J Gastroenterol
DOI 10.1007/s00535-017-1371-9
REVIEW
Meta-analysis: proton pump inhibitors moderately increase
the risk of small intestinal bacterial overgrowth
Tingting Su1,2 • Sanchuan Lai1,2 • Allen Lee3 • Xingkang He1,2 • Shujie Chen1,2
Received: 8 June 2017 / Accepted: 15 July 2017
Ó Japanese Society of Gastroenterology 2017
Abstract The use of proton pump inhibitors (PPIs) may
potentially predispose to the development of small
intestinal bacterial overgrowth (SIBO), but this association
is controversial due to conflicting results from studies
conducted to date. The aim of this meta-analysis was to
evaluate the association between the use of PPIs and the
risk of SIBO. We systematically searched the online
PubMed, Embase, and Cochrane Library databases and
Web of Science for relevant articles published up to
November 2016. Two researchers identified and extracted
data independent of each other. The pooled analysis was
performed using the generic inverse-variance random-ef-
fects model. Subgroup and sensitivity analysis were con-
ducted to assess the stability and heterogeneity of the
pooled results. The risk of publication bias was evaluated
by assessing for funnel plot asymmetry and by Egger’s test
and Begg’s test. A total of 19 articles met the eligibility
criteria for the meta-analysis, reporting on 7055 subjects.
Electronic supplementary material The online version of this
article (doi:10.1007/s00535-017-1371-9) contains supplementary
material, which is available to authorized users.
& Allen Lee
allenlee@med.umich.edu
& Shujie Chen
4094108@qq.com
1
Department of Gastroenterology, Sir Run Run Shaw
Hospital, Zhejiang University School of Medicine,
Hangzhou 310016, Zhejiang, China
2 lating evidence suggests that PPIs have long-term risks
Institute of Gastroenterology, Zhejiang University,
Hangzhou 310016, Zhejiang, China
3 small intestinal bacterial overgrowth (SIBO), which is
Division of Gastroenterology and Hepatology, Department of
Medicine, University of Michigan Hospital, Ann Arbor, MI,
USA
The pooled odds ratio (OR) showed a statistically signifi-
cant association between increased risk of SIBO and PPI
use (OR 1.71, 95% confidence interval 1.20–2.43). Sub-
group analyses demonstrated an association between SIBO
and PPI use in studies that employed small bowel aspirates
culture and glucose hydrogen breath tests (GHBT) as
diagnostic tests for SIBO. Our meta-analysis suggests that
the use of PPI moderately increases the risk of SIBO,
thereby highlighting the need for appropriate prescribing of
PPIs.
Keywords Small intestinal bacterial overgrowth
Proton
pump inhibitors
Irritable bowel syndrome
Introduction
Proton pump inhibitors (PPIs) have become the mainstay of
medical treatments for a number of conditions, including
gastroesophageal reflux disease (GERD), peptic ulcer dis-
ease, and Helicobacter pylori infection [1]. Due to their
efficacy and overall safety profile, PPIs have become one
of the most commonly prescribed classes of medications
[2]. Indeed, one study demonstrated that the use of PPIs in
Australia increased by 1318% from 1995 to 2006, partic-
ularly in elderly patients [3]. Other studies have consis-
tently indicated that PPIs are overprescribed in both the
inpatient and outpatient settings, with 25–70% of patients
having no indication to be on a PPI [4–6].
Although PPIs are generally well tolerated, accumu-
[7]. One potential risk of PPI use is the development of
defined as [105 bacteria colony-forming units per milli-
liter upon culturing upper gut aspirates. SIBO is
123

