Suspension Training Added 4.8 cm to Jump Height in 71 Untrained Teenagers

A three-arm randomized trial in untrained 14-year-old boys tested suspension training against plain bodyweight resistance work and a control group. Suspension training produced the largest gains in jump height and 30-metre sprint time, but it did not beat bodyweight training outright.

Suspension Training raised countermovement jump height by an adjusted 4.822 cm over a non-exercising control group in untrained male adolescents, according to a randomized controlled trial published on September 21, 2026 in BMC Sports Science, Medicine and Rehabilitation. Sprint times improved as well, while agility did not move.

The trial was conducted by Yu Wang of Beijing Sport University and Tsinghua University High School-Shangdi, with Mingyue Sun of Beijing City University, Yong Yang of Chengdu Sport University, Weiqiang Xu of Beijing Sport University and the Gdansk University of Physical Education and Sport, and Alexios Batrakoulis of European University Cyprus and Democritus University of Thrace.

Seventy-two participants with a mean age of 14.74 years were randomized equally to suspension training, bodyweight resistance training or a control condition. One participant in the resistance arm withdrew because of an illness unrelated to the intervention, and because that individual data was unavailable the team could not run an intention-to-treat analysis. The reported results come from a per-protocol, complete-case sample of 71.

Countermovement jump was the primary outcome. Medicine ball throw, 30-metre sprint and a T-Agility test were secondary. Between-group inference used baseline-adjusted ANCOVA models for each outcome, followed by three prespecified contrasts with Holm correction applied within each outcome.

How the Suspension Training Trial Was Run

Three-arm designs are the useful kind in this corner of the literature, because they answer two questions at once. Does the intervention beat doing nothing extra, and does it beat the cheaper, simpler alternative? Suspension training was compared against both.

The analysis choices are more careful than is typical for a trial this size. Baseline-adjusted ANCOVA is the appropriate model when groups may differ slightly at the start, and Holm correction guards against the false positives that accumulate when three comparisons are run on each of four outcomes. The authors prespecified the contