Effects of eccentric exercise in patients with subacromial impingement syndrome: a systematic review and meta-analysis

Key points

1. 01Post-treatment pain was statistically lower with eccentric exercise (MD -12.3 mm on a 0-100 VAS) but fell below the 15 mm minimal important difference threshold
2. 02No significant improvement in function was found at any time point (SMD -0.10, 95% CI -0.79 to 0.58, p = 0.76)
3. 03Painful eccentric exercise showed no advantage over pain-free eccentric exercise
4. 04Six to eight weeks of eccentric exercise produced similar pain reduction to 12-week programmes
5. 05GRADE certainty was low for pain outcomes and very low for function outcomes

How it was conducted

Design

Systematic review and meta-analysis of RCTs (PROSPERO CRD42019126917)

Databases searched

PubMed, Cochrane Library, PEDro; searched March 2019

Included studies

7 studies reported in 8 articles, published 2012-2018; 6 studies included in meta-analyses

Participants

303 total (281 in meta-analyses); adults with subacromial impingement syndrome, mean age 39-60 years

Primary outcomes

Pain (VAS or NPRS) and function (DASH, Constant-Murley, WORC, SPADI, Oxford Shoulder Score) post-treatment and at 6-12 months

Quality assessment

PEDro scale (median score 7/10, range 5-8); GRADE applied to certainty of evidence

What they found

• Post-treatment pain (6 studies, n=281): eccentric exercise reduced pain by MD -12.3 mm (95% CI -17.8 to -6.8, I2 = 7%, p < 0.001) compared with other exercise, below the 15 mm minimal important difference
• 6-8 week subgroup pain (3 studies): MD -13.5 mm (95% CI -28.5 to 1.4, I2 = 55%, p = 0.08), not significant
• 12-week subgroup pain (3 studies): MD -11.9 mm (95% CI -18.2 to -5.5, I2 = 0%, p < 0.001), significant but not clinically important
• Intermediate to long-term pain (3 studies, n=167, 6-12 months): MD -4.9 mm (95% CI -15.4 to 5.6, I2 = 50%, p = 0.36), not significant
• Post-treatment function (6 studies, n=281): SMD -0.10 (95% CI -0.79 to 0.58, I2 = 85%, p = 0.76), not significant
• Intermediate to long-term function (3 studies, n=167): SMD 0.28 (95% CI -0.67 to 1.24, I2 = 87%, p = 0.56), not significant
• Painful vs pain-free eccentric exercise (1 study, n=22, 4 weeks): no significant difference in pain or function

Limitations

• All studies lacked blinding of participants and therapists, and most did not blind outcome assessors, introducing high risk of bias
• High heterogeneity in exercise protocols and outcome measures (I2 up to 87% for function) limits pooling and interpretation
• Most included trials were small (range n=11-97), reducing statistical power and precision of estimates
• Only 3 of 7 studies confirmed that pain originated from rotator cuff tendon pathology, making it unclear whether results apply specifically to tendinopathy

Why it matters

For patients

Eccentric shoulder exercises are unlikely to provide meaningfully greater pain relief than standard resistance exercise, so patients should not feel obliged to seek out a specific eccentric programme if other exercise is available.

For clinicians

Either eccentric or standard resistance exercise can be prescribed for subacromial impingement syndrome; higher training intensity may matter more than exercise type, and 6-8 weeks appears sufficient for initial pain management.

For readers

This review highlights the need for better-designed trials using verified tendinopathy inclusion criteria and clearly defined high-intensity eccentric loading protocols before firm recommendations can be made.