Small Intestinal Bacterial Overgrowth May Increase the Likelihood of Lactose and Sorbitol but Not Fructose Intolerance False Positive Diagnosis

Home , Hydrogen breath test

Available online at www.annclinlabsci.org Annals of Clinical & Laboratory Science, vol. 47, no. 4, 2017 447 Small Intestinal Bacterial Overgrowth May Increase the Likelihood of Lactose and Sorbitol but not Fructose Intolerance False Positive Diagnosis

Tsachi Tsadok Perets1, Dalal Hamouda1, Olga Layfer1, Olga Ashorov1, Doron Boltin2, Sigal Levy3, Yaron Niv2, and Ram Dickman2

1Gastroenterology Laboratory and 2The Neurogastroenterology Service, Department of Gastroenterology, Rabin Medical Center, Beilinson Campus and Sackler Faculty of Medicine, Tel Aviv University, and 3The Academic College of Tel Aviv-Jaffa, Tel Aviv, Israel

Abstract. Background. Small intestinal bacterial overgrowth (SIBO) is defined as a bacterial count of more than 105 colony-forming units per milliliter in duodenal aspirate. It shares many symptoms with carbohydrate intolerance, which makes the clinical distinction of the disorders difficult. The aim of the study was to examine the relationship between a positive carbohydrate breath test and the presence of SIBO suggested by a positive lactulose hydrogen breath test. Methods. The electronic database of the gastroenterology laboratory of a tertiary medical center was searched for all patients clinically tested for SIBO in 2012-2013 for whom previous results for lactose, fructose, and/or sorbitol breath test were available. The correlation between positive findings for carbohydrate intolerance and for SIBO was statistically analyzed. Results. The study group included 349 patients, 231 female and 118 male, of mean age 53±19 years. All had symptoms of abdominal bloating and gas. There was a statistically significant difference in rates of a positive breath test for lactose and sorbitol at ≤90 minutes between patients who were positive and negative for SIBO [χ2(1)=12.8, p<0.01 and χ2(1)=9.5, p<0.01 respectively]. Findings for fructose were not significant. There was no effect of age or gender. Conclusions. SIBO may represent an important reversible cause of carbohydrate intolerance. It may be especially prudent to exclude SIBO patients with an early peak (≤90 minutes) in H2 excretion.

Key words: SIBO, Carbohydrate intolerance, Hydrogen breath test.

Introduction reaching an estimated 50% in the over-75-year age group, owing to age-related disturbances in gastric Small intestinal bacterial overgrowth (SIBO) is acid production and intestinal motility and other characterized by a high bacterial count in duodenal natural defense mechanisms [5]. Patients with an aspirate. The generally accepted cutoff is105 or underlying gastrointestinal pathology, such as more colony-forming units per milliliter of duode- chronic pancreatitis, immune system dysfunction, nal aspirate [1]. The clinical presentation varies functional dyspepsia, gastroparesis, celiac disease, from mild nonspecific symptoms such as abdomi- and inflammatory webo l disease (IBD), are more nal pain, flatulence, and bloating to a full-blown susceptible [6-11]. malabsorption syndrome with diarrhea and weight loss [2,3]. Lactulose hydrogen breath test (LaBT) is SIBO is treated with antibiotics, namely, neomy- an important means of clinical diagnosis of SIBO. cin, ciprofloxacin, and especially rifaximin [12]. It is simple, non-invasive and demands short period Rifaximin, a semi-synthetic antibiotic based on ri- of fasting [4].The actual prevalence of SIBO is cur- famycin, is highly suitable for SIBO treatment be- rently unclear. Rates are higher in older individuals, cause it is not absorbed by the intestine. When taken orally, it is poorly absorbed into the blood- stream and operates locally in the gut lumen against Address correspondence to Dr. Tsachi Tsadok Perets, PhD; Head of the Gastroenterology Laboratory, Rabin Medical Center, Beilinson susceptible bacteria [4,13] exerting activity against Campus, 39 Jabotinsky Street, Petah Tikva, Israel 49100; phone: 972 3 9377297; fax: 972-3-9377280; e mail: [email protected], both gram-positive and gram-negative aerobes and [email protected] anaerobes [4,13-14]. Rifaximin has been associated

Table 1. Comparison between patients positive and negative forSIBO for carbohydrate intolerance, including logistic regres- sion for gender and sex.

A Lactose Lactose Odds Signifi- >90 min ≤90 min TotalN ratio cance

SIBO Negative 41(41.8%) 57 (58.2%) 98 SIBO 3.128 p<0.01 Total Positive 19 (18.6%) 83 (81.4%) 102 Gender 1.254 NS 60 140 200 Sex 0.994 NS

B Sorbitol Sorbitol Odds Signifi- >90 min ≤90 min TotalN ratio cance

SIBO Negative 18 (85.7%)3(14.3%) 21 SIBO 7.306 p<0.01 Total Positive 20 (45.5%) 24 (54.5%) 44 Gender 3.099 NS 38 27 65 Sex 1.006 NS

C Fructose Fructose Odds Signifi- >90 min ≤90 min TotalN ratio cance

SIBO Negative 2 (33.0%)4(67.0%)6 SIBO 2.538 NS Total Positive 3 (17.6%) 14 (82.4%) 17 Gender 1.137 NS 5 1 23 Sex 1.009 NS with a wide range of eradication rates, from 16.7% We also sought to identify predictors of carbohy- to 100% [13,15-19]. Treatment generally lasts from drate malabsorption in patients with suspected 7 to 14 days [13-16], although a recent study sug- SIBO. gested that a longer period of up to 12 weeks is needed for eradication, corresponding to the time to nor- Materials and Methods malization of the LaBT [4]. Subjects.Aretrospective cohort study design was used. Evaluating carbohydrate intolerance using hydrogen The electronic database of the gastroenterology laboratory at Rabin Medical Center, a tertiary hospital, was reviewed breath tests (HBTs) is complex. Lactose, fructose, for 349 consecutive patients aged more than 18 years and sorbitol are inherently difficult to digest, such (mean age 53±19 years; 231 females and 118 males). All that a small unabsorbed proportion is expected to patients were referred by a consultant gastroenterologist, reach the distal small intestine. Studies have reported had abdominal symptoms such as bloating, abdominal that up to 8% of ingested lactose reaches the ileum pain, flatulence, and diarrhea and underwent LaBT for in normal subjects [20]. If bacterial overgrowth is suspected SIBO from April 2012 to July 2013. The pa- present in the distal small intestine, this residual car- tients' previous results of prior HBTs for lactose, fructose, bohydrate becomes fermented, resulting in the pro- and/or sorbitol were also available. duction of gases, including hydrogen, carbon diox- ide, and methane, leading to a falsely abnormal Exclusion criteria were the use of antibiotics or probiotics in the previous four weeks, current use of laxatives or pro- carbohydrate breath test [7,21-23]. Furthermore, motility agents, or bowel preparation for colonoscopy or lactose, fructose, and sorbitol intolerance share many capsule endoscopy within the previous 30 days. gastrointestinal symptoms with irritable bowel syndrome and SIBO [24-26], which can lead to over Hydrogen Breath Test Procedure. All breath tests were diagnosis of carbohydrate malabsorption and under performed according to standard protocols [21,28-29]. In diagnosis of SIBO [27]. brief, patients were asked to consume a low carbohydrate diet for 24 hours prior to the test in order to minimize the The aim of this study was to examine the relation- basic hydrogen discharge. They were then asked to fast ship between carbohydrate breath test positivity and overnight for 12 hours and rinse with a chlorhexidine the presence of SIBO indicated by a positive LaBT. mouthwash on the morning of the test. Smoking and physical activity were prohibited on the day of the test. SIBO and Carbohydrates Intolerance 449 [χ2(1)=9.5, p <0.01]. Controlling for age and gen- Table 2. Correlations of demographic variables with hy- der had no effect on the significance of these results drogen breath tests (HBTs). (Table 1A&B). For fructose, no significant relationship was found with SIBO (Fisher's exact test, HBT Gender Age Significance P=NS), regardless of patient age or gender (Table 1C). SIBO (N=349) 0.054 0.018 NS Lactose (N=280) -0.001 -0.001 NS Fructose (N=196) -0.0115 -0.014 NS When bivariate analysis was performed to evaluate Sorbitol (N=195) 0.266 -0.005 p<0.01 the effect of demographic variables, the only significant finding was a higher rate of female than male patients with sorbitol intolerance (Table 2). To test for SIBO, baseline hydrogen level was measured, and patients were given 15 gr of lactulose dissolved in 50 Discussion ml of water. Samples of end-expiratory air were collected after the baseline measurement and then every 15 min- An accurate diagnosis of SIBO is crucial for suc- utes for up to 90 minutes. To test for carbohydrate intol- cessful eradication of the excess bacteria. However, erance, after baseline hydrogen was measured, patients SIBO is often misdiagnosed or underdiagnosed ow- were given either 50 gr of lactose or 25 grams of fructose or 15 grams of sorbitol dissolved in 250 ml of water, and ing to the diversity of symptoms and their similar- samples of end-expiratory air were collected every 30 ity to those of other gastrointestinal disorders such minutes for up to 180 minutes. Test results were consid- as carbohydrate intolerance [2,3,24-26]. The pres- ered positive when hydrogen concentration exceeded ent retrospective study shows that SIBO is signifi- baseline by 10 PPM for the LaBT and 20 PPM for the cantly associated with positive findings onTs HB other HBTs. for lactose and sorbitol, but not fructose, at T≤90. These findings suggest that lactose and sorbitol in- Statistical Analysis. All analyses were performed using tolerance may be incorrectly diagnosed in patients SPSS version 21.0 (IBM, Chicago, IL, USA). Chi- with SIBO. We attribute the no significant findings square test or Fisher's exact test, as appropriate accord- for time to positivity for fructose intolerance to the ing to sample size, was used to test the correlation between a positive LaBT (presence of SIBO) and positive smaller sample size (Table 1C) rather than a physi- lactose, fructose and sorbitol tests at 90 minutes or less ological mechanism. (T≤90). The analysis was repeated using logistic regres- sion to control for age and sex. Bivariate Pearson correla- There are several hypotheses that may explain why tions were calculated between the demographic variables SIBO might cause carbohydrate malabsorption. (sex and age) and each type of carbohydrate intolerance First, the excess bacteria may cause fermentation of and SIBO. carbohydrates, including lactose and sorbitol, thereby contributing to the increase in expired hy- Ethics statement. The study was performed in accor- drogen and a positive carbohydrate breath test. dance with the principles of the Declaration of Helsinki Second, bacterial overgrowth may cause mucosal and was approved by the Institutional Review Board of Rabin Medical Center - Beilinson Hospital damage within the small intestine, leading to a (#0036-15). short-term deficiency in mucosal enzymes and transporters (such as GLUT 5 or lactase) which Results may precipitate carbohydrate malabsorption. Both these mechanisms support the presumption that The results showed that in the subgroup of patients carbohydrate malabsorption is reversible in a pro- who were positive for SIBO, 81% were also posi- portion of patients after successful antibiotic treat- tive for lactose at T≤90, compared to 58.2% of the ment for SIBO. subgroup of patients negative for SIBO. This difference was statistically significant [χ2(1)=12.8, This study has several limitations. We used a retro- p<0.01]. A significant difference was also found for spective design which has inherent drawbacks and, sorbitol: 54% of patients who were positive for owing to our exclusion criteria, the cohort was SIBO at T≤90 were also positive for sorbitol, com- highly heterogeneous with respect to the underly- pared to 14.3% of patients negative for SIBO ing pathogenesis of SIBO. For example, SIBO due 450 Annals of Clinical & Laboratory Science, vol. 47, no. 4, 2017 Lactulose is associated with a high variability in sensitivity and specificity for diagnosis of SIBO, but it is preferred in patients with diabetes [2]. Furthermore, a recent study showed that the glucose breath test has a high false-positive rate for SIBO [30]. Lactulose also decreases oro-cecal transit time, making it difficult to identify the dual hydrogen peak profile usually indicative of SIBO [28,31]. However, this problem is counteracted by the delayed small intestinal transit time in patients with SIBO [31]. It has been shown that normal oro-cecal transit time (OCTT) post oral admin- istration of 10 grams of lactulose is 60 to 120 minutes. The presence of excess bacteria in the small intestine prolongs OCTT [31]. Therefore we pre- sume that our 90 minutes LaBT is a reliable indica- tor for the presence of SIBO. Our laboratory does not measure methane in addition to hydrogen, raising concerns of false-negative results, as this misses the 20%-30% of the population that produce methane as the main byproduct of carbohydrate fermentation. Finally, it must be empha- sized that although SIBO Figure 1. Proposed algorithm for the evaluation of patients with excessive flatus and may be responsible for bloating. carbohydrate intolerance in a proportion of pa- to a blind loop may be a different entity from SIBO tients, causation was not established in this study. A pro- due to a transient impairment in gastric defenses. spective study is needed to examine whether following The LaBT is also imperfect. Our laboratory uses successful antibiotic treatment for SIBO, breath test for lactulose rather than glucose as the substrate. carbohydrates malabsorption becomes negative. SIBO and Carbohydrates Intolerance 451 In summary, we found that patients with a positive 14. Sharara AI, Aoun E, Abdul-Baki H, Mounzer R, Sidani S, Elhajj I. A randomized double-blind placebo-controlled trial of rifaxi- breath test for carbohydrate malabsorption within 90 min in patients with abdominal bloating and flatulence. Am J minutes have a greater likelihood of LaBT positivity. Gastroenterol 2006;101:326-333. This suggests that SIBO might be the cause of carbohy- 15. Biancone L, Vernia P, Agostini D, Ferrieri A, Pallone F. Effect of rifaximin on intestinal bacterial overgrowth in Crohn's disease as drate malabsorption in these patients. Based on these assessed by the H2-glucose breath test. Curr Med Res Opin findings, we recommend that when clinically mandat- 2000;16:14-20. ed; patients should be tested (and treated) for SIBO 16. Lauritano EC, Gabrielli M, Lupascu A, Santoliquido A, Nucera G, Scarpellini E, Vincenti F, Cammarota G, Flore R, Pola P, before undergoing testing for carbohydrate malabsorp- Gasbarrini G, Gasbarrini A. Rifaximin dose-finding study for the tion as illustrated in our proposed algorithm (Figure treatment of small intestinal bacterial overgrowth. Aliment Pharmacol Ther 2005;22:31-35. 1). If carbohydrate malabsorption is tested first, a rapid 17. Lauritano EC, Gabrielli M, Scarpellini E, Ojetti V, Roccarina D, (before 90 minutes) rise in expired hydrogen should Villita A, Fiore E, Flore R, Santoliquido A, Tondi P, Gasbarrini G, alert the physician to the possibility that SIBO is pres- Ghirlanda G, Gasbarrini A. Antibiotic therapy in small intestinal bacterial overgrowth: rifaximin versus metronidazole. Eur Rev ent and prompt a referral for an appropriate test. Med Pharmacol Sci 2009;13:111-116. Gastroenterologists should be aware of this phenome- 18. Furnari M, Parodi A, Gemignani L, Giannini EG, Marenco S, non and exercise care in interpreting HBT results. Savarino E, Assandri L, Fazio V, Bonfanti D, Inferrera S, Savarino V. Clinical trial: The combination of rifaximin with partially hy- drolysed guar gum is more effective than rifaximin alone in eradi- References cating small intestinal bacterial overgrowth. Aliment Pharmacol Ther 2010;32:1000-1006. 1. Khoshini R, Dai SC, Lezcano S, Pimentel M. A systematic review of 19. Collins BS, Lin HC. Double-blind, placebo-controlled antibiotic diagnostic tests for small intestinal bacterial overgrowth. Dig Dis Sci treatment study of small intestinal bacterial overgrowth in chil- 2008;53:1443-1454. dren with chronic abdominal pain.JPediatr Gastroenterol Nutr 2. Gasbarrini A, Corazza GR, Gasbarrini G, Montalto M, Di Stefano 2011;52:382-386. M, Basilisco G, Parodi A, Usai-Satta P, Vernia P, Anania C, 20. Nucera G, Gabrielli M, Lupascu A, Lauritano EC, Santoliquido Astegiano M, Barbara G, Benini L, Bonazzi P, Capurso G, Certo M, A, Cremonini F, Cammarota G, Tondi P, Pola P, Gasbarrini G, Colecchia A, Cuoco L, Di Sario A, Festi D, Lauritano C, Miceli E, Gasbarrini A. Abnormal breath tests to lactose, fructose and sor- Nardone G, Perri F, Portincasa P, Risicato R, Sorge M, Tursi A. bitol in irritable bowel syndrome may be explained by small intes- Methodology and indications of H2-breath testing in gastrointesti- tinal bacterial overgrowth. Aliment Pharmacol Ther nal diseases: the Rome Consensus Conference. Aliment Pharmacol 2005;21:1391-1395. Ther 2009;29(suppl 1):1-49. 