Sie sind auf Seite 1von 11

Evidence of the Impact of Diet, Fluid Intake, Caffeine, Alcohol

and Tobacco on Lower Urinary Tract Symptoms: A


Systematic Review
Catherine S. Bradley,* Bradley A. Erickson, Emily E. Messersmith,
Anne Pelletier-Cameron,† H. Henry Lai, Karl J. Kreder, Claire C. Yang,
Robert M. Merion, Tamara G. Bavendam and Ziya Kirkali, for the Symptoms
of Lower Urinary Tract Dysfunction Research Network (LURN)
From the University of Iowa Carver College of Medicine, Iowa City, Iowa (CSB, BAE, KJK), Arbor Research Collaborative for Health
(EEM, RMM) and University of Michigan (AP-C), Ann Arbor, Michigan, Washington University School of Medicine, St. Louis, Missouri (HHL),
University of Washington, Seattle, Washington (CCY), and National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes
of Health, Bethesda, Maryland (TGB, ZK)

Purpose: Diet, fluid intake and caffeine, alcohol and tobacco use may have effects
Abbreviations
on lower urinary tract symptoms. Constructive changes in these modifiable
and Acronyms
nonurological factors are suggested to improve lower urinary tract symptoms. To
AUA-SI ¼ American Urological better understand the relationship between nonurological factors and lower
Association Symptom Index
urinary tract symptoms, we performed a systematic literature review to
BPH ¼ benign prostatic examine, grade and summarize reported associations between lower urinary
hyperplasia tract symptoms and diet, fluid intake and caffeine, tobacco and alcohol use.
LUTS ¼ lower urinary tract Materials and Methods: We performed PubMedÒ searches for eligible articles
symptoms
providing evidence on associations between 1 or more nonurological factors and
NS ¼ nonsignificant association lower urinary tract symptoms. A modified Oxford scale was used to grade the
NUF ¼ nonurological factor evidence.
OAB ¼ overactive bladder Results: We reviewed 111 articles addressing diet (28 studies), fluid intake (21)
RCT ¼ randomized controlled trial and caffeine (21), alcohol (26) and tobacco use (44). The evidence grade was
SS ¼ statistically significant generally low (6% level 1, 24% level 2, 11% level 3 and 59% level 4). Fluid intake
association and caffeine use were associated with urinary frequency and urgency in men and
SUI ¼ stress urinary incontinence women. Modest alcohol use was associated with decreased likelihood of benign
prostatic hyperplasia diagnosis and reduced lower urinary tract symptoms in
UI ¼ urinary incontinence
men. Associations between lower urinary tract symptoms and ingestion of
UUI ¼ urgency urinary certain foods and tobacco were inconsistent.
incontinence
Conclusions: Evidence of associations between lower urinary tract symptoms
and diet, fluid intake and caffeine, alcohol and tobacco use is sparse and mostly
observational. However, there is evidence of associations between increased fluid

Accepted for publication April 30, 2017.


No direct or indirect commercial incentive associated with publishing this article.
The corresponding author certifies that, when applicable, a statement(s) has been included in the manuscript documenting institutional
review board, ethics committee or ethical review board study approval; principles of Helsinki Declaration were followed in lieu of formal ethics
committee approval; institutional animal care and use committee approval; all human subjects provided written informed consent with
guarantees of confidentiality; IRB approved protocol number; animal approved project number.
This is publication number 3 of the Symptoms of Lower Urinary Tract Dysfunction Research Network (LURN) and is supported by Grants
DK097780, DK097772, DK097779, DK099932, DK100011, DK100017, DK097776 and DK099879 from the National Institute of Diabetes and
Digestive and Kidney Diseases through cooperative agreements.
Supplementary references 51 to 126 for this article can be obtained at http://jurology.com/.
* Correspondence: Department of Obstetrics and Gynecology, University of Iowa Hospitals and Clinics, 200 Hawkins Dr., Iowa City, Iowa
52242 (e-mail: catherine-bradley@uiowa.edu).
† Financial interest and/or other relationship with Medtronic.

0022-5347/17/1985-1010/0 http://dx.doi.org/10.1016/j.juro.2017.04.097

1010 j www.jurology.com
THE JOURNAL OF UROLOGY®
Ó 2017 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
Vol. 198, 1010-1020, November 2017
Printed in U.S.A.
LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS 1011

and caffeine intake and urinary frequency/urgency, and between modest alcohol intake and decreased benign
prostatic hyperplasia diagnosis and lower urinary tract symptoms. Given the importance of these non-
urological factors in daily life, and their perceived impact on lower urinary tract symptoms, higher quality
evidence is needed.

Key Words: lower urinary tract symptoms, risk factors, urination disorders, health surveys, self report

