Mediation analysis and causal mechanisms of an intervention to promote physical activity among COPD patients

Abstract

Listen

Translate article

Rationale: Previously, we have shown that a behavioral intervention (BI) to promote physical activity (PA) increased self-reported PA (SRPA) among COPD patients. The purpose of this investigation was to study potential mediators on the effect of the BI on SRPA. Methods: Data from a trial of usual care (UC) vs. BI were used to examine patient characteristics as potential mediators (mMRC dyspnea, self-efficacy [SE], depression, SF-12, Chronic Respiratory Questionnaire [CRQ], readiness to change PA) measured at 6 months. SRPA, dichotomized as active vs under/inactive, was measured at 6 and 12 months. Mediation analysis was used to estimate the proportion mediated on the effect of the BI on SRPA at 12 months adjusting for potential confounders. Results: Of 305 patients randomized, SRPA results were available for 84%. After the BI a higher proportion of the BI group were classified as active (BI=74% vs UC=58%) (p=0.03). Of the potential mediators, readiness to change PA, SE for managing dyspnea with physical exertion, depression, and SF-12 vitality were found to mediate the association of the BI on SRPA over 12 months (Figure). Conclusion: These results suggest that the mechanism of the BI for enhancing SRPA may partly be through selected patient characteristics. Targeting these factors may enhance the effectiveness of future interventions to promote PA among patients with COPD.