associated with a series of clinical symptoms, including
bloating, abdominal pain, diarrhea, nutrient malabsorption
and weight loss [8, 9]. Since gastric acid is an important
barrier that prevents bacterial colonization of the stomach
and small intestine, PPI therapy may promote the growth
of small intestinal microbiota through chronic acid sup-
pression and subsequent hypochlorhydria [10]. However,
while several studies have demonstrated a relationship
between the development of SIBO and intake of PPIs,
other studies have published conflicting results [11–29]. A
meta-analysis which reviewed 11 observational studies
published before 2013 reported a statistically significant
increased risk of SIBO among PPI users [odds ratio (OR)
2.282, 95% confidence interval (CI) 1.238–4.205] [30].
However, the meta-analysis was limited by significant
publication bias, and the adjusted OR value was no longer
statistically significant after potential publication bias was
accounted for using the Duval and Tweedie trim-and-fill
method [30]. Consequently, the relationship between PPI
use and risk of SIBO remains unanswered and contro-
versial. Several high-quality cohort and case–control
studies have recently been published in an attempt to
clarify the risk of SIBO associated with PPI use [14–20].
The aim of this meta-analysis is to review available
published peer-reviewed evidence in order to evaluate the
risk of SIBO among PPI users.
Methods
Search strategy
This meta-analysis was conducted according to the
PRISMA (preferred reporting items for systematic reviews
and meta-analysis) guidelines [31]. The search for relevant
articles published before November 2016 was performed
using the online PubMed, Embase, and Cochrane Library
databases and Web of Science. No restrictions with regards
to the year of publication were applied. The keywords
‘‘proton pump inhibitor,’’ ‘‘omeprazole’’, ‘‘lansoprazole,’’
‘‘pantoprazole,’’ ‘‘esomeprazole,’’ ‘‘rabeprazole,’’
‘‘dexlansoprazole,’’ ‘‘bacterial overgrowth’’, and ‘‘SIBO’’
were used as search terms. Our meta-analysis was regis-
tered in PROSPERO with registration number
CRD42016051069.
Selection of eligible studies
Two of the authors (STT and LSC) screened for eligible
studies independently of each other using a two-step
strategy. First, studies that were obviously unsuitable were
excluded by screening of the title and abstract. Second, a
full-text review was performed. Eligible studies were
123
J Gastroenterol
included in this meta-analysis if they met the following
criteria: (1) the original study evaluated the association
between the diagnosis of SIBO and the intake of PPI; (2)
the study population included PPI users and a non-PPI
control group of adults from the general population
([18 years); (3) studies clearly reported the incidence of
SIBO diagnosed by one of the clinical tests, including
small bowel aspirate culture or breath tests. There was no
language limitation. Studies where only abstracts were
available were not included.
Data extraction and quality assessment of selected
studies
All data extraction and quality assessment of the included
studies were performed by two of the authors (STT and
LSC) independently of each other, and any discrepancies
were resolved by discussion with the other authors. The
following data were collected from eligible studies:
authors; publication year; country; study design; mean age
of subjects; patient population; type, dose and duration of
PPI therapy; the type of test used to diagnose SIBO; cut-off
values of breath testing; total number of subjects with and
without SIBO among both PPI and non-PPI users. The
Newcastle–Ottawa Scale was applied for quality assess-
ment in observational studies [32].
Data synthesis and statistical analysis
As the outcome parameters were presented as dichotomous
values (PPI users vs non-PPI users), pooled ORs were
chosen for the meta-analysis. Considering the variability of
the design, testing modality, and types of control groups
among the eligible studies, we chose a generic inverse-
variance random effects model for the pooled analysis.
Since most of the eligible studies did not provide an
adjusted ratio, the OR was calculated from the raw data.
The OR was considered to be statistically significant if the
95% CI did not span a value of 1 with a p value of \0.05.
The heterogeneity across pooled studies was quantified
using the Cochran Q test and I2 statistics [33]. Hetero-
geneity was considered to be statistically significant if
p was B0.1 and/or I2 was [50%. Subgroup analyses were
performed to investigate sources of heterogeneity by
grouping eligible studies according to study design, testing
modality, control type, and study quality. We also con-
ducted a sensitivity analysis in which one study was
removed each time and the analysis re-run to estimate the
effect of single large studies. A cumulative analysis was
performed to exhibit time trends by publication year.
Potential publication bias was assessed by a funnel plot,
Egger’s test, and Begg’s test (p \ 0.05 was considered to
be significant). All statistical analyses were performed
J Gastroenterol
using Stata software version 13.0 (StataCorp LP, College
Station, TX)
Results
Study selection
A total of 761 initial studies were identified by the search
strategy, of which 36 were selected for full-text review
after screening of the title and abstract. Among these 36
studies, 13 for which only the abstract was available were
not included. Four additional studies were excluded,
including two studies [34, 35] due to unconventional
methods of defining SIBO, one study for not including non-
PPI users as a control group [36], and a fourth study [37]
due to an overlapping study population with a subsequent
publication [15]. Ultimately, 19 original studies met the
inclusion and exclusion criteria and were subsequently
analyzed (Fig. 1).
Characteristics and quality of eligible studies
The characteristics of the included studies are shown in
Table 1. These studies cumulatively included 2175 cases of
Fig. 1 Flow diagram summarizing study identification and selection.
SIBO Small intestinal bacterial overgrowth
SIBO in 7055 subjects. Ten studies were case–control
studies, while the remaining nine were cohort studies.
Sixteen studies used an independent non-PPI user popula-
tion as control group, whereas the other three studies
compared the incidence of SIBO in the same cohort before
and after PPI exposure. SIBO was diagnosed in eight
studies using small bowel aspirate culture, in seven studies
using the glucose hydrogen breath test (GHBT), in three
studies using the lactulose hydrogen breath test (LHBT),
and in one study using the D-xylose breath test. The cut-off
value for the diagnosis of SIBO using small bowel aspirate
culture was 105 CFU/mL while the cut-off thresholds for
the various breath tests varied among studies. The qualities
of eligible studies were moderate or high, based on
assessment by the Newcastle–Ottawa scale tool, indicating
they were of acceptable methodological quality for inclu-
sion in the meta-analysis [Electronic Supplementary
Material (ESM) Tables S1a, S1b).
The risk of SIBO and the intake of PPIs
Among the 19 eligible studies, seven found a positive
association between PPI intake and increased risk for
SIBO, whereas the remaining 12 studies did not show a
statistically significant relationship. The pooled estimate of
the 19 studies demonstrated that PPI use was significantly
associated with a moderately increased risk of SIBO (OR
1.71, 95% CI 1.20–2.43) (Fig. 2).
There was considerable heterogeneity (I2 = 84.1%,
p = 0.000) among studies, as anticipated (Fig. 2). To
explore the sources of the heterogeneity, we performed
subgroup analyses by grouping studies by study design,
type of control group, testing modality, and study quality
(Table 2). The subgroup analysis revealed that study
quality was an important source of heterogeneity. When the
pooled analysis was limited to high-quality studies, statis-
tically significant increased risk of SIBO among PPI users
was found without significant heterogeneity (OR 1.31, 95%
CI 1.01, 1.69, I2 = 27.8%, p = 0.172). Study design,
testing modalities, and type of control group had fewer
effects on heterogeneity. The significant association
between increased risk of SIBO and PPI therapy persisted
after stratification by study design or by control type
(Table 2). Analysis of the effect of SIBO testing modality
revealed a significantly increased risk of SIBO among PPI
users in both the GHBT subgroup (OR 1.84, 95% CI
1.03–3.30) and aspirate culture subgroup (OR 2.22, 95% CI
1.33–3.68).
A sensitivity analysis removing one study at a time
confirmed the stability of our results (Fig. 3a). Three
studies [11, 13, 22] showed a very high OR that was not in
line with the OR of the other studies (Fig. 1). When these
three studies were omitted, the positive association
123
 J Gastroenterol

