Abstract P349: Association of Lower Fasting Glucose and Visit-to-visit Glucose Variability With Cognitive Performance: The Multi-ethnic Study of Atherosclerosis

Abstract

Background: Diabetes is a well-known risk factor for dementia. Despite emerging evidence that higher glucose variability is associated with cognitive decline and hypoglycemia is associated with increased risk for dementia, few studies have investigated the relationship of lower fasting glucose and glucose variability with cognitive performance in the general population. Hypothesis: Diabetic and lower levels of fasting glucose and higher visit-to-visit glucose variability over 10 years are associated with worse cognitive performance. Methods: Participants in the Multi-Ethnic Study of Atherosclerosis (N = 4,591; mean age 69.7 ± 9.4 years) completed a cognitive assessment at Exam 5 including the Cognitive Abilities Screening Instrument (CASI), Digit Symbol Coding (DSC), and Digit Span (DS, Forward and Backward combined) tests measuring global cognitive performance, executive function/processing speed, and working memory, respectively. We defined fasting glucose categories as low (<80 mg/dL), normal (80-99 mg/dL), impaired (100-125 mg/dL), and diabetic (≥126 mg/dL). Participants with diabetes were included in the diabetic category. We examined the relationship of Exam 1 and Exam 5 fasting glucose categories and visit-to-visit glucose variability (coefficient of variation across Exams 1-5) to scores on the CASI (0-100), DSC (0-133), and DS (0-28). We report regression coefficients ( β [95% confidence limits]) from regression models adjusted for age, sex, race/ethnicity, education, income, cardiovascular risk factors, and APOE genotype. Results: Relative to normal glucose, participants with diabetic fasting glucose performed significantly worse on the CASI (Exam 1 β = -0.75 [-1.36, -0.15]; Exam 5 β = -0.30 [-0.83, 0.23]), DSC (Exam 1 β = -2.97 [-4.20, -1.74]; Exam 5 β = -1.26 [-2.34, -0.18]), and DS (Exam 1 β = -0.39 [-0.72, -0.07]; Exam 5 β = -0.29 [-0.58, -0.003]). In contrast, participants with low Exam 1 fasting glucose performed significantly better on the DSC (Exam 1 β = 1.55 [0.45, 2.65]) and similarly on the CASI and DS compared to those with normal fasting glucose. Adjusting for glucose lowering medication attenuated associations with the DSC and DS but not the CASI. Accounting for heterogeneity in the low glucose category by hypoglycemia (<70 mg/dL) did not significantly alter results. Higher visit-to-visit glucose variability from Exam 1 to Exam 5 was associated with worse performance on the CASI ( β = -0.36 [-0.60, -0.12]) and DSC ( β = -1.14 [-1.63, -0.65]). Associations were consistent across race and sex. Conclusions: Results support the hypothesis that hyperglycemia and higher glucose variability are associated with worse cognitive performance. Despite previous suggestion of a link between hypoglycemia and incident dementia, lower antecedent fasting glucose may be associated with better executive function/processing speed in this population-based sample.