Brain activation during episodic scene encoding is associated with amyloid and tau levels in amyloid‐positive older adults

Abstract

AbstractBackgroundThe relationship between amyloid and tau deposition and episodic memory‐related brain activity is understudied. Our goal was to determine the association between brain activation on fMRI during encoding of complex scenes and amyloid and tau deposition on PET.Methods91 individuals from the Indiana Memory and Aging Study (36 cognitively normal (CN), 27 subjective cognitive decline (SCD), 28 mild cognitive impairment (MCI)) underwent amyloid ([18F]florbetapir or [18F]florbetaben) and [18F]flortaucipir PET, structural and functional MRI, and clinical and cognitive testing. The fMRI task was a scene encoding paradigm in which the participant is asked to try to remember a set of complex scenes relative to viewing a scrambled control image. Amyloid Centiloid (CL) and tau SUVR images were generated using standard methods. Amyloid positivity was defined as global cortical CL≥21. Areas activated during the task (“main effect” (ME)) across all participants were determined (Figure 1A). Mean BOLD signals from ME regions, the entorhinal cortex (EC), and the inferior parietal lobule (IPL) were extracted. Mean cortical amyloid CL and tau SUVR from the bilateral EC, IPL, and lateral temporal lobe (LTL) were also extracted. Tau and amyloid were non‐normally distributed, so Spearman correlation models were used to evaluate the relationship of scene encoding activation and PET. Age, diagnosis, and sex were evaluated as covariates but did not change the observed pattern of results.ResultsSignificant association between tau and scene encoding activation was observed (Figure 1B). When limited to amyloid‐positive individuals only (n=28), the observed associations were considerably stronger between scene encoding activation and both amyloid CL level and tau deposition (Figure 1C‐E). Similar results were observed when limiting to the CN and SCD groups.ConclusionAmyloid and tau deposition are associated with reduced brain activity during episodic memory encoding in older adults at risk for AD. These findings suggest that brain function is another important biomarker to consider in both the temporal sequencing of biomarkers (Jack et al. 2013) and the A/T/N framework. Future studies in larger samples are warranted. Reference: Jack et al. (2013) Lancet Neurology.