• breath test;
  • constipation;
  • irritable bowel syndrome;
  • methane;
  • psychological;
  • small intestinal bacterial overgrowth;
  • visceral hypersensitivity


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Abstract Small intestinal bacterial overgrowth (SIBO) has been implicated in the pathogenesis of irritable bowel syndrome (IBS), although the issue is still under debate. The aim of this study was to determine the prevalence of SIBO in those with IBS and its association with colonic motility, bowel symptoms and psychological distress. Sucrose hydrogen and methane breath tests were performed in 158 IBS patients and 34 healthy controls (HC). Thresholds for pain and urgency were tested by barostat in the descending colon. The motility index (MI) was calculated as the average area under the curve for all phasic contractions. Questionnaires assessed psychological distress, IBS symptom severity (IBS-SS), IBS quality of life (IBS-QOL) and self-reported bowel symptoms. Fifty-two of 158 (32.9%) IBS patients had abnormal breath tests compared with six of 34 (17.9%) HC (χ2 = 0.079). SIBO (SIBO+) and non-SIBO (SIBO−) patients did not differ in the prevalence of IBS subtypes, IBS-SS, IBS-QOL and psychological distress variables. IBS patients had a greater post-distension increase in MI than HC, but there was no difference between SIBO+ and SIBO− patients. Predominant methane producers had higher urge thresholds (28.4 vs 18.3, P < 0.05) and higher baseline MI (461 vs 301.45, P < 0.05) than SIBO− IBS patients, and they reported more ‘hard or lumpy stools’ when compared with predominant hydrogen producers (P < 0.05) and SIBO− IBS patients (P < 0.05). SIBO is unlikely to contribute significantly to the pathogenesis of IBS. Methane production is associated with constipation.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Small intestinal bacterial overgrowth (SIBO) is characterized by the presence of an abnormally high number of bacteria in the small bowel. SIBO may be associated with a broad range of predisposing conditions including decreased peristaltic motility in the small intestine or surgical procedures that promote stasis.1,2 The usual symptoms are diarrhoea, flatulence, abdominal pain and bloating.3 SIBO has been proposed to be common in patients with irritable bowel syndrome (IBS), with reported prevalence figures ranging from 10% to 70%.4,5 Galatola et al.6 showed that 29% of IBS patients without diarrhoea, and 56% of those with diarrhoea of obscure origin had an abnormal 1g-14C xylose breath test. However, Pimentel et al. reported a SIBO prevalence of 80% among IBS patients, and suggested that: (a) SIBO may be the aetiology for all or most IBS, and (b) that antibiotics are an effective treatment option for IBS.5 However, these assertions are controversial.7 Critics have suggested that differences as to how SIBO is diagnosed, and especially differences regarding the substrate used, may account for inconsistencies between studies.

In early studies of SIBO, only excess hydrogen in breath following ingestion of sugar was tested, and those studies usually reported diarrhoea to be the primary symptom. However, recently, researchers have also tested for elevated methane levels and found that this gas, which is liberated by anaerobic bacteria, is more often associated with constipation. Pimentel et al. showed in animal model studies that exogenous methane slows transit and augments non-propagating contractions in the small bowel.8 Further studies are needed to evaluate methane’s effect on colonic motility and its association with constipation.

Pimentel et al. recommend that IBS be treated with broad-spectrum antibiotics.9 However, treatment with antibiotics is not completely safe: antibiotics may worsen symptoms of IBS by altering the delicate balance between ‘good’ and ‘bad’ (e.g. probiotic vs harmful) bacteria in the gut.10 Whether antibiotic choices should differ for patients who are predominant hydrogen vs predominant methane producers remains to be determined.

No study has looked at psychological distress as a moderating variable in SIBO patients. However, some studies revealed increased number of positive breath tests in fibromyalgia patients, which may indicate an association between psychological distress factors like somatization and SIBO.11

The aims of this study were to: (a) determine the prevalence of SIBO in a large sample of patients with IBS and in controls. (b) Compare IBS patients with and without SIBO for differences in clinical symptoms and physiological and psychological variables. The clinical parameters tested were IBS subtype, IBS quality of life and IBS symptom severity. The physiological variables studied were pain and urge thresholds, and phasic contractions under conditions of fasting, response to intraluminal distension and response to a high-fat meal. Psychological variables studied were anxiety, depression and somatization. (c) Subdivide IBS patients with SIBO into predominant methane producers (PMP) and predominant hydrogen producers (PHP) and compare these groups with respect to predominant IBS subtype, IBS symptom severity and quality of life impact, pain and urge thresholds, and phasic contractions of the descending colon.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References


Subjects were recruited by advertizement or physician referrals and screened by telephonic interviews. The study population included 158 patients with physician diagnoses of IBS confirmed by Rome II criteria. Patients who conform to Rome II criteria have abdominal pain or discomfort at least 25% of weeks in the preceding 12 months, plus two or more of the following: (1) relief of pain or discomfort with defecation; (2) onset of discomfort associated with a change in frequency of stool; and/or (3) onset of discomfort associated with a change in form (appearance) of stool.12 These subjects had no history of gastrointestinal resection (other than appendectomy, or cholecystectomy), inflammatory bowel disease, heart disease or diabetes mellitus, and they were not pregnant at the time of study.

The control population included 34 subjects without any significant gastrointestinal problems. Exclusion criteria were average stool frequency less than 3 week−1 or more than 3 day−1, abdominal pain or use of a laxative or anti-diarrhoeal agent on more than two occasions over the previous year, history of alcohol or substance abuse, a psychiatric diagnosis, or any of the medical conditions listed above for the IBS patients. The study was approved by the Institutional Review Board of the University of North Carolina (UNC) and all subjects provided written informed consent prior to participation.

