REF-1145 · Rio et al., 2017 · Clinical Journal of Sport Medicine · Paper · PubMed

Isometric contractions are more analgesic than isotonic contractions for patellar tendon pain in-season

Over 4 weeks of 4 sessions a week, isometric leg-extension holds eased decline-squat pain within each session more than isotonic leg extensions (P < 0.002), both programmes reduced pain, n=20; the one positive in-season RCT; a pre-registered single-session crossover (Holden 2020, a replication of Rio 2015) did not confirm the advantage

B+moderate evidenceMeasured on trained to highly trained athletes

Rio and colleagues asked whether isometric or isotonic quadriceps work gives more immediate pain relief to jumping athletes with patellar tendinopathy who keep competing.

Twenty sub-elite volleyball and basketball players (18 men, mean 22.5 ± 4.7 years) were randomised to isometric leg-extension holds at 60 degrees of knee flexion at 80% of their maximal voluntary isometric contraction (MVIC), or to isotonic leg extensions at 80% of their 8-repetition maximum, 4 times a week for 4 weeks, with time under load and rest matched.

Pain during a single-leg decline squat (SLDS, 0-10) was rated before and after each session. Baseline median pain was 5/10 in both groups. Isometric sessions produced significantly greater immediate analgesia throughout the 4 weeks (P < 0.002). The size of week-1 relief correlated with the improvement in VISA-P at 4 weeks (r = 0.64).

The authors conclude that both programmes reduce in-season pain and that isometrics give more immediate relief, which "may increase the ability to load or perform". This is the only in-season RCT in the base where isometrics beat isotonic work; the pre-registered crossover in 21 patients (REF-1147) and Clifford's meta-analysis (REF-1149) found no difference.

Protocol

Participants (from the abstract): 20 jumping athletes over 16 years old, clinically diagnosed patellar tendinopathy, taking part in games or training 3 times a week in sub-elite volleyball and basketball competitions; isotonic n = 10, isometric n = 10.
Isometric arm: leg-extension holds at 60 degrees of knee flexion at 80% of MVIC. The abstract does not say whether the holds were against a machine weight (yielding) or a fixed stop (overcoming), and gives no hold duration, sets or rest.
Isotonic arm: leg extensions at 80% of 8RM.
Dose matching: time under load and rest between sets matched between groups.
Frequency and duration: 4 sessions a week for 4 weeks, during the competitive season. The abstract does not say whether sport load was changed; in van Ark 2016 (REF-1146), from the same group, it was not.
Outcomes: SLDS pain (0-10 numerical rating) before and after each session; VISA-P at baseline and after 4 weeks.
Analysis: conducted blinded to group.
Not in the abstract: whether sessions were supervised, the statistical model, dropouts, adverse events, the per-group VISA-P change.

Key Results

OutcomeIsometric (n=10)Isotonic (n=10)Statistics
Participants18 men of 20, mean 22.5 ± 4.7 years(pooled)—
Baseline SLDS pain, median (range)5/10 (2-8)5/10 (1-8)—
Immediate analgesia per session, across 4 weeksgreatersmallerP < 0.002
Mean SLDS pain before → after a session, as extracted by Clifford 2020 (REF-1149, Table 6)5 → 3.25 → 4.1treatment effect −0.9 (−1.1 to −0.7), Clifford's calculation; reductions 1.8 vs 0.9 points in Clifford's discussion
Week-1 analgesia vs VISA-P gain at 4 weeks——r = 0.64 (groups not separated in the abstract)
Pain over 4 weeksreducedreduced"both protocols appear efficacious"; no numbers in the abstract
Size of the effect: the abstract gives a P value but no effect size. Clifford's extraction puts the mean extra relief from isometrics at about 0.9 of 10 points per session, under the 2-point change usually taken as clinically important (REF-1147 uses 2 points). Our reading: a statistically clear but small gap, far from the 6.8 vs 2.6 point gap in Rio 2015 (REF-627).

Mechanisms

Measured: week-1 within-session relief correlated with the 4-week VISA-P improvement (r = 0.64).
Authors' interpretation: greater analgesia may let athletes load or perform more. Not tested here: training load and performance were not reported as outcomes in the abstract.
Not measured here: the cortical inhibition change that Rio 2015 (REF-627) proposed as the mechanism.
Inference: the correlation may mean early responders do better, or that the same people improve on both measures for reasons unrelated to the exercise (regression to the mean, expectation). It does not show that the analgesia caused the VISA-P gain.

Practical Application

Isometric holds for patellar tendon pain? In an in-season RCT of 20 jumpers they eased pain within a session more than leg extensions (P < 0.002, about 1 of 10); a pre-registered one-session crossover found no difference ([REF-1147]).
Offer holds as something an athlete may try, not as a reliable pain-reliever. Either programme is compatible with playing on: both reduced pain over 4 weeks without the athletes stopping their sport (same group: REF-1146, where the two programmes did not differ). Let the athlete's preference and response decide, which is also what the authors of REF-1147 advise.
The dose tested (from the abstract): leg-extension holds at 60 degrees of knee flexion at 80% of MVIC, 4 times a week for 4 weeks, matched for time under load with 80% of 8RM leg extensions. Hold length and sets are not in the abstract; 5 × 45 s is the Rio 2015 dose (REF-627).
Not from this paper; the base's own practice: for a runner or cyclist with front-of-knee tendon pain, isometrics are one form of tendon loading inside a progressive plan (REF-626 heavy slow resistance; REF-624 load, not rest), under the pain-monitoring rule in SYNTH-73. The transfer from jumpers is untested.

Limitations

Abstract only: hold duration, sets, rest, supervision, dropouts, adverse events, the effect size and its confidence interval could not be checked.
Small (n=20) and short (4 weeks, no longer follow-up); 18 of 20 were men; jumping athletes only.
Pain is self-rated and participants could not be blinded to which exercise they did; the abstract states only that the analysis was blinded. Clifford 2020 rated the trial poor quality (high risk for blinding of participants and assessors, attrition, external validity and precision).
Possible overlap with van Ark 2016 (REF-1146): same research group, same competitions, same two programmes; Clifford tabulates both with the same cohort (29 athletes, 13 isometric and 16 isotonic, 62% follow-up). Whether these 20 are a subset of van Ark's 29 cannot be settled from the abstract; if they are, the two papers are one trial, not two.
Not confirmed: a pre-registered single-session crossover in 21 patients found no difference between isometric and dynamic exercise (REF-1147).
Funding and conflicts of interest are not stated in the abstract.
How this entry was checked and changed
Why this level

(from the abstract) — parallel RCT with n=20 (10 + 10), the bottom of the B+ band: no open full text, so dose details, the between-group effect size and its confidence interval, dropouts and blinding of participants and testers are unchecked. Clifford 2020 (REF-1149) rated it poor quality and tabulates it with the same 29-athlete cohort as van Ark 2016 (REF-1146), so it may not be independent evidence