Setting: The increasing numbers of elderly people highlight the question of smoking effects in this age group.
Objective: To investigate whether there is a relationship between smoking and 8-year mortality in a representative elderly cohort aged 65 years and over, residing in the department of Gironde, in the south-west of France.
Methods: Data were collected by a questionnaire administered during home visits. Every death was systematically recorded during the 8-year follow-up.
Results: The 2786 (99.8%) subjects included in the study were categorised into current smokers, former smokers and never smokers. Mortality was higher (P < 0.0001) in men (39.6%) than in women (29.4%). In men, the risk of mortality was higher (P = 0.01) among current (44.5%) and former smokers (41.4%) than in never smokers (32.5%). The risks were similar in never smokers and former smokers who had stopped more than 20 years previously. Tobacco consumption in pack-years was higher (P = 0.02) in those people who died during the follow-up period than in those who were still alive. Mortality due to respiratory disease (P = 0.008) and lung cancer (P < 0.0001) was significantly higher in current smokers than in former and never smokers. Adjusting for potential confounding factors, smoking remains significantly associated with mortality both in current (relative risk [RR] = 1.7, P < 0.001) and former smokers (RR = 1.3, P = 0.004).
Conclusion: These results confirm the consequences of smoking even in the elderly, and raise the question of smoking prevention after 65 years of age.