INTRODUCTION
Small-intestinal bacterial overgrowth (SIBO) is being increasingly acknowledged as a clinically relevant condition in gastrointestinal and systemic disorders, with particular implications for patients with Crohn’s disease (CD).1 In CD, dysbiosis and structural alterations of the small intestine, particularly strictures, prolong intestinal transit and alter the intraluminal environment, predisposing patients to bacterial overgrowth and promoting symptoms such as bloating, diarrhea, and malnutrition.2
The diagnostic approaches for SIBO frequently rely on hydrogen–methane breath testing, for which the North American Consensus provides standardized criteria (≥20 ppm increase in hydrogen or ≥10 ppm increase in methane by 90 minutes).1 However, breath testing has known limitations, including variable sensitivity and potential false negatives when overgrowth is localized distally due to proximal glucose absorption.1,2 The utility of glucose breath testing for detecting CD-related strictures and secondary SIBO3 and its association with abdominal symptoms in patients with inflammatory bowel disease (IBD) have been previously reported.4
In a Western meta-analysis, the prevalence of SIBO in patients with CD ranged from approximately 19% to 50%, depending on the disease phenotype and surgical history5; the overall prevalence was 25.4% in patients with CD, increasing to 31.8% following intestinal surgery and up to 33% after ileocecal valve resection.5 An updated recent meta-analysis reported a pooled prevalence of 32.2% in patients with CD, with the odds of SIBO in patients with CD being >7-fold higher than that in healthy controls (HCs).6 However, to our knowledge, prospective reports from Japan on SIBO in patients with CD are lacking. This gap is clinically relevant, because dietary practices differ substantially between Japan and Western countries. Furthermore, in Japan, elemental diets (EDs; e.g., Elental® [EA Pharma Co., Ltd.]) are extensively utilized even during remission, potentially reducing fermentable substrates and influencing the prevalence of SIBO in comparison with Western cohorts.7
Despite the pathophysiological plausibility that small-bowel strictures foster SIBO, prospective data confirming this association remain limited, and the effects of therapeutic interventions such as endoscopic balloon dilation (EBD) or ileocolic resection on SIBO dynamics have not been systematically evaluated. Therefore, using glucose hydrogen–methane breath testing, we conducted a prospective study of Japanese patients with CD to clarify the prevalence of SIBO, its association with strictures, and the changes following anatomical interventions.
MATERIALS AND METHODS
Study Design
This cross-sectional analysis of a prospective cohort was conducted at Niigata University Medical and Dental Hospital. The overall study design and patient flow are summarized in Figure 1. This study was approved by the Institutional Review Board of Niigata University (IRB No.
2023-0286) and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from patients for publication of the report and accompanying images.
Participants
This study enrolled 32 patients with CD between February 2024 and March 2026 (CD cohort). Patients with small-bowel strictures confirmed using double-balloon enteroscopy (DBE) were assigned to the stricture group (n = 17). Patients without strictures, as verified through cross-sectional imaging (computed tomography enterography or magnetic resonance enterography), capsule endoscopy, and clinical evaluation, were categorized into the non-stricture group (n = 15).
All DBE examinations were performed by two experienced endoscopists, and stricture classification was confirmed by consensus. Thirty-three asymptomatic individuals without prior gastrointestinal surgery or chronic disease were included as HCs for comparison of SIBO prevalence.
Clinical and Demographic Data
The following data were collected: age, sex, body mass index (BMI), CD duration, medication use (biologics, thiopurines, 5-aminosalicylic acid), history of surgery, ED use, and proton-pump inhibitor (PPI) use (Table 1). ED use was defined as regular intake of Elental® (EA Pharma) for ≥1 month before breath testing, regardless of daily dose. None of the patients, including those who underwent EBD or surgery, received antibiotics within 2 weeks before SIBO testing.
Disease activity was assessed using the Crohn’s Disease Activity Index (CDAI). Upper gastrointestinal symptom burden was evaluated using the Gastrointestinal Symptom Rating Scale (GSRS).
SIBO Diagnosis
SIBO was diagnosed on the basis of breath tests. To ensure diagnostic accuracy, participants were instructed to abstain from consuming bread, pasta, and noodles, which are known to increase hydrogen production, for 24 hours before the breath test. Additionally, the patients were instructed to fast with only minimal water intake for at least 12 hours before the test.
