Smoking is thought to have contradictory roles for Crohn's disease (CD) and ulcerative colitis (UC). CD patients are more likely to be smokers at diagnosis and UC patients former smokers. However, the majority of evidence is based on findings from Western countries. We compared the effect of smoking on IBD prevalence and outcomes between Chinese and American populations.Methods:
Participants diagnosed with CD or UC and their friend or family controls were recruited from the sixth Affiliated Hospital of Sun Yat-sen University in China, and Johns Hopkins University and University of California, Irvine in USA. Participants completed a questionnaire administered by healthcare professional or online about environmental risk factors and IBD (bit.ly/IBD-MIMAS). Smoking status at IBD diagnosis was determined using self-reported start and stop year of smoking and year of IBD diagnosis, categorized into current, former, and never smoking. We used multivariable logistic regressions to analyze association of smoking and IBD prevalence, adjusting for sex; and association of smoking and IBD outcomes, adjusting for sex, age of diagnosis, and disease duration at questionnaire.Results:
Eligible participants included 223 CD cases (China 62, USA 161), 159 UC cases (China 32, USA 127), and 261 controls (China 102, USA 159), recruited between May 2014 and August 2016. In China, the percentage of current and former smoking at diagnosis was 11.3% and 6.5% for CD cases, compared with 19.6% and 1.0% for controls (adjusted odds ratio [OR] and 95% confidence interval [CI] versus Never: Current 0.44 [0.16–1.21]; Former 5.0 [0.52–47.73]); 3.1% of UC cases were current and 25% were former smokers compared with 19.6% and 2.9% of controls (Current 0.47 [0.04–4.82]; Former 24.89 [4.19–148]). In USA, 18.6% of CD cases were current and 11.2% were former smokers at diagnosis, compared with 28.9% and 9.4% of controls (Current 0.57 [0.33–0.96]; Former 1.04 [0.50–2.18]); 20.5% of UC cases were current and 15.8% were former smokers compared with 25.2% and 13.2% of controls (Current 0.79 [0.44–1.40]; Former 1.14 [0.58–2.25]). For Chinese CD cases, current and former smoking at diagnosis had a tendency to increase need for immunosuppressant and biologic therapy (Immunosuppressant: Current 5.94 [0.63–55.73], Former 3.16 [0.17–57.95]; Biologics: Current 2.37 [0.34–16.6], Former 2.36 [0.25–22.29]), and decrease need for CD-related surgery (Current 0.37 [0.05–2.77], Former 0.24 [0.02–2.82]). For American CD cases, current or former smoking had a tendency to be associated with increased need of biologic therapy and CD-related surgery (Biologics: Current 1.3 [0.47–3.64], Former 1.59 [0.47–5.41]; Surgery: Current 1.56 [0.64–3.81], Former 1.14 [0.36–3.65]). Stratified analysis and tests for interaction in logistic models suggested that sex, IBD family history, and country of residence do not significantly modify the association between smoking and IBD prevalence and outcomes.Conclusions:
In both China and USA, CD and UC cases were less likely to be current smokers at diagnosis and more likely to be former smokers. However, both current and former smokers tended to have a more severe disease course including having increased need for immunosuppressant and biologic therapy in Chinese CD population and for biologic therapy and surgery in American CD population.