Gastroenterology 2016;150:866874

Depression and Somatization Are Associated With Increased

Postprandial Symptoms in Patients With Irritable
Bowel Syndrome
CLINICAL AT

Lukas Van Oudenhove,1,2 Hans Trnblom,3,4 Stine Strsrud,3 Jan Tack,1 and
Magnus Simrn3,4
1

Translational Research Center for Gastrointestinal Disorders, Department of Clinical and Experimental Medicine, University of
Leuven, Leuven, Belgium; 2Consultation-Liaison Psychiatry, University Psychiatric Centre KU Leuven, University Hospitals
Leuven, Leuven, Belgium; 3Department of Internal Medicine and Clinical Nutrition, Institute of Medicine, 4University of
Gothenburg Centre for Person-Centered Care, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
BACKGROUND & AIMS: Patients with irritable bowel syndrome
(IBS) have increased postprandial symptom responses and more
psychosocial morbidities than healthy individuals. However, the
relationship between psychosocial status and postprandial
symptom responses in patients with IBS is unclear. We investigated this relationship in a prospective study of patients with
IBS. METHODS: A total of 193 IBS patients, diagnosed according
to Rome II (n 126) or Rome III (n 67) criteria, consumed a
standard breakfast (540 kcal: 36% fat, 15% proteins, 49% carbohydrates, and 8.9 g ber). They completed visual analogue
scales assessing the severity of 5 gastrointestinal symptoms
(abdominal pain, bloating, nausea, gas, and fullness) before
breakfast and every 30 minutes, up to 240 minutes after eating.
All patients completed validated self-report questionnaires for
their current levels of anxiety and depression; patients diagnosed based on Rome II criteria also completed a somatization
questionnaire. The relationship between these variables and the
course of gastrointestinal symptom scores over time was
analyzed using linear mixed models, controlling for comorbid
functional dyspepsia. RESULTS: We observed a main effect of
anxiety levels on fullness and bloating (P < .04), and of
depression levels on abdominal pain (P .007), reecting a
general upward shift of the entire symptom curve. Depressionby-time interactions were seen for nausea and gas (P < .03).
Somatization levels had a main effect on all 5 symptoms
(P < .0001), independent of anxiety and depression. We
observed somatization-by-time interaction effects for bloating
(P .005), and nausea (P .02), and a nonsignicant trend was
found for pain (P .054), reecting a steeper early postprandial
increase in symptoms among subjects with higher levels of
somatization. CONCLUSIONS: Based on a prospective study of
patients with IBS, psychosocial morbidities are associated with
increased levels of gastrointestinal symptoms in general.
Depression and somatization levels are associated specically
with increased postprandial symptoms.

defecation.1 Despite its high global prevalence (11.2%),2

IBS remains a poorly understood condition with a multifactorial and heterogeneous etiology and pathogenesis.3,4
IBS is characterized by high levels of psychiatric comorbidity, particularly anxiety, depressive, and somatoform
disorders.57
A majority of IBS patients perceive their symptoms as
being related to meal intake.8 Perceived intolerance to specic foods also is common but has been shown to correspond
poorly with malabsorption testing.911 We recently showed
that IBS patients, compared with healthy controls, show
increased gastrointestinal (GI) symptom responses to a
standard breakfast during a 4-hour postprandial period,
particularly for bloating.12 Altered postprandial smallbowel13 and colonic14,15 motor responses as well as malabsorption of certain nutrients,16 altered colonic fermentation,17 and/or abnormal gas production or handling18 all
have been suggested to be involved. Taken together, this
suggests that meal intake may interact with GI sensory and
motor (dys)functions involved in IBS pathophysiology, but
the exact mechanisms underlying this phenomenon remain
unclear. This represents a huge gap in IBS research given the
high prevalence of postprandial exacerbations of IBS symptoms, and the difculties in managing these symptoms.8
However, because IBS should be regarded as a disorder
of the braingut axis,19 postprandial symptom responses
should be looked at within the context of gutbrain
signaling. More specically, altered central processing and
modulation of (nutrient-related) visceral sensory signals
also may contribute to the increased postprandial symptom
generation found in IBS. Although unexplored, this represents a plausible mechanism because increasing levels of
psychological distress are risk factors to develop IBS20,21 as

Keywords: Psychologic Factors; Discomfort; Psychiatric; Nervous System.

Abbreviations used in this paper: ANS, autonomic nervous system; CCK,

cholecystokinin; FD, functional dyspepsia; GI, gastrointestinal; IBS, irritable bowel syndrome; IBS-C, constipation-predominant irritable bowel
syndrome; IBS-D, diarrhea-predominant irritable bowel syndrome; IBSSSS, IBS severity scoring system; PHQ, Patient Health Questionnaire;
PYY, peptide tyrosine tyrosine.

rritable bowel syndrome (IBS) is dened by the Rome

III criteria as a functional bowel disorder in which
abdominal pain or discomfort is associated with defecation
or a change in bowel habit, and with features of disordered

Most current article

2016 by the AGA Institute
0016-5085/$36.00
http://dx.doi.org/10.1053/j.gastro.2015.11.010

well as associated with increasing GI symptom levels in

general (ie, regardless of their timing relative to meal
intake) in IBS patients.22,23 These associations can be
explained by the fact that processing and modulation of
visceral sensory signals is inuenced heavily by psychosocial factors.24 However, our understanding of the relationship between levels of psychological symptoms and
postprandial sensations or symptoms in IBS is incomplete.
The aim of this study therefore was to study the relationship between anxiety, depression, and somatization
levels on the one hand and GI symptom levels before and
after a standard breakfast on the other hand, in IBS patients.
We hypothesized that the severity of anxiety, depression,
and somatization symptoms would be associated with
increased GI symptom levels at the preprandial baseline, as
well as an increased postprandial symptom response.

