Secondary Logo

Journal Logo

Original Articles: Gastroenterology

Double-blind, Placebo-controlled Antibiotic Treatment Study of Small Intestinal Bacterial Overgrowth in Children With Chronic Abdominal Pain

Collins, Brynie Slome*; Lin, Henry C

Author Information
Journal of Pediatric Gastroenterology and Nutrition: April 2011 - Volume 52 - Issue 4 - p 382-386
doi: 10.1097/MPG.0b013e3181effa3b
  • Free


Chronic abdominal pain (CAP) is a frequent and often debilitating childhood complaint. The overwhelming majority of children with CAP have a functional gastrointestinal disorder with symptoms not explained by anatomical or biochemical abnormalities. Although the exact prevalence of CAP is unknown, this diagnosis is responsible for 2% to 4% of all of the pediatric office visits in the United States, (1), with 13% to 19% of American school-age children experiencing weekly abdominal pain (2). Moreover, the abdominal pain and associated symptoms experienced by these children are often so severe that school performance, psychosocial function, and overall quality of life are impaired (3–5).

Although a unifying pathogenesis has yet to be identified for CAP, several mechanisms have been proposed to explain the symptoms of this condition, including altered intestinal motility (6), abnormal visceral perception (7), and psychological factors (8). Recently, evidence has also pointed to an underlying gut microbial mechanism for CAP (9,10).

Normally, in humans, the concentration of gut microbes is greatest in the colon, dropping steeply from 1010–12 organisms/mL in the cecum to 105–8 organisms/mL in the proximal ileum and 100–4 in the jejunum and duodenum (10). Hydrogen produced by bacterial fermentation is eliminated, in part, by methanogens or sulfate-reducing bacteria that convert hydrogen to methane or hydrogen sulfide (11,12). These microorganisms are competitive because the stool of an individual usually produces high concentrations of either methane or hydrogen sulfide.

When the microbial population native to the large intestine migrates proximally into the small intestine, a shift in the host–gut microbial relationship occurs known as small intestinal bacterial overgrowth (SIBO) (13). In patients with SIBO, abnormal microbial fermentation can occur in the small intestine and may be detected by measuring hydrogen and methane in the exhaled breath after ingestion of a fermentable substrate such as lactulose. This response is the basis for the lactulose breath test (LBT) (14). Because methanogens reside in the large intestine, with their concentration highest in the left colon (15), methane is normally absent from the exhaled breath during a 180-minute LBT.

Similar to patients with CAP, patients with SIBO complain of postprandial bloating, abdominal pain, nausea, vomiting and diarrhea, and constipation, as well as a variety of extraintestinal symptoms such as fatigue. In our recently published study (9), abnormal microbial fermentation suggesting SIBO was found in 91% of patients with CAP compared with 35% in healthy controls. These data are similar to that of Scarpellini et al (16), who found that 65% of children with irritable bowel syndrome (IBS) had SIBO. Abnormal microbial fermentation has also been reported in 78% to 84% of adult patients with IBS compared with 20% in healthy controls (17,18). In addition, in a randomized, double-blind, placebo-controlled trial, patients with IBS who responded to treatment with a nonabsorbable antibiotic, as shown by normalization of the LBT, reported 75% global improvement in symptoms (18). To date, there are no published data on the effect of antibiotic treatment in children with CAP. Therefore, the aim of the present study was to test the hypothesis that treatment of SIBO with a nonabsorbable antibiotic may reduce symptoms in children with CAP.


