Do Cardiovascular Risk Factors and Cardiovascular Disease Explain Sex Differences in Cognitive Functioning in Old Age?

Abstract

Background: Sex differences in cognitive functioning in old age are known to exist yet are still poorly understood.

Objective: This study examines to what extent differences in cardiovascular risk factors and cardiovascular disease between men and women explain sex differences in cognitive functioning.

Methods: Data from 2,724 older adults from the Longitudinal Aging Study Amsterdam were used. Information processing speed and episodic memory, measured three times during six years of follow-up, served as outcomes. The mediating role of cardiovascular risk factors and cardiovascular disease was examined in single and multiple mediator models. Determinant-mediator effects were estimated using linear or logistic regression, and determinant-outcome and mediator-outcome effects were estimated using linear mixed models. Indirect effects were estimated using the product-of-coefficients estimator.

Results: Women scored 1.58 points higher on information processing speed and 1.53 points higher on episodic memory. Several cardiovascular risk factors had small mediating effects. The sex difference in information processing speed was mediated by smoking, depressive symptoms, obesity, and systolic blood pressure. The sex difference in episodic memory was mediated by smoking, physical activity, and depressive symptoms. Effects of smoking, LDL cholesterol, and diabetes mellitus on information processing speed differed between men and women.

Conclusion: Differences in cardiovascular risk factors between women and men partially explained why women had better cognitive functioning. A healthy cardiovascular lifestyle seems beneficial for cognition and sex-specific strategies may be important to preserve cognitive functioning at older age.

Keywords: Aged; behavior; cardiovascular diseases; cognition; epidemiology; episodic memory; sex.

Fig. 1

Single mediator model.

Fig. 2

Sex differences in cognitive functioning over time. The sex difference in information processing speed was 1.58 (95% C.I.: 1.07–2.09, p < 0.01), the sex difference in episodic memory was 1.53 (95% C.I.: 1.36–1.71, p < 0.01). For models with blood pressure, total cholesterol, HDL cholesterol or LDL cholesterol, total effects of sex on information processing speed were 1.57 (95% C.I.: 1.06–2.07, p < 0.01), 1.32 (95% C.I.: 0.67–1.97, p < 0.01), 0.67 (95% C.I.: –0.19–1.53, p = 0.13) and 0.66 (95% C.I.: –0.19–1.53, p = 0.13), respectively, and the total effects of sex on episodic memory were 1.53 (95% C.I.: 1.35–1.70, p < 0.01), 1.45 (95% C.I.: 1.22–1.67, p < 0.01), 1.55 (95% C.I.: 1.24–1.85, p < 0.01), and 1.55 (95% C.I.: 1.24–1.85, p < 0.01), respectively.

Fig. 3

Multiple mediator models. Arrows pointing from sex towards the mediator represent a-paths, arrows pointing from the mediator towards cognitive functioning represent b-paths. The indirect effect estimates (95% C.I.) are placed above or below the mediator. Mediators in bold indicate statistically significant indirect effect estimates. c = total effect, c’=direct effect. A) model for information processing speed, also includes a sex-by-time interaction (0.05 (0.03), p = 0.05) and mediator-by-time interactions for systolic blood pressure and physical activity. B) model for episodic memory, also includes a mediator-by-time interaction for systolic blood pressure.