Materials and Methods

Subjects
We prospectively included patients fullling Rome II25 and
later Rome III1 criteria for IBS, who consulted at a single secondary/tertiary care outpatient unit (Sahlgrenska University
Hospital, Gothenburg, Sweden) during the time period from
2005 until 2008. A proportion of the subjects (n 67) also
were included in our recent publication on postprandial
symptom response in IBS patients vs controls.12
The diagnosis was based on a typical clinical presentation
and additional investigations if considered necessary on clinical
grounds. However, all subjects underwent a physical examination as well as routine laboratory blood testing including
transglutaminase antibodies for celiac disease.
The fact that not all patients were diagnosed according to
the same version of the diagnostic criteria rendered subdividing patients according to bowel habits difcult. However, we
subdivided the entire patient sample (Rome II and Rome III
subgroups combined) into 3 categories based on Bristol Stool
Form data: constipation-predominant IBS (IBS-C), diarrheapredominant IBS (IBS-D), and IBS-nonCnonD (corresponding
to Rome II IBS-A [alternating IBS]25 and Rome III IBS-M [mixed
IBS] and IBS-U [unsubtyped IBS]1).
All patients were given study-specic verbal and written
information before giving their written consent to participate in
the studies. The Regional Ethical Review Board in Gothenburg
approved the study before the start of patient inclusion.

Standard Breakfast and Postprandial

Symptom Response Assessment
The procedure was described in detail in our previous
publication.12
All medications with known effects on the gastrointestinal
tract were discontinued at least 48 hours before the study,
including antidepressants. None of the included subjects were
on anxiolytics, and the number of patients on antidepressants
was small (<20 patients). After an overnight fast, the subjects
presented to the laboratory at 7.30 am. Demographic data
including height and weight were collected to calculate body
mass index. Subjects were served a breakfast of oat bran (27 g)
with 50 g applesauce and 2 dL milk, 2 crispbreads with 10 g

Postprandial Symptoms in IBS

867

margarine and 2 slices of cheese (20 g), and 1 dL apple juice

(540 kcal; 36% fat, 15% proteins, 49% carbohydrates, and 8.9 g
ber).
Subjects were instructed to ingest the entire meal and nish
within 10 minutes. Gastrointestinal symptoms (abdominal pain,
bloating, nausea, gas, and fullness) were assessed before
breakfast (preprandial baseline) and every 30 minutes up to
240 minutes postprandially using 100-mm visual analogue
scales ranging from no sensation to very severe sensation.
Before the meal, all patients completed questionnaires to assess
anxiety, depression, somatization, IBS symptoms, and comorbid
functional dyspepsia (see later).

Psychosocial Assessment
For the assessment of anxiety and depression symptom
levels during the past week, patients completed the validated
Hospital Anxiety and Depression Scale.26 The Hospital Anxiety
and Depression Scale consists of 14 items in total (7 anxiety
items and 7 depression items), each scored on a Likert scale
ranging from 0 to 3. Subscale scores for anxiety and depression
are calculated by summing the 7 respective items.
Somatization (the tendency to report multiple somatic
symptoms) during the past 4 weeks was assessed using the
somatic symptom severity module of the Patient Health Questionnaire (PHQ), consisting of the 15 most common somatic
symptoms (PHQ-15).27 The severity of each of these symptoms
and their associated distress is scored on a Likert scale ranging
from 0 to 2, resulting in a total somatization score ranging from
0 to 30. To avoid confounding of the relationship between
somatization and our GI symptom outcome variables as well as
with the comorbid functional dyspepsia (FD) variable (see
later) by the GI items of the PHQ-15, these 3 items were
omitted as previously described and validated (PHQ-12).28
Because to the best of our knowledge no validated cut-off
values for the PHQ-12 exist, we also report the PHQ-15
scores for descriptive purposes. Somatization was measured
only in the Rome II subsample (n 127; 65.3%).

IBS Symptom Severity and Comorbid

Functional Dyspepsia Assessment
All patients also completed the IBS severity scoring system
(IBS-SSS).29 This is a well-validated IBS symptom score that is
based on 5 items (severity of pain, duration of pain, abdominal
distension, bowel dissatisfaction, and interference with life in
general) and uses visual analogue scales. The maximum score is
500 and patients can be categorized as having mild (75175),
moderate (176300), or severe (>300) IBS.
Comorbid FD was assessed by Rome II30 or Rome III31
modular questionnaires, depending on the time of recruitment (before or after the publication of Rome III criteria)
yielding information on the presence (yes/no) of comorbid FD.

