Lactulose breath testing, bacterial overgrowth, and IBS: just a lot of hot air?

Abstract

Pimentel M, Chow EJ, Lin HC (Cedars-Sinai Medical Center, Los Angeles, California; and University of California School of Medicine, Los Angeles, California). Normalization of lactulose breath testing correlates with symptom improvement in irritable bowel syndrome: a double-blind, randomized, placebo-controlled study. Am J Gastroenterol 2003;98:412–419.The pathogenesis of irritable bowel syndrome (IBS) is poorly understood. Recently, small intestinal bacterial overgrowth was proposed to contribute to IBS symptoms. In the current study, Pimentel and colleagues present findings of a double-blind trial of antibiotic therapy in IBS patients and correlate symptoms to lactulose breath test results.One hundred eleven community subjects with IBS satisfying the Rome I criteria were recruited. Exclusion criteria included recent antibiotic or probiotic use, prior breath testing, extraintestinal diseases (diabetes, thyroid disease, connective tissue disease), prior abdominal surgery, and narcotic use. Fifteen gender-matched healthy controls without IBS also participated. Subjects graded IBS symptoms from 0 to 5. Composite scores were generated by summing pain, diarrhea, and constipation scores. Breath testing was performed after ingesting 10 grams of lactulose. A normal test was defined as no rise in breath hydrogen or methane in the first 90 minutes after lactulose and no rise >20 parts per million after 3 hours. Subjects received neomycin 500 mg twice daily versus placebo for 10 days in blinded fashion. Seven days after completing therapy, symptoms were reassessed and breath testing was repeated. Clinical responses were defined by 50% decreases in composite scores.Eighty-four percent of IBS patients exhibited abnormal breath tests versus 20% of controls (P < 0.00001). Neomycin produced 35% ± 5% composite score reductions compared with 11% ± 9% with placebo (P < 0.05). In the neomycin group, 40% ± 5% reported bowel habit normalization versus 15% ± 4% for placebo (P < 0.001). Clinical responses were observed in 43% of neomycin and 23% of placebo patients (P < 0.05). In those with an abnormal breath test, 55% on neomycin had a clinical response versus 11% of placebo patients (P < 0.0001). Twenty percent of neomycin patients and 2% in the placebo group normalized their breath test. Those who achieved breath test normalization had greater symptom reductions (62% ± 9%) than those who did not (34% ± 6%, P < 0.01). Breath hydrogen profiles were similar in diarrhea- and constipation-predominant patients. However, breath methane was increased in 39% of constipated patients versus no patient with diarrhea. Constipation scores were twice as high in those with elevated breath methane compared with constipated patients with normal methane.In conclusion, this study confirmed a high prevalence of abnormal lactulose breath tests in IBS. Furthermore, antibiotics reduced symptoms as well as breath hydrogen and methane. Finally, expired gas profiles were different in the distinct IBS subgroups. In the discussion, the investigators concluded these findings are consistent with the hypothesis that bacterial overgrowth underlies symptoms in most IBS patients and that increased methane production may selectively contribute to constipation.CommentEvidence is accumulating that altered gut flora play pathogenic roles in irritable bowel syndrome (IBS). Hydrogen production by enteric bacteria after lactulose is increased in IBS patients (Lancet 1998;352:1187–1189). Furthermore, excluding dietary fermentable carbohydrates decreases hydrogen production. However, some report poor correlations between hydrogen excretion and symptoms (Am J Gastroenterol 1989;84:375–378). Therapies that modify gut flora may improve symptoms in some patients. Probiotics containing active cultures decrease gas-related IBS symptoms (Am J Gastroenterol 2000;95:1231–1238). Recently, metronidazole was reported to reduce symptoms by >50% in Indian IBS patients versus 25% on placebo (Indian J Gastroenterol 1997;16:137–139).One group has promoted an extension of this theory. In their initial publication in 2000, Pimentel and colleagues noted positive lactulose breath tests in 78% of IBS patients, which they attributed to small intestinal bacterial overgrowth (Am J Gastroenterol 2000;95:3503–3506). Forty-seven of 202 patients were given antibiotics and restudied. Twenty individuals normalized their breath tests with associated decreases in diarrhea and pain. More recently, positive lactulose breath tests were demonstrated in 78% of fibromyalgia patients (J Musculoskel Pain 2001;9:107–113). Antibiotics reduced joint pains and fatigue, symptoms the investigators associated with bacterial overgrowth. To define the mechanism underlying bacterial overgrowth, Pimentel et al. compared small