Advertisement

AMYLOID ACCUMULATION WITHIN THE NEGATIVE RANGE IS LINKED TO MEMORY DECLINE IN COGNITIVELY NORMAL OLDER INDIVIDUALS

      Background

      The earliest phase of amyloid accumulation, before deposition is widespread throughout cortex, is of great interest for understanding the etiology of Alzheimer’s disease and for developing amyloid-modifying treatments. A critical question that has emerged is whether amyloid accumulation can be detected prior to amyloid positivity, and whether this accumulation is accompanied by any subtle changes in cognition.

      Methods

      We identified 160 normal older individuals from the Alzheimer’s Disease Neuroimaging Initiative who were negative on 18F-florbetapir PET based on global uptake in cortical summary regions. We contrasted the baseline florbetapir scans of subjects who subsequently increased or did not increase over 2 or 3 scans (2-4 years) of longitudinal florbetapir imaging in order to define a set of early accumulation regions at baseline. Then we determined whether amyloid accumulation within these early accumulation regions was related to concurrent memory change.

      Results

      Early accumulation regions (regions elevated at baseline for subjects who subsequently increased) were located in bilateral precuneus, posterior cingulate, and superior parietal lobes. Within these regions, 103/160 (64%) of baseline-negative normals had positive florbetapir slopes. 17 subjects (4 APOE4+) converted to florbetapir positive status over 3.4+-1.0 yrs (3.1% per year). Negative normals who increased on florbetapir did not differ from nonincreasing subjects with respect to age, sex, education, ApoE4 status, baseline temporoparietal glucose metabolism, or hippocampal volume but they did have slightly higher baseline cortical florbetapir, indicating that accumulation was ongoing. Increasing florbetapir was associated with memory decline regardless of whether florbetapir change was assessed as a dichotomous (p=0.007) or continuous (p=0.03) variable. Finally, the effect was stronger for subjects with longer florbetapir followup (3 scans over 4.1+/-0.3yrs compared with 2 scans over 2.4+/-0.8yrs).

      Conclusions

      Despite considerable variability in longitudinal amyloid and memory measurements in amyloid negative healthy older individuals, those with increasing amyloid had poorer longitudinal memory performance than the nonincreasing group. These effects, though subtle, may represent the earliest consequences of amyloid deposition that are detectable prior to the onset of amyloid positivity and any clinically-relevant cognitive dysfunction.
      Figure thumbnail fx1