Objective: To investigate whether serum cortisol levels are associated with cognitive performance and cognitive decline in elderly persons and whether this association differs by age, sex, and depression status.
Design: Data from the Longitudinal Aging Study Amsterdam, with repeated measurements of cognitive performance after 3 and 6 years.
Participants: A total of 1,154 persons, aged 65-88 years.
Measurements: Serum concentrations of total cortisol (CRT) and corticosteroid binding globulin (CBG) were measured at baseline, and from these free cortisol index (CRT/CBG) was computed. At baseline and 3 and 6 years of follow-up, global cognitive functioning, verbal memory performance, and speed of information processing were assessed.
Results: After adjustment for demographics, health, and life style variables, a significant association between high levels of free cortisol and poorer performance on verbal learning (B = -0.32; 95% confidence interval: -0.64 to -0.01) was found in both women and men. Additional adjustment for depression did not change this association. In women, but not in men, high levels of free cortisol (B = -0.85; 95% confidence interval: -1.40 to -0.31) were associated with slower speed of information processing. The associations between cortisol and cognitive performance were the same for the younger and the older old and for depressed and nondepressed persons. Higher levels of cortisol were not associated with cognitive decline over a period of 6 years.
Conclusion: Our study provides further evidence that high levels of cortisol measured during the day are associated with lower memory function and speed of information processing but not with decline in cognitive functioning over 6 years of time.