Statistical Analysis
SAS 9.3 software (SAS Institute, Cary, NC) was used to
analyze the data, which are expressed as means SEM. Signicance was set at a P value of less than .05.
To test our hypothesis that levels of psychological symptoms (anxiety, depression, and somatization) are associated
with an increased postprandial symptom response in IBS patients, we used marginal linear mixed models with main effects

CLINICAL AT

April 2016

868

Van Oudenhove et al

CLINICAL AT

of time (categoric variable, 9 time points in total: 1 preprandial,

8 postprandial) and the psychological variable (continuous
variable), as well as a time-by-psychological variable interaction effect. All analyses were controlled for comorbid FD
(dichotomous variable). The psychological variables were
standardized with a mean of 0 and a standard deviation of 1 to
facilitate the interpretation of the main effect of time in the
models containing the time-by-psychological variable interaction effect (because the main effect of time is estimated at the
average level of the psychological variable). The effects of
anxiety and depression levels were tested together on the full
sample (n 193) in one model per GI symptom. The effects of
the somatization level were tested on the Rome II subsample
(n 125), controlling for putative effects of anxiety and
depression levels (both main effects and interaction effects
with time). These 2 models were estimated for each of the 5 GI
symptoms, these visual analogue scale ratings constituted the
dependent variables in these models.
A signicant main effect of the psychological variable indicated a signicant association between the psychological variable and the GI symptom dependent variable, which is
independent of time (ie, upward or downward shift of the
entire timesymptom curve with increasing levels of psychological variable, without change in the shape of this curve). A
signicant time-by-psychological variable interaction effect
indicates that the effect of the psychological variable on the GI
symptom dependent variable differs over time (change in shape
of the timesymptom curve, eg, in the case of a steeper early
postprandial increase in patients with increasing levels of
psychological symptoms, without differences at baseline or in
later phases). To illustrate these effects visually, the time
symptom curves will be plotted at different levels of the
psychological variable (mean -1 SD, mean, mean 1 SD,
mean 2 SD).

Results
Descriptive Results
We included a total of 193 IBS patients (mean age, 41.3
1.0 y; body mass index, 24.0 0.4 kg/m2; 143 [74.1%]
women). A total of 126 (65.3%) patients were recruited
before the introduction of the Rome III criteria1 and, hence,
were diagnosed according to the Rome II criteria.25 Somatization data were available only in this subgroup. The
remaining 67 (34.7%) patients were diagnosed according to
the Rome III criteria because these patients were included
after their introduction.1 Rome III patients had slightly,
although signicantly, higher IBS-SSS total scores (270.0
97.4 vs 303.5 88.3; t[189] 2.41; P .017). Given the
small effect size for this difference (Cohens d 0.36), we
believe this does not represent a major problem for pooling
subjects from both subgroups in our mixed-model analysis
on the effect of anxiety and depression on postprandial
symptoms reported later, especially because controlling for
IBS-SSS score in these models did not change the effects of
anxiety and depression for any of the GI symptom outcomes
(data not shown). The fact that patients were recruited
prospectively from a single center by 2 gastroenterologists
experienced in IBS (H.T., M.S.), and that Rome II and Rome
III criteria for IBS are relatively similar, further ensures

Gastroenterology Vol. 150, No. 4

sufcient similarity in this sample. Further descriptive data

are given in Table 1.

Effect of IBS Subtype

No signicant main effect of subtype (IBS-C, IBS-D, or
IBS-nonCnonD) was found for abdominal pain, fullness,
nausea, or gas, and there was no signicant subtype-by-time
interaction effect for any of these outcome variables. For
bloating, no signicant main effect of subtype was found,
but there was a signicant subtype-by-time interaction effect (F[16,187] 1.81; P .033). However, post hoc 1-way
analysis of variance comparing the 3 subtypes at all
different time points did not show any signicant differences at any of these time points, although the differences
Table 1.Descriptive Data
Mean SEM
Age, y
Sex
Women
Men
BMI, kg/m2
Diagnosis
Rome II
Rome III
Subtypea
IBS-C
IBS-D
IBS-nonCnonD
IBS-SSS score
Pain severity
Pain frequency
Bloating severity
Dissatisfaction bowel habits
Interference daily life
Total
IBS severityb
Mild
Moderate
Severe
Functional dyspepsia diagnosisc
Hospital Anxiety and Depression score
Anxiety
Depression
Anxiety disorder diagnosisd
Possible
Probable
Depression diagnosisd
Possible
Probable
PHQ-12 somatization scoree

n (%)

41.3 1.0
143 (74.1)
50 (25.9)
24.0 0.4
126 (65.3)
67 (34.7)
54 (28.3)
81 (42.4)
56 (29.3)
43.2
55.9
57.6
68.9
65.0
291.5

2.0
2.4
1.9
1.9
1.8
6.7
13
86
94
116

(6.7)
(44.6)
(48.7)
(60.1)

5.9 0.3
5.8 0.3
51 (26.4)
21 (10.9)
49 (25.4)
9 (4.1)
7.1 0.4

BMI, body mass index.

a
Entire patient sample (Rome II and Rome III subgroups
combined) based on Bristol Stool Form data: IBS-C, IBS-D,
and IBS-nonCnonD (corresponding to Rome II IBS-A and
Rome III IBS-M and IBS-U).
b
Based on validated IBS-SSS cut-off values.
c
Based on Rome II or Rome III modular questionnaire.
d
Based on validated HAD cut-off values.
e
Obtained in Rome II subgroup only.

