Access to this full-text is provided by Wiley.
Content available from Alimentary Pharmacology & Therapeutics
This content is subject to copyright. Terms and conditions apply.
Aliment Pharmacol Ther. 2024;59:157–174.
|
157wileyonlinelibrary.com/journal/apt
Received: 9 May 2023
|
First decision: 15 June 2023
|
Accepted: 13 Oc tober 2 023
DOI: 10.1111/apt.17782
Systematic review and meta-analysis: Foods, drinks and diets
and their effect on chronic constipation in adults
Alice Van Der Schoot | Zoi Katsirma | Kevin Whelan | Eirini Dimidi
This is an op en access arti cle under the ter ms of the Creative Commons Attribution-NonCommercial-NoDerivs License, whi ch permits use a nd distribution in
any medium, provided the original work is properly cited, the use is no n-commercial and no modi ficat ions or adaptat ions are made.
© 2023 The Authors . Alimentary Pharmacology & Therapeutics publishe d by John Wiley & Sons Ltd.
Alice Van D er Schoot and Zoi K atsirma shou ld be considere d as joint f irst author.
As par t of AP&T's peer-review p rocess , a technical ch eck of thi s meta-a nalysis was per formed by Dr Y Yuan . The Handling Ed itor for t his article wa s Profes sor Coli n Howden, and it wa s
accepte d for publicat ion aft er full p eer-revie w.
Depar tment of Nutrit ional Sciences, King's
College L ondon , London, UK
Correspondence
Eirini Dimidi, Department of Nutritional
Science s, King's College London, 150
Stamfor d Street , London SE1 9NH, UK .
Email: eirini.dimidi@kcl.ac.uk
Funding information
Gener al and Edu cation Trust Fund; Br itish
Dietetic Association
Summary
Background: Dietary approaches are recommended for the management of chronic
constipation. Until now, there has been no systematic review and meta-analysis on
foods, drinks and diets in constipation.
Aims: To investigate the effect of foods, drinks and diets on response to treatment,
stool output, gut transit time, symptoms, quality of life, adverse events and compli-
ance in adults with chronic constipation via a systematic review and meta-analysis.
Methods: Studies were identified using electronic databases (12th July 2023).
Intervention trials (randomised controlled trials [RCTs], non-randomised, uncon-
trolled) were included. Risk of bias was assessed using Cochrane 2.0 (RCTs) or JBI
Critical Appraisal (uncontrolled trials). Data from RCTs only were synthesised using
risk ratios (RRs), mean differences (MDs), standardised mean differences (95% CI)
using random-effects.
Results: We included 23 studies (17 RCTs, 6 uncontrolled; 1714 participants): kiwi-
fruit (n = 7), high-mineral water (n = 4), prunes (n = 2), rye bread (n = 2), mango, fig, ce-
real, oat bran, yoghurt, water supplementation, prune juice, high-fibre diet, no-fibre
diet (n = 1). Fruits resulted in higher stool frequency than psyllium (MD: +0.36 bowel
movements [BM]/week, [0.25–0.48], n = 232), kiwifruits in particular (MD: +0.36
BM/week, [0.24–0.48], n = 192); there was no difference for prunes compared with
psyllium. Rye bread resulted in higher stool frequency than white bread (MD: +0.43
BM/week, [0.03–0.83], n = 48). High-mineral water resulted in higher response to
treatment than low-mineral water (RR: 1.47, [1.20–1.81], n = 539).
Conclusions: Fruits and rye bread may improve certain constipation-related out-
comes. There is a scarcity of evidence on foods, drinks and diets in constipation and
further RCTs are needed.
158
|
VAN DER SCHOOT ET AL.
1 | INTRODUCTION
Chronic constipation is a prevalent functional bowel disorder
that negatively affects quality of life.1–3 Furthermore, it poses
a considerable financial burden to individuals and healthcare
systems.4,5
Diet plays an impor tant role in the management of chronic con-
stipation. Current guidelines recommend increasing dietary fibre
(e.g. fibre supplementation); however, they include no consistent
recommendations for specific foods, drinks or diets.6,7 Other guid-
ance encourages adequate fluid intake, the consumption of vegeta-
bles, whole grains, and fruits and their juices that have a high sorbitol
content (e.g. apples).8 Fibre may increase stool bulk or soften stools
due to a high-water holding capacity, or result in the production of
fermentation-derived by-products, regulating gut muscular contrac-
tions and reducing gut transit time (GTT ).9 Sorbitol, a nonabsorbable
sugar alcohol found in fruits, may exhibit an osmotic ef fect, increas-
in g lu m in al wa ter conten t and soft eni ng st ool s.10, 11 However, there is
a scarcity of evidence to support the use of specific foods and drinks
in chronic constipation to inform self-management and clinical care,
which may contribute to the high dissatisfaction rates reported by
people with constipation regarding the effectiveness of treatment
options.12
While systematic reviews have been conducted on the effec-
tiveness of certain dietar y interventions in chronic constipation,
these focus on interventions in supplement forms rather than their
natural forms. For example, systematic reviews show improvements
in cardinal symptoms of constipation from fibre supplements,13 but
none have been conducted in relation to fibre-containing foods.
Others focused on probiotics14 and fo od sup plements (e.g. fr uit ex-
tracts),15 but to date, there have been none on the effect of whole
foods, drinks and diets in chronic constipation.
The aim of this systematic review and meta-analysis was to in-
vestigate the effect of foods, drinks, and diets on response to treat-
ment, stool output, GTT, symptoms, quality of life, adverse events
and compliance in adults with chronic constipation.
2 | MATERIALS AND METHODS
This systematic review and meta-analysis was undertaken following
the Cochrane Handbook for Systematic Reviews of Interventions16
and Preferred Reporting Items for Systematic reviews and Met a-
Analyses updated guidelines.17 A protocol was published prior its
conduct (PROSPERO CRD42021241072).
2.1 | Eligibility criteria
Eligibilit y criteria were developed using PICOS (Patient,
Intervention, Comparators, Outcome, Study design) (Table 1).
Briefly, the inclusion criteria were interventio n trials reporting t he
effect of foods (including herbs and spices), drinks, and diets in
adults with chronic constipation that measured constipation out-
comes. Uncontrolled trials were eligible due to limited RCTs, and
the unique challenges of designing and blinding controls in food
and diet interventions.18
2.2 | Search strategy
Studies were identified through a systematic search of electronic
databases, hand-searching conference abstracts and back-searching
reference lists of eligible studies and relevant review articles.
Three electronic databases were searched: MEDLINE (1946
to July 2023; OvidSP), EMBASE (1974 to July 2023; OvidSP),
The Cochrane Central Register of Controlled Trials (all years; The
Cochrane Library). Combinations of terms related to foods, herbs,
spices, drinks, and diets were searched as medical subject head-
ings and free text terms (Appendix S1). No restrictions were ap-
plied to language or publication date. The US National Institute
of Health clinical trials register (w w w . c l i n i c a l t r i a l s . g o v / ) was
searched to identify unpublished trials. The final search date was
12th July 2023.
Abstracts from annual conferences were searched: Digestive
Disease Week (2011 to 2021, Gastroenterology), British Dietetic
Association (2011 to 2021, J Hum Nutr Diet), British Society of
Gastroenterology (2011 to 2021, Gut), and the European Society
for Clinical Nutrition and Metabolism (2004 to 2021, Clin Nutr; Clin
Nutr Supp, e-SPEN).
2.3 | Selection process
References were imported into a reference manager to assess eli-
gibility (EndNote X9; Thomson Reuters). Two reviewers (AvdS, ZK)
independently screened titles and abstracts, then full-text articles
against the pre-defined eligibility criteria. Foreign-language articles
were translated. Discrepancies in the selection of eligible studies
were resolved by a third and fourth reviewer (ED, KW)
2.4 | Data collection process
Two reviewers (AvdS, ZK) independently extracted data from eligi-
ble studies onto a standardised form (Table 1). Extracted data were
compared between reviewers and discrepancies were resolved.
Where an article provided insufficient data, authors were contac ted
to provide information. Data were extracted as intention-to-treat
analyses when possible. For dichotomous data, drop-outs were as-
sumed to be treatment failures. If this information was not clear,
analysis was undertaken on all participant s with reported evalu-
able data. When required, data were estimated from figures using
PlotDigitizer 2.6.9.19
|
159
VAN DER SCHOOT E T AL.
2.5 | Risk of bias assessment
Risk of bias was assessed independently by two reviewers (AvdS,
ZK). For RCTs, the Cochrane risk-of-bias 2.0 tool was used, assessing
bias arising from randomisation, deviations from intended interven-
tions, missing outcome data, outcome measurement, and selective
reporting.20 The highest level of bias across individual domains de-
termined the overall bias for each outcome of the study (‘low risk’,
‘some concerns’ or ‘high risk’). For single-arm trials, the JBI Critical
Appraisal checklist was used.21 Disagreements were resolved
through discussion with a third reviewer (ED). Available trial regis-
trations and protocols were obtained to ensure that pre-specified
outcomes matched the final publication.
2.6 | Data synthesis
Bot h quant it ative (meta-analysis) and narrati ve synthe sis was und er-
taken, based on the study design. Given the heterogeneity of the
interventions, comparators, and designs of the included studies, a
TABLE 1 Table of inclusion and exclusion criteria for participants, intervention, comparator, outcomes and study designs.
Characteristic Inclusion and exclusion criteria Data extracted
Participants Adult s (≥18 years old) of any sex or ethnicity, with chronic idiopat hic constipation,
identif ied by: (1) clinical diagnostic criteria (e.g. Rome); (2) author or clinician-
defined criteria/diagnosis; (3) participant-def ined criteria (e.g. self-report); or
(4) presence of at least one of the following constipation symptoms: <3 bowel
movement s/week, hard/lumpy stools, sensation of incomplete evacuation,
straining, manual manoeuvres, physiological markers (e.g. slow gut transit time)
or an evacuation disorder. Community or outpatient set tings were eligible.
Studies were excluded if all participants had secondary constipation or belonged
to specif ic clinical population groups (e.g. all pregnant women, inpatients).
However, a study was eligible if only a subset (e.g. 30% of the participants)
belonged to one of these specific population groups.
