Comparison of Lactose Versus Lactulose Breath Tests in Subjects Suspected with Small Intestinal Bacterial Overgrowth

Abstract

Purpose: Evidence for comparative utility of various carbohydrate substrates in hydrogen breath testing (HBT) for the diagnosis of small intestinal bacterial overgrowth (SIBO) is scant. The aim of this study was to compare the yields for lactulose vs. lactose HBT in suspected SIBO. We determined: 1) the frequency of SIBO; 2) the level of rise in hydrogen (H) & methane (CH3); & 3) the time to rise in H or CH3 with the highest sensitivity for SIBO with each carbohydrate substrate. Methods: All subjects who underwent HBT from 2003–06 at Rush University Medical Center were identified using the HBT database, cross referenced to a billing database. Subjects were excluded if they had HIV, immunosuppression, antibiotic use, widespread malignancy or terminal illness. For HBT, 25 gm of lactose or 20 gm of lactulose were used. Two baseline samples & samples every 15 mins were taken up to 165 mins and analyzed using the Quinton Microlyzer for H, CH3 & carbon dioxide (CO2). The clinical data & the HBT results were obtained via chart review and were analyzed with chi- square, Fisher exact tests for categorical data, & Kruskal-Wallis tests for continuous variables, using SPSS v. 11.5 & S+7.0. HBT were considered positive for SIBO if there was a double peak in H or CH3 in lactose or lactulose HBT; or the hydrogen level was > 20 ppm before 90 min in lactulose HBT. The tests were read by a single experienced gastroenterologist. Results: Patient related variables such as age, gender, pre- or post- test symptoms or diagnosis, past medical or surgical history, medications, endoscopic or histological findings from EGD &/or colonoscopy, hemoglobin & albumin did not differ significantly between the patients receiving lactose HBT (N = 110) vs. lactulose HBT (N = 85). There was no statistically significant difference in the frequency of SIBO in subjects who underwent lactose (36.7%) versus lactulose (40%) HBT (P= 0.221). There was no statistically significant difference between mean CH3 levels at any time interval with either substrate, although there was a trend towards higher CH3 levels at all time intervals in lactulose HBT. Mean CH3 levels in lactose vs lactulose peaked at 30 mins vs 60 mins, respectively in those with SIBO. Conclusion: Lactose and lactulose HBT yield similar results in SIBO diagnosis. Use of the lactose substrate could be considered in those patients suspected to have lactose intolerance and SIBO.