Pain intensity and symptoms of depression are correlated and individually associated with decreased physical function. We compared two explanatory mediation models; one with depression as mediator of the association of pain intensity with physical function and the other one with pain intensity as the mediator of the effect of depression on physical function. In a cross-sectional study, 102 patients with upper extremity musculoskeletal illness completed measures of pain intensity, PROMIS depression CAT, PROMIS physical function-upper extremity CAT and demographics. We determined that pain intensity and symptoms of depression were partial mediators of their respective and independent effects on physical function. While depression had a larger standardized mediation effect and a bigger kappa-squared (κ 2) effect size compared to pain intensity, the actual proportion of variance in physical function that could be explained by the mediated effects (i.e., R 2 mediation effect size) was equal in both models suggesting that the two mediation models have equal ability to explain variations in physical function. The bidirectional mediation effect suggests a reinforcement mechanism, in which, pain intensity and symptoms of depression reciprocally influence their negative impacts on physical function among patients with musculoskeletal illness.
Keywords: Bidirectional mediation; Depression; Musculoskeletal; Pain; Physical function; Upper extremity.