Dynamic associations between tau aggregation, atrophy, and cognitive decline in Alzheimer’s disease

Tau aggregation measured with PET and neurodegeneration measured with MRI are closely associated with cognitive decline in Alzheimer’s disease, but the independent contributions of baseline and longitudinal measures of these imaging markers to cognitive decline remain unclear. Here, we tested a) independent associations between baseline and rate of change in tau-PET or MRI and cognitive decline, b) relative contributions of baseline and rate of change in tau-PET or MRI vs cognitive decline and c) effects of simulated treatment-induced reductions of tau aggregation on cognitive trajectories.

We included n =761 amyloid-positive individuals from the Swedish BioFINDER-2 study ( n =253 cognitively unimpaired; n =508 cognitively impaired), with [ ¹⁸ F]RO948-tau-PET, 3T structural MRI and cognition, including MMSE and a cognitive composite, of whom n =322 participants had longitudinal imaging data ( n =120 cognitively unimpaired; n =202 cognitively impaired). Tau-PET SUVR and cortical thickness measures were quantified in entorhinal, amygdala (tau-PET only), inferior middle temporal, neocortical regions and hippocampal volume (MRI only). Associations between imaging measures and cognitive decline were assessed using linear mixed models with random intercepts and slopes or linear regressions, adjusting for age, sex, education, diagnosis and other imaging modality.

Baseline and longitudinal tau-PET showed stronger associations with cognitive decline than MRI, with the strongest effects for inferior-middle-temporal and neocortical regions for baseline tau-PET (MMSE: inferior-middle-temporal: β =-0.18±0.01, p <0.001; neocortical: β =-0.17±0.01, p <0.001; cognitive-composite: inferior-middle-temporal: β =-0.22±0.02, p <0.001; neocortical: β =-0.25±0.02, p <0.001), and neocortical composite for longitudinal tau-PET (MMSE: β =-0.56±0.04, p <0.001; and cognitive-composite: β =-0.62±0.05, p <0.001). When correcting for the other imaging modality, associations for tau-PET showed smaller reductions than MRI at baseline (14.9% for tau-PET vs 46.7% for MRI). Serial mediation models showed that baseline tau-PET explained the largest proportion of cognitive decline (54.0-94.0%), with modest mediation effects for longitudinal tau-PET or MRI pathways (2.0-15.0%). Simulated treatment-induced partial reduction of tau aggregation showed marginal effects on cognitive trajectories (MMSE: placebo vs 70% slope reduction: Δ _{rate-of-change} =0.35[0.18-0.53], p <0.001, placebo vs 50%: Δ _{rate-of-change} =0.25[0.08-0.43], p <0.001, cognitive-composite: placebo vs 70%: Δ _{rate-of-change} =0.10[0.04-0.16], p <0.001) compared to halting tau progression (MMSE: placebo vs no-progression: Δ _{rate-of-change} =1.03[0.85-1.22], p <0.001, cognitive-composite: placebo vs no-progression: Δ _{rate-of-change} =0.33[0.27-0.39], p <0.001).

In conclusion, our results demonstrated the greater utility of tau-PET than MRI as a prognostic and disease monitoring marker, across the clinical spectrum of Alzheimer’s disease. In addition, baseline tau load showed stronger associations with future cognitive decline than tau progression. Furthermore, simulated reductions of tau progression showed limited effects on cognitive trajectories compared to halting tau progression and may point towards the need for early tau interventions.