, outpatient hospital, and home visit encounters.
For hospital readmissions, we separately investigated 30-
day all-cause, unplanned readmissions as defined by the
Center for Medicare and Medicaid Services (CMS), trans-
ferred admissions, and planned readmissions. CMS defines
“readmission” as an unplanned admission to an hospital
within 30 days after discharge from the same or another
hospital.15 Hospital admissions that represent planned (in-
cluding in-patient rehabilitation) or transferred admissions
are excluded from 30-day all-cause readmissions. To identify
planned readmissions, we used the CMS planned readmis-
sion algorithm (version 2.1, March 2013). This algorithm
defines whether readmissions are planned admissions, that
is, if the primary readmission diagnosis is not acute or
not a complication of care. The algorithm lists diagnoses
and procedures using the Clinical Classification Software.
We used the Clinical Classification Software provided by
the Agency for Healthcare Research and Quality for back
translation into ICD-9-CM diagnosis and procedure codes
(http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp).
To test our previously published prediction model, we
used multivariable logistic regression analysis to, first, de-
termine whether PEG tube placement is an independent
predictor for 30-day unplanned readmissions and, second,
556 Nutrition in Clinical Practice 33(4)

16,461 stroke patients in 2011

(Dx 1 = 431.xx or 434.xx)

134 excluded
- patients <18 years

7375 excluded based on baseline 1 year prior index

hospital stay
- 3 with missing data
- 3842 not enrolled for ≥ 1 year prior the index hospital stay
- 3527 not meeting eligibility criteria (3304 diagnosed with prior
stroke, 379 diagnosed with prior dysphagia, 120 had prior
feeding tube placement, 100 had prior SLP utilization)

8955 index stroke patients

44 excluded
- patients with 2 hospital
admissions at the index stroke date

8911 index stroke patients

- 148 patients with PEG tube placement
- 8763 patients without PEG tube placement

Figure 1. Flowchart for patient enrollment for the index patient cohort. Dx 1, primary admission diagnosis; PEG, percutaneous
endoscopic gastrostomy; SLP, speech and language pathologist.

to predict the outcome of 30-day unplanned readmissions
in patients with a stroke who had a PEG tube placement
during their index hospital stay. We adopted our previously
published model that included the following 4 predictors:
(1) LOS (expressed as a binary variable with a cutoff
point of, median, 11 days), (2) stroke type (hemorrhagic vs
ischemic stroke as the admission diagnosis), (3) comorbidi-
ties (measured by the CCI), and (4) discharge destination
from the index hospital stay. For the predictor “discharge
destination” we tested discharges to any destination (eg,
home, IRF, hospice) against discharges to nursing facilities
(eg, SNF), because nursing facilities were the discharge
destinations for the majority of the patients. In addition,
to the previously identified significant predictors, we tested
whether stroke severity (measured by the SASI) is a signifi-
cant predictor for hospital readmissions and whether adding
this predictor would improve our model. We assessed model
fit with the Hosmer and Lemeshow test and the model’s
discriminatory ability to differentiate between patients who
will and will not experience a 30-day hospital readmission
by the area under the receiver operating characteristic curve
(AUROC).
All analyses were conducted by using SAS statistical
software (version 9.3; SAS Institute, Inc., Cary, NC). The
study was reviewed by our institutional review board who
determined the study was not human subject research.
Results
Patients
In the MarketScan databases from 2011, 16,461 patients had
at least 1 inpatient stay with a primary diagnosis for ischemic
or hemorrhagic stroke (Figure 1). After excluding patients
younger than 18 years, patients not fulfilling our 1-year
baseline criteria and patients with 2 hospital admissions at
the same index stroke date, our final sample included 8911
stroke patients. Of this sample, 148 patients (1.66%) had a
PEG tube placement during their index hospital stay.
For the follow-up period, the number of enrolled stroke
patients decreased from 8911 patients at the time of the
index hospital stay to 5791 (64.99%) patients at the time of
the last follow-up at 12 months postdischarge. Table 1 shows
how many patients were included in each group at each
follow-up time point. At 12 months postdischarge, 39.19%
of all patients with a PEG tube placement and 65.42% of all
patients without a PEG tube placement were still enrolled
Wilmskoetter et al 557

Table 1. Enrolled and Eligible Patients.

Stroke Patients With Stroke Patients Without

PEG Tube Placement PEG Tube Placement
All Stroke During Index Hospital During Index Hospital
Time point Patients, n (%) Stay, n (%) Stay, n (%)

Index hospital stay 8911 (100) 148 (100) 8763 (100)

1 month postdischarge 8330 (93.48) 133 (89.86) 8197 (93.54)
3 months postdischarge 7773 (87.23) 112 (75.68) 7661 (87.42)
6 months postdischarge 7025 (78.48) 86 (58.11) 6939 (79.19)
12 months postdischarge 5791 (64.99) 58 (39.19) 5733 (65.42)

PEG, percutaneous endoscopic gastrostomy.

Table 2. Patient Demographics and Characteristics.

Group 1: Stroke Group 2: Stroke Absolute Statistical

Patients With PEG Patients Without Difference Difference
Tube Placement, PEG Tube Between the Between the
Variable n = 148 Placement, n = 8763 Groups Groups, P Value

Age (years) Mean (SD) 55.35 (7.2) 54.36 (8.39) 0.99 .2985b
Range 29–64 18–64 NA
Female gender, (vs male) n (%) 58 (39.19) 3,632 (41.45) 3,574 (2.26) .5803a
Ischemic stroke (vs n (%) 94 (63.51) 7,611 (86.85) 7,517 (23.34) <.001a
hemorrhagic)
Length of stay (days) Mean (SD) 17.08 (10.93) 4.35 (5.61) 12.73 <.001b
Median (Range) 14 (1-60) 3 (1-214) 11
Died during hospital stay n (%) 7 (4.73) 328 (3.75) 321 (1.02) .5368a
Charlson Comorbidity Index Mean (SD) 1.55 (1.56) 0.99 (1.49) 0.56 <.001b
(min 0, max 24) Median (Range) 2 (0–10) 0 (1–12) 2
Coded for dysphagia n (%) 69 (46.62) 437 (4.99%) 368 (41.63) <.001a
Stroke administrative severity Mean (SD) 5.22 (5.63) 2.17 (3.54) 3.05 <.001b
index (min 0, max 56) Median (Range) 5 (0–28) 0 (0–35) 5

max, maximum; min, minimum; NA, not applicable; PEG, percutaneous endoscopic gastrostomy.
a χ 2 test.
b Wilcoxon-rank-sum test.

in insurance and, thus, available for the 12-month follow-up
analysis.
Index Hospital Stay
Patient demographics and characteristics. Table 2 shows the
demographic and medical characteristics of patients with
and without PEG tube placement during their index hospi-
tal stay. The average age and gender distribution matched
between the 2 groups and was not statistically different.
About the same proportion of patients in both groups died
during the index hospital stay. Stroke patients with PEG
tube placement were more likely to have had a hemorrhagic
stroke, longer hospital stays, more comorbidities, higher
stroke severity, and were more often coded for dysphagia
when compared with stroke patients without PEG tube
placement.
Discharge destinations. Figure 2 shows the proportions
of patients with and without PEG tube placement being
discharged to specific destinations. The largest proportion,
one third (36.5%), of stroke patients with a PEG tube
placement were directly discharged from the index hospital
stay to IRFs. About a quarter (24.3%) of stroke patients
with PEG tube placement were directly discharged to SNFs,
and about a sixth (15.5%) were discharged to home or home
healthcare. The vast majority, two thirds (68.3%), of stroke
patients without a PEG tube placement were discharged to
home or home healthcare. Furthermore, 14% went to IRFs
and 2.9% of all discharged patients without a PEG were
discharged to SNFs.
Using multivariable logistic regression, PEG tube place-
ment during the index hospital stay was an independent
positive predictor for discharge to IRFs, SNFs, and long-
term care hospitals, but a negative predictor for discharge
to home after controlling for age, gender, stroke severity
558
15.5
Home
Inpatient rehabilitation facility
14
24.3
Skilled nursing facility
2.9
12.2
Other destination
10.7
6.8
Long term care hospital
0.4
4.7
Inhospital death
3.7
0 10 20
Figure 2. Discharge destinations from the index hospital stay
for stroke patients with and without percutaneous endoscopic
gastrostomy (PEG) tube placement.
(SASI score), comorbidities (CCI score), and stroke type.
Patients with PEG tube placement had higher odds for
being discharged to IRFs, SNFs, or long-term care hos-
pitals, but lower odds for being discharged home when
compared with stroke patients without PEG tube placement
(Figure 3).
Stroke severity varied for patients discharged to different
destinations (Table 3). Overall patients who were discharged
home had the lowest severity and patients discharged to
long-term care hospitals the highest stroke severity. Patients
with PEG tube placement who were discharged home had
significantly lower stroke severity than patients discharged
to IRFs. Patients without PEG tube placement who were
discharged home had significantly lower stroke severity than
patients discharged to other destinations, such as SNFs,
IRFs, or long-term care hospitals.
The 1-Year Care Trajectory
IRFs. Within 1 month after discharge from the acute care
hospital, 45.11% (60 of 133) of all stroke patients who
had a PEG tube placement during their index hospital
stay had been to an IRF. This was a significantly larger
proportion than patients without a PEG tube placement
of whom 17.32% (1420 of 8197) had been to an IRF
(Table 4). PEG tube placement increased the odds 2-fold
for patients going to IRFs while controlling for confounders
(Figure 4). The cumulative LOS in IRFs was significantly
longer for stroke patients with PEG tube placement than for
patients without PEG tube placement. PEG tube placement
was an independent predictor of longer LOS in IRFs,
even after controlling for stroke severity and stroke type
(adjusted mean rehabilitation LOS estimate for patients
with PEG tube placement was 3.21 days and was 2.77 days;
P < .0001 without PEG). Age, gender, and comorbidities
Nutrition in Clinical Practice 33(4)
68.3
36.5
Stroke patients with PEG
tube placement (N=148)
Stroke patients without PEG
tube placement (N=8,763)
30 40 50 60 70 80
Percentages
Figure 3. Unadjusted and adjusted odds ratios for discharge
destinations from the index hospital stay for patients with
percutaneous endoscopic gastrostomy (PEG) tube placement
when compared with patients without PEG tube placement.
Unadjusted: simple logistic regression with PEG tube
placement as primary predictor; adjusted: multivariable
logistic regression with PEG tube placement as primary
predictor and age, gender, stroke severity (Stroke
Administrative Severity Index score), comorbidities (Charlson
Comorbidity Index score), and stroke type as control variables
(length of hospital stay and dysphagia were not included
because of moderate correlations with the main predictor
PEG tube placement [correlation coefficient r > 0.2; P <
.0001]).
did not significantly contribute to the model for LOS in
IRFs and, thus, were dropped from the final parsimonious
model.
Hospital readmissions. In total, 35% of all stroke patients
with a PEG tube placement and 13% without a PEG tube
placement were admitted to an inpatient facility—excluding
IRFs—during the first month after discharge (Figure 5).
This included transfer admission, 30-day hospital readmis-
sion as defined by CMS, and planned hospital readmissions.
The 30-day all-cause, unplanned readmission rate (exclud-
ing rehabilitation, transfer, and planned admissions) was
18% for patients with PEG tube placement and 9% for
patients without PEG tube placement.
After controlling for age, gender, stroke severity (SASI
score), comorbidities (CCI score), and stroke type, stroke
patients with PEG tube placement had 1.8 times higher odds
for a 30-day all-cause, unplanned hospital readmission and
3.8 higher odds for being transferred to another hospital
when compared with stroke patients without PEG tube
Table 3. Average Stroke Severity (Measured by the Stroke

Statistical Difference Test Between Stroke

months

<.001b
<.001a
Patients With and Without PEG Tube
Administrative Severity Index) of Patients Discharged From

12
the Index Hospital Stay to Different Destinations (Ordered by
Increasing Severity of Group 1).

Placement, P Value

months

<.001a

.001b
Group 2: Stroke

6
Discharge Group 1: Stroke Patients Without
Destination From Patients With PEG PEG Tube
Index Hospital Tube Placement, n = Placement, n =

months

<.001b
<.001a
Stay 148; Mean (SD) 8763; Mean (SD)

3
Home 3.91 (5.92) 1.50 (2.68)
Others 4.34 (6.81) 2.41 (3.73)b

<.001b
<.001a
month
Skilled nursing 4.89 (5.24) 3.46 (4.40)b

1
facility
Inpatient 6.00 (4.73)a 4.25 (3.81)b
rehabilitation

(n = 8197) (n = 7661) (n = 6939) (n = 5733)

Stroke Patients With PEG Tube Placement Stroke Patients Without PEG Tube Placement

months

(17.63)

(13.69)
facility

1,011

18.05

1–96
12
Long-term care 7.90 (9.10) 4.06 (4.41)b
hospital

PEG, percutaneous endoscopic gastrostomy.

months

(17.91)

(13.41)
1–136
1,243

17.86
a Significant difference (P ࣘ .05) in stroke severity of patients

6
discharged to this location when compared with patients of the same
group discharged to home.
b Highly significant difference (P < .001) in stroke severity of patients

discharged to this location when compared with patients of the same

months
Table 4. Stays in Inpatient Rehabilitation Facilities After Discharge From the Index Hospital Stay.

(17.88)

(12.81)

IRF, inpatient rehabilitation facility; LOS, hospital length of stay; PEG, percutaneous endoscopic gastrostomy.
1–136
1,370

17.51
group discharged to home.

3
placement. The likelihood for planned readmissions did not

(17.32)

(11.50)
month

1,420

16.93

1–88
differ between the 2 groups (Figure 6).
The most frequent primary readmission diagnosis for 30- 1
day all-cause, unplanned hospital readmissions for stroke (n = 58)
patients with and without PEG tube placement (n = 779)
months

(23.64)
8–111
(48.2)
32.5
12

was cerebral artery occlusion (27.21%), followed by in- 28

tracerebral hemorrhage (4.88%), transient cerebral ischemia
(3.34%), and cerebral embolism with cerebral infarction
(n = 133) (n = 112) (n = 86)
months months

(47.67)

(21.46)
8–111
33.54
(3.21%). The most frequent primary readmission diagnoses
41
6

were the same for patients with and without PEG tube
placement.
(52.68)

(18.12)

Within 1 month after being discharged from the index

30.24

8–82
59
3

hospital stay, patients with a PEG tube placement were,

on average, readmitted to an acute care hospital after 8.04
days (SD 9.04, minimum 0, maximum 30) when com-
(45.11)

(17.16)
month

29.12

6–82

pared with patients without a PEG tube placement who

60
1

were readmitted after 9.14 days (SD 8.73, minimum 0,

maximum 30). The difference between patients with and
Cumulative LOS Mean (SD)

without PEG tube placement was not statistically significant

Range
n (%)

(P = .5028).
b Wilcoxon-Rank-Sum test.