Table 1 Characteristics of included studies

Reference Year Country Study Control type SIBO diagnostic Mean age Total subjects SIBO cases
design test (years) (n) (n)

Enko et al. [16] 2016 Austria Case– Separate GHBT 44.0 109 35
control control
Liang et al. [14] 2016 China Case– Separate GHBT 50.7 200 125
control control
Giamarellos et al. 2016 Greece Case– Separate Aspirate culture 64.4 897 95
[15] control control
Schatz et al. [17] 2015 USA Case– Separate XBT 57.4 932 513
control control
Franco et al. [18] 2015 USA Case– Separate Aspirate culture 52.0 1260 384
control control
Senderovky et al. 2014 Argentina Case– Separate LHBT 57.5 225 69
[19] control control
Jacobs et al. [20] 2013 USA Case– Separate Aspirate culture 43.0 150 20
control control
Ratuapli et al. [21] 2012 USA Case– Separate GHBT 60.9 1191 257
control control
Compare et al. [22] 2011 Italy Cohort Self-control GHBT 36.0 42 11
Choung et al. [23] 2011 USA Case– Separate Aspirate culture 53.0 675 54
control control
Lombardo et al. 2010 Italy Cohort Separate GHBT 37.7 450 152
[24] control
Law et al. [25] 2009 USA Cohort Separate LHBT 43.7 555 302
control
Majewski et al. [26] 2007 USA Case– Separate GHBT 46.4 204 93
control control
Pereira et al. [27] 1998 UK Cohort Separate Aspirate culture 76.0 14 6
control
Hutchinson et al. 1997 UK Cohort Separate GHBT 78.6 44 23
[28] control
Thorens et al. [29] 1996 Swizerland Cohort Separate Aspirate culture 42.0 37 13
control
Lewis et al. [13] 1996 South Cohort Self-control Aspirate culture 43.1 20 7
Africa
Gough et al. [12] 1995 UK Cohort Self-control LHBT 55.0 10 2
Fried et al. [11] 1994 Switzerland Cohort Separate Aspirate culture 50.6 40 14
control
SIBO Small intestinal bacterial overgrowth, GHBT glucose hydrogen breath test, LHBT lactulose hydrogen breath test, XBT D-xylose breath test,

between the risk of SIBO and PPI therapy did not change
significantly (OR 1.52, 95% CI 1.08–2.15) (ESM Fig. S1).
A cumulative analysis was also performed in which studies
were added over time, which showed a persistent and
significantly increased risk of SIBO with PPI use. Fur-
thermore, the OR value tended to be robust in studied
performed after 2014 (Fig. 3b). No statistically significant
publication bias was found based on the Egger’s test
(p = 0.149) or Begg’s test (p = 0.115); however, visual
inspection of the funnel plot suggested an asymmetry to the
plot (Fig. 3c).
Discussion
In this meta-analysis of observational studies, we have
shown that PPI therapy is associated with a moderately
increased risk of SIBO. The sensitivity analysis supports
the stability of our results. As expected, there is statistically
significant heterogeneity across the studies included in our
meta-analysis, and we have demonstrated that study quality
is an important source of this heterogeneity. When the
pooled analysis was limited to high-quality studies, a sta-
tistically significant increased risk of SIBO among PPI

