REF-1148 · O'Neill et al., 2019 · Knee Surgery, Sports Traumatology, Arthroscopy · Paper · PubMed · Free full text
Acute sensory and motor response to 45-s heavy isometric holds for the plantar flexors in Achilles tendinopathy
Five 45-s calf holds at 70% MVC gave no immediate pain relief (functional pain 4.2 → 4.9 of 10, not significant, in the 9 of 16 who had pain to test), no change in pressure pain, and the most painful got worse; uncontrolled, n=16
After Rio 2015 (REF-627) reported large immediate pain relief from heavy quadriceps holds in patellar tendinopathy, textbooks began recommending heavy isometrics as stage-1 rehabilitation for the Achilles too. O'Neill and colleagues tested the same dose on the calf.
Sixteen adults with unilateral mid-portion Achilles tendinopathy (symptoms over 3 months, VISA-A 61.3) did five 45-s isometric plantar-flexor contractions at 70% of their maximal voluntary contraction (MVC), seated with the knee bent to favour the soleus, with real-time force feedback.
Pain on a graded functional task did not fall: 4.2 before and 4.9 after on a 0-10 scale (not significant), and only 9 of the 16 had any pain to measure. People with the highest baseline pain got worse; those below about 4 improved by only about 1 point. Pressure pain sensitivity at the tendon did not change.
Concentric torque at 90°/s rose by 5.9 Nm, below the 12.7 Nm the authors take as the smallest real change; the other torque measures did not change. The authors conclude that heavy 45-s isometric holds cannot be recommended for immediate pain relief or improved motor output in Achilles tendinopathy.
Protocol
Participants: 16 of 19 examined (5 women), unilateral mid-portion Achilles tendinopathy for more than 3 months diagnosed clinically and on ultrasound tissue characterisation; age 48.6 ± 8.9 years, height 173.4 cm, weight 81.6 kg, VISA-A 61.3 ± 23.0, symptom duration 46.5 months (range 3.5-177.8). Excluded: plantaris-related or insertional pain, tears, VISA-A above 90, any current or previous treatment for the tendinopathy.
Intervention (overcoming isometric): five 45-s plantar-flexor contractions at 70% of MVC with 2 min rest, on a force platform (Fysiometer) with the ankle fixed by a belt; seated, hip and knee at 90 degrees, ankle at 10 degrees of plantar flexion. The bent knee targets the soleus; the force trace was shown on screen with a target line. MVC was the peak of three maximal efforts after three warm-up contractions: 101.9 ± 23.6 kg, 127.7 ± 29.4% of body weight. One session.
Expectation setting: every participant heard the same statement that "an Australian group have shown this to be excellent for knee tendons"; no attempt was made to blind them to the hypothesis.
Outcomes (before and immediately after): pain (0-10) during a graded functional task — 10 two-leg heel raises, then 10 one-leg heel raises, then 10 one-leg hops, stopping at the first that hurt and repeating the same task after; mechanical pressure pain threshold at the most tender spot (electronic von Frey, mean of three); isokinetic plantar-flexor torque with the knee straight, concentric at 90 and 225°/s and eccentric at 90°/s.
Analysis: paired t-tests on available data, no imputation.
Key Results
| Outcome | Before, mean (SD) | After, mean (SD) | Change, mean (95% CI) as printed (before − after) | n |
|---|---|---|---|---|
| Pain during the functional task, 0-10 | 4.2 (1.9) | 4.9 (3.2) | −0.7 (−1.82 to 0.49), not significant | 9 (7 had no pain on any task) |
| Pressure pain threshold, g/mm² | 446.5 (248.5) | 411.8 (211.8) | 34.8 (−49.2 to 118.8), not significant | 16 |
| Concentric torque at 90°/s, Nm | 47.1 (14.5) | 53.0 (18.5) | −5.9 (−11.4 to 0.33); p = 0.039 in the abstract | 16 |
| Concentric torque at 225°/s, Nm | 33.6 (11.6) | 33.4 (6.6) | 0.2 (−4.8 to 5.3), not significant | 16 |
| Eccentric torque at 90°/s, Nm | 98.5 (34.2) | 106.0 (41.4) | −7.5 (−24.1 to 9.1), not significant | 16 |
Who had pain to test: those with pain on the functional task had worse VISA-A scores (54.3 ± 15.8) than those without (70.3 ± 9.3); no other baseline differences.
