Predicting functional decline in aging and Alzheimer’s disease with PET-based Braak staging

Abstract

Abstract The progression of PET-based Braak stages correlates with cognitive deterioration in aging and Alzheimer’s disease. Here, we investigate the association between PET-based Braak stages and functional impairment and assess whether PET-based Braak staging predicts a longitudinal decline in the performance of activities of daily living. In this cohort study, we evaluated cognitively unimpaired individuals and individuals with mild cognitive impairment or Alzheimer’s disease dementia. Participants underwent [18F]MK6240 tau-PET, were assigned a PET-based Braak stage at baseline and were followed for a mean (SD) of 1.97 (0.66) years. Functional performance was evaluated with the Functional Activities Questionnaire, Everyday Cognition and functional Clinical Dementia Rating sum of boxes. Multiple linear regressions assessed the association of PET-based Braak stages with baseline functionality and with the longitudinal rate of change in functional scores, adjusting for age, sex and amyloid-β load. We employed voxel-based regression models to investigate the association between functionality and tau-PET signal and assessed the voxel overlap with Braak regions of interest. We included 291 individuals (181 cognitively unimpaired, 56 amyloid-β+ mild cognitive impairment and 54 amyloid-β+ Alzheimer’s disease) aged 70.60 (7.48) years. At baseline, PET-based Braak stages III–IV (β = 0.43, P = 0.03) and V–VI (β = 1.20, P < 0.0001) showed associations with poorer Functional Activities Questionnaire scores. Similarly, stages III–IV (β = 0.43, P = 0.02) and V–VI (β = 1.15, P < 0.0001) were associated with worse Everyday Cognition scores. Only stages V–VI were associated with higher functional Clinical Dementia Rating sum of boxes (β = 1.17, P < 0.0001) scores. Increased tau-PET signals in all Braak regions of interest were linked to worse performance in all tools. The voxelwise analysis showed widespread cortical associations between functional impairment and tau-PET and high voxel overlap with Braak regions of interest. Baseline PET-based Braak stages V–VI predicted significant longitudinal functional decline as assessed by the Functional Activities Questionnaire (β = 1.69, P < 0.0001), the Everyday Cognition (β = 1.05, P = 0.001) and the functional Clinical Dementia Rating sum of boxes (β = 1.29, P < 0.0001). Our results suggest that functional impairment increases with the severity of tau accumulation. These findings also indicate that PET-based Braak staging is a good predictor of functional impairment in the Alzheimer’s disease continuum. Finally, our study provides evidence for the clinical significance of the PET-based Braak staging framework.