Age, sex , location, type of constipation,
diagnosis method, inclusion and
exclusion criteria, number of
participant s in groups, number
of participant s with se condar y
constipation or from a specific
population group (if present).
Intervention Studies were eligible if they administered individual foods, drinks , herbs, spices, or
diets (defined as a change in habitual dietar y pattern), that are widely available
and applicable in practice.
Studies were excluded if the intervention was a composite dish, due to difficulty
in replic ating multiple ingredients in a specific dosage in a clinical setting (e.g.
a specific vegetable is eligible, but not a mixed vegetable soup), a combination
of produc ts that are not widely available in that form (e.g. prunes are eligible,
but not a combination of yoghur t with pr unes and linseeds), or a diet based on
specially formulated composite foods (e.g. a Mediterranean diet is eligible, but
not a diet based on a specially formulated soup).
Intervention t ype, ingredients, form,
dose, schedule, duration.
Comparators Trials with or without a control arm were eligible, as follows: (i) no control arm;
(ii) control arm wit h no inter vention (e.g. no intervention or habitual diet); (iii)
placebo control (e.g. placebo capsule, inert food, sham diet); (iv) comparator
arm with a potentially active inter vention (e.g. psyllium supplement,
acupuncture, laxatives).
Type, form, dose, sche dule, duration.
Outcomes Studies reporting dichotomous or continuous data on response to treatment
(e.g. proportion of participants no longer constipated according to Rome III
criteria), stool frequency (e.g. daily bowel movement diary), stool consistenc y
(e.g. Bris tol Stool Form Scale), physiological outcomes (e.g. stool weight using
weighing scales or visual guides, gut transit time using radio-opaque markers),
frequency or severity of individual symptoms (e.g. straining, bloating using a
symptom diary, visual analogues sc ales, etc.).
Integrative symptom scores (e.g. Patient Assessment of Constipation Symptoms ,
Gastrointestinal Symptom Rating Scale questionnaires), symptom response
(e.g. proportion of participants with ≥3 complete spontaneous bowel
movement s/week), qualit y of life (e.g. Patient A ssessment of Constipation
Qualit y of Life questionnaire), laxative use (e.g. daily diary), adverse events,
compliance (e.g. counting number of unused product).
Outcomes, method of measurement ,
baseline, midpoint and endpoint
values or change from baseline, det ails
of adverse events, compliance.
Study design Intervention trials, including controlled and comparative trials (randomised or non-
randomised) or uncontrolled trials (e.g. single-arm studies). Parallel group and
cross-over trials were eligible.
Study design, washout period duration,
intention to treat analysis, number
of excluded participant s, reasons for
exclusion, randomisation method,
allocation concealment, blinding,
funding source, funder involvement,
conflicts of interest.
160
|
VAN DER SCHOOT ET AL.
hierarchy was developed to ensure that any quantitative synthesis
was meaningful. Studies were only considered for met a-analysis if
they were RCTs, with a duration of intervention ≥2 weeks, and if a
cross-over design, a washout period of ≥2 weeks. Cross-over stud-
ies without an adequate washout period were only considered for
meta-analysis if the data from the first period could be ex trac ted.
Data from studies not included in the meta-analysis (uncontrolled or
single-arm studies, when a single study evaluated the inter vention
type, duration of intervention <2 weeks, or cross-over study with in-
adequate washout period for which data from the first period could
not be extracted) were described in a narrative synthesis.
For RCTs considered for meta-analysis, this was per formed
where data for the same outcome were available from two or more
studies of the same intervention type, using Review Version 5.4 (The
Cochrane Collaboration, 2020). Foods, drinks and diets were me-
ta-analys ed sepa rately. If two or more elig ible inter ventions were in-
vestigated in one study (e.g. different doses), these were treated as
separate entries in the meta-analysis; each arm was compared with
the control separately, and the sample size of the control group was
divided by the number of intervention arms to reduce unit-of-anal-
ysis error.16
Dichotomous outcomes were expressed as risk ratio (RR) and
95% confidence intervals (CIs). Mean difference (MD) was calcu-
lated for continuous outcomes that were measured using the same
assessment tool and reported in the same units. Standardised mean
difference (SMD) was calculated for continuous outcomes that were
measured or repor ted dif ferently. Means, standard deviations (SDs),
sample size and p-values were used for analysis. If medians and
ranges were reported, means and SDs were estimated using meth-
ods from Wan et al.22
Meta-analyses were performed using a random-effects model.
Statistical heterogeneity was assessed using the Chi-squared test
and quantified using the I2 statistic. Thresholds of 50% and 75%
were considered to represent substantial and considerable hetero-
geneity, respectively.16 Subgroup analys es were per formed to inves-
tigate heterogeneity and explore the effects of intervention types
(e.g. fruits), where appropriate. For subgroup analyses, p < 0.1 was
considered statistically significant.23
3 | RESULTS
A tot al of 10,905 non-duplicated records were identified, of which
103 were potentially eligible (Figure 1). Of these, 80 records were
excluded (Figure 1, Table S2). In total, 23 studies fulfilled the in-
clusion criteria, involving 1714 participants with chronic constipa-
tion.24–4 6 Table 2 displays the characteristics of included studies.
There were 17 RCTs (13 parallel-group,25,30–33,36–41,43,46 four cross-
over24,29,44,45), and six uncontrolled trials.26–28,34,35,42 Of the uncon-
trolled trials, five were single-arm, and one trial (constipation vs.
healthy group) was treated as single-arm as data were extracted for
constipation only.26 Studies administered kiwifruit (n = 7),25–29,44,45
prunes (n = 2),24,2 5 mango (n = 1),30 fig paste (n = 1),31 rye bread
(n = 2),32,33 cereal (n = 1),34 oat bran biscuits (n = 1),35 pasteurised
yoghurt (n = 1),36 high-mineral water (n = 4),3 7–4 0 water supplemen-
tation (n = 1),41 prune juice (n = 1),46 a high-fibre diet (n = 1),43 and
a ‘no-fibre’ diet (n = 1).42 There were no eligible studies of herbs
or spices in chronic constipation. Twenty-one publications were
English, one Spanish43 and one Dutch.27 There were 22 full text
articles and one commercial report.27 Twenty-one authors were
contacted to provide additional information24,25,27–41,43–46; nine re-
plied,25–27,38,39,43–46 and five provided additional data.
39, 43 –4 6 The
outcomes of the meta-analyses are reported in Table 3.
4 | FOODS
4.1 | Fruits
Ten studies administered fruits.24–31,44,45 Six studies investigating
fruits (kiwifruit, prunes, or mango) compared with psyllium (compar-
ator) were combined for meta-analysis.24,25,29,30,44,45 One three-arm
study was par tially randomised; only the randomised groups (prunes
vs. psyllium) were included in the meta-analysis, whereas data from
the non-randomised group (kiwifruit) was narratively described.25
Three uncontrolled studies investigated kiwifruit,26–2 8 and one RCT
investigated fig compared with placebo.31
4.1.1 | Fruits (kiwifruit, prunes and mango) compared
with psyllium: RCTs
Response to treatment
Response to treatment was defined as proportion of participants no
longer constipated (Rome III criteria), or with an increase ≥1 or ≥1.5
complete spontaneous bowel movement (CSBM)/week: 54/118
(46%) resp ond e d to frui ts an d 43/11 5 (37% ) to psyl li um (R R: 1.1 8; 95%
CI: 0.89–1.58, p = 0.25; I2 = 0%, p = 0.64) (Figure 2A, Figure S1).2 5, 29, 4 4
In the partially randomised study, 13/29 responded to kiwifruit
(non-randomised group; not included in meta-analysis) and 14/22 to
psyllium (p = 0.19; increase ≥1 CSBM/week).25 Another study was
not included in the meta-analysis due to inadequate washout pe-
riod and data for this outcome could not be extracted for the first
period; 28/40 responded to prunes and 22/40 to psyllium (p = 0.10;
improvement in global symptoms).24
Stool frequency
Stool frequency (bowel movements [BM]/week) was reported in
four studies (n = 232).24,29,44,45 Fruits resulted in higher stool fre-
quency than psyllium (MD: +0.36 BM/week, 95% CI: 0.25–0.48,
p < 0.00001; I2 = 0%, p = 0.78). In the subgroup analysis, kiwifruit
resulted in higher stool frequency (MD: + 0.36 BM/week, 95% CI:
0.24–0.48, p < 0.00001; I2 = 0% , p = 0.73), but there was no differ-
ence in the effect of prunes compared to psyllium (MD: +0.75 BM/
week, 95% CI: −0.40 to 1.90, p = 0.20). No subgroup differences
were detected (p = 0.51) (Figure 2B, Figure S1).
|
161
VAN DER SCHOOT E T AL.
The partially randomised study not included in the meta-analysis
reported stool frequency as mean CSBM/week (kiwifruit 2.2, prunes
2.1, psyllium 2.9; no between-groups analysis).25
Stool consistency
Stool consistenc y (Bristol Stool Form Scale [BSFS]) was reported
in six studies (n = 315).24,25,29,30,44,45 Fruits resulted in softer stools
than psyllium (MD: +0.48 points, 95% CI: 0.11–0.86, p = 0.01) and
substantial heterogeneity was detected (I2 = 51%, p = 0.07).24,25,29,30
In the subgroup analysis, there was no difference in the effect of
kiwifruit (MD: +0.32 points, 95% CI: −0.12 to 0.76, p = 0.16; I2 = 55%,
p = 0.11), or prunes compared with psyllium (MD: +0.45 points,
95% CI: −0.24 to 1.14, p = 0.20; I2 = 0%, p = 0.80); however, mango
resulted in softer stools compared to psyllium (MD: +1.41 points,
95% CI: 0.58–2.24, p = 0.00 08), with a significant subgroup effect
(p = 0.07) (Figure 2C, Figure S1).