Once readmitted, stroke patients with PEG tube place-

ment had a significantly longer LOS for their first un-
least 1 IRF stay

planned all-cause hospital readmission within 1 month

Patients with at

after discharge when compared with stroke patients without

IRF stays

PEG tube placement (average 12.17 days and 7.68 days,

respectively; P = .0304). However, after controlling for
for all
Variable

a χ 2 test.

stroke severity and stroke type, PEG tube placement did

not have a statistically significant effect the LOS of the first

559
560 Nutrition in Clinical Practice 33(4)

Figure 4. Unadjusted and adjusted odds ratios for admissions to inpatient rehabilitation facilities within 30 days (1 month) after
discharge from the acute hospital for patients with percutaneous endoscopic gastrostomy (PEG) tube placement when compared
with patients without PEG tube placement. Unadjusted: simple logistic regression with PEG tube placement as primary
predictor; adjusted: multivariable logistic regression with PEG tube placement as primary predictor and age, gender, stroke
severity (Stroke Administrative Severity Index score), comorbidities (Charlson Comorbidity Index score), and stroke type as
control variables (length of hospital stay and dysphagia were not included because of moderate correlations with the main
predictor PEG tube placement [correlation coefficient r > 0.2; P < .0001]).

unplanned all-cause hospital readmission as exemplified

by the data from the first month after discharge (P =
.1897; adjusted means estimate for patients with PEG tube
placement was 2.27 days, for patients without PEG tube
placement 1.98 days).

SNF encounter, outpatient hospital visits, and home visits.

Table 5 shows the number and proportion of stroke patients
with and without a PEG tube placement who were admitted
to a SNF and/or had an outpatient hospital visit and/or
home visit. At all follow-up time points, significantly more
stroke patients with a PEG tube were admitted to a SNF
when compared with stroke patients without a PEG tube

25% Figure 6. Unadjusted and adjusted odds ratios for all-cause,

unplanned readmissions as defined by Center for Medicare
% of all index hospital paents with

and Medicaid Services, transfer admissions and planned

one or more inpaent admission

20%

hospital readmissions within 30 days (1 month) after

15% discharge from the acute hospital for patients with
Transfer admissions percutaneous endoscopic gastrostomy (PEG) tube placement
30-day hospital readmission as when compared with patients without PEG tube placement.
10% defined by CMS
Unadjusted: simple logistic regression with PEG tube
Planned hospital readmissions (other
than inpaent rehabiliaon) placement as primary predictor; adjusted: multivariable
5%
logistic regression with PEG tube placement as primary
predictor and age, gender, stroke severity (Stroke
0% Administrative Severity Index score), comorbidities (Charlson
All stroke paents with PEG All stroke paents without
tube placement PEG tube placement Comorbidity Index score), and stroke type as control variables
Type of hospitalizaon (length of hospital stay and dysphagia were not included
because of moderate correlations with the main predictor
Figure 5. Hospitalizations within 1 month after discharge PEG tube placement [correlation coefficient r > 0.2; P <
from the index hospital stay for all patients who were enrolled .0001]).
at 1 month postdischarge. CMS, Center for Medicare and
Medicaid Services; PEG, percutaneous endoscopic
gastronomy.
Wilmskoetter et al 561

Table 5. Number and Proportion of Patients With and Without PEG Tube Placement Who Had a Skilled Nursing Facility Stay,
Outpatient Visit or Home Visit at 1-Month, 3-Month, 6-Month, or 12-Month Follow-Ups.

Visit Patient Group 1 month, n (%) 3 months, n (%) 6 months, n (%) 12 months, n (%)

Skilled nursing facility stay Stroke patients with 30 (23) 40 (36) 32 (37) 23 (40)
PEG tube placement
Stroke patients without 258 (3) 322 (4) 313 (5) 273 (5)
PEG tube placement
Outpatient hospital visit Stroke patients with 40 (30) 78 (70) 67 (78) 52 (90)
PEG tube placement
Stroke patients without 3607 (44) 4887 (64) 5041 (73) 4,581 (80)
PEG tube placement
Home visit Stroke patients with 15 (11) 40 (36) 38 (44) 24 (41)
PEG tube placement
Stroke patients without 753 (9) 916 (12) 904 (13) 790 (14)
PEG tube placement

PEG, percutaneous endoscopic gastrostomy.

Table 6. Multivariable Logistic Regression Model for the Event of an All-Cause, Unplanned 30-Day Readmission for Stroke
Patients With a PEG Tube Placement During Their Index Hospital Stay.

Variable β SE β P Value OR (95% CI)

Intercept −1.55 0.61 .0110 NA

Hemorrhagic stroke 1.25 0.57 .0271 3.50 (1.15–10.61)
Charlson Comorbidity Index 0.22 0.15 .1544 1.25 (0.92–1.68)
Length of hospital stay (ࣙ11 d) −0.95 0.57 .0960 0.39 (0.13–1.18)
Discharge to home 1.32 0.65 .0412 3.73 (1.05–13.21)
Discharge to rehab facility −0.89 0.61 .1468 0.41 (0.12–1.37)
Discharge to other destinations −1.75 1.13 .1197 0.17 (0.02–1.58)

The variable “discharge to” was not included in the model because only 2 patients were coded for this event. β, regression coefficient; NA, not
applicable; PEG, percutaneous endoscopic gastrostomy.

placement (χ 2 test, P < .0001). After controlling for age,
gender, stroke severity, stroke type, and comorbidities, the
odds for being admitted to a SNF were 6–8 times higher
for patients with a PEG tube than for patients without a
PEG tube (1 month: OR = 5.84, 95% CI, 3.65–9.34; 3
months: OR = 7.23, 95% CI, 4.71–11.11; 6 months: OR =
7.48, 95% CI, 4.64–12.07; 12 months: OR = 7.79, 95% CI,
4.39–13.82).
Patients with PEG tube placement had significantly more
outpatient hospital visits in the first month (P = .0014), but
did not differ at the 3-month, 6-month, and 12-month follow
ups. After controlling for age, gender, stroke severity, stroke
type, and comorbidities, PEG tube patients had significantly
lower odds for outpatient hospital visits at 1 month (OR =
0.33, 95% CI, 0.21–0.49) and 3 months follow-up (OR =
0.573, 95% CI, 0.369–0.890).
The number of home visits did not differ between the
2 groups for the 1-month follow up (P > .05), but did for
the 3-month, 6-month, and 12-month follow-up time points,
when significantly more patients with PEG tube placement
had home visits (P < .001; Table 5). After controlling for age,
gender, stroke severity, stroke type, and comorbidities, PEG
tube patients had significantly higher odds for home visits at
the last 3 follow-up time points (3 months: OR = 2.50, 95%
CI,1.66–3.76; 6 months: OR = 3.37, 95% CI, 2.15–5.30; 12
months: OR = 2.69, 95% CI, 1.55–4.68).
Prediction Model for All-Cause, Unplanned
30-Day Readmissions
30-day readmission prediction model for stroke patients with
PEG tube placement. We tested our previously developed
prediction model for 30-day readmissions of stroke patients
with PEG tube placement in the stroke patient cohort of
the study presented here. The model showed a good model
fit with a Hosmer-Lemeshow test P value of .95 (P > .05).
Significant predictors in the model were hemorrhagic stroke
and discharge destination (Table 6). The AUROC was 0.75
(Figure 7).
Adding the SASI to the prediction model to control
for stroke severity—and deleting the predictor discharge to
hospice because SASI itself predicts 1-year mortality or
562 Nutrition in Clinical Practice 33(4)

discharge to hospice—improved the discriminative ability

of the model. Adding all SASI indicators separately re-
vealed the best discriminative ability with an AUROC of
0.85 (Hosmer-Lemeshow test: P = 0.69 [P > .05]; Figure 8)
when compared with adding the total SASI score (AUROC
= 0.77; Hosmer-Lemeshow test: P = .83 [P > .05]).
Significant predictors for 30-day readmission for stroke
patients with a PEG tube placement were comorbidities,
hemorrhagic stroke, and presence of hemiplegia and/or
monoplegia (Table 7).

30-day readmission prediction model for stroke patients

without PEG tube placement. We tested the same model
for all-cause, unplanned 30-day readmissions in patients
without PEG tube placement. The Hosmer-Lemeshow test
showed a good model fit with a P value of .48 (P > .05). The
predictors stroke type, CCI, LOS, and discharge destination
were significant (P < .05; Table 8). The AUROC was 0.61
(Figure 9).
Adding the SASI to the prediction model to control for
stroke severity did not improve the discriminative ability of
the model (adding the total SASI score, Hosmer-Lemeshow
Figure 7. Area under the receiver operating curve (ROC) for
test: P = .81 [P > .05]; AUROC = 0.61; adding all SASI
the model to predict all-cause, unplanned 30-day readmissions
indicators separately, Hosmer-Lemeshow test: P = .42 [P > in stroke patients with a percutaneous endoscopic
.05]; AUROC = 0.61). gastrostomy tube placement during their index hospital stay.

Discussion
The purpose of our study was to better understand the
prognosis and trajectory of care of stroke patients who have
a PEG tube placement during their acute hospital stay. Here,
we sought to improve clinicians’ ability to counsel these
vulnerable and high-risk patients and their families when
deciding whether to place a PEG tube, and to identify needs,
plan resources, and prevent potential complications after
placement.
We found that in total only 1.7% of all stroke patients in
our study received a PEG tube during their index hospital
stay, compared with 5.22% of stroke patients admitted to
community hospitals in Florida during 2012 in a previous
study.4 A higher proportion of hemorrhagic than ischemic
stroke patients received a PEG tube (4.5% and 1.2%,
respectively). This is consistent with our previous find-
ings. However, the proportion of ischemic stroke patients
receiving a PEG tube during their index hospital stay
has been reported higher in other studies, with approx-
imately 8% of all ischemic stroke patients.21 The rather
small proportion in this study may be a result of several
factors. Through the availability and linkage of 3 years of Figure 8. Area under the receiver operating curve (ROC) for
MarketScan data, we were able to apply rigorous inclusion the model including the individual Stroke Administrative
and exclusion criteria (eg, exclusion of patients with a prior Severity Index indicators to predict all-cause, unplanned
stroke or prior diagnosis for dysphagia) that we had not 30-day readmissions in stroke patients with a percutaneous
endoscopic gastrostomy tube placement during their index
been able to apply in our previous study. Furthermore, the
hospital stay.
MarketScan databases only include commercially insured
Wilmskoetter et al 563

Table 7. Multivariable Logistic Regression Model Including the Individual SASI Indicators for the Event of an All-Cause
Unplanned 30-Day Readmission for Stroke Patients With a PEG Tube Placement During Their Index Hospital Stay.

Variable β SE β P Value OR (95% CI)

Intercept −1.66 0.84 .0475 NA

Hemorrhagic stroke 1.44 0.64 .0233 4.22 (1.22–14.65)
Charlson Comorbidity Index 0.48 0.18 .0067 1.62 (1.14–2.28)
Length of hospital stay (ࣙ11 d) −1.12 0. 64 .0817 0.33 (0.09–1.15)
Discharge to home 1.02 0.69 .1377 2.77 (0.72–10.63)
Discharge to rehab facility −1.21 0.72 .0917 0.30 (0.07–1.22)
Discharge to other destinations −1.72 1.18 .1444 0.18 (0.02–1.80)
SASI aphasia −0.75 0.83 .3704 0.47 (0.09–2.43)
SASI coma −0.17 1.27 .8916 0.84 (0.07–10.19)
SASI dysphagia and/or dysarthria 1.13 0.63 .0724 3.08 (0.90–10.52)
SASI hemi- and/or monoplegia −2.29 0.78 .0034 0.10 (0.02–0.47)

The 3 SASI indicators tracheostomy and/or ventilation, nutritional infusion, and neglect were not included in the model because ࣘ2 patients were
coded for these indicators. β, regression coefficient; NA, not applicable; PEG, percutaneous endoscopic gastrostomy; SASI, Stroke Administrative
Severity Index.

Table 8. Multivariable Logistic Regression Model for the Event of an All-Cause Unplanned 30-Day Readmission for Stroke
Patients Without a PEG Tube Placement During Their Index Hospital Stay.

Variable β SE β P Value OR (95% CI)

Intercept −2.27 0.13 <.0001 NA

Hemorrhagic stroke 0.32 0.11 .0028 1.38 (1.12–1.71)
Charlson Comorbidity Index 0.11 0.02 <.0001 1.11 (1.06–1.17)
Length of hospital stay (ࣙ11 d) 0.77 0.13 <.0001 2.16 (1.67–2.79)
Discharge to home −0.30 0.13 .0234 0.74 (0.58–0.96)
Discharge to rehab facility −0.50 0.16 .0018 0.61 (0.44–0.83)
Discharge to hospice −0.76 0.76 .3204 0.47 (0.11–2.09)
Discharge to other destinations 0.41 0.17 .0149 1.50 (1.08–2.08)

β, regression coefficient; NA, not applicable; PEG, percutaneous endoscopic gastrostomy.