123
J Gastroenterol

Fig. 2 Meta-analysis of risk for

SIBO with proton pump
inhibitor use. OR Odds ratio, CI
confidence interval

Table 2 Subgroup analyses of Subgroup analysis Number of studies Pooled OR (95% CI) Heterogeneity I2 (%) p
risk for small intestinal bacterial
overgrowth with proton pump Study design
inhibitor use
Case–control 10 1.43 (1.00, 2.03) 84.6 0.000
Cohort 9 3.83 (1.41, 0.96) 84.6 0.000
Overall 19 1.71 (1.20, 2.43) 84.1 0.000
Control type
Self-control 3 17.31 (3.09, 97.02) 0 0.741
Separate control 16 1.57 (1.10, 2.22) 85.7 0.000
Overall 19 1.71 (1.20, 2.43) 84.1 0.000
SIBO diagnostic test
Aspirate culture 8 2.22 (1.33, 3.68) 64.0 0.007
GHBT 7 1.84 (1.03, 3.30) 84.1 0.000
LHBT 3 0.99 (0.50, 1.96) 59.7 0.084
XBT 1 0.74 (0.57, 0.96) – –
Overall 19 1.71 (1.20, 2.43) 84.1 0.000
Study quality
C6 12 1.31 (1.01, 1.69) 27.8 0.172
\6 7 2.20 (1.10, 4.40) 93.6 0.000
Overall 19 1.71 (1.20, 2.43) 84.1 0.000
OR Odds ratio, CI confidential interval, PPI proton pump inhibitor

users was found without significant heterogeneity (OR
1.31, 95% CI 1.01, 1.69, I2 = 27.8%, p = 0.172).
Although the funnel plot appears to be asymmetrical
upon visual inspection, Egger’s test and Begg’s test
showed no statistically significant public bias. While funnel
plot asymmetry has long been equated with publication
bias, the funnel plot should be seen as a generic means of
displaying small-study effects—i.e., a tendency for the
intervention effects estimated in smaller studies to differ
from those estimated in larger studies [38]. Indeed, as the
funnel plot showed in our research (Fig. 3c), all five out-
lying studies in this meta-analysis were smaller studies.

123
 J Gastroenterol

Fig. 3 a One-study removed

analysis of SIBO risk to assess
the single large effect.
b Cumulative analysis of SIBO
risk to assess the consistency of
results over time. c Funnel pot
of standard error by SIBO risk

123
J Gastroenterol
According to the Cochrane Handbook for Systematic
Review of Interventions, small-study effects other than
publication bias might contribute to the asymmetry of our
funnel plot. Smaller studies tend to be conducted and
analyzed with less methodological rigor than larger studies,
which may lead to spuriously inflated effects [39]. When
the three smaller studies [11, 13, 22] with markedly high
OR values were omitted from the pooled estimate, the
positive association between the risk of SIBO and PPI
therapy did not change much (OR 1.52, 95% CI 1.08–2.15)
(ESM Fig. S1), indicating that these smaller studies were
not highly influential in the meta-analysis. In the previous
meta-analysis on SIBO and PPI use [30], the funnel plot
showed four outlying studies at the right margin of the plot.
The Duval and Tweedie trim-and-fill analysis was then
performed in that study to identify and correct for funnel
plot asymmetry arising from publication bias. The result
showed four studies being imputed with an adjusted OR of
1.44 (95% CI 0.778–2.667), suggesting no significant
association between risk of SIBO and PPI intake after
accounting for potential publication bias. However,
according to the Cochrane Handbook for Systematic
Review of Interventions (http://training.cochrane.org/hand
book), this method does not take into account reasons for
funnel plot asymmetry other than publication bias. Fur-
thermore, the trim-and-fill method is known to perform
poorly in the presence of substantial between-study
heterogeneity [40, 41]. Due to the high heterogeneity
across the included studies in our meta-analysis, we did not
perform the trim-and-fill method.
Our meta-analysis includes seven new observational
studies [14–20] that were published which were not
included in the previous meta-analysis [30]. Although our
meta-analysis shows similar risks of SIBO in PPI users as
reported in previous studies, there are important new
findings in our study. First, we demonstrated an increased
risk of SIBO in PPI users that persisted in the subgroup
analyses when studies were included that compare the
incidence of SIBO before and after treatment with PPIs, as
well as PPI users and an independent non-PPI control
group. In the prior meta-analysis, a positive association
was only shown when four studies were analyzed
[11, 22, 27, 42] that compared the incidence of SIBO
before and after PPI therapy. These four self-control
studies were all small studies with OR values that were
markedly higher than those in the remainder of the inclu-
ded studies (a total of 126 subjects were included in the
four studies, which only accounted for 4% of the overall
subjects included in the meta-analysis). When these four
studies were removed from the analysis, the association
between SIBO and PPI use was no longer statistically
significant, which suggests that the overall result was
instable. Second, the subgroup analysis performed in our
meta-analysis exhibited a clear association between the
increased risk of SIBO and PPI therapy in both the aspirate
culture subgroup and the GHBT subgroup. In the previous
meta-analysis, the increase in risk for SIBO was only noted
in those studies using small bowel aspirate culture and not
in those using the GHBT. Only four studies using the
GHBT as diagnostic modality were included in the previ-
ous meta-analysis, and the lack of a positive association
between SIBO and PPI therapy in the pooled analysis of
these four studies using GHBT may be due to the inade-
quate number of studies.
Bacterial counts exceeding 105 CFU per milliliter of
small bowel aspirate is considered the gold standard for the
diagnosis of SIBO [43]. However, due to the invasiveness
of the test and requirement for an upper endoscopy, non-
invasive breath tests with various substrates (glucose, lac-
tulose, D-xylose, and C-glycocholate) have been developed
as indirect diagnostic methods. The most frequently used
substrates in hydrogen breath tests for diagnosis of SIBO
are glucose and lactulose [44]. Our meta-analysis showed a
positive association between the risk of SIBO and PPI
among studies using GHBT, which was consistent with the
pooled analysis of studies using small bowel aspirate.
However, in the studies using LHBT as diagnostic
approach, there was no significant association between
SIBO and PPI therapy. Compared with the GHBT, the
LHBT exhibits a limited diagnostic accuracy
[43–45].Studies performing the LHBT and small bowel/
scintigraphy simultaneously have shown that the LHBT is
a measure of small bowel transit as opposed to small bowel
bacterial burden [46]. Therefore, the absence of a clear
association between SIBO and PPI therapy in the pooled
analysis of studies using LHBT as the diagnostic method
might be attributed to misclassification of SIBO diagnosis
due to the low sensitivity and specificity of the LHBT.
However, one should be careful in drawing this ‘‘negative’’
conclusion since only a few published studies have used
the LHBT. Furthermore, the lack of unified cut-off
thresholds for breath tests is one of the essential factors
contributing to the inconsistent findings and heterogeneity
across studies. Thus, standardization of the cut-off value
for the breath test is urgently needed to diagnose SIBO.
SIBO is believed to play a vital role in irritable bowel
syndrome (IBS), and as such it is considered to be a
potential therapeutic target among IBS patients [47–49].
However, Spiegel et al. [50] recently proposed a ‘‘PPI
hypothesis’’ in which the linkage between IBS and SIBO is
more of an epiphenomenon due to the chronic intake of
PPIs. Many studies indicate a clear overlap between IBS
and GERD [51, 52], and given the high degree of this
overlap between these two diseases, IBS patients may be
more likely to take PPIs than a control group. Hence, the
increased risk of SIBO among PPI users highlights the
123