intestinal manometric findings in IBS patients and healthy controls (Dig Dis Sci 2002;47:2639–2643). Those with IBS exhibited delayed cycling of fasting motor complexes, and complex durations were shortened by one third. It was concluded that bacterial overgrowth in IBS results from small intestinal hypomotility.The present study by Pimentel et al. is an effort to address shortcomings of their initial investigation. The efficacy of neomycin in reducing symptoms and decreasing breath hydrogen and methane excretion was compared in rigorous double blind, placebo-controlled fashion in a moderately large IBS population. Similar numbers of constipation- and diarrhea-predominant patients were included. Breath test findings in IBS patients were compared with a matched control group. As in the initial study, most IBS patients exhibited abnormal breath tests. The important new findings of this study included the demonstration that neomycin reduces IBS symptoms more than placebo and that symptom reductions correlate with breath test normalization.These findings, if validated, have enormous implications for our understanding of IBS pathophysiology and its management. If IBS symptoms largely result from bacterial colonization of a normally nearly sterile gut region, the current focus on pharmaceuticals to modulate visceral motor and sensory function would be misguided. Rather, exploration of new means of eradicating small bowel overgrowth would be warranted. The notion of small bowel bacterial fermentation as a cause of IBS symptoms is attractive because it is easy to picture hungry organisms devouring meal residues and producing gaseous and liquid by-products which then distend and irritate the distal intestine. With a condition like IBS in which true progress seems slow and incremental, one is tempted to jump on a new bandwagon that so completely rejects the existing dogma. Parallels can be drawn between this theory and early data supporting Helicobacter pylori as a cause of peptic ulcers. However, in contrast to the convincing research on H. pylori, this investigation teases the reader with methodologies that infer but do not confirm the presence of small intestinal organisms in IBS.The use of lactulose breath testing in IBS generates more questions than answers. Most importantly, what and where are the organisms that cause symptoms in IBS? The authors explain that IBS is different from other conditions with bacterial overgrowth in that only the distal small intestine is colonized, thus jejunal cultures are not useful. Indeed, a recent study using jejunal sampling detected pathogenic bacteria in only 4 of 33 IBS patients (Gastroenterology 2003;124:S1152). However, manometric abnormalities described in IBS by the investigators involve the proximal small bowel like other dysmotility syndromes, thus it is not clear why the jejunum should not be colonized in IBS (Scand J Gastroenterol 1996;31:875–880). Reliance on lactulose breath testing as a sole test of bacterial overgrowth can be questioned. Compared with jejunal cultures, lactulose breath testing exhibits sensitivities as low as 17% in documented bacterial overgrowth (Am J Gastroenterol 1996;91:1795–1803, J Am Geriat Soc 1992;40:692–696). More troublesome is its poor specificity (<50%) showing that many positive tests do not result from overgrowth at all (Gastroenterology 1990;98:302–309). Most false positives likely are attributable to colonic fermentation of the sugar. Because lactulose is not absorbed, anyone with hydrogen-generating bacteria will produce a hydrogen peak after ingestion. Setting the standards for the test involves determining how much hydrogen is abnormal and when it should appear in the breath. In the present study, a normal lactulose breath test is defined as one with no increase in hydrogen or methane in the first 90 minutes after lactulose consumption. In studies using lactulose breath testing as a measure of transit, orocecal transit times using the same lactulose dose as in the current study (10 grams) range from 94 to 97 minutes in healthy volunteers (J Lab Clin Med 1975;85:546–555). Based on these data, one would have expected 50% of healthy controls in this study to exhibit hydrogen or methane peaks in the initial 90 minutes after lactulose instead of the observed 20%.It is unclear why symptoms responded so poorly to antibiotics in this study. In bacterial overgrowth caused by delayed transit, antibiotics produce prompt improvements in diarrhea, reductions in hydrogen, and increased weight (Gastroenterology 1999;117:794–797, Am J Gastroenterol 1999;94:1327–1331, Age Aging 1992;21:13–19, J Clin Gastroenterol 2000;31:63–66). The authors justify the 7-day follow-up interval because of their observation that breath test abnormalities recur within 2 weeks in IBS after antibiotic eradication. It is counterintuitive to postulate that healthy young