Regular medications were administered 2 hours before the test, if necessary, and the participants were advised to refrain from exercising or smoking during this period. Oral rinsing was performed to prevent the metabolism of sugar substrates by oral bacteria. Breath measurements, specifically measurements of hydrogen and methane concentrations, were performed three times before sugar substrate loading, and the average score was used as the baseline value.
Subsequently, breath measurements were recorded every 15 minutes post-loading for up to 120 minutes. The sugar substrate used was glucose (50 g), which is commonly used in clinical practice in Japan and several Asian studies. Although the North American Consensus recommends a 75-g glucose load, the same diagnostic cut-off values were applied in the present study.
Exhaled hydrogen and methane concentrations were measured using a BGA2000D instrument.
According to the North American Consensus,1,8 the diagnostic criteria for SIBO include an increase in hydrogen ≥20 ppm from baseline by 90 minutes and a methane level ≥10 ppm. Cases with elevated hydrogen levels were classified as hydrogen-type SIBO (H-SIBO), whereas those with elevated methane levels were categorized as methane-type SIBO (M-SIBO; intestinal methanogen overgrowth). Both hydrogen- and methane-positive cases were included in the positive group for each sub-analysis.
Intervention and Reassessment
In the stricture group, the patients who required intervention underwent EBD or surgical resection. To evaluate the short-term changes in SIBO status, patients who provided consent for retesting repeated the breath test within 3 days after EBD. Conversely, postoperative breath testing was performed approximately 2 months after surgery, once patients had clinically stabilized and completed perioperative antibiotic treatment.
All patients who underwent follow-up breath testing had a single clinically dominant stricture; in all cases, the treated stricture was successfully relieved by EBD or surgery. Balloon diameter and surgical details were recorded (Table 2).
Statistical Analysis
Due to the small sample size, all between-group comparisons were performed using a uniformly nonparametric approach. Continuous variables (age, BMI, disease duration, CDAI, and GSRS score) were compared using the Mann–Whitney U test, and effect sizes were estimated exclusively using Cliff’s delta. Categorical variables (sex, medication use, ED, PPI use, and SIBO positivity) were compared using Fisher’s exact test.
Odds ratios with Haldane–Anscombe correction and 95% confidence intervals were reported when binary outcomes such as SIBO prevalence were analyzed. Statistical significance was defined as a two-sided P-value of <0.05. All analyses were performed using R (version 4.3.1) and Python (SciPy and Pandas).
RESULTS
Patient Characteristics
The characteristics of patients with CD, with and without small-bowel strictures, based on a uniform nonparametric analysis are summarized in Table 1. Median age was higher in the stricture group than in the non-stricture group (42 years [interquartile range (IQR), 35.0–53.0 years] vs. 32 years [IQR, 29.0–40.5 years]).
BMI and disease duration were comparable between the groups. A history of intestinal surgery was more frequent in the stricture group (12/17, 70.6%) than in the non-stricture group (5/15, 33.3%), with most procedures consisting of ileal resection, including segmental ileal or jejunal resection. The use of PPIs, biologics, thiopurines, and 5-aminosalicylic acid was similar between the groups.
Conversely, HCs had a median age of 30 years (range, 25–72 years), and their sex distribution did not differ significantly from that of the CD group. None of the HCs were receiving any regular medications.
Clinical Scores
Patients with strictures had significantly higher GSRS scores than those without strictures (median, 39 [IQR, 29–45] vs. 22 [IQR, 17–25], P < 0.001) (Fig. 2A).
CDAI values were also higher in the stricture group (median, 80 [IQR, 56–146] vs. 69 [IQR, 13–86], P = 0.041) (Fig. 2B).
The correlation between the GSRS score and CDAI was weak overall (Spearman’s ρ = 0.30) and minimal within the stricture group (ρ = 0.11), whereas a moderate positive trend was observed in the non-stricture group (ρ = 0.46). Among SIBO-positive patients, no clear correlation was observed. The median GSRS scores were similar between the SIBO-positive and SIBO-negative patients within the stricture group (39 vs.
38), despite the higher CDAI values in SIBO-positive cases (146 vs. 70). These dissociation patterns suggest that the symptom burden in stricture-type CD may be partially independent of inflammatory activity.
SIBO Prevalence
H-SIBO was more frequent in patients with CD than in HCs, although the difference was not statistically significant when only the results of pre-intervention breath tests were analyzed (3/32 vs. 0/33; P = 0.114) (Fig. 2C).