patients and do not include patients from other providers,
such as Medicare or Medicaid. Consequently, the patients
in the cohort of our present study were younger overall
(54.37 years compared with 71.36 years), had fewer co-
morbidities (CCI was 1.00 compared with 3.04), shorter
LOS (mean 4.56 days compared with 5.62 days), and a
smaller proportion died during the index hospital stay
(3.77% compared with 5.77%). These differences suggest
that the MarketScan patient cohort was generally healthier,
which likely contributed to the lower rate of PEG tube
placements.22,23
In our study, patients with and without PEG tube place-
ment during their acute hospital stay differed with respect
to several medical characteristics. A higher proportion of
patients with PEG tube placement had hemorrhagic strokes,
more severe strokes (SASI), more comorbidities, and stayed
longer in the hospital than patients without PEG tube
placement. Patients who received a PEG tube were overall
in a more severe acute medical condition and, as a result of
this, may demand higher levels of specialized and intense
follow-up care, thus, supporting our focus on this group of
patients.
For our follow-up analyses, we observed a high dropout
of patients. At 12 months postdischarge, 35% of all patients
without and 61% of all patients with PEG tube placement
were no longer enrolled in the private insurance and conse-
quently lost for follow-up. The remarkably higher dropout
in the group of PEG tube patients likely reflects the higher
severity of this group. A dropout of the insurance can, for
example, be caused through job loss, causing the patient to
switch from the job-provided private insurance to another
insurance, such as Medicare.
Our main objective was to investigate the trajectory of
care for stroke patients with PEG tube placement. The
trajectory of care after the acute care hospital stay starts
with the discharge destination. Discharge destinations are
crucial for the outcome of stroke patients because they can
significantly impact patients’ recovery as a result of varia-
tions in treatment access. Of the patients with PEG tube
placement, 60%—compared with only 16.9% of patients
564 Nutrition in Clinical Practice 33(4)

without PEG tube placement—were directly discharged

from the acute care hospital to higher intense care/treatment
facilities, with 36.5% to IRFs and 24.3% being discharged
to SNFs. These discharge patterns differ from our previous
research of stroke patients discharged from community
hospitals in Florida in 2012. In our previous study, only
15.4% of stroke patients with a PEG tube placement were
discharged to IRFs and 54.3% to SNFs.14 Again, we specu-
late that this difference originates from differences in stroke,
age, and overall medical conditions. Whether a patient is
discharged to IRFs or SNFs is influenced by their recovery
potential. Patients with a better prognosis and who can
participate in a daily program of intense therapy (usually at
least 3 hours of rehabilitation training a day) will be more
likely to be discharged to IRFs. Thus, we speculate that
patients from the MarketScan cohort presented here had
a better recovery potential and prognosis based on factors
such as younger age and fewer comorbidities and, thus,
more patients were eligible for IRFs. Future research should
address factors that contribute to the decision making on
discharge destinations for stroke patients with PEG tubes.
When combining patients who were directly discharged
to IRFs and patients who were admitted later, we found that Figure 9. Area under the receiver operating curve (ROC) for
half of all stroke patients with a PEG tube placement had the model to predict all-cause, unplanned 30-day readmissions
a least 1 stay in an IRF within 1 year after their discharge in stroke patients without a percutaneous endoscopic
gastrostomy tube placement during their index hospital stay.
from the acute care hospital. Stroke patients with PEG
tube placement were also twice as likely to have at least
1 stay in an IRF and that stay was significantly longer example, Medicare reimburses SLP treatment for stroke
when compared with stroke patients without PEG tube patients in SNF, rehabilitation facilities, and for home
placement, after controlling for stroke severity, comorbidi- healthcare up to 100 days poststroke (www.medicare.gov).
ties, and other confounders. Thus, our findings suggest that In contrast, patients who are discharged to other destina-
patients with PEG tube placement are preferably discharged tions, had their stroke >100 days ago, or have private insur-
(or later admitted) to IRFs when compared with their ance do not necessarily receive coverage for SLP services.
counterparts without a PEG tube placement. This finding Unfortunately, the MarketScan database does not specify
warrants verification in other studies, but suggests that swallowing treatment—or SLP treatment in general—and,
healthcare providers identify the high need of rehabilitation thus we were not able to assess if and how often patients
in patients with PEG tube placement. However, still 50% received swallowing treatment.
of stroke patients with PEG tube placement did not receive Our findings suggest opportunities for those patients
rehabilitation treatment at an IRF. Further clarification is who may not receive rehabilitation through inpatient or
needed on whether these patients had received rehabilitation outpatient services. Our data showed that 35% of PEG
through another facility (eg, subacute rehabilitation in a tube patients were admitted to some kind of inpatient
SNF) or through outpatient rehabilitation. Because of the hospital stay within 1 month, meaning that 1 of 3 patients
associated risks, complications, and decrease in quality life, with an initial PEG tube placement will have access to
PEG tube removal in conjunction with swallow recovery is a inpatient healthcare professionals within a few weeks after
main goal of poststroke rehabilitation. Swallowing therapy discharge from their index hospital stay. Some of these
is an agreed-on facilitator of feeding tube removal, even in inpatient admissions were a result of complications and
chronic stroke patients.12,13,24-27 Therefore, any patient with worsening of the medical status with a recurrent stroke
a PEG tube who does not receive swallowing therapy is less being the most frequent readmission reason. Therefore,
likely to be weaned of their PEG tube and consequently is patients may not present their full recovery potential at that
at higher risks for PEG tube–associated complications than time. Nevertheless, we believe that the high frequency of
patients who receive swallowing therapy. inpatient encounters of PEG tube stroke patients could
Indeed, not all stroke patients receive swallowing treat- serve as an ideal opportunity to provide reevaluations of
ment because access to SLP treatment depends on multiple their swallowing ability and need of PEG tube. Some
factors, such as insurance and discharge destination. For patients may not receive reevaluations otherwise (as this
Wilmskoetter et al
service is often limited outside of hospitals) and, as a con-
sequence, will not be identified as candidates for swallowing
therapy or PEG tube removal.
Besides transfers to other hospitals, the largest propor-
tion of hospital admissions within 1 month after discharge
from the index hospital stay were unplanned all-cause
readmissions. Twice as many patients with PEG tube place-
ments (18%) had an all-cause, unplanned readmission when
compared with patients without PEG tube placement (9%).
These readmissions excluded rehabilitation, transfer, and
planned admissions. These findings replicate our previous
study of hospital readmissions of stroke patients who were
discharged from community hospitals in Florida in 2012,
where we also found that twice as many patients with PEG
tube placement (21.06%) were readmitted within 30 days
when compared with patients without PEG tube placement
(10.84%).14 Building on our previous analysis, we were able
to show that PEG tube placement is an independent predic-
tor increasing the odds 2-fold for 30-day readmissions even
after controlling for stroke severity and several other patient
and medical factors. Previous studies had to consider PEG
tube placement as a surrogate for stroke severity instead of
an independent predictor because of the lack of a proper
stroke severity measure. To our knowledge, our study is
the first study that can verify the independent contribution
of PEG tube placement to 30-day readmissions. The most
common causes for 30-day unplanned readmissions were
recurrent strokes (cerebral artery occlusions, hemorrhages,
transient ischemic attacks) in both groups of patients
with and without PEG tube placement. In our previous
study, cerebral artery occlusion and septicemia followed
by transient ischemic attacks were the leading causes of
hospital readmission for stroke patients without PEG tube
placement, and septicemia and food/vomit pneumonitis
followed by cerebral artery occlusion for stroke patients with
PEG tube placement.14 The slight differences in the most
common readmission diagnoses between our 2 studies indi-
cate limitations for generalizations across different patient
cohorts and datasets.
The 1-year care trajectory revealed that stroke patients
with PEG tube placement were 8 times more likely to have
a SNF stay after controlling for age, gender, comorbidities,
stroke severity, and stroke type. The reasons for this striking
difference between the 2 groups remain speculative. It is
to be expected that factors such as family support, level
of dependency, and recovery prognosis play a role in the
decision to admit patients to skilled nursing facilities. The
same factors might also affect the initial decision to place
a PEG in the acute care hospital stay. Unfortunately, we
were not able to assess these factors within the MarketScan
data. Future studies are warranted to investigate medical
and nonmedical factors contributing to poststroke facility
stays.
565
In a previous study, we were able to develop a prediction
model for 30-day all-cause, unplanned readmissions of
stroke patients who had PEG tube placement during their
index hospital stay. The model showed a good discriminative
capability as tested in stroke patients admitted to commu-
nity hospitals in Florida with an AUROC of 0.81.14 In the
study presented here, we were able to test our model in
a patient cohort of commercially insured younger stroke
patients with PEG tube placement. The model showed
a comparable discriminative capability for 30-day read-
missions with an AUROC of 0.76. Thus, we successfully
validated our prediction model in an independent dataset
with a different stroke patient population. Adding SASI to
our prediction model improved its discriminative ability for
30-day all cause unplanned readmissions for stroke patients
with PEG tube placement. However, the same model did
not show a satisfying discriminative ability for 30-day all
cause unplanned readmissions for stroke patients without
PEG tube placement. This was not surprising as the model
was originally developed for stroke patients with PEG
tube placement. Thus, our findings suggest that develop-
ing prediction models for hospital readmissions should be
population specific to reach the best possible discriminative
ability.
Limitations
Our study has several limitations that are a characteristic
to administrative databases. There is a potential for as-
certainment bias with possible mismatches between what
is coded in the database and true clinical events. As an
example, in our study, 5.7% of 8911 stroke patients were
coded for dysphagia at discharge from their index hospital
stay. Strikingly, this is in contrast to various clinical studies
that have shown that 50%–78% of all acute stroke pa-
tients present dysphagia.28,29 Undercoding of dysphagia in
archival data is a well-known problem30 that has prevented
us from including this variable in any of our outcome
analyses. Another example is the removal of PEG tubes.
Few patients were coded for PEG tube removal in either
inpatient, outpatient, or facility billing files. We suspect that
1 reason for this may be that PEG tube removal might not be
coded during an office visit or can happen accidentally apart
from a medical encounter. The absence of reliable coding
for PEG tube removal in administrative data excludes that
variable from analyses.
Furthermore, we had to use changing cohorts in the 1-
month, 3-month, 6-month, and 12-month follow-up analy-
ses as a result of patients dropping out of the commercial
insurance and, thus, being lost to longer term follow-up in
our analysis. As a consequence, a direct comparison of the
inpatient care that stroke patients received at the different
follow-up time points is not possible in these data.
566
Conclusions
To our knowledge this is the first study investigating the
care trajectory for stroke patients with and without PEG
tube placement using administrative databases. Our study
provides unique insights into differences between these 2
patient groups. The study results improve our understand-
ing of what clinicians and—more important—patients can
expect within 1 year following a stroke and can aide in
predicting recovery, planning resources, and identifying
gaps, risks, and points for improvement in stroke care
itself.
Statement of Authorship
Janina Wilmskoetter, Annie Simpson and Heather Bonilha
contributed to the conception and design of the research; Kit
Simpson contributed to the design of the research; Janina
Wilmskoetter, Annie Simpson, Sarah Logan and Kit Simp-
son contributed to the acquisition and analysis of the data;
Janina Wilmskoetter, Annie Simposn and Heather Bonilha
contributed to the interpretation of the data. Janina Wilm-
skoetter drafted the manuscript; all authors critically revised
the manuscript, agree to be fully accountable for ensuring the
integrity and accuracy of the work, and read and approved the
final manuscript.
References
1. Mozaffarian D, Benjamin EJ, Go AS, et al. Heart disease and stroke
statistics—2015 update: a report from the American Heart Association.
Circulation. 2015;131(4):e29-e322.
2. Centers for Disease Control and Prevention (CDC). Prevalence and
most common causes of disability among adults—United States, 2005.
MMWR Morb Mortal Wkly Rep. 2009;58(16):421-426.
3. Dennis M, Lewis S, Cranswick G, Forbes J. FOOD: a multicentre
randomised trial evaluating feeding policies in patients admitted to
hospital with a recent stroke. Health Technol Assess. 2006;10(2):iii-iv,
ix-x, 1–120.
4. Wilmskoetter J, Simpson AN, Simpson KN, Bonilha HS. Practice
patterns of percutaneous endoscopic gastrostomy tube placement in
acute stroke: are the guidelines achievable. J Stroke Cerebrovasc Dis.
2016;25(11):2694-2700.
5. Janes SEJ, Price CSG, Khan S. Percutaneous endoscopic gastros-
tomy: 30-day mortality trends and risk factors. J Postgrad Med.
2005;51(1):23-29.
6. Allison MC, Morris AJ, Park RH, Mills PR. Percutaneous endoscopic
gastrostomy tube feeding may improve outcome of late rehabilitation
following stroke. J R Soc Med. 1992;85(3):147-149.
7. Powell-Tuck J, van Someren N. Enterostomy feeding for patients with
stroke and bulbar palsy. J R Soc Med. 1992;85(12):717-719.
8. Scolapio JS, Romano M, Meschia JF, Tarrosa V, Chukwudelunzu FE.
PEG feeding tube placement following a stroke: when to place, when to
wait. Nutr Clin Pract. 2000;15(1):36-39.
9. Ickenstein GW, Kelly PJ, Furie KL, et al. Predictors of feeding
gastrostomy tube removal in stroke patients with dysphagia. J Stroke
Cerebrovasc Dis. 2003;12(4):169-174.
10. Yi Y, Yang EJ, Kim J, Kim WJ, Min Y, Paik NJ. Predictive factors for
removal of percutaneous endoscopic gastrostomy tube in post-stroke
dysphagia. J Rehabil Med. 2012;44(11):922-925.
Nutrition in Clinical Practice 33(4)
11. Wilmskoetter J, Herbert TL, Bonilha HS. Factors associated with
gastrostomy tube removal in patients with dysphagia after stroke. Nutr
Clin Pract. 2017;32(2):166-174.
12. Buchholz AC. Weaning patients with dysphagia from tube feeding
to oral nutrition: a proposed algorithm. Can J Diet Pract Res.
1998;59(4):208-214.
13. Crary MA, Groher ME. Reinstituting oral feeding in tube-fed adult
patients with dysphagia. Nutr Clin Pract. 2006;21(6):576-586.
14. Wilmskoetter J, Simpson KN, Bonilha HS. Hospital readmissions
of stroke patients with percutaneous endoscopic gastrostomy feeding
tubes. J Stroke Cerebrovasc Dis. 2016;25(10):2535-2542.
15. Centers for Medicare & Medicaid Services (CMS). Readmissions
Reduction Program (HRRP) 2016. https://www.cms.gov/medicare/
medicare-fee-for-service-payment/acuteinpatientpps/readmissions-
reduction-program.html. Accessed March 7, 2016.
16. Lichtman JH, Leifheit-Limson EC, Jones SB, et al. Predictors
of hospital readmission after stroke: a systematic review. Stroke.
2010;41(11):2525-2533.
17. Winstein CJ, Stein J, Arena R, et al. Guidelines for adult stroke
rehabilitation and recovery: a guideline for healthcare professionals
from the American Heart Association/American Stroke Association.
Stroke. 2016;47(6):e98-e169.
18. Simpson AN, Wilmskoetter J, Hong I, et al. Stroke administrative
severity index: Using administrative data for 30-day poststroke out-
comes prediction. J Comp Eff Res. 2017. https://doi.org/10.2217/cer-
2017-0058 [Epub ahead of print].
19. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson
Comorbidity Index and score for risk adjustment in hospital discharge
abstracts using data from 6 countries. Am J Epidemiol. 2011;173(6):676-
682.
20. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for
defining comorbidities in ICD-9-CM and ICD-10 administrative data.
Med Care. 2005;43(11):1130-1139.
21. George BP, Kelly AG, Schneider EB, Holloway RG. Current practices
in feeding tube placement for US acute ischemic stroke inpatients.
Neurology. 2014;83(10):874-882.
22. Dubin PH, Boehme AK, Siegler JE, et al. New model for predicting
surgical feeding tube placement in patients with an acute stroke event.
Stroke. 2013;44(11):3232-3234.
23. Li J, Zhang J, Li S, Guo H, Qin W, Hu WL. Predictors of percutaneous
endoscopic gastrostomy tube placement after stroke. Can J Neurol Sci.
2014;41(1):24-28.
24. Carnaby G, Hankey GJ, Pizzi J. Behavioural intervention for dys-
phagia in acute stroke: a randomised controlled trial. Lancet Neurol.
2006;5(1):31-37.
25. Huckabee ML, Cannito MP. Outcomes of swallowing rehabilitation in
chronic brainstem dysphagia: A retrospective evaluation. Dysphagia.
1999;14(2):93-109.
26. Shaker R, Easterling C, Kern M, et al. Rehabilitation of swallowing by
exercise in tube-fed patients with pharyngeal dysphagia secondary to
abnormal UES opening. Gastroenterology. 2002;122(5):1314-1321.
27. Elmstahl S, Bulow M, Ekberg O, Petersson M, Tegner H. Treatment of
dysphagia improves nutritional conditions in stroke patients. Dyspha-
gia. 1999;14(2):61-66.
28. Martino R, Foley N, Bhogal S, Diamant N, Speechley M, Teasell R.
Dysphagia after stroke: incidence, diagnosis, and pulmonary compli-
cations. Stroke. 2005;36(12):2756-2763.
29. Daniels SK, Foundas AL. Lesion localization in acute stroke patients
with risk of aspiration. J Neuroimaging. 1999;9(2):91-98.
30. Gonzalez-Fernandez M, Gardyn M, Wyckoff S, Ky PK, Palmer JB.
Validation of ICD-9 code 787.2 for identification of individuals with
dysphagia from administrative databases. Dysphagia. 2009;24(4):398-
402.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Strategies for the Enhancement of Nutrition Practice August 2018 567–575