Study design

Subjects were admitted to the General Clinical Research Center for a 24–30 h period. They were asked to fast for at least 4 h prior to admission. On the day of admission, a medical history, physical examination and hydrogen breath test (HBT) for SIBO were performed. The symptom severity scale, quality of life scale and psychological questionnaires were completed during the breath test. At the end of day 1, bowel preparation for next day’s procedure with 1.5 oz of Fleet’s phosphosoda, repeated twice, was performed. On day 2, physiological testing was performed (see Colonic sensory and motility testing below).


The Irritable Bowel Syndrome Quality of Life Scale13 (IBS-QOL) is a validated, disease-specific scale to measure the impact of IBS symptoms on quality of life. It is a 34-item self-report questionnaire with eight subscales: dysphoria, interference with activity, body image, health worry, food avoidance, social reaction and sexual relationship. Subjects are asked to answer questions about their feelings over the past 30 days on an ordinal scale (1 – not at all to 5 – a great deal).

The Irritable Bowel Severity Scoring System14 (IBS-SS) is a validated scale for measuring the overall severity of IBS symptoms. It consists of five equally weighted questions about symptoms occurring in the past 10 days: the average intensity of abdominal pain, the number of days with any abdominal pain, the average severity of abdominal distension, dissatisfaction with bowel habits, and the degree to which bowel symptoms interfered with usual activities. Responses to all except the pain frequency question were on a 0–100 numeric scale (‘none’ to ‘worst ever’), and the number of days of pain in the past 10 days was multiplied by 10 to arrive at a 0–100 score for this item. The scores on these five questions were added to arrive at a total score of 0–500.

The Brief Symptom Inventory-18 (BSI-18)15 is a validated questionnaire for measuring the degree of psychological distress over the past week. Subjects are asked how much they were bothered by each of 18 symptoms, and they respond on a five-point ordinal scale (‘not bothered at all’ to ‘extremely bothered’). The BSI includes three subscales – anxiety, depression and somatization – and a Global Symptom Index (GSI) scale. Sum scores of each subscale and the GSI were converted to standardized scores where the mean for the healthy population is a score of 50 and each standard deviation from the mean is equal to 10 scale points; thus a score of 60 is one standard deviation above the mean. These standardized scores adjust for sex differences in the reporting of psychological symptoms, so that BSI-18 scores for males and females can be pooled together.

The Recent Physical Symptom Questionnaire16 (RPSQ) is an IBS-specific somatization scale. It consists of 26 items to which subjects respond on a five-point frequency scale, describing how often each symptom occurred in the last month. The Catastrophizing Scale17 measures negative self-statements and catastrophizing thoughts related to pain.

Subjective assessment of bowel patterns was also performed by means of questions adopted from the Rome III modular questionnaire18–‘In last three months how often did you have: less than 3 bowel movements (BM) (0–2) in a week, more than 3 BM (≥4) in a day, hard or lumpy stools, or loose, mushy or watery stools’.


The barostat is a computer-controlled pump (Distender II model; G&J Electronics, Willodale, ON, Canada) used for testing sensory thresholds in the gut lumen. It inflates a plastic bag to a predefined pressure and holds this pressure constant for a fixed period of time by adding or subtracting air at 38 mL s−1. Volumes and pressures are recorded 16 times per second and are displayed graphically in real time.

The motility catheter (Model C7-CB-0026; Mui Scientific, Mississauga, ON, Canada) is 5-mm-wide bundle of smaller polyethylene tubes bonded together. It includes a central lumen which accommodates a guide wire, two lumens which open inside the bag (one to inflate/deflate the bag and a second to monitor pressure inside the bag), plus four small catheters used to measure pressures 2.5 and 5 cm from the proximal and distal edges of the bag. A disposable, 10-cm-long, 600-mL-capacity polyethylene bag (Model CT-BP600R, Mui Scientific) was attached to the surface of the motility catheter.

The pneumohydraulic pump (eight-channel hydraulic capillary infusion system; Arndorfer Inc., Greendale, WI, USA) uses a tank of compressed air to force de-gassed sterile water from a reservoir through four capillary catheters which are connected to four pressure transducers. These pressure transducers are also connected to the four small catheters in the motility catheter which open above and below the barostat bag. Water is perfused through the pressure transducers and the perfusion catheters at a rate of 0.37 mL min−1. A continuous column of water connecting the pressure transducer to the openings on the outside of the motility catheter allows transmission of pressure changes occurring at the catheter openings to the pressure transducers. The outputs of these pressure transducers were continuously recorded (see below).

A physiological recorder (Synectics Polygram; Medtronic Inc., Minneapolis, MN, USA) was used to record phasic motility changes above and below the balloon and store them in a digital file. A research nurse marked these recordings to indicate which experimental condition was in effect.

Colonic sensory and motility testing

All physiological and sensory testing was performed on day 2. A flexible sigmoidoscope was used to guide the placement of the barostat catheter over the guide wire in the descending colon and the position was confirmed by fluoroscopy. Following this, the subject rested for 90 min before testing began. No sedation was used during sigmoidoscopy. Subjects were not permitted to have food until the test meal.

