Late Life Cognitive Function Trajectory Among the Chinese Oldest-Old Population-A Machine Learning Approach

Area of Science:

• Gerontology
• Cognitive Science
• Biopsychosocial Health

Background:

• Understanding cognitive function trajectories in the oldest-old (80+) is crucial for healthy aging.
• The biopsychosocial framework provides a comprehensive lens for examining aging processes.

Purpose of the Study:

• To identify distinct cognitive function trajectories among the oldest-old using the CLHLS dataset.
• To explore factors influencing these cognitive aging patterns.

Main Methods:

• Utilized the Chinese Longitudinal Healthy Longevity Survey (CLHLS) dataset.
• Employed K-means clustering to identify latent cognitive groups.
• Applied Lasso regression to determine predictive factors.

Main Results:

• Identified two groups: High Stability (HS) with maintained cognitive function and Low Stability (LS) with consistently low function.
• Socioeconomic status, biological conditions, mental health, lifestyle, and psychological/behavioral factors predicted group membership.

Conclusions:

• Cognitive aging among the oldest-old is not uniform, with distinct stability patterns observed.
• A multifactorial approach incorporating biopsychosocial elements is essential for understanding and supporting cognitive health in aging.
• Pioneered non-parametric machine learning for analyzing cognitive trajectories in longevity research.