Eccentric exercise improves joint flexibility in adults: a systematic review update and meta-analysis

Key points

1. 01Pooling 27 trials, eccentric exercise improved joint flexibility by a moderate standardised mean difference (Hedges' g = 0.54, 95% CI 0.34 to 0.74).
2. 02The effect was similar whether flexibility was measured as joint range of motion (g = 0.52) or muscle fascicle length (g = 0.57).
3. 03Lower limb flexibility improved clearly (g = 0.57), but the upper limb effect was inconclusive (g = 0.37, 95% CI 0 to 0.74), with only three upper-limb trials.
4. 04Eccentric exercise helped whether compared against no intervention (g = 0.48) or against another intervention (g = 0.66).
5. 05Trial quality was moderate on average (mean PEDro 5.2 out of 10) and heterogeneity was borderline (I2 = 50%), so overall GRADE confidence ranged from low to high depending on the outcome.

How it was conducted

Design

Systematic review and random effects meta-analysis (PRISMA; PROSPERO CRD42020151303), updating and extending a 2012 review

Data sources

AMED, CINAHL, MEDLINE, EMBASE, SportDiscus, searched from inception to Nov 29, 2021

Included studies

Randomised controlled trials of eccentric exercise (>=4 weeks) in adults aged over 18, versus no intervention or a different intervention

Participants

32 trials, 1122 participants (108 lost to follow-up), mean age 31.6 (SD 14.0) years; 27 trials (911 participants) meta-analysed

Primary outcome

Joint flexibility, measured as passive or active joint range of motion or muscle fascicle length, at first post-intervention follow-up

Quality assessment

PEDro scale for risk of bias and methodological quality; GRADE for overall certainty of evidence

What they found

• Eccentric exercise improved joint flexibility overall (pooled random effects Hedges' g SMD = 0.54, 95% CI 0.34 to 0.74) across 27 trials and 35 comparisons.
• Heterogeneity was borderline (I2 = 50%; Q = 67.6, d.f. = 34, p = 0.001; T2 = 0.17), with a wide 95% prediction interval of -0.33 to 1.42.
• Effect was similar for joint range of motion (g = 0.52, 95% CI 0.31 to 0.74, I2 = 0%) and muscle fascicle length (g = 0.57, 95% CI 0.28 to 0.87, I2 = 60%); meta-regression difference g = 0.07, 95% CI -0.42 to 0.55.
• Lower limb flexibility improved (g = 0.57, 95% CI 0.34 to 0.79) but upper limb did not clearly (g = 0.37, 95% CI 0 to 0.74); subgroup difference g = -0.21, 95% CI -0.93 to 0.51.
• Effect held against no intervention (g = 0.48, 95% CI 0.28 to 0.69) and against another intervention (g = 0.66, 95% CI 0.28 to 1.03); subgroup difference g = 0.17, 95% CI -0.29 to 0.63.
• No conclusive small-study bias (Egger's test = 0.39, 95% CI -1.57 to 2.35, p = 0.69); excluding the largest-effect trial barely changed the result (g = 0.49, 95% CI 0.34 to 0.64).

Limitations

• Average trial quality was only moderate (mean PEDro 5.2 out of 10, range 3 to 8), and more than half of trials scored below 6, contributing to risk-of-bias downgrades.
• Between-trial heterogeneity was borderline substantial (I2 = 50%) and the wide 95% prediction interval (-0.33 to 1.42) means the true effect in any single setting could be positive, zero, or negative.
• Only three trials examined the upper limb, leaving the upper-limb effect inconclusive.
• Trials varied widely in exercise type, load, frequency and comparators, and effects in people with a pathological loss of range of motion remain largely untested.

Why it matters

For patients

If you want to improve joint flexibility, structured eccentric exercise over several weeks can give a moderate, reliable improvement, especially in the legs.

For clinicians

Eccentric exercise is a reasonable option to increase and maintain joint flexibility in adults, with confident evidence for range of motion but only suggestive evidence in clinical contracture and the upper limb.

For readers

This update of an earlier review now quantifies a moderate, precisely estimated flexibility benefit from eccentric exercise, though certainty varies by outcome from low to high.