Sample distension and determination of individual operating pressure  Subjects were instructed to give separate ratings of the intensity of pain and urgency to defecate experienced at the end of each distension, on a six-point scale (0 = no sensation; 1 = weak; 2 = mild; 3 = moderate; 4 = strong; 5 = intense). The scale was visible to subjects during the procedure. Sample distensions were then performed during which the barostat bag was inflated in a stepwise fashion by increasing pressure by 4 mmHg every 15 s until the subject reported moderate pain (rating of 3). These distensions were done to: (a) ensure that the barostat bag was unfolded; (b) teach the subject how to use the rating scale; and (c) lower anticipatory anxiety. The barostat bag was then slowly inflated with 30 mL of air and the pressure was allowed to equilibrate for 3 min. The average bag pressure during the last 15 s defined the individual operating pressure (IOP),19 which is the minimum pressure required to overcome mechanical forces and inflate the bag with 30 mL of air. All sensory and motility testing was performed with the subject in prone position to avoid overlying body tissues from compressing the bowel.

Ascending method of limits  Pain thresholds in the colon were assessed using the ascending method of limits (AML).19 Phasic distensions were 30 s long, separated by 30-s rest intervals starting at the IOP and progressively increasing in 2 mmHg steps until either the subject requested the research nurse to stop the protocol or 48 mmHg was reached. The pain threshold was defined as the amount of pressure above the IOP at which the subject first reported moderate pain (absolute distending pressure minus the IOP). If the subject reached 48 mmHg without reporting moderate pain, then the pain threshold was defined as 50 mmHg minus the IOP. After measuring pain thresholds, there was a 15 min rest period.

Colonic phasic motility  Phasic contractions were measured from the perfusion ports above and below the bag under the following conditions: (a) during fasting baseline for 10 min at the IOP, (b) during distension for 10 min at a pressure of IOP + 20 mmHg, (c) during a recovery period after intraluminal distension for 15 min at the IOP, and (d) following the meal for 30 min at the IOP. These tracings were visually screened to exclude artefacts (i.e. wave amplitudes <5 mmHg or durations <6 s). The beginning and ending inflection points for each individual contraction were identified visually and the area under the curve was calculated using computer software (Polygram, Lower GI Edition, Version 5.06; Synectics Medical, now Medtronic, Inc.). These areas were added together, then divided by recording time in seconds (excluding the time occupied by movement artefact), and multiplied by 100. The motility index (MI) was the average of phasic contractions at four perfusion ports.

Postprandial motility  The standardized test meal contained 810 kcal and 38 g of fat to be consumed within 10 min. Immediately after completing the meal, the patient returned to the prone position and bag pressure was maintained at the IOP for 30 min (the postprandial period). Phasic motility was recorded throughout this period.

Hydrogen breath test

To minimize basal hydrogen production, subjects were asked to have a carbohydrate-restricted dinner a day before and fasted overnight prior to the hydrogen breath test (HBT). Smoking and physical exercise were not allowed 2 h prior to and during the test. A baseline end-expiratory breath sample was obtained prior to ingestion of 50 g of sucrose (dissolved in 250 mL of water). Additional breath samples were obtained every 15 min for 2 h. The samples were analysed for hydrogen and methane concentrations in parts per million (ppm) using a Quintron microanalyser (Quintron Instrument Company, Milwaukee, WI, USA). Breath tests were considered to indicate SIBO based on the following criteria.20

  • 1
    A baseline breath hydrogen of ≥20 parts per million (ppm) (elevated methane levels at baseline was not interpreted as evidence of SIBO because this most likely reflects colonic bacterial fermentation.).
  • 2
    An increase of breath hydrogen and/or methane of ≥12 ppm above the baseline that occurs within the first 60 min.
  • 3
    If the first peak (≥12 above baseline) occurs more than 60 min after ingestion of sucrose and is followed by a second peak after a gap of at least 15 min. The second peak should be ≥20 ppm above the baseline and should be clearly separated by a dip in concentration from the first peak.

Subjects with IBS with a positive breath test were further subdivided into predominantly hydrogen producers (PHP) and predominantly methane producers (PMP) based on whether changes in hydrogen or methane were used to diagnose SIBO. As criterion 1 involves only hydrogen, patients with SIBO diagnosed with that criterion were classified as PHP.

Data and statistical analysis

Residual plots of all data sets were examined to decide whether parametric or non-parametric procedures were appropriate. Summary statistics of normally distributed data are expressed as means and standard deviations (SD). Student t-tests were used to compare the groups. Analysis of variance was performed if more than two groups were compared. For all analyses, a P-value of <0.05 without adjustment for multiple comparisons defined statistical significance. Categorical data were analysed using chi-squared test with continuity correction.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

One hundred and fifty-eight IBS patients (25 males; mean age 34.6 years) and 34 healthy controls (HC) (eight males; 37.2 years) underwent colonic sensory, motility studies and hydrogen breath testing. Table 1 shows the demographic characteristics of the sample. There were no significant differences between IBS patients and HC in gender, age or race.

Table 1.   Demographics, IBS symptom severity and psychological distress scales
 IBS (n = 158)Controls (n = 34)
SIBO (n = 52)Non-SIBO (n = 106)
  1. Data are shown as mean ± SEM.

  2. *P < 0.05, **P < 0.01, ***P < 0.1 compared with controls.

  3. P < 0.1 compared with non-SIBO.

  4. SIBO, small intestinal bacterial overgrowth; IBS-SS, IBS severity scale; RPSQ somatization, recent physician symptom questionnaire somatization; BSI-18, Brief Symptom Index-18.

