Identifying cognitive subgroups in older adults from community data with hierarchical cluster analysis

Area of Science:

• Neuroscience
• Gerontology
• Psychology

Background:

• Conventional mild cognitive impairment (MCI) criteria may not fully capture cognitive heterogeneity in aging populations.
• Understanding cognitive subgroups is crucial for characterizing aging-related cognitive changes.
• Existing diagnostic categories may oversimplify the spectrum of cognitive performance in older adults.

Purpose of the Study:

• To identify distinct cognitive subgroups in a community-recruited sample of older adults using unsupervised clustering.
• To characterize multidomain cognitive heterogeneity beyond standard MCI classifications.
• To validate the identified cognitive subgroups using a cross-validation approach.

Main Methods:

• Ward's hierarchical clustering was applied to neuropsychological data from 180 older adults.
• Twenty-two demographically adjusted z-scores across memory, language, executive function, and attention/processing speed were analyzed.
• A k-means analysis was used for cross-validation of the four-cluster solution.

Main Results:

• Four reproducible cognitive subgroups were identified: Average Balanced Profile (ABP), Average with Relative Non-Memory Weakness (A-nonMEM↓), Average with Memory-Specific Weakness (A-MEM↓), and Global Multidomain Weakness (G-MD↓).
• The dissociation between A-nonMEM↓ and A-MEM↓ mirrored amnestic versus non-amnestic MCI distinctions.
• Cluster membership explained more variance in cognitive performance than the Montreal Cognitive Assessment (MoCA).

Conclusions:

• Unsupervised, data-driven clustering can identify reproducible cognitive phenotypes in aging populations.
• Hierarchical clustering offers a valuable framework for characterizing cognitive heterogeneity beyond predefined diagnostic labels.
• These findings highlight the utility of clustering for understanding the nuanced spectrum of cognitive aging.