Testing Differential Associations Between Smoking and Chronic Disease Across Socioeconomic Groups: Pooled Data From 15 Prospective Studies

Abstract

Background: For the same quantity of cigarettes smoked, relative to more affluent people, socioeconomically disadvantaged people have higher levels of smoking biomarkers. This may be ascribed to inhaling cigarette smoke more deeply and more frequently and/or choosing higher tar-containing brands. We investigated whether this increased tobacco load, as captured using cotinine measurements, is associated with a greater risk of mortality in lower social groups. Methods: We used Cox proportional hazards models stratified by socioeconomic position to calculate hazard ratios in a pooled sample of 15 English and Scottish prospective cohort studies (N = 81,476). Results: During a mean (SD) follow-up of 10.3 (4.4) years, 8234 deaths occurred. Risk of total mortality (hazard ratio; 95% confidence interval) for smokers relative to never-smokers in the high (2.5; 2.1, 3.1), intermediate (2.1; 1.8, 2.4), and low (2.0; 1.9, 2.2) educational groups did not differ markedly (P for interaction=0.61). Similar findings emerged when using cause-specific outcomes and occupational social class and housing tenure as socioeconomic indices. Conclusion: Contrary to our hypothesis, we found no indication that chronic disease mortality associated with smoking was higher in disadvantaged people.