Systematic review and meta-analysis finds pain and function improve over time in patellofemoral pain, but recovery is variable

Objective: To investigate the long-term (defined as >12 months) prognosis of knee pain and knee function in adults and adolescents with patellofemoral pain (PFP). Design: Systematic review with meta-analysis and meta-regressions. Data sources: MEDLINE, OVID, CENTRAL, Web of Science, OpenGrey, and International Patellofemoral Research Retreat abstract books. Eligibility criteria for selecting studies: Prospective studies of patients clinically diagnosed with PFP, aged <40 years, with a long-term follow-up (minimum of 12 months). Primary outcomes were self-reported pain intensity (worst, during activity, and usual) and function. Meta-analyses and meta-regressions were performed where appropriate. Narrative synthesis was performed for those not included in the metanalysis. Risk of bias was assessed using the Quality In Prognosis Studies (QUIPS) tool, and certainty of evidence using GRADE. Results: A total of 42 studies (n = 3,230) were included. At 12 months, meta-analysis indicated reduction in worst pain (SMD 1.36; 95% CI 0.85-1.86), pain during activity (SMD 1.36; 95% CI 0.61-2.11), and resting pain (SMD 0.91; 95% CI: 0.75- 1.08). No significant reduction was found for usual pain. We found improvement in self-reported function (investigated using the Anterior Knee Pain Scale (AKPS) MD 14.60; 95% CI 11.60-17.61), FIQ (MD 3.33; 95% CI: 2.46-4.20) and the Western Ontario McMaster Universities Osteoarthritis Index (WOMAC) (MD -7.73; 95% CI: -10.36 to -5.10). Extended follow-up (more than 60 months) suggested more variable improvements. Meta-regression showed no association between age and 12-month function, while older age was modestly associated with greater improvement in activity-related pain at extended follow-up. Overall, a considerable proportion of participants continued to report persistent symptoms, and heterogeneity across studies was substantial. Certainty of evidence ranged from very low to moderate across outcomes investigated. Conclusion: Pain and self-reported function generally improve over time, particularly within the first 12 months. However, substantial heterogeneity and persistent symptoms in a considerable proportion of patients at extended follow-up indicate that recovery is not universal and trajectories are highly variable.