In the partially randomised study, stool consistency was not dif-
ferent between kiwifruit and psyllium (mean 3.6 vs. 3.1 BSFS point s;
p = 0.19).25
Gut transit time
One study comparing kiwifruit to psyllium measured GTT using
smart pill or radio-opaque markers, and there was no difference be-
tween groups (p > 0.05).44
Gastrointestinal symptoms
Integrative symptoms were reported in two studies using the
Gastrointestinal Symptom Rating Scale (GSRS) total score44 or the
Constipation Scoring System (CSS) (n = 130); there was no differ-
ence in the effect of fruits compared with psyllium (SMD: −1.87,
95% CI: −4.88 to 1.14, p = 0.22) and considerable heterogeneity was
detected (I2 = 98%, p < 0.00001). In the subgroup analysis, kiwifruit
resulted in lower integrative symptom scores than psyllium (SMD:
−3.40, 95% CI: −4.03 to −2.77, p < 0.000 01), but there was no differ-
ence in the effect of mango compared with psyllium (SMD: −0.33,
95% CI: −1.00 to 0.34, p = 0.33), with a significant subgroup effect
(p < 0.00 001) (Figure S2).30,44
Severity of straining was reported in two studies of prunes com-
pared with psyllium (n = 86)24,2 5; there was no difference in effect
FIGURE 1 PRISMA flow diagram of studies included in the systematic review.
Records identified through
database searching
(n = 12,980)
Screening
Included EligibilityIdentification
Additional records identified
through other sources
(n = 49)
Records after duplicates removed
(n = 10,905)
Records screened
(n = 10,905)
Excluded (n = 80)
Reasons (not mutually exclusive):
Not chronic constipation: 27
Not otherwise healthy: 6
Not community/outpatient: 5
Not an individual
food/drink/herb/spice/diet: 13
Intervention in supplement form: 10
Intervention is probiotic product: 14
Not widely available or applicable in
practice: 10
Not relevant outcomes: 8
Conference abstract of published
paper: 6
Exploratory protocol: 1
Incomplete study and unpublished
data not shared: 2
Studies included in
qualitative synthesis
(n = 23)
Full-text articles assessed for
eligibility
(n = 103)
Records excluded
(n = 10,802)
Studies included in
quantitative synthesis
(n = 12)
Excluded from meta-analysis (n = 11)
Uncontrolled study: 6
Single study administered
intervention: 5
162
|
VAN DER SCHOOT ET AL.
TABLE 2 Characteristics of studies investigating the effect of foods, drinks and diet chronic constipation in adults.
Study, year
(references) Study design Sampl e size (% female)
Age (yea rs), mean ± SD
(range)aConstipation diagnosis Intervention (dose) Duration Comparator (dose) Outcomes
Attal uri et al, 201124 Cros s-over RCT,
single-blind
Total: 40 (93% )
Prunes-psyllium: 20
Psyllium-prunes: 20
38, NR (N R) Rome III c riteria Prune s (100 g/day) 3 weeks Fibre su pplement:
psylli um (22 g/d)
with water
Meta-a nalysis: stool f requency, sto ol
consis tency, sev erity of stra ining;
Narra tive sy nthesi s: response to
treatment, bloating.
Chey et al , 202125 Parallel R CT
(part ially
randomisedb),
no blinding
Total: 79 (87%)
Kiwif ruit: 3 0
Prune s: 26
Psyllium: 23
Kiwif ruit: 4 3.1 ± 14.9
Prune s: 41.9 ± 17
Psyll ium: 43. 0 ± 17.2, (18–76)
Rome IV cr iteria for chro nic
consti pation or IBS- C
Prune s (100 g/d)
Kiwif ruit, g reen (2
fruits/day)c
4 weeks Fibre su pplement:
psylli um (12 g/
day) diss olved in
water
Meta-a nalysis (prun es vs. psyllium
groups o nly): respons e to
treatm ent, stool con sistency,
severi ty of str aining; Narr ative
synth esis: respons e to treat ment,
stool f reque ncy, stoo l consistency,
straining, abdominal pain.
Chan et al , 200726 Case cont rol study;
no contr ol
interventiond
Total: 33 (73% ) 49.9 ± 12, ( NR) Rome III c riteria and Chi nese
constipation questionnaire
Kiwif ruit, g reen (2
fruits/day)c
4 weeks –N arrat ive synt hesis: respon se to
treatm ent, stool fre quency, stool
consistency, bloating.
Hiele, 2 01027 Uncontrolled trial Tota l: 38 (NR) NR, (20 –70) Rome III crite ria Kiwi fruit (3 fruit s/day) 3 wee ks –Narra tive sy nthesi s: stool freque ncy,
stool co nsistency, ease o f
defecation.
Cunill era et al, 201528 Uncontrolled trial Tota l: 46 (91%) 49.5 (m edian), NR, (NR ) Rome III c riteria Kiwif ruit, g reen (3
fruits/day)c
3 weeks –N arrat ive synt hesis: respon se to
treatm ent, stool fre quency, stool
consistency.
Gearr y et al, 202344 Cr oss-over R CT,
single-blind
Total: 60e ( 72%)
Kiwifruit-psyllium: NR
Psyllium-kiwifruit: NR
38.1 ± 15 .6, (18–65) Rome III c riteria Kiwif ruit, g reen (2
fruits/day)c
4 weeks Fibre su pplement:
psylli um (7.5 g/
day)
Meta-a nalysis: respo nse to treatme nt,
stool f reque ncy, stoo l consistency,
integr ative symptoms , frequency of
strai ning; Narrat ive synt hesis: GTT
Bayer et al , 202245 Cross-over RC T,
single-blind
Tot al: 17 f (NR)
Kiwifruit-psyllium: NR
Psyllium-kiwifruit: NR
40.2 ± 4 .1, (NR) Rome IV cr iteria Kiwifruit , gold (2
fruits/day)c
4 weeks Fibre su pplement:
psylli um (7.5 g/
day)
Meta-a nalysis: stool f requency, sto ol
consis tency, fr equen cy of strainin g
Eady et a l, 201929 Cross-over RC T,
single-blind
Total: 35 (100% )
Kiwifruit-psyllium: NR
Psyllium-kiwifruit: NR
51 (median), NR , (21–65) Rome III c riteria for chro nic
consti pation or IBS- C
Kiwif ruit, g old
(3 fruit s/day)c
4 weeks Fibre su pplement:
psylli um (14.75 g/
day)
Meta-a nalysis: stool f requency, sto ol
consis tency, fr equen cy of strainin g;
Narra tive sy nthesi s: response to
treatment.
Venancio e t al, 201830 Parall el RCT, no
blinding
Total: 48 (52%)
Mango: 24
Psyllium: 24
Mango: 23 .5 ± 4.4, P sylliu m:
28.9 ± 8.9, (NR )
Rome III c riteria Mango (3 00 g/day) 4 weeks Fibre sup plement:
psylli um (5.8 g/
day)
Meta-a nalysis: stool c onsistency,
integrative symptoms
Baek et al , 201631 Paralle l RCT,
double-blind
Total: 80 (89 %)
Fig pas te: 40
Control: 40
Fig pas te: 24.0 ± 4.1
Contro l: 24.7 ± 4.5, (NR)
Rome III c riteria Fig pas te (300 g /day) 8 week s Placebo pa ste of
water, suga r and
modif ied starch
(300 g /day)
Narra tive sy nthesi s: stool freque ncy,
stool co nsistency, stoo l quantity,
GTT, abd ominal pain, ab dominal
discomf ort, effor t for evacuatio n,
incomplete evacuation, defecation
time, s atisf actor y relief.
Hongis to et al, 200632 P arallel RCT, no
blinding
Total: 29 (100% )g
Rye bread: 15
Control: 14
41, NR, (18–57) Self-reporte d constipatio n:
reduced/less-frequent bowel
moveme nts, st raining at
defecation
Rye brea d (300 g/day) 3 weeks White br ead (144 g/
day)
Meta-a nalysis: stool f requency, GT T,
integr ative symptoms ; Narrative
synth esis: stool cons istency,
difficulty in defecation
|
163
VAN DER SCHOOT E T AL.
(Continued)
Study, year
(references) Study design Sampl e size (% female)
Age (yea rs), mean ± SD
(range)aConstipation diagnosis Intervention (dose) Duration Comparator (dose) Outcomes
Holma et a l, 201033 Paralle l RCT, no
blinding
Total: 20 (95%)h
Rye bread: 10
Control: 10
Rye brea d: 51, NR (31–78)
Contro l: 43, NR, (24–55)
Self-rep orted consti pation:
part icipan t's own re port of
consti pation and <5 bowel
movements/week without
laxati ves or ≤7 bowel
moveme nts/wee k with laxative s
Rye brea d (240 g/day) 3 weeks White br ead (192 g/
day)
Meta-a nalysis: stool f requency, gut
transi t time, integra tive symptoms;
Narra tive sy nthesi s: stool
consis tency, st ool weig ht, dif ficulty
in defec ation.
Sharia ti et al, 200834 Un controlled trial Total: 41 (100 %) 6 0, NR, (33–77) Pelvic f loor disorder a nd meeting
Rome II cr iteria
High-fi bre cere al
(7 g/day in we ek
1, follow ed by
gradu al increase
to 28 g/day)
6 weeks –N arrat ive synt hesis: stool fr equency,
integr ative symptoms , straining ,
hard/lu mpy stools, in comple te
evacuation, anorectal obstruction,
use of man ual mano euvres, lax ative
use.
Valle-Jones, 198535 Uncontrolled trial Total: 50 (64% ) 70.3 ± 7.0, (NR ) Diagno sis from genera l practition er
based on i nfrequent bowe l
evacua tion, u nusual ly hard
stools
or pain on p assing stools
Oat bran b iscuits
(808 g /day, 2
biscuits)
12 weeks –N arrative syn thesis : stool freque ncy,
stool co nsistency, pain o n
defecation.
Liu et al, 2 01536 Parallel RC T,
double-blind
Total: 120 (50 %)
Pasteu rised yoghur t: 60
Contro l: 60
NR, (30 –60) Rome III c riteria Pasteu rised yoghur t
(220 mL /day)
7 weeks Pasteu rised milk
(220 mL /day)
Narra tive sy nthesi s: stool freque ncy,
strai ning, lumpy/hard s tools,
incomplete evacuation, anorectal
obstruction.
Koyama et a l, 202246 Pa rallel RCT,
double-blind
Total: 84 (75%)
Prune ju ice: 42
Control: 42
Prune ju ice: 50.8 ± 11.1
Contro l: 50.5 ± 10.5, (NR )
Rome IV cr iteria Prune juice,
concentrated
(54 g/day)
8 weeks Placeb o juice of
sugar s, malic a cid,
flavour ing and
colour ing (54 g/
day)
Narra tive sy nthesi s: stool freque ncy,
stool co nsistency, hard s tools,
flatulence, incomplete evacuation,
urgenc y, diarr hoea, l oose stools.
Dupont e t al, 201437 Parall el RCT,
double-blind
Total: 244 (100 %)
High-mi neral water low
dose: 85
High dos e: 82
Control: 77
Low dose : 44.4 ± 10 .5
High dos e: 41.7 ± 10.3
Contro l: 40.8 ± 12.4, (N R)
Rome III c riteria High-mi neral water,
high dos e (1 L/
day high- miner al
water an d 0.5 L/d
low-mineral
water)
High-mi neral water,
low dose ( 0.5 L/
day high- miner al
water +1 L/day
low-mineral
water)
4 weeks Low-min eral water
(1.5 L/day)
Meta-a nalysis: respo nse to treatme nt,
stool f reque ncy, abdo minal pain,
laxati ve use, QoL ; Narrat ive
synthesis: stool consistency
Nauman n et al, 201638 Paral lel RCT,
double-blind
Total: 100 (85% )
High-mi neral water: 5 0
Control: 50
High-mi neral water:
47. 2 ± 10. 5
Contro l: 42.3 ± 11.9 (NR)
Rome III c riteria and onl y 2–4 bowel
moveme nts/wee k during
preceding months
High-mi neral water
(1 L/day)
6 weeks Low-min eral tap
water (1 L/day)
Meta-a nalysis: stool f requency;
Narra tive sy nthesi s: stool
consis tency, pa in on defe cation.
Bothe et a l, 201739 Parallel R CT,
double-blind
Total: 76 (82.9%)
High-mi neral water: 3 8
Contro l: 38
High-mi neral water:
46.8 ± 1 2.9
Contro l: 46.0 ± 12.5, (N R)
Rome III c riteria and onl y 2–4 bowel
moveme nts/wee k during
preceding months
High-mineral sparkling
water (0 .5 L/day)
6 weeks Low-min eral sparkli ng
water (0 .5 L/day)
Meta-a nalysis: respo nse to treatme nt,
stool f reque ncy, integ rative
sympto ms, abdomina l pain,
QoL; Nar rative s ynthe sis: stool
consistency.
TABLE 2 (Continued)
164
|
VAN DER SCHOOT ET AL.
Study, year
(references) Study design Sampl e size (% female)
Age (yea rs), mean ± SD
(range)aConstipation diagnosis Intervention (dose) Duration Comparator (dose) Outcomes
Dupont e t al, 201940 Paralle l RCT,
double-blind
Total: 226 (100 %)
High-mi neral water: N R
Control: NR
High-mi neral water:
41.1 ± 11.4
Contro l: 41.3 ± 10.6 , (NR)
Rome III c riteria High-mi neral water
(1 L/day high-
miner al water
and 0.5 L /d low-
miner al water)
2 weeks Low-min eral water
(1.5 L/day)
Meta-a nalysis: respo nse to treatme nt,
stool f reque ncy, integ rative
sympto ms, abdomina l pain, laxativ e
use; Nar rative synth esis: st ool
consistency.
Anti et al , 199841 Pa ralle l RCT, no
blinding
Total: 141 (63%)i
Water supplementation:
NR
Control: NR
Water supplementation:
35.5 ± 11. 3
Contro l: 42.5 ± 13.6, (NR )
Rome III c riteria Water
supplementation
with min eral
water (2 L /day)
and high -fibre
diet (25 g /day)
8 weeks Ad libit um fluid intake
+ high-fi bre diet
(25 g/day)
Narra tive sy nthesi s: stool freque ncy,
laxati ve use.
Mego et al , 202343 Paralle l RCT,
double-blind
Total: 44 (95%)
High-fi bre diet : 22
Low-fibr e diet: 22
High-fi bre diet : 48 (median),
(23–5 9)
Low-fibr e diet: 45 (median) ,
(32–72)
Rome IV an d abnormal ano rectal
evacuation pattern with
manome try for chron ic
consti pation with dys synergic
defecation
High-fi bre diet
(25–30 g/
day). Diet itian
specified foods
for brea kfast
(wholegrain
cookie s), lunch
(vegetables
or legum es,
wholemeal bread,
meat, f ruit) and
dinner (c ream of
mixed veg etables,
wholemeal bread,
meat, f ruit).
Biofee dback
sessions
(3 × 30–45 mi n at
week 2, 5, 8 ).
9 weeks Low-fibr e diet
(15–20 g/
day). Diet itian
specified foods
for brea kfast
(non-wh ole grain
cereals/crackers,
white bread
sandw ich), lunch
(vegeta bles or
legume s, white
bread , meat, f ruit)
and dinn er (pasta,
rice, br oth or
salad , meat, fruit).
Biofee dback
sessions
(3 × 30–45 mi n at
week 2, 5, 8 ).
Narra tive sy nthesi s: response to
treatm ent, stool fre quency, stool
consis tency, fl atulen ce, abdominal
disten sion, abdomin al pain,
borbo rygmi, stra ining, incomp lete
evacua tion, di gestive well-b eing.
TABLE 2 (Continued)
|
165
VAN DER SCHOOT E T AL.
Study, year
(references) Study design Sampl e size (% female)
Age (yea rs), mean ± SD
(range)aConstipation diagnosis Intervention (dose) Duration Comparator (dose) Outcomes
Ho et al, 20 1242 Prospective
longit udinal, no
blinding
Total: 63 (75%) 47 (median), ( 20–80) Strai ning to ex pel bulky lar ge stool s,
or <1 bowel mo vement p er
3 days for at l east 3 months.
No-fibr e diet: for
2 weeks s top
intake of f ibre.
For 6 mont hs
contin ue with
as litt le fibre as
comfortable.
6 months –N arrative syn thesis : stool freque ncy,
bloating, straining.
Abbreviations: CSBM, complete spontaneous bowel movements; GT T, gut transit time; IBS-C , irritable bowel sy ndrome with constipation; NR , not repor ted; QoL , qualit y of life; RCT, randomised
controlled trial; Ref, reference.
aValues for whole study population unless intervention groups are specified.
bThe first 30 participants were assigned to the kiwifruit group to accommodate the limited growing season and fruit availability. The remainder of participants were r andomised through computer
generation to the prunes or psyllium groups.
cParticipant s told to consume kiwifruit without skin.
dConstipated compared with healthy population; dat a for cons tipated participants only included in this review.
eTotal population n = 184, data extracted for chronic constipation only.
fTotal population n = 73, dat a extracted for chronic constipation only.
gTotal population n = 59, data extracted for rye bread and white bread (control) groups only.
hTotal population n = 51, data extracted for rye bread and white bread (control) groups only.
iPropor tion female reported for those who completed study only (n = 117 out of 141).
TABLE 2 (Continued)
(SMD: −0.13, 95% CI: −0.69 to 0.43, p = 0.66; I2 = 41%, p = 0.19)
(Figure S2). In the par tially randomised study, straining was not
different between kiwifruit and psyllium (p = 0.36). 25 Frequency of
straining was reported in three studies of kiwifruit compared with
psyllium (n = 193); there was no difference in effect (SMD: −1.62,
95% CI: −4.45 to 1.21, p = 0. 26) and heterogeneit y was considerable
(I2 = 98%, p < 0.00001).29,44,45
Severity of bloating or abdominal pain was repor ted as not sig-
nificantly different between kiwifruit, prunes, or psyllium groups in
the par tially randomised study, however no p-value was provided.
The proportion of bowel movements with incomplete evacuation
was not different across groups (kiwifruit 50%, prunes 54%, psyllium
63%; p = 0 .49). 25 In another study not included in the meta-analysis
(inadequate washout period and data could not be extracted for the
first period for this outcome), there was no difference in bloating
between prunes and psyllium (mean points on 11-point scale: 4.1 vs.
3.5, higher score denoting more severe; p > 0.1). 24
Adverse events
No adverse event s occurred in one study each of kiwifruits, prunes
or mango compared with psyllium.24 ,2 9, 30 In the partially randomised
study, adverse events were reported (kiwifruit: 32% of participants,
prunes: 45%, psyllium: 66.6%, p = 0.06), including abdominal pain
(lower in kiwifruit 0%, vs. prunes 18% or psyllium 33%, p = 0.02),
abdominal discomfort (kiwifruit 3.7%, prunes 13.7%, psyllium 9.5%,
p = 0.44), bloating (lower in kiwifruit 10.5% and psyllium 9.5% vs.
prunes 36.3%, p = 0.03) and gas (kiwifruit 0%, prunes 18%, psyllium
19%, p = 0.16).25 In a study comparing kiwifruit with psyllium, 10/60
participants reported adverse events; however, details were not avail-
able44; in another, 4 events were treatment-related in the psyllium
group (struggle to swallow, abdominal discomfort and constipation).45
Compliance
In one RCT, compliance (self-repor ted) was 90%, however data for
kiwifruit and psyllium groups were not repor ted separately.29 In an-
other, all participants were reported compliant except for one with a
sensory dislike of psyllium.44 Compliance was reported to be higher
wi t h man go co m par ed wi th ps yll ium , but va l ues we re no t prov i ded . 30
Risk of bias
Bias in studies comparing fruits with psyllium was ‘high risk’, as par-
ticipants were not blinded.24,25,29,30,44,45 One study was open and
partially randomised25 and others had missing data.
30,44,45 One
cross-over study was at ‘high risk’ of bias arising from carryover ef-
fects due to an inadequate washout period (1 week); however, for
outcomes where data were extracted for the first period only, these
were not affected by carryover effects (Figures S1 and S2).
4.1.2 | Kiwifruit – uncontrolled trials
Three uncontrolled studies administered kiwifruit.26 –28 In one,
55% responded (increase of ≥1 CSBM/week)26; in another, 44%
166
|
VAN DER SCHOOT ET AL.
responded (increase of ≥1 BM/week).26 –28 Further information is
presented in Appendix S1.
4.1.3 | Fig paste – RCT
One RCT investigated fig paste (300 g /day) compared with placebo
for 8 weeks (n = 80).31 Fig resulted in softer stool consistency com-
pared with placebo (6.3 ± 3.6 vs. 2.8 ± 0.8 BSFS points, p = 0.024).
Further information is presented in Appendix S1.
4.2 | Rye bread – RCTs
Two RCTs administering rye bread (6 or 8 slices) compared with
white bread (control) for 3 weeks were combined for meta-analysis
(n = 48).32,33
Stool frequency
Rye bread resulted in higher stool frequency compared with white
bread (MD: +0.43 BM/week, 95% CI: 0.03–0.83, p = 0.03; I2 = 0%,
p = 0.66) (Figure 3A, Figure S3).
Stool consistency
Both studies administering rye bread measured stool consistency
(3-point scale; −1 loose, 0 normal, 1 hard); however, the studies re-
ported this outcome differently and data could not be combined for
meta-analysis. In one study, a higher proportion of participants had
softer stools with rye bread compared with white bread (6/10 (60%)
vs. 3/10 (30%); p = 0.037)33; in the other, stools were softer with r ye
bread (p < 0.001).32
TABLE 3 Result s of meta-analyses comparing (1) fruits vs. psyllium (comparator), (2) rye bread vs white bread (control) and (3) high-
mineral water vs. low-mineral water (control) for response to treatment, stool output, gut transit time, symptoms and qualit y of life in adults
with chronic constipation.
Outcome
Number of
studies in meta-
analysis (ref)
Results Heterogeneity
Participants
(n)
Meta-analysis overall
estimate (95% CI)apχ2 value p I2 (%)
Fruits vs psyllium
Response to treatment 3b25, 2 9,4 4 233 RR 1.18 (0.89 to 1.58) 0.25 0.90 0.64 0
Stool frequency 424,29,44,45 232 MD 0.36 (0.25 to 0.48) <0.00001 1.07 0.78 0
Stool consistency 6b24,25,29,30,44,45 315 MD 0.48 (0.11 to 0.86) 0.01 10.18 0.07 51
Gastrointestinal symptoms
(integrative score)
230,44 130 SMD −1.87 (−4.88 to 1.14) 0.22 43.23 <0.00001 98
Severit y of straining 2b24,25 86 SMD −0.13 (−0.69 to 0.43) 0.66 1.70 0.19 41
Frequency of straining 329,44,45 193 SMD −1.62 (−4.45 to 1.21) 0.26 106 .89 <0.00001 98
Rye bread vs. white bread
Stool frequency 232,33 48 MD 0.43 (0.03 to 0.83) 0.03 0.19 0.66 0
Gut transit time 232,33 48 MD −20.94 (−29.35 to −12.53) <0.00001 1.19 0.28 16
Gastrointestinal symptoms
(integrative score)
232,33 48 MD 2.00 (0.4 8 to 3.53) 0.01 0.38 0.54 0
High-mineral water vs low-mineral water
Response to treatment 3c37, 39, 4 0 539 RR 1.47 (1.20 to 1.81) 0.0002 2.18 0.54 0
Stool frequency 4c37– 40 638 MD 0.41 (−0.05 to 0.88) 0.08 9.4 4 0.05 58
Gastrointestinal symptoms
(integrative)
239,4 0 296 SMD −0.0 4 (−0.27 to 0.18) 0.70 0.46 0.50 0
Severit y of abdominal pain 3c3 7,3 9, 40 538 SMD −0.17 (−0.42 to 0.0 8) 0.18 5.58 0.13 46
Laxative use 2c37, 40 439 RR 0.47 (0.21 to 1.07) 0.07 4.58 0.10 56
Qualit y of life 2c37, 39 317 SMD 0.13 (−0.10 to 0.36) 0.27 1.08 0.58 0
Abbreviations: CI, confidence interval; MD, mean difference; RR, risk ratio; SMD, standardised mean difference.
aData were meta-analysed using a random-ef fect s model. Statis tical heterogeneity was assessed by using the chi-square test and quantified by using
the I2 statistic . p-values in bold are statistically significant (p < 0.05).
bPartially randomised, 3-arm s tudy (kiwifruit, prunes, psyllium) with only the randomised arms (pr unes, psyllium) included in the meta-analysis. 25
cAnalysis includes a study with two different doses of mineral water.37
|
167
VAN DER SCHOOT E T AL.
FIGURE 2 Forest plot of subgroup analysis based on type of fruit for (A) response to treatment (n = 233), (B) stool frequency (n = 232) and
(C) stool consistency (n = 315) in randomised controlled trials comparing fruits with psyllium (comparator) in adults with chronic constipation.
Values were calculated as (A) risk ratio; (B,C) mean dif ference (95% CIs) using a random-effects model. *Data extracted for first period only
for this crossover study.
(B)
Stool frequency
(C)
Stool consistency
(A)
Response to treatment
*
*
168
|
VAN DER SCHOOT ET AL.
Stool weight
In one study, stool weight was greater with rye bread compared with
white bread (marginal mean [95% CI]: 159 g/day [128–189] vs. 145 g/
day [114–175]; p = 0.008).33
Gut transit time
Rye bread resulted in shorter GTT compared with white bread (MD:
−20.94 h, 95% CI: −29.35 to −12.53, p < 0.00001; I2 = 16%, p = 0.28)
(Figure 3B, Figure S3).32,33
Gastrointestinal symptoms
Rye bread resulted in higher integrative symptom scores compared
with white bread (MD: +2.00 points, 95% CI: 0.48–3.53, p = 0.01;
I2 = 0%, p = 0.54) measured as a total score of 18 (Figure 3C, Figure S3).
Di ffic ul t y in de fec ati on (3 -p o in t scal e: −1 easy, 0 norm a l, 1 st rai n-
ing at defecation), was reported differently in studies, and data could
not be combined for meta-analysis. In one study, more participants
had easing of defecation with rye bread compared with white bread
(6/10 (60%) vs. 3/10 (30%); p = 0.018)33; in the other, mean difficulty
of defecation was 0.1 ± 0.3 with rye bread, compared with 0.5 ± 0.35
with white bread (p-value not reported).32
Adverse events
One study repor ted that gastrointestinal symptoms including flatu-
lence and bloating, were ‘strongest’ in the first week of rye bread
and then diminished.32 Another reported that adverse effects did
not differ significantly between groups.33
Compliance
In one study, participants consumed less rye bread than instructed
(mean intake 7.5, instructed to eat ≥8 slices/day). Participant s in
the white bread group adhered to the intervention (mean intake 6,
instructed to eat ≤8 slices/day).32 In another, compliance (dietary
records) was 96% in week 1 and 85% in week 3 with rye brea d; how-
ever, values were not reported for white bread.33
FIGURE 3 Forest plot of (A) stool frequency (n = 48), (B) gut transit time (n = 48) and (C) integrative symptom scores (n = 48) in
randomised controlled trials comparing rye bread with white bread (control) in adult s with chronic constipation. Values were calculated as
mean difference (95% CIs) using a random-effects model.
(A)
Stool frequency
(B)
Gut transit time
(C)
Integrativesymptoms
|
169
VAN DER SCHOOT E T AL.
Risk of bias
Bias was ‘high risk’ in studies administering rye bread, as both were
unblinded,32,33 and one did not provide information on missing
data32 (Figure S3).
4.3 | OTHER FOODS
4.3.1 | Pasteurised yoghurt – RCT
One RCT investigated pasteurised yoghur t (220 mL/day) compared
with pasteurised milk (control) for 7 weeks (n = 120).36 Yoghurt re-
sulted in higher stool frequency than control (p < 0.01). Fur ther in-
formation is presented in Appendix S1 .
4.3.2 | High-fibre cereal – uncontrolled trial
One uncontrolled trial investigated high-fibre cereal (increased
from 7 to 28 g/day) for 6 weeks (n = 41) and stool frequency
increased (p < 0.01).34 Further information is presented in
Appendix S1.
4.3.3 | Oat bran biscuits – uncontrolled trial
One uncontrolled trial investigated oat bran biscuits (2/day) for
12 weeks (n = 50) and the proportion of participants with stool fre-
quency ≥1 BM/day increased (p < 0.01).35 Further infor mat ion is pre-
sented in Appendix S1.
5 | DRINKS
5.1 | High-mineral water – RCTs
Four RCTs administering high-mineral water compared with low-
mine ral wat er (co nt rol ) we re comb in ed fo r met a- an aly si s (n = 755).37–
40 One study administered two doses of high-mineral water.37
5.1.1 | Response to treatment
Response to treatment was defined as proportion of participants
with ≥4 BM/week, an increase of ≥2 BM compared with baseline plus
<25% lumpy/hard stools,37, 4 0 or softer stools.
39 Overall, 144/313
(46%) responded to high-mineral water and 73/226 (32%) to low-
mineral water (RR: 1.47, 95% CI: 1.20–1.81, p = 0.0002; I2 = 0%,
p = 0.54) (Figure 4A, Figure S4).
5.1.2 | Stool frequency
Stool frequency (BM/week) in four studies administering high-mineral
water (n = 638)37–4 0 was not different compared with low-mineral
water (MD: +0.41 BM/week, 95% CI: −0.05 to 0.88, p = 0.08) and het-
erogeneity was substantial (I2 = 58%, p = 0.05) (Figure 4B, Figure S4).
5.1.3 | Stool consistency
Stool consistenc y was measured in four studies (BSFS); how-
ever, one did not report endpoint38 and three reported data in
FIGURE 4 Forest plot of (A) response to treatment (n = 539) and (B) stool frequency (n = 638) in randomised controlled trials comparing
high-mineral water with low-mineral water (control) in adults with chronic constipation. Values were calculated as (A) risk ratio, (B) mean
difference (95% CIs) using a random-effect s model.
(A)
Response to treatment
(B)
Stool frequency
170
|
VAN DER SCHOOT ET AL.
forms that could not be combined for meta-analysis.37, 3 9, 40 In one
study, high-mineral water resulted in softer stools compared to
low-mineral water (p < 0.001).39 In another, 72.2% of participants
had normal stools with high-mineral water compared with 65.4%
with low-mineral water, but no p-value was reported.40 In an ot h er,
change in stool consistency was not different between groups
(p > 0.05).37
5.1.4 | Gastrointestinal symptoms
Integrative symptom scores were reported in two studies using
the GSRS39 or Rome III criteria40 (n = 296); there was no dif fer-
ence for high-mineral water compared with low-mineral water
(SMD: −0.04, 95% CI: −0.27 to 0.18, p = 0.70; I2 = 0%, p = 0.50)
(Figure S4).
Severity of abdominal pain was reported in three studies
(n = 538)3 7,3 9, 4 0; there was no difference for high-minera l water com-
pared with low-mineral water (SMD: −0.17, 95% CI: −0.42 to 0.08,
p = 0.18; I2 = 46%, p = 0.13) (Figure S4). Severit y of pain on defeca-
tion (scale 0–10) was measured in one study (n = 100); 80% of par-
ticipant s had less pain after high-mineral water, compared with 74%
after low-mineral water (p = 0.33).38
5.1.5 | Laxative use
Laxative use in two studies (n = 439) was not different for high-
mineral water compared with low-mineral water (RR: 0.47, 95% CI:
0.21–1.07, p = 0.07; I2 = 56%, p = 0.10) (Figure S5).3 7,4 0
5.1.6 | Quality of life
Quality of life, measured using PAC-QoL37 or 12-Item Short Form
Survey (SF-12)39 (n = 317), was not different for high-mineral water
compared with low-mineral water (SMD: 0.13, 95% CI: −0.10 to 0.36,
p = 0.27; I2 = 0%, p = 0.58) (Figure S5).
5.1.7 | Adverse events
In one study, 20 adverse events were reported; two (abdominal
bloating, gas) were considered related to high-mineral water.40 In
another, there was no difference in the number of adverse events
bet wee n groups (p = 0. 571). 39 In another, 10 adverse events were re -
ported and two (diarrhoea, gas) were possibly treatment-related.38
One study reported 74 adverse events (high-dose: 21, low-dose: 28,
control: 25), and diarrhoea was most frequently reported, occur-
ring in 7 participant s each in the high- and low-dose groups, and 5
in control.37
5.1.8 | Compliance
Compliance (unconsumed product) was 100% for high- and low-
mineral water groups in two studies,37, 3 9 and 93.4% ± 10.8% for
high-mineral water compared to 93.1% ± 9.9% for low-mineral water
in another.40 One study reported 2.2 ± 4.9% returned product with
high-mineral water compared with 3.7 ± 5.8% with low-mineral
water.38
5.1.9 | Risk of bias
Bias in two studies administering high-mineral water was ‘high
risk ’, 37, 3 8 while two were of ‘some concerns’3 9,4 0 (Figures S4 and S5).
5.2 | Water supplementation – RCT
One RCT investigated water supplementation (2 L/day) compared
with control ad libitum fluid intake for 2 months (n = 141), both
alongside a high-fibre diet (25 g/day).41 Water supplementation
resulted in higher stool frequency (4.2 ± 1.3 vs. 3.3 ± 1.8/week ,
p < 0.001) compared with control. Further information is pre-
sented in Appendix S1.
5.3 | Prune juice – RCT
One RCT investigated prune juice (54 g/day) compared with pla-
cebo for 8 weeks (n = 84). Prune juice resulted in softer stool
consistency compared with placebo (3. 57 ± 0.81 vs. 3.03 ± 1.10
BSFS points, p = 0.012). Further information is presented in
Appendix S1.
6 | DIET
6.1 | High-fibre diet – RCT
One RCT investigated a high-fibre diet (25–30 g/day) compared with
a low-fibre diet (control, 15–20 g/day) for 9 weeks (n = 44). The high-
fibre diet had no impact on stool frequency (p = 0.12) and resulted in
worse flatulence (p = 0.03) compared with the low-fibre diet. Further
information is presented in Appendix S1 .
6.2 | No-fibre diet – uncontrolled trial
One uncontrolled trial investigated reducing dietary fibre intake
for 6 months (n = 63),42 and stool frequency increased (p < 0.001).
Further information is presented in Appendix S1.
|
171
VAN DER SCHOOT E T AL.
7 | DISCUSSION
This systematic review and meta-analysis established that a small
number of foods, drinks and diet inter ventions have been investi-
gated in chronic constipation, and few as RCTs. Only kiwifruits,
prunes, rye bread, and high-mineral water have been investigated
in more than one RCT. Mango, fig product, pasteurised yoghurt and
prune juice have been investigated in only one RCT each, and high-
fibre cereal and oat bran biscuits in one uncontrolled study each.
Astonishingly, only two whole diet interventions have been inves-
tigated: a high-fibre diet (RCT) and a ‘no-fibre’ diet (uncontrolled
study).
Fruits improved stool frequency and consistency compared
with psyllium, itself a fibre supplement with established effec-
tiveness in constipation.13,47 No met a-analyses favoured psyl-
liu m over fr uits, which may indic ate that fr uit s were as ef fe ctive
at improving constipation outcomes as psyllium, although type
II errors cannot be ruled out even in a meta-analysis. Fibre sup-
plements also worsen flatulence; thus, fruits may be considered
as an alternative treatment option. Kiwifruits (2–3/day), in par-
ticular, are well-tolerated, and improved stool frequency com-
pared with psyllium. Kiwifruit s are high in soluble and insoluble
fibres and increase stool water content and colonic volume in
MRI studies.48–51 Kiwifruits also contain a proteolytic enzyme,
actinidin, which is shown in animal studies to accelerate gas-
tric emptying, and is suggested to accelerate gut motility and
affect pain reception and anti-inflammatory responses in the
gut through protease-activated receptors, which could explain
effects in improving gastrointestinal symptoms.52–54 Kiwifruit
may also have butyrogenic, bifidogenic, and anti-inflammatory
effects, observed in a human gastrointestinal model of consti-
pation.55 Pr unes we r e s t udi e d in tw o tr ial s ; o ne w as exp lor ato r y,
partially randomised and not powered to detect effectiveness,
and in another, it was unclear whether the study was powered
as a non-inferiority or equivalence trial.24,2 5 Prunes were no
more effective at improving stool output compared with psyl-
lium. However, a previous adequately powered study in indi-
viduals with infrequent bowel movements, but not meeting
formal constipation diagnostic criteria, showed that prunes (80,
120 g/day) plus water (300 mL/day) increased stool frequency
compared with water alone, but GTT was unchanged.56 Further
adequately powered active control equivalence or placebo-con-
trolled studies are needed to establish the effect of prunes in
constipation.
Rye bread (7 slices/day) increased stool frequency by +0.4 bowel
movements/week, which may not represent a clinically meaningful
change, and accelerated GTT (−21 h) compared with white bread.
Rye bread worsened flatulence, which may be attributed to its high
fructan and fructose content contributing to fermentation-induced
luminal gas production.5 7–59 The introduction of this high-fibre food
was not gradual, as recommended to avoid side effects.60 Both stud-
ies were at high risk of bias; therefore, findings must be interpreted
with caution.32,33
Diet interventions were defined as a change in habitual di-
etary pattern (i.e. including all or most aspects of the diet including
meals and food groups) in a form that could be widely applied in
practice. Surprisingly, only one RCT administering a high-fibre diet
(25–30 g/day) was identified, and this was halted before reaching
the expected sample size due to lack of efficacy, after having no
impact on response to treatment or stool output, and worsening
flatulence and abdominal distention compared with a low-fibre diet
(15–20 g/day).55 This review is limited to include studies in chronic
constipation only; however, studies administering a high-fibre diet
in secondary constipation have been performed, for example, an
RCT in patients on peritoneal dialysis, also showing no effect when
compared with placebo.61 The other eligible diet study in this re-
view, assessing a ‘no-fibre’ diet (i.e. reducing fibre intake), showed
a beneficial effec t on constipation outcomes; however, this was
uncontrolled. Both single studies were at high risk of bias and not
meta-analysed; therefore, no meaningful conclusions can be drawn
on their effectiveness in chronic constipation. Interestingly, one RCT
showed that water supplementation (2 L/day) alongside a high-fibre
diet (25 g/day) may improve constipation.41 However, since the con-
trol group also consumed a high-fibre diet, it remains unclear if this is
the result of increasing fluid and fibre intake together, or fluid intake
alone. There is a need for high-quality RCTs of feeding or dietary
advice studies for the whole diet approach in chronic constipation.
Conversely, increasing fibre intake through fibre supplementation is
well-studied, and systematic reviews and meta-analyses show that
certain fibre supplements beneficially impact cardinal constipation
symptoms.13,47 Therefore, increasing fibre in the diet with particular
attention to the types and combinations of fibres may be more ef-
fective in the management of chronic constipation than increasing
fibre through a range of different foods, though this remains to be
confirmed.
High-mineral water (0.5–1.0 L/day), included within usual fluid in-
take, increased response to treatment compared low-mineral water,
but did not impact other constipation outcomes. High-mineral water
is rich in magnesium sulphate, which induces osmosis in the gut
lumen, potentially leading to accelerated GTT and softer stools.62,6 3
However, studies included in this review did not measure GTT, and
reported varied findings on stool consistency. The only other drink
included in this review was prune juice, which improved stool out-
put compared with placebo.46 This may be attributed to sorbitol,
whi ch draws wate r into the gut lu me n, potentiall y soft ening stools.11
Further RCTs are required to establish the effect of prune juice in
chronic constipation.
The findings of this sys tematic review and met a-anal ysis are cru-
cial for clinical practice. People with constipation experience high
rates of dissatisfaction and treatment failure.12,64–66 Current guide-
lines recommend foods that have not been tested (e.g. apples, apri-
cots), or are not shown to be effective (e.g. high-fibre diet), whereas
interventions showing effectiveness in chronic constipation are not
recommended specifically (e.g. kiwifruits).6–8,60 This sh ow s an urg ent
need to conduc t fur ther high-quality RCTs on dietary approaches,
and update guidelines for the management of chronic constipation
172
|
VAN DER SCHOOT ET AL.
to reflect current evidence to increase the effectiveness of clinical
advice and improve patient care.
This is the first ever systematic review and meta-analysis to
comprehensively assess the effect of all available intervention
trials of whole foods, drinks and diets (controlled, uncontrolled,
randomised, non-randomised) in chronic constipation that can be
widely applied in practice; 23 trials were identified, administering
13 different interventions.16 Effort was made to search grey lit-
erature with no language restrictions, reducing publication bias.
Important limitations include the source literature, the extremely
small number of studies included, and significant heterogeneit y
an d hi gh ri sk of bia s amo ng tr ial s, wh ic h may lead to an ov er est ima-
tion of the effect size. However, the challenge of placebo-control
groups in food an d diet inter ventions may be dif ficult to f ull y ove r-
come.18,67 This review's strict criteria to include only interventions
that can be widely applied in practice means that studies adminis-
tering a composite dish were not included, due difficulty to repli-
cate these in the real-life setting; however, this would be valuable
to address in future work.
In conclusion, this systematic review and meta-analysis high-
lights that few studies exist on whole food, drink, and diet inter-
ventions for individuals with chronic constipation, thus findings
must be interpreted with caution. The small number of studies
indicate a potentially beneficial effect of fruits on stool output.
In particular, 2–3 kiwifruits daily are considered a well-tolerated
treatment option and may improve stool frequency. While rye
bread (7 slices) may increase stool frequency and GTT, it also
worsens flatulence. High-mineral water (0.5–1 L/day) may in-
crease overall response to treatment; however, there is no evi-
dence of an effect on specific constipation symptoms. There is
a need for further high-quality RCTs that use appropriate con-
trols to establish the effect of specific foods, drinks and diets
(e.g. prunes, water supplementation, high-fibre diets) in chronic
constipation.
Registration and protocol
The protocol for this review is registered on PROSPERO (ref
CRD42021241072) available on h t t p s : / / w w w . c r d . y o r k . a c . u k / p r o s p
e r o / d i s p l a y _ r e c o r d . p h p ? R e c o r d I D = 241072
AUTHOR CONTRIBUTIONS
Alice van der Schoot: Conceptualization (equal); methodology
(equal); writing – original draft (equal); writing – review and edit-
ing (equal). Zoi Katsirma: Conceptualization (equal); methodology
(equal); writing – original draft (equal); writing – review and edit-
ing (supporting). Kevin Whelan: Conceptualization (equal); fund-
ing ac qu is it ion (su pp ort ing ); method ology (equal); writin g – review
and editing (equal). Eirini Dimidi: Conceptualiz at io n (l ead); funding
acquisition (lead); methodology (equal); writing – review and edit-
ing (lead).
ACKNO WLE DGE MENTS
Declaration of personal interests: AV has been supported by a doc-
toral fellowship from the Almond Board of California. ZK has been
supported by a doctoral fellowship from the International Nut and
Dried Fruit Council. KW has received research funding from gov-
ernment bodies including National Institute of Health Research and
Medical Research Council, charities including Crohn's and ColitisUK ,
ForCrohns, The Leona M. and Harry B. Helmsley Charitable Trust,
Kenneth Rainin Foundation, as well as from industrial sources in-
cluding Almond Board of California, Danone and the International
Dried Fruit and Nut Council, and is the coinventor of biomarkers in
the diagnosis and management of IBS. ED has received an educa-
tion grant from Alpro, research funding from the British Dietetic
Association, Almond Board of California, the International Nut and
Dried Fruit Council and Nestec Ltd and has served as a consultant
for Puratos. This systematic review was part of a project funded by
the General and Education Trust Fund, British Dietetic Association
(grant ID: 19/14). Funders had no role in the design, conduct or pub-
lishing of this review.
FUNDING INFORMATION
Funding provided by the General and Education Trust Fund, British
Dietetic Association. Funders had no role in the design, conduct or
publishing of this review.
SYSTEMATIC REVIEW REGISTRATION
PROSPERO registration number CRD42021241072.
AUTHORSHIP
Guarantor of the article: ED.
ORCID
Alice Van Der Schoot https://orcid.org/0000-0002-2889-7900
Kevin Whelan https://orcid.org/0000-0001-5414-2950
Eirini Dimidi https://orcid.org/0000-0003-0140-0199
REFERENCES
1. Sperber AD, Bangdiwala SI, Drossman DA, Ghoshal UC, Simren M,
Tack J, et al. Worldwide prevalence and burden of functional gas-
trointestinal disorders, results of Rome Foundation Global Study.
Gastroenterology. 2021;160:99–114.e3.
2. Belsey J, Greenfield S, Candy D, Geraint M. Systematic review: im-
pact of constipation on qualit y of life in adults and children . Aliment
Pharmacol Ther. 2010;31:938–49.
3. L acy BE, Mearin F, Chang L , Chey WD, Lembo AJ, Simren M, et al.
Bowel disorders. Gastroenterology. 2016;150(6):1393–407.e5.
4. Cai Q, Buono JL , Spalding WM, Sarocco P, Tan H , Stephenson JJ,
et al. Healthcare costs among patients with chronic constipation: a
retrospective claims analysis in a commercially insured population .
J Med Econ. 2014;17:148–58.
5. Peer y AF, Crocket t SD, Murphy CC, Jensen ET, Kim HP, Egberg MD,
et al. Burd en an d cost of gastrointes ti nal, liv er, and pancre atic dis eases in
the United States: update 2021. Gastroenterolog y. 2022;162:621–44.
6. Serra J, Pohl D, A zpiroz F, Chiarioni G, Ducrotté P, G ourcerol G,
et al. European society of neurogastroenterology and motility
guidelines on functional constipation in adults. Neurogastroenterol
Motil. 2020;32(2):e13762.
7. Bharucha AE, Dorn SD, Lembo A, Pressman A . American
Gastroenterological Association medical position statement on
constipation. Gastroenterology. 2013;144(1):211–7.
|
173
VAN DER SCHOOT E T AL.
8. National Institute for Health and Care E xcellence. Clinical knowl-
edge summary: constipation in adults. 2023 h t t p s : / / c k s . n i c e . o r g .
u k / c o n s t i p a t i o n # ! s c e n a r i o
9. Gill SK, Rossi M, Bajka B, Whelan K . Dietary fibre in gastro-
intestinal healt h and disease. Nat Rev Gastroenterol Hepatol.
2021;18:101–16.
10. Katsirma Z, Dimidi E, Rodriguez-Mateos A, Whelan K. Fruits and
their impact on the gut microbiota, gut motility and constipation.
Food Funct. 2021;12:8850–66.
11. Lu Y. Humect ancies of D-tagatose and D-sorbitol. Int J Cosmet Sci.
2001;23:175–81.
12. Johanson JF, Krals tein J. Chronic constipation: a survey of the pa-
tient perspec tive. Aliment Pharmacol Ther. 2007;25(5):599–608.
13. van der Schoot A, Drysdale C, Whelan K, Dimidi E. The effect of
fiber supplementation on chronic constipation in adults: an updated
systematic review and meta-analysis of randomized controlled tri-
als. Am J Clin Nutr. 2022;116(4):953–69.
14. van der Schoot A, Helander C, Whelan K, Dimidi E. Probiotics
and synbiotics in chronic constipation in adults: a systematic re-
view and meta-analysis of randomized controlled trials. Clin Nutr.
2022;41(12):2759–77.
15. van der Schoot A, Creedon A, Whelan K, Dimidi E. The effect of
food, vitamin, or mineral supplements on chronic constipation in
adults: a systematic review and meta-analysis of randomized con-
trolled trials. Neurogastroenterol Motil. 2023;e14613. h t t p s : / / d o i .
o r g / 1 0 . 1 1 1 1 / n m o . 1 4 6 1 3
16. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ,
et al. Cochrane handbook for systematic reviews of interventions
version 6.4 (updated august 2023). Cochrane; 2023. w w w . t r a i n i n g .
cochr ane. org/ handbook
17. Page MJ, McKenzie JE, B ossuy t PM, Boutron I , Hoffmann TC,
Mulrow CD, et al. The PRISMA 2020 statement: an updated guide-
line for reporting systematic reviews. BMJ. 2021;372:n71.
18. Staudacher HM, Yao CK, Chey WD, Whelan K . Optimal design of
clin ical trials of dietar y interventions in dis ord ers of gut-brain int er-
action. Am J Gastroenterol. 2022;117(6):973–84.
19. Huwaldt JA, Steinhorst S. PlotDigitizer Sof tware version 2.6.9.
2020 h t t p s : / / p l o t d i g i t i z e r . s o u r c e f o r g e . n e t /
20. Sterne JAC, Savović J, Page MJ, Elbers RG , Blencowe NS, Boutron
I, et al. RoB 2: a revised tool for assessing risk of bias in randomised
trials. BMJ. 2019;366:I4898.
21. Munn Z, Moola S, Lisy K, Riitano D, Tufanar u C. Methodological
guidance for systematic reviews of observational epidemiological
studies repor ting prevalence and incidence data. Int J Evid Based
Healthc. 2015;13:147–53.
22. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and
standard deviation from the sample size, median, range and/or in-
terquar tile range. BMC Med Res Methodol. 2014;14:135.
23. Richardson M, Garner P, Donegan S. Interpretation of subgroup
analyses in systematic reviews: a tutorial. Clin Epidemiol Glob
Health. 2019;7(2):192–8.
24. Attaluri A, Donahoe R, Vales tin J, Brown K, Rao SSC . Randomised
clinical trial: dried plums (prunes) vs . psyllium for constipation.
Aliment Pharmacol Ther. 2011;33:822–8.
25. Chey SW, Chey WD, Jackson K, Eswaran S. Exploratory comparative
ef fe c tive ne s s tr ial of g re e n kiwif ru i t , ps y l l ium, or pr un e s in US pa tient s
with chronic constipation. Am J Gastroenterol. 2021;116:1304–12.
26. Chan AOO, Leung G , Tong T, Wong N YH. Increasing dietary fiber
intake in terms of kiwifruit improves constipation in Chinese pa-
tients . World J Gas troenterol. 20 07;13:4771–5.
27. Hiele. Effect of consumption of kiwifruit on constipation in adults.
Confidential report to Zespri International Ltd Gastroenterology
Depar tment , Catholic University of Leuven . 2010.
28. Cunillera O, A lmeda J, Mascort JJ, Basora J, Marzo-Castillejo M.
Improvement of func tional constipation with kiwifruit intake in a
Mediterranean patient population: an open, non-randomized pilot
study. Rev Esp Nutr Hum Diet. 2015;19:58–67.
29. Eady SL, Wallace AJ, Butts C A, Hedderley D, Drummond L, A nsell
J, et al. The effect of “Zesy002” kiwifruit (Actinidia chinensis var.
chinensis) on gut health function: a randomised cross-over clinic al
trial. J Nutr Sci. 2019;8:e18.
30. Venancio VP, Kim H, Sirven MA, Tekwe CD, Honvoh G, Talcott
ST, et al. Polyphenol-rich mango (Mangifera indica L.) ameliorate
functional constipation symptoms in humans beyond equivalent
amount of fiber. Mol Nutr Food Res. 2018;62:1701034.
31. Baek HI, Ha KC, Kim HM, Choi EK, Park EO, Park BH, et al.
Randomized, double-blind, placebo-controlled trial of Ficus carica
paste for the management of functional constipation. Asia Pac J
Clin Nutr. 2016;25:487–96.
32. Hongis to SM, Paajanen L, Saxelin M , Korpela R. A combination of
fibre-rich rye bread and yoghurt containing Lactobacillus GG im-
proves bowel function in women with self-reported constipation.
Eur J Clin Nutr. 2006;60:319–24.
33. Holma R, Hongisto SM, Saxelin M, Korpela R. Constipation is re-
lieved more by Rye bre ad th an wh eat brea d or la xat ives without in-
creased adver se ga str oint est inal effect s. J Nutr. 2010 ;140:5 34– 41.
34. Shariati A, Maceda JS, Hale DS. High-fiber diet for treatment of
constipation in women with pelvic floor disorders. Obstet Gynecol.
2008;111:9 08–13.
35. Valle-Jones JC. An open study of oat bran meal biscuits (‘Lejfibre’)
in the treatment of constipation in the elderly. Curr Me d Res Opin.
1985;9:716–20.
36. Liu ZM , Xu Z Y, Han M, Guo BH. Efficacy of pasteurised yoghurt in
improving chronic constipation: a randomised, double-blind, place-
bo-controlled trial. Int Dairy J. 2015;40:1–5.
37. Dupont C, Campagne A, Constant F. Efficac y and safety of a mag-
nesium sulfate-rich natural mineral water for patients with func-
tional constipation. Clin Gastroenterol Hepatol. 2014;12:1280–7.
38. Naumann J, Sadaghiani C, Alt F, Huber R. Effects of sulfate-rich
mineral water on functional constipation: a double-blind, ran-
domized, placebo-controlled study. Forsch Komplementmed.
2016;23:356–63.
39. Bothe G, Coh A, Auinger A. Efficacy and safety of a natural min-
eral water rich in magnesium and sulphate for bowel function: a
double-blind, r andomized, placebo-controlled study. Eur J Nutr.
2017;56:491–9.
40. Dupont C, Constant F, Imber t A , Hébert G, Zourabichvili O, Kapel
N. Time to treatment response of a magnesium- and sulphate-rich
natural mineral water in functional constipation. Nutrition.
2019;65:167–72.
41. Anti M, Pignataro G, Armuzzi A, Valenti A, Lascone E, Marmo R,
et al. Water supplement ation enhances the effect of high-fiber diet
on stool frequency and laxative consumption in adult patients with
functional constipation. Hepatogastroenterology. 1998;45:727–32.
42. Ho KS, Tan CYM, Daud MAM, Seow-Choen F. Stopping or reducing
dietar y fiber intake reduces constipation and it s associated symp-
toms. World J Gastroenterol. 2012;18:4593–6.
43. Mego M, Huaman JW, Videla S, Jansana M, Tinoco K, Saperas E.
Effectiveness of a high fiber diet in improving constipation in pa-
tients with defecatory dyssynergy under treatment with anorrectal
biofeedback. Exploratory, randomized clinical trial. Gastroenterol
Hepatol. 2023. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . g a s t r o h e p . 2 0 2 3 . 0 1 . 0 1 2
44. Gearry R, Fukudo S, Barbara G, Kuhn-Sherlock B, Ansell J,
Blatchford P, et al. Consumption of 2 green kiwifruits daily im-
proves constipation and abdominal comfort-results of an interna-
tional multicenter randomized controlled trial. Am J Gastroenterol.
2023;118(6):1058–68.
45. Bayer SB, Heenan P, Frampton C , Wall CL, Drummond LN, Roy NC,
et al. Two gold kiwifruit daily for effec tive treatment of constipation
in adult s-a randomized clinical trial. Nut rients. 2022;14(19):4146.
174
|
VAN DER SCHOOT ET AL.
46. Koyama T, Nagata N, Nishiura K , Miura N, Kawai T, Yamamoto H.
Prune juice containing sorbitol, pectin, and polyphenol ameliorates
subjective complaints and hard feces while normalizing stool in
chronic constipation: a randomized placebo-controlled trial. Am J
Gastroenterol. 2022;117(10):1714–7.
47. Christodoulides S, Dimidi E, Fragkos KC, Farmer AD, Whelan K,
Scott SM. Systematic review with meta-analysis: effect of fibre
supplementation on chronic idiopathic constipation in adults.
Aliment Pharmacol Ther. 2016;44(2):103–16.
48. Sutherland P, Hallett I, Redgwell R , Benhamou N , MacRae E.
Localization of Cell Wall polysaccharides during kiwifruit ( Actinidia
deliciosa) ripening. Int J Plant S ci. 1999;160:1099–109.
49. Dawson DM, Melton LD. Two pectic polys accharides from kiwifruit
cell walls. Carbohydr Poly m. 1991;15:1–11.
50. Redgwell RJ, Melton LD, Brasch DJ. Cell-wall polysaccharides of
kiwifruit (Actinidia deliciosa): chemical features in different tissue
zones of the fruit at harvest. Carbohydr Res. 1988;182:241–58.
51. Wilkinson-Smith V, Dellschaf t N, Ansell J, Hoad C, Marciani L,
Gowland P, et al. Mechanisms underlying effects of kiwifruit on
intestinal function shown by MRI in healthy volunteers. Aliment
Pharmacol Ther. 2019;49:759–68.
52. Montoya CA, Hindmarsh JP, Gonzalez L, Boland MJ, Moughan PJ,
Rutherfurd SM. Diet ar y ac ti nid in from kiwi fr uit (Actinidia deliciosa cv.
Hay ward) inc re as es gas tr ic diges tion and the ga st ric em pt ying rate of
several dietar y proteins in growing rat s. J Nutr. 2014;144:440–6.
53. Bayer SB, Frampton CM, G earr y RB, Barbara G. Habitual green ki-
wifruit consumption is associated with a reduction in upper gas-
trointestinal symptoms: a systematic scoping review. Adv Nutr.
2022;13(3):846–56.
54. Bayer SB, Gearr y RB, Drummond LN. Putative mechanisms of kiwi-
fruit on maintenance of normal gastrointestinal func tion. Crit Rev
Food Sci Nutr. 2018;58:2432–52.
55. Goya-Jorge E, Bondue P, Gonza I, Laforêt F, Antoine C, Boutaleb
S, et al. Butyrogenic, bif idogenic and slight anti-inflammatory
effec ts of a green kiwifr uit powder (K iwi FFG®) in a human gas-
trointestinal model simulating mild constipation. Food Res Int.
2023;173:113348.
56. Lever E, Scott SM, Louis P, Emer y PW, Whelan K. The effect of
prune s on stool outp ut, gut transit time and gas trointestinal micro-
biota: a randomised controlled trial. Clin Nutr. 2019;38:165–73.
57. Karppinen S, Myllymäki O, For ssell P, Poutanen K. Fructan content
of rye and r ye products. Cereal Chem. 2003;80:168–71.
58. Shepherd SJ, Parker FC, Muir JG, Gibson PR. Dietary triggers of ab-
dominal symptoms in patients wit h irritable bowel syndrome: ran-
domized placebo-controlled evidence. Clin Gastroenterol Hepatol.
2008;6:765–71.
59. Bruhw yler J, Carreer F, Demanet E, Jacobs H. Digestive toler-
ance of inulin-type fructans: a double-blind, placebo-controlled,
cross-over, dose-r anging , r andomized study in healthy volunteers.
Int J Food Sci Nutr. 2009;60:165–75.
60. Ford AC, Moay yedi P, Lacy BE, Lembo AJ, Saito YA, Schiller LR,
et al. American College of Gastroenterology monograph on the
management of irritable bowel syndrome and chronic idiopathic
constipation. A m J Gastroenterol. 2014;109(Suppl 1):S2–S26.
61. Sutton D, Ovington S, Engel B. A multi-Centre, randomised t rial to
assess whether increased dietary fibre intake (using a fibre supple-
ment or high-fibre foods) produces healthy bowel performance and
reduces laxative requirement in free living patients on peritoneal
dialysis . J Ren Care. 2014;40(3):157–63.
62. Vu MK, Nouwens MAG, Biemond I, Lamers CBHW, Masclee AAM.
The osmotic laxative magnesium sulphate ac tivates the ileal brake.
Aliment Pharmacol Ther. 2000;14:587–95.
63. Izzo AA , Gaginella TS, Capasso F. The osmotic and intrinsic mech-
anisms of the pharmacologic al laxative action of oral high doses of
magnesium sulphate. Impor tance of the release of digestive poly-
peptides and nitric oxide. Magnes Res. 1996;9:133–8.
64. Oh SJ, Fuller G , Patel D, Khalil C, Spalding W, Nag A, et al. Chronic
constipation in the United States: results from a population-based
survey assessing healthcare seeking and use of pharmacotherapy.
Am J Gastroenterol. 2020;115:895–905.
65. Lacy BE, Shea EP, Manuel M, Abel JL, Jiang H, Taylor DCA. Lessons
learned: chronic idiopathic constipation patient experiences with
over-the-counter medications. PLoS One. 2021;16:e0243318.
66. Guerin A, Carson R, Lewis B, Yin D, Kaminsky M, Wu E. The economic
burden of treatment failure amongst patients with irritable bowel
syndrome with constipation or chronic constipation: a retrospective
analysis of a Medicaid population. J Med Econ. 2014;17:577–86.
67. Staudacher HM, Irving PM, Lomer MCE, W helan K . The challenges
of control groups, placebos and blinding in clinical trials of dietary
interventions. Proc Nutr Soc. 2017;76:203–12.
SUPPORTING INFORMATION
Additional supporting information will be found online in the
Supporting Information section.
How to cite this article: Van Der Schoot A, Kat sirma Z,
Whelan K, Dimidi E. Systematic review and meta-analysis:
Foods, drinks and diets and their effect on chronic
constipation in adults. Aliment Pharmacol Ther.
2024;59:157–174. http s://doi.o rg /10.1111/apt.17782
