Results
Baseline and Demographic Characteristics
During the 6-year period from 2004 to 2010 a total of 1,291 GHBT were performed to evaluate for the presence of SIBO. After applying exclusion criteria, 1,191 patients were included in the study. The primary indications for breath testing were bloating, dyspepsia, and/or other gas-related symptoms (n=507), diarrhea (n=337), abdominal pain (n=166), nausea, and vomiting (n=55). The remaining patients had overlapping indications that included constipation, IBS, dyspepsia, and gastroparesis and weight loss. The primary indication was listed as GERD in 12 patients, and each of these patients had secondary symptoms of bloating, dyspepsia, or abdominal pain. Only four patients studied had scleroderma and these all had diarrhea. On the basis of the chart review, a clinical history of IBS and GERD was found in 80.7% (961) and 64.8% (772) of the patients, respectively. The use of probiotics 7.2% (86) and fiber supplements 11.5% (137) was observed only in a small minority of study patients. The use of medications that alter intestinal motility, such as antidiarrheals 8.7% (104) and laxatives 10.1% (120), was also seen in only a small minority of the study population. Clinical conditions that alter GI motility, namely diabetes and hypothyroidism, were observed in 9.6 % (114) and 19.4% (231), respectively.
The patients were subcategorized into two groups: PPI users and PPI nonusers. PPI usage was present in 48% (566) of the study sample. The two groups were generally similar in demographics, but the PPI user group was slightly older (63 vs. 59 years) with a slightly higher BMI (27 vs. 25 kg/m) compared with the PPI nonusers. Patients with a clinical diagnosis of IBS were evenly distributed between these two groups (Table 1).
PPI Use and GHBT Positivity
The GHBT were evaluated as positive using four different cutoff values for exhaled level of H2 and CH4: (a) increase in H2>20 p.p.m. over baseline, (b) sustained rise H2>10 p.p.m. over baseline, (c) CH4>15 p.p.m. over baseline, and (d) either rise H2>20 p.p.m. over baseline or CH4>15 p.p.m. As shown in Table 2, no significant differences were found in the proportion of patients testing positive between PPI users and PPI nonusers applying any of the four criteria for a positive hydrogen breath test. Subgroup analysis to evaluate the influence of PPI dose frequency on the proportion of patients with a positive breath test also showed no significant differences between once daily PPI use and twice daily PPI use based on any of the four criteria evaluated (Table 3). The proportion of patients on once daily PPIs that had a positive GHBT was slightly higher than the proportion of patients on twice daily PPIs, but the difference was not statistically significance.
Multivariable Analysis
Multivariable logistic regression was used to evaluate the effect of PPI use, age, gender, BMI, IBS, GERD, probiotic use, fiber supplements, histamine receptor antagonists, antidiarrheal agents, diarrhea, bloating, and diabetes mellitus on the odds of having a positive GHBT (Table 4). The odds of a positive GHBT using the criterion of H2>20 p.p.m. over baseline was significantly associated with age (OR 1.03; 95% CI 1.01, 1.04) and the use of antidiarrheal medications (OR 1.96; 95% CI 1.14, 3.38). A similar pattern was seen when the H2>10 p.p.m. over baseline criterion for an abnormal breath test was used; increasing age (OR 1.01; 95% CI 1.01, 1.02) and an indication of diarrhea (OR 1.52; 95% CI 1.11, 2.06) for the GHBT were associated with the presence of an abnormal GHBT. Only increasing age (OR 1.01; 95% CI 1.01, 1.02) was associated with an abnormal GHBT using the criterion of H2>20 p.p.m. or CH4>15 p.p.m. over baseline. No significant associations were found between these predictors and an abnormal GHBT using the CH4>15 p.p.m. criterion alone. Importantly, PPI use was not significantly associated with a positive GHBT using any of the tested criteria either alone or after controlling for the other potential confounders.