Prognostic factors for adolescent knee pain: an individual participant data meta-analysis of 1281 patients

Adolescent knee pain has a propensity for chronicity, impacting physical activity and health into adulthood. The aim of this study is to investigate prognostic factors in adolescents with knee pain using individual participant data (IPD) meta-analysis. Studies were identified through a systematic search and a collaborative group. We included IPD from prospective studies of adolescents (age 10-19 years) with nontraumatic knee pain (13 studies and 1516 adolescents with 1281 unique participants). Primary outcomes were pain intensity and function (Knee Injury and Osteoarthritis Outcome Score Sport/Rec subscale). Primary endpoint was 12 months. Risk of bias was appraised with Quality in Prognosis Studies tool. Harmonised IPD was analysed by multilevel modelling. Fifty-one percent reported knee pain after 12-months. Lower baseline pain frequency was associated with lower pain intensity at 12 months (less than weekly; 12 (95% confidence interval [CI] 7-17) and monthly; 15 (95% CI 9-22) points lower on a 100-point pain scale, compared with almost daily pain). Other factors most strongly associated with 1-year pain prognosis were lower quality of life (30, 95% CI 19-42 points per unit change in the EQ5D index score), female sex (8 points, 95% CI 4-12 higher compared with males), and bilateral pain (7, 95% CI 1-13 points higher pain). Similar factors were associated with function. Body mass index, pain sensitivity, and knee strength were not associated with prognosis of pain or function. Adolescent knee pain is associated with clinically relevant long-term pain and functional deficits. Self-reported characteristics may help identify those at risk of poor prognosis.

Automated summary

Adolescent knee pain tends to become chronic, affecting physical activity and health in adulthood. Lower baseline pain frequency, lower quality of life, female sex, and bilateral pain were the most strongly associated factors with 1-year pain prognosis. Self-reported characteristics may help identify individuals at risk of poor prognosis.