The examination of physical function and cognitive outcomes in middle-to-older high-risk adults: an unsupervised clustering method

Alzheimer's disease rates are expected to triple by 2050. Early detection and specific mitigation efforts are warranted to blunt the alarming increase. Physical function index (PFI) declines with age; additionally, higher PFI is associated with better cognitive functioning in middle-to-older age individuals. However, most studies utilize one domain of PFI to examine associations with cognition. Therefore, using clustering methods, the purpose of this investigation was to determine if high-risk individuals with higher PFI have better cognitive outcomes compared to individuals with lower PFI. Participants (<i>n</i> = 215; 73.1% female; 45-75 years) completed a body mass scan, venous blood draw, 7 PFI tasks, and 7 cognitive tests. A <i>k</i>-means cluster analysis was utilized to identify PFI cluster for participants, one-way ANCOVAs were used to assess differences in cognition among clusters. Cluster 1 (C1; <i>n</i> = 29) was characterized as the highest strength/power, faster dual-task walking time, and higher aerobic capacity, Cluster 3 (C3; <i>n</i> = 113) had the lowest values between PFI groups, Cluster 2 (C3; <i>n</i> = 74) was in-between C1 and C3. Individuals in C1 had significantly higher global cognitive, visuospatial scores, digital executive functioning and associative learning compared to individuals in C3 (<i>p</i> < 0.05). Individuals in C1 and C2 had significantly higher values on orientation task and figure recall than individuals in C3 (<i>p</i> < 0.05). The results from this current study demonstrate that individuals with higher combined PFI output have higher global cognitive scores than individuals with lower combined PFI output. Examining PFI variables together may be a valuable tool when assessing cognition among cognitively at-risk individuals.