Medical Histories Associated With Absence of Alzheimer Disease Neuropathologic Changes in the Oldest-Old: The 90+ Study.

Abstract

Exploration of medical histories and medications associated with Alzheimer disease neuropathologic change (ADNC) absence and potential resistance may identify protective factors against ADNC. This was a retrospective examination of data from participants age ≥90 years who enrolled in <i>The 90+ Study</i>, a longitudinal study based in California. Participants underwent neuropathologic analysis for the presence of neuritic amyloid plaques (NPs) (any), beta amyloid plaques (Thal phase > 0), and neurofibrillary tangles (>2). Odds ratios (ORs) for the presence of pathologic changes to <i>APOE</i> genotype, self-reported medical histories, and medication use were estimated by logistic regression and (a)djusted for sex, age at death (continuous), and education. We examined 267 participants; 75% were female, with a mean age at death of 98 (± 3.5) years. Variables associated with ADNC absence/presence were for NPs: heart disease (adjusted OR [aOR] = 0.46), <i>APOE</i>-ε4 (aOR = 3.48), angiotensin-converting enzyme (ACE) inhibitors (aOR = 0.46), cataracts (aOR = 0.26), and beta-blockers (aOR = 0.51); for Aβ: heart disease (aOR = 0.33) and <i>APOE</i>-ε4 (aOR = 8.10); and for neurofibrillary tangles: ACE inhibitors (aOR = 0.46), glaucoma (aOR = 0.29), HTN (aOR = 0.44), seizures (aOR = 0.16), <i>APOE</i>-ε2 presence (aOR = 0.39), and vasodilators (aOR = 0.36). These observations would not survive multiple comparison corrections. Cataract formation, <i>APOE</i> status, ACE inhibitors, and beta-blockers may be associated with ADNC absence because of differential lipid trafficking. Survival bias and type 1 errors warrant consideration.