Age (yr)32.73 ± 1.5036.59 ± 1.1637.21 ± 2.18
 Female 44 8926
 Male  8 17 8
 White 38 7917
 Non-white 14 2717
IBS subtype
 IBS-D 16 31
 IBS-C 16 27
 IBS-M 20 48
 Overall272.94 ± 13.88**278.45 ± 9.58**25.54 ± 8.6
 Pain severity42.78 ± 3.65**49.03 ± 2.53**4.11 ± 2.09
 Pain frequency4.67 ± 0.37**4.78 ± 0.23**0.42 ± 0.20
 Distension46.84 ± 4.37**47.43 ± 2.94**2.87 ± 1.34
 Interference55.48 ± 4.22**58.07 ± 2.78**2.44 ± 1.64
 Bowel dissatisfaction18.90 ± 3.10**24.23 ± 2.49**83.67 ± 5.18
IBS-QOL69.72 ± 2.55**69.23 ± 2.00**99.11 ± 0.49
RPSQ somatization11.10 ± 0.62**10.5 ± 0.47**4.70 ± 0.72
Catastrophizing7.60 ± 0.94**7.30 ± 0.70**2.25 ± 0.69
 Global severity54.53 ± 1.24**51.77 ± 0.84**41.70 ± 1.37
 Somatization55.77 ± 0.98**53.29 ± 0.78**44.24 ± 1.07
 Depression52.46 ± 1.62**51.20 ± 1.04**45.82 ± 1.29
 Anxiety52.82 ± 1.44**50.31 ± 0.88**42.72 ± 1.08

IBS patients vs healthy controls

The prevalence of SIBO was 32.9% in IBS patients (52/158) and 17.9% (6/34) in HC (Pearson’s χ2 = 0.079). In IBS subjects who were diagnosed as having SIBO, 48.0% (25/52) fulfilled criterion 1 for diagnosis, 46.2% (24/52) criterion 2, and 5.8% (3/52) criterion 3 (see Hydrogen breath test above). In controls, 50% (3/6) were diagnosed with SIBO with both criteria 1 and 2. Overall, 48.3% (28/58) of patients with SIBO were diagnosed using criterion 1, 46.5% (27/58) using criteria 2 and 5.2% (3/58) using criterion 3.

The sample size of IBS (158) vs HC (34) achieved 74% power and sample size of SIBO+ (52) vs SIBO− (106) achieved 84% power to detect 0.5 standard deviation with a significance level (alpha) of 0.05 using a two-sided t-test.


Clinical symptoms  The distribution of IBS subtypes in the SIBO+ population was 30.8% IBS-D, 30.8% IBS-C and 38.5% IBS-M; and in the SIBO− group there were 29.2% IBS-D, 25.5% IBS-C and 45.3% IBS-M. There were no significant differences in the prevalence of various IBS subtypes in SIBO+ or SIBO− IBS populations.

Patients with IBS scored significantly higher than HC on the IBS-SS. However, there were no significant differences between SIBO+ and SIBO− patients on the IBS-SS. Both SIBO+ and SIBO− IBS groups scored lower on the IBS-QOL than HC, but there were no significant differences between these two groups (Table 1).

Psychological determinants  As shown in Table 1, IBS patients scored significantly higher than HC on all subscales of the BSI-18. SIBO+ IBS tended to score higher than SIBO− IBS patients on the Global Symptom Index and somatization subscale (P < 0.10), but this did not lead to the achievement of statistical significance. There were no significant differences on the RPSQ or the Catastrophizing Scale between SIBO+ and SIBO− IBS patients; however, both scored significantly higher than HC (Table 1).

Physiological parameters  SIBO+ and SIBO− IBS groups had lower AML pain thresholds (IOP-corrected) compared with HC (Table 2). There were no significant differences between SIBO+ and SIBO− IBS groups, although there was a trend (P < 0.10) favouring lower pain thresholds in SIBO− compared with SIBO+ subjects. SIBO+ and SIBO− patients also had significantly lower urge thresholds compared with HC, but there was no statistically significant difference between the SIBO+ and SIBO− IBS groups.

Table 2.   Sensory thresholds and colonic motility responses
  1. Data are shown as mean ± SEM.

  2. *P < 0.05, **P < 0.01, ***P < 0.1 compared with controls.

  3. P < 0.1 compared with non-SIBO.

  4. SIBO, small intestinal bacterial overgrowth.

Pain threshold (mmHg)30.08 ± 1.52*25.95 ± 1.32**36.61 ± 1.82
Urge thresholds (mmHg)21.36 ± 1.73**18.27 ± 1.24**29.85 ± 2.57
Motility Index
 Baseline361.61 ± 37.87***301.45 ± 20.97264.63 ± 30.58
 Distension987.22 ± 114.02**943.13 ± 89.59*577.98 ± 71.65
 Recovery463.31 ± 48.90*469.61 ± 49.19***316.45 ± 28.14
 Post-meal505 ± 47.43474.99 ± 30.37476.16 ± 54.38
 Reactivity625.61 ± 113.48*642.87 ± 84.41*313.34 ± 75.46

Phasic colonic motility  There was no difference in the MI (phasic contractions) at baseline (fasting, no intraluminal distension) between SIBO+ and SIBO− IBS subjects and none when compared with HC (Table 2). In response to intraluminal distension, both IBS and HC showed a significant increase in phasic contraction compared with their respective baselines; this is referred to in Table 2 as ‘reactivity’. Reactivity was greater in both SIBO+ and SIBO− IBS patients compared with HC (P < 0.05), but there were no significant differences in reactivity between SIBO+ and SIBO− IBS groups. During the recovery period following intraluminal distension, MI declined nearly to baseline levels in all groups. The SIBO+ group showed a higher MI during recovery than HC but was not significantly different from SIBO− IBS. All groups showed significant increases in MI following the meal, but the magnitude of the meal-stimulated increase was not different in IBS vs HC nor was there a significant difference between SIBO+ and SIBO− IBS groups (Table 2).


