UCSF

BACKGROUND: Oldest-old individuals exhibit heterogeneous cognitive function ranging from superior cognitive performance to dementia. Most SuperAger definitions rely on cross-sectional memory scores. Here, we extend the definition by identifying distinct subgroups based on longitudinal trajectories for memory. We also evaluated longitudinal trajectories for executive function. METHOD: The 774 study participants were from the LifeAfter90 study (LA90), a multi-ethnic cohort of individuals aged 90 or older. Verbal episodic memory (VEM) and executive function (EF) were assessed approximately every 6 months using the Spanish and English Neuropsychological Assessment Scale from at least two visits (mean = 4; range: 2-7). Using group-based trajectory modelling, we identified unique trajectories separately for VEM and EF, accounting for baseline age, sex, education, and race/ethnicity. RESULT: 62% of participants were women, the overall mean baseline age was 92.4 (±2.3), and 26% had ≤high school education (Table 1). We identified four distinct trajectories for VEM (Figure 1). Group 4-High stable performers: with the highest baseline performance, largest practice effects, and no decline; Group 3-Stable performers: with a lower baseline than group 4, moderate practice effects, and slight decline; Group 2-Slow decliners: with lower baseline than groups 4 and 3, slight practice effects, and a slow declining trajectory; and Group 1-Accelerated decliners: with the lowest baseline performance, no practice effects, and rapid decline. Thus, we identified one subgroup with no decline in longitudinal performance and with practice effects that could be classified as SuperAgers. For EF, we also identified four groups (Figure 2). Baseline EF scores varied for Groups 4, 2, and 1 (Group 4 being the highest) and followed relatively stable trajectories. Group 3, which had the second highest baseline score, had a slowly declining trajectory. For EF, stable trajectories were not limited to the top baseline performing groups. The two domains had distinct patterns with limited overlap in group memberships. CONCLUSION: Using longitudinal patterns, we identified SuperAgers based on superior memory performance and no decline. In future research, we will further explore the use of longitudinal performance trajectories and the use of multiple cognitive domains in the definition of SuperAgers. We will also identify defining characteristics of each subgroup.