Smoking and mortality in a prospective cohort study of elderly Chinese in Hong Kong

Background and AimsLarge cohort studies on smoking and mortality in elderly people are scarce, and few studies examined smokers aged 85+ years separately. We estimated the risks of all-cause and cause-specific mortality due to smoking in an elderly Chinese cohort in Hong Kong. DesignA population-based prospective cohort of 65510 Chinese enrolled from 1998 to 2001 and followed until May 2012. SettingAll 18 Elderly Health Service centres in Hong Kong, China. ParticipantsElderly people aged 65+ years. MeasurementsSelf-reported smoking status was assessed at baseline interview and categorized as never, former and current smokers. FindingsCompared with never smokers, after adjustment for sex, age, education, social security assistance, housing type, monthly expenditure, alcohol use, depressive symptoms and health status, the hazard ratio (HR) for current smokers was 1.89 [95% confidence interval (CI)=1.81-1.98] for all participants aged 65+ years at baseline, corresponding to an attributable fraction (AF) of about 50%, which is based on AF=(HR-1)/HR. As the effect of smoking varied with age (P for age interaction <0.001), subgroup analysis by age group showed that the adjusted HR for current smokers aged 65-84 years was 1.93 (95% CI=1.84-2.03), and for 85+ years was 1.29 (95% CI=1.05-1.58). All the risk estimates did not vary by sex (P for sex interaction ranged 0.74-0.89). ConclusionsIn Hong Kong, the risk of death from smoking appears to be the same for Chinese women as it is for men. Half of all deaths in Chinese smokers aged 65 years and older and a quarter of all deaths in Chinese smokers aged 85 years and older are caused by smoking-attributable diseases.