effect of PPI therapy in studies linking SIBO and IBS.
However, most studies have failed to include PPI therapy
as a potential predictor of SIBO in IBS. Ford et al. [53]
performed a meta-analysis of 12 studies and reported that
the pooled OR for any positive test for SIBO was three- to
fivefold greater in IBS than in controls. Among these 12
studies, only two reported data on the rates of PPI use in
the involved subjects [26, 54]. Since the publication of
Ford et al.’s meta-analysis [53], least eight observational
studies have been published on the frequency of SIBO
among patients with IBS [15, 24, 45, 47, 48, 55–57]; of
these, four studies [24, 48, 55, 57] excluded subjects with
PPI consumption, and one study [15] showed that PPI
intake is not an independent predictor for SIBO. Interest-
ingly, all of the studies taking the factor of PPIs into
consideration demonstrated a positive association between
SIBO and IBS. Taken together, these studies suggest that
the search for predictor(s) of PPIs is attracting increasing
interest. However, more high-quality original studies are
still needed to answer the question of whether the linkage
between IBS and SIBO is independent of PPI therapy.
Furthermore, future studies should examine whether the
intake of PPIs is related to the symptom(s) of SIBO and
whether PPI therapy has a negative effect on the symptoms
of IBS.
Various adverse effects of PPI use have been reported
recently. PPIs are frequently prescribed together with non-
steroidal anti-inflammatory drugs (NSAIDs) to minimize
NSAID-related adverse effects in the upper gastrointestinal
tract. However, recent studies indicate that PPIs increase
the incidence of NSAID-induced small bowel injury
[58, 59]. In addition, PPIs have been associated with
medication change and increased risk of hospitaliza-
tion/surgery in patients with inflammatory bowel disease
(IBD) [60, 61]. These adverse effects may partly be
attributed to changes in gut microbiota and SIBO induced
by PPI therapy [61–63]. Interestingly, SIBO has also been
linked to the two diseases mentioned above in several
clinical studies. The results of one retrospective study
indicate that SIBO was significantly associated with an
increased OR (6.54; 95% CI 40–30.50, p = 0.02) for
NSAID-induced small intestinal damage after adjusting for
the use of PPI [64]. However, this study was limited by a
small study population. Several other studies have reported
that the rate of SIBO was significantly higher in their IBD
patients than in the controls [65–67]. However, all these
studies did not take the effect of PPI intake into consid-
eration. Whether SIBO is an independent risk factor for the
development of NSAID-induced small bowel injury and
IBD, or just an epiphenomenon, is still unknown. Future
studies should be designed to clarify the relationship
between PPI therapy, SIBO, and NSAID-induced small
bowel injury or IBD.
123
J Gastroenterol
Our meta-analysis has several limitations. These include
not determining whether dose, duration and type of PPI
exposure had an effect on the risk of developing SIBO.
These factors were not examined due to a lack of data in
previous studies. Despite the cumulative analysis which
showed that in terms of temporal progress the pooled OR
value tended to be stable after 2014, more high-quality
evidence is still required to clarify the effect of different
dose, duration and type of PPIs on the diagnosis of SIBO,
which is of great significance to guide mediation use.
In conclusion, this meta-analysis of 19 eligible studies
indicates that PPI therapy is associated with a moderately
increased risk of SIBO. These findings persisted even on
subgroup analysis when studies were examined using dif-
ferent control populations as well as different modalities to
diagnose SIBO, such as culture of small bowel aspirates
and GHBT. These findings highlight the need for careful
and judicious prescription of PPIs.
Acknowledgments This work was supported by the Zhejiang Pro-
vince Key Science and Technology Innovation Team (2013TD13),
the National Natural Science Foundation of China (81472214), and
the Zhejiang Provincial Medical and Health Research Plan
(2015126452).
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
References
1. Corleto VD, Festa S, Di Giulio E, et al. Proton pump inhibitor
therapy and potential long-term harm. Curr Opin Endocrinol
Diabetes Obes. 2014;21(1):3–8.