C 2017 American Society for

in a New York State Level 1 Trauma Center: A Hospital’s Parenteral and Enteral Nutrition
DOI: 10.1177/0884533617724144
Journey wileyonlinelibrary.com

Lisa Musillo, MS, RDN, CNSC1 ; Laryssa Marie Grguric, RDN, CNSC1 ;
Edward Coffield, PhD2 ; Frank Aversano, MD3 ; Jeremy Bosworth, MD3 ;
and Richard Batista, MD3

Abstract
Background: Provision of enteral nutrition (EN) support is historically inadequate in the critically ill population. An interdisci-
plinary approach utilizing various strategies has been shown to improve initiation of timely EN support. The purpose of this
study was to examine whether the implementation of a series of interventions led by an interdisciplinary team was associated
with changes in the initiation of nutrition support in a level 1 trauma center. Methods: Patients admitted between 2009 and 2013
with isolated closed head trauma injuries were identified through the hospital’s trauma center database. The initial population
consisted of 159 patients; after exclusion criteria, 141 patients were included in the statistical analyses. Two statistical analyses
were conducted. The first calculated the average days to the initiation of nutrition start by admission year. The second estimated
the association between admission year and time to nutrition initiation with a generalized linear model. Results: Time to initiate
nutrition therapy was estimated to decrease by 1.46 days (47.31%) from 2009 to 2013. The time to initiate nutrition in 2013 was
1.63 days. A significant association was found between the time to initiate nutrition and the 2012 and 2013 binary variables while
controlling for confounding variables. The time frame was estimated to be 1.09 (P = .008) and 1.75 (P = .000) days shorter in 2012
and 2013 relative to 2009. Conclusions: An interdisciplinary effort utilizing multiple strategies identified and addressed barriers,
resulting in a reduction of variability and a proactive approach to early EN. (Nutr Clin Pract. 2018;33:567–575)

Keywords
clinical protocols; closed head injuries; critical illness; enteral nutrition; nutritional support

Nutrition during critical illness is 1 of many therapies
that influence patient outcomes. As described in the 2016
guidelines for the provision and assessment of nutrition
support therapy, “delivery of early nutrition support ther-
apy, primarily by the enteral route, is seen as a proac-
tive therapeutic strategy that may reduce disease severity,
diminish complications, decreased length of stay in the
intensive care unit (ICU) and favorably impact patient
outcomes.”1 The initiation and provision of nutrition during
clinical treatment fall well below established benchmarks.1,2
Addressing the deficiencies in critical care nutrition support
requires universal and local strategies to minimize and
possibly prevent complications.3,4
From a local or single hospital perspective, the strategies
within our ICU to address nutrition insufficiency during
critical care were developed by an interdisciplinary team
and implemented and refined over a multiyear period.
After the critical care dietitian identified deficiencies in
nutrition provision, it was apparent that study participation
was necessary to compare the facility against international
nutrition practice. The dietitian then became instrumental
in arrangement of survey participation in a multisite in-
ternational survey. This facility participated in the 2007
study “Improving the Practice of Nutrition Therapy in the
Critically Ill: An International Quality Improvement Project
(IQIP).”5 Participation in this project allowed an assessment
of our processes for initiating and providing adequate
nutrition therapy to our patients, relative to other sites. Our
performance in this survey was used as a springboard for
change within our facility.
From the 1 Department of Food and Nutrition, Nassau University
Medical Center, East Meadow, New York, USA; 2 Department of
Health Professions, Hofstra University, Hempstead, New York, USA;
and 3 Department of Surgery, Nassau University Medical Center,
East Meadow, New York, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on September 27, 2017.
Corresponding Author:
Lisa Musillo, MS, RDN, CNSC, Department of Food and Nutrition,
Nassau University Medical Center, 2201 Hempstead Tpk, East
Meadow, NY 11554, USA.
Email: lmusillo@numc.edu
568 Nutrition in Clinical Practice 33(4)

Table 1. Challenges in Providing Adequate Nutrition Therapy in a Level 1 Trauma Center.

Challenge Description

Knowledge deficit r Inadequate knowledge of nutrition therapy policies and practices across all disciplines
r Unnecessary delays and/or withholding of tube feeding based on gastric residual volume
r Inadequate advancement of tube feeding per provider order of “as tolerated” or “ramp up”
r Infrequent use of promotility agents to enhance tube feeding tolerance
r Insufficient nursing communication between shifts regarding tube feeding tolerance and advancement
r Infrequent usage of rectal tube system, prompting withholding of feeding for diarrhea
r Lack of interest or awareness of the importance of timely nutrition therapy
Time lost to r Lost hours/days to anticipated extubation, lengthy tests and procedures/operations
clinical care r Preoperative/procedure fasting
r Canceled cases resulting in unnecessary fasting
r Tube feeding practices during hemodialysis and sustained low-efficiency dialysis
Equipment issues r Tube feeding pump availability and usage issues
r Lack of availability of jejunally placed feeding tubes

This retrospective study details the processes, challenges,
and barriers that we encountered while identifying barriers
to adequate nutrition therapy (Table 1) and developing
and implementing interventions intended to address these
barriers (Table 2) through an interdisciplinary team ap-
proach. The association between the interventions and time
to nutrition initiation was examined with a sample of 141
admitted patients with isolated closed head trauma injuries
from 2009 to 2013. The results from this study illustrate
how an internal focus on quality improvement, led by
an interdisciplinary team, has changed our practices to
improve nutrition support and patient care.
Background and Changes
Nutrition provision challenges within our facility were first
identified with IQIP. The IQIP study results were released
to our hospital administration in 2007, which prompted
the formation of an interdisciplinary team consisting of
members from the nutrition, surgery, medicine, and nursing
departments. The critical care dietitian was the lead team
member. This team was tasked with identifying barriers
to nutrition therapy and then formulating a quality im-
provement plan. After identification of barriers to nu-
trition implementation in 2008, intervention mechanisms
were applied to improve patient care. The team evaluated
progress as time passed, and it adjusted the improvement
interventions as required. Hospital leadership and team
members supported this evaluation-intervention-evaluation
approach, although it was challenging from time and re-
source perspectives.
The team identified a number of barriers to nutrition
therapy through IQIP (eg, time to initiation of nutrition
therapy and adequacy of provision; Table 1). The main
barriers were inadequate nutrition education across all
disciplines, a lack of awareness regarding nutrition initiation
and/or adequacy, access to feeding devices and tools, and the
delay/interruption of nutrition provision due to common
ICU events. Examples include patients having undergone
major surgery or trauma, immediate preextubation and pos-
textubation, patients receiving dialysis, patients with severe
sepsis requiring vasopressor support, perceived high gastric
residual volumes (GRVs), infrequent use of promotility
agents, lack of diet advancement, inefficient staff commu-
nication, no postpyloric feeding tube access, and lack of
enteral nutrition (EN) pump availability.4 The interventions
intended to address these barriers are listed in Table 2. For
the IQIP study, each participating center identified its goal
provision/adequacy based on its facility’s standards.
One of the first interventions, introduced in 2008, ad-
dressed the education component. Hospital staff were pro-
vided with nutrition lectures conducted by the critical care
dietitian. These lectures stressed the importance of prompt
initiation and advancement of nutrition therapy. In 2008,
the hospital also introduced “tube feeding pump centers.”
The purpose of these centers was to increase access to and
availability of EN pumps and tube feeding products and
accessories. It was hypothesized that facilitated access via
these pump centers would minimize the amount of time that
staff allocated to search for the required items to initiate and
maintain nutrition support.
The third intervention introduced in 2008 was the pur-
chase, by the Department of Surgery, of an electromag-
netic feeding tube placement device to facilitate postpyloric
feeding tube placements.6 The purchase of this device was
intended to increase the ease and efficiency of nasojeju-
nal/orojejunal tube placements. The expectation was that
the device would facilitate accessibility and efficacy and
thereby increase the quantity of feeding tubes placed at
an earlier time. Placements were done by surgical residents
Musillo et al 569

Table 2. Nutrition Therapy Quality Improvement Tasks and Interventions by Admission Year.

Year Task/Intervention

2007 Participation in “Improving the Practice of Nutrition Therapy in the Critically Ill: An International Quality
Improvement Project” (IPNT)
2008 IPNT results released and disseminated to administration
From IPNT results, identified deficiencies in nutrition therapy practice relative to other sites and critical care guidelines
Multidisciplinary team formed
Identification of barriers to adequate nutrition therapy
Development and implementation of improvement plan
Tube feeding pump centers created
Purchase and introduction of electromagnetic feeding tube placement device to assist in placement of jejunally placed
feeding tubes at bedside
Updated preoperative fasting policy
Nutrition therapy educational lectures for all clinical staff
Preoperative fasting guideline liberalized to 2 h with jejunally placed feeding tubes (Oct 2009)
2009 New physician hired as director of Department of Medicine and Critical Care Committee
New director of Department of Medicine advocated for change in nutrition practices (Nov 2009)
2010 Critical care nutrition snapshot (eg, review of practices) completedResults presented at Critical Care Committee
meetings
Repeat data collection presented at each meeting moving forward (Feb 2010–Mar 2010)
“Pending verification” order status added, allowing registered dietitian nutritionists to place nutrition-specific orders
into the electronic medical record for physician approval (Mar 2010)
2011 Enrollment in the “Enhanced Protein-Energy Provision via the Enteral Route in Critically Ill Patients (PEP uP
Protocol)” trial (Jul 2011)
2012 Official “rollout” of modified feeding protocol to burn center, coronary care unit, medical intensive care unit, surgical
intensive care unit (fall 2011–winter 2012)
Preoperative fasting guideline further liberalized to no tube feeding withholding when provided by jejunally placed
feeding tubes (Oct 2012)
2013 Registered dietitian nutritionists granted privilege to place jejunally placed feeding tubes using electromagnetic feeding
tube placement device (fall 2013)

(in conjunction with a surgical technician) or surgical
physician’s assistants under the supervision of a surgical
attending.
The intervention mechanisms implemented in 2008 were
maintained once implemented and included continued nu-
trition therapy education lectures. The continuation of these
education sessions provided new staff with nutrition therapy
education and updated existing staff on any procedure or
policy change. The educational sessions were seen as a key
component to ensure that all staff members worked together
to improve the timeliness and adequacy of nutrition therapy.
This culture was further supported by the addition of a
new Department of Medicine chair in 2009, who strongly
advocated for improvements in nutrition therapy for the
critically ill based on our hospital’s performance in the IQIP
study. The final intervention implemented in 2009 was an
update of the anesthesia preoperative fasting policy. This
updated policy shortened the time frame of EN withholding
to 2 hours prior to any surgical or invasive procedure for
patients with jejunally placed feeding tubes.7-9 Prior to this
update, the withholding time was a minimum of 8 hours.
In 2010, the team conducted a systematic review of
the nutrition therapy program. This included a repeat
data collection to compare with preintervention outcomes.
Improvement was evident; however, the team identified fur-
ther improvements that were needed to achieve acceptable
benchmarks. One of the first identified steps needed to fur-
ther improve nutrition initiation was a more constant data
surveillance procedure. Moving forward, data collections
were completed every 2 months by the Nutrition Service.
The key metrics collected were the time to initiate nutrition
and provision with an EN pump. Nutrition provision data
were collected with EN pumps with long-term memory
features, which could be accessed to collect the actual
volume of EN delivered. The use of the EN pumps for this
purpose also initiated the need to ensure the accuracy of the
data collected. There are discrepancies between documented
EN delivery and the actual amount provided to patients.10
The bimonthly time to initiation of nutrition therapy data
was started in 2010 and is ongoing and regularly presented
at committee meetings.
In an attempt to further improve outcome metrics, an
additional change in 2010 was a “pending verification”
order status in the electronic medical record (EMR). The
team worked with the information technology department
to develop an order entry field that allowed registered
570
dietitian-nutritionists (RDN) to input nutrition-related
orders (eg diets, tube feedings, and supplements). This
process was expected to improve the timing of diet order
changes and the implementation of nutrition support and
decrease the amount of nutrition-related order errors.
The review of outcomes and implementation of inter-
ventions or adjustments to existing practices continued in
2011. In this year, permission was granted by the hospital’s
administration for participation in a multicenter cluster-
randomized trial entitled “Enhanced Protein-Energy Pro-
vision via the Enteral Route in Critically Ill Patients: A
Single Center Feasibility Trial of the PEP uP Protocol.”11
This feeding strategy was different from the method of
EN administration being used at the time in the ICU.
The standard method was a continuous static rate with no
compensation for lost volume due to EN withholding. The
new protocol implemented a volume-based feeding strategy.
The purpose of the PEP uP protocol was to expedite
the initiation of nutrition support and adequate calorie and
protein provision for critically ill patients in an attempt to
improve patient outcomes. The protocol emphasizes early
enteral access, initiation and formula choice, dosing by
volume with patient weight, trophic EN during inotropic
therapy, built-in method of infusion rate changes based on
the actual volume delivered, proactive use of prokinetic
agents, and the use of GRVs and diarrhea algorithms. The
PEP uP protocol also empowered nursing to achieve the
desired daily EN goal. Institutional Review Board approval
was obtained, and the study commenced in summer of
2011. The medical ICU was the pilot location. The team
received positive feedback on the PEP uP protocol during its
testing. The results from the formal PEP uP protocol study
illustrated improvement in adequacy and total nutrition
across all study sites.11
After completion of the trial, the team decided to use
a modified version of the PEP uP protocol with different
enteral formulas and weight-based volumes.11 The modified
enteral feeding protocol (EFP) was implemented and ex-
panded to all critical care areas (surgery, coronary care, and
burn center) after a successful rollout in the medical ICU by
late fall/early winter of 2012. As part of the implementation,
a new EMR order panel was developed to assist doctors in
correctly ordering the EFP. A “nurse champion” is impera-
tive to the success of the protocol. A nurse champion was
designated to assist with continuous protocol coordination
and to help educate staff on appropriate protocol use. A
breach in EFP would trigger a reeducation to the staff
involved to reinforce concepts of the protocol provided
by the nurse champion. As this is largely a nurse-driven
protocol, a strong leader in nursing is imperative to success.
Further changes were made to the anesthesia preopera-
tive fasting policy in 2012. The policy was liberalized to state
that EN withholding is unnecessary with jejunally placed
feeding tubes. This change was based on evidence-based
Nutrition in Clinical Practice 33(4)
research.7-9 Coming into 2013, the dietitians were granted
privileges to place postpyloric feeding tubes with the elec-
tromagnetic feeding tube device. This was implemented to
increase the number of feeding tubes placed in a timely
manner.
The changes implemented throughout the quality im-
provement initiative were intended to improve the hospital’s
nutrition practice. This study examined the timing of nutri-
tion support in an isolated closed head injury population to
examine if a relationship existed between quality improve-
ment measures implemented and current nutrition practice.
Specifically, the study examined whether time to initiate
nutrition was associated with binary year variables, which
presumably measured differences in nutrition practice over
time. The interventions were viewed as a collective whole
and were examined as such, although days to nutrition
initiation were estimated annually across the examined
period.
Methods
This was an expedited study approved by the Nassau
University Medical Center Institutional Review Board. The
sample population was drawn from a 500-bed regional
level 1 trauma center in Nassau County, New York. For
inclusion in this retrospective study, patients had to be
admitted between 2009 and 2013 with isolated closed head
trauma injuries. Patients meeting these criteria were iden-
tified through the hospital’s trauma center database. Once
identified, the medical records of each patient were reviewed
to confirm that each patient’s diet information was complete
and accurate. Required information included type of diet
(ie, oral, nonoral [EN vs parenteral nutrition]) and time to
initiation of nutrition therapy.
Time to initiation of nutrition was recorded in number
of days and calculated from the time that patients were
admitted to the ICU to the time that nutrition therapy was
initiated. The time-based nature of this variable entailed
that it was in noninteger form (eg, 1.4 days). Clinical and
demographic variables used in the analyses were extracted
from the patients’ medical records. Per admission dates, each
patient was assigned 1 of 5 binary variables indicating the
patient’s admission year (2009–2013). Each of the previ-
ously discussed quality improvement measures aligned with
1 of the admission year periods (Table 2). Accordingly, the
admission year binary variables allowed an examination of
whether the quality improvement measures were associated
with changes in the time to initiate nutrition. Specifically, did
days to the initiation of nutrition differ by admission year?
Two statistical analyses were conducted. The first anal-
ysis calculated the average days to start of nutrition by
admission year. This analysis allowed a preliminary inves-
tigation into whether the initiation of nutrition changed
annually from 2009 to 2013. The second analysis estimated
Musillo et al 571