LOWER urinary tract symptoms are common and performed to identify publications studying associations
bothersome, affecting 20% to 50% of men and women between LUTS and each of the 5 NUFs. A search string
and negatively impacting health related quality of was developed for LUTS and each factor, including
life.1e3 Patients seeking care for lower urinary tract MeSHÒ terms and key words for text searches, limited to
English language publications (supplementary table 1,
symptoms are frequently instructed to modify daily
http://jurology.com/).
behaviors to reduce symptoms. For example pro-
All citations and abstracts were screened using previ-
viders may recommend that patients change fluid ously developed eligibility criteria (table 1). If the initial
intake, or use less caffeine or alcohol.4 The quantity screener was unsure whether to include a citation, a
and quality of evidence to support such recommen- second investigator reviewed it. When uncertainty per-
dations are unclear. Lifestyle changes, while typi- sisted, the citation was included for additional review at
cally low risk, may be obtrusive to the lives of patients the full text stage. Each article considered eligible after
and may increase anxiety or stress. What patients screening was reviewed (full text) by 2 investigators. All
eat, drink and ingest depends on culture, region, articles confirmed eligible were assigned a level of evi-
employment, socioeconomic status and other factors. dence by both reviewers using a system based on the
These behaviors are part of the daily human experi- Oxford Centre for Evidence-Based Medicine Level of
Evidence scale (2009 version),7 and International
ence, and as such, a better understanding of their
Consultation on Urological Diseases steps for developing
impact on lower urinary tract symptoms is critical.
and grading guideline recommendations modified to
LURN (Symptoms of Lower Urinary Tract include cross-sectional studies as level 4b evidence (see
Dysfunction Research Network) is a cooperative Appendix).8 If initial grades differed, investigators
network supported by the NIDDK (National Insti- arrived at a grade by consensus.
tute of Diabetes and Digestive and Kidney Diseases) Data from each article were reviewed and abstracted
with objectives to improve the measurement of LUTS using a standard form. Information collated included
and identify important LUTS subtypes.5 In concep- study design, population, LUTS outcome (eg OAB), NUF
tualizing the scope of lower urinary tract dysfunction exposure (eg caffeine), summary measure of association
and its resultant symptoms we considered multiple and analysis performed. Meta-analyses were not per-
factors that potentially contribute to LUTS. The ob- formed given the heterogeneous study designs, outcomes
and exposures identified.
jectives of this study were to identify, grade and
summarize peer-reviewed literature examining as-
sociations between LUTS and diet, fluid intake and RESULTS
caffeine, alcohol and tobacco use. In addition to Electronic searches were performed through
identifying evidence-based associations between January 4, 2016. Results of the searches, screening
these factors and LUTS, the results will help identify and selection process, and reasons for exclusion are
gaps where future efforts may be focused. presented in tables 1 and 2. We reviewed 111
unique articles in the areas of diet (28 studies), fluid
Table 1. Criteria for excluding articles from systematic review
METHODS
This systematic review was designed to answer the Reasons for Exclusion No. Articles
question, “Are diet, fluid intake and caffeine, alcohol and
No relevant nonurological factor studied 158
tobacco use associated with the prevalence and/or severity No relevant LUTS or lower urinary tract condition 83
of LUTS in men and women?” The review used findings (including prostate Ca)
from RCTs, cohort, case-control and case series, and cross- LUTS studied as treatment result or adverse effect 24
sectional studies that could provide evidence related to (eg postprostatectomy incontinence)
Sample size smaller than 25 pts (unless RCT design) 3
these associations. Research focused on bladder pain and
Editorial/commentary/nonsystematic review 76
conditions such as interstitial cystitis/bladder pain syn- Nonrelevant research type (eg qualitative studies, 6
drome was excluded. This systematic review was based on instrument development)
guidelines put forth by PRISMA (Preferred Reporting Not human subject research 9
Items for Systematic Reviews and Meta-Analyses).6 Pediatric population 5
Pregnant population 1
PubMed searches were developed with assistance from
a health science librarian. Five separate searches were More than 1 reason for exclusion may be listed for an individual article.
1012 LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS

Table 2. Publications from PubMed searches included in review

No. Publications Eligible No. Publications Eligible No. Additional Articles Total No. Publications Included
Nonurological Factor No. References after Screening Citation/Abstract after Full Text Review Identified during Review in Systematic Review
Diet 128 31 28 0 28
Fluid intake 207 30 19 2 21
Caffeine use 36 22 20 1 21
Alcohol use 66 30 26 0 26
Tobacco use 126 45 44 0 44
Totals 563 158 137 3 111*

* Some publications provided results for more than 1 nonurological factor.

intake (21) and caffeine (21), alcohol (26) and allium vegetables, eg onion and garlic) and poly-
tobacco use (44). A total of 22 articles provided unsaturated fats (including specifically eicosapentae-
results on more than 1 factor. The evidence grade noic and docosahexaenoic acids) and low in saturated
was generally low (6% of studies were level 1, 24% fat may be associated with decreased risk.15e20 Studies
level 2, 11% level 3 and 59% level 4). We summarize on micronutrients have implicated carotene to
the publications reviewed and synthesize the decrease risk, while zinc may increase risk.21,22
results related to the association between each NUF
Diet and Urinary Incontinence. Many studies that
and LUTS.
assess diet and UI indirectly evaluate diet through
weight loss. Our review found 2 RCTs conducted in
Diet and LUTS
women, of which 1 showed that intensive lifestyle
A total of 28 publications met criteria and provided
therapy (including a low calorie diet) decreased UI
information related to the association between diet
in prediabetic women9 and 1 demonstrated that
and LUTS (table 3). Diet was assessed by a food-
modest weight reduction (mean 7.8 kg) decreased
frequency questionnaire in almost all studies. There
SUI episodes (but not urge).12 In addition, weight
were 5 interventional RCTs.9e13 Only 2 of these series
loss of 5% to 10% of body weight is sufficient to
specifically addressed adding a particular component
significantly decrease UI episodes.14 Another RCT
to improve LUTS.10,13 The remaining studies
specifically examined urgency incontinence
analyzed diet and LUTS in the setting of interventions
episodes in nursing home patients and found that
concerning diabetes prevention,9 constipation reduc-
with combined toileting assistance, exercise and
tion11 and weight loss (supplementary tables 2A to
an increase in caloric intake UI episodes decreased
2D, http://jurology.com/).12,14
significantly.11 Another RCT evaluating a diet rich
Diet and BPH. Overall there appears to be a weak in soy, which is hypothesized to increase
association between diet and surgical BPH therapy. A circulating estrogens via phytoestrogens, revealed
myriad of food types, food groups, micronutrients and no improvement in overall LUTS or UI when
macronutrients were evaluated. Consumption of a compared to a control diet.10
high calorie diet that is high in starches and red meat Studies assessing the association of dietary com-
may be weakly associated with BPH risk, while a lower ponents and incontinence found that consump-
calorie diet that is high in vegetables (specifically tion of saturated and monounsaturated fats, and

Table 3. Studies on diet and LUTS

LUTS

Nonspecific Urinary
BPH15e22,92,93 Incontinence9e12,14,23,24,94,95 Symptoms13,25e29,96 OAB30e32
No. articles 9 9 7 3
No. study 2 Cohort, 6 case-control, 4 RCT, 4 cohort, 1 cross-sectional 1 RCT, 1 cohort, 5 cross-sectional 3 Cohort
design 1 cross-sectional
No. population 9 Men 8 Women only, 1 men þ women 5 Men only, 1 women only, 1 2 Women only, 1 men only
study types men þ women
Results Mixed (pos SS in 7 studies, Mixed (pos SS in 6 studies, Mixed (pos SS in 3 studies, Mixed (pos SS in 1 study, neg SS
neg SS in 7, NS in 7) neg SS in 3, NS in 5) neg SS in 2, NS in 2) in 2, NS in 3)
Comments Diet may indirectly affect prostate Most effects of diet occur Overall weak associations; low caloric Weak associations noted,
growth through influences on indirectly through wt alteration intake, high in polyunsaturated fats þ mostly indirect through wt
androgens; low caloric intake, vegetables may be protective against gain/loss; some evidence of
high in polyunsaturated fats LUTS; vitamin C may be protective effects on OAB through diabetes
may be protective when obtained from food mechanism (neurovascular)

Some studies tested more than 1 association.


LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS 1013

carbonated beverages may increase the risk of SUI, Fluid Intake and BPH. It is unclear from the few pub-
while intake of breads/starches and vegetables may lished articles whether there is an association be-
decrease the risk.23,24 Interestingly, similar to the tween fluid intake and BPH. In 1 RCT 138 men with
association seen with BPH in men, zinc intake was BPH were randomized to increase fluid intake by 1.5
associated with SUI in women, as was vitamin B12. L per day vs placebo.33 There was no difference in
Consumption of phytoestrogens did not affect SUI. AUA-SI total, voiding or quality of life scores
between the groups at 6 months. However, AUA-SI
Diet and General Urinary Symptoms. Validated ques-
storage scores were worse in those who increased
tionnaires were commonly used to test associations
fluid intake (effect size 1.3, p <0.001). In a
between diet and LUTS. Most studies were cross-
nonrandomized uncontrolled study AUA-SI
sectional, and thus determining causality becomes
increased from 7.9 to 8.9 (p ¼ 0.028) after subjects
more difficult given that researchers believe lifetime
increased fluid intake by 2 L daily for 8 weeks.34
exposure is more important for health than current
However, the magnitude of symptom worsening
diet. Clinically significant LUTS were associated
was small and likely clinically insignificant.
with poor overall diet, dietary variety and sodium
intake in men,25,26 and with increased total caloric Fluid Intake and Nocturia. There is not a clear asso-
intake (adjusting for weight) in women.27 Protein ciation between fluid intake and nocturia. In a large
intake may decrease the risk in men but increases RCT 307 women were randomized to receive tol-
the risk of storage symptoms in women. In elderly terodine vs tolterodine plus behavioral therapy that
men consumption of isoflavone (a phytoestrogen) included pelvic floor muscle exercise training,
showed a strong correlation with LUTS.28 However, bladder control techniques and fluid management.35
a single RCT evaluated the effects of flaxseed extract No difference in numbers of voids at night was
(which is high in phytoestrogens) on LUTS and noted found between treatment groups at 10 weeks.
a dose dependent, significant decrease in LUTS in However, this RCT did not specifically examine
men on flaxseed.13 In a longitudinal cohort study fluid management since other behavioral therapies
greater vitamin C intake at baseline was associated were included and all subjects received general
with less progression of storage LUTS but vitamin C information to avoid excessive fluid intake.
supplementation was associated with worse LUTS at Similarly a large cohort study conducted in men
5-year followup in women.29 found no association between nighttime fluids and
incident nocturia.36 In contrast, 2 uncontrolled
Diet and OAB. Few studies have directly assessed OAB
case series demonstrated that fluid restriction
and diet. There appears to be a weak association with
improved nocturia.37,38
potato/starch consumption.30 Evaluation of
micronutrients suggests increased consumption of Fluid Intake and OAB Symptoms. Six of 7 studies of
vitamin D, protein and potassium may be protective OAB symptoms, including 2 small crossover RCTs,
against OAB in women.31 High energy/caloric intake revealed a positive association between fluid intake
in the setting of high glycemic indices and low and urinary frequency/urgency. One series ran-
physical activity may also be a risk factor in women.32 domized 69 women to caffeine restriction plus fluid
increase to 3 L daily, compared to caffeine restric-
Fluid Intake and LUTS tion plus fluid decrease to 750 ml daily.39 Another
Table 4 summarizes the 21 publications reviewed. series included 24 men and women randomized to
Supplementary tables 3A to 3D (http://jurology.com/) increase vs decrease daily fluid intake by 25%
provide additional details. compared to individual baseline.40 Both trials

Table 4. Studies on fluid intake and LUTS

LUTS

OAB/Non-UI
BPH33,34,97 Nocturia35e38,97,98 Symptoms24,35,39,40,98e100 UI11,24,35,39e42,56,59,61,98,100e102

No. articles 3 6 7 14
No. study design 1 RCT, 1 outcomes research, 1 RCT, 1 cohort, 2 outcomes 3 RCT, 1 cohort, 3 cross-sectional 5 RCT, 2 cohort, 1 case series,
1 cross-sectional research, 2 cross-sectional 6 cross-sectional
No. population 3 Men 3 Men only, 2 men þ women, 5 Women only, 2 men þ 10 Women only, 1 men only,
study types 1 women only women 3 men þ women
Results Mixed (pos SS in 1 study, Mixed (pos SS in 2 studies, Pos association in most studies Mixed (pos SS in 6 studies,
NS in 2) NS in 4) (pos SS in 6 studies, NS in 1) neg SS in 1, NS in 7)
Comments Fluid intake worsens storage Evidence inconclusive on whether Most studies showed pos association Results inconclusive for association
symptoms, has no impact fluid restriction reduces nocturia between fluid intake and frequency between fluid intake and UI
on voiding symptoms and/or urgency symptoms
1014 LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS

found significantly increased frequency and urgency to coffee intake. Three studies of men with BPH yiel-
symptoms with fluid increase, and decreased ded conflicting results. Two older case-control studies
frequency and urgency with fluid reductions. of men with surgically treated BPH revealed
Fluid Intake and UI. A total of 14 articles showed nonsignificant associations between coffee intake
mixed results regarding the association between and BPH,45,46 while a large, population based, cross-
fluid intake and UI, with 6 demonstrating positive, sectional study showed increasing coffee
1 negative and 7 no correlation. In a small RCT consumption was positively associated with BPH.47
increasing fluid intake worsened weekly UI epi- Caffeine and Nocturia. Coffee consumption was not
sodes, while decreasing fluid intake improved associated with nocturia in a cohort study of men and
weekly incontinence episodes. In contrast, an RCT a large cross-sectional study of women.48,49 In
by Zimmern et al did not reveal any correlation.35 contrast, the same cross-sectional study found that
One RCT did not yield any useful results since tea intake was associated with increasing nocturia,
most patients were unable to adhere to the fluid although the increased risk was small (OR 1.2) and
protocols.41 Two cohort studies examined new was seen only if drinking 3 or more cups of tea
onset UI, with 1 finding that consumption of daily. Another large cross-sectional study found
carbonated drinks was associated with new onset that women with nocturia were less likely to drink
SUI,24 and 1 not finding an association between caffeinated beverages after 6 p.m.50 This finding
fluid intake and new UI.42 highlights the limitations of cross-sectional
Caffeine Intake and LUTS analyses since women who have nocturia may be
The results of 21 studies of caffeine intake and LUTS likely to avoid caffeinated beverages in the evening.
are outlined in table 5, with detailed information Caffeine and OAB/LUTS. Small, randomized, inter-
provided in supplementary tables 4A to 4E (http:// ventional studies provide limited evidence that
jurology.com/). Most series were observational but 2 caffeine reduction may decrease OAB symptoms in
small RCTs were reviewed. Most assessed caffeine women. In 2 RCTs of mostly women caffeine
intake as self-reported cups of coffee or milligrams of reduction was associated with decreased urinary
caffeine ingested daily, estimated using a self- frequency, urgency and OAB quality of life
reported food-frequency questionnaire or other non- scores.51,52 A large population based cohort study
validated questions. Two studies in women focused found that caffeine intake was associated with
on urodynamic test findings related to caffeine. One LUTS progression in men and with urgency in
small study found that detrusor pressure during women. Increasing intake of coffee and soda at
filling increased (but other parameters were un- 5 years compared to baseline was also associated
changed) after caffeine ingestion.43 Another demon- with progression of LUTS, especially storage
strated that detrusor overactivity in women with UI symptoms.53 Caffeine restriction was not
was associated with greater caffeine intake.44 associated with changes in frequency/urgency in a
Caffeine and BPH. It is unclear if caffeine intake and third uncontrolled study.39 Coffee consumption in
BPH are related, and evidence reviewed was limited a large cross-sectional study was not associated
Table 5. Studies on caffeine intake and LUTS

LUTS

OAB/Non-UI
BPH45e47 Nocturia48,50,54 Symptoms39,51e55 UI39,49,51,52,56e61,103,104 Urodynamic Parameters43,44
No. articles 3 3 6 12 2
No. study design 2 Case-control, 1 1 Cohort, 2 2 RCT, 1 cohort, 1 case 2 RCT, 2 cohort, 1 case 1 Case series, 1
cross-sectional cross-sectional series, 2 cross-sectional series, 7 cross-sectional cross-sectional
No. population 3 Men only 1 Men only, 2 3 Women only, 2 men þ 9 Women only, 2 men þ 2 Women only
study types women only women, 1 men only women, 1 men only
Results Mixed (pos þ neg NS in Mixed (pos SS in 1 study, Pos association in higher Mixed (pos SS in 5 studies, Pos SS in 2 studies
2 studies, pos SS in 1) neg SS in 1, pos NS in 1, evidence studies (pos SS pos NS in 2, pos þ neg
pos þ neg NS in 1) in 4 studies, pos þ neg NS in 4, neg NS in 1)
NS in 2)
Comments Unclear if caffeine or other Tea (but not coffee) Caffeine restriction is Overall evidence may Pos results but
coffee constituents intake is associated associated with small suggest weak pos conflicting findings
influence BPH with nocturia improvements in frequency association between caffeine (varying end points)
and urgency in women; and UI but conflicting results
caffeine (coffee) associated for UI types; studies
with symptom progression in men are lacking
and urgency in men and
women

Some studies tested more than 1 association.


LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS 1015

with urgency, but women reporting tea intake of 3 articles reviewed, with a decrease in BPH diagnosis
or more cups daily were more likely to report or surgery in men who drank alcohol compared to
urgency.54 A large cross-sectional study in men nondrinkers. The association between alcohol
found that caffeine intake was not associated with intake and reduction in BPH diagnosis and surgery
increased LUTS.55 was particularly strong for modest alcohol con-
sumption (defined as 1 to 3 drinks daily), with this
Caffeine and UI. Overall evidence may suggest a
consumption level having the greatest BPH reduc-
weak positive association between caffeine and UI
tion when compared to nondrinkers.
but there are conflicting results for UI types, and
LUTS, UI and nocturia in men alone were
studies in men are lacking. Four interventional se-
assessed in 14 articles and had relatively consistent
ries (2 randomized, 2 uncontrolled) found no impact
findings, with modest drinking being associated
of caffeine reduction on UI.39,51,52,56 A longitudinal
with fewer symptoms when compared to non-
study in women found caffeine associated with
drinkers in all but 2 articles. However, heavy
frequent UI and UUI but only in women with the
alcohol consumption (defined as self-reported alco-
greatest caffeine intake.57 Caffeine intake was not
holism, ie more than 72 gm or 5.1 drinks daily, or
associated with UI progression when analyzed
more than 40 gm or 2.9 drinks daily) appears to
using the same longitudinal data.58 Mixed results
have a negative effect, with an increase in inconti-
were found in several large cross-sectional studies
nence and obstructive and irritative LUTS.62 This
in women, with 1 finding coffee/tea intake
association (J-shaped curve) of alcohol intake and
associated with SUI and tea intake with overall UI
LUTS was most clearly seen in a cross-sectional
(but not UUI), and 1 finding no association
study of 30,196 Korean men participating in a
between coffee/tea consumption and any type of
comprehensive health examination.63 In that series
UI.53,59,60 One large cross-sectional study of men
the odds of moderate or severe AUA-SI scores were
found the highest level of caffeine intake
lowest among modest drinkers (0 to 10 gm daily)
associated with moderate to severe UI.61
compared to nondrinkers, and highest in men who
drank more than 40 gm daily.
Alcohol and LUTS
A total of 26 articles on alcohol intake and LUTS were Alcohol and Nocturia. Only 2 articles assessed noc-
identified, reviewed and graded (table 6). Study turia. In a cohort consisting of both genders no
details are presented in supplementary tables 5A association was found between alcohol intake and
to 5E (http://jurology.com/). Most publications nocturia.64 By comparison, in a single article
assessed alcohol intake as self-reported drinks per including only men modest alcohol intake carried
day, week or month, or grams of alcohol consumed the lowest risk of moderate or severe nocturia.48
based on self-report. Four articles analyzed alcohol
Alcohol and OAB/UI. Among 3 articles assessing OAB
type (beer, wine, spirits) as well as total consumption.
there were inconsistent findings. Results from the
Alcohol and BPH and LUTS in Men. Results for BPH BACH (Boston Area Community Health) Survey
and BPH surgery were consistent among 7 of 8 demonstrated inconsistent findings by intake level