Study Population

All of the study subjects were between the ages of 8 and 18 years. Children with CAP were recruited from the gastroenterology (GI) clinics at Children's Hospital Los Angeles (CHLA) as well as from the CHLA GI outreach clinics in Pasadena and San Luis Obispo, California. The diagnosis of CAP was made using the Rome II Criteria for Functional Bowel Disorders Associated With Abdominal Pain or Discomfort in Children (19). Children were included if they met the criteria for functional dyspepsia, IBS, functional abdominal pain (FAP), or abdominal migraine. Patients were excluded if they had a history of inflammatory bowel diseases, diabetes, cirrhosis or other liver disease, juvenile rheumatoid arthritis, systemic lupus, or bowel resection because all of these conditions could predispose them to the development of bacterial overgrowth. Subjects were also excluded if they had been treated with antibiotics or probiotics within 2 months before enrollment, because these treatments could affect gut microflora. They were also excluded if they had an allergy to rifaximin or rifampin. Female subjects who were sexually active were required to use some form of birth control. Approval was obtained from the institutional review board at CHLA before initiation of the study, and consent and assent were obtained from all of the participants and their legal guardians. Additionally, this project was reviewed and approved for support by the National Institutes of Health–supported General Clinical Research Center (GCRC) at CHLA.

Study Design

The first 5 patients were seen and breath samples collected in the GI clinic at CHLA. However, shortly after initiation of the study, we received approval for support by the GCRC. Therefore, all subsequent patients were seen and breath samples collected in the GCRC.

All of the participants were called the day before their scheduled appointment and instructed to eat a light meal for dinner with nothing further to eat or drink after midnight. Upon arrival the next morning, patients and their guardians filled out a detailed history form. In addition, they were asked to fill out a questionnaire to assess the frequency and severity of their gastrointestinal symptoms of CAP. The questionnaire consisted of 10 visual analog scales (VAS) ranging from 0 to 10, with 0 representing absence of symptoms. The VAS assessed bloating, excess gas, incomplete evacuation, abdominal pain, diarrhea, constipation, urgency to have a bowel movement, passage of mucus, straining, and fecal soiling of underpants. In addition, the questionnaire contained 8 multiple-choice questions addressing the location, frequency, and duration of the pain as well as its effect on the patient's daily life.

All of the participants then underwent a lactulose breath hydrogen test (LBT). For this test, they exhaled during a single breath exchange through a straw into a 750-mL gas collection bag (Quintron Instrument Co, Milwaukee, WI) for 1 to 2 seconds to obtain baseline hydrogen and methane concentrations. They then drank 10 g of lactulose (Inalco Spa., Milano, Italy, packaged by Xactdose Inc, South Beloit, IL) in 100 ml of water and exhaled through a straw into a collection bag every 15 minutes for 180 minutes (for a total of 13 collections). End expiratory breath samples were taken to enhance alveolar gas sampling.

All of the LBT samples were analyzed for hydrogen and methane using either a Quintron SC MicroLyzer or a digital Quintron DP Breathtracker gas chromatograph (Quintron Instrument Co). Similar readings were confirmed on both machines by paired comparisons of multiple samples. A normal LBT was operationally defined as a clear peak of hydrogen >90 minutes after ingestion of the lactulose with a peak hydrogen concentration <20 ppm in 180 minutes (17,20). An abnormal LBT was based on a profile that failed to meet this operational definition.

Patients were then randomized, by personnel not associated with the study, in a 2:1 ratio, double-blind fashion, to receive either 550 mg of rifaximin (a gift of Salix Pharmaceuticals, Morrisville, NC) or matching placebo t.i.d. for 10 days. One week after initiation of treatment, each participant received a telephone call inquiring about compliance with medications and adverse effects.

Approximately 2 weeks after completing their randomized treatment, participants returned for a follow-up visit in which they were asked to fill out a questionnaire to assess the frequency and severity of their gastrointestinal symptoms of CAP in the preceding 2 weeks. The follow-up questionnaire consisted of the same 10 VAS as the first questionnaire. Participants were also asked to subjectively rate overall symptom improvement they experienced as a percent (100% meaning complete improvement). Compliance was assessed by pill count. They then underwent a repeat LBT to assess for a successful treatment of SIBO. The same testing protocol was used as described above. Patients were blinded to the results of their first LBT.