IBS patients have such refractory bacterial colonization, whereas many sicker individuals with true small bowel dysmotility clear so easily. Indeed, the limited symptom responses and rapid recurrences are more plausibly explained by pathogenic organisms in the colon that would be expected to reconstitute more quickly.Comment should be made on discrepancies between the authors’ data and literature reports. In their experience, manometry findings are similar in IBS patients with diarrhea and constipation, whereas others observed increased cycling in diarrhea-predominant patients (Dig Dis Sci 2002;47:2639–2643, Dig Dis Sci 1995;40:2383–2389). This study noted similar intestinal transit times in diarrhea and constipation, whereas others measured more rapid transit in diarrhea-predominant IBS (Gut 1983;21:405–411). Finally, it is unusual that placebo reduced symptoms in only 11%–15% of patients. Most IBS drug trials report placebo responses in 30%–70% of patients—values similar to the response to neomycin in the current study (Gastroenterology 1988;95:232–241).The present study observed selective methane production only in constipated patients. Methane is produced by colonic bacteria in 40% of healthy humans. Diverticulosis and encopresis have increased methane excretion, whereas inflammatory bowel disease patients produce less methane (Gut 1986;27:698–704, Scand J Gastroenterol 1983;18:289–298, J Ped Gastroenterol Nutr 1990;10:473–477). It is uncertain if these are epiphenomena or if methane retards gut transit. Furthermore, some persons lose the capacity to generate methane, whereas others acquire this ability with time (J Clin Invest 1992;89:1304–1311, Dig Dis Sci 1994;39:494–497). It is unknown if this could be a mechanism for the alternating bowel pattern in some IBS patients.Even with limitations, this investigation presents intriguing data. This study does not prove that IBS symptoms are caused by small intestinal bacteria, but it does infer an important alteration of gut flora. It should be possible to devise an experimental technique to confirm if distal small intestinal bacterial overgrowth is present in IBS. Placement of a flexible capillary tube or other absorbent substance could facilitate collection of ileal secretions for culture. Alternatively, lactulose breath testing could be performed with an occlusive balloon in the distal ileum to prevent colonic fermentation. Finally, concurrent scintigraphy could ascertain if hydrogen and methane levels increase when the lactulose is in the ileum or after it reaches the colon. Regardless, this provocative work challenges current theories about the pathogenesis of functional bowel disorders and forces those of us in the field to consider alternate hypotheses. Pimentel M, Chow EJ, Lin HC (Cedars-Sinai Medical Center, Los Angeles, California; and University of California School of Medicine, Los Angeles, California). Normalization of lactulose breath testing correlates with symptom improvement in irritable bowel syndrome: a double-blind, randomized, placebo-controlled study. Am J Gastroenterol 2003;98:412–419. The pathogenesis of irritable bowel syndrome (IBS) is poorly understood. Recently, small intestinal bacterial overgrowth was proposed to contribute to IBS symptoms. In the current study, Pimentel and colleagues present findings of a double-blind trial of antibiotic therapy in IBS patients and correlate symptoms to lactulose breath test results. One hundred eleven community subjects with IBS satisfying the Rome I criteria were recruited. Exclusion criteria included recent antibiotic or probiotic use, prior breath testing, extraintestinal diseases (diabetes, thyroid disease, connective tissue disease), prior abdominal surgery, and narcotic use. Fifteen gender-matched healthy controls without IBS also participated. Subjects graded IBS symptoms from 0 to 5. Composite scores were generated by summing pain, diarrhea, and constipation scores. Breath testing was performed after ingesting 10 grams of lactulose. A normal test was defined as no rise in breath hydrogen or methane in the first 90 minutes after lactulose and no rise >20 parts per million after 3 hours. Subjects received neomycin 500 mg twice daily versus placebo for 10 days in blinded fashion. Seven days after completing therapy, symptoms were reassessed and breath testing was repeated. Clinical responses were defined by 50% decreases in composite scores. Eighty-four percent of IBS patients exhibited abnormal breath tests versus 20% of controls (P < 0.00001). Neomycin produced 35% ± 5% composite score reductions compared with 11% ± 9% with placebo (P < 0.05). In the neomycin group, 40% ± 5% reported bowel habit normalization versus 15% ± 4% for placebo (P < 0.001). Clinical responses were observed in 43% of neomycin and 23% of placebo patients (P < 0.05). In those with an abnormal breath test, 55% on