This finding differs from most previously reported data from Western populations. In the CD cohort, H-SIBO was detected in 3/17 patients with strictures and 0/15 patients without strictures (17.6% vs. 0%; P = 0.229) (Fig.
2D). Although not significant, the effect estimate suggested a tendency toward a higher prevalence in the stricture group (Haldane–Anscombe corrected odds ratio, 7.48). In an additional sensitivity analysis, all patients who tested positive for H-SIBO at any time point, including those who changed from negative to positive after EBD or surgery, were considered SIBO-positive within the CD cohort.
Under these conditions, H-SIBO was detected in 6/32 patients with CD in comparison with 0/33 HCs (Fisher’s exact test, P = 0.011) (Fig. 2E). Although this analysis incorporated post-intervention data and therefore does not represent a primary cross-sectional comparison, the findings suggest that anatomical factors such as strictures may mask SIBO detection and that relief from obstruction may unmask previously undetectable hydrogen-producing bacterial overgrowth.
In contrast, M-SIBO (intestinal methanogen overgrowth) was detected in 1 of 33 HCs (3.0%) but was absent in all 32 patients with CD.
Impact of EBD and Surgery
Eight patients underwent EBD, with a mean maximum dilation diameter of 11.3 mm (median, 11.5 mm; standard deviation, 1.9 mm; range, 8.0–13.5 mm). Among the four patients who were reassessed post-EBD, two transitioned from SIBO-negative to SIBO-positive, one remained negative, and one remained positive throughout the study (Table 2). Notably, both patients who converted to SIBO-positive underwent transanal DBE, indicating that the dilated strictures were located in the ileum.
These findings suggest that severe stenosis may result in false-negative breath-test results before dilation, and that anatomical alterations in the distal small-bowel may profoundly influence microbial redistribution and breath-test results. Additionally, among the two patients who were re-evaluated after surgery, one who underwent segmental ileal resection with anastomosis developed H-SIBO postoperatively, suggesting that postoperative anatomical alterations in the ileum may influence the small-intestinal microbial environment. Importantly, these interventions (EBD and ileal segmental resection) only affected H-SIBO.
Representative endoscopic and radiologic images, along with breath-test results before and after EBD, are presented in Figure 3. The individual patient-level data before and after EBD or surgery are summarized in Table 2.
DISCUSSION
In our study, the prevalence of H-SIBO was 9.4% (3/32) on pre-intervention testing, which increased to 18.8% (6/32) when cases detected after endoscopic or surgical intervention were included. Nevertheless, the observed prevalence of H-SIBO in our cohort was lower than that reported in most previous studies (typically 19%–50%).5 A potential contributing factor may be the widespread use of EDs and fermentable carbohydrate restriction in Japan, even during remission, which could reduce small-intestinal fermentative substrates and suppress overgrowth.6,9 However, this interpretation remains speculative because we did not perform a direct analysis of the association between ED use and SIBO prevalence, nor did we adjust for potential confounders. In addition, ED use was more frequent in the stricture group, which may have further complicated the interpretation of its relationship with SIBO prevalence.
Although our exploratory data cannot establish causality, previous studies have suggested that EDs may reduce SIBO by limiting fermentable substrates and residual luminal nutrients.7,10
M-SIBO was detected in one HC (3.0%), but was absent in all patients with CD. Previous meta-analyses indicated that M-SIBO is uncommon in patients with IBD (pooled prevalence, ≈5%–6%), and patients with CD are particularly unlikely to produce excess methane in comparison with non-IBD populations.11 Our findings are consistent with global evidence, suggesting that hydrogen-dominant overgrowth predominates in CD, whereas intestinal methanogen overgrowth is rare.
Previous cross-sectional studies or meta-analyses have reported a higher prevalence of SIBO in patients with prior intestinal surgery, particularly after ileocecal resection.12 In contrast, our study uniquely demonstrated dynamic changes in the SIBO status within the same patient following stricture relief. To our knowledge, no previous study has prospectively evaluated the SIBO status before and after anatomical intervention in the same individual with CD. These observations suggest two plausible mechanisms.
