Eccentric Training Added 2.92 cm to Countermovement Jump in 11-Trial Review

A review published today pooled 11 randomized trials to test what eccentric training does for explosive power in trained athletes. Vertical jump height moved by nearly 3 cm; sprint times over 10, 20 and 30 metres did not move at all.

Eccentric training has spent two decades migrating from injury clinics into mainstream strength programmes, and a systematic review published on October 12, 2026 in Frontiers in Physiology asked what that migration has actually bought in terms of power. Across 11 randomized controlled trials and 348 athletes, the countermovement jump improved by a pooled 2.92 cm, while sprint performance showed no effect the pooled data could confirm in any direction.

The review was carried out by researchers at Yuncheng University, Chongqing Normal University, Baoshan University and Northwest Minzu University, and registered with PROSPERO as CRD420261380375. The literature search was run on July 20, 2026 across six databases indexing Chinese-language and English-language publications.

Of the 348 athletes pooled, 51 were female. Methodological quality was assessed with the PEDro scale and rated low to moderate, a judgement the authors use to frame the entire review as exploratory rather than conclusive.

The jump result was the clearest thing in the paper: a mean difference of 2.92 cm favouring the eccentric training groups, with a 95% confidence interval from 1.26 to 4.59 cm, a Z value of 3.44 and a p value below 0.001. Heterogeneity was moderate at an I-squared of 52%.

Why Eccentric Training Is Supposed to Build Power in the First Place

Every lift has a lowering phase, and muscle can resist more load while lengthening than it can lift while shortening. Eccentric training exploits that gap, loading the lowering phase heavily through slow tempos, flywheel devices, drop-and-catch work or an overloaded descent with assistance on the way up.

The mechanical argument is that this stresses tendon and connective tissue, shifts muscle architecture and improves the stretch-shortening cycle, which is precisely the mechanism a countermovement jump depends on. An athlete dips, stores elastic energy in a rapid eccentric phase, then reverses it. If eccentric training improves that reversal, jump height should rise.

It did. The 2.92 cm gain is larger than most strength interventions produce in trained athletes over a comparable period, which is part of why the result is interesting and part of why the heterogeneity deserves attention.

The Eight-Week Line That Split the Jump Results

The most practically useful analysis in the review is a subgroup split by programme length. Trials running 8 weeks or longer produced a pooled jump gain of 3.41 cm, with a 95% confidence interval from 1.38 to 5.44 cm and a p value of 0.001. Trials shorter than 8 weeks pooled to a loss of 1.26 cm.

The formal test for a difference between those subgroups returned a p value of 0.009, which is about as clean a dose signal as a review of 11 trials can generate. Short eccentric training blocks, on this evidence, do not simply produce a smaller benefit; they may produce none.

That fits what coaches already report about unaccustomed eccentric work. The first weeks bring soreness, stiffness and depressed force output, and a trial that stops mid-adaptation can measure the cost without ever reaching the return.

Why the Eccentric Training Sprint Numbers Came Back Empty

Sprinting is the outcome most coaches actually care about, and it is where eccentric training produced nothing. The 20-metre sprint shifted by 0.10 s with a confidence interval running from 0.21 to 0.02 and a p value of 0.11. The 30-metre sprint gave a standardised effect of 0.42, interval 1.12 to 0.27, p value 0.23.

Both intervals comfortably include zero. The direction of the point estimates favours eccentric training, which is suggestive, but neither reaches a threshold where a coach should expect a stopwatch to notice.

There is a plausible physiological reason. A vertical jump is a single, bilateral, largely vertical expression of force against a long ground-contact time. Sprinting is a cyclical, unilateral, horizontal task with contact times under a fifth of a second. Transfer between the two has never been automatic, and this review offers no evidence that eccentric training bridges the gap.

What an I-Squared of 99% Actually Tells You

The 10-metre sprint analysis reached an I-squared of 99%, and to their credit the authors declined to pool it, summarising results at the level of individual studies instead. An I-squared that high means the included trials were not measuring the same thing in the same population under the same conditions.

Ten-metre times are short enough that starting technique, timing-gate placement and reaction protocol can dominate the signal. A pooled number from trials that differ on those details would look authoritative and mean very little.

Where the Review Reports Itself Inconsistently

Readers going to the paper should know that it states the 8-week subgroup effect as 3.41 cm in the results text and as 1.91 cm in the abstract, and that the number of included trials appears as 12 in the methods and 11 in the results after a PEDro-based exclusion. The 30-metre outcome is also labelled a mean difference in one place and a standardised effect in another.

None of this overturns the headline finding, and the direction of every estimate stays the same. It does mean the precise magnitude of the eccentric training subgroup effect should be read with caution until the version of record settles it.

What the Review Could Not Test

The authors wanted to examine moderators and could not. Sport type, competitive level, training intensity, sex and age all went untested because most of the primary studies never reported stratified statistics, so the pooled estimates may be hiding real differences between groups of athletes.

Participants were also overwhelmingly male. Only one included trial used a mixed-sex sample, and it did not report results separately by sex. With 51 female athletes across the entire pool, this review says very little about how women respond to eccentric training for power.

Change-of-direction performance, arguably the outcome closest to field sport reality, could not be pooled at all. Only the countermovement jump and the three sprint distances had enough data.

How to Read This as a Lifter or a Coach

The practical reading is narrow but usable. If vertical jump is the target, eccentric training looks worth programming, and it looks worth programming for at least 8 weeks. If sprint time is the target, this evidence gives no reason to expect a change. General, non-individualized points:

• Plan eccentric blocks long enough to pass through the soreness phase, since trials under 8 weeks showed no jump benefit.
• Expect elevated muscle soreness early, and schedule heavy eccentric work away from competition and testing dates.
• Keep eccentric training as one component of a programme rather than a replacement for conventional lifting and sprint work.
• Jump height and sprint time are different qualities; test whichever one you are actually trying to change.
• Heavy supramaximal lowering generally needs a spotter, a machine or a device built for it, not improvisation.

Anyone using jump height as a monitoring tool should read this alongside our coverage of the scoping review that found countermovement jump fatigue testing varies widely across 117 studies, because a 2.92 cm change only means something if the test itself is stable. The pattern of strong jump gains without sprint transfer also appeared in a review of sprint interval training and explosive performance in high-level athletes.

For comparison on dose and timing, a trial of resistance priming that raised jump height 5.5% at 30 hours works on a far shorter timescale, while suspension training added 4.8 cm of jump height in 71 untrained teenagers, a reminder that untrained populations move much further than the trained athletes pooled here.