Patterns and predictors of smoking cessation in an elderly cohort.

Objectives: To identify subject characteristics that predict smoking cessation and describe patterns of cessation and recidivism in a cohort of elderly smokers.

Design: Prospective cohort study.

Setting: Piedmont region, North Carolina.

Participants: Five hundred seventy-three subjects enrolled in the North Carolina Established Populations for Epidemiologic Studies of the Elderly who responded "yes" to question 179 on the baseline survey (Do you smoke cigarettes regularly now?) and survived at least 3 years, until the next in-person follow-up (1989/90). Subjects were classified as quitters (n=100) or nonquitters (n=473) based on subsequent smoking behavior.

Measurements: Reported smoking behavior, demographic characteristics of the smokers at baseline or subsequent interviews, 7-year mortality.

Results: After controlling for all characteristics studied, subjects who quit smoking were more likely to be female (P=.03) and showed a trend toward greater likelihood of a recent cancer diagnosis (P=.11). Recidivism was observed in only 16% (19/119) of subjects who reported no smoking in 1989/90. The percentage of subjects who died during 7 years of follow-up was 44.0% of quitters, compared with 51.6% of nonquitters. Smoking cessation was not associated with a statistically significant decrease in risk of death after controlling for other variables (odds ratio=0.78, 95% confidence interval=0.48-1.26).

Conclusion: Smoking cessation in this elderly cohort was associated with different subject characteristics from those that predict successful cessation in younger populations, suggesting that older smokers may have unique reasons to stop smoking. Further study is needed to assess potential motives and benefits associated with smoking cessation at an advanced age.