Diverse pathophysiologic mechanisms, alone or in combination,

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2007;5: Risk of Irritable Bowel Syndrome After an Episode of Bacterial Gastroenteritis in General Practice: Influence of Comorbidities ANA RUIGÓMEZ,* LUIS ALBERTO GARCÍA RODRÍGUEZ,* and JULIÁN PANÉS *Centro Español de Investigación Farmacoepidemiológica, Madrid, Spain; and Servei de Gastroenterologia, Hospital Clínic de Barcelona, Institut d Investigacions Biomèdiques August Pi i Sunyer, CIBERHEPAD, Barcelona, Spain Background & Aims: Given the discrepant estimates of incidence rates of irritable bowel syndrome (IBS) after gastroenteritis (GE), we performed a cohort study to quantify this risk in community subjects and to identify factors acting as modifiers of this effect. Methods: In a previous study, we identified patients aged years with a first ever episode of bacterial GE during We excluded patients with a history of bowel disease and cancer, resulting in a cohort of 5894 individuals with GE. From the same source population, a control group free of GE was sampled. We followed up the 2 cohorts to identify incident cases of IBS. A nested case-control analysis was performed to quantify the role of potential risk factors. Results: During a mean follow-up period of 4.1 years, 1105 patients developed IBS. The incidence rate of IBS after an episode of bacterial GE was 98.2 per 10,000 person-years, and 45.3 per 10,000 person-years in the comparison cohort. The adjusted relative risk of IBS associated with bacterial GE was 2.2 ( ) compared with the control cohort. The nested casecontrol analysis adjusting for additional risk factors produced similar results (odds ratio, 1.8; 95% confidence interval, ). The risk of IBS after GE was increased significantly in patients with depression, anxiety, stress or sleep disorders, prior gastrointestinal morbidity, or prolonged use of antibiotics. Conclusions: The risk of IBS in community individuals after having bacterial GE was two fold greater in the general population. Pre-existing psychologic and gastrointestinal comorbidities independently increase this risk of developing IBS. Diverse pathophysiologic mechanisms, alone or in combination, appear to contribute to irritable bowel syndrome (IBS), including altered motility, visceral hyperalgesia, brain-gut disturbances, psychosocial disturbances, and genetic and environmental factors. 1 It has long been recognized that an episode of acute gastroenteritis (GE) may precede the onset of IBS. 2,3 Many groups have attempted to determine the incidence of IBS after an episode of GE, but, for the most part, previous studies have assessed relatively small population cohorts with limited power to define risks and factors that modify the effect. Retrospective studies have reported incidence rates between 6% 4 and 32%. 5 In some studies poor patient recall of their premorbid bowel habits may have biased the results toward a higher incidence rate. 6 However, prospective studies also have produced discrepant estimates of incidence rates that ranged between 10% 7 and 36%. 8 The majority of prospective studies have concentrated on severe outbreaks of GE, which may have contributed to the high incidence rates found in some of them. A previous study from our group among community subjects who had suffered an episode of GE found the incidence of postinfectious IBS to be 4.4%. 9 A number of studies were aimed at identifying factors that may be associated with the development of IBS after an infectious episode. Among those, psychologic alterations have received the most attention; psychologic symptoms of anxiety and depression were found to be more common in IBS patients than in either healthy volunteers 10 or patients with organic gastrointestinal diseases, 11 and stressful life events often preceded the onset of chronic bowel symptoms. 12,13 However, most of these studies were conducted after IBS had been present for some time, and although they indicate an association, they do not permit us to establish causation. In other studies the presence of psychologic alterations were assessed retrospectively, which is subject to considerable recall bias. Given the discrepant estimates of incidence rates of IBS after GE, we undertook a large cohort study to quantify this risk in community subjects. We also took advantage of the information extracted from a primary care database to identify factors that could act as modifiers of this effect, in particular common comorbidities including general, gastrointestinal, and psychologic. Patients and Methods Study Population We used data from the U.K. General Practice Research Database. This database contains clinical information on more than 3 million residents enrolled with general practitioners in the United Kingdom. These physicians have been trained to record their patients medical and demographic information in a standard manner, and have agreed to supply it anonymously for research purposes. The database is managed by the United Kingdom s Medicines and Healthcare Products Regulatory Agency and its completeness, validity, and quality of data have been widely reported by several studies in the gastrointestinal area. 14,15 Study Cohorts From a previous study we identified 6414 patients aged years with a first ever episode of infectious GE during All had documented bacterial GE with a specific bacteria isolated. In this previous study we applied the follow- Abbreviations used in this paper: CI, confidence interval; GE, gastroenteritis; IBS, irritable bowel syndrome; RR, relative risk by the AGA Institute /07/$32.00 doi: /j.cgh

2 466 RUIGÓMEZ ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 5, No. 4 Table 1. Incidence Rate and RR of IBS in the GE Cohort Compared With the Control Cohort Control cohort (N 46,996) GE cohort (N 5894) Person-years 190,623 24,555 Cases of IBS Incidence per 10, ( ) 98.2 ( ) person-years (95% CI) RR (95% CI) a ( ) RR during first year of ( ) follow-up evaluation (95% CI) a RR after the first year (95% CI) a ( ) NOTE. The mean follow-up period was 4.1 years (SD, 2.5 y; range 0 10 y). a The RR was estimated by Cox regression including age, sex, and calendar year. ing eligibility criteria: patients had to be free of cancer, alcohol abuse, prior gastroenteritis, inflammatory bowel disease, and related gastrointestinal infectious disease or enteritis/colitis at any time before the episode of GE. For the current study, we further identified all individuals with prior IBS or 5 or more prescriptions of any IBS/inflammatory bowel disease treatment (antidiarrheal, antispasmodics, laxatives, chronic bowel disorders treatment: balsalazide, mesalamine, olsalazine, sulfasalazine, infliximab, prednisolone, hydrocortisone, budesonide) before the study entry date and excluded them from the study. The remaining patients formed our final cohort of GE patients (N 5894). We randomly sampled a comparison cohort of 50,000 individuals free of GE who were frequency-matched by age, sex, and calendar year to the GE cohort from the same source population (where the cohort of GE initially was ascertained). After applying the same eligibility criteria as used in the ascertainment of the GE cohort, we ended up with a comparison cohort of 46,996 individuals. Follow-Up Evaluation to Ascertain Cases of Irritable Bowel Syndrome We followed up all members in our 2 study cohorts from the date of GE diagnosis or the random date in the comparison cohort free of GE. Follow-up evaluation ended with the first occurrence of IBS, inflammatory bowel disease, cancer, death, date of last data collection, or December 31, 2001, whichever came first. We reviewed computerized patient profiles of all patients identified with a code of IBS (N 1198), and considered a patient to have IBS when there was a recorded diagnosis of IBS together with specific treatment, when there were repeated entries for IBS, or when there was a diagnosis of IBS together with related gastrointestinal symptoms. An individual with only 1 computerized entry of IBS in the absence of treatment or other gastrointestinal symptoms was not considered to have IBS. After independent review by all 3 authors, 93 patients were not retained as IBS cases. Analysis Incidence rates of IBS in the 2 cohorts were calculated by dividing the number of incident cases over the total follow-up experience in each study cohort. The relative risk (RR) of developing IBS during the first year as well as during the whole follow-up period in the GE cohort compared with the comparison cohort was estimated using a Cox regression analysis adjusting for age, sex, and calendar year. In a secondary nested case-control analysis, we used all 1105 incident cases of IBS occurring in the 2 cohorts and considered the date of first IBS diagnosis as the index date. A date during the study period was generated at random for every member of the 2 study cohorts. If the random date of a study member was included in his/her eligible person-time of follow-up evaluation, we used this random date as the index date and marked that patient as an eligible control. This mechanism (ie, incidence density sampling) allows that the likelihood of being selected as a control is proportional to the person-time at risk. Five thousand controls were frequency-matched by sex, age, and calendar year to the IBS cases. Estimates of odds ratios and 95% confidence intervals (CIs) of IBS, assumed to be valid estimates of the RR, were computed using unconditional logistic regression, adjusting for the matching factors and a number of independent risk factors. Results During a mean follow-up period of 4.1 years (range, 0 10 y; SD, 2.5 y), we ascertained 241 patients with a first-ever diagnosis of IBS in the GE cohort and 864 in the comparison cohort (Table 1). Figure 1 shows the incidence rate of IBS in the 2 cohorts. The overall incidence rate of IBS in patients after an episode of bacterial GE was 98.2 per 10,000 person-years. The corresponding estimate in the comparison cohort free of GE was 45.3 per 10,000 person-years. The adjusted RR of IBS associated with bacterial GE was 2.2 (95% CI, ). When we stratified by length of follow-up time, we observed that the risk associated with bacterial GE was rather similar during the first year after the infective intestinal episode (RR, 2.6; 95% CI, ) than after the first year (RR, 2.0; 95% CI, ). Females carried an increased risk of developing IBS (RR, 3.1; 95% CI, ), and individuals younger than 40 years old presented a RR of 2.6 (95% CI, ) compared with those age 60 years and older. Table 2 presents a summary description of IBS cases occurring in the GE and comparison cohorts. In the GE cohort, Figure 1. Proportion developing IBS during the follow-up period in the GE cohort ( ) and control cohort ( ).

3 April 2007 RISK OF IBS AFTER GASTROENTERITIS 467 Table 2. Summary of Incident Cases of IBS in the GE and Comparison Control Cohorts Comparison cohort (N 864) (%) GE cohort (N 241) (%) Sex Male 224 (26) 71 (29.5) Female 640 (74) 170 (70.5) Age, y (15.3) 44 (18.3) (25.0) 56 (23.2) (27.2) 56 (23.2) (18.3) 52 (21.6) (11.5) 21 (8.7) (2.8) 12 (5.0) GE type Salmonella 67 (27.8) Campylobacter 162 (67.2) Shigella and other 12 (5.0) Interval between the start date and IBS occurrence a, (mo) 1 15 (1.7) 5 (2.1) (24.2) 69 (28.6) (18.8) 63 (26.1) (17.2) 32 (13.3) (38.1) 72 (29.9) a Interval from random date in the comparison cohort or date of GE episode in the bacterial GE cohort unit IBS occurrence. Campylobacter was the microorganism most commonly detected (162 patients), followed by 67 patients with Salmonella infection, 11 patients with Shigella isolated, and 1 patient with Clostridium difficile. Table 3 presents the estimates of odds ratios of IBS associated with GE after adjusting for a number of risk factors in the nested case-control analysis. The corresponding odds ratio of IBS was 1.8 (95% CI, ) in the GE cohort. We found an association of IBS with previously identified risk factors such as depression, anxiety, stress, or sleep disorders. Prior gastrointestinal morbidity (GERD, dyspepsia, diarrhea) was associated with an increased risk of developing IBS. Among patients with bacterial GE, we observed that infection caused by Campylobacter carried a slightly greater risk of IBS compared with Salmonella infection, although this result was not statistically significant (Table 4). We also found an association with chronic use of antibiotics after the episode of GE: patients receiving 5 or more prescriptions had a 2-fold increased risk of IBS compared with those patients not receiving a single antibiotic prescription (Table 4). Discussion The incidence rate of IBS in our cohort of 5894 patients with an episode of microbiologically documented bacterial GE was 10 per 1000 person-years, resulting in a doubling of the risk observed in the comparison cohort of individuals free of GE. The nested case-control analysis gave results similar to the cohort analysis, with the sole difference of slightly smaller estimates of risk after adjusting for a number of additional risk factors. Follow-up evaluations of other smaller cohorts with bacterial GE have estimated that 4% 32% of patients go on to develop IBS. 4,5,7,8,17,18 In the current study this proportion was 4.1%, which was lower than some previous studies but in keeping with studies using similar methodology. 