Associations between lifestyle and cognitive function over time in women aged 40-79 years

Smoking, excessive drinking, and physical inactivity are associated with reduced cognitive function but the independence, domain specific cognitive effects, and trajectories of these associations are not firmly established.

Our aim was to examine these lifestyle-cognitive function associations in middle-to-older aged women across time.

Cohort study design with repeat surveys (2001, 2005, and 2008). Participants were volunteers from a random sample of Australian women on the Brisbane electoral roll; mean (±SD) age 60 ± 11 years in 2001. Outcome measures were the Mini-Mental State Examination (MMSE), Auditory Delayed Index (ADI), Visual Delayed Index (VDI), Working Memory Index (WMI), and Processing Speed Index (PSI).

489 women completed cognitive testing in 2001, 451 in 2005, and 376 in 2008. Mean (±SD) cognitive scores in 2001 were MMSE: 29.1 ± 1.2, ADI: 104.6 ± 13.4, VDI: 107.2 ± 14.0, WMI: 104.1 ± 12.3, and PSI: 102.7 ± 11.8. Multivariate adjusted mean scores (95% CI) over the 7-year study period were higher for moderate drinkers than non-drinkers for the MMSE (β = 0.32; 0.04, 0.61), the VDI (β = 4.33; 0.96, 7.70), and the WMI (β = 3.21; 0.34, 6.07). Current smokers performed worse than never-smokers for the MMSE (β = -0.35; 0.64, -0.06), the VDI (β = -3.91; -7.57, -0.26), the WMI (β = -3.42; -6.67, -0.18), and the PSI (β = -5.89; -8.91, -2.87). PSI was higher in women performing strenuous physical activity compared to inactive women (β = 2.14; 0.37, 3.90). None of the three lifestyle parameters influenced the changes in cognition across time.

Alcohol and exercise were associated with selective protective effects and tobacco with selective harmful effects on cognitive function in middle-to-older aged women. Associations remained consistent across time.