the association between the binary admission year variables Table 3. Descriptive Statistics for 141 Patients With Isolated
and time (measured in days) to nutrition initiation through Closed Head Trauma Injuries Who Received Nutrition
a generalized linear model (GLM) with a γ distribution and Therapy.a
identity link. A GLM with a γ distribution and identity Variable Percentageb SD Min Max
link was used in the second analysis; using an ordinary
least squares for this analysis may result in faulty estimates Age, y, mean 50.80 21.14 19 92
due to the strictly positive nature of the response variable <65 73.76 44.15
(days to nutrition initiation). The GLM-based analysis ࣙ65 26.24 44.15
allowed an investigation into whether the annual quality Sex
Female 21.28 41.07
improvement measures (binary admission year variables)
Male 78.72 41.07
were associated with changes in days to nutrition initiation, Race
while controlling for confounding factors.12 African American 17.02 37.72
The primary independent variables in the GLM were the Otherc 3.55 18.56
binary admission years. The base admission year was 2009. White 79.43 40.56
Accordingly, if the annual quality improvement measures Ethnicity
were associated with decreases in time to nutrition therapy Non-Hispanic 82.98 37.72
Hispanic 17.02 37.72
initiation, the coefficient on the other binary admission year
Admission year, mean 2011 1.38 2009 2013
variables should be negative. Covariates included in the 2009 20.57 40.56
GLM were age (ࣙ65 years), race (African American, other, 2010 21.28 41.07
white), Hispanic ethnicity, inpatient mortality, number of 2011 21.28 41.07
comorbidities and complications, diet type (oral, nonoral), 2012 20.57 40.56
Glasgow Coma Score, injury severity score, admission unit 2013 16.31 37.08
(surgical, nonsurgical), and cause of injury (fall, motor Clinical outcome
Alive 73.76 44.15
vehicle accident, other, unknown). The covariates were
Deceased 26.24 44.15
included to control for other factors that may influence time Diet type
to nutrition therapy initiation. All statistical analysis were Oral 18.44 38.92
conducted in STATA 14.12 Nonorald 81.56 38.92
Clinical variables, mean
Days to nutrition 2.67 1.57 0.39 8.65
Results initiatione
Total comorbidities 4.74 1.88 0.00 6.00
The initial population consisted of 159 patients. However,
Total complications 1.47 1.58 0.00 6.00
18 of these patients were removed due to incomplete diet Glasgow Coma Scale 9.71 4.62 3.00 15.00
documentation (n = 10), incorrect classification of admis- Injury Severity Score 26.23 11.79 4.00 75.00
sion to the adult ICU (n = 4), and incomplete covariate Cause of injury
information (n = 4). Statistical differences were found be- Otherf 31.21 46.50
tween the omitted patients and sample patients on a number Fall 36.17 48.22
of factors. As illustrated in Table 3, the majority of the Motor vehicle accidentg 24.82 43.35
Unknown 7.80 26.92
analytic sample was male (73.76%) with an average age of
50.80 years. Approximately 20% of the patients identified as a Statisticswere calculated on a sample with full covariate information
nonwhite and 17.02% as being of Hispanic ethnicity. From (N = 141).
b Values are presented as percentages unless noted otherwise.
a clinical perspective, the average number of comorbidities c Individuals who identified as American Indian/Eskimo/Aleut, Asian
among the sample was 4.74, and 26.24% of the sample died
or Pacific Islander, or other.
while in the facility. d Individuals who received enteral or parenteral feeding.

The majority of patients (81.56%) received EN or par- e The difference in days between when patients were admitted and

enteral nutrition, and the average time to initiation of
nutrition therapy was 2.67 days overall. As illustrated in
Table 4, the average days to nutrition initiation steadily
decreased with each succeeding year from 2009 to 2013.
The average days to nutrition initiation was 3.095 in 2009,
while it was 1.631 days in 2013. This 47.31% reduction in
days to the initiation of nutrition therapy could be due to
a number of factors (eg, differences in patients’ injury types
or comorbidities between the admission years) beyond the
quality improvement measures. The GLM was intended to
when they began receiving nutrition.
f Individual’s injury caused by assault, bicyclist/pedestrian struck, boat
accident, cutting instrument, fight/brawl, found down, hanging,
railway accident, sports, penetrating trauma (knife, gun, impalement).
g Individual’s injury caused by motor vehicle collision.
control for many of these factors and isolate the association
between the changes in nutrition therapy practice (Table 2),
measured through the binary admission year variables, and
days to nutrition initiation.
572 Nutrition in Clinical Practice 33(4)

Table 4. Mean Days to Nutrition Initiation for 141 Patients Table 5. Association Between the Number of Days to
With Isolated Closed Head Trauma Injuries Stratified by Nutrition Initiation and the Admission Year and Other
Admission Year.a Characteristics.a

Yearb No. Meanc SD Median Min Max Variable (Referent) Coefficient P Value

2009 29 3.095 1.43 2.91 0.98 7.18 Admission yearb (2009)

2010 30 3.288 1.64 2.74 1.12 6.94 2010 −0.47 .220
2011 30 2.985 1.78 2.70 0.74 8.65 2011 −0.72 .089
2012 29 2.094 1.29 1.78 0.53 5.82 2012 −1.09 .008
2013 23 1.631 0.95 1.54 0.39 4.44 2013 −1.75 .000
Age group (<65 y): ࣙ65 y 0.20 .564
a Nutrition initiation: the difference in days between when patients Sex (male): female −0.21 .465
were admitted and when they began receiving nutrition therapy. Race (white)
Statistics were calculated on a sample with full covariate information African American 0.18 .568
(N = 141). Otherc 0.97 .285
b Admission year or year of treatment is intended to indicate the
Ethnicity (non-Hispanic): Hispanic −0.33 .289
quality improvement interventions that were implemented in the year
that a patient was admitted or treated. Clinical outcome (alive): deceased −0.32 .252
c Means are rounded, which adjusts the percentage change statistic Diet type (nonorald ): oral 1.15 .002
cited in the article between 2009 and 2013. Based on these rounded Clinical variables
figures, the percentage change between these 2 years is 47.30%; Total comorbidities −0.11 .119
unrounded, the percentage change is 47.31%. Total complication 0.06 .535
Glasgow Coma Scale 0.04 .070
Injury Severity Score 0.01 .387
The GLM results are illustrated in Table 5. The re- Cause of injury (fall)
sults are reported in days. For instance, holding all else Othere −0.20 .504
constant, it was estimated that the time to the initiation Motor vehicle accidentf 0.20 .638
of nutrition therapy was 1.09 days (P = .008) shorter in Unknown −0.91 .037
2012 relative to 2009. Likewise, it was estimated that the Constant 3.15 .000
time to the initiation of nutrition therapy was 1.75 days a Coefficients (days) estimated from a generalized linear model with a
(P = .000) shorter in 2013 relative to 2009. These 2 results γ distribution and identity link; the sample included 141 patients with
support the hypothesis that an association exists between isolated closed head trauma injuries who received nutrition therapy.
the implemented quality improvement interventions and Nutrition initiation: the time difference (in days) between when
patients were admitted and when they began receiving nutrition
days to initiation of nutrition. Although the coefficients for
therapy.
the other 2 binary year variables (2010, 2011) were in the b Admission year or year of treatment is intended to indicate the
expected direction, a significant association was not found quality improvement interventions that were implemented in the year
between days to nutrition initiation and these 2 variables. that a patient was admitted or treated.
c Individuals who identified as American Indian/Eskimo/Aleut, Asian
The results also illustrate that diet type was associated with
or Pacific Islander, or other.
days to nutrition initiation. Patients who received their first d Individuals who received enteral or parenteral feeding.

feeding orally were estimated to have a time to nutrition e Individual’s injury caused by assault, bicyclist/pedestrian struck, boat

initiation that was 1.15 days (P = .002) longer than patients accident, cutting instrument, fight/brawl, found down, hanging,
railway accident, sports, penetrating trauma (knife, gun, impalement).
whose first feeding was administered nonorally. f Individual’s injury caused by motor vehicle collision.
The results from the GLM support the trends illustrated
in Table 4; the difference between these 2 analyses (eg, no
statistical differences in days to nutrition therapy initiation
closed head trauma injuries. There was a 47.31% reduction
between 2009 and 2010 or 2009 and 2011 in the GLM mod-
(or 1.46 days) in the time to initiate nutrition therapy from
els) is likely due to the inclusion of confounding factors in
2009 to 2013. The time to initiate in 2013 was 1.631 days, an
the GLM. Results from an ordinary least squares regression
outcome consistent with recommendations for EN to begin
model (Appendix Table A1) were similar to those estimated
within 2 days.2 These changes are also in line with the new
in the GLM. However, the significance of the coefficient on
2016 nutrition critical care guidelines,1 which recommends
the 2012 binary year was sensitive to the patients included
a volume-based feeding or top-down multistrategy protocol
in the ordinary least squares regression.
be used. Protocols such as these empower nursing to alter
the run rates to achieve the goal volume by the end of the day
Discussion
instead of traditional static run rates, which were previously
Our multi-intervention strategy has demonstrated signifi- used in our facility.13,14
cant improvement, starting in 2012, in the timeliness of Over the course of 5 years, different strategies were
initiation of nutrition among a population with isolated implemented and maintained (Table 2), and their effects
Musillo et al
here are demonstrated in a cumulative fashion. Although
significant associations were found between the days to
nutrition initiation and the 2012 and 2013 binary admission
year variables, these associations are likely due to all inter-
ventions and not just the interventions that occurred in these
2 years. In other words, significant changes that were found
are likely due to a comprehensive effect—one that builds
on the interventions implemented in previous years, instead
of just the interventions introduced in 2012 and 2013.
Heyland et al articulated a strategy to begin to motivate
surgeons and other providers to overcome the prevailing
persistence to change traditional or historic practice and
adopt evidence-based perioperative nutrition therapy, thus
narrowing the knowledge practice gap. The 8 phases of
the action cycle discussed various activities needed to move
knowledge into practice. Our interventions were not based
on this model, but many interventions and processes mirror
the cycle discussed here. This would be a good resource in
addition to this article to bring change to other facilities’
historic practice.15 Although not examined with these data,
the largest barrier to overcome when implementing the new
protocol was historic practice by nursing in regard to EN
“ramp up” protocols and use of static run rates. It was
also challenging to overcome fears regarding perceived high
GRVs and “high” run rates. After a number of education
sessions, staff were seemingly less reluctant to implement
change, due to gained experience with the new protocols.
Stooling has historically been a barrier to EN provision.
However, with the use of improved fecal management
systems and diarrhea-specific protocols, there has been a
concomitant decrease in nurses’ hesitation to provide EN.
It is imperative to have the support of nursing and hos-
pital administration to effect change and maintain success
while implementing similar protocols. Nurse champions
who understand and appreciate the importance of timely
and adequate nutrition are essential in the change of culture
within the critical care arena.
Another barrier to initiation of EN in the ICU was the
use of antiquated preoperative fasting policies. Patients were
maintained nil per os with a minimum of 8 hours prior
to any interventional procedure. At times, the procedures
were delayed or postponed if nutrition was not withheld
within a designated 8-hour time frame, thereby increasing
fasting time. With placement of a postpyloric feeding tube,
invasive procedures were not delayed, due to preoperative
fasting protocols by anesthesia. The purchase and use of
a bedside electromagnetic feeding tube device allowed for
expedited postpyloric feeding tube placement. This in turn
allowed for EN to be provided with minimal need for x-
ray confirmation, with no required preoperative fasting,
and with less EN withholding for “high” GRVs or other
intolerance issues.
From 2013 to present, the strategies implemented over
the examined years culminated into a proactive nutrition
573
environment. Strengths of this quality improvement process
include that it was an interdisciplinary effort with numer-
ous stakeholders assisting to elicit ongoing improvements.
Implemented changes were supported by the group and
maintained from inception. In addition, despite the results
of this study and the possible improvement in our hospital’s
nutrition policies and practices, the continuous improve-
ment nature of our system entails that the monitoring, ed-
ucation, and search for other ways to improve our nutrition
policies and practices continues. For instance, deficiencies
were identified in the protocol’s order panel in the EMR,
and changes were made to optimize protocol compliance
with the order panel. Working as an interdisciplinary team,
hospital staff continue to identify deficiencies and provide
corrections to illustrate that this is an ongoing and fluid
process.
While this was a retrospective study to examine the
improvement measures performed at this facility, the results
presented in this analysis are subject to a number of limi-
tations. Importantly, the results are based on associations,
not causation. While attempts were made to isolate the
relationship between days to nutrition initiation and the
interventions, we were unable to directly connect changes in
days to initiation to the discussed intervention. In addition,
binary admission year variables may serve as indicators
for factors beyond the interventions, such as unmeasured
changes within the hospital during the examined year.
However, in the absence of a control group or hospital, this
rather noisy method was used. The single-center nature of
this study is thus an additional limitation.
Many of the strategies used here were demonstrated
to be effective in other facilities.3,6,8,11 We “fine-tuned”
the interventions to best fit the needs of our hospital.
Individual centers should examine their present cultures
and own needs prior to implementing new interventions.
Overall, it is important to note that the results of this study
may be due to this facility and its own evolving nutrition
culture.
Conclusion
Our participation in board-approved research projects has
shaped our evidence-based best practice, and significant
change has occurred as a result. Barriers to adequate nutri-
tion provision were identified and addressed, resulting in a
reduction of variability and a proactive approach to early
nutrition initiation. This approach was novel in the fact that
the process began with just the identification that an issue
existed in the timing of initiation of nutrition. A cascade
of interventions ensued based on identified needs of the
facility. In retrospect, it may appear the interventions were
not novel; however, they were in 2007 in this facility. Many
hospitals continue to struggle with the barriers that we
have identified, and they are still searching for their specific
574 Nutrition in Clinical Practice 33(4)

