Refining small intestinal bacterial overgrowth diagnosis by means of carbohydrate specificity: a proof-of-concept study.

BACKGROUND

Diagnosis of small intestinal bacterial overgrowth (SIBO) remains challenging. This study aimed at proving the diagnostic concept of carbohydrate-specific SIBO (cs-SIBO) using glucose, fructose and sorbitol hydrogen (H2)/methane (CH4) breath testing (HMBT).

METHODS

In this study 125 patients referred to our outpatient clinic for SIBO testing were included. All individuals underwent glucose (50 g), fructose (25 g) and sorbitol (12.5 g) HMBT at 3 consecutive days. Patients with cs-SIBO (i.e. early H2/CH4 peak) were given rifaximin (600 mg/day) in a 10-day treatment. HMBT results were reassessed in a subset of patients 3-6 months after antibiotic therapy. In view of cs-SIBO diagnosis, agreements between HMBT results obtained for different sugars were calculated using Cohen's kappa (κ) with 95% confidence intervals (CIs).

RESULTS

A total of 59 (47.2%) patients presented an early H2/CH4 peak with one or more sugars. Among these, 21 (16.8%), 10 (8.0%) and 7 (5.6%) individuals had a positive HMBT result with either glucose, fructose or sorbitol, respectively. Another 21 (16.8%) patients with a positive glucose HMBT result were also found positive with an early H2/CH4 peak obtained after ingestion of fructose and/or sorbitol. Fair agreement was observed between glucose and fructose (κ = 0.26, p = 0.0018) and between glucose and sorbitol (κ = 0.18, p = 0.0178) HMBT results. Slight agreement was observed between fructose and sorbitol (κ = 0.03, p = 0.6955) HMBT results only. Successful antibiotic therapy with rifaximin could be demonstrated in 26/30 (86.7%) of patients as indicated by normal HMBT results and symptom remission.

CONCLUSIONS

Combined glucose, fructose and sorbitol HMBT has the potential to optimize cs-SIBO diagnosis. Furthermore, the majority of patients with cs-SIBO seem to benefit from rifaximin therapy regardless of its carbohydrate specificity.