17 A careful exclusion of subjects with premorbid IBS from the study cohort based on diagnoses and previous symptoms and/or medication recorded in the database probably had a significant effect on the measure of this risk. Patients with IBS are 4.3 times more likely to seek medical attention for infectious GE than controls. 18 As a result, it is likely that prevalent IBS cases may be misclassified as incident cases of IBS. In our study, 520 patients (8.1%) from the initially identified cohort had a diagnosis of IBS before GE and thus were excluded from the study cohort. The RR estimate derived from the current study also was lower than that from a previous study by our group, which also used the General Practitioner Research Database. 9 Several factors may have contributed to this effect. Sample size is one factor: the current study was much larger and consequently its results afforded greater confidence. Another significant factor was the characteristics of the source population (users of acid-suppressing drugs) in which the cohort of GE originally was ascertained in the earlier study. 9 It is likely that a population receiving acidsuppressing therapy shares a number of gastrointestinal conditions, not all amenable to an easy characterization, that may predict a greater risk of IBS than in the general population, such as the source population used in the current study to ascertain the cohort of GE. Indeed, we obtained similar estimates of the incidence of IBS in the comparison cohort (4 5 per 1000 person-years) in both studies, whereas the corresponding estimate in the GE cohort was higher in the earlier study (40 per 1000 person-years) vs 10 in the current study. In agreement with previous community-based reports, 17,18 the majority of GE cases were caused by Campylobacter infection, followed by Salmonella and Shigella species, without a major difference in risk of IBS according to the underlying bacteria. The current study also indicated that the excess risk of IBS occurrence was similar during the first year after the infective intestinal episode compared with subsequent years. Although given the nature of our study we cannot rule out the possibility that a considerable time lag may have existed between the initiation of symptoms of IBS and patient consultation or the establishment of the definitive diagnosis, the prolonged period of excess risk of developing IBS after an episode of GE is in keeping with data from Marshall et al 8 showing a 3-fold RR after an interval of at least 2 years. A variety of risk factors associated with IBS after GE were evaluated. Putative risk factors identified in previous research included female sex, younger age, longer duration of diarrhea, 4 presence of fever during the GE episode, 17 and smoking. 19 We found the well-described association of greater risk among women and among younger individuals. Our observation that prolonged use of antibiotics was associated with a higher risk of developing IBS after GE is in line with the premise that more severe GE increases the likelihood of developing IBS, as suggested by surrogate indicators of GE severity such as presence of fever or duration of symptoms. We hypothesized that antibiotics were used for more prolonged periods in patients with more severe GE. In contrast, we did not find any association between smoking status and the risk of IBS after the episode of GE. The possible role of psychologic disturbances as susceptibility factors for the development of IBS after GE has been assessed in several recent well-designed prospective studies. The pioneering study by Gwee et al 20 showed that after controlling for possible

