Smoking and risk of vascular and non-vascular mortality in old age: 15-year follow-up of 7000 men in the Whitehall resurvey

Abstract

Objective: To assess the relevance of tobacco smoking to cause-specific mortality in older men. Methods: In a prospective study of surviving participants of the Whitehall study of male civil servants initially recruited in 1967-1970, 7044 were resurveyed in 1997-8 (mean age 77 years; range 66-97), of which 7024 reported information on smoking habits and had follow-up information on cause-specific mortality up to the end of 2012. Hazard ratios (HR) in relation to smoking habits in old age were estimated using Cox proportional hazards models after adjustment for age, last known employment grade and previous diagnoses of vascular disease or cancer. Results: At resurvey in 1997-98, 13% of men were current smokers (median cigarette consumption 9 per day), 58% were former-smokers (median time since quitting 25 years) and 23% were never-smokers. The remaining 5% reported being a never-smoker in the resurvey, but not in the initial survey in 1967-70, and were handled as a separate category. During a median follow-up of 15 years, there were 4965 deaths (74/1000/year), of which 2063 were from cardiovascular disease \[CVD\] (31/1000/yr), 1167 from cancer (17/1000/yr]), 802 from respiratory (12/1000/yr) and 933 (14/1000/yr) from other causes. Compared with never smokers, average mortality rates among former smokers were 15% higher (HR 1.15, 1.07-1.23), due chiefly to increased risks of death from cancer (1.24, 1.07-1.44) and respiratory disease (1.58, 1.29-1.93). The excess risk associated with being a former smoker varied depending on years since quitting; among men who had quit within the preceding 25 years (median 14 years), mortality rates were 28% higher than those observed among never smokers (HR 1.28, 1.18-1.39), whereas among men who quit 25 or more years earlier (median 35 years earlier), no significant excess risk was observed when compared with never smokers (HR 1.05, 0.97-1.13). Compared with older men who had never regularly smoked, older men who remained current smokers in their seventies had 50% higher total mortality rates (HR 1.50, 95% CI 1.36-1.64), due to increased mortality from CVD (1.34, 1.16-1.55), cancer (1.74, 1.44-2.11), and respiratory disease (2.39, 1.87-3.04). Among men who survived to age 70, the probability of survival to age 85 was 65% for never smokers compared with 48% for current smokers, with current smokers losing, on average, 3 to 4 years of life expectancy. Conclusions: Among men who survive into their seventies, continuing to smoke is associated with persistent excess vascular and non-vascular mortality which translate into important differences in remaining lifespan.