The role of small intestinal bacterial overgrowth in the pathophysiology of irritable bowel syndrome.

Small intestinal bacterial overgrowth (SIBO) is a clinical condition caused by excessive number of bacteria in the small bowel. SIBO is characterized by symptoms of diarrhea, abdominal pain, or bloating which may be associated with excessive gas of small intestine due to increased production by bacterial fermentation in the gut. In the last decade, SIBO has been occupied as a hot topic of interests by a group of researchers because of its potential role in the development of irritable bowel syndrome (IBS).1

In the diagnosis of SIBO, a bacterial count greater than 105 colony-forming units/mL by small bowel culture is believed to be the gold standard.2 But small bowel culture through jejunal aspirate is time-consuming, invasive, and potential for contamination. Due to several flaws of small bowel culture, a number of non-invasive and indirect tests for diagnosing SIBO have been developed. Breath test, the most common indirect method for evaluating SIBO, utilizes the metabolism of carbohydrates by bacterial enzymes, in turn indirectly measures bacterial activity of small bowel. Different carbohydrates have different properties. As lactulose is a non-absorbable carbohydrate, it easily reaches the site of potential overgrowth in the small bowel, although non-absorbed carbohydrate has limitations by the intestinal transit time. In patients with rapid transit, lactulose may produce an early hydrogen peak which makes it difficult to discriminate the true SIBO from a phenomenon by rapid intestinal transit, therefore specificity of lactulose hydrogen breath test (LHBT) is relatively low.3

Pimentel and colleagues4 suggested a concept that SIBO is a major pathogenic mechanism underlying IBS in 2000. They have found 78% of 202 IBS subjects to be positive for LHBT which is suggestive of SIBO. They also demonstrated a significant improvement in symptoms including abdominal pain, bloating, and diarrhea by treating SIBO with antibiotics and converted many IBS patients from Rome criteria positive to negative.

Since then, SIBO has been proposed as an etiologic factor in IBS, but also the potential role of SIBO in IBS has been strongly debated about the methods used in diagnosing SIBO. A prevalence of SIBO in subjects meeting diagnostic criteria for IBS varies from 4% to 64%, depending on the kind of test and criteria used to define a positive result.5 In addition, a recent well-designed study from the US6 using LHBT to diagnose SIBO could not discriminate patients with IBS from healthy subjects, which was a consistent finding the authors demonstrated previously.7 Therefore, the proposed etiologic role of SIBO diagnosed by LHBT in IBS patients is controversial and the correlation between the normalization of the LHBT following antibiotic treatment and symptom improvement has become less clear.

Glucose, on the other hand, is rapidly absorbed in the proximal small bowel. Thereby detecting hydrogen in this test means SIBO in a proximal location. Glucose hydrogen breath test (GHBT) is the most commonly used test in the diagnosis of SIBO, although GHBT may be less sensitive for the diagnosis of SIBO since it is absorbed in the proximal small bowel. GHBT has been reported to have a overall sensitivity between 25% to 93% and a specificity from 34% to 96% in detecting SIBO.3,5

In this issue of the Journal of Neurogastroenterology and Motility, Ghoshal et al. reported the frequency of SIBO in patients with IBS and chronic non-specific diarrhea (CNSD) with comparison to healthy controls by using GHBT. The authors found that SIBO was more common in patients with CNSD (21.9%) than those with IBS (8.5%) and healthy controls (2%). Since CNSD including diarrhea predominant IBS (D-IBS) had higher frequency of SIBO, the authors suggested clinicians to consider testing for SIBO among them. However, since false positive GHBT could result from rapid intestinal transit in patients with diarrhea, the frequency of SIBO in patients with CNSD and D-IBS demonstrated in the author's study might have been overestimated in part.