April 2016

Postprandial Symptoms in IBS

869

Table 2.Overview of Mixed-Model Analyses: Anxiety and Depression

F
b SEa
Time
Anxiety
Anxiety-by-time
Depression
Depression-by-time
Functional dyspepsia

3.1
0.8
1.3 1.5
1.1
7.35
3.9 1.5
0.8
4.4
5.1 2.5

P
.003
.37
.36
.007
.64
.038

Fullness
F
b SEa
24.4
6.7
4.8 1.9
0.3
0.1
-0.7 1.9
1.4
10.3
9.9 3.1

Bloating
P

<.0001
.010
.96
.70
.19
.002

F
b SEa
8.3
4.3
3.6 1.8
0.9
0.3
0.9 1.7
1.5
4.0
5.7 2.8

Nausea
P

<.0001
.039
.55
.59
.16
.047

F
b SEa
2.5
0.02
0.21 1.5
1.81
2.8
2.5 1.5
2.4
0.8
2.2 2.5

Gas
P
.012
.89
.077
.097
.017
.38

F
b SEa
3.18
2.2
2.4 1.7
1.2
0.0
0.06 1.6
2.3
1.5
3.2 2.6

P
.002
.14
.28
.97
.021
.23

NOTE. Signicant effects (P < .05) are shown in bold.

a
Shown when a single b was available (ie, for main effects of continuous or dichotomous variables).

between the 3 groups increased numerically toward the

later postprandial period, with the quickest recovery in
IBS-C patients and the slowest recovery in IBS-D patients
(details not shown).

Association Between Psychological Symptom

Levels and Symptom Response to a Meal
An overview of the results of all the marginal mixedmodel analyses is shown in Tables 2 and 3. The analyses
with somatization level were performed on the Rome II
subgroup (n 125), in which somatization data were
available, controlling for putative effects of anxiety and
depression levels (Table 3).

Abdominal Pain
A signicant main effect of depression was found (P
.007), reecting an upward shift of the entire curve with
increasing levels of depression, as illustrated in Figure 1A.
For somatization, a main effect was found (P<.0001), as
well as a nonsignicant trend (P .054) for the interaction
effect with time. Figure 2A indeed shows that the entire
curve not only is shifted upward with increasing levels of
somatization, but that increasing somatization levels also
are associated with a different shape of the curve, particularly a steeper increase in pain rating in the early postprandial period (with the curve being virtually at in
subjects with low somatization levels). The b coefcients for
the time-by-somatization interaction effect conrm that the
effect of somatization is signicantly stronger compared
with the preprandial time point up to the 90th minute
postprandially (details not shown). The effect of comorbid
FD was signicant in the model with anxiety and depression, but not in the model with somatization.

Fullness
Signicant main effects were found for anxiety (P .01)
and somatization (P < .0001), as shown in Figures 1B and
2B, respectively. The effect of comorbid FD was signicant
in both models.

Bloating
Signicant main effects were found for anxiety (P
.039) (Figure 1C) and somatization (P < .0001). For somatization, in addition to the main effect, a signicant time-bysomatization interaction effect was found (P .005).
Figure 2C shows that this effect again reects a steeper increase in bloating ratings in the early postprandial period as
well as an additional increase in bloating ratings in the later
postprandial period in subjects who score higher on somatization, with the curve being virtually at in subjects with
low somatization levels. The b coefcients for the time-bysomatization interaction effect conrm that the effect of
somatization is signicantly stronger compared with the
preprandial time point up to the 60th minute postprandially
and again at the 180-minute postprandial time point (details
not shown). The effect of comorbid FD was (borderline)
signicant in both models.
Adding the main effect of subtype and its interaction
effect with time to the models did not change the signicance of any of the other effects (details not shown). Similar
to the model without psychological variables reported
earlier, the subtype-by-time interaction effect was signicant in both models, whereas the main effect of subtype was
not, indicating that the effect of subtype and the effects of
psychological variables on the time course of the bloating
response are independent.

Nausea
A signicant depression-by-time interaction effect was
found (P .017), reecting an enhanced early (up to the
60th minute) and late (from the 210th minute) symptom
response in patients with higher depression scores
(Figure 1D). For somatization, signicant main and interaction effects were found (P < .0001 and P .018,
respectively). The interaction effect reects steeper early
postprandial increases in nausea ratings with increasing
levels of somatization (Figure 2D). The b coefcients for the
time-by-somatization interaction effect show that the effect
of somatization is signicantly stronger compared with the
preprandial time point at the 60th minute postprandially

CLINICAL AT

Abdominal pain

870

Van Oudenhove et al

Gastroenterology Vol. 150, No. 4

Table 3.Overview of Mixed-Model Analyses: Somatization

Abdominal pain
F
b SEa
Time
Somatization
Somatization-by-time
Functional dyspepsia
CLINICAL AT

2.72
23.4
7.9 1.6
2.0
0.6
2.2 2.9

P
.009
<.0001
.054
.45

Fullness
F
b SEa
11.7
29.6
11.5 2.1
1.5
8.0
10.6 3.8

Bloating
P

<.0001
<.0001
.18
.005

F
b SEa
4.0
20.1
9.0 2.0
3.0
3.9
6.7 3.4

Nausea
P

.0003
<.0001
.005
.052

F
b SEa
1.1
29.4
9.8 1.8
2.4
0.02
0.5 3.3

Gas
P

.41
<.0001
.018
.88

F
b SEa
1.8
19.6
8.3 1.9
1.8
1.2
3.4 3.1

P
.086
<.0001
.079
.28

NOTE. The somatization analyses were performed on the Rome II subgroup only (n 125; 65.3% of the total sample); all
models controlled for main effects of anxiety and depression, as well as anxiety-by-time and depression-by-time interaction
effects. Signicant effects (P < .05) are shown in bold.
a
Shown when a single b was available (ie, for main effects of continuous or dichotomous variables).

(details not shown). The effect of comorbid FD was not

signicant in any of these models.