Statistical Analysis

All of the analyses were made using the intention-to-treat population, which included all of the enrolled subjects who took at least 1 pill. Comparisons of continuous variables between the patients with CAP with and without abnormal LBT and between the treatment and placebo groups were made using the Student t test and the nonparametric median test. Comparisons of noncontinuous variables were made using the Fisher exact test and median test. All of the results were expressed as mean ± SE, and a P value of <0.05 was considered statistically significant.

Sample Size Analysis

Because there was no published data at the time of the present study to reflect either the prevalence of SIBO in children with CAP or their response to antibiotic treatment, the power calculations were based on adult IBS data. With an anticipated prevalence of 75% in children with CAP and 20% in healthy controls, enrolling 75 children with CAP and 40 healthy controls would allow us >99% statistical power (2-sided alpha 0.05) to determine a difference in the prevalence rates between groups. If the prevalence of SIBO in children with CAP was as low as 47.4%, we had at least an 80% statistical power to detect a difference (2-sided alpha 0.05).

Again on the basis of adult data (18), we anticipated that at least 60% of the children treated with antibiotics would normalize their breath tests and 40% would still have abnormal breath tests suggestive of SIBO after treatment with rifaximin. Among the antibiotic-treated patients who normalized their breath tests, we expected an average of 75% improvement in symptoms. For placebo-treated subjects, we expected an average of 15% improvement. According to the initial sample size calculation, with the 75 children with CAP randomized to receive antibiotic versus placebo in a 2:1 ratio, there was a 98% statistical power to detect a difference in average symptom improvement between the antibiotics-treated patients who normalized their LBT and the placebo-treated patients (2-sided alpha 0.05; pooled standard deviation 50%). With an average symptom improvement as low as 57% for successfully treated subjects, we still had at least an 80% statistical power to detect a difference.


Patient Population

Seventy-five patients with CAP were enrolled in the study. Of the children with CAP, 11 (15%) met the Rome II criteria for functional dyspepsia, 41 (55%) met the criteria for IBS, and 23 (31%) met the criteria for FAP. Forty-nine of the 75 children with CAP received rifaximin 550 mg t.i.d. for 10 days, and 26 received matching placebo administered in the same way. Of the 49 children who received rifaximin, 44 (90%) had an abnormal initial LBT suggestive of SIBO compared with 24 of 26 (92%) of those with CAP who received placebo (NS). No differences were found in the demographical features of children with CAP who had an abnormal LBT and those with a normal LBT (Table 1). In addition, 23 of 75 (30%) of those with CAP were methane excreters. One patient from the treatment group withdrew from the study because of abdominal pain after taking 1 day of rifaximin. Four additional children, 2 from the treatment group and 2 from the placebo group, did not return for their follow-up visits. Their data were handled in accordance with intent-to-treat analysis. No differences were found in the demographical features of children who received rifaximin and those who received placebo (Table 2).

Comparison of demographics between CAP with abnormal LBT and CAP with normal LBT
Comparison of demographics between rifaximin and placebo

Response to Treatment

When the 46 children who received rifaximin were compared with the 24 children who received placebo, no significant differences were found between the 2 groups after treatment (NS for all of the individual symptoms and for overall symptom improvement). Analysis of compliance revealed no difference in the distribution of returned pills between groups. No significant differences were found between the groups in the mean baseline or maximum levels of hydrogen or methane production before and after treatment (NS) (Figs 1 and 2). Abnormal LBTs persisted after completion of treatment in 37 of 46 (80%) children with an abnormal initial LBT who received rifaximin and 19 of 22 (86%) with an abnormal initial LBT who received placebo (Fig. 3).

Mean baseline and maximum H2 excretion before (visit 1) and after (visit 2) treatment.
Mean baseline and maximum CH4 excretion before (visit 1) and after (visit 2) treatment.
Percentage of patients with an abnormal LBT before and after treatment. LBT = lactulose breath test.

