Exploring the pain and disability continuum in adolescents with non-traumatic anterior knee pain: a mediation analysis using individual participant data of prospective studies.

Abstract

To use individual patient data (IPD) to investigate if the effect of pain on sports-related disability is mediated through physical (lower extremity isometric strength) or psychological (depression/anxiety and knee confidence) factors in adolescents with non-traumatic anterior knee pain. This study included four datasets from a previously harmonised IPD dataset. Prior to analysis, the protocol and analysis approach were predefined and published on Open Science Framework. Potential mediators were pre-sepcified as isometric knee and hip strengths, self-reported anxiety/depression and confidence in the knee, allmeasured at 12 weeks after baseline evaluation. Mediation analyses were undertaken using the CMAVerse package in RStudio using the regression-based approach to decompose the total effect of the exposure (pain at baseline evaluation) on the outcome (sports-related disability at 6 months) into the 'indirect effect' (the portion of the total effect acting through the mediators) and the 'direct effect'. Two-hundred and seventy-nine adolescents with non-traumatic knee pain were included in the analysis. Median age was 13 (range 10-19), and 72% were women. Baseline pain was associated with sports-related disability at 6 months. There was no evidence of the association being mediated by any of the proposed mediators (total natural indirect effect for strength 0.01 (-1.14 to 1.80) and psychological factors 0.00 (-0.66 to 2.02)). We found an effect of pain on sports-related disability at 6 months which appears to be independent of lower extremity muscle strength, or depression/anxiety and knee confidence in adolescents with non-traumatic anterior knee pain.