neomycin had a clinical response versus 11% of placebo patients (P < 0.0001). Twenty percent of neomycin patients and 2% in the placebo group normalized their breath test. Those who achieved breath test normalization had greater symptom reductions (62% ± 9%) than those who did not (34% ± 6%, P < 0.01). Breath hydrogen profiles were similar in diarrhea- and constipation-predominant patients. However, breath methane was increased in 39% of constipated patients versus no patient with diarrhea. Constipation scores were twice as high in those with elevated breath methane compared with constipated patients with normal methane. In conclusion, this study confirmed a high prevalence of abnormal lactulose breath tests in IBS. Furthermore, antibiotics reduced symptoms as well as breath hydrogen and methane. Finally, expired gas profiles were different in the distinct IBS subgroups. In the discussion, the investigators concluded these findings are consistent with the hypothesis that bacterial overgrowth underlies symptoms in most IBS patients and that increased methane production may selectively contribute to constipation. CommentEvidence is accumulating that altered gut flora play pathogenic roles in irritable bowel syndrome (IBS). Hydrogen production by enteric bacteria after lactulose is increased in IBS patients (Lancet 1998;352:1187–1189). Furthermore, excluding dietary fermentable carbohydrates decreases hydrogen production. However, some report poor correlations between hydrogen excretion and symptoms (Am J Gastroenterol 1989;84:375–378). Therapies that modify gut flora may improve symptoms in some patients. Probiotics containing active cultures decrease gas-related IBS symptoms (Am J Gastroenterol 2000;95:1231–1238). Recently, metronidazole was reported to reduce symptoms by >50% in Indian IBS patients versus 25% on placebo (Indian J Gastroenterol 1997;16:137–139).One group has promoted an extension of this theory. In their initial publication in 2000, Pimentel and colleagues noted positive lactulose breath tests in 78% of IBS patients, which they attributed to small intestinal bacterial overgrowth (Am J Gastroenterol 2000;95:3503–3506). Forty-seven of 202 patients were given antibiotics and restudied. Twenty individuals normalized their breath tests with associated decreases in diarrhea and pain. More recently, positive lactulose breath tests were demonstrated in 78% of fibromyalgia patients (J Musculoskel Pain 2001;9:107–113). Antibiotics reduced joint pains and fatigue, symptoms the investigators associated with bacterial overgrowth. To define the mechanism underlying bacterial overgrowth, Pimentel et al. compared small intestinal manometric findings in IBS patients and healthy controls (Dig Dis Sci 2002;47:2639–2643). Those with IBS exhibited delayed cycling of fasting motor complexes, and complex durations were shortened by one third. It was concluded that bacterial overgrowth in IBS results from small intestinal hypomotility.The present study by Pimentel et al. is an effort to address shortcomings of their initial investigation. The efficacy of neomycin in reducing symptoms and decreasing breath hydrogen and methane excretion was compared in rigorous double blind, placebo-controlled fashion in a moderately large IBS population. Similar numbers of constipation- and diarrhea-predominant patients were included. Breath test findings in IBS patients were compared with a matched control group. As in the initial study, most IBS patients exhibited abnormal breath tests. The important new findings of this study included the demonstration that neomycin reduces IBS symptoms more than placebo and that symptom reductions correlate with breath test normalization.These findings, if validated, have enormous implications for our understanding of IBS pathophysiology and its management. If IBS symptoms largely result from bacterial colonization of a normally nearly sterile gut region, the current focus on pharmaceuticals to modulate visceral motor and sensory function would be misguided. Rather, exploration of new means of eradicating small bowel overgrowth would be warranted. The notion of small bowel bacterial fermentation as a cause of IBS symptoms is attractive because it is easy to picture hungry organisms devouring meal residues and producing gaseous and liquid by-products which then distend and irritate the distal intestine. With a condition like IBS in which true progress seems slow and incremental, one is tempted to jump on a new bandwagon that so completely rejects the existing dogma. Parallels can be drawn between this theory and early data supporting Helicobacter pylori as a cause of peptic ulcers. However, in contrast to the convincing research on H. pylori, this investigation teases the reader with methodologies that infer but do not confirm the presence of small intestinal organisms in IBS.The use of