Of the 9 children with an abnormal initial LBT who received rifaximin and had a normal repeat LBT, 6 had IBS and 3 had FAP. Of the 3 children with an abnormal initial LBT who received placebo and had a normal repeat LBT, 1 had IBS and 2 had FAP.


CAP, also called recurrent abdominal pain or RAP, first appeared in the literature in 1958 to describe children who had experienced at least 3 bouts of pain that was severe enough to affect daily activities during a period of at least 3 months (21). Although this definition is nonspecific, CAP continues to be used to describe all children with abdominal pain in whom a specific organic etiology cannot be identified. In March 2005, an American Academy of Pediatrics Subcommittee and the North American Society of Pediatric Gastroenterology, Hepatology, and Nutrition Committee on Chronic Abdominal Pain published a technical report supporting the use of the Rome II Criteria for Functional Bowel Disorders Associated with Abdominal Pain or Discomfort in Children as diagnostic criteria for CAP (22). These criteria have the subcategories of functional dyspepsia, IBS, FAP, and abdominal migraine. Although these symptom-based criteria can provide the clinician with an organized way to approach the diagnosis of children with CAP, no unifying pathogenesis has been identified to explain the varying presentations associated with CAP.

In our previously published prevalence data of the present study population, we found that 91% of children with CAP, regardless of their symptoms, have abnormal microbial fermentation as demonstrated by an abnormal LBT (9). These data are consistent with other recently published findings of an increased prevalence of abnormal microbial fermentation in children with IBS. Scarpellini et al (16) reported that 65% of children with IBS had abnormal LBT compared with 7% of controls. These findings support an underlying gut microbial mechanism in CAP.

In adults, the contributing role of a gut microbial mechanism in IBS has been supported by clinical studies showing a significant reduction in both gastrointestinal and extraintestinal symptoms in adult patients with IBS, when a previously abnormal LBT was normalized by a short course of antibiotic. Specifically, in a double-blind, placebo-controlled study of 111 patients drawn from the general IBS population, Pimentel et al (17) found that there was a graded response to antibiotic treatment, whereby the mean improvement of global symptoms within 1 week of treatment was 11% for placebo-treated patients, 36.7% for antibiotic-treated patients who were not successfully treated, and 75% for antibiotic-treated patients who achieved a normal LBT after treatment (P < .001). In another double-blind, randomized, placebo-controlled study using rifaximin, the subjects achieved an improvement of IBS symptoms that was sustained for 10 weeks after completing a 10-day course of treatment (23).

The aim of the present study was to test the hypothesis that treatment of SIBO with a nonabsorbable antibiotic may reduce symptoms in children with CAP. Unlike the response in adults with IBS, our results show no significant difference in any symptom improvement between children treated with rifaximin and those treated with placebo. However, 80% of children who received rifaximin had persistently abnormal LBTs after completion of their treatment course, suggesting that the treatment was ineffective in this group. This response is in contrast to the previously reported findings in adults that rifaximin, at the dose of 1200 mg/day for 7 days, had a 60% rate of normalizing glucose breath tests (24), and at a higher dose of 1600 mg/day for 7 days, had an 80% rate of normalizing breath tests (25). In our study, rifaximin at the dose of 1650 mg/day for 10 days had a rate of normalizing LBTs in only 20% of the children. It remains unclear whether a different treatment with higher efficacy would lead to symptom improvement in patients with CAP.

There are several possible explanations as to why rifaximin may be less effective in children than in adults. One possibility is that children have a more resistant gut bacterial population that may require either a longer treatment course or a higher antibiotic dose than adults. Another possibility is that children are colonized with bacteria with different antibiotic susceptibility than are adults.