Table 6. Studies on alcohol and LUTS

LUTS

Male
BPH17,46,47,69,105e109 Symptoms19,55,62,63,96,110e114 Nocturia48,64 OAB/Non-UI Symptoms64e66 UI60,64,96,113,115

No. articles 9 10 2 3 5
No. study design 2 Cohort, 3 case-control, 2 Cohort, 8 cross-sectional 1 Cohort, 1 1 Cohort, 2 cross-sectional 3 Cross-sectional,
3 cross-sectional, 1 meta- cross-sectional 1 case series
analysis
No. population 9 Men only 10 Men only 1 Men only, 1 women 3 Women þ men 1 Men only, 3 women only, 1
study types þ men women þ men
Results Consistent (pos SS in Consistent (pos SS in Inconsistent (pos SS Inconsistent (NS in 2 Pos SS in 1 study, pos NS
8 studies, NS in 1) 7 studies, ie neg association in 1 subgroup, studies, SS for subgroup in 2, pos þ neg NS in 2
with modest alcohol use, NS in 1 study) in 1)
pos association with heavy
alcohol use; NS in 2,
variable results in 1)
Comments Reduced BPH surgery in Fewer symptoms in modest Inconsistent results Inconsistent findings No significant
pts who drank alcohol drinkers compared to association noted
compared to nondrinkers nondrinkers in all but 2
in all but 1 study studies; alcoholism had neg
effect on urinary symptoms
1016 LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS

and symptom subtype, with few groups achieving symptoms in men or voiding symptoms in men or
statistical significance.65 However, an interview women, but a positive association with obstructive
study of 833 elderly individuals found increased symptoms in men and storage symptoms in
odds of urgency and frequency among current women.62,75
drinkers compared to nondrinkers but not former
Tobacco and UI. We found no evidence to review
drinkers.66 In 4 articles no association was found
regarding UI in men. In women studies have shown
between any type of UI and alcohol intake.
inconsistent results, with some providing evidence of
a positive association between tobacco use and SUI,
Tobacco and LUTS UUI and mixed incontinence, motor incontinence and
A total of 44 articles on tobacco use and LUTS were incontinence of any (unspecified) type.60,76e79 Six
systematically reviewed (table 7, with detailed studies demonstrated no association and 1 found a
summary included in supplementary tables 6A to negative association between occasional UI and
6E, http://jurology.com/). Most series were cross- current smoking.70,72,74,79e82 In addition, Hannestad
sectional but some cohort and case-control studies et al observed mixed results between current, former
were reviewed. Tobacco use was almost always and heavy smoking and various measures of
studied as self-reported current cigarette smoking. incontinence.60 The 2 studies that examined severe
Tobacco and BPH. Of 12 articles identified 8 revealed UI revealed a positive association.60,79
no association between BPH and tobacco use. Four Tobacco and Other LUTS Measures. A few studies
studies found a negative association between heavy focused on other parameters. One study each
or current smoking and BPH but no trend in the as- showed positive associations between smoking and
sociation with quantity of cigarettes smoked.47,67e69 maximum cough spike in women,83 cough leak point
Tobacco and Nocturia. Evidence regarding nocturia pressure and maximum intravesical pressures
was inconsistent. Of 6 studies 1 showed a positive generated by cough in women,84 and male estradiol
association between current smoking and nocturia levels.85 Two studies demonstrated evidence that
in women,50 2 demonstrated a negative association male smokers were at decreased risk for low urinary
in men69 and women,70 and 1 revealed a negative flow rates.86,87 However, another study revealed no
association with heavy current smoking (but not association.85 No associations were found between
lighter current smoking) in a sample of men and tobacco use and female maximum closure pressure,
women.71 Two studies showed no association.48,72 mean pressure transmission ratio or degree of
urethral axis with stress,83 or male testosterone,
Tobacco and OAB. Within the broad category of evi- dehydroepiandrosterone or dehydroepiandrosterone
dence of OAB or LUTS in general there are mixed sulfate levels, or prostate weight.85
findings. A small amount of evidence suggests
former and/or current smoking is related to fre- Key Points
quency in women. Two studies demonstrated a Expert consensus of the reviewed literature sug-
positive association between urgency and current gests a balanced low calorie/low saturated fat diet in
tobacco use, while 2 did not.70,72e74 A single study a physically active, nonobese person will decrease
each revealed no association with irritative the lifetime likelihood of LUTS and/or BPH.

Table 7. Studies on tobacco use and LUTS

LUTS

OAB/Non-UI Other Urodynamic


Symptoms55,62,70, Parameters/Plasma
BPH17,45e47,67e69,86,87,116e118 Nocturia48,50,69e72 72e75,86,87,96,110,112,114,119e124
UI60,70,72,74,76e82,119,125 Levels83e87
No. articles 12 6 19 13 5
No. study design 4 Cohort, 6 cross-sectional, 5 Cross-sectional, 18 Cross-sectional, 1 cohort 2 Case-control, 1 Case-control,
2 case-control 1 cohort 11 cross-sectional 1 case series,
3 cross-sectional
No. population 12 Men only 3 Women only, 6 Women only, 11 men 13 Women only 2 Women only,
study types 2 men only only, 2 women þ men 3 men only
Results Mixed (NS in 7 studies, Mixed (NS in 2 studies, Mixed (NS in 7 studies, pos Mixed (NS in 6 studies, Mixed (pos NS/SS in 2
pos NS/SS in 1, neg neg NS/SS in 1, pos NS/SS in 3, pos SS in 7, pos/neg NS/SS in 1, studies, pos SS in 1,
NS/SS in 2, neg SS in 2) SS in 1, neg SS in 2) not reported in 2) pos SS in 3, pos/neg SS neg SS in 2)
in 1, not reported in 2
Comments Some evidence of neg Inconsistent results Some evidence of pos Some evidence of pos Outcomes vary,
association, more evidence association, other association, other some evidence
of no association evidence demonstrates evidence shows of associations
no association no association
LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS 1017

