Effect of adherence to exercise-based injury prevention programmes on the risk of sports injuries: a systematic review and meta-analysis of RCTs

Key points

1. 01Top adherence (>=84% of sessions) cut injury rate by 47% vs. controls (RR 0.53, p<0.001)
2. 02Middle adherence (74-83%) cut injury rate by 50% vs. controls (RR 0.50, p<0.001)
3. 03Bottom adherence (<74%) showed only a non-significant 13% reduction (RR 0.87, p=0.08)
4. 04Meta-regression confirmed each 1% increase in adherence independently lowered injury rate (beta=-0.013, p<0.001)
5. 05Findings were consistent across adolescents, adults, males, and females

How it was conducted

Design

Systematic review and meta-analysis of randomised and cluster-randomised controlled trials

Databases searched

PubMed, EMBASE, SPORTDiscus, PEDro, Cochrane CENTRAL, Web of Science (inception to May 2023)

Studies included

52 in systematic review; 35 in meta-analysis

Participants

Data from 35 RCTs covering 2,288,113 exposure hours and 6,579 injuries

Adherence grouping

Bottom <74%, middle 74-83%, top >=84% of planned sessions completed

Primary outcome

Sports injury rate ratio (intervention vs. control group)

What they found

• Top-adherence group: RR=0.53 (95% CI 0.42 to 0.67, p<0.001) - 47% lower injury rate than control
• Middle-adherence group: RR=0.50 (95% CI 0.39 to 0.65, p<0.001) - 50% lower injury rate than control
• Bottom-adherence group: RR=0.87 (95% CI 0.75 to 1.02, p=0.08) - non-significant 13% reduction
• Meta-regression: beta=-0.013 (p<0.001), each percentage point increase in adherence reduces RR by 0.013 units
• Subgroup differences: top vs. bottom (p<0.001) and middle vs. bottom (p<0.001) were significant; top vs. middle was not (p=0.76)
• Adolescents - top: RR=0.48 (95% CI 0.38 to 0.62, p<0.001); middle: RR=0.63 (95% CI 0.52 to 0.77, p<0.001); bottom: RR=0.65 (95% CI 0.40 to 1.06, p=0.09)
• Adults - top: RR=0.62 (95% CI 0.43 to 0.89, p=0.01); middle: RR=0.43 (95% CI 0.31 to 0.58, p<0.001); bottom: RR=0.95 (95% CI 0.83 to 1.09, p=0.49)
• Female participants - top: RR=0.38 (95% CI 0.18 to 0.81, p=0.01); middle: RR=0.45 (95% CI 0.31 to 0.63, p<0.001); bottom: RR=0.99 (95% CI 0.81 to 1.21, p=0.95)
• Sensitivity analysis removing high-risk-of-bias studies: top RR=0.49 (95% CI 0.40 to 0.61), middle RR=0.44 (95% CI 0.30 to 0.64), bottom RR=0.94 (95% CI 0.84 to 1.05) - conclusions unchanged
• Statistical heterogeneity was high: I2=68% (top), 90% (middle), 64% (bottom), 90% overall

Limitations

• High statistical heterogeneity across all adherence groups limits confidence in pooled estimates
• Large variation in how adherence was defined and measured across studies complicated pooling; 20 studies had to be excluded from meta-analysis for this reason
• Many studies used cluster-randomisation, which may reduce statistical efficiency compared with individual randomisation
• Female top-adherence subgroup comprised only one study, preventing definitive conclusions for female athletes

Why it matters

For patients

If you are given an injury prevention exercise programme, completing at least three-quarters of the planned sessions appears necessary to actually lower your injury risk.

For clinicians

Prescribing an exercise-based injury prevention programme is not enough; monitoring and supporting athlete adherence above the 74% threshold is essential to achieve the protective effect shown in trials.

For readers

This meta-analysis of 35 RCTs is the first to pool adherence effects across all injury types, and it confirms a clear dose-response relationship between session completion rate and injury rate reduction.