Criticism has been aimed at the sensitivity of the LBT to test for SIBO compared with the criterion standard of small bowel aspiration and culture (26). Adult studies based on culture of duodenal aspirates report the sensitivity of the LBT to be between 16% and 61% (26,27); however, pediatric data report the sensitivity of the LBT to be 86% (28). The problem with all of these assessments of the LBT is that aspirates from the proximal bowel were used as the criterion standard for comparison, even though it is well established that there are fundamental problems in accessing, sampling, and culturing the small bowel (10). Only the proximal 60 cm of the small intestine can be reached during upper endoscopy, and because the duodenum is exposed to strongly acidic gastric juice, this region of the gut is the least likely place for bacteria to reside. In addition, because the resident gut microbes are embedded in the mucous layer atop the mucosa, aspirating luminal content is inadequate for sampling. Moreover, an estimated 60% to 80% of gut microbes are not culturable by any means, which significantly limits this method of detection. The LBT, however, is not limited by these considerations. As a nonabsorbable carbohydrate, lactulose travels across the full length of the gut and is available for fermentation by bacteria, wherever the microbial population may be. It is not surprising, therefore, that direct culture has a reproducibility of only 38% compared with 92% for LBTs (27). We elected to use the LBT because it is noninvasive and easily performed in the pediatric population.

Whether an abnormal LBT is indicative of SIBO has also been questioned. Hydrogen and methane are gaseous byproducts of microbial fermentation and metabolism (29). Therefore, when an abnormal LBT is based on the profiles of these gases in the exhaled breath, this demonstrates, without controversy, abnormal microbial fermentation. The sustained symptom improvement for 10 weeks in adults after completing a 10-day course of antibiotics (23) further supports the role of gut microbes in IBS.

In summary, in an earlier publication (9), we showed that similar to adults with IBS, children with CAP have a high prevalence of abnormal microbial fermentation suggesting SIBO. In the present study, we tested the role of gut microbes using a randomized antibiotic treatment. We found that treatment with 10 days of rifaximin has low efficacy in normalizing LBT in this group. Additional studies are needed to determine whether a treatment approach directed at gut microbes with higher efficacy would lead to symptom improvement in patients with CAP.


We thank the nurses in the GCRC at CHLA for their technical support in performing the breath tests. We also thank the other physicians in the Division of Gastroenterology and Nutrition for their assistance with patient recruitment. Finally, we thank Colleen Azen for assistance with statistical analysis.