2. Bartholow M. Top 200 prescription drugs of 2009. Pharmacy
Times. 2010. http://www.pharmacytimes.com/publications/issue/
2010/may2010/rxfocustopdrugs-0510.
3. Hollingworth S, Duncan EL, Martin JH. Marked increase in
proton pump inhibitors use in Australia. Pharmacoepidemiol
Drug Saf. 2010;19(10):1019–24.
4. Batuwitage BT, Kingham JG, Morgan NE, et al. Inappropriate
prescribing of proton pump inhibitors in primary care. Postgrad
Med J. 2007;83(975):66–8.
5. Ramirez E, Lei SH, Borobia AM, et al. Overuse of PPIs in patients
at admission, during treatment, and at discharge in a tertiary
Spanish hospital. Curr Clin Pharmacol. 2010;5(4):288–97.
6. Forgacs I, Loganayagam A. Overprescribing proton pump inhi-
bitors. BMJ. 2008;336(7634):2–3.
7. Schnoll-Sussman F, Katz PO. Clinical implications of emerging
data on the safety of proton pump inhibitors. Curr Treat Options
Gastroenterol. 2017;15(1):1–9.
8. Posserud I, Stotzer PO, Bjornsson ES, et al. Small intestinal
bacterial overgrowth in patients with irritable bowel syndrome.
Gut. 2007;56(6):802–8.
9. King T. Small intestinal bacterial overgrowth and irritable bowel
syndrome. JAMA. 2004;292(18):2213 (author reply 2213–4).
10. Laine L, Ahnen D, McClain C, et al. Review article: potential
gastrointestinal effects of long-term acid suppression with proton
pump inhibitors. Aliment Pharmacol Ther. 2000;14(6):651–68.
J Gastroenterol
11. Fried M, Siegrist H, Frei R, et al. Duodenal bacterial overgrowth
during treatment in outpatients with omeprazole. Gut.
1994;35(1):23–6.
12. Gough A, Andrews D, Bacon PA, et al. Evidence of omeprazole-
induced small bowel bacterial overgrowth in patients with scle-
roderma. Br J Rheumatol. 1995;34(10):976–7.
13. Lewis SJ, Franco S, Young G, et al. Altered bowel function and
duodenal bacterial overgrowth in patients treated with omepra-
zole. Aliment Pharmacol Ther. 1996;10(4):557–61.
14. Liang SF, Xu L, Zhang DS, et al. Effect of probiotics on small
intestinal bacterial overgrowth in patients with gastric and col-
orectal cancer. Turk J Gastroenterol. 2016;27(3):227–32.
15. Giamarellos-Bourboulis EJ, Pyleris E, Barbatzas C, et al. Small
intestinal bacterial overgrowth is associated with irritable bowel
syndrome and is independent of proton pump inhibitor usage.
BMC Gastroenterol. 2016;16(1):67–73
16. Enko D, Kriegshauser G. Functional 13C-urea and glucose
hydrogen/methane breath tests reveal significant association of
small intestinal bacterial overgrowth in individuals with active
Helicobacter pylori infection. Clin Biochem. 2016;50(1-2):46–9.
17. Schatz RA, Zhang Q, Lodhia N, et al. Predisposing factors for
positive D-Xylose breath test for evaluation of small intestinal
bacterial overgrowth: a retrospective study of 932 patients. World
J Gastroenterol. 2015;21(15):4574–82.
18. Franco DL, Disbrow MB, Kahn A, et al. Duodenal aspirates for
small intestine bacterial overgrowth: yield, PPIs, and outcomes
after treatment at a Tertiary Academic Medical Center. Gas-
troenterol Res Pract. 2015;2015:971582.
19. Senderovky M, Lasa J, Dima G, et al. Influence of proton pump
inhibitors on intestinal fermentative profile: a case–control study.
Acta Gastroenterol Latinoam. 2014;44(3):205–9.
20. Jacobs C, Coss Adame E, Attaluri A, et al. Dysmotility and
proton pump inhibitor use are independent risk factors for small
intestinal bacterial and/or fungal overgrowth. Aliment Pharmacol
Ther. 2013;37(11):1103–11.
21. Ratuapli SK, Ellington TG, O’Neill MT, et al. Proton pump
inhibitor therapy use does not predispose to small intestinal
bacterial overgrowth. Am J Gastroenterol. 2012;107(5):730–5.
22. Compare D, Pica L, Rocco A, et al. Effects of long-term PPI
treatment on producing bowel symptoms and SIBO. Eur J Clin
Invest. 2011;41(4):380–6.
23. Choung RS, Ruff KC, Malhotra A, et al. Clinical predictors of
small intestinal bacterial overgrowth by duodenal aspirate cul-
ture. Aliment Pharmacol Ther. 2011;33(9):1059–67.
24. Lombardo L, Foti M, Ruggia O, et al. Increased incidence of
small intestinal bacterial overgrowth during proton pump inhi-
bitor therapy. Clin Gastroenterol Hepatol. 2010;8(6):504–8.
25. Law D, Pimentel M. Proton pump inhibitor therapy does not
affect hydrogen production on lactulose breath test in subjects
with IBS. Dig Dis Sci. 2010;55(8):2302–8.
26. Majewski M, McCallum RW. Results of small intestinal bacterial
overgrowth testing in irritable bowel syndrome patients: clinical
profiles and effects of antibiotic trial. Adv Med Sci.