lactulose breath testing in IBS generates more questions than answers. Most importantly, what and where are the organisms that cause symptoms in IBS? The authors explain that IBS is different from other conditions with bacterial overgrowth in that only the distal small intestine is colonized, thus jejunal cultures are not useful. Indeed, a recent study using jejunal sampling detected pathogenic bacteria in only 4 of 33 IBS patients (Gastroenterology 2003;124:S1152). However, manometric abnormalities described in IBS by the investigators involve the proximal small bowel like other dysmotility syndromes, thus it is not clear why the jejunum should not be colonized in IBS (Scand J Gastroenterol 1996;31:875–880). Reliance on lactulose breath testing as a sole test of bacterial overgrowth can be questioned. Compared with jejunal cultures, lactulose breath testing exhibits sensitivities as low as 17% in documented bacterial overgrowth (Am J Gastroenterol 1996;91:1795–1803, J Am Geriat Soc 1992;40:692–696). More troublesome is its poor specificity (<50%) showing that many positive tests do not result from overgrowth at all (Gastroenterology 1990;98:302–309). Most false positives likely are attributable to colonic fermentation of the sugar. Because lactulose is not absorbed, anyone with hydrogen-generating bacteria will produce a hydrogen peak after ingestion. Setting the standards for the test involves determining how much hydrogen is abnormal and when it should appear in the breath. In the present study, a normal lactulose breath test is defined as one with no increase in hydrogen or methane in the first 90 minutes after lactulose consumption. In studies using lactulose breath testing as a measure of transit, orocecal transit times using the same lactulose dose as in the current study (10 grams) range from 94 to 97 minutes in healthy volunteers (J Lab Clin Med 1975;85:546–555). Based on these data, one would have expected 50% of healthy controls in this study to exhibit hydrogen or methane peaks in the initial 90 minutes after lactulose instead of the observed 20%.It is unclear why symptoms responded so poorly to antibiotics in this study. In bacterial overgrowth caused by delayed transit, antibiotics produce prompt improvements in diarrhea, reductions in hydrogen, and increased weight (Gastroenterology 1999;117:794–797, Am J Gastroenterol 1999;94:1327–1331, Age Aging 1992;21:13–19, J Clin Gastroenterol 2000;31:63–66). The authors justify the 7-day follow-up interval because of their observation that breath test abnormalities recur within 2 weeks in IBS after antibiotic eradication. It is counterintuitive to postulate that healthy young IBS patients have such refractory bacterial colonization, whereas many sicker individuals with true small bowel dysmotility clear so easily. Indeed, the limited symptom responses and rapid recurrences are more plausibly explained by pathogenic organisms in the colon that would be expected to reconstitute more quickly.Comment should be made on discrepancies between the authors’ data and literature reports. In their experience, manometry findings are similar in IBS patients with diarrhea and constipation, whereas others observed increased cycling in diarrhea-predominant patients (Dig Dis Sci 2002;47:2639–2643, Dig Dis Sci 1995;40:2383–2389). This study noted similar intestinal transit times in diarrhea and constipation, whereas others measured more rapid transit in diarrhea-predominant IBS (Gut 1983;21:405–411). Finally, it is unusual that placebo reduced symptoms in only 11%–15% of patients. Most IBS drug trials report placebo responses in 30%–70% of patients—values similar to the response to neomycin in the current study (Gastroenterology 1988;95:232–241).The present study observed selective methane production only in constipated patients. Methane is produced by colonic bacteria in 40% of healthy humans. Diverticulosis and encopresis have increased methane excretion, whereas inflammatory bowel disease patients produce less methane (Gut 1986;27:698–704, Scand J Gastroenterol 1983;18:289–298, J Ped Gastroenterol Nutr 1990;10:473–477). It is uncertain if these are epiphenomena or if methane retards gut transit. Furthermore, some persons lose the capacity to generate methane, whereas others acquire this ability with time (J Clin Invest 1992;89:1304–1311, Dig Dis Sci 1994;39:494–497). It is unknown if this could be a mechanism for the alternating bowel pattern in some IBS patients.Even with limitations, this investigation presents intriguing data. This study does not prove that IBS symptoms are caused by small intestinal bacteria, but it does infer an important alteration of gut flora. It should be possible to devise an experimental technique to confirm if distal small intestinal bacterial overgrowth is present in IBS. Placement of a flexible capillary tube or other absorbent substance