Gas
A signicant time-by-depression interaction effect was
found (P .021), reecting a late postprandial increase in
gas rating in subjects scoring high on depression, with the
inverse pattern observed during the early postprandial
period (Figure 1E). Furthermore, a signicant main effect of
somatization was found (P < .0001) (Figure 2E). The effect
of comorbid FD was not signicant in any of these models.

Discussion
In this study we demonstrate that increasing levels of
anxiety, depression, and somatization are associated with
higher preprandial and/or postprandial GI symptom levels
in IBS patients. More specically, increasing levels of anxiety
were associated with an upward shift of the timesymptom
curve over its entire course (ie, to the same extent at the
preprandial baseline and the postprandial time points) for
fullness and bloating. Increasing depression levels were
associated with abdominal pain levels in a similar way. For
nausea, however, increasing levels of depressive symptoms

Figure 1. Overview of the signicant associations between anxiety, depression, and preprandial and postprandial gastrointestinal symptoms ratings in Rome II/Rome III IBS patients (n 193). (A) Depression and abdominal pain, (B) anxiety and
fullness, (C) anxiety and bloating, (D) depression and nausea, and (E) depression and gas. VAS, visual analog scale. Error bars
represent the standard error of the mean; time 0 represents the preprandial time point, and other time points are in minutes
after the meal.

Postprandial Symptoms in IBS

871

CLINICAL AT

April 2016

Figure 2. Associations between somatization and preprandial and postprandial gastrointestinal symptom ratings in Rome II
IBS patients (n 125). (A) Abdominal pain, (B) fullness, (C) bloating, (D) nausea, and (E) gas. VAS, visual analog scale. Error
bars represent standard error of the mean; time 0 represents the preprandial time point, and other time points are in minutes
after the meal. All analyses are controlled for the effects of anxiety and depression (main effects and interaction effects with
time).

were associated with a change in the shape of the postprandial symptom curve, more specically, with a steeper
early postprandial increase as well as the presence of a
smaller late postprandial increase. For gas, increasing
depressive symptom levels were associated with increased
symptom responses in the late postprandial phase only. The
strongest effects, however, were found for somatization.
These effects were independent of those of anxiety and
depression, but it should be noted that somatization data
were available in the Rome II subsample only. Besides signicant main effects of somatization for all 5 GI symptoms
(indicating an overall higher symptom level over all time
points), increasing levels of somatization were associated
with steeper early postprandial increases for abdominal
pain (although this effect did not reach signicance, P
.054), bloating, and nausea. Taken together, these results
indicate that, irrespective of time relative to meal intake (ie,
both preprandially and postprandially), increasing anxiety
and depression levels are associated with higher levels of
specic individual GI symptoms (fullness and bloating for
anxiety, pain for depression), whereas increasing somatization levels are associated with higher levels of all GI
symptoms. In addition, however, for some symptoms, higher
depression and somatization levels are associated with
increased postprandial symptom levels specically. These
relationships occurred independently of comorbid FD, with
FD comorbidity being on the high end of the range reported
in the literature.32 Because interaction effects were
found primarily for symptoms that often are present in FD
(fullness, bloating, nausea), and these effects were driven

primarily by the early postprandial time period (hence,

putatively by altered foregut responses) and FD symptoms
often are meal-related, it would be interesting to test these
relationships in a separate FD cohort.
The nding of an association between psychological
symptom severity and GI symptom severity in general is in
line with reports in the literature showing signicant associations between these and other psychological symptoms
on the one hand and self-reported overall IBS symptom
severity (typically scored over weeks or months) on the
other.22,23 In line with previous ndings in FD, we showed
that the somatization level is associated with increased
reports for all symptoms, whereas anxiety and depression
levels are associated with the severity of specic GI symptoms.33 This ts with the denition of somatization as a
psychological tendency to report multiple somatic symptoms for which no organic explanation can be found.5,34 The
nding that, for some GI symptoms, depression and somatization levels are associated specically with increased
postprandial responses is, to the best of our knowledge,
novel. However, these ndings are in line with earlier work
by Elsenbruch et al,35 showing that postprandial GI symptom levels and negative affectivity responses to an acute
stressor are correlated. Furthermore, our ndings are in line
with a recent report of another IBS sample from the same
center, showing that symptom levels during a lactulose
challenge test were associated with anxiety levels.36 However, because symptom responses were quantied as area
under the time-symptom curve (including preprandial time
points) in this study, no distinction could be made between

