Antibiotic use and the development of Crohn’s disease
- 1Division of Epidemiology and Public Health, University of Nottingham, Nottingham, UK
- 2London School of Hygiene and Tropical Medicine, London, UK
- 3London School of Hygiene and Tropical Medicine, London, UK, and Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK
- Correspondence to:
Dr T Card
University of Nottingham, Division of Epidemiology and Public Health, Queen’s Medical Centre, Nottingham NG7 2UH, UK;
- Accepted 26 August 2003
Background: Few environmental determinants of Crohn’s disease are well established. Some observational data exist to implicate antibiotic use as a risk factor but these are derived from studies using questionnaires to assess reported antibiotic use that were susceptible to recall bias. We have therefore explored this relationship in prospectively gathered data.
Methods: We selected incident cases of Crohn’s disease from the General Practice Research Database with at least five years of data prior to diagnosis. Controls with five years of complete data were randomly selected. Data were extracted on smoking, drug prescriptions, age, sex, and a variety of symptoms and diagnoses that might be indicative of occult Crohn’s disease. Logistic regression was used to investigate the relationship between antibiotic use and Crohn’s disease.
Results: A total of 587 Crohn’s disease cases and 1460 controls were available for analysis. We found that antibiotic use 2–5 years pre-diagnosis occurred in 71% of cases compared with 58% of controls (p<0.001), and the median number of courses was two in the cases and one in the controls (p<0.001). Adjusting for age, sex, smoking, and use of other drugs, antibiotic use had an odds ratio of 1.32 (1.05–1.65). We were unable to show specificity to any subgroup of antibacterials. Associations similar to that with antibiotics were also found with oral contraceptive, cardiovascular, and neurological drugs.
Conclusions: We found a statistically significant association between Crohn’s disease and prior antibiotic use. This cannot be explained by recall bias, but due to lack of specificity it is unclear whether it is causal.
- GPRD, general practice research database
- PAF, population attributable fraction
- BNF, British National Formulary
- OR, odds ratio