P3-160: empirically defining trajectories of late-life cognitive and functional decline

The objective was to define trajectories of cognitive and functional decline, and characteristics associated with distinct trajectories, using Growth Mixture Modeling (GMM). We studies healthy controls (HC, n=325), and persons with early mild cognitive impairments (EMCI, n=279), late MCI (LMCI, n=372), or Alzheimer's disease (AD) dementia (n=216) from the Alzheimer's Disease Neuroimaging Initiative (ADNI-1, GO and 2). We focused on participants with available baseline beta-amyloid measurements who were assessed on the Alzheimer's Disease Assessment Scale-cognitive (ADAS-Cog 14) and Functional Activities Questionnaire (FAQ) for up to 24 months. Cerebrospinal fluid (CSF) Aβ 1-42, or composite SUVR for [11 C]PiB or [18 F]florbetapir PET, determined amyloid status and the time of baseline for our analyses. GMM was conducted on ADAS-Cog and FAQ. Subsequently, baseline variables for each class were compared including diagnosis, demograpics, co-morbidities, clinical and cognitive measures, and APOE allele status. We identified 3 trajectories of cognitive and functional change (Table): Class 1 (n=162, 13.6%) “moderately progressive” with highest baseline scores on FAQ and ADAS-Cog and a similar rate of cognitive and functional worsening; Class 2 (n=211, 17.7%) “mildly progressive” with intermediate baseline scores on FAQ and ADAS-Cog, worsening on both, but faster functional decline; Class 3 (n=819, 68.7%) “nonprogressive” lowest baseline scores on FAQ and ADAS-Cog with minimal change over time. Class 3 participants were less likely to be amyloid positive or carry an APOE 4 allele and more likely to be an HC or have EMCI at baseline. Cognitive and functional change appears to follow 3 distinct patterns regardless of baseline diagnosis. Amyloid and APOE4 carrier status predict faster decline.