872

Van Oudenhove et al

CLINICAL AT

effects on symptom reporting in general (ie, over the preprandial and postprandial time period together) and postprandial symptom responses specically.
We believe that the present ndings are clinically
important because they suggest that assessing levels of
depression and somatization and, to a lesser extent, anxiety,
may be indicated in case of postprandial worsening of
symptoms in IBS. It obviously remains to be determined
whether successful treatment of psychological morbidity
would normalize postprandial symptom responses, as has
been shown for mechanosensitivity.37
Given the lack of earlier data on this matter and the lack
of physiological measurements in the present study, we can
only speculate on the potential mechanisms involved in the
associations we found. It is known from the pain literature
that normal psychological processes such as emotions
(anxiety, sadness) and, a fortiori, their pathologic equivalents such as anxiety and depressive disorders can inuence
pain perception and symptom reporting by interfering with
systems involved in processing and modulating noxious
(visceral) afferent signals.24,3840 It is conceivable that
similar mechanisms are involved in modulating the processing and, hence, the perception of physiological gut
brain signals such as nutrient-induced signals. However, this
speculative interpretation would require testing using, for
example, functional brain imaging approaches. Furthermore,
it remains to be elucidated why depression levels are
associated specically with postprandial symptom levels for
some symptoms, whereas anxiety levels only seem to be
associated with symptom reporting irrespective of time
relative to meal intake. In any case, these ndings are in line
with the observation that both depressive and anxiety disorders are associated with increased levels of GI symptoms
in general.41 Another candidate mechanism to explain the
association between depression levels and postprandial GI
symptom levels in IBS patients is the (afferent and efferent)
autonomic nervous system (ANS), which acts as the principal interface of the gutbrain axis. Abnormal ANS
responses to a meal indeed have been described in IBS,42
and depressive disorders are associated with abnormal
(efferent) ANS function,43 which may in turn inuence GI
symptom perception by interfering with motor and other
responses of the GI tract to meal ingestion. Finally, we may
speculate that gastrointestinal hormones such as peptide
tyrosine tyrosine (PYY) and/or cholecystokinin (CCK) may
play a role in the associations we found because IBS patients
have abnormal levels of PYY44 and abnormal CCK responses
to a meal.45 Furthermore, CCK may increase rectal mechanosensitivity,46 and mood disorders have been associated
with CCK genetic polymorphisms47 as well as abnormal PYY
levels.48
The mechanisms underlying the associations between
somatization level and (postprandial) GI symptom levels
(which was studied in the Rome II subsample only) may be
even more difcult to speculate. Evidence is growing that
increased somatic symptom reporting (ie, somatization) is a
multidimensional phenomenon. More specically, interrelated psychobiological processes such as negative affectivity (including personality traits such as neuroticism as well

Gastroenterology Vol. 150, No. 4

as mood and anxiety disorders), somatosensory amplication, hypervigilance, fear learning, symptom-associated anxiety, attributions, cognitions, and behaviors may constitute the
key underpinnings of somatization.4951 Applied to our results, we may speculate that most of these processes may
increase the perception of postprandial GI signals, as already
has been shown convincingly for (somatic) pain signals and
symptoms.3,5255 For example, sensory amplication of
physiological nutrient-induced signals as well as hypervigilance, anxiety, or catastrophizing thoughts toward the sensations elicited by these signals may increase their perception
as symptoms. However, because the PHQ-12 operationalizes
somatization as the distress elicited by somatic symptoms,
the associations with somatization we found in our study
may reect a more general association between (psychological) distress and (postprandial) GI symptom reporting, even
though we controlled for levels of anxiety and depression in
the somatization analyses.
This study had a number of limitations that should be
addressed. First, this was a cross-sectional study, thereby
inherently limiting the conclusions that can be drawn
regarding the directionality or causality of the associations
found. Therefore, it remains possible that the higher (postprandial) GI symptom levels are the cause rather than the
effect of increased levels of depression, anxiety, or somatization. However, based on the literature on psychological
amplication of bodily signals outlined earlier, as well as
epidemiologic studies,20,21 we consider the latter interpretation (ie, increasing psychological distress levels induce
increased [postprandial] GI symptom responses) most likely
to be correct. Second, the study was performed on a mixed
Rome II/Rome III sample. Therefore, our results on subtype
should be interpreted with sufcient caution. Furthermore,
somatization data were available only in the Rome II subgroup, representing 65.3% of the total sample. Therefore,
our ndings on somatization cannot be generalized to Rome
III IBS patients, and should be interpreted with somewhat
more caution given the smaller sample size. However, even
this subgroup still can be considered a large sample for this
type of research, which should be considered a strength.
Third, patients were recruited in secondary and tertiary
care, which implies that caution is warranted when
extrapolating these results to other populations of IBS
patients (primary care, nonhealth careseeking patients).
Fourth, because we only included IBS patients in this study,
we cannot exclude the possibility that the associations between anxiety, depression, and somatization levels and
postprandial GI symptom responses are not specic for IBS.
Testing this would require performing a similar study
in healthy volunteers and/or other patient cohorts.
Regarding healthy volunteers, however, we previously
showed that postprandial symptom responses to the standard meal, which also was used in the present study, are
very limited except for fullness,12 leaving very little variability to be putatively explained by psychological symptom
levels. Finally, the time-by-somatization interaction effect
for abdominal pain only showed a nonsignicant trend
(P .054) and therefore should be interpreted with more
caution.

In conclusion, we present evidence that meal ingestion

and level of psychological distress not only contribute
independently to increased GI symptom reporting in IBS,
but that they may in some cases interact with psychological
symptoms, particularly depression and somatization, being
associated with enhanced postprandial GI symptom responses. These ndings are relevant to increase our understanding of the multifactorial nature of IBS, as well as
clinically in that postprandial worsening of symptoms
should lead not only to dietary, but also psychological,
assessment focusing on depression and somatization.