solutions. This topic needs to continue to be discussed in 4. The interdisciplinary team approach was important.
the literature to help bring awareness and foster change It is recommended that other systems follow a team-
to abandon historic practices. Optimization of nutrition based approach that develops solutions based on
therapy requires the coordination of multiple disciplines. their needs. The interventions introduced here were
The interdisciplinary effort applied in our facility was a specific to the needs and identified barriers of this
key component of our efforts to improve nutrition therapy examined facility.
practices. Increased communication among team members
regarding nutrition goals and actual support provided Appendix
has increased. A culture change has occurred over time. Table A1. Association Between the Number of Days to
This nurse-directed protocol, with associated changes, has Nutrition Initiation and the Admission Year and Other
become second nature to the critical care team. Elevation of Characteristics.a
the RDN within the institution is imperative for successfully
improving the quality of nutrition care. The RDN’s scope Variable (Referent) Coefficient P Value
of practice needs to be expanded to include contribution to Admission yearb (2009)
evidence-based nutrition practice, feeding tube placements, 2010 −0.31 .496
physical assessment, and nutrition-specific order entry in 2011 −0.46 .314
the EMR for best outcomes. Physician knowledge regarding 2012 −1.08 .042
nutrition has been enhanced throughout this process with 2013 −1.53 .000
resulting improvement in nutrition therapies. The RDN has Age group (<65 y): ࣙ65 y 0.51 .137
Sex (male): female 0.06 .860
also been recognized by other healthcare team members as
Race (white)
an integral part of total patient care. African American 0.18 .654
Nutrition therapy is now viewed in a proactive manner, Otherc 1.75 .044
striving for enhancement in the time frame from the time Ethnicity (non-Hispanic): Hispanic 0.10 .763
of admission to the start of EN. EN provision has been Clinical outcome (alive): deceased −0.47 .222
significantly improved despite the multiple common barriers Diet type (nonorald ): oral 0.97 .016
that historically plague this aspect of care. Simplicity of the Clinical variables
Total comorbidities −0.07 .352
application and tools that were developed can be modified
Total complication 0.07 .442
to meet the requirements of other areas/facilities. Glasgow Coma Scale 0.06 .068
The results presented here were from our institution’s Injury Severity Score 0.01 .665
directive to improve nutrition therapy. This is an example Cause of injury (fall)
of what an institution can do when it looks into its own Othere 0.03 .935
nutrition policies, relative to standards and through knowl- Motor vehicle accidentf 0.44 .234
edge gained through participation in research studies, and Unknown −0.73 .171
Constant 2.46 .003
then implements strategies to address challenges through an
interdisciplinary team-based approach with the support of a Coefficients (days) estimated from an ordinary least squares
hospital administration. regression model; the sample included 141 patients with isolated
closed head trauma injuries who received nutrition therapy. Nutrition
initiation: the time difference (in days) between when patients were
Key Points admitted and when they began receiving nutrition therapy.
b Admission year or year of treatment is intended to indicate the

1. Not just 1 intervention but all of them together likely
explain our results.
2. The results are likely partially explained by buy-
in from across the hospital and a continuation
of support for adequate and timely nutrition. The
interdisciplinary team approach allowed us to not
only account for the medical knowledge needed
but also understand the barriers from multiple
perspectives.
3. The overall understanding of barriers, as well as pos-
sible challenges with implementing the interventions,
allowed the development of solutions to challenges
that arose from a complete perspective instead of
1 solution from only 1 disciplinary or departmental
perspective.
quality improvement interventions that were implemented in the year
that a patient was admitted or treated.
c Individuals who identified as American Indian/Eskimo/Aleut, Asian
or Pacific Islander, or other.
d Individuals who received enteral or parenteral feeding.
e Individual’s injury caused by accident, accident at home, assault,
bicyclist, boat accident, cutting instrument, fight/brawl, found down,
gunshot wound, gunshot, hanging, hit by blunt instrument, impaled
by object, other, pedestrian struck, poison, railway accident, sports,
stabbing, or stab wound.
f Individual’s injury caused by either a motor vehicle collision or a
motorcycle collision.
Acknowledgments
We thank Kathy Hill, MA, RDN, for her ongoing support and
assistance with implementing the nutrition-related strategies
presented here.
Musillo et al
Statement of Authorship
L. Musillo contributed to the conception/design of the re-
search; all authors contributed to the acquisition, analysis, or
interpretation of the data; L. Musillo, L. M. Grguric, and R.
Batista drafted the manuscript; L. Musillo, L. M. Grguric, E.
Coffield, and R. Batista critically revised the manuscript; and
L. Musillo and L. M. Grguric agree to be fully accountable for
ensuring the integrity and accuracy of the work. All authors
read and approved the final manuscript.
References
1. McClave SA, Taylor BE, Martindale RG et al. Guidelines for the
provision and assessment of nutrition support therapy in the adult
critically ill patient: Society of Critical Care Medicine (SCCM) and
American Society for Parenteral and Enteral Nutrition (ASPEN).
JPEN J Parenter Enteral Nutr. 2016;40(2):159-211.
2. Lee ZY, Barakatun-Nisak MY, Airini IN, Heyland DK. Enhanced
Protein-Energy Provision via the Enteral Route in Critically Ill Patients
(PEP uP protocol): a review of evidence. Nutr Clin Pract. 2016;31(1):68-
79.
3. Wilson S, Madisi NY, Bassily-Marcus A, Manasia A, Oropello J,
Kohli-Seth R. Enteral nutrition administration in a surgical intensive
care unit: achieving goals with better strategies. World J Crit Care Med.
2016;4(3):108-186.
4. Peev MP, Yeh DD, Quraishi SA et al. Causes and consequences of
interrupted enteral nutrition: a prospective observational study in criti-
cally ill surgical patients. JPEN J Parenter Enteral Nutr. 2015;39(1):21-
27.
5. Cahill NE, Dhaliwal R, Day AG et al. Nutrition therapy in the
critical care setting: what is “best achievable” practice? An international
multicenter observational study. Crit Care Med. 2010;38(2):395-401.6.
575
6. Rivera R, Campana J, Hamilton C, Lopez R, Seidner D. Small
bowel feeding tube placement using an electromagnetic tube placement
device: accuracy of tip location. JPEN J Parenter Enteral Nutr.
2011;35(5):636-642.
7. Moncure M, Samaha E, Moncure K et al. Jejunostomy tube feedings
should not be stopped in perioperative patient. JPEN J Parenter Enteral
Nutr. 1999;23(6):356-359.
8. Pousman RM, Pepper C, Panharipande P et al. Feasibility of imple-
menting a reduced fasting protocol for critically ill trauma patients
undergoing operative and nonoperative procedures. JPEN J Parenter
Enteral Nutr. 2009;33(2):176-180.
9. Parent BA, Mandell SP, Maier RV et al. Safety of minimizing preoper-
ative starvation in critically ill and intubated trauma patients. J Trauma
Acute Care Surg. 2016;80(6):157-163.
10. Musillo L, Grguric-Smith LM, Coffield E, Totino K, DiGiacomo JC. Is
there a discrepancy? Comparing enteral nutrition documentation with
enteral pump volumes [published online September 1, 2016]. Nutr Clin
Pract.
11. Heyland DK, Cahill NE, Dhaliwal R et al. Enhanced protein-energy
provision via the enteral route in critically ill patients: a single center
feasibility trial of the PEP uP protocol. Crit Care Med. 2010;14(2):R78.
12. StataCorp. Stata Statistical Software: Release 14. College Station, TX:
StataCorp LP; 2015.
13. Marshall AP, Cahill NE, Gramlich L, MacDonald G, Alberda C,
Heyland DK. Optimizing nutrition in intensive care units: empowering
critical care nurses to be effective agents of change. Am J Crit Care.
2012;21(3):186-194.
14. McClave SA, Mohamed SA, Esterle M et al. Volume-based feeding in
the critically ill patient. JPEN J Parenter Enteral Nutr. 2015;39(6):707-
712.
15. Heyland DK, Dhaliwal R, Cahill NE et al. Driving perioperative
nutrition quality improvement processes forward! JPEN J Parenter
Enteral Nutr. 2013;37(1):83S-98S.
Clinical Research

Nutrition in Clinical Practice

Volume 33 Number 4
Gastrostomy Tube Feeding in Children With Developmental August 2018 576–583

C 2018 American Society for

or Acquired Disorders: A Longitudinal Comparison Parenteral and Enteral Nutrition

DOI: 10.1002/ncp.10056
on Healthcare Provision and Eating Outcomes 4 Years wileyonlinelibrary.com
After Gastrostomy

Ellen Backman, MSc1,2 ; Ann-Kristin Karlsson, PhD3 ; and Lotta Sjögreen, PhD4

Abstract
Background: Studies on long-term feeding and eating outcomes in children requiring gastrostomy tube feeding (GT) are scarce.
The aim of this study was to describe children with developmental or acquired disorders receiving GT and to compare longitudinal
eating and feeding outcomes. A secondary aim was to explore healthcare provision related to eating and feeding. Methods: This
retrospective cohort study reviewed medical records of children in 1 administrative region of Sweden with GT placement between
2005 and 2012. Patient demographics, primary diagnoses, age at GT placement, and professional healthcare contacts prior to
and after GT placement were recorded and compared. Feeding and eating outcomes were assessed 4 years after GT placement.
Results: The medical records of 51 children, 28 boys and 23 girls, were analyzed and grouped according to “acquired” (n = 13) or
“developmental” (n = 38) primary diagnoses. At 4 years after GT placement, 67% were still using GT. Only 6 of 37 (16%) children
with developmental disorders transferred to eating all orally, as opposed to 10 of 11 (91%) children with acquired disorders. Children
with developmental disorders were younger at the time of GT placement and displayed a longer duration of GT activity when
compared with children with acquired disorders. Conclusions: This study demonstrates a clear difference between children with
developmental or acquired disorders in duration of GT activity and age at GT placement. The study further shows that healthcare
provided to children with GT is in some cases multidisciplinary, but primarily focuses on feeding rather than eating. (Nutr Clin
Pract. 2018;33:576–583)

Keywords
enteral nutrition; feeding and eating disorders of childhood; gastrostomy; interdisciplinary; nutritional support; patient care teams;
pediatrics; tube feeding

Introduction “To eat” is defined according to the International

Gastrostomy tube feeding (GT) is necessary when children
with severe feeding or eating disorders require prolonged
nonoral enteral nutrition (EN) due to acquired or develop-
mental disorders.1,2 Indications for GT include digestive dis-
eases inducing a decrease in nutrient absorption, disorders
affecting the ability to eat, or inadequate caloric intake.3
Depending on the underlying disorder, GT can be short
term or long term. A clear difference in the length of GT
programmes when comparing children in need of GT due to
neurological impairments or to transient organ failure has
been noted.4,5 However, this difference has only been briefly
explored in relation to demographic characteristics and in
established healthcare guidelines.6
Feeding and eating are complex, interrelated activities
and for caregivers of children with GT, both of these
concepts have been shown to be of importance.7 “To feed”
can be defined as “to give food to” and relates to nutrition
to sustain vital bodily functions. Undernutrition may result
in poor cognitive development, impaired wound healing,
reduced immuno-competence, and/or growth inhibition.8,9
Classification of Functioning, Disability and Health as a
multifaceted activity including actions of eating food that
has been served, bringing it to the mouth, and consuming it
in culturally acceptable ways.10 Eating difficulties can cause
an imbalance in the caregiver–child interaction and may
lead to significant distress and frustration in the child and
From the 1 School of Health and Welfare, Halmstad University,
Sweden; 2 Regional Habilitation Center, Region Halland,
Kungsbacka, Sweden; 3 Department of Research and Development,
Region Halland, Halmstad, Sweden; and 4 Mun-H-Center Orofacial
Resource Center for Rare Diseases, Public Dental Service,
Gothenburg, Sweden.
Financial disclosure: This work was supported by the Department of
Research and Development, Region Halland, Sweden.
Conflicts of interests: None declared.
This article originally appeared online on March 30, 2018.
Corresponding Author:
Ellen Backman, MSc, School for Health and Welfare, Halmstad
University, Box 823 SE-301 18 Halmstad, Sweden.
Email: ellen.backman@hh.se
Backman et al
the caregiver.11 Traditional biomedical approaches focusing
on physical health outcomes is therefore insufficient for
interventions such as GT with broad and long-term
consequences.12 A multidisciplinary team approach has
been advised to improve nutrition management in children
with GT, including the areas of psychology, nutrition,
speech-language pathology, and medicine.13-15
The use of nonoral EN has increased in the pediatric
population in western Europe during the past 20 years,4,16,17
yet the healthcare related to feeding and eating practices
and follow-up provided are still nonstandardized and highly
variable.14,18,19 More important, as little is known about
the long-term feeding and eating outcomes correlated with
underlying disorders, this limits our capacity to provide
professional healthcare adapted to whether feeding and
eating functions have been lost or not yet developed.
The aim of this study was therefore to describe children
with developmental or acquired disorders receiving GT and
to explore healthcare provision related to eating and feed-
ing. The aim was to also compare longitudinal eating and
feeding outcomes in children with developmental disorders
with those in children with acquired disorders.
Methods
This descriptive and longitudinal study retrospectively re-
viewed medical records belonging to children receiving GT.
Data were reviewed up to 4 years after GT placement.
Medical records belonging to 51 children in 1 adminis-
trative region in the western part of Sweden were retrieved
from specialized pediatric healthcare providers in December
2016. Approximately 3% of Sweden’s 10 million inhabitants
live in the region. All children using GT were eligible for
taking part in the study, and inclusion criteria were (a) age
at start of GT <19 years, (b) GT duration >1 month, and
(c) GT placement between January 1, 2005, to December
31, 2012. The complete search strategy is available online
(Supporting Information Figure S1).
A protocol was used for data collection from the medical
records prior to and after GT placement (Supporting Infor-
mation Table S1). A pilot protocol was initially tested on the
records of 15 children to evaluate content and its use and
reduce the amount of missing data. Pilot testing proved the
protocol to include ambiguous definitions of variables re-
lated to feeding and healthcare as well as numerous free-text
options. Following pilot testing, many of the free-text op-
tions were replaced with tick boxes as saturation of possible
medications, diagnoses, and number and category of health-
care professionals in contact with the family was established.
The child’s primary diagnosis (the diagnostic codes from
the International Statistical Classification of Diseases and
Related Health Problems–10th revision; ICD-1020 ) stated in
the medical record by the pediatrician in charge of referring
to GT surgery was used to classify the underlying condition.
577
This represents a slightly different classification than has
been used earlier but was chosen as ICD-10 is the interna-
tional standard for reporting diseases and health conditions
for all clinical and research purposes.20 The ICD-10 has
been used in Sweden since 1997 with a Swedish update in
2011.21 The primary diagnoses were classified as “develop-
mental” (DD), relating to, for example, neurodevelopmental
disabilities or congenital malformations, or “acquired”
(AD), relating to conditions that were contracted after
birth.2
Data were collected retrospectively using the before-
mentioned instrument beginning at 12 months prior to GT
placement and continuing until year 4 following GT place-
ment. The medical records of all children were accessed
through the region’s electronic medical records system.
The descriptions of healthcare actions of each healthcare
professional was divided into care provided prior to and
after GT placement and was compiled at group level until
saturation was reached.
The head of the clinics responsible for the children’s
healthcare were informed about the study, and they ap-
proved a review of the medical records. All families eligible
for the study received written information and were given
the possibility to withdraw from the study at any time. The
study adhered to the ethical guidelines of the Declaration
of Helsinki22 and was ethically approved by the Regional
Ethics Committee in Lund (dnr 2016/93).
Data Analysis
Data were mainly presented using descriptive statistics and
expressed as medians, minimum–maximum, and percent-
ages. Nonparametric statistical tests were used to analyze
differences as the data were nonnormally distributed. Com-
parison of categorical data was performed using Fisher’s
exact test. Group differences in continuous variables were
analyzed using the Mann–Whitney U test for 2 indepen-
dent samples. A P value <.05 was considered statisti-
cally significant. The analyses have been performed us-
ing SPSS software, version 24.0 (IBM Corp., Armonk,
NY). Because of the retrospective nature of the study, the
data for all variables for every child were not available.
The total amount of children for each analysis is clearly
reported.
Results
Characteristics of Study Cohort
The medical records of 51 children, 28 boys and 23 girls,
were analyzed in the study. The characteristics of the
children are given in Table 1. The median age at the time
of GT placement was 35 months (minimum–maximum:
4–196), and 51% of the children (n = 26) were aged 3
or younger. A total of 3 children died during the study
578 Nutrition in Clinical Practice 33(4)

