Improving statistical modeling with a neuropathology‐based <i>APOE</i> genetic risk score

Background The apolipoprotein E gene (APOE) is the predominant genetic risk factor for late-onset Alzheimer disease (AD), with three alleles contributing to and strongly associated many AD endophenotypes: ε2 (reduced risk), ε3 (reference), ε4 (increased risk). Researchers often include APOE carrier status (ε4+/-) in analyses account effect; however, this does not adequately protective effects of or heterogeneous effect different genotypes (ε2ε2, ε2ε3, ε2ε4, ε3ε3, ε3ε4, ε4ε4). We hypothesized that accounting full genotype would better recapitulate influence statistical than status. Method To generate a weighted score (APOE-npscore), we log-transformed odds ratios from recent study reporting neuropathology-confirmed AD. regressed cerebrospinal fluid (CSF) endophenotypes on APOE-npscore using data Wisconsin Registry Alzheimer’s Prevention (WRAP: 233 cognitively unimpaired (CU), 3 mild cognitive impairment (MCI-AD)) Disease Research Center (WADRC: 287 CU, 42 MCI-AD, 48 dementia-AD). In preliminary analyses, APOE-npscores were tested against CSF amyloid-beta1-42 (Aβ42) phosphorylated tau 181 (pTau181) measured Elecsys® immunoassays (Roche Diagnostics International Ltd, Rotkreuz, Switzerland). Result After adjusting sex age at lumbar puncture, was Aβ42 (β = -115.9, p 5.7e-17) pTau181 2.05, 4.7e-8). also -238.0, 1.9e-13) 4.25, 1.1e-6); models explained 2.4% more variation 0.8% These findings similar subset participants. Conclusion Our results demonstrate benefit pseudo-continuous variable like neuropathology-based instead binary allows researchers use continuous measure collapse important categories. This increases power, avoids modeling issues result small cell sizes, nuanced interpretation analyses.