However, current evidence of associations between a BPH diagnosis and reduced LUTS. As alcohol
diet and individual dietary constituents and LUTS intake of this level falls within federal guidelines,
is mixed and suggests that associations, if present, modest alcohol consumption may be considered a
are weak. reasonable recommendation in clinical practice.
Fluid intake is associated with urinary frequency Fluid intake was positively associated with urinary
and urgency in men and women, and the association frequency and urgency symptoms in men and
is bidirectional. Evidence supports the use of fluid women in 2 small interventional studies and in
reduction to manage urinary urgency (as in the AUA observational studies of mixed quality. Caffeine
guideline on OAB).4 Given potential risks of dehy- intake was also positively associated with urinary
dration, recommendations that patients reduce fluid storage symptoms (frequency and urgency) in men
intake by 25%, provided they do not drink less than and women in observational and small interven-
1 L daily, seem reasonable.88 Relationships between tional studies. These findings support inclusion of
UI, nocturia and fluid intake are less conclusive. fluid management and caffeine reduction within the
We found inconsistent associations between behavioral strategies recommended as first-line
caffeine intake and BPH and nocturia. Evidence treatments for OAB.4 In other areas of our review
suggests that caffeine intake is associated with inconsistent results or lack of evidence precluded
urinary frequency and urgency in men and women, conclusions about associations between NUFs
and that caffeine reduction may decrease urinary and LUTS.
frequency and urgency in women (small effects). Strengths of this effort include our standardized
Conflicting results related to caffeine and overall UI protocol used for screening citations, determining
and UI subtypes suggest any association, if present, eligibility for inclusion and abstracting study re-
is weak. sults. We used an accepted and widely applied
Moderate/modest alcohol consumption in men is system for grading the evidence of the included
associated with a reduced risk of BPH and BPH studies. When possible, we attempted to synthesize
surgery as well as decreased LUTS when compared results to help clarify clinical usefulness of the
to nondrinkers. However, excessive alcohol intake literature.
above the recommended threshold of healthy con- One limitation is that most of the evidence was
sumption is associated with worse LUTS in men. from observational studies, and we did not attempt
These results are consistent with the U.S. Depart- meta-analyses due to the heterogeneous search
ment of Health and Human Services recommenda- results. Few series in any single area included
tion regarding alcohol consumption for men (2 or sufficiently similar outcomes and exposures such
fewer drinks daily).89 In contrast, we found no evi- that statistical integration would be useful or
dence in women that avoidance of alcohol decreases valid. Given that our results included lower evi-
the risk of UI or LUTS. dence levels, caution must be taken in making
We did not find strong evidence that smoking clinical recommendations based on these find-
increases UI. Indeed, we found mixed and thus ings.90,91 However, systematic review of observa-
weak evidence of any associations between tobacco tional studies may be an important alternative
use and LUTS. There was perhaps a positive asso- when RCTs either cannot produce the evidence
ciation between smoking and urinary frequency in needed or would be unethical.91 For example a
women, based on limited evidence. dietary factor may require an extended duration
of exposure to cause LUTS, which could not be
feasibly assessed in a RCT. In another example
DISCUSSION RCTs measuring the impact of tobacco on LUTS
We systematically reviewed the literature exam- would be unethical.
ining associations between LUTS and daily behav- Another weakness of much of the evidence
iors, including diet, fluid intake and caffeine, alcohol included in this review is use of self-report mea-
and tobacco use. We identified 111 articles meeting surement to assess exposures. Most of the reviewed
our eligibility criteria, graded their evidence, and articles evaluated exposures using self-report, often
summarized the findings by population and lower by interview or questionnaire. Thus, the summa-
urinary tract condition studied. Overall relatively rized associations between the NUFs and LUTS are
few, largely observational, studies were eligible subject to limitations of participant memory and
(fewer than 50 per factor), and evidence quality social desirability effects. Other tools such as elec-
was low. tronic diaries and biomarkers may provide more
Based on our review, few definitive conclusions valid assessment of exposure in future studies.
about associations could be drawn. In observational However, any research into lifestyle factors is likely
studies of men modest alcohol use (compared to to face challenges in obtaining accurate and unbi-
nonuse) was associated with decreased likelihood of ased measurements of these factors.
1018 LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS

The challenges in performing research on life- Menendez, Alexis Siurek, Meera Tavathia and
style factors may partly explain the lack of higher Veronica Venezuela.
quality evidence available. Despite this fact (or NorthShore University Health System, Evan-
perhaps because of it), we believe that a systematic ston, Illinois (Grant DK097779)dprincipal investi-
review of these observational data is important to gator: Dr. Brian T. Helfand; study coordinators:
summarize the evidence available (even if lower Jasmine Nero and Pooja Talaty.
quality) and to highlight the lack of evidence in this University of Michigan Health System, Ann
important topic area. We hope that results from this Arbor, Michigan (Grant DK099932)dprincipal
review will spur additional research into lifestyle investigator: Dr. J. Quentin Clemens; co-
changes that may modify and/or prevent LUTS. On investigators: Drs. Mitch Berger, John DeLancey,
the other hand, we also acknowledge that strong Dee Fenner, Rick Harris, Steve Harte, Anne
evidence may never exist for some of these research Pelletier-Cameron and John Wei; study co-
questions. Thus, health care providers should make ordinators: Morgen Barroso, Linda Drnek, Greg
a practical assessment of the evidence available and Mowatt and Julie Tumbarello.
consider the individual situation of each patient in University of Washington, Seattle, Washington
making clinical recommendations. (Grant DK100011)dprincipal investigator: Dr.
Claire Yang; coinvestigator: Dr. John L. Gore; study
coordinators: Alice Liu and Brenda Vicars.
CONCLUSIONS Washington University in St. Louis, St. Louis,
A systematic literature review showed that evidence Missouri (Grant DK100017)dprincipal in-
supporting associations between LUTS and diet, vestigators: Drs. Gerald Andriole and Henry Lai;
fluid intake and caffeine, alcohol and tobacco use is coinvestigator: Dr. Joshua Shimony; study co-
sparse. The data available are largely observational ordinators: Susan Mueller, Heather Wilson and
and generally of lower quality. Given that these Aleksandra Klim.
behavioral factors are often modifiable and are National Institute of Diabetes and Digestive and
frequently included in management recommenda- Kidney Diseases, Division of Kidney, Urology and
tions by LUTS care providers, more and higher Hematology, Bethesda, Marylanddproject scien-
quality evidence is needed to better understand tist: Dr. Ziya Kirkali; project officer: Dr. John
their impact on LUTS. Kusek; National Institutes of Health personnel:
Drs. Tamara Bavendam and Robert Star, and Jenna
ACKNOWLEDGMENTS Norton.
The following individuals were instrumental in the Arbor Research Collaborative for Health, Data
planning and conduct of this study at each of the Coordinating Center, Ann Arbor, Michigan (Grants
participating institutions. DK097776 and DK099879)dprincipal investigator:
Duke University, Durham, North Carolina Dr. Robert Merion; coinvestigators: Drs. Brenda
(Grant DK097780)dprincipal investigators: Drs. Gillespie and Victor Andreev; project manager:
Cindy Amundsen and Kevin Weinfurt; co- Melissa Fava; clinical study process manager: Peg
investigators: Drs. Kathryn Flynn, Matthew O. Hill-Callahan; clinical monitor: Timothy Buck;
Fraser, Todd Harshbarger, Aaron Lentz, Drew research analysts: Margaret Helmuth and Jon
Peterson, Nazema Siddiqui and Alison Weidner; Wiseman; project associate: Julieanne Lock.
study coordinators: Carrie Dombeck, Robin Gilliam,
Akira Hayes and Shantae McLean.
University of Iowa, Iowa City, Iowa (Grant APPENDIX
DK097772)dprincipal investigators: Drs. Karl Levels of evidence used in grading articles
Kreder and Catherine S. Bradley; coinvestigators:
Level Study Design or Type
Drs. Bradley A. Erickson, Susan K. Lutgendorf,
Vince Magnotta, Michael A. O’Donnell and Vivian 1a Systematic review of RCTs
1b Individual RCT (good quality)
Sung; study coordinators: Ahmad Alzubaidi, Andrea 2a Systematic review of cohort studies
Lopez and Linda Moss. 2b Individual cohort study, low quality RCT
Northwestern University, Chicago, Illinois 3a Systematic review of case-control studies
3b Individual case-control study
(Grant DK097779)dprincipal investigator: Dr. 4a Good quality case series, poor quality cohort or case-control study
David Cella; coinvestigators: Drs. James Griffith, 4b Cross-sectional study
Kimberly Kenton, Christina Lewicky-Gaupp, Todd 5 Expert opinion
Parrish and Jennie Yu Fan Chan; study co- Modified from 2009 Oxford Centre for Evidence-Based Medicine Levels of
ordinators: Sarah Buono, Maria Corona, Beatriz Evidence7 and International Consultation on Urological Diseases.8
LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS 1019

