The role of cerebrovascular pathology in the association between amyloid‐β and tau in cognitively unimpaired and impaired individuals

Abstract

AbstractBackgroundAlthough cerebrovascular pathology often co‐exists with Alzheimer’s disease pathology, it remains incompletely understood to what extent it can contribute to the progression of Alzheimer’s disease pathology. We investigated whether cerebrovascular pathology (white matter lesions [WML] and microbleeds) modified the association between amyloid‐ß (Aß) and both cross‐sectional and longitudinal tau in cognitively unimpaired (CU) and cognitively impaired (CI) individuals.MethodWe included 1240 participants from the Swedish BioFINDER‐2 cohort including 402 Aß‐negative CU, 147 Aß‐positive CU, 257 Aß‐negative CI (MCI and non‐AD dementia) and 434 Aß‐positive CI participants (MCI and AD dementia) (Table 1). All underwent baseline [18F]RO948 (tau)‐PET, and a subset (n = 678) underwent longitudinal tau‐PET (2.6±1.0‐years). We extracted tau‐PET SUVr in a temporal meta‐region‐of‐interest. For Aß, we used global [18F]flutemetamol‐PET SUVr for CU and Aß‐status (+/‐) for CI participants. Global WML volume was calculated using the SPM lesion segmentation tool, corrected for intra‐cranial volume and log‐transformed. Presence of microbleeds was dichotomized into absent or present ( = 1) by visual read. Primary analyses included age‐, sex‐ and APOE e4‐corrected linear mixed models between tau‐PET (outcome) and interactions between time, Aß and WML volume/microbleeds (predictors).ResultUpon correcting for age, sex and e4‐carriership, Aß‐positive CU and CI participants showed significantly more microbleeds compared to Aß‐negative participants of the same cognitive stage (both p<0.05), but no differences in WML volume was observed (both p>0.05). In CU, we observed a significant interaction effect between = 1 microbleed and higher Aß‐load with both greater cross‐sectional tau (ß = 0.67 [CI: 0.45‐0.89], p<0.001) and longitudinal tau (ß = 0.11 [CI: 0.07‐0.16], p<0.001) (Fig‐1, Table‐2). In CU, we also observed a significant interaction effect between larger WML volumes and higher Aß‐load with greater cross‐sectional tau (ß = 0.14 [CI: 0.06‐0.22], p = 0.001). When including both microbleeds and WML volume in the same model, only effects of microbleeds remained significant (Table‐2). No significant effects were observed in CI participants.ConclusionIn CU (but not CI) individuals, cerebrovascular pathology may modify the association between Aß and tau by accelerating tau accumulation in the presence of Aß. These results might suggest an early effect of cerebrovascular pathology on the interplay between Aß and tau.