The distribution of bowel habit subtypes was 38.1% IBS-D, 23.8% IBS-C and 38.1% IBS-M in PHP compared with 0% IBS-D, 60% IBS-C and 40% IBS-M in PMP. There were no significant differences among PHP, PMP and SIBO− IBS patients in the IBS-SS total score or individual symptoms (Table 3).

Table 3.   IBS Subtype, Symptom severity and quality of life of PHPs & PMPs
 PHP (SIBO-IBS) (n = 42)PMP (SIBO-IBS) (n = 10)Non-SIBO IBS (n = 106)
  1. Data are shown as mean ± SEM.

  2. P < 0.1 compared with non-SIBO.

  3. PHP, predominant hydrogen producers; PMP, predominant methane producers; SIBO, small intestinal bacterial overgrowth; IBS (D – diarrhoea; C – constipation; M – mixed), QOL, quality of life.

IBS subtype
 Overall269.35 ± 13.61288.00 ± 45.83278.45 ± 9.58
 Pain severity40.47 ± 3.6152.50 ± 11.4549.03 ± 2.53
 Pain frequency4.55 ± 0.405.20 ± 0.994.78 ± 0.23
 Distension46.33 ± 4.5949.00 ± 12.6047.43 ± 2.94
 Interference55.35 ± 4.4956.00 ± 11.8258.07 ± 2.78
 Bowel dissatisfaction18.28 ± 3.0321.50 ± 10.3024.23 ± 2.49
QOL70.51 ± 2.6766.43 ± 7.2969.23 ± 2.00

As shown in Table 4, AML pain thresholds were similar among PMP, PHP and SIBO− IBS patients. However, the AML urge threshold was higher in PMP than in SIBO− IBS, which is in keeping with this being a predominantly constipated group.

Table 4.   Sensory thresholds and colonic motility responses of PHPs and PMPs
  1. Data are shown as mean ± SEM.

  2. *P < 0.05, **P < 0.01, ***P < 0.1 compared with non-SIBO.

  3. P < 0.1 compared with PHP-SIBO.

  4. PHP, predominant hydrogen producers; PMP, predominant methane producers; SIBO, small intestinal bacterial overgrowth.

Pain threshold (mmHg)29.62 ± 1.6532.00 ± 3.9525.95 ± 1.32
Urge threshold (mmHg)19.54 ± 1.7928.44 ± 4.27*18.27 ± 1.24
Motility Index
 Baseline337.79 ± 45.03461.64 ± 45.85*301.45 ± 20.97
 Distension926.05 ± 110.571244.14 ± 374.28943.13 ± 89.59
 Recovery463.31 ± 57.21463.32 ± 88.47469.61 ± 49.19
 Post-meal497.96 ± 56.13536.43 ± 78.78474.99 ± 30.37
 Reactivity588.26 ± 113.37782.50 ± 360.78642.87 ± 84.41

Motility Indices  As shown in Table 4, the baseline MI was significantly higher in PMP patients compared with that in SIBO− IBS patients. However, PHP and PMP were not significantly different from each other during baseline, intraluminal distension, recovery from distension, or the postprandial period.

Bowel pattern  As shown in Table 5, PMP subjects were significantly more likely to report hard or lumpy stools compared with PHP and SIBO− IBS patients. There was also a trend for PMP to report more often that their bowel movements occurred less than three times per week compared with PHP and SIBO− IBS patients (P = 0.055). Self-described bowel habits in the PHP group did not appear to differ from those in the SIBO− IBS group.

Table 5.   Subjective bowel patterns of PHPs and PMPs
  1. Data are shown as mean ± SEM.

  2. *P < 0.05, **P < 0.01, ***P < 0.1 compared with non-SIBO and PHP-SIBO.

  3. PHP, predominant hydrogen producers; PMP, predominant methane producers; SIBO, small intestinal bacterial overgrowth.

Less than three bowel movements in a week (0–2)0.36 ± 0.070.70 ± 0.15***0.39 ± 0.04
More than three bowel movements in a week (≥4)0.40 ± 0.070.40 ± 0.160.47 ± 0.04
Hard and lumpy stools0.52 ± 0.070.90 ± 0.10*0.53 ± 0.04
Loose, mushy and watery stools0.71 ± 0.070.50 ± 0.160.71 ± 0.04

Relationship between gas production and IBS-SS  The IBS-SS scale was not significantly correlated with peak levels of hydrogen in the PHP group (rho = −0.06) or with peak methane production in the PMP group (rho = −0.11). Similarly, the IBS-SS was not significantly correlated with the area under the curve for hydrogen in PHP patients or with area under the curve for methane in the PMP group.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

These data fail to confirm that the prevalence of SIBO is higher in IBS patients (32.9%) compared with that in HC (17.9%), although there was a trend in this direction (P = 0.079). We also observed a lower prevalence of SIBO in our IBS population when compared with some other studies, probably due to differences in the assessment method and criteria used to define SIBO.21