4 468 RUIGÓMEZ ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 5, No. 4 Table 3. RR of IBS Associated With GE and Other Risk Factors in a Nested Case-Control Analysis Controls IBS cases (N 5000) (%) a (N 1105) (%) Odds ratio b 95% CI Cohort status Cohort free of GE 4458 (89.2) 864 (78.2) 1 GE cohort 542 (10.8) 241 (21.8) Gastrointestinal comorbidity Gastroesophageal reflux 348 (7.0) 178 (16.1) Dyspepsia 624 (12.5) 320 (29.0) Peptic ulcer 105 (2.1) 39 (3.5) Appendicitis 257 (5.1) 75 (6.8) Diarrhea 889 (17.8) 412 (37.3) Central nervous system comorbidity Depression 813 (16.3) 301 (27.2) Anxiety 610 (12.2) 264 (23.9) Stress 230 (4.6) 114 (10.3) Sleep disorders 335 (6.7) 141 (12.8) General comorbidity Asthma 495 (9.9) 134 (12.1) Chronic pulmonary disease 59 (1.2) 14 (1.3) Ischemic heart disease 117 (2.3) 36 (3.3) Diabetes 86 (1.7) 15 (1.4) Smoking Nonsmoker 2735 (54.7) 601 (54.4) 1 Smoker 1120 (22.4) 281 (25.4) Ex-smoker 414 (8.3) 108 (9.8) Unknown 731 (14.6) 115 (10.4) Body mass index (6.1) 87 (7.9) (34.2) 388 (35.1) (23.6) 286 (25.9) (11.6) 132 (11.9) Unknown 1243 (24.9) 212 (19.2) Alcohol consumption, U/wk None 1533 (30.7) 369 (33.4) (27.1) 303 (27.4) (13.9) 164 (14.8) (3.2) 50 (4.5) Unknown 1257 (25.1) 219 (19.8) a Controls were frequency matched to cases by age, sex, and calendar year. b Estimates were adjusted for age, sex, calendar year, and visits to the general practitioner in the year prior. confounding effects of the acute illness, there was a significant association between high anxiety, depression, and somatization scores and the subsequent development of IBS after an episode of GE. This observation has been confirmed in subsequent studies by the same 13 and other groups. 21 Our study, similar to that of Dunlop et al, 21 provides an estimation of the risk of developing IBS after an episode of GE in the general population in an attempt to avoid selection bias derived from factors that determine consultation to gastroenterologists, including not only symptom severity but also patient anxiety about serious underlying disease. 22 Series from secondary and tertiary care centers therefore tend to be enriched with anxious patients, which may overemphasize the role of psychologic factors in the condition. We found that the risk for presenting with IBS after an episode of GE was increased significantly in individuals with a previous history of either anxiety, depression, stress, or sleep disorders. The fact that other general conditions such as asthma, chronic pulmonary disease, ischemic heart disease, or diabetes were not associated with IBS after GE is an indication that reporting bias is probably not the cause of the association found between psychologic factors and higher risk of IBS in our study, suggesting that psychologic factors may have a predictive role in the development of IBS symptoms. This study had several methodologic strengths that included its large sample size, the reliability of the clinical information contained in the General Practitioners Research Database 14,15 that allowed for an analysis of a number of plausible risk factors, and the long-term follow-up period. One of the weaknesses of the current study was that the group of patients with GE still may represent a limited subset of individuals with GE. For example, only subjects who sought medical attention for their symptoms and had stool studies collected were eligible to be members of our GE cohort, and this may have lead to misclassification studying the more severe cases. Another limitation was that the pattern of diagnosis and symptom recording also was likely to vary between physicians because there are no uniform criteria for establishing the diagnosis of IBS among family practitioners participating in the database, this may have a significant impact on the proportion of patients diagnosed as having IBS. 17,23

5 April 2007 RISK OF IBS AFTER GASTROENTERITIS 469 Table 4. RR of IBS in the Cohort of GE Associated With Type of Bacteria and Use of Antibiotics Controls (N 542) (%) IBS cases (N 241) (%) Odds ratio a 95% CI Bacteria isolated Salmonella 169 (31.2) 67 (27.8) 1 Campylobacter 335 (61.8) 162 (67.2) Shigella and other b 38 (7.0) 12 (5.0) Use of antibiotics (number of prescriptions between episode of bacterial GE and index date) None 152 (28.0) 44 (18.3) (36.5) 70 (29.0) (28.4) 95 (39.4) (7.0) 32 (13.3) a Estimates were adjusted for age, sex, calendar year, and visits to the general practitioner in the year prior. b Other includes Staphylococus (4), Clostridium difficile (5), and E coli (2). In summary, the risk of IBS in community individuals suffering an episode of bacterial GE is twice the risk in the general population. Previous psychologic disturbances, dyspepsia, and prolonged