References
1. Longstreth GF, Thompson WG, Chey WD, et al. Functional bowel disorders. Gastroenterology 2006;
130:14801491.
2. Lovell RM, Ford AC. Global prevalence of and risk factors
for irritable bowel syndrome: a meta-analysis. Clin Gastroenterol Hepatol 2012;10:712721.e4.
3. Elsenbruch S. Abdominal pain in irritable bowel syndrome: a review of putative psychological, neural and
neuro-immune mechanisms. Brain Behav Immun 2011;
25:386394.
4. Ohman L, Simren M. Pathogenesis of IBS: role of
inammation, immunity and neuroimmune interactions.
Nat Rev Gastroenterol Hepatol 2010;7:163173.
5. Whitehead WE, Palsson O, Jones KR. Systematic review
of the comorbidity of irritable bowel syndrome with other
disorders: what are the causes and implications?
Gastroenterology 2002;122:11401156.
6. Henningsen P, Zimmermann T, Sattel H. Medically
unexplained physical symptoms, anxiety, and depression: a meta-analytic review. Psychosom Med 2003;
65:528533.
7. Henningsen P, Herzog W. Irritable bowel syndrome and
somatoform disorders. J Psychosom Res 2008;
64:625629.
8. Morcos A, Dinan T, Quigley EMM. Irritable bowel syndrome: role of food in pathogenesis and management.
J Dig Dis 2009;10:237246.
9. Tomba C, Baldassarri A, Coletta M, et al. Is the subjective perception of lactose intolerance inuenced by the
psychological prole? Aliment Pharmacol Ther 2012;
36:660669.
10. Dainese R, Galliani EA, De Lazzari F, et al. Discrepancies
between reported food intolerance and sensitization test
ndings in irritable bowel syndrome patients. Am J
Gastroenterol 1999;94:18921897.
11. Monsbakken KW, Vandvik PO, Farup PG. Perceived food
intolerance in subjects with irritable bowel syndrome etiology, prevalence and consequences. Eur J Clin Nutr
2005;60:667672.
12. Posserud I, Strid H, Strsrud S, et al. Symptom pattern
following a meal challenge test in patients with irritable
bowel syndrome and healthy controls. United European
Gastroenterol J 2013;1:358367.
13. Simren M, Castedal M, Svedlund J, et al. Abnormal
propagation pattern of duodenal pressure waves in the

Postprandial Symptoms in IBS

14.

15.

16.

17.

18.

19.
20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

873

irritable bowel syndrome (IBS). Dig Dis Sci 2000;

45:21512161.
Trnblom H, Van Oudenhove L, Tack J, et al. Interaction
between preprandial and postprandial rectal sensory and
motor abnormalities in IBS. Gut 2014;63:14411449.
Di Stefano M, Miceli E, Missanelli A, et al. Meal induced
rectosigmoid tone modication: a low caloric meal
accurately separates functional and organic gastrointestinal disease patients. Gut 2006;55:14091414.
Rumessen JJ, Gudmand-Hoyer E. Functional bowel
disease: malabsorption and abdominal distress after
ingestion of fructose, sorbitol, and fructose-sorbitol
mixtures. Gastroenterology 1988;95:694700.
King TS, Elia M, Hunter JO. Abnormal colonic fermentation in irritable bowel syndrome. Lancet 1998;
352:11871189.
Serra J, Salvioli B, Azpiroz F, et al. Lipid-induced intestinal gas retention in irritable bowel syndrome. Gastroenterology 2002;123:700706.
Mayer EA, Tillisch K. The brain-gut axis in abdominal pain
syndromes. Ann Rev Med 2011;62:381396.
Sykes MA, Blanchard EB, Lackner J, et al. Psychopathology in irritable bowel syndrome: support for a psychophysiological model. J Behav Med 2003;26:361372.
Koloski NA, Jones M, Kalantar J, et al. The brain-gut
pathway in functional gastrointestinal disorders is bidirectional: a 12-year prospective population-based study.
Gut 2012;61:12841290.
van Tilburg MAL, Palsson OS, Whitehead WE. Which
psychological factors exacerbate irritable bowel syndrome? Development of a comprehensive model.
J Psychosom Res 2013;74:486492.
Thijssen AY, Jonkers DM, Leue C, et al. Dysfunctional
cognitions, anxiety and depression in irritable bowel
syndrome. J Clin Gastroenterol 2010;44:e236e241.
Van Oudenhove L, Aziz Q. The role of psychosocial
factors and psychiatric disorders in functional dyspepsia.
Nat Rev Gastroenterol Hepatol 2013;10:158167.
Thompson WG, Longstreth GF, Drossman DA, et al.
Functional bowel disorders and functional abdominal
pain. Gut 1999;45:II43II47.
Bjelland I, Dahl AA, Haug TT, et al. The validity of the
Hospital Anxiety and Depression Scale: an updated
literature review. J Psychosom Res 2002;52:6977.
Kroenke K, Spitzer RL, Williams JBW. The PHQ-15:
validity of a new measure for evaluating the severity of
somatic symptoms. Psychosom Med 2002;64:258266.
Spiller RC, Humes DJ, Campbell E, et al. The Patient
Health Questionnaire 12 Somatic Symptom scale as a
predictor of symptom severity and consulting behaviour
in patients with irritable bowel syndrome and symptomatic diverticular disease. Aliment Pharmacol Ther 2010;
32:811820.
Francis CY, Morris J, Whorwell PJ. The irritable bowel
severity scoring system: a simple method of monitoring
irritable bowel syndrome and its progress. Aliment
Pharmacol Ther 1997;11:395402.
Talley NJ, Stanghellini V, Heading RC, et al. Functional
gastroduodenal disorders. Gut 1999;45:37ii42ii.

