This umbrella review evaluated the reporting of exercise prescription variables, assessed the methodological quality, and provided exercise prescription recommendations.
This prospectively registered umbrella review of systematic reviews (CRD42021287440) included studies on knee pain. English-language, peer-reviewed journal articles were searched from inception to February 14, 2023 using PubMed, SPORTDiscus, Cochrane Database of Systematic Reviews, EMBASE, and CINAHL. Systematic reviews and meta-analyses of randomized controlled trials that compared exercise training to non-pharmacological, non-surgical conservative treatments, placebo, usual care, true control or other active exercise interventions were included. Outcomes were physical function, pain intensity, and mental health in adults (>18 years) with knee pain. We categorized the extracted data into two groups: reviews that reported all exercise prescription data and reviews that quantitatively analyzed the relationship between exercise prescription variables and outcomes. We assessed methodological quality using AMSTAR-2 and evaluated the overlap of primary studies using overlap analysis.
Of 6757 records, 65 systematic reviews addressing knee pain conditions were included. Only fifteen (23%) reviews reported all exercise prescription variables, and eight (12%) quantitatively analyzed the relationship between different dosages and outcome(s). The most commonly reported exercise prescription variables were duration (N=63) and frequency (N=57). Among the 65 reviews, all assessed pain intensity, 55 evaluated physical function, but only six considered mental health. Subgroup analysis for intervention duration of exercise training showed that Tai Chi for 10-17 weeks (SMD [95%CI]: -1.05 [-1.31, -0.79]) and therapeutic land-based exercises for 8-11weeks (-1.27 [-1.74, -0.81]) had the largest effect on physical function and pain intensity respectively. The overall methodological quality was rated as 'critically low' in 91% of reviews due to insufficient information on critical domains. Overlap analysis indicated a high degree of overlap among the included primary studies, with a corrected covered area of 12%.
There is a lack of high methodological quality evidence for optimal exercise training prescription variables for individuals with knee pain. Future research could address key gaps by (a) performing high-quality reviews per AMSTAR-2 criteria, (b) focusing on improved reporting using exercise reporting checklists (e.g. TIDieR and CERT), and performing more detailed analyses such as a dose-response network meta-analysis, to better understand the relationship between exercise prescription variables and outcomes which will lead to more precise exercise training recommendations.
Poor reporting of exercise prescription variables and high overlap in the available evidence limits the generalisability of findings and translation of results into clinical practice. Thus, there is a need to improve reporting of exercise prescription variables to optimize the management of knee pain conditions.
patellofemoral pain syndrome
musculoskeletal pain
