REF-1789 · Soliman et al., 2026 · Clinical Therapeutics, · Paper · PubMed

Isometric exercise for tendon pain has little support as a better option: a systematic review of 13 randomised trials (336 participants, six tendon sites, 3 Achilles and 4 patellar) found pain and function improved significantly in only 3 trials, rated 10 of the 13 trials poor quality, and could not pool them; holds looked tolerable, with short-term pain relief possible

Main caveat: Mostly poor-quality trials (10 of 13), small on average (336 participants across 13 trials, about 26 per trial by our arithmetic).
fairly strong evidenceMeasured on non-athletes and recreational peopleRead from the abstract only

Soliman and colleagues reviewed controlled evidence on isometric exercise for tendinopathy at any site, motivated by uncertainty about whether holds have a place for pain relief and recovery beside the established eccentric approach.

Of 304 records, 13 randomised trials met the criteria, with 336 participants (40% female), mostly at the patellar (4 trials) and Achilles (3) tendons, the rest at the lateral elbow, rotator cuff, wrist extensors and gluteal tendons. Symptoms had lasted a median of 23 months (IQR 3-51). Programmes ran from one 45-minute session to 4 months.

Adherence started near 100% and fell over time: 70% of the isometric groups and 58% of the isotonic control groups completed at least 80% of sessions. Pain and function improved significantly in only 3 trials (the abstract does not say whether this means against the comparator or within the group). Imaging was reported inconsistently, and the trials differed too much to pool.

By the Cochrane tool, 3 trials were of good and 10 of poor methodological quality. The authors conclude that the evidence gives limited support for isometrics being superior; they appear tolerable and may give short-term pain relief, while their long-term worth against eccentric or isotonic exercise is uncertain.

Participants

13 RCTs, 336 participants, 40% female.
Sites: patellar 4, Achilles 3, lateral elbow 2, rotator cuff 2, wrist extensor 1, gluteal 1.
Median symptom duration 23 months (IQR 3-51).
Age, sport, activity level and how tendinopathy was diagnosed: not in the abstract.

Protocol

Search: PubMed, MEDLINE, Embase and the Cochrane Library, inception to 1 August 2025; 304 records identified.
Eligibility: RCTs and prospective or retrospective cohorts of isometric exercise for tendinopathy, with or without a comparator; case series and case reports excluded. Only RCTs were included in the end.
Data: demographics, site, symptom and treatment duration, adherence; outcomes VISA (A, P or G), visual analogue pain, Patient-Specific Functional Scale, SF-36, EQ-5D, and ultrasound or MRI where available.
Appraisal: Cochrane Risk of Bias tool.
Synthesis: narrative; meta-analysis not done because of heterogeneity.
Not in the abstract: the isometric doses (angle, intensity, hold time), the comparators in each trial beyond "isotonic", effect sizes, follow-up times, and which 3 trials were positive.

Key Results

ItemResult
Trials with significant pain and function improvement3 of 13
Methodological quality (Cochrane tool)good 3, poor 10
Completed ≥80% of sessions70% isometric vs 58% isotonic control
Adherence at the startnearly 100%
Intervention lengthone 45-min session to 4-month programmes
Imaging outcomesinconsistently reported
Pooled effectnone (heterogeneity)

Practical Application

Isometric holds are a tolerable way to load a painful tendon, not a proven better treatment than other tendon exercise. Across 13 trials (336 patients, six tendon sites), only 3 showed significant gains in pain and function, 10 were of poor quality, and the authors found limited support for holds over other exercise.
This is not treatment advice; a physiotherapist or physician diagnoses the tendon problem and chooses the loading.
Agrees in direction with [SYNTH-88]: the guide calls holds a tolerable load, not a reliable painkiller, with long-term capacity from heavy slow resistance or eccentric loading; this review, with a later search than REF-1149, likewise finds limited support for holds over other exercise, while its authors leave room for short-term pain relief, which the base's trials (REF-1147, REF-1148) did not find. SYNTH-73 lists isometric calf holds as one form of the loading.Its finding that more of the isometric groups kept to their programme (70% vs 58% completing at least 80% of sessions) may mean holds are easier to stick with; the abstract gives no test of that difference.
Inference, not tested here: for an athlete whose tendon is too irritable for heavy dynamic work, holds can keep some loading going while a clinician plans the progression; whether they speed recovery is not shown.
Not settled by this design: the size of any effect, the best hold dose, and results for runners or cyclists specifically.

Limitations

Mostly poor-quality trials (10 of 13), small on average (336 participants across 13 trials, about 26 per trial by our arithmetic).
Six tendon sites with different loading demands pooled in one narrative synthesis; no meta-analysis, no effect sizes.
Interventions from a single session to 4 months, so short-term analgesia and longer-term rehabilitation are mixed together.
Adherence fell over time in both arms.
No certainty grading (such as GRADE) is mentioned in the abstract.
Written from the abstract only: doses, comparators, follow-up, which trials were positive, whether "significant" means between groups, registration and funding are unknown.
How this entry was checked and changed
Why this level

a systematic review of randomised trials whose abstract reports its risk-of-bias appraisal and the result (good in 3 trials, poor in 10), so the abstract-only cap does not apply; held at A- because that appraisal is serious (mostly poor quality) and the trials were too heterogeneous to pool