Insulin and Cognitive Function in an Elderly Population: The Rotterdam Study

Abstract

OBJECTIVE To examine the association between insulin and cognitive function in the population-based Rotterdam Study. RESEARCH DESIGN AND METHODS Serum insulin was measured 2 h after an oral glucose load, while global cognitive function was assessed by the Mini Mental State Examination in 5,510 subjects, aged 55 years and over. RESULTS An increase in postload insulin only in women was associated with a decrease in cognitive function (age-adjusted regression coefficient −0.10 per 50 mU/l insulin; 95% CI −0.16 to −0.04). The association between insulin resistance, assessed by the ratio of postload insulin over glucose, and cognitive function was not statistically significant. Further adjustment for the individual risk factors of serum glucose, BMI, HDL, systolic blood pressure, smoking, or use of estrogen did not change the results. The association was present in women with and without cardiovascular disease and present after excluding subjects with diabetes. Women with dementia, the extreme of cognitive dysfunction, had increased age-adjusted insulin levels (76.3 ± 4.9 vs. 66.8 ± 1.0 mU/l [means ± SE], P = 0.06). CONCLUSIONS The results of this study suggest that increased serum insulin may be associated with decreased cognitive function and dementia in women. These findings are more compatible with a direct effect of insulin on the brain than with an effect through an increase in cardiovascular risk factors.