Hongchun Wei, Tianhao Zhang, Zhigang Liang, Min Li, Min Kong, Maowen Ba
Aβ pathology modifies the longitudinal cognitive trajectory associated with APOE ε4 in pre-dementia adults over a 2-year period. While consistent with the antagonistic pleiotropy framework, this observation requires independent replication, and alternative explanations including cognitive reserve and survivor bias warrant careful consideration.
BACKGROUND: The apolipoprotein E (APOE) ε4 allele is the strongest genetic risk factor for late-onset Alzheimer's disease (AD), yet emerging evidence suggests its cognitive effects may be age- and pathology-dependent. Whether and how amyloid-β (Aβ) pathology modifies the longitudinal cognitive trajectory associated with APOE ε4 in the pre-dementia stage remains unclear.
METHODS: We conducted a 2-year longitudinal study in 70 pre-dementia adults from the Alzheimer's disease Neuroimaging Initiative (ADNI), assessing APOE genotype, Aβ status via positron emission tomography (PET), multi-domain cognition, structural magnetic resonance imaging (MRI) volumes, and resting-state functional connectivity of a predefined hippocampus-auditory network. Linear mixed-effects models evaluated the three-way interaction of time, APOE ε4 carrier status, and Aβ status on longitudinal trajectories. Sensitivity and exploratory mediation analyses were performed.
RESULTS: A significant three-way interaction (Time × APOE ε4 × Aβ status) was observed for global cognitive decline (ADAS13: β = -4.54, p = 0.001, P_FDR = 0.003), indicating that Aβ pathology moderates the effect of APOE ε4 on cognitive trajectories. Post-hoc analyses revealed that among Aβ-positive individuals, APOE ε4 carriers exhibited a slower rate of cognitive decline compared to non-carriers, whereas no such difference was evident in Aβ-negative individuals. This moderating effect was robust to sensitivity analyses and was consistently observed across multiple cognitive domains (MMSE, CDRSB, FAQ, MoCA; all P_FDR < 0.01). However, no significant three-way interactions survived multiple comparison correction for regional brain volumes or hippocampus-auditory functional connectivity, though nominally significant trends were observed in middle temporal gyrus volume and specific temporal lobe connections. Longitudinal hippocampal atrophy did not mediate the observed association.
CONCLUSION: Aβ pathology modifies the longitudinal cognitive trajectory associated with APOE ε4 in pre-dementia adults over a 2-year period. While consistent with the antagonistic pleiotropy framework, this observation requires independent replication, and alternative explanations including cognitive reserve and survivor bias warrant careful consideration.