Serum Insulin and Cognitive Performance in Older Adults: A Longitudinal Study

Purpose The aim of this study was to examine the association of serum glucose, insulin, and insulin resistance with cognitive functioning 7 years later in a longitudinal population-based study of Finnish older adults. Methods Serum glucose and insulin were measured at baseline in 269 dementia-free individuals aged 65-79 years, from the Cardiovascular Risk Factors, Aging, and Dementia (CAIDE) study. Insulin resistance was estimated with the homeostasis model assessment (HOMA-IR). Participants were reexamined 7 years later, and global cognition, episodic memory, executive functioning, verbal expression, and psychomotor speed were assessed, both at baseline and at follow-up. Multiple linear regression was used to investigate the associations with cognitive performance at follow-up, after adjusting for several potential confounders, including common vascular risk factors. Results In the multivariable-adjusted linear regression models, no associations of insulin resistance with cognitive functioning were observed. After excluding 19 incident dementia cases, higher baseline HOMA-IR values were related to worse performance in global cognition (β [standard error (SE)] -.050 [0.02]; P = .043) and psychomotor speed (β [SE] -.064 [.03]; P = [.043]) 7 years later. Raised serum insulin levels were associated with lower scores on global cognition (β [SE] -.054 [.03]; P = .045) and tended to relate to poorer performance in psychomotor speed (β [SE] -.061 [.03]; P = .070). Conclusions Serum insulin and insulin resistance may be independent predictors of cognitive performance 7 years later in elderly individuals without dementia. Randomized controlled trials are needed to determine this issue.