Physical activity personas of older adults based on the capability, opportunity, motivation-behavior model: a...

To construct physical activity (PA) personas of older adults using a data-driven health persona approach based on the Capability, Opportunity, Motivation-Behavior (COM-B) model.

Older adults were recruited from two community health centers. PA was measured using the Physical Activity Scale for the Elderly. Variables potentially associated with PA were investigated based on the COM-B model, including capability, motivation, and opportunity factors. These variables were first screened through univariate analysis and subsequently underwent dimensionality reduction using exploratory factor analysis. Then, k-means clustering was employed to construct PA personas by integrating PA levels with the key factors derived from factor analysis. PA personas were further visualized.

A total of 313 valid questionnaires were collected for data analysis. Nine variables were identified as correlated with PA through univariate analysis, and subsequently reduced to three underlying factors via exploratory factor analysis: the motivation-opportunity factor, the psychological capacity factor, and the physical capacity factor. Considering both PA levels and the three factors, clustering analysis revealed four distinct PA personas: the highly active cluster (25.6%; with the highest PA levels), the physically constrained inactive cluster (18.8%; with the poorest physical capability), the knowledge-skill deficient inactive cluster (32.6%; with the lowest psychological capability), and the motivation-opportunity deficient inactive cluster (23.0%; with the lowest motivation and opportunity).

This study identified four distinct PA personas with typical characteristics among older adults based on the COM-B model. These findings suggest that personalized interventions targeting specific barriers are warranted to promote PA in this population.