use of antibiotics after the episode of GE are associated with a higher risk. Novel management strategies that target this subset of patients deserve further study. References 1. Barbara G, De Giorgio R, Stanghellini V, et al. New pathophysiological mechanisms in irritable bowel syndrome. Aliment Pharmacol Ther 2004;20(Suppl 2): Stewart GT. Post-dysenteric colitis. BMJ 1950;1: Chaudhary NA, Truelove SC. The irritable colon syndrome. A study of the clinical features, predisposing causes, and prognosis in 130 cases. QJM 1962;31: Neal KR, Hebden J, Spiller R. Prevalence of gastrointestinal symptoms six months after bacterial gastroenteritis and risk factors for development of the irritable bowel syndrome: postal survey of patients. BMJ 1997;314: McKendrick MW, Read NW. Irritable bowel syndrome post salmonella infection. J Infect 1994;29: Maxwell P, Mendall M. Prevalence of gastrointestinal symptoms after bacterial gastroenteritis. Patients with the irritable bowel syndrome may underreport historical symptoms. BMJ 1997;314: Mearin F, Perez-Oliveras M, Perello A, et al. Dyspepsia and irritable bowel syndrome after a Salmonella gastroenteritis outbreak: one-year follow-up cohort study. Gastroenterology 2005;129: Marshall JK, Thabane M, Garg AX, et al. Incidence and epidemiology of irritable bowel syndrome after a large waterborne outbreak of bacterial dysentery. Gastroenterology 2006;131: García Rodríguez LA, Ruigómez A. Increased risk of irritable bowel syndrome after bacterial gastroenteritis: cohort study. BMJ 1999; 318: MacDonald AJ, Bouchier IA. Non-organic gastrointestinal illness: a medical and psychiatric study. Br J Psychiatry 1980;136: Walker EA, Roy-Byrne PP, Katon WJ, et al. Psychiatric illness and irritable bowel syndrome: a comparison with inflammatory bowel disease. Am J Psychiatry 1990;147: Dunlop SP, Jenkins D, Neal KR, et al. Relative importance of enterochromaffin cell hyperplasia, anxiety, and depression in postinfectious IBS. Gastroenterology 2003;125: Gwee KA, Leong YL, Graham C, et al. The role of psychological and biological factors in postinfective gut dysfunction. Gut 1999; 44: Lewis JD, Brensinger C, Bilker WB, et al. Validity and completeness of the General Practice Research Database for studies of inflammatory bowel disease. Pharmacoepidemiol Drug Saf 2002; 11: Ruigómez A, Wallander MA, Johansson S, et al. One-year follow-up of newly diagnosed irritable bowel syndrome patients. Aliment Pharmacol Ther 1999;13: García Rodríguez LA, Ruigómez A, Panés J. Acute gastroenteritis is followed by an increased risk of inflammatory bowel disease. Gastroenterology 2006;130: Borgaonkar MR, Ford DC, Marshall JK, et al. The incidence of irritable bowel syndrome among community subjects with previous acute enteric infection. Dig Dis Sci 2006;51: Parry SD, Stansfield R, Jelley D, et al. Does bacterial gastroenteritis predispose people to functional gastrointestinal disorders? A prospective, community-based, case-control study. Am J Gastroenterol 2003;98: Parry SD, Barton JR, Welfare MR. Factors associated with the development of post-infectious functional gastrointestinal diseases: does smoking play a role? Eur J Gastroenterol Hepatol 2005;17: Gwee KA, Graham JC, McKendrick MW, et al. Psychometric scores and persistence of irritable bowel after infectious diarrhoea. Lancet 1996;347: Dunlop SP, Jenkins D, Spiller RC. Distinctive clinical, psychological, and histological features of postinfective irritable bowel syndrome. Am J Gastroenterol 2003;98: Thompson WG, Heaton KW, Smyth GT, et al. Irritable bowel syndrome in general practice: prevalence, characteristics, and referral. Gut 2000;46: Talley NJ, Weaver AL, Zinsmeister AR, et al. Onset and disappearance of gastrointestinal symptoms and functional gastrointestinal disorders. Am J Epidemiol 1992;136: Address requests for reprints to: Luis Alberto García Rodríguez, Centro Español de Investigación Farmacoepidemiológica, Almirante 28, 2, Madrid, Spain. lagarcia@ceife.es; fax: (34) Supported in part by a research grant from AstraZeneca R&D (Sweden) and by a grant (C03/02) from the Instituto de Salud Carlos III (J.P.). The authors thank the staff at the General Practice Research Database, the participating general practitioners for their collaboration, and the Boston Collaborative Drug Surveillance Program for providing access to the database.