1. Starfield B, Hoeklman RA, McCormick M, et al. Who provides health care to children and adolescents in the United States? Pediatrics 1984; 74:991–997.
2. Hyams JS, Burke G, Davis PM, et al. Abdominal pain and irritable bowel syndrome in adolescents: a community based study. J Pediatr 1996; 129:220–226.
3. Drossman DA, Li Z, Andruzzi E, et al. U.S. householder survey of functional gastrointestinal disorders: prevalence, sociodemography and health impact. Dig Dis Sci 1993; 38:1569–1580.
4. Apley J. The child with recurrent abdominal pain. Pediatr Clin North Am 1967; 14:63–72.
5. Youssef NN, Murphy TG, Langseder AL, et al. Quality of life for children with functional abdominal pain: a comparison study of patients' and parents' perceptions. Pediatrics 2006; 117:54–59.
6. Hyman PE, Napolitano JA, Diego A, et al. Antroduodenal manometry in the evaluation of chronic functional gastrointestinal symptoms. Pediatrics 1990; 86:39–44.
7. Van Ginkel R, Voskuijl WP, Benninga MA, et al. Alterations in rectal sensitivity and motility in childhood irritable bowel syndrome. Gastroenterology 2001; 120:31–38.
8. Hyman PE, Bursch B, Sood M, et al. Visceral pain-associated disability syndrome: a descriptive analysis. J Pediatr Gastroenterol Nutr 2002; 35:663–668.
9. Collins BS, Lin HC. Chronic abdominal pain in children is associated with high prevalence of abnormal microbial fermentation. Dig Dis Sci 2010; 55:124–130.
10. Lin HC. Small intestinal bacterial overgrowth: a framework for understanding irritable bowel syndrome. JAMA 2004; 292:852–858.
11. Levitt MD. Production and excretion of hydrogen gas in man. N Engl J Med 1969; 281:122–127.
12. Strocchi A, Furne J, Ellis C, et al. Methanogens outcompete sulphate reducing bacteria for H2 in the human colon. Gut 1994; 35:1098–1101.
13. King CE, Toskes PP. Small intestine bacterial overgrowth. Gastroenterology 1979; 76:1035–1055.
14. Rhodes JM, Middleton P, Jewell DP. The lactulose hydrogen breath test as a diagnostic test for small-bowel bacterial overgrowth. Scand J Gastroenterol 1979; 14:333–336.
15. Macfarlane GT, Gibson GR, Cummings JH. Comparison of fermentation reactions in different regions of the human colon. J Appl Bacteriol 1992; 72:57–64.
16. Scarpellini E, Giorgio V, Gabrielli M, et al. Prevalence of small intestinal bacterial overgrowth in children with irritable bowel syndrome: a case-control study. J Pediatr 2009; 155:416–420.
17. Pimentel M, Chow EJ, Lin HC. Eradication of small intestinal bacterial overgrowth reduces symptoms in irritable bowel syndrome. Am J Gastroenterol 2000; 95:3503–3506.
18. Pimentel M, Chow EJ, Lin HC. Normalization of lactulose breath testing correlates with symptom improvement in irritable bowel syndrome: a double blind, randomized controlled study. Am J Gastroenterol 2003; 98:412–419.
19. Rasquin-Weber A, Hyman PE, Cucchiara S, et al. Childhood functional gastrointestinal disorders. Gut 1999; 45(Suppl 2):II60–II68.
20. Joseph F Jr, Rosenberg AJ. Breath hydrogen testing: diseased versus normal patients. J Pediatr Gastroenterol Nutr 1988; 7:787–788.
21. Apley J, Naish N. Recurrent abdominal pain: a field survey of 1000 school children. Arch Dis Child 1958; 33:165–170.
22. Chronic abdominal pain in children: a technical report of the American Academy of Pediatrics and the North American Society of Pediatric Gastroenterology, Hepatology and Nutrition. Pediatrics 2005;115:e370–81.
23. Pimentel M, Park S, Mirocha J, et al. The effect of a nonabsorbed oral antibiotic (rifaximin) on the symptoms of the irritable bowel syndrome: a randomized trial. Ann Intern Med 2006; 145:557–563.
24. Lauritano EC, Gabrielli M, Lupascu A, et al. A. Rifaximin dose-finding study for the treatment of small intestinal bacterial overgrowth. Aliment Pharmacol Ther 2005; 22:31–35.
25. Scarpellini E, Gabrielli M, Lauritano CE, et al. A. High dosage rifaximin for the treatment of small intestinal bacterial overgrowth. Aliment Pharmacol Ther 2007; 25:781–786.
26. Riordan SM, McIver CJ, Walker BM, et al. The lactulose breath hydrogen test and small intestinal bacterial overgrowth. Am J Gastroenterol 1996; 91:1795–1803.
27. Corazza G, Strocchi A, Sorge M, et al. Prevalence and consistency of low breath H2 excretion following lactulose ingestion. Possible implications for the clinical use of the H2 breath test. Dig Dis Sci 1993; 38:2010–2016.
28. Mendoza E, Crismatt C, Matos R, et al. Diagnosis of small intestinal bacterial overgrowth in children: The use of lactulose in the breath hydrogen test as a screening test. Biomedica 2007; 27:325–332.
29. Tillman R, King C, Toskes P. Continued experience with the xylose breath test: evidence that the small bowel culture as a gold standard for bacterial overgrowth may be tarnished. Gastroenterology 1981; 80:1304.

chronic abdominal pain; functional abdominal pain; functional dyspepsia; irritable bowel syndrome; small intestinal bacterial overgrowth

Copyright 2011 by ESPGHAN and NASPGHAN