could facilitate collection of ileal secretions for culture. Alternatively, lactulose breath testing could be performed with an occlusive balloon in the distal ileum to prevent colonic fermentation. Finally, concurrent scintigraphy could ascertain if hydrogen and methane levels increase when the lactulose is in the ileum or after it reaches the colon. Regardless, this provocative work challenges current theories about the pathogenesis of functional bowel disorders and forces those of us in the field to consider alternate hypotheses. Evidence is accumulating that altered gut flora play pathogenic roles in irritable bowel syndrome (IBS). Hydrogen production by enteric bacteria after lactulose is increased in IBS patients (Lancet 1998;352:1187–1189). Furthermore, excluding dietary fermentable carbohydrates decreases hydrogen production. However, some report poor correlations between hydrogen excretion and symptoms (Am J Gastroenterol 1989;84:375–378). Therapies that modify gut flora may improve symptoms in some patients. Probiotics containing active cultures decrease gas-related IBS symptoms (Am J Gastroenterol 2000;95:1231–1238). Recently, metronidazole was reported to reduce symptoms by >50% in Indian IBS patients versus 25% on placebo (Indian J Gastroenterol 1997;16:137–139). One group has promoted an extension of this theory. In their initial publication in 2000, Pimentel and colleagues noted positive lactulose breath tests in 78% of IBS patients, which they attributed to small intestinal bacterial overgrowth (Am J Gastroenterol 2000;95:3503–3506). Forty-seven of 202 patients were given antibiotics and restudied. Twenty individuals normalized their breath tests with associated decreases in diarrhea and pain. More recently, positive lactulose breath tests were demonstrated in 78% of fibromyalgia patients (J Musculoskel Pain 2001;9:107–113). Antibiotics reduced joint pains and fatigue, symptoms the investigators associated with bacterial overgrowth. To define the mechanism underlying bacterial overgrowth, Pimentel et al. compared small intestinal manometric findings in IBS patients and healthy controls (Dig Dis Sci 2002;47:2639–2643). Those with IBS exhibited delayed cycling of fasting motor complexes, and complex durations were shortened by one third. It was concluded that bacterial overgrowth in IBS results from small intestinal hypomotility. The present study by Pimentel et al. is an effort to address shortcomings of their initial investigation. The efficacy of neomycin in reducing symptoms and decreasing breath hydrogen and methane excretion was compared in rigorous double blind, placebo-controlled fashion in a moderately large IBS population. Similar numbers of constipation- and diarrhea-predominant patients were included. Breath test findings in IBS patients were compared with a matched control group. As in the initial study, most IBS patients exhibited abnormal breath tests. The important new findings of this study included the demonstration that neomycin reduces IBS symptoms more than placebo and that symptom reductions correlate with breath test normalization. These findings, if validated, have enormous implications for our understanding of IBS pathophysiology and its management. If IBS symptoms largely result from bacterial colonization of a normally nearly sterile gut region, the current focus on pharmaceuticals to modulate visceral motor and sensory function would be misguided. Rather, exploration of new means of eradicating small bowel overgrowth would be warranted. The notion of small bowel bacterial fermentation as a cause of IBS symptoms is attractive because it is easy to picture hungry organisms devouring meal residues and producing gaseous and liquid by-products which then distend and irritate the distal intestine. With a condition like IBS in which true progress seems slow and incremental, one is tempted to jump on a new bandwagon that so completely rejects the existing dogma. Parallels can be drawn between this theory and early data supporting Helicobacter pylori as a cause of peptic ulcers. However, in contrast to the convincing research on H. pylori, this investigation teases the reader with methodologies that infer but do not confirm the presence of small intestinal organisms in IBS. The use of lactulose breath testing in IBS generates more questions than answers. Most importantly, what and where are the organisms that cause symptoms in IBS? The authors explain that IBS is different from other conditions with bacterial overgrowth in that only the distal small intestine is colonized, thus jejunal cultures are not useful. Indeed, a recent study using jejunal sampling detected pathogenic bacteria in only 4 of 33 IBS patients (Gastroenterology 2003;124:S1152). However, manometric abnormalities described in IBS by the investigators involve the proximal small bowel like other dysmotility