142 Chapter 6 ABSTRACT Background We aimed to assess the association between (change in) tau pathology and relative cerebral blood flow (rCBF) with cognitive decline over time in individuals along the Alzheimer’s disease (AD) spectrum. Methods Fifty-four cognitively unimpaired (CU; mean age 66 ± 7, 54% female, 37% amyloid-β positive (Aβ+)) and 75 cognitively impaired (CI; mean age 65 ± 8, 52% female, 100% Aβ+, MCI/dementia = 6/69) individuals underwent dynamic 18F-flortaucipir PET and longitudinal neuropsychological testing (mean follow-up 4.4 ± 2.1 years). Fourty-nine CU and 37 CI individuals additionally underwent follow-up 18F-flortaucipir PET after approximately 2 years. Binding potential (BPND) and R1 were extracted as measures for tau pathology and rCBF. Regional BPND (Braak I, III/IV and V/VI) or rCBF (global) were used as predictors in linear mixed models, adjusted for age-at-PET, sex and education. Neuropsychological test scores on MMSE, RAVLT immediate recall and TMT-B (covering global cognition, memory, and executive functioning) were used as outcome. Separate models were run for each region, predictor (baseline and annual change) and in CU and CI individuals. Lastly, a model including both tau PET BPND and rCBF as predictors was performed. All results were corrected for multiple comparisons using the Bejamini Hochbergs false discovery rate (FDR). Results Higher baseline tau PET strongly predicted decline in most cognitive tests in both CI and CU individuals and these effects were more apparent after correction for rCBF. For rCBF, lower baseline rCBF was predictive of steeper RAVLT in CU, but not of the other cognitive tests. Furthermore, a higher rate of accumulation in tau PET also strongly predicted multi-domain cognitive decline in both CI and CU individuals, but only when corrected for (change in) rCBF. A steeper decrease in rCBF predicted a steeper decline on MMSE in CI individuals only. Conclusion Baseline and change in tau load are predictive for multi-domain longitudinal cognitive trajectories in both CI and CU individuals, while reduced rCBF is only subtly predictive. This underlines that tau PET is a valuable tool in predicting domain-specific cognition and indicates that tau pathology and changes in cerebral perfusion are (partly) independent factors predicting cognitive decline, with tau pathology outperforming blood flow.
RkJQdWJsaXNoZXIy MjY0ODMw