We used sucrose as the substrate for HBT which has advantages over the popularly used lactulose. Sucrose is broken down into glucose and fructose. Glucose is completely absorbed in the small intestine and fructose is also absorbed through GLUT-5-mediated transport in the jejunum. The glucose breath test is more specific (78–83%vs 44–70%) and more sensitive (62–93%vs 17–68%) than the lactulose breath test.2,22 Consequently, any peak in hydrogen and methane levels after sucrose ingestion is considered diagnostic of SIBO. However with lactulose, because it is not absorbed in the small intestine, the timing and amplitude of the peaks (small intestinal vs colonic) have to be considered, which reduces the inter-observer agreement from a κ-value of 1.0 (with glucose) to 0.86.20 A disadvantage of sucrose (used in our study) is that there can be a delayed peak in hydrogen or methane because of either sucrose malabsorption as seen individuals with sucrase deficiency23,24 or rarely in fructose malabsorption.25,26 However, fructose in the form of sucrose is better absorbed because of facilitation by glucose.27 Hence, it makes it important to differentiate a second colonic peak to prevent misdiagnosing sucrose or fructose intolerance as SIBO. However, sucrose is unlike lactulose, which will always have a colonic peak that needs to be separated from small intestinal peak and that may give a higher number of false positives. This is evident by the fact that only 5.2% (3/58) of total SIBO population was diagnosed using the double peak criteria (criteria 3: see Hydrogen breath testing above) in our study.

The second goal of our study was to compare SIBO+ IBS patients with SIBO− IBS patients to determine whether SIBO is associated with specific bowel symptoms and specific physiological mechanisms. Although diarrhoea is often regarded as the hallmark symptom of SIBO,3 we found no differences between SIBO+ and SIBO− IBS patients in the proportion who met criteria for IBS-D (30.8%vs 29.2%). However, with the recent recognition of methane in SIBO and its association with constipation, the association of SIBO with only diarrhoeal symptoms is being questioned.28 Another symptom frequently thought to be associated with SIBO is bloating and distension. However, SIBO+ did not differ from SIBO− IBS patients on the abdominal distension scale of the IBS-SS (46.8 vs 49.9), nor did they differ in overall IBS-SS (272.9 vs 278.4).

Our IBS patients scored significantly higher on psychological distress variables including anxiety, depression, somatization and catastrophizing compared with HC. However, there were no differences in psychological distress between SIBO+ and SIBO− patients. Therefore, it is unlikely that psychological distress could mediate the association between SIBO and bowel symptoms.

We anticipated that SIBO+ IBS patients might have lower pain thresholds than IBS− patients because SIBO is a type of bacterial infection, and bacterial infections that produce inflammation have been shown to lower sensory thresholds.29 However, our analysis showed an opposite trend: SIBO+ IBS patients tend to have higher AML pain thresholds compared with SIBO− IBS patients. (The AML pain and urge thresholds of both SIBO+ and SIBO− IBS were significantly lower than HC).

Results have been inconsistent as to whether IBS patients have more phasic contractions than HC under baseline conditions (i.e. fasting and no stimulation by distension, stress or exogenous hormones): there have been reports that IBS patients show more baseline motility,30 less motility,31 and about the same amount32 compared with HC. We found no difference when comparing HC with either SIBO+ or SIBO− IBS patients.

Sustained intraluminal distension by a balloon simulates physiological distension of the colon by faecal material or gas. We observed an increase in phasic motility in response to this distension stimulus in both HC and IBS patients with a significantly greater increase in IBS patients when compared with HC. This is consistent with our previous observations and defines the ‘hyper-reactivity’ characteristic of IBS patients.33,34 However, SIBO+ and SIBO− patients showed similar increases in phasic contractions in response to intraluminal distension. This response is reversible, with MI returning to baseline when the distending stimulus is removed in SIBO+, SIBO− IBS patients as well as HC.

Eating also stimulates an increase in the MI in both HC and IBS, as has been shown in other studies.35 However, there were no differences in the post-meal MI increase between our IBS patients and HC, or between SIBO+ and SIBO− IBS patients.

A third goal of our study was to compare methane producers with hydrogen producers. Methane production is limited to only a few species of bacteria, and two of these methanogens, Methanobrevibacter smithii and Methanobacterium ruminatum, are only found in the left colon, which is the major site of methane production. Other anaerobic species, Bacteroides and Clostridium, can also liberate methane and could easily be part of the small intestinal flora. We showed a 19% prevalence of PMP in our SIBO+ IBS population. There were no IBS-D patients in the PMP subgroup, and most PMP patients had IBS-C. We also found that PMP patients were more likely to report hard/lumpy stools and stools less often than 3 week−1. This is in accordance with the emerging concept of methane production in SIBO, and its subjective as well as objective association with constipation.28

Methane has been shown to be associated with increased numbers of small intestinal non-propagating contractions in studies of animal models.8 Increased numbers of non-propagating contractions and/or decreased numbers of migrating motor complexes, which normally sweep bacteria and food residues out of the small intestine during the inter-digestive period, could permit bacterial overgrowth in the small intestine.36 Our data support the observations reported in animals by showing higher baseline colonic phasic contractions (which are primarily non-propagating contractions) in PMP. This would predispose patients towards further bacterial overgrowth. Further research is needed to characterize the small intestinal and colonic motility of these different gas producers and their association with clinical symptoms and predominant bowel patterns.

Our data also show that PMP patients have higher AML urge thresholds than PHP or SIBO- IBS subjects. This is a novel finding. The pain thresholds of PHP and PMP were, however, similar to each other.

There was no significant correlation between peak hydrogen and/or methane production during the HBT, and the IBS-SS or its subscales. This further suggests that IBS and SIBO are independent dimensions, and those individuals producing more gas do not as a group have greater severity of IBS symptoms.