First, false-negative breath tests before intervention: In severe stenosis, glucose may be absorbed proximally before reaching distal segments harboring overgrowth, leading to underdiagnosis, which is well-documented, and strictures can alter glucose test performance.1,3 Second, microbial redistribution after stricture relief, dilation, or resection may restore luminal continuity, allowing previously isolated distal microbiota to migrate proximally and alter ecological niches, thereby enabling the detection of hydrogen-producing bacteria.2,13 Since symptom-based assessments such as the GSRS were not evaluated in patients undergoing SIBO conversion, these findings should be interpreted as physiological or microbiological changes in breath-test profiles rather than direct evidence of clinically symptomatic SIBO.
Prior intestinal surgery, particularly ileocecal resection, increases the risk of SIBO by altering motility, disrupting the ileocecal valve, and creating blind-loop–like segments.14 In this study, surgical history was more frequent in the stricture group (70.6% vs. 33.3%), indicating that the observed association between strictures and SIBO may be partially confounded by postoperative anatomical changes. However, SIBO positivity also occurs after EBD in patients without a history of intestinal surgery.
Thus, although prior surgery remains a plausible confounder, our findings indicate that stricture-related impairment of glucose-substrate delivery may contribute to false-negative pre-intervention results. Nevertheless, given the substantial imbalance in the surgical history between the groups and the small number of SIBO-positive events, we could not reliably disentangle the relative contributions of prior surgery and stricture severity. Therefore, the independent effects of surgery versus stricture on the development of SIBO could not be adequately assessed.
SIBO has been linked to future clinical relapse in quiescent CD, reinforcing the value of detection even during remission, where symptoms are subtle.15 Our results support targeted breath testing with planned retesting after EBD or surgery and gas-profile–guided, guideline-concordant management, along with appropriate dietary strategies.1,7,15
The strengths of this study include its prospective enrollment, standardized breath-test protocols based on consensus criteria, direct confirmation of small-bowel strictures using DBE, evaluation of SIBO status before and after EBD and surgery, and its distinction as the first prospective cohort study in Japanese patients with CD to address these questions. The limitations of this study include its modest sample size, single-center design, and lack of multivariable adjustment. Owing to the limited sample size, age-adjusted multivariable analysis could not be performed; therefore, unadjusted comparisons may have overestimated the true difference in SIBO prevalence and should be interpreted with caution.
Another important limitation was the glucose dose used in breath testing. Although the North American Consensus recommends a 75-g glucose load, we used a 50-g dose, which is commonly used in Japanese clinical practice. A lower glucose dose may reduce test sensitivity and increase the likelihood of false-negative results, particularly in patients with small-bowel strictures in whom proximal glucose absorption may prevent the delivery of the substrate to distal overgrowth sites.
This methodological difference may have contributed to the relatively low prevalence of H-SIBO observed in our cohort and an increased likelihood of false-negative results before stricture relief. Additionally, our observations should be interpreted as short-term shifts rather than durable changes, since post-intervention breath testing was performed within 3 days after EBD. Therefore, the rapid conversion from SIBO-negative to SIBO-positive status may reflect false-negative results before dilation and transient microbial redistribution or procedure-related effects, rather than stable microbial colonization.
Longitudinal follow-up analyses are required to determine whether these changes persist or normalize over time.
Future multicenter studies with larger cohorts and integrated microbiome profiling are crucial for validating these findings and elucidating the mechanisms underlying SIBO dynamics. Longitudinal retesting before and after EBD or surgery, along with transit assessments and multi-omics profiling, is necessary to determine the optimal timing for retesting and understand how stricture relief reshapes the luminal ecology and symptom burden. Trials that integrate nutritional strategies (EDs and low-fermentable substrates) with SIBO-directed therapy can define optimal care pathways and identify patient subsets most likely to benefit.10,16-18
In summary, this study provides novel observational evidence of SIBO in Japanese patients with CD using DBE for precise stricture confirmation and unique tracking of changes in the SIBO status before and after anatomical interventions (EBD and surgery). Our findings indicate that H-SIBO predominates in patients with CD, whereas M-SIBO is rare, consistent with global meta-analytic evidence. However, the prevalence of hydrogen-type overgrowth in our Japanese cohort was lower than that reported in most previous studies.
The observation that SIBO can emerge after stricture dilation or resection suggests two mechanisms: 1) false-negative breath tests before intervention due to proximal glucose absorption and 2) microbial redistribution after restoration of luminal continuity. These insights highlight the importance of breath-test interpretation in stricturing CD and suggest a retesting paradigm post-intervention, while also suggesting diet and anatomy as modifiable contributors to small-intestinal microbial ecology.