syndromes, thus it is not clear why the jejunum should not be colonized in IBS (Scand J Gastroenterol 1996;31:875–880). Reliance on lactulose breath testing as a sole test of bacterial overgrowth can be questioned. Compared with jejunal cultures, lactulose breath testing exhibits sensitivities as low as 17% in documented bacterial overgrowth (Am J Gastroenterol 1996;91:1795–1803, J Am Geriat Soc 1992;40:692–696). More troublesome is its poor specificity (<50%) showing that many positive tests do not result from overgrowth at all (Gastroenterology 1990;98:302–309). Most false positives likely are attributable to colonic fermentation of the sugar. Because lactulose is not absorbed, anyone with hydrogen-generating bacteria will produce a hydrogen peak after ingestion. Setting the standards for the test involves determining how much hydrogen is abnormal and when it should appear in the breath. In the present study, a normal lactulose breath test is defined as one with no increase in hydrogen or methane in the first 90 minutes after lactulose consumption. In studies using lactulose breath testing as a measure of transit, orocecal transit times using the same lactulose dose as in the current study (10 grams) range from 94 to 97 minutes in healthy volunteers (J Lab Clin Med 1975;85:546–555). Based on these data, one would have expected 50% of healthy controls in this study to exhibit hydrogen or methane peaks in the initial 90 minutes after lactulose instead of the observed 20%. It is unclear why symptoms responded so poorly to antibiotics in this study. In bacterial overgrowth caused by delayed transit, antibiotics produce prompt improvements in diarrhea, reductions in hydrogen, and increased weight (Gastroenterology 1999;117:794–797, Am J Gastroenterol 1999;94:1327–1331, Age Aging 1992;21:13–19, J Clin Gastroenterol 2000;31:63–66). The authors justify the 7-day follow-up interval because of their observation that breath test abnormalities recur within 2 weeks in IBS after antibiotic eradication. It is counterintuitive to postulate that healthy young IBS patients have such refractory bacterial colonization, whereas many sicker individuals with true small bowel dysmotility clear so easily. Indeed, the limited symptom responses and rapid recurrences are more plausibly explained by pathogenic organisms in the colon that would be expected to reconstitute more quickly. Comment should be made on discrepancies between the authors’ data and literature reports. In their experience, manometry findings are similar in IBS patients with diarrhea and constipation, whereas others observed increased cycling in diarrhea-predominant patients (Dig Dis Sci 2002;47:2639–2643, Dig Dis Sci 1995;40:2383–2389). This study noted similar intestinal transit times in diarrhea and constipation, whereas others measured more rapid transit in diarrhea-predominant IBS (Gut 1983;21:405–411). Finally, it is unusual that placebo reduced symptoms in only 11%–15% of patients. Most IBS drug trials report placebo responses in 30%–70% of patients—values similar to the response to neomycin in the current study (Gastroenterology 1988;95:232–241). The present study observed selective methane production only in constipated patients. Methane is produced by colonic bacteria in 40% of healthy humans. Diverticulosis and encopresis have increased methane excretion, whereas inflammatory bowel disease patients produce less methane (Gut 1986;27:698–704, Scand J Gastroenterol 1983;18:289–298, J Ped Gastroenterol Nutr 1990;10:473–477). It is uncertain if these are epiphenomena or if methane retards gut transit. Furthermore, some persons lose the capacity to generate methane, whereas others acquire this ability with time (J Clin Invest 1992;89:1304–1311, Dig Dis Sci 1994;39:494–497). It is unknown if this could be a mechanism for the alternating bowel pattern in some IBS patients. Even with limitations, this investigation presents intriguing data. This study does not prove that IBS symptoms are caused by small intestinal bacteria, but it does infer an important alteration of gut flora. It should be possible to devise an experimental technique to confirm if distal small intestinal bacterial overgrowth is present in IBS. Placement of a flexible capillary tube or other absorbent substance could facilitate collection of ileal secretions for culture. Alternatively, lactulose breath testing could be performed with an occlusive balloon in the distal ileum to prevent colonic fermentation. Finally, concurrent scintigraphy could ascertain if hydrogen and methane levels increase when the lactulose is in the ileum or after it reaches the colon. Regardless, this provocative work challenges current theories about the pathogenesis of functional bowel disorders and forces those of us in the field to consider alternate hypotheses. ReplyGastroenterologyVol. 125Issue 6Preview Full-Text PDF