There has been a significant controversy regarding the prevalence of SIBO in IBS, the means of testing it and its importance in the overall symptomatology of IBS patients. Pimentel et al. proposed diagnosing SIBO by means of the lactulose HBT and reported positive tests in 78–84% of patients with IBS. However, more recent reports using jejunal cultures found SIBO in only 4% of the IBS population and detected no differences from a HC population.37 Lack of uniformity in the tests used to diagnose SIBO has led to significant controversy around the whole topic. Our studies looking at demographic, symptomatic, psychological and physiologic parameters have failed to suggest that SIBO+ individuals are distinct, thereby supporting the notion that SIBO in IBS is either an artefact of testing or an epiphenomenon. The methane–IBS C association is vindicated and is the likely explanation for any trends towards differences between SIBO+ and SIBO− individuals. It becomes important to understand and correctly diagnose SIBO to prevent exposure to unnecessary antibiotic therapy which often can be a cause of perpetuation of gastrointestinal symptoms.

Study limitations

There can be alternative explanations for a positive HBT other than SIBO. Poor adherence to the breath testing protocol, especially dietary and smoking restrictions prior to the test, can lead to high baseline breath hydrogen levels.38 Forty-eight per cent of positive tests in our study were due to high baseline levels. In addition, the sucrose HBT may rarely give false-positive peaks because sucrose is broken down into glucose and fructose, and it is possible for either sucrose or fructose to give rise to a second colonic peak in hydrogen or methane if it passes through the small intestine unabsorbed. However, only three of 58 SIBO+ patients in our study showed a delayed (>60 min) rise in hydrogen or methane levels followed by a second peak. Thus, the sucrose HBT appears to be significantly more sensitive and specific for detection of SIBO in the proximal small intestine than is the lactulose HBT. A second limitation is that the use of an invasive test protocol (e.g. barostat testing) may have biased recruitment, as is suggested by the relatively young age of our sample. However, there is no reason to believe that the association between SIBO and IBS differs in young vs older patients. Thirdly, the physiologic testing was performed in the colon and was not designed to define effects of SIBO, but rather, to seek for evidence of overall differences; it is possible that SIBO is related more to motility changes in the small intestine than to motility changes in the colon. Further research into the link between SIBO and small intestinal motility is needed.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This study was supported by grants RO1 DK31369, R24 DK067674 and MO1 RR00046.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Gasbarrini A, Lauritano EC, Gabrielli M et al. Small intestinal bacterial overgrowth: diagnosis and treatment. Dig Dis 2007; 25: 23740.
  • 2
    Corazza GR, Menozzi MG, Strocchi A et al. The diagnosis of small bowel bacterial overgrowth. Reliability of jejunal culture and inadequacy of breath hydrogen testing. Gastroenterology 1990; 98: 3029.
  • 3
    Quigley EM, Quera R. Small intestinal bacterial overgrowth: roles of antibiotics, prebiotics, and probiotics. Gastroenterology 2006; 130(2 Suppl 1): S78S90.
  • 4
    Lin HC. Small intestinal bacterial overgrowth: a framework for understanding irritable bowel syndrome. JAMA 2004; 292: 8528.
  • 5
    Pimentel M, Park S, Mirocha J, Kane SV, Kong Y. The effect of a nonabsorbed oral antibiotic (rifaximin) on the symptoms of the irritable bowel syndrome: a randomized trial. Ann Intern Med 2006; 145: 55763.
  • 6
    Galatola G, Grosso M, Barlotta A et al. [Diagnosis of bacterial contamination of the small intestine using the 1g [14C] xylose breath test in various gastrointestinal diseases]. Minerva Gastroenterol Dietol 1991; 37: 16975.
  • 7
    Drossman DA. Treatment for bacterial overgrowth in the irritable bowel syndrome. Ann Intern Med 2006; 145: 6268.
  • 8
    Pimentel M, Lin HC, Enayati P et al. Methane, a gas produced by enteric bacteria, slows intestinal transit and augments small intestinal contractile activity. Am J Physiol Gastrointest Liver Physiol 2006; 290: G108995.
  • 9
    Corazza GR, Di SM, Scarpignato C. Treatment of functional bowel disorders: is there room for antibiotics? Digestion 2006; 73(Suppl 1): 3846.
  • 10
    Maxwell PR, Rink E, Kumar D, Mendall MA. Antibiotics increase functional abdominal symptoms. Am J Gastroenterol 2002; 97: 1048.
    Direct Link:
  • 11
    Pimentel M, Wallace D, Hallegua D et al. A link between irritable bowel syndrome and fibromyalgia may be related to findings on lactulose breath testing. Ann Rheum Dis 2004; 63: 4502.
  • 12
    Thompson WG, Longstreth G, Drossman DA, Heaton K, Irvine EJ, Muller-Lissner S. Functional bowel disorders and functional abdominal pain. In: DrossmanDA, CorazziariE, TalleyNJ, ThompsonWG, WhiteheadWE, eds. Rome II: The Functional Gastrointestinal Disorders. 2nd edn. McLean, VA: Degnon Associates, 2000: pp. 351432.
