Measurement Properties of Hand-Held Dynamometry for Ankle Strength Assessment: A COSMIN-Based Systematic Review and Meta-Analysis.
Matheus Hissa Lourenço Ferreira, João Victor Völker Oliveira, Nelson Carvas Junior et al.
✦ AI-curated · Sources linked
In 30 seconds
This COSMIN-based systematic review and meta-analysis assessed handheld dynamometry (HHD) for measuring isometric ankle muscle strength in healthy individuals. It included 24 observational and methodological studies reporting reliability, measurement error, or construct validity against isokinetic dynamometry. Inter-rater reliability was good across all four muscle groups, with ICCs from 0.75 for evertors to 0.89 for dorsiflexors; intra-rater reliability ranged from good to excellent (ICC = 0.88-0.91). Construct validity was strong for dorsiflexors but uncertain for plantar flexors, while measurement error could not be determined.
Key findings
- Inter-rater reliability was good for dorsiflexors (ICC = 0.89, 95% CI 0.80-0.94) and plantar flexors (ICC = 0.86, 95% CI 0.72-0.93).
- Inter-rater reliability was ICC = 0.79, 95% CI 0.66-0.87 for invertors and ICC = 0.75, 95% CI 0.46-0.90 for evertors.
- Intra-rater reliability ranged from good to excellent (ICC = 0.88-0.91); external fixation showed numerically higher inter-rater ICCs for some groups, without statistically significant subgroup differences.
- Construct validity was strong for dorsiflexors (r = 0.87, 95% CI 0.42-0.98) but uncertain for plantar flexors (r = 0.72, 95% CI -0.20-0.96); measurement error remained indeterminate.
Why it matters
Reliable strength measurement supports consistent assessment of ankle muscle function. These findings suggest HHD can provide reproducible isometric measurements across ankle muscle groups, including without external fixation. However, reliability does not establish that measurements accurately reflect strength in all groups or that observed changes are meaningful, so validity and measurement error remain important considerations.
What to keep in mind
Evidence for construct validity was low-certainty, and measurement error could not be established because a minimal important change was not established. The review concerned healthy individuals, limiting direct inference to other populations.