2007;52:139–42.
27. Pereira SP, Gainsborough N, Dowling RH. Drug-induced
hypochlorhydria causes high duodenal bacterial counts in the
elderly. Aliment Pharmacol Ther. 1998;12(1):99–104.
28. Hutchinson S, Logan R. The effect of long-term omeprazole on
the glucose-hydrogen breath test in elderly patients. Age Ageing.
1997;26(2):87–9.
29. Thorens J, Froehlich F, Schwizer W, et al. Bacterial overgrowth
during treatment with omeprazole compared with cimetidine: a
prospective randomised double blind study. Gut. 1996;39(1):54–9.
30. Lo WK, Chan WW. Proton pump inhibitor use and the risk of
small intestinal bacterial overgrowth: a meta-analysis. Clin
Gastroenterol Hepatol. 2013;11(5):483–90.
31. Stewart LA, Clarke M, Rovers M, et al. Preferred reporting items
for systematic review and meta-analyses of individual participant
data: the PRISMA-IPD statement. JAMA.
2015;313(16):1657–65.
32. Wells GA, Shea B, O’Connell D et al. The Newcastle–Ottawa
Scale (NOS) for assessing the quality of nonrandomised studies
in meta-analyses. Ottawa Hospital Research Institute, Ottawa
33. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-
analysis. Stat Med. 2002;21(11):1539–58.
34. Tsuda A, Suda W, Morita H, et al. Influence of proton-pump
inhibitors on the luminal microbiota in the gastrointestinal tract.
Clin Transl Gastroenterol. 2015;6:e89.
35. Freedberg DE, Toussaint NC, Chen SP, et al. Proton pump
inhibitors alter specific taxa in the human gastrointestinal
microbiome: a crossover trial. Gastroenterology.
2015;149(4):883–5.e9.
36. Ardatskaia MD, Loginov VA, Minushkin ON. Syndrome of
bacterial overgrowth in patients with the reduced stomach acid
secretion: some aspects of the diagnosis. Eksp Klin Gastroenterol.
2014;12:30–6.
37. Pyleris E, Giamarellos-Bourboulis EJ, Tzivras D, et al. The
prevalence of overgrowth by aerobic bacteria in the small intes-
tine by small bowel culture: relationship with irritable bowel
syndrome. Dig Dis Sci. 2012;57(5):1321–9.
38. Sterne JA, Gavaghan D, Egger M. Publication and related bias in
meta-analysis: power of statistical tests and prevalence in the
literature. J Clin Epidemiol. 2000;53(11):1119–29.
39. Egger M, Juni P, Bartlett C, et al. How important are compre-
hensive literature searches and the assessment of trial quality in
systematic reviews? Empirical study. Health Technol Assess.
2003;7(1):1–76.
40. Terrin N, Schmid CH, Lau J, et al. Adjusting for publication bias
in the presence of heterogeneity. Stat Med. 2003;22(13):2113–26.
41. Peters JL, Sutton AJ, Jones DR, et al. Performance of the trim and
fill method in the presence of publication bias and between-study
heterogeneity. Stat Med. 2007;26(25):4544–62.
42. Nelis GF, Engelage AH, Samson G. Does long-term inhibition of
gastric acid secretion with omeprazole lead to small intestinal
bacterial overgrowth? Neth J Med. 1994;45(3):93–100.
43. Hawkey CJ, Richter JE, et al., editors. Textbook of clinical
gastroenterology and Hepatology. Hoboken: Wiley; 2012.
44. Gasbarrini A, Corazza GR, Gasbarrini G, et al. Methodology and
indications of H2-breath testing in gastrointestinal diseases: the
Rome consensus conference. Aliment Pharmacol Ther.
2009;29(Suppl 1):1–49.
45. Rana SV, Sharma S, Kaur J, et al. Comparison of lactulose and
glucose breath test for diagnosis of small intestinal bacterial
overgrowth in patients with irritable bowel syndrome. Digestion.
2012;85(3):243–7.
46. Yu D, Cheeseman F, Vanner S. Combined oro-caecal scintigra-
phy and lactulose hydrogen breath testing demonstrate that breath
testing detects oro-caecal transit, not small intestinal bacterial
overgrowth in patients with IBS. Gut. 2011;60(3):334–40.
47. Chu H, Fox M, Zheng X, et al. Small intestinal bacterial over-
growth in patients with irritable bowel syndrome: clinical char-
acteristics, psychological factors, and peripheral cytokines.
Gastroenterol Res Pract. 2016;2016:3230859.
48. Abbasi MH, Zahedi M, Moghadam SD, et al. Small bowel bac-
terial overgrowth in patients with irritable bowel syndrome: the
first study in iran. Middle East. J Dig Dis. 2015;7(1):36–40.
49. Moraru IG, Moraru AG, Andrei M, et al. Small intestinal bac-
terial overgrowth is associated to symptoms in irritable bowel
syndrome. Evidence from a multicentre study in Romania. Rom J
Intern Med. 2014;52(3):143–50.
50. Spiegel BM, Chey WD, Chang L. Bacterial overgrowth and
irritable bowel syndrome: unifying hypothesis or a spurious
123