21. Montalto M, Gallo A, Ojetti V, Gasbarrini A. Fructose, trehalose 3. Bures J, Cyrany J, Kohoutova D, Förstl M, Rejchrt S, Kvetina J, and sorbitol malabsorption. Eur Rev Med Pharmacol Sci Vorisek V, Kopacova M. Small intestinal bacterial overgrowth syn- 2013;17(suppl 2):26-29. drome. WorldJGastroenterol 2010;16:2978-2990. 22. Gabrielli M, D'Angelo G, Di Rienzo T, Scarpellini E, Ojetti V. 4. Bae S, Lee KJ, Kim YS, Kim KN. Determination of rifaximin treat- Diagnosis of small intestinal bacterial overgrowth in the clinical ment period according to lactulose breath test values in nonconsti- practice. Eur Rev Med Pharmacol Sci 2013;17(suppl 2):30-35. pated irritable bowel syndrome subjects.JKorean Med Sci 23. Compare D, Pica L, Rocco A, De Giorgi F, Cuomo R, Sarnelli G, 2015;30:757-762. Romano M, Nardone G. Effects of long-term PPI treatment on 5. Elphick DA, Chew TS, Higham SE, Bird N, Ahmad A, Sanders DS. producing bowel symptoms and SIBO. EurJClin Invest Small bowel bacterial overgrowth in symptomatic older people: can 2011;41:380-386. it be diagnosed earlier? Gerontology 2005;51:396-401. 24. 24. Bohmer CJ, Tuynman HA. The clinical relevance of lactose 6. Trespi E, Ferrieri A. Intestinal bacterial overgrowth during chronic malabsorption in irritable bowel syndrome. EurJGastroenterol pancreatitis. Curr Med Res Opin 1999;15:47-52. Hepatol 1996;8:1013-1016. 7. Belistos PC, Greenson JK, Yardley JH, Sisler JR, Bartlett JG. 25. Ledochowski M, Widner B, Bair H, Probst T, Fuchs D. Fructose- Association of gastric hypoacidity with opportunistic enteric infec- and sorbitol-reduced diet improves mood and gastrointestinal dis- tions in patients with AIDS.JInfect Dis 1992;166:277-284. turbances in fructose malabsorbers. ScandJGastroenterol 8. Lombardo L, Foti M, Ruggia O, Chiecchio A. Increased incidence of 2000;35:1048-1052. small intestinal bacterial overgrowth during proton pump inhibitor 26. Zhao J, Fox M, Cong Y, Chu H, Shang Y, Fried M, Dai N. Lactose therapy. Clin Gastroenterol Hepatol 2010;8:504-508. intolerance in patients with chronic functional diarrhoea: the role 9. George NS, Sankineni A, Parkman HP. Small intestinal bacterial of small intestinal bacterial overgrowth. Aliment Pharmacol Ther overgrowth in gastroparesis. Dig Dis Sci 2014;59:645-652. 2010;31:892-900. 10. Rubio-Tapia A, Barton SH, Rosenblatt JE, Murray JA. Prevalence of 27. Sachdev AH, Pimentel M. Gastrointestinal bacterial overgrowth: small intestine bacterial overgrowth diagnosed by quantitative cul- pathogenesis and clinical significance. Ther Adv Chronic Dis ture of intestinal aspirate in celiac disease. J Clin Gastroenterol 2013;4:223-231. 2009;43:157-156. 28. Riordan SM, McIver CJ, Walker BM, Duncombe VM, Bolin TD, 11. Rana SV, Sharma S, Kaur J, Prasad KK, Sinha SK, Kochhar R, Thomas MC. The lactulose breath hydrogen test and small intes- Malik A, Morya RK. Relationship of cytokines, oxidative stress and tinal bacterial overgrowth. Am J Gastroenterol GI motility with bacterial overgrowth in ulcerative colitis patients.J 1996;91:1795-1803. Crohns Colitis 2014;8:859-865. 29. Ledochowski M, Ledochowski E. Hydrogen breath tests. 1st ed. 12. Shah SC, Day LW, Somsouk M, Sewell JL. Meta-analysis: antibiotic Innsbruck: Verlab Ledochowski, 2008; 53 p. therapy for small intestinal bacterial overgrowth. Aliment Pharmacol 30. Lin EC, Massey BT. Scintigraphy demonstrates high rate of false- Ther 2013;38:925-934. positive results from glucose breath tests for small bowel bacterial 13. Scarpellini E, Gabrielli M, Lauritano CE, Lupascu A, Merra G, overgrowth. Clin Gastroenterol Hepatol 2016;14:203-208. Cammarota G, Cazzato IA, Gasbarrini G, Gasbarrini A. High dos- 31. Roland BC, Ciarleglio MM, Clarke JO, Semler JR, Tomakin E, age rifaximin for the treatment of small intestinal bacterial over- Mullin GE, Pasricha PJ. Small intestinal transit time is delayed in growth. Aliment Pharmacol Ther 2007;25:781-786. small intestinal bact erial overgrowth.JClin Gastroenterol 2015;49:571-576.