Whilst smoking is established as a protective modifiable environmental risk factor for the development of Ulcerative Colitis (UC), the evidence for its impact on subsequent disease activity is conflicting. We therefore aimed to investigate the impact of smoking on clinical outcomes in the disease course of UC.Methods
Using a nationally representative clinical practice research database (CPRD), we identified incident cases of UC diagnosed between 2005 and 2014. Patients were grouped as: smokers, never-smokers or ex-smokers at UC diagnosis based on medical record codes for smoking status in the two years preceding UC diagnosis. Medical record codes were also examined to determine change in smoking status following diagnosis. We compared corticosteroid dependency (as defined in ECCO guidelines), thiopurine use and colectomy rates between these defined groups. Survival analysis, Cox proportional hazards analysis and logistic regression were used determine the risk of first thiopurine use, corticosteroid dependency and colectomy given smoking status.Result
We identified 4069 cases of UC over the study period. There were 1678 never smokers (41%), 329 smokers (8%), 1541 ex-smokers (38%) and 521 patients whose follow-up smoking status changed (13%). Multivariate regression analysis, adjusting for all covariates listed in table 1, demonstrated smokers had a significantly lower risk of thiopurine use compared to both never smokers (HR 0.52, 95% CI 0.27–0.97, p=0.04) and ex-smokers (HR 0.51, 95% CI 0.27–0.98, p=0.04). In contrast there was no difference in corticosteroid dependency (OR 0.83, 95% CI 0.44–1.75) or rates of colectomy (HR 0.4, 95% CI 0.53–3.02) in a multivariate analysis.Conclusions
Smoking is associated with a decreased requirement for thiopurines however it does not impact on the risk of corticosteroid dependency or colectomy. The risks associated with smoking outweigh any benefits and smoking cessation should be encouraged.