consequence of proton pump inhibitors? Am J Gastroenterol.
2008;103(12):2972–6.
51. Talley NJ, Dennis EH, Schettler-Duncan VA, et al. Overlapping
upper and lower gastrointestinal symptoms in irritable bowel
syndrome patients with constipation or diarrhea. Am J Gas-
troenterol. 2003;98(11):2454–9.
52. Locke GR 3rd, Zinsmeister AR, Fett SL, et al. Overlap of gas-
trointestinal symptom complexes in a US community. Neuro-
gastroenterol Motil. 2005;17(1):29–34.
53. Ford AC, Spiegel BM, Talley NJ, et al. Small intestinal bacterial
overgrowth in irritable bowel syndrome: systematic review and
meta-analysis. Clin Gastroenterol Hepatol. 2009;7(12):1279–86.
54. Nucera G, Gabrielli M, Lupascu A, et al. Abnormal breath tests to
lactose, fructose and sorbitol in irritable bowel syndrome may be
explained by small intestinal bacterial overgrowth. Aliment
Pharmacol Ther. 2005;21(11):1391–5.
55. Yakoob J, Abbas Z, Khan R, et al. Small intestinal bacterial
overgrowth and lactose intolerance contribute to irritable bowel
syndrome symptomatology in Pakistan. Saudi J Gastroenterol.
2011;17(6):371–5.
56. Sachdeva S, Rawat AK, Reddy RS, et al. Small intestinal bac-
terial overgrowth (SIBO) in irritable bowel syndrome: frequency
and predictors. J Gastroenterol Hepatol. 2011;26[Suppl 3]:135–8.
57. Ghoshal UC, Kumar S, Mehrotra M, et al. Frequency of small
intestinal bacterial overgrowth in patients with irritable bowel
syndrome and chronic non-specific diarrhea. J Neurogastroenterol
Motil. 2010;16(1):40–6.
58. Washio E, Esaki M, Maehata Y, et al. Proton pump inhibitors
increase incidence of nonsteroidal anti-inflammatory drug-in-
duced small bowel injury: a randomized placebo-controlled trial.
Clin Gastroenterol Hepatol. 2016;14(6):809–15.
123
J Gastroenterol
59. Watanabe T, Tanigawa T, Nadatani Y, et al. Risk factors for
severe nonsteroidal anti-inflammatory drug-induced small
intestinal damage. Dig Liver Dis. 2013;45(5):390–5.
60. Juillerat P, Schneeweiss S, Cook EF, et al. Drugs that inhibit
gastric acid secretion may alter the course of inflammatory bowel
disease. Aliment Pharmacol Ther. 2012;36(3):239–47.
61. Shah R, Richardson P, Yu H, et al. Gastric acid suppression is
associated with an increased risk of adverse outcomes in
inflammatory bowel disease. Digestion. 2017;95(3):188–93.
62. Otani K, Tanigawa T, Watanabe T, et al. Microbiota plays a key
role in non-steroidal anti-inflammatory drug-induced small
intestinal damage. Digestion. 2017;95(1):22–8.
63. Wallace JL, Syer S, Denou E, et al. Proton pump inhibitors
exacerbate NSAID-induced small intestinal injury by inducing
dysbiosis. Gastroenterology. 2011;141(4):1314–22, 1322 e1–5.
64. Muraki M, Fujiwara Y, Machida H, et al. Role of small intestinal
bacterial overgrowth in severe small intestinal damage in chronic
non-steroidal anti-inflammatory drug users. Scand J Gastroen-
terol. 2014;49(3):267–73.
65. Castiglione F, Del Blanco GDV, Rispo A, et al. Orocecal transit
time and bacterial overgrowth in patients with Crohn’s disease.
J Clin Gastroenterol. 2000;31(1):63–6.
66. Rana SV, Sharma S, Malik A, et al. Small intestinal bacterial
overgrowth and orocecal transit time in patients of inflammatory
bowel disease. Dig Dis Sci. 2013;58(9):2594–8.
67. Greco A, Caviglia GP, Brignolo P, et al. Glucose breath test and
Crohn’s disease: diagnosis of small intestinal bacterial over-
growth and evaluation of therapeutic response. Scand J Gas-
troenterol. 2015;50(11):1376–81.