The association between adult-life smoking and age-related cognitive decline in Danish men

BACKGROUND: Most previous studies of effects of smoking on age-related cognitive decline have compared cognitive decline in current smokers, former smokers, and never smokers rather than investigating the effects of pack-years. The aim of the present study was to analyze the association between smoking and age-related cognitive decline in a sample of men administered the same intelligence test in young adulthood and late midlife, using pack-years between the two assessments as the primary measure of exposure to smoking.

METHODS: In 5052 men, scores on a military intelligence test (BPP, Børge Priens Prøve) were available from young adulthood and a late midlife follow-up assessment including the same intelligence test and a comprehensive questionnaire on socio-demographic factors, lifestyle, and health. Information on smoking was self-reported at follow up for eight age periods, and pack-years were calculated from age 19 based on information on daily smoking and the duration of each age period. The differences in cognitive decline between adult-life smokers and non-smokers and the differences between light, moderate, and heavy smokers defined by pack-years were analyzed in linear regression models.

RESULTS: All smoking variables were only weakly associated with cognitive decline. Comparison of adult-life smokers and non-smokers showed less cognitive decline among smokers (1.12 IQ points, p <  0.001). Among smokers, analyses of pack-years suggested a weak dose-response relationship with more decline in heavy smokers than in light smokers (1.33 IQ points, p =  0.001). Independent of pack-years, current smoking was associated with larger cognitive decline than former smoking (1.73 IQ points, p <  0.001).

CONCLUSION: Smoking explained negligible fractions of the variance in cognitive decline, and thus our results did not indicate that smoking is a strong predictor of cognitive decline. The effects of pack-years suggest a relatively weak, possibly cumulative effect of smoking across the adult lifespan. The difference in decline between smokers and non-smokers may reflect participation bias and selective attrition at follow-up while the effects of current smoking may reflect either temporary effects of smoking or individual and life-style characteristics associated with continuation of smoking into late midlife.