REFERENCES
1. Kupelian V, Wei JT, O’Leary MP et al: Prevalence and obese women through modest weight loss. 29. Curto TM, Giovannucci EL, McKinlay JB et al:
of lower urinary tract symptoms and effect on Obstet Gynecol 2010; 116: 284. Associations between supplemental or dietary
quality of life in a racially and ethnically diverse intake of vitamin C and severity of lower urinary
random sample: the Boston Area Community 15. Araki H, Watanabe H, Mishina T et al: High-risk tract symptoms. BJU Int 2015; 115: 134.
Health (BACH) Survey. Arch Intern Med 2006; group for benign prostatic hypertrophy. Prostate
166: 2381. 1983; 4: 253. 30. Dallosso HM, Matthews RJ, McGrother CW
et al: The association of diet and other lifestyle
2. Coyne KS, Sexton CC, Thompson CL et al: The 16. Bravi F, Bosetti C, Dal Maso L et al: Food groups
factors with the onset of overactive bladder: a
prevalence of lower urinary tract symptoms and risk of benign prostatic hyperplasia. Urology
longitudinal study in men. Public Health Nutr
2006; 67: 73.
(LUTS) in the USA, the UK and Sweden: results 2004; 7: 885.
from the Epidemiology of LUTS (EpiLUTS) study. 17. Chyou PH, Nomura AM, Stemmermann GN et al:
BJU Int 2009; 104: 352. 31. Dallosso HM, McGrother CW, Matthews RJ
A prospective study of alcohol, diet, and other
et al: Nutrient composition of the diet and the
3. Coyne KS, Wein AJ, Tubaro A et al: The burden lifestyle factors in relation to obstructive urop-
development of overactive bladder: a longitudi-
athy. Prostate 1993; 22: 253.
of lower urinary tract symptoms: evaluating the nal study in women. Neurourol Urodyn 2004; 23:
effect of LUTS on health-related quality of life, 18. Galeone C, Pelucchi C, Talamini R et al: Onion 204.
anxiety and depression: EpiLUTS. BJU Int, suppl., and garlic intake and the odds of benign pros-
2009; 103: 4. tatic hyperplasia. Urology 2007; 70: 672.
32. McGrother CW, Donaldson MM, Thompson J
et al: Etiology of overactive bladder: a diet and
4. Gormley EA, Lightner DJ, Burgio KL et al: Diag- 19. Kristal AR, Arnold KB, Schenk JM et al: Dietary lifestyle model for diabetes and obesity in older
nosis and treatment of overactive bladder (non- patterns, supplement use, and the risk of women. Neurourol Urodyn 2012; 31: 487.
neurogenic) in adults: AUA/SUFU guideline. symptomatic benign prostatic hyperplasia: re-
J Urol 2012; 188: 2455. sults from the Prostate Cancer Prevention Trial.
33. Spigt M, van Schayck O, Knipschild P et al: Is it
possible to improve elderly male bladder func-
5. Yang CC, Weinfurt KP, Merion RM et al: Symp- Am J Epidemiol 2008; 167: 925.
tion by having them drink more water? A ran-
toms of Lower Urinary Tract Dysfunction 20. Suzuki S, Platz EA, Kawachi I et al: Intakes of domized trial of effects of increased fluid intake/
Research Network. J Urol 2016; 196: 146. energy and macronutrients and the risk of benign urine output on male lower urinary tract func-
6. Moher D, Liberati A, Tetzlaff J et al: Preferred prostatic hyperplasia. Am J Clin Nutr 2002; 75: tion. Urology 2006; 68: 1031.
reporting items for systematic reviews and 689.
34. Spigt MG, Knottnerus JA, van de Beek C et al:
meta-analyses: the PRISMA statement. PLoS 21. Lagiou P, Wuu J, Trichopoulou A et al: Diet and Short-term effects of increased urine output on
Med 2009; 6: e1000097. male bladder function and lower urinary tract
benign prostatic hyperplasia: a study in Greece.
7. Oxford Centre for Evidence-Based MedicinedLevels Urology 1999; 54: 284. symptoms. Urology 2004; 64: 499.
of Evidence (March 2009). Available at http://www.
22. Tavani A, Longoni E, Bosetti C et al: Intake of 35. Zimmern P, Litman HJ, Mueller E et al: Effect of
cebm.net/oxford-centre-evidence-based-medicine-
selected micronutrients and the risk of surgically fluid management on fluid intake and urge in-
levels-evidence-march-2009/. Accessed February
treated benign prostatic hyperplasia: a case- continence in a trial for overactive bladder in
15, 2017.
control study from Italy. Eur Urol 2006; 50: 549. women. BJU Int 2010; 105: 1680.
8. Abrams P and Khoury S: International Consul-
23. Dallosso H, Matthews R, McGrother C et al: Diet 36. Johnson TM II, Sattin RW, Parmelee P et al:
tation on Urological Diseases: evidence-based
as a risk factor for the development of stress Evaluating potentially modifiable risk factors for
medicine overview of the main steps for devel-
urinary incontinence: a longitudinal study in prevalent and incident nocturia in older adults.
oping and grading guideline recommendations.
women. Eur J Clin Nutr 2004; 58: 920. J Am Geriatr Soc 2005; 53: 1011.
Neurourol Urodyn 2010; 29: 116.
24. Dallosso HM, McGrother CW, Matthews RJ 37. Soda T, Masui K, Okuno H et al: Efficacy of
9. Brown JS, Wing R, Barrett-Connor E et al:
et al: The association of diet and other lifestyle nondrug lifestyle measures for the treatment of
Lifestyle intervention is associated with lower
prevalence of urinary incontinence: the Diabetes
factors with overactive bladder and stress in- nocturia. J Urol 2010; 184: 1000.
continence: a longitudinal study in women. BJU
Prevention Program. Diabetes Care 2006; 29: 38. Tani M, Hirayama A, Torimoto K et al: Guidance
Int 2003; 92: 69.
385. on water intake effectively improves urinary
25. Erickson BA, Vaughan-Sarrazin M, Liu X et al: frequency in patients with nocturia. Int J Urol
10. Manonai J, Songchitsomboon S, Chanda K et al:
The effect of a soy-rich diet on urogenital atro-
Lower urinary tract symptoms and diet quality: 2014; 21: 595.
findings from the 2000-2001 National Health and
phy: a randomized, cross-over trial. Maturitas
Nutrition Examination Survey. Urology 2012; 79: 39. Swithinbank L, Hashim H and Abrams P: The
2006; 54: 135. effect of fluid intake on urinary symptoms in
1262.
11. Schnelle JF, Leung FW, Rao SS et al: A controlled women. J Urol 2005; 174: 187.
26. Maserejian NN, Giovannucci EL and McKinlay
trial of an intervention to improve urinary and
JB: Dietary macronutrients, cholesterol, and so- 40. Hashim H and Abrams P: How should patients
fecal incontinence and constipation. J Am
dium and lower urinary tract symptoms in men. with an overactive bladder manipulate their fluid
Geriatr Soc 2010; 58: 1504.
Eur Urol 2009; 55: 1179. intake? BJU Int 2008; 102: 62.
12. Subak LL, Wing R, West DS et al: Weight loss to
27. Maserejian NN, McVary KT, Giovannucci EL et al: 41. Dowd TT, Campbell JM and Jones JA: Fluid
treat urinary incontinence in overweight and
Dietary macronutrient intake and lower urinary intake and urinary incontinence in older
obese women. N Engl J Med 2009; 360: 481.
tract symptoms in women. Ann Epidemiol 2011; community-dwelling women. J Community
13. Zhang W, Wang X, Liu Y et al: Effects of dietary 21: 421. Health Nurs 1996; 13: 179.
flaxseed lignan extract on symptoms of benign
28. Wong SY, Lau WW, Leung PC et al: The asso- 42. Townsend MK, Jura YH, Curhan GC et al: Fluid
prostatic hyperplasia. J Med Food 2008; 11: 207.
ciation between isoflavone and lower urinary intake and risk of stress, urgency, and mixed
14. Wing RR, Creasman JM, West DS et al: tract symptoms in elderly men. Br J Nutr 2007; urinary incontinence. Am J Obstet Gynecol 2011;
Improving urinary incontinence in overweight 98: 1237. 205: 73.e1.
1020 LIFESTYLE FACTORS AND LOWER URINARY TRACT SYMPTOMS

43. Creighton SM and Stanton SL: Caffeine: does it 46. Morrison AS: Risk factors for surgery for prostatic 49. Tettamanti G, Nyman-Iliadou A, Pedersen NL
affect your bladder? Br J Urol 1990; 66: 613. hypertrophy. Am J Epidemiol 1992; 135: 974. et al: Influence of smoking, coffee, and tea
consumption on bladder pain syndrome in female
44. Arya LA, Myers DL and Jackson ND: Dietary 47. Gass R: Benign prostatic hyperplasia: the oppo- twins. Urology 2011; 77: 1313.
caffeine intake and the risk for detrusor insta- site effects of alcohol and coffee intake. BJU Int
bility: a case-control study. Obstet Gynecol 2000; 2002; 90: 649.
96: 85. 50. Asplund R and Aberg HE: Nocturia in relation
48. Shiri R, Hakama M, Hakkinen J et al: The effects to body mass index, smoking and some other
45. Morrison AS: Prostatic hypertrophy in greater of lifestyle factors on the incidence of nocturia. J life-style factors in women. Climacteric 2004;
Boston. J Chronic Dis 1978; 31: 357. Urol 2008; 180: 2059. 7: 267.

Das könnte Ihnen auch gefallen