Objectives: Multiple established Crohn's disease (CD) and ulcerative colitis (UC) risk factors including family history, tobacco use, Jewish ethnicity, urban residency, and CARD15/NOD2 mutations have been evaluated singly and in hospital-based observational studies. The goal of this study was to assess the relative contributions of all these risk factors jointly in a nonreferral, population-based cohort derived from a population epidemiologic database.
Methods: CD (N = 232) and UC (N = 121) subjects were ascertained from our population-based IBD Registry derived from Manitoba Health, the single provincial insurer. Healthy controls (HC) (N = 336) were recruited via a 10:1 mailing matched for age, sex, and postal code. Ethnicity, tobacco use, family history, residency, and CARD15/NOD2 genotype status were determined.
Results: In both univariate analyses and analyses adjusted for all risk factors, CD was influenced independently by CARD15/NOD2 heterozygote and homozygote/compound-heterozygote status (adjusted odds ratios [OR] 3.7 and 40.0, respectively), Jewish ethnicity (OR 18.5), CD family history (OR 6.2), and smoking (OR 3.0 current and 1.7 ex-smoker, respectively). Penetrance for homozygote/compound-heterozygotes was 4.9%, heterozygotes 0.54%, and wild types 0.184%. Population attributable risk for CARD15 was 26.7% and current tobacco use was 46.8%. A tobacco-CARD15 interaction was not observed. UC was influenced by Jewish ethnicity (OR 37.1), and by family history (OR 2.6), ex-smoker status (OR 1.9), and CARD15/NOD2 heterozygote or homozygote/compound-heterozygote status (OR 1.9 and 6.4, respectively) in adjusted analyses only.
Conclusions: CARD15/NOD2, family history, smoking, and Jewish ethnicity are independent risk factors for CD. Examination of these risk factors together in a single population-based cohort has provided initial data for population epidemiological characterization and genetic counseling uses.