CLINICAL AT

April 2016

874

Van Oudenhove et al

CLINICAL AT

31. Tack J, Talley NJ, Camilleri M, et al. Functional gastroduodenal disorders. Gastroenterology 2006;130:14661479.
32. Cremonini F, Talley NJ. Review article: the overlap between functional dyspepsia and irritable bowel syndrome
- a tale of one or two disorders? Aliment Pharmacol Ther
2004;20:4049.
33. Clauwaert N, Jones MP, Holvoet L, et al. Associations
between gastric sensorimotor function, depression, somatization, and symptom-based subgroups in functional
gastroduodenal disorders: are all symptoms equal?
Neurogastroenterol Motil 2012;24: 1088e565.
34. Lipowski ZJ. Somatization: the concept and its clinical
application. Am J Psychiatry 1988;145:13581368.
35. Elsenbruch S, Lovallo WR, Orr WC. Psychological and
physiological responses to postprandial mental stress in
women with the irritable bowel syndrome. Psychosom
Med 2001;63:805813.
36. Le Neve B, Posserud I, Bohn L, et al. A combined nutrient
and lactulose challenge test allows symptom-based
clustering of patients with irritable bowel syndrome. Am
J Gastroenterol 2013;108:786795.
37. Guthrie E, Barlow J, Fernandes L, et al. Changes in
tolerance to rectal distension correlate with changes in
psychological state in patients with severe irritable bowel
syndrome. Psychosom Med 2004;66:578582.
38. Bushnell MC, Ceko M, Low LA. Cognitive and emotional
control of pain and its disruption in chronic pain. Nat Rev
Neurosci 2013;14:502511.
39. Goesling J, Clauw D, Hassett A. Pain and depression: an
integrative review of neurobiological and psychological
factors. Curr Psychiatry Rep 2013;15:18.
40. Asmundson GJG, Katz J. Understanding the cooccurrence of anxiety disorders and chronic pain:
state-of-the-art. Depress Anxiety 2009;26:888901.
41. Mussell M, Kroenke K, Spitzer RL, et al. Gastrointestinal
symptoms in primary care: prevalence and association
with depression and anxiety. J Psychosom Res 2008;
64:605612.
42. Ng C, Malcolm A, Hansen R, et al. Feeding and colonic
distension provoke altered autonomic responses in irritable bowel syndrome. Scand J Gastroenterol 2007;
42:441446.
43. Kemp AH, Quintana DS, Gray MA, et al. Impact of
depression and antidepressant treatment on heart rate
variability: a review and meta-analysis. Biol Psychiatry
2010;67:10671074.
44. Simren M, Stotzer PO, Sjovall H, et al. Abnormal levels of
neuropeptide Y and peptide YY in the colon in irritable
bowel syndrome. Eur J Gastroenterol Hepatol 2003;
15:5562.

Gastroenterology Vol. 150, No. 4

45. Van Der Veek PPJ, Biemond I, Masclee AAM. Proximal
and distal gut hormone secretion in irritable bowel syndrome. Scand J Gastroenterol 2006;41:170177.
46. van der Schaar PJ, van Hoboken E, Ludidi S, et al. Effect
of cholecystokinin on rectal motor and sensory function
in patients with irritable bowel syndrome and healthy
controls. Colorectal Dis 2013;15:e29e34.
47. Wilson J, Markie D, Fitches A. Cholecystokinin system
genes: associations with panic and other psychiatric
disorders. J Affect Disord 2012;136:902908.
48. Gimenez-Palop O, Coronas R, Cobo J, et al. Fasting
plasma peptide YY concentrations are increased in
patients with major depression who associate weight
loss. J Endocrinol Invest 2012;35:645648.
49. Duddu V, Isaac MK, Chaturvedi SK. Somatization, somatosensory amplication, attribution styles and illness
behaviour: a review. Int Rev Psychiatry 2006;18:2533.
50. De Gucht V, Maes S. Explaining medically unexplained
symptoms: toward a multidimensional, theory-based
approach to somatization. J Psychosom Res 2006;
60:349352.
51. Naliboff BD. Towards a nondualistic approach to multisystem illness. Am J Gastroenterol 2007;102:27772780.
52. Sturgeon J, Zautra A. Psychological resilience, pain
catastrophizing, and positive emotions: perspectives on
comprehensive modeling of individual pain adaptation.
Curr Pain Headache Rep 2013;17:19.
53. Lucchetti G, Oliveira A, Mercante J, et al. Anxiety and
fear-avoidance in musculoskeletal pain. Curr Pain
Headache Rep 2012;16:399406.
54. Wiech K, Ploner M, Tracey I. Neurocognitive aspects of
pain perception. Trends Cogn Sci 2008;12:306313.
55. Crombez G, Van Damme S, Eccleston C. Hypervigilance
to pain: an experimental and clinical analysis. Pain 2005;
116:47.

Received May 4, 2015. Accepted November 11, 2015.

Reprint requests
Address requests for reprints to: Magnus Simrn, MD, PhD, Department of
Internal Medicine and Clinical Nutrition, Institute of Medicine, Sahlgrenska
Academy, University of Gothenburg, 41345 Gothenburg, Sweden. e-mail:
magnus.simren@medicine.gu.se; fax: 46 31 17412917.
Conicts of interest
The authors disclose no conicts.
Funding
Supported by the Swedish Medical Research Council (grants 13409, 21691
and 21692), the Marianne and Marcus Wallenberg Foundation, the University
of Gothenburg Centre for Person-Centered Care (Sahlgrenska Academy,
University of Gothenburg, Gothenburg, Sweden), and by the Faculty of
Medicine (University of Gothenburg). Also funded by the KU Leuven Special
Research Fund (Bijzonder Onderzoeksfonds to L.V.O.).