Abstract
We examined the relationship of diabetes and hemoglobin A1C (A1C) to 2 common causes of dementia. The study included 1228 subjects who underwent annual clinical evaluations and a brain autopsy at death, as part of a Rush longitudinal cohort study of aging. A total of 433 subjects had A1C data available. Neuropathologic evaluations documented the size and location of infarcts. Modified silver stain-based Alzheimer disease (AD) measures included global and regional scores. We used regression analyses to examine associations of diabetes and A1C with overall and regional neuropathology. Diabetes [odds ratio (OR)=0.94; 95% confidence interval (CI), 0.73-1.20) and A1C (OR=0.83; 95% CI, 0.62-1.10) were not associated with global AD pathology across the brain, nor with overall or individual measures of neuropathology in mesial temporal or neocortical regions separately (all P>0.05). Diabetes was associated with a higher odds of any infarct (OR=1.43; 95% CI, 1.07-1.90), and particularly with gross (OR=1.53; 95% CI, 1.14-2.06) but not microinfarcts (P=0.06), and subcortical (OR=1.79; 95% CI, 1.34-2.39) but not cortical infarcts (P=0.83). In summary, we found no relationship of diabetes or A1C with global or regional AD pathology, including in the mesial temporal lobe. Diabetes is associated with gross subcortical infarcts. Our results suggest that the diabetes-dementia link is based on subcortical vascular pathology and not on regional AD pathology.