Modeling Causal Pathways Among Physical Activity, Pain, and Self-Efficacy in Older Adults with Chronic Pain: A Simulation-Based Analysis

Chronic pain in older adults is a complex condition often linked to reduced physical activity and dimin-ished self-efficacy. This simulation-based study modeled the causal pathways among these variables us-ing synthetic data representative of older adults with chronic pain. A longitudinal cohort of 200 subjects was simulated over 30 days, with daily measures of physical activity (steps), pain intensity (0–10 scale), and self-efficacy (0–10 scale). Advanced statistical techniques, including marginal structural modeling (MSM), mixed-effects regression, and mediation analysis, were employed to examine bidirectional and indirect effects. Results revealed that higher pain levels predicted subsequent reductions in physical ac-tivity (β = -350 steps/day, *p* < 0.001), while greater self-efficacy was associated with increased activity (β = +260 steps/day, *p* < 0.001) and lower pain. Mediation analysis indicated that 27% of the beneficial effect of physical activity on pain was mediated by self-efficacy. Sensitivity analyses confirmed robust-ness to missing data. These findings align with real-world studies, underscoring the role of self-efficacy as both a predictor of activity and a mediator of pain reduction. The study demonstrates the utility of syn-thetic data for causal inference in chronic pain research, offering a replicable framework for future inves-tigations. Methodological innovations, such as MSM and dynamic mediation modeling, provide insights for clinical interventions targeting pain, activity, and psychosocial factors in aging populations.