Exercise for Patellar Tendinopathy: Cochrane’s Uncertain Verdict

Exercise is a sensible first treatment for patellar tendinopathy, but the best available review still cannot tell you whether to do eccentric decline squats, heavy slow resistance (HSR), or isometric holds—or how much of any of them to do. That is an awkward gap for a condition that commonly sidelines jumping athletes.

A 2025 Cochrane review found only seven randomized trials with 211 total participants. None directly compared the loading protocols most often debated in clinics and gym programs. Exercise may help, but the evidence does not establish a winning protocol, dose, frequency, or progression scheme.

That does not make rehabilitation pointless. It means confidence should stop where the data stop.

Exercise may reduce pain, but certainty is very low

Against no treatment, strengthening exercise reduced pain by a mean 34.94 points on a 0–100 visual analog scale in one trial of 39 participants (95% CI 20.94 to 48.94). A difference that large could matter to an athlete. Cochrane rated the certainty of that finding very low, however, so the estimate could change substantially with better trials.

Function told a less decisive story. Across two studies with 95 participants, exercise improved function by 7.04 points versus no treatment, but the confidence interval ran from 6.94 points worse to 21.02 points better. Because it crossed zero, the result was not statistically significant. Cochrane rated this evidence low certainty.

The distinction matters. A promising pain result does not prove that one loading program restores sport-specific function, prevents recurrence, or works equally well outside the young competitive athletes studied.

Exercise performed similarly to injection and surgery in small trials

In one 38-person trial, exercise did not show a statistically significant advantage over glucocorticoid injection at follow-up. Pain was 6.04 points worse with exercise on average (95% CI 8.19 points worse to 20.26 points better), and function was 5.75 points worse (95% CI 17.41 points worse to 5.93 points better). Both findings had low certainty. This does not show that injection is better; the study was too imprecise to settle the comparison.

The surgery comparison gives conservative care its strongest practical rationale. At 12 months, one 40-person study found no statistically meaningful difference between exercise and surgery in pain: exercise was 4.00 points worse on average, with a 95% CI from 4.06 points better to 12.06 points worse. Function was 7.30 points better with exercise, but the 95% CI ranged from 5.02 points worse to 19.62 points better. Return to sport was 85% with exercise and 86% with surgery (risk ratio 1.02, 95% CI 0.78 to 1.34). Cochrane rated pain and function evidence very low certainty and return-to-sport evidence low certainty.

Those results support trying progressive rehabilitation before an operation, given surgery's recovery burden. They do not prove equal long-term durability, equal recurrence rates, or equal results among people who have already failed a well-delivered rehabilitation program.

The eccentric-versus-HSR debate has no direct trial answer

The headline limitation is simple: none of the seven Cochrane-included trials compared eccentric, HSR, and isometric programs head to head. The field has protocols; it lacks the comparative randomized evidence needed to rank them.

HSR has become a familiar rehabilitation option. In a 60-athlete, randomized, double-blind, placebo-controlled trial with one-year follow-up, Olesen et al. gave both groups HSR and tested whether local IGF-1 injection added benefit. It did not. Both groups improved, but the study compared IGF-1 with placebo on top of HSR—not HSR with eccentric or isometric training. It supports skipping that adjunct in this setting; it cannot establish HSR as superior.

Eccentric training also remains an active comparator. A randomized trial by Sannasi et al. compared fascial manipulation with eccentric training plus static stretching. Using eccentric work as the comparison arm shows that researchers regard it as meaningful care. It does not show that eccentric loading beats HSR or isometric holds.

Some confidence in eccentric work also comes from Achilles-tendon practice. A systematic review of mid-portion Achilles tendinopathy found good evidence for eccentric training versus no therapy and evidence that eccentric and HSR training may have similar effects in that condition. Achilles and patellar tendons are not interchangeable clinical trials. Extrapolation can guide practice when evidence is thin, but it is still extrapolation.

Why the evidence stays uncertain

These trials have more problems than small sample sizes. Participants and clinicians cannot realistically be blinded to exercise, and all seven Cochrane-included trials had performance bias. None used a sham exercise program that matched attention and contact time. Therefore, exercise-versus-no-treatment results cannot isolate tendon loading from expectations, therapeutic contact, or other nonspecific effects.

Five of the seven trials had attrition bias. Four had problems with random sequence generation, three lacked adequate allocation concealment, and two had selective reporting bias. Each flaw can distort a result; together they explain Cochrane's cautious certainty ratings.

Harms are also a blind spot. The review reported zero adverse events across all seven trials, but no trial systematically captured, defined, and reported them. "No reported adverse events" means exactly that. It does not quantify the safety of high-load HSR or isometric regimens in a painful tendon.

Generalizability is narrow. All 211 participants were competitive athletes; 88% were male, their mean age was 26 years, and their mean symptom duration was 41.6 months. Yet patellar tendinopathy affects an estimated 7% to 18.3% of sporting populations and 1.6% of the general population, according to a 2023 overview. We do not know whether the same results apply to recreational athletes, women, older adults, or non-athletes.

Mechanism remains a hypothesis, not a prescription

A 2025 causal mediation analysis examined whether physical or imaging changes explained benefits from progressive tendon loading in athletes with patellar tendinopathy. Its findings raise the possibility that symptom improvement does not depend solely on visible structural tendon remodeling. That is useful mechanistic context, not proof that imaging changes—or their absence—should determine a program.

A 2025 narrative review has proposed neuroplastic periodization, incorporating cortical, spinal, and peripheral mechanisms. The proposal challenges eccentric-only thinking, but it does not supply the missing comparative trials. A theory can generate a better trial; it cannot replace one.

The problem extends beyond the patellar tendon. A mixed-methods synthesis across tendinopathy sites described small, heterogeneous exercise trials with inconsistent outcomes and frequent limitations in blinding. Non-patentable exercise interventions are hard to fund and hard to blind, but those realities do not make uncertain estimates more certain.

What to do with this evidence

If patellar tendon pain is limiting training, progressive exercise-based rehabilitation is a reasonable first step before surgery. Eccentric loading and HSR are both defensible choices because each is used as active treatment in the literature. The evidence does not justify claiming that one beats the other, and it provides no validated Monday-through-Sunday loading recipe.

Choose a structured program you can perform consistently, progress it with a qualified clinician when symptoms or training demands require it, and judge it by pain, function, and return-to-sport progress—not by a promise that one exercise has solved the protocol question.

The next useful study is not another untested variation with a dramatic name. It is a sufficiently powered randomized comparison of eccentric, HSR, and isometric loading with standardized parameters, blinded outcome assessment, systematic harm reporting, and participants who resemble the broader population with this condition. Until then, exercise is a practical first-line choice, while claims of protocol supremacy remain ahead of the evidence.