Individual responses (from the authors' figure): participants with the highest pain scores got worse; those under about 4 of 10 improved by about 1 point, below the 2-point minimal important difference.
Torque: the 5.9 Nm rise at 90°/s is below the 12.7 Nm minimal detectable change; only 2 participants exceeded it.
Mechanisms
Measured: no group-level change in functional pain or pressure pain sensitivity, with changes in both directions between individuals.
Authors' interpretation: Achilles tendinopathy may involve cortical excitation rather than the cortical inhibition reported in patellar tendinopathy, so the mechanism proposed by Rio 2015 may not apply; the two tendons also change differently (the patellar tendon becomes stiffer, the Achilles less stiff).
Authors' interpretation: the positive patellar studies were very small and may simply not be robust.
Inference: with no control session, a small true effect cannot be excluded or confirmed; but a positive expectation was set and the dose was heavy (about 128% of body weight at MVC), and pain still did not fall.
Practical Application
Heavy calf holds for instant Achilles pain relief? In 16 running-club patients, 5 × 45 s at 70% MVC did not ease pain (4.2 → 4.9 of 10, not significant; 9 had pain to test) and the most painful got worse.
Do not use them as a same-day painkiller, but calf holds still load the tendon. This study tested acute relief only, not holds as part of a progressive programme. Not from this paper; the base's own practice: for Achilles rehabilitation the tested routes are heavy slow resistance or eccentric loading over about 12 weeks (REF-626, REF-625), with running continued under the pain-monitoring rule (REF-628, SYNTH-73).
Watch the high-pain athlete: those with the highest pain worsened after heavy holds. Not from this paper; the base's own practice, prompted by these individual data: if an athlete's tendon flares after holds, stop using them as a warm-up.
The dose tested: five 45-s plantar-flexor holds at 70% MVC, 2 min rest, seated with the knee bent (soleus bias), ankle at 10 degrees of plantar flexion.
Limitations
Single arm with no control session and no blinding; participants were primed with a positive statement about the exercise.
Small: 16 participants, and the primary pain outcome rests on the 9 who had pain during testing. The functional test stopped at one-leg hops, which may not provoke pain in runners who had already reduced their activity.
Reporting inconsistencies: Table 3 gives the pressure-threshold change as 34.8 (−49.2 to 118.8) but the text as −32.4 (−46.3 to 111.0); the 90°/s torque change is called significant (abstract p = 0.039) while its printed interval (−11.4 to 0.33) crosses zero.
Torque was tested with the knee straight, the holds with the knee bent; MVC was not re-tested after.
Older participants (mean 48.6) with long symptom duration; mid-portion tendinopathy only.
Funding: none. The authors declare no conflict of interest.
How this entry was checked and changed
Why this level
uncontrolled pre-post study, n=16 (rubric B); the pain outcome rests on only the 9 participants who had pain during the functional test, participants were told the exercise had been "excellent for knee tendons", and the paper's own tables and text disagree in two places (Limitations); read from the full text
Full text (checked 2026-10-16)
16 individuals (5 female) met the inclusion criteria; age 48.6 (± 8.9), VISA-A 61.3 (± 23.0), symptom duration 46.5 (± 55.2; 3.5 to 177.8) months (full text).
Intervention: five 45-s isometric plantar flexor contractions at 70% MVC separated by 2-min rests; ankle at 10° of plantar flexion, knee and hip at 90° flexion (full text).
MVC 101.9 kg (± 23.6), 127.7 (± 29.4) percent of body weight (full text).
Only 9 of the 16 participants reported pain on the functional tests; VISA-A 54.3 (± 15.8) with pain vs 70.3 (± 9.3) without (full text).
Pain during a functional task 4.2 (± 1.9) to 4.9 (± 3.2), change − 0.7 (− 1.82–0.49) (full text).
Pressure pain threshold 446.5 (± 248.5) to 411.8 (± 211.8), change 34.8 (− 49.2–118.8) in Table 3; text: a decrease of − 32.4 g/mm2 (95%CI − 46.3–111.0) (full text).
Concentric 90°/s 47.1 (± 14.5) to 53.0 (± 18.5) Nm, change − 5.9 (− 11.4–0.33); minimal detectable change 12.7 Nm (full text).
Participant statement: "an Australian group have shown this to be excellent for knee tendons" (full text).
Funding: "No funding was provided for completion of this study"; "The authors have no conflict of interest" (full text).