BRAIN INSULIN RESISTANCE AND ALZHEIMER'S DISEASE NEUROPATHOLOGY AMONG OLDER AUTOPSIED PERSONS WITH AND WITHOUT DIABETES

Abstract

Diabetes mellitus increases risk of dementia, yet little is known about the role of insulin resistance in the aging brain and its relation to the most common cause of dementia. We examined the relation of brain phosphorylated insulin receptor substrate 1 (IRS-1) and related markers of the leptin pathway, to Alzheimer’s disease (AD) pathology in community-dwelling elderly individuals with and without diabetes. We used data from 124 deceased and autopsied participants of the Rush Religious Orders Study (mean age-at-death =87 years; 52% female), 62 with and 62 without diabetes. Neuropathologic data (mean post-mortem interval=6.6 hours) included a continuous, overall AD pathology measure based on counts of neuritic plaques, diffuse plaques, and neurofibrillary tangles (NFT) on silver-stained sections, as previously published. We used ELISAs to assess insulin and leptin signaling measures (Cell Signaling Technology, Danvers MA) in frozen midfrontal gyrus cortical specimens. Linear and ordinal logistic regressions were used to examine the relation of brain measures and AD pathology. For analyses, the overall AD outcome measure was square-root transformed and other AD measures were categorized into quartiles. In the total group, IRS-1ps307 and IRS-1 were weakly correlated (rs=0.22; p=0.01), and STAT3pY705 and STAT3 were moderately correlated (rs=0.66; p<0.0001). Adjusting for age, sex, and education, a higher IRS-1ps307/IRS-1 ratio was associated with higher overall AD pathology (estimate=0.52, SE=0.24; p=0.03), and with the specific measure of neuritic plaques (OR=2.40, logOR SE=1.18; p=0.04) but not with diffuse plaques or NFT (both p>0.11). The STAT3pY705/STAT3 ratio, a marker of leptin sensitivity, was not associated with measures of overall AD, or neuritic or diffuse plaques or NFT (all p>0.08). Additional analyses examining for effect modification by diabetes, did not show any interaction of diabetes with either ratio (all p>0.09). This study of individuals with and without diabetes suggests that brain insulin resistance, as indicated by phosphorylated IRS-1, is associated with increased AD pathology, and neuritic plaques in particular. Furthermore, diabetes does not appear to modify observed associations of insulin resistance with AD. Funding: The National Institutes of Health grants P30AG010161, R01AG015819, R01NS084965.