Using Mixture Modeling to Construct Subgroups of Cognitive Aging in the Wisconsin Longitudinal Study

Area of Science:

• Gerontology
• Cognitive Neuroscience
• Biostatistics

Background:

• Longitudinal studies on aging increasingly include cognitive assessments.
• Few studies utilize mixture modeling to identify distinct cognitive aging trajectories.
• Understanding interindividual variability in cognitive aging is crucial.

Purpose of the Study:

• To apply mixture modeling to identify subgroups of cognitive aging patterns.
• To examine intraindividual change in cognitive performance over time.
• To advance empirical evidence on cognitive aging variability.

Main Methods:

• Latent class analysis and latent transition analysis were used.
• Data from 3,713 participants in the Wisconsin Longitudinal Study (WLS) were analyzed.
• Cognitive performance in verbal fluency and episodic memory was assessed at ages 65 and 72.

Main Results:

• Four distinct cognitive aging subgroups were identified at each time point.
• Approximately 75% of participants showed continuity in their cognitive profile.
• 18.7% of participants exhibited a marked decline in episodic memory over the 7-year period.

Conclusions:

• Mixture modeling effectively identifies distinct subgroups of cognitive aging.
• Results highlight the utility of population health data for cognitive aging research.
• Findings have implications for understanding and intervening in cognitive aging processes.