Abstract
OBJECTIVE: To evaluate whether APOE ε4 is associated with severity of Lewy body (LB) pathology, independently of Alzheimer disease (AD) pathology. METHODS: Six hundred fifty-two autopsy-confirmed LB disease (LBD) cases and 660 clinical controls were genotyped for APOE. In case-control analysis, LBD cases were classified into 9 different groups according to severity of both LB pathology (brainstem, transitional, diffuse) and AD pathology (low, moderate, high) to assess associations between APOE ε4 and risk of different neuropathologically defined LBD subgroups in comparison to controls. In LBD cases only, we also measured LB counts from 5 cortical regions and evaluated associations with ε4 according to severity of AD pathology. RESULTS: As expected, APOE ε4 was associated with an increased risk of transitional and diffuse LBD in cases with moderate or high AD pathology (all odds ratios ≥3.42, all p ≤ 0.004). Of note, ε4 was also associated with an increased risk of diffuse LBD with low AD pathology (odds ratio = 3.46, p = 0.001). In the low AD pathology LBD subgroup, ε4 was associated with significantly more LB counts in the 5 cortical regions, independently of Braak stage and Thal phase (all p ≤ 0.002). CONCLUSIONS: Our results indicate that APOE ε4 is independently associated with a greater severity of LB pathology. These findings increase our understanding of the mechanism behind reported associations of ε4 with risk of dementia with Lewy bodies and Parkinson disease with dementia, and suggest that ε4 may function as a modifier of processes that favor LB spread rather than acting directly to initiate LB pathology.