Plasma C-peptide and cognitive performance in older men without diabetes.

BACKGROUND: Emerging evidence suggests that type 2 diabetes may be related to diminished cognition, but little data are available directly regarding the role of insulin levels. OBJECTIVE: The objective of this prospective cohort study was to examine the relation of insulin secretion to cognitive function among men without diabetes. SETTING: The study setting was the Physicians' Health Study-U.S. male physicians. PARTICIPANTS: Three hundred sixty-seven men who provided blood samples in 1982, when they had no lifetime history of diabetes and ranged in age from 47-65 years (mean age: 57 years). MEASUREMENTS: The authors assayed plasma C-peptide, reflecting insulin secretion, in the stored blood samples. Beginning in 2001, an average 18 years after blood collection, the authors administered telephone interviews, including tests of general cognition (Telephone Interview of Cognitive Status [TICS]), verbal memory, and category fluency. The authors used regression models to estimate mean differences in cognitive performance across levels of C-peptide controlling for a wide variety of potential confounding factors. RESULTS: On the TICS, men in the top tertile of C-peptide performed significantly worse than those in the bottom (multivariable-adjusted mean difference: -1.01 points, 95% confidence interval: -1.78 to -0.24); this apparent impact of C-peptide on cognition was equivalent to the cognitive differences the authors observed between men 6 years apart in age. Performance on the global score (combining results from all the individual tests) and verbal memory score (combining results from four tests of verbal memory) appeared lower among men in the highest C-peptide tertile, but results were not statistically significant. CONCLUSION: Higher midlife insulin secretion may be related to decreased later-life cognitive function, even among men without diabetes.