  • 13
    Patrick DL, Drossman DA, Frederick IO, Dicesare J, Puder KL. Quality of life in persons with irritable bowel syndrome – development and validation of a new measure. Dig Dis Sci 1998; 43: 40011.
  • 14
    Francis CY, Morris J, Whorwell PJ. The irritable bowel severity scoring system: a simple method of monitoring irritable bowel syndrome and its progress. Aliment Pharmacol Ther 1997; 11: 395402.
  • 15
    Derogatis LR. BSI 18 Brief Symptom Inventory 18: Administration, Scoring, and Procedures Manual. Minneapolis, MN: NCS Pearson, Inc., 2000.
  • 16
    Palsson OS, Jones KR, Turner MJ, Drossman DA, Whitehead WE. Development of specific questionnaires to assess somatization and comorbid diagnoses in irritable bowel syndrome (IBS). Gastroenterology 2002; 122(Suppl 1): A502.
  • 17
    Keefe FJ, Brown GK, Wallston KA, Caldwell DS. Coping with rheumatoid arthritis pain: catastrophizing as a maladaptive strategy. Pain 1989; 37: 516.
  • 18
    Longstreth GF, Thompson WG, Houghton LA, Mearin F, Spiller RC. Functional bowel disorders. In: DrossmanDA, CorazziariE, DelvauxM, SpillerRC, TalleyNJ, ThompsonWG, WhiteheadWE, eds. Rome III: The Functional Gastrointestinal Disorders. 3rd edn. McLean, VA: Degnon Associates; 2006. pp. 487555.
  • 19
    Whitehead WE, Delvaux M. Standardization of barostat procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. The Working Team of Glaxo-Wellcome Research, UK. Dig Dis Sci 1997; 42: 22341.
  • 20
    Romagnuolo J, Schiller D, Bailey RJ. Using breath tests wisely in a gastroenterology practice: an evidence-based review of indications and pitfalls in interpretation. Am J Gastroenterol 2002; 97: 111326.
    Direct Link:
  • 21
    Pimentel M, Chow EJ, Lin HC. Eradication of small intestinal bacterial overgrowth reduces symptoms of irritable bowel syndrome. Am J Gastroenterol 2000; 95: 35036.
    Direct Link:
  • 22
    Riordan SM, McIver CJ, Walker BM, Duncombe VM, Bolin TD, Thomas MC. The lactulose breath hydrogen test and small intestinal bacterial overgrowth. Am J Gastroenterol 1996; 91: 1795803.
  • 23
    Davidson GP, Robb TA, Kirubakaran CP. Bacterial contamination of the small intestine as an important cause of chronic diarrhea and abdominal pain: diagnosis by breath hydrogen test. Pediatrics 1984; 74: 22935.
  • 24
    Lembcke B, Folsch UR, Caspary WF, Ebert R, Creutzfeldt W. Influence of metronidazole on the breath hydrogen response and symptoms in acarbose-induced malabsorption of sucrose. Digestion 1982; 25: 18693.
  • 25
    Kneepkens CM, Jakobs C, Douwes AC. Apple juice, fructose, and chronic nonspecific diarrhoea. Eur J Pediatr 1989; 148: 5713.
  • 26
    Rumessen JJ. Fructose and related food carbohydrates. Sources, intake, absorption, and clinical implications. Scand J Gastroenterol 1992; 27: 81928.
  • 27
    Truswell AS, Seach JM, Thorburn AW. Incomplete absorption of pure fructose in healthy subjects and the facilitating effect of glucose. Am J Clin Nutr 1988; 48: 142430.
  • 28
    Chatterjee S, Park S, Low K, Kong Y, Pimentel M. The degree of breath methane production in IBS correlates with the severity of constipation. Am J Gastroenterol 2007; 102: 83741.
    Direct Link:
  • 29
    Coelho AM, Fioramonti J, Bueno L. Systemic lipopolysaccharide influences rectal sensitivity in rats: role of mast cells, cytokines, and vagus nerve. Am J Physiol Gastrointest Liver Physiol 2000; 279: G78190.
  • 30
    Rogers J, Henry MM, Misiewicz JJ. Increased segmental activity and intraluminal pressures in the sigmoid colon of patients with the irritable bowel syndrome. Gut 1989; 30: 63441.
  • 31
    Misiewicz JJ, Connell AM, Pontes FA. Comparison of the effect of meals and prostigmine on the proximal and distal colon in patients with and without diarrhoea. Gut 1966; 7: 46873.
  • 32
    Whitehead WE, Engel BT, Schuster MM. Irritable bowel syndrome: physiological and psychological differences between diarrhea-predominant and constipation-predominant patients. Dig Dis Sci 1980; 25: 40413.
  • 33
    Whitehead WE, Drescher VM. Perception of gastric contractions and self-control of gastric motility. Psychophysiology 1980; 17: 5528.
  • 34
    Whitehead WE, Holtkotter B, Enck P et al. Tolerance for rectosigmoid distention in irritable bowel syndrome. Gastroenterology 1990; 98(5 Pt 1): 118792.
  • 35
    Sullivan MA, Cohen S, Snape WJ Jr. Colonic myoelectrical activity in irritable-bowel syndrome. Effect of eating and anticholinergics. N Engl J Med 1978; 298: 87883.
  • 36
    Pimentel M, Soffer EE, Chow EJ, Kong Y, Lin HC. Lower frequency of MMC is found in IBS subjects with abnormal lactulose breath test, suggesting bacterial overgrowth. Dig Dis Sci 2002; 47: 263943.
  • 37
    Posserud I, Stotzer PO, Bjornsson ES, Abrahamsson H, Simren M. Small intestinal bacterial overgrowth in patients with irritable bowel syndrome. Gut 2007; 56: 8028.
  • 38
    Thompson DG, Binfield P, De BA, O’Brien J, Warren S, Wilson M. Extra intestinal influences on exhaled breath hydrogen measurements during the investigation of gastrointestinal disease. Gut 1985; 26: 134952.