Cognitive function in an elderly population with persistent impaired glucose tolerance.

To study cognitive function in an elderly population with persistent impaired glucose tolerance (IGT). Fasting and postload 2-h plasma glucose and insulin levels were determined at baseline in a population-based sample of 1,300 people and repeated an average of 3.5 years later in 980 subjects. At follow-up, cognitive function was evaluated in subjects with persistent normal glucose tolerance (NGT; n = 506) and IGT (n = 80) with a brief neuropsychological test battery. Subjects with persistent IGT scored lower in the Mini-Mental State Examination (MMSE) and in the Buschke Selective Reminding Test long-term memory scores. Multiple linear regression analysis revealed that age, education, and insulin levels (either fasting or 2-h value) were associated with the MMSE score in subjects with persistent IGT. Other potential risk factors for impaired cognitive function were not significantly associated with the MMSE score. Our study showed that persistent IGT in the elderly is associated with mildly impaired cognitive function, and hyperinsulinemia may account for this association.