Age at Diagnosis of Crohnʼs Disease May Explain NOD2-Smoking Interactions

Abstract

Introduction: A recent meta-analysis demonstrated that patients with Crohn's disease (CD) who carry a 1007fs variant in the NOD2 gene were significantly less likely to be a current or former smoker at the time of CD diagnosis. We hypothesize that the NOD2-smoking interaction is explained by the differential prevalence of NOD2 variants and smoking status based on age at diagnosis. Methods: A case-only study was used to assess the interaction between smoking status at diagnosis and the 1007fs allele of the NOD2 gene among patients with CD (N=733). Smoking status was defined as current, former (quit smoking more than 1 year prior to diagnosis), and lifelong non-smoker prior to diagnosis. The 1007fs variant of the NOD2 gene was identified using a golden gate custom chip. Chi-squared tests were used to assess the association between the 1007fs variant, smoking status, and age at diagnosis as defined by the Montreal Classification: A1 (≤16 years of age); A2 (17 to 40 years); and A3 (> 40 years). Logistic regression was used to estimate the odds ratio (OR) of being a carrier of the 1007fs variant and an ever smoker (current or former) at diagnosis for patients more than 40 years old at diagnosis (A3) compared to those less than 40 at diagnosis (A1+A2). Results: Among the733 patients with CD 76 (10.4%) were carriers for the 1007fs variant; 227 (31.0%) were current smokers at diagnosis and 89 (12.1%) were former smokers. Overall, 107 (14.9%) patients were A1; 487 (67.9%) were A2; and 123 (17.2%) were A3. The prevalence of the 1007fs variant varied significantly with age of diagnosis (Figure 1; p < 0.01). In contrast, the proportion of patients that were ever smokers at diagnosis was significantly higher among A3 patients (Figure 1; p < 0.001). Diagnosis after the age of 40 years significantly decreased the odds of being a 1007fs carrier (OR 0.24, 95% CI 0.09 to 0.68) and increased the odds of having ever smoked (OR 3.36, 95% CI 2.22 to 5.08).Figure 1Conclusion: Although meta-analysis data has demonstrated a significant negative interaction between the 1007fs variant and cigarette smoking, the mechanism for the protective effect of these two risk factors remains unknown. Our data suggests that this interaction may result from differences in the prevalence of the 1007fs mutation in NOD2 and smoking status at age of diagnosis. Consequently, gene-environment interaction studies should be powered to examine age-specific interactions.