Table 1. Characteristics of Study Cohort.

Developmental Acquired
Category Disorders Disordersa P Value

Number of patients, n (%)b 38 (75) 13 (25)

Males, n (%) 18 (47) 10 (77) .11
Median age at gastrostomy placement in months, n 24 (4–180) 109 (4–196) .041
(minimum–maximum)
Number of children with nasogastric tube prior to 22 (58) 8 (62) >.99
gastrostomy placement, n (%)
Medication, n (%)
Intestinal regulators 19 (50) 9 (69) .34
Antibiotics 11 (29) 11 (85) .001
Anticonvulsants 14 (37) 2 .19
Respiratory medication 11 (29) 3 >.99
Gastroesophageal reflux disease medication 10 (26) 3 >.99
Chemotherapy – 10 (77) <.001
No medication 4 – .56
Other 10 (26) 11 (85) <.001
a Neoplasms and intracranial injuries were classified as acquired disorders.
bN = 51.

Table 2. Primary Diagnoses Comprising the Study Group
According to the Chapters of the International Statistical
Classification of Diseases and Related Health Problems–10th
Revision and Allocation to Subgroups, N = 51.
Number of
Category Children, n (%)
Developmental disorders, n = 38
Chapter E, metabolic disorders (E71)
Chapter F, mental, behavioural, and
neurodevelopmental disorders (F79, F90,
F98)
Chapter G, diseases of the nervous system
(G40, G71, G80)
Chapter P, certain conditions originating in
the perinatal period (P07)
Chapter Q, congenital malformations,
deformations and chromosomal
abnormalities (Q03, Q04, Q21, Q37, Q75,
Q79, Q85, Q87, Q90, Q92, Q93, Q99)
Acquired disorders, n = 13
Chapter C, malignant neoplasms (C41, C71,
C71, C91, C92)
Chapter D, diseases of the blood and
blood-forming organs and certain disorders
involving the immune mechanism (D71)
Chapter S, intracranial injury (S 06)
period, all attributed to causes other than GT placement.
Postoperative details for these children were not included.
A total of 35 primary medical diagnoses in 8 different
chapters of the ICD-10 were noted among the children
(Table 2).
A total of 17 different indications for commencing GT
was stated in the medical records when referring to gas-
trostomy surgery, with “swallowing disorders” from ICD-
10 chapter R being the most frequent indication for the
study cohort (n = 18, 35%). Examples of other indications
were from chapters C, neoplasm; E, malnutrition; F, other
feeding disorders of infancy and childhood; K, intesti-
nal malabsorption; and P, feeding problems of newborn.
1 (2) Among the children with DD, “swallowing disorders” from
3 (5)
ICD-10 chapter R was the most frequent indication, corre-
sponding to 37% of the indications for commencing GT (n
12 (24) = 14). Among the children with AD, malignant neoplasm
from ICD-10 chapter C was the most frequent indication
4 (8) (n = 5, 38%).
18 (35)
Professional Healthcare
Healthcare professionals represented in the children’s
medical records in regard to feeding, and/or eating in-
10 (20)
cluded pediatrician, registered nurse, registered dietitian,
1 (2) speech-language pathologist, pediatric dentist, occupational
therapist, physiotherapist, social worker in medical and
healthcare, psychologist, and special needs educator. The
2 (4) types of healthcare actions are further described in the
supplementary material available online (Supporting Infor-
mation Table S2).
Of the 51 children, 26 children representing with both
DD and AD also had specialized care outside the admin-
istrative region, which was sometimes only briefly noted in
the provided medical records. This care took place in clinics
specialized in cancer treatment or centers for rare diseases,
for example.
Backman et al 579

Table 3. Number of Children in Contact With Each Healthcare Professional.

Prior to Gastrostomy Placement After Gastrostomy Placement

Developmental, Acquired, n = 13; Developmental, Acquired, n = 11;

Healthcare Contacta n = 38; n (%) n (%) n = 37; n (%) n (%)b

Pediatrician 35 (92) 12 (92) 35 (95) 11 (100)

Registered nurse 26 (68) 11 (85) 34 (92) 11 (100)
Registered dietitian 27 (71) 7 (54) 36 (97) 11 (100)
Speech-language pathologist 26 (68) 2 30 (81) 1
Paediatric dentist 14 (38) 5 (38) 24 (65) 9 (82)
Social worker in medical and 13 (34) – 14 (38) –
healthcare and/or psychologistc
Occupational therapist 11 (29) – 14 (38) 1
Physiotherapist 8 (21) – 10 (27) –
Special needs educators 3 – 3 –
Specialized multidisciplinary feeding 1 – 4 1
team
a Healthcare contacts related to feeding and/or eating within the administrative region 12 months prior to gastrostomy placement and for 4 years
after.
b A total of 3 children died during follow-up; data for these children are not included after gastrostomy placement.
c Healthcare provided by social workers or psychologists was grouped as 1 category as the counseling was often given by the 2 professionals

together.

The analysis of the professional healthcare contacts
involved in eating and feeding demonstrated a higher de-
gree of multidisciplinary care, defined as including areas
of psychology, nutrition, speech-language pathology and
medicine in the group of children with DD both prior to
and after GT placement when compared with the children
with AD as seen in Table 3.
A total of 5 specialized multidisciplinary feeding teams
were mentioned in the medical records: 1 in this region, 3 in
other administrative regions in Sweden, and 1 abroad. The
feeding teams were comprised differently, and the therapy
was provided in some cases as intensive in-patient care, and
sometimes as outpatient follow-up.
No formal assessments of the children’s eating abilities
were found in the medical records prior to GT place-
ment, but 27 children (53%) had undergone an informal
assessment, and 6 children (12%) were assessed using a
Videofluoroscopic Swallowing Study; all had DD. Prior to
GT placement, 24 children (47%, 11 DD/13 AD) were not
assessed for eating ability.
After GT placement, 1 child with DD was formally
assessed using a Swedish scale measuring activities of daily
living, with eating ability as 1 part of the scale. A total of 26
children (54%, 24 DD/2 AD) were informally assessed. A
total of 21 children (44%, 13 DD/8 AD) were not assessed
for eating ability after GT placement.
Feeding and Eating Outcomes
GT activity 4 years after GT placement is depicted in Table
4. Discontinuation of GT for eating sufficiently orally oc-
curred after a median of 16 months (minimum–maximum:
2–43) after GT placement and was more often seen in
children with AD (P < .001). The use of nasogastric tube
feeding in the year prior to GT placement was not related
to a specific eating outcome.
Commercially manufactured enteral products were the
source of nutrition for the majority of the children. Home-
made, blenderized food was given to 11 children (23%)
during the study period, all of whom had DD. Homemade
food was most frequently used in combination with manu-
factured products. GT was permanently removed for 1 child
receiving homemade food due to eating sufficiently orally,
and 2 children increased their oral intake.
Feeding habits were documented after GT placement
based on the amount of food eaten orally or in tube and are
depicted in Figure 1. The degree of oral intake improved
for 19 children (40%), remained the same for 18 children
(38%), and worsened for 6 children (13%). For 3 children,
the degree of oral intake varied during the study period.
Data on feeding habits were missing for 1 child (1 DD) the
first and fourth year, for 1 child (1 DD) the first year, and 1
child (1 AD) received all nutrition orally, using GT only for
additional fluid and/or medication.
Further analyses of the changes in degree of oral intake
revealed the following:
r None of the 15 children (14 DD/1 AD) receiving “all
in tube” during the first year after GT placement
transferred to managing “all orally.” Of these chil-
dren, 6 increased their oral intake to either “most in
tube” (4 DD/1 AD) or “mostly orally”(1 DD).
r A total of 16 children (8 DD/8 AD) received “most
in tube” in the first year after GT placement, and 9
580 Nutrition in Clinical Practice 33(4)

Table 4. Follow-Up of Gastrostomy Tube Feeding Activity 4 Years After Gastrostomy Placement.

Ongoing Gastrostomy Eating Sufficiently P

Category Tube Feeding Orally Value

Number of children, N = 48, n (%) 32 (67) 16 (33)

Male, n (%) 16 (50) 11 (73) .136
Acquired disorders, n (%)a 1 (3) 10 (63)
Developmental disorders, n (%) 31 (97) 6 (37)
Median age at gastrostomy placement in 24 (4–152) 113 (4–196) .03
months, n (minimum–maximum)
a Neoplasms and traumatic brain injuries were classified as acquired disorders.

100%

90%

80%

70%

60%
All orally
50% Mostly orally
Most in tube
40%
All in tube
30%

20%

10%

0%
Developmental Developmental Acquired disorders, Acquired disorders,
disorders, year 1, n= 36 disorders, year 4, n=37 year 1, n=11 year 4, n=11

Figure 1. Feeding habits first and fourth year after gastrostomy placement distributed by acquired (“neoplasms” or “intracranial
injuries”) or developmental disorders: most in tube, >50% of total intake in the tube; mostly orally, >50% of oral intake from
liquids and solid foods orally. Data are missing for 2 children the first year and for 1 child the fourth year as a result of care
outside of the administrative region.

of these (1 DD/8 AD) transferred to managing “all
orally” during the study period. A total of 2 children
decreased their oral intake (2 DD), and 1 child (1
DD) increased the oral intake to “mostly orally.”
r A total of 14 children (13 DD/1 AD) received
“mostly orally” in the first year after GT placement.
Of these children, 4 (3 DD/1 AD) transferred to man-
aging “all orally” during the study period, whereas
5 children had decreased oral intake at the end of
follow-up to “most in tube” (3 DD) or “all in tube”
(2 DD).
“Oral stimulation using different tastes” was given to
nearly all children with “all in tube” (83%–100%) during the
4 years following GT placement.
Discussion
This longitudinal follow-up of children receiving GT shows
that at the end of the study period, 4 years after GT
placement, two thirds of the study population still required
GT. One main finding is the apparent difference in eating
outcomes for children requiring GT for DD or AD. Only
6 of 37 (16%) children with DD transferred to eating all
orally, as opposed to 10 of 11 (91%) children with AD.
Diagnoses relating to malignancies dominated in the group
with AD (11 of 13 children). The need for supplemental nu-
trition in this group stemmed from intensive chemotherapy
leading to, for example, loss of appetite, food aversion, and
vomiting, and not from an unsafe swallow or poor eating
skills. The use of GT in this group is consistent with prior
studies showing the use of GT as an important part of the
care to improve somatic health issues and to reduce family
frustration during the intensive phase of treatment.23,24 A
return to oral feeding after completion of treatment would
therefore be expected.
Another important difference was that children with
DD were younger at the time of GT placement. These
findings relating to age and GT activity are not in line with
earlier studies reporting that children who were weaned
from GT were younger at the time of GT placement than
Backman et al
children with GT still in place.25 However, the study did
not explore differences related to primary diagnoses,25 a
factor found to be of great importance in the present
study. The current findings are further strengthened by
previous studies indicating that differences in GT activity
were related to underlying disease,4,5,18 with neurological
disorders associating with a prolonged need for GT.
In the present findings, feeding habits were associated
with long-term eating outcomes. Children receiving “all in
tube” the first year after GT placement were more likely
to continue to require nutrition support by GT after 4
years. Children with GT as a complementary source of
nutrition during the first year, that is, receiving “most in
tube” or “mostly orally,” were more likely to increase their
oral intake. This was particularly true for children with
AD, most probably related to the secondary nature of their
feeding difficulties. These findings can also be interpreted
as the development of sensory functions and oral skills was
facilitated for children continuing receiving food orally dur-
ing GT.26 The long-term need for GT in children with DD
further highlights the necessity of this nutrition support for
children who otherwise could be at risk of forced feeding,
severe malnutrition, or growth impairment, an assumption
corresponding to earlier findings on positive outcomes of
GT.3,5,19
With regard to the primary medical diagnosis of children
with GT, the most frequent category was from ICD-10
chapter Q, congenital malformations, deformations, and
chromosomal abnormalities. This was somewhat surprising
as earlier studies have described neurological disease as
being the most common indication for GT.4,5,17,25 One ex-
planation could be the use of diverse classification systems
for primary diagnoses previously. The present study is the
first to use the international ICD-10 system for classifying
the primary medical diagnoses of children with GT. This
approach would simplify comparisons between studies in
the future.
Complications related to GT including gastric perfora-
tions, tissue granulation, or site infections have been well
described in previous studies3,5,20 and have therefore not
been included in the present study.
A multidisciplinary approach in the care of children
with GT has been advised.14,15 Sharp and colleagues15
recommend that all care offered to children with severe
feeding difficulties should include psychology, nutrition,
speech-language pathology, and medicine. This would offer
the necessary oversight and guidance needed to address the
feeding difficulties. The findings from the present study do
indeed show a wide range of healthcare professionals in-
volved in the care of the children, both prior to and after GT
placement. The type of care actions addressed basic abilities
such as swallowing, chewing, growth, and positioning, as
well as the following aspects of well-being: pain, nausea,
and feelings related to eating. However, distribution of the
581
contact with different healthcare professionals indicated a
preponderance of actions related to aspects of “feeding”:
medical status, nutrition, and growth. These results corre-
spond to those of previous studies that documented parent
perceptions of healthcare provided to children with GT as
primarily focusing on medical, not social, needs.27 Despite
the indications when referring to GT surgery, another
finding of the healthcare reported in the present study,
was the lack of formal tools used to assess eating abilities.
This corresponds to earlier stated inconsistencies in the use
of clinical measures for eating, drinking, and feeding.28
Moreover, the children’s participation in mealtimes, whether
eating orally or not, was seldom a documented area of
concern.
The present study is the first to compare the healthcare
provided to children with GT as a factor of the type of
primary diagnosis. The results reveal a discrepancy between
the care provided to children with DD and to those with
AD. A multidisciplinary team approach with care focused
on the broader concept of eating was more often used in
the former group. Despite this, the latter group showed a
greater increase in oral intake and in reestablishing adequate
eating. Increase in oral intake was expected because of the
transient nature of the feeding difficulties for the majority
of children with AD. Nevertheless, healthcare professionals
need to be aware of the multidimensional aspects of food
also for this patient group as food has been shown to carry
cultural representations of the life outside the hospital aside
from mere nutrition aspects.29 This difference also stresses
the importance of the primary diagnosis when anticipating
and communicating expected eating outcomes after GT
placement. The results highlight a question of whether
it is fruitful to strive for the same healthcare guidelines,
regardless of the child’s primary diagnosis, or if separate
guidelines relating to classification of the disorder would be
more useful in the clinical context. Lalanne et al5 stated that
as neurological impairment represents the main indication
for prolonged gastrostomy feeding, caregivers should be
provided with information regarding not only short term
practical and medical aspects of GT but also the potential
duration on GT related to the child’s underlying disorder.
This study was performed in 1 administrative region,
reflecting that particular demographic, geographic, and
healthcare setting. This affects the generalizability of the
results, and conclusions must therefore be interpreted with
caution. However, restricting the data collection to 1 admin-
istrative region had the advantage of identifying children
who were not formally registered in the medical records’
system, by using personal contacts, thus giving a richer data
material. Another limitation was the retrospective design
that bound the data collection to what was noted in the
medical records. This likely affected the variable of “feeding
habits,” for example, that often had to be estimated based
on free text because no standardized classification had
582
been used. Despite the retrospective nature of the data
collection, the findings are important as they are based on
legislated, professional documentation, and thereby should
reflect the healthcare provided and communicated across
different healthcare providers.
The present study found examples of children with AD,
as well as DD, increasing their oral intake. Future research
challenges lie in gaining a deeper understanding this in-
crease related to the natural history of different disorders,
identifying possible predictors of feeding outcomes, and in
exploring healthcare that stimulates eating development to
prevent tube dependency.
In summary, the findings of the present study show a
clear difference in age at GT placement and eating outcomes
for children with either DD or AD. Children receiving
GT as a result of AD will most likely begin to eat orally
when the intensive phase of treatment or acute phase of
illness is over, whereas children with DD receiving GT at
early age require healthcare that addresses feeding-related
and eating-related issues for an extensive period of time. A
more systematic and coherent use of available classification
systems and multidimensional assessment tools may help
clinicians and researchers to meet the needs of the indi-
vidual child when developing treatment programs, selecting
functional outcome measures, and communicating across
disciplines. Furthermore, the social aspects of mealtimes,
whether eating orally or not, must not be forgotten given the
positive implications this everyday interaction has on child
development.30,31
Acknowledgments
The authors wish to thank Anders Holmén, Ola Andersson,
and Amir Baigi at the Department of Research and Develop-
ment, Region Halland, for input in the design of the study as
well as for statistical advice. Thank you also to Ebba Sundin,
Halmstad University, and Mats Granlund, Jönköping Univer-
sity, for wise comments greatly improving the manuscript.
Statement of Authorship
E. Backman conceptualized the study, collected data, col-
laborated on the analysis and interpretation of data, and
drafted the initial manuscript. A.-K. Karlsson participated in
the conception and design of the study, collaborated on the
analysis of data, and reviewed and revised the manuscript.
L. Sjögreen participated in the conception and design of the
study, supervised data collection, collaborated on the analysis
and interpretation of data, and reviewed and revised the
manuscript. All authors read and approved the final manuscript
as submitted, and agree to be accountable for all aspects of the
work.
Supplementary Information
Additional supporting information may be found online in the
supporting information tab for this article.
Nutrition in Clinical Practice 33(4)
References
1. Gauderer MW. Percutaneous endoscopic gastrostomy—20 years later:
a historical perspective. J Pediatr Surg. 2001;36:217-219.
2. Adams RC, Elias E, Council on Children with Disabilities. Non-
oral feeding for children and youth with developmental or acquired
disabilities. Pediatrics. 2014;134:1745-1762.
3. Fröhlich T, Richter M, Carbon R, Barth B, Köhler H. Review article:
percutaneous endoscopic gastrostomy in infants and children. Aliment
Pharmacol Ther. 2009;31:788-801.
4. Diamanti A, Di Ciommo VM, Tentolini A, et al. Home enteral
nutrition in children: a 14-year multicenter survey. Eur J Clin Nutr.
2013;67:53.
5. Lalanne A, Gottrand F, Salleron J, et al. Long-term outcome of
children receiving percutaneous endoscopic gastrostomy feeding. J
Pediatr Gastroenterol Nutr. 2014;59:172-176.
6. Heuschkel RB, Gottrand F, Devarajan K, et al. ESPGHAN position
paper on management of percutaneous endoscopic gastrostomy in
children and adolescents. J Pediatr Gastroenterol Nutr. 2015;60:131-
141.
7. Petersen MC, Kedia S, Davis P, Newman L, Temple C. Eating and
feeding are not the same: caregivers´ perceptions of gastrostomy
feeding for children with cerebral palsy. Dev Med Child Neurol.
2006;48:5.
8. Engle P, Huffman SL. Growing children’s bodies and minds: max-
imizing child nutrition and development. Food Nutr Bull. 2010;31(2
suppl):S186.
9. Holmes S. Enteral nutrition: an overview. Nurs Stand. 2012;26:41.
10. World Health Organization. International Classification of Function-
ing, Disability and Health (ICF). http://apps.who.int/classifications/
icfbrowser/. Published 2011. Accessed September 28, 2016.
11. Davies WH, Satter E, Berlin KS, et al. Reconceptualizing feeding and
feeding disorders in interpersonal context: the case for a relational
disorder. J Fam Psychol. 2006;20:409-417.
12. Nelson KE, Lacombe-Duncan A, Cohen E, et al. Family experiences
with feeding tubes in neurologic impairment: a systematic review.
Pediatrics. 2015;136:E140-E151.
13. Agostoni FC, Axelsson WLI, Colomb WLV, et al. The need for
nutrition support teams in pediatric units: a commentary by the
ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr.
2005;41:8-11.
14. Edwards S, Davis AM, Bruce A, et al. Caring for tube-fed children: a
review of management, tube weaning, and emotional considerations.
JPEN J Parenter Enteral Nutr. 2016;40:616-622.
15. Sharp WG, Volkert, VM, Scahill L, McCracken CE, McElhanon B. A
systematic review and meta-analysis of intensive multidisciplinary in-
tervention for pediatric feeding disorders: how standard is the standard
of care? J Pediatr. 2017;118:116-124.e4
16. Daveluy W, Guimber D, Uhlen S, et al. Dramatic changes in home-
based enteral nutrition practices in children during an 11-year period.
J Pediatr Gastroenterol Nutr. 2006;43:240-244.
17. Pedron-Giner C, Navas-Lopez VM, Martinez-Zazo AB, et al. Analysis
of the Spanish National Registry for Pediatric Home Enteral Nutrition
(NEPAD): implementation rates and observed trends during the past 8
years. Eur J Clin Nutr. 2013;67:318-323.
18. Ricciuto A, Baird R, Sant’Anna A. A retrospective review of enteral
nutrition support practices at a tertiary pediatric hospital: a compari-
son of prolonged nasogastric and gastrostomy tube feeding. Clin Nutr.
2015;34:652-658.
19. Dahlseng MO, Andersen GL, Da Graca Andrada M, et al. Gas-
trostomy tube feeding of children with cerebral palsy: variation
across six European countries. Dev Med Child Neurol. 2012;54:938-
944.
Backman et al
20. World Health Organization. International Statistical Classification of
Diseases and Related Health Problems, Tenth Revision (ICD-10). World
Health Organization: Malta; 2011.
21. The National Board of Health and Welfare. ICD-10-SE, Internationell
statistisk klassifikation av sjukdomar och relaterade hälsoproblem–
systematisk f örteckning [International statistical classification of
diseases and related health problems–a systematic list]. http://www.
socialstyrelsen.se/klassificeringochkoder/diagnoskodericd-10.
Published 2011. Accessed June 10, 2016.
22. World Medical Association. Declaration of Helsinki—ethical princi-
ples for medical research involving human subjects 2013. http://www.
wma.net/en/30publications/10policies/b3/. Published 2013. Accessed
May 13, 2016.
23. Skolin I, Hernell O, Larsson MV, Wahlgren C, Wahlin YB. Percuta-
neous endoscopic gastrostomy in children with malignant disease. J
Pediatr Oncol Nurs. 2002;19:154-163.
24. Bauer J, Jürgens H, Frühwald M. Important aspects of nutrition in
children with cancer. Adv Nutr. 2011;2:67-77.
25. Åvitsland TL, Kristensen C, Emblem R, Veenstra M, Mala T,
Bjørnland K. Percutaneous endoscopic gastrostomy in children: a safe
technique with major symptom relief and high parental satisfaction. J
Pediatr Gastroenterol Nutr. 2006;43:624-628.
583
26. van den Hoek L, Harding C, van Gerven M, Cockerill H. Pediatric
feeding and swallowing rehabilitation: an overview. J Pediatr Rehabil
Med. 2017;10:95-105.
27. Brotherton AM, Abbott J, Aggett PJ. The impact of percutaneous
endoscopic gastrostomy feeding in children; the parental perspective.
Child Care Health Dev. 2007;33:539-546.
28. Sellers D, Pennington L, Mandy A, Morris C. A systematic review
of ordinal scales used to classify the eating and drinking abilities of
individuals with cerebral palsy. Dev Med Child Neurol. 2014;56(4):313-
322.
29. Skolin I, Wahlin Y, Broman DA, et al. Altered food intake and
taste perception in children with cancer after start of chemotherapy:
perspectives of children, parents and nurses. Support Care Cancer.
2006;14:369-378.
30. Fiese BH, Foley KP, Spagnola M. Routine and ritual elements in family
mealtimes: contexts for child well-being and family identity. New Dir
Child Adolesc Dev. 2006;Spring:67-89.
31. Harrison ME, Norris ML, Obeid N, Fu M, Weinstangel H, Sampson
M. Systematic review of the effects of family meal frequency on
psychosocial outcomes in youth. Can Fam Physician. 2015;61:96-
106.
Letter to the Editor
Four-Oil Lipid Emulsion (Smoflipid) as a Tool in Managing
Parenteral Nutrition Shortages
Drug shortages call for institutions to be creative to provide
optimal care to patients. We read with interest the article
by Holcombe et al1 reviewing the impact of shortages
and strategies to address them. Our inpatient pharmacy
is using similar approaches to manage the critical amino
acid shortage with the addition of incorporating a four-oil
lipid injectable emulsion (ILE; Smoflipid) into some of our
inpatient care plans. Our institution has found adding four-
oil ILE to formulary particularly helpful during times of
parenteral nutrition (PN) shortages.
All patient protein orders were decreased as a con-
servation measure institution-wide. In patients who could
not tolerate additional dextrose, the dose of four-oil ILE
was increased to >1 g/kg/d and in some cases up to
2 g/kg/d not exceeding 30% total kilocalories. Recom-
mended dosing practices allow higher doses of four-oil ILE
compared with soybean–oil derived lipids. Our institution
has worked closely with pharmacy information technology
building order panels for PN linked to four-oil ILE and PN
linked to conventional soybean–oil ILE to avoid dispensing
the wrong ILE product. We also created a report for our
intravenous (IV) room pharmacist to double-check that
patients are not receiving >1 ILE because having 2 ILE
products on formulary that look similar could introduce
error. Since these processes have been in place, our institu-
tion has reported zero medication safety events related to
dispensing errors.
There have been no documented adverse effects from
using this higher dose of ILE. At our institution, the four-
oil ILE is implemented in patients receiving standardized
commercially available 2-in-1 PN products for initiation or
de-escalation of PN, chronic PN or those deemed chronic
during admission, patients with elevated liver function tests
3 times normal limit, triglycerides (TGs) between 300 and
500 mg/dL, and patients experiencing hyperglycemia refrac-
Nutrition in Clinical Practice
Volume 33 Number 4
August 2018 584

C 2018 American Society for
Parenteral and Enteral Nutrition
DOI: 10.1002/ncp.10106
wileyonlinelibrary.com
tory to insulin therapy. Given this product is new to the U.S.
market, there have been questions regarding its safety and
efficacy. Our institution has been using this product since
April 2017 and recently completed a retrospective review in-
cluding 24 patients over a 4-month time period. Our results
were similar to previously published studies that showed a
trend toward decreased aspartate aminotransferase (AST;
165 vs 56 units/L; P = .46), alanine transaminase (ALT;
96 vs 43 units/L; P = .32), and C-reactive protein (CRP;
104 vs 61 mg/L; P = .65). TGs were unchanged before and
after (174 vs 192 mg/dL; P = .65). The duration of four-oil
ILE varied, with 46% of patients receiving 1–7 days, 25%
receiving 8–14 days, 17% receiving 15–21 days, and 12%
receiving ࣙ21 days. Given the majority of patients received
a short duration of four-oil ILE and still experienced a
decrease in AST/ALT and CRP is promising.
In summary, incorporating a four-oil ILE into patient
care plans to deliver more calories from fat has allowed us
to maintain caloric intake despite necessary reductions in
protein because of the amino acid shortage. This has been
one creative approach our institution is taking in response
to critical PN shortages in addition to those described by
the authors.
Lisa Mostafavifar, PharmD, BCPS, BCNSP
Department of Pharmacy, The Ohio State University
Wexner Medical Center, Columbus, Ohio, USA
David C. Evans, MD, FACS
Department of Surgery, The Ohio State University,
Columbus, Ohio, USA
Reference
1. Holcombe B, Mattox TW, Plogsted S. Drug shortages: effect on par-
enteral nutrition therapy. Nutr Clin Pract. 2018;33(1):53-61.