Social cognitive or learning theory use to improve self-efficacy in musculoskeletal rehabilitation: a systematic review and meta-analysis

Key points

1. 01Meta-analysis of 25 studies (20 contributing to effect size) found a large overall mean effect size (Cohen's d = 0.98, 95% CI 0.42-1.86) favouring self-efficacy-focused interventions
2. 0210 of 20 effect-size studies showed large effects (Cohen's d >= 0.8), 5 moderate (0.5-0.8), and 5 small (<0.5)
3. 03Most participants were older women (78.9% female, mean age 64.8 years) with hip or knee osteoarthritis or post-arthroplasty
4. 04No study reported participant attrition or adverse reactions directly caused by self-efficacy-focused behavioural change techniques
5. 05Authors propose early application of these techniques may also benefit younger, more athletically active individuals before conditions reach end-stage severity

How it was conducted

Design

Systematic review and meta-analysis following PRISMA 2009 guidelines

Databases

OVID and SPORTDiscus; no date restrictions applied

Search terms

'Social cognitive theory' OR 'social learning theory' combined with 'rehabilitation'

Included studies

25 peer-reviewed English-language studies involving participants with orthopaedic or musculoskeletal conditions

Quality assessment

McMaster Quantitative or Qualitative Critical Review Forms; inter-rater agreement Cohen's kappa 0.72-0.79 (p < 0.0001)

Participants

1947 total; 78.9% women; mean age 64.8 +/- 17 years; predominantly hip or knee OA or post-arthroplasty

What they found

• Overall mean effect size: Cohen's d = 0.98 (95% CI 0.42-1.86), classified as large
• Of 20 studies contributing to effect size: 10 large (>= 0.8), 5 moderate (0.5 to <0.8), 5 small (<0.5)
• Two studies showed unfavourable outcomes: one with effect size 0.26 (95% CI -0.3 to 0.84) using a potentially biased survey instrument (AESOP), and one with effect size 0.2 (95% CI -0.67 to 1.09) in an ACL injury cohort with possible contamination and small final sample (n = 12 per group)
• 23 of 25 studies provided patient outcome data; very few reported injury or adverse reactions leading to attrition
• 44% of studies (11/25) involved participants with hip or knee OA or post-hip or knee arthroplasty; 24% (6/25) involved post-femur or tibia fracture
• Study origins: 14 USA, 3 UK, 2 Canada, 2 Thailand, 2 Sweden, 1 Netherlands, 1 Iran

Limitations

• Review restricted to studies explicitly mentioning social cognitive or social learning theory; studies applying equivalent principles under different terminology were excluded
• Most participants were older women with severe OA, limiting generalisability to younger, more active, or male populations
• McMaster Critical Review Form criteria may not provide the rigor needed to conclusively evaluate study methodologies
• Heterogeneity in self-efficacy measurement tools and outcome constructs across studies complicates direct comparison

Why it matters

For patients

Patients recovering from hip or knee problems, fractures, or sports injuries may benefit from therapy programmes that actively build their confidence in performing movement tasks, not just physical exercises alone.

For clinicians

Incorporating social cognitive or learning theory principles into musculoskeletal rehabilitation is supported by consistent moderate-to-large effect sizes and a strong safety profile, making it a practical low-cost addition to standard programmes.

For readers

This first meta-analysis on the topic demonstrates that self-efficacy-enhancing behavioural strategies are effective across diverse orthopaedic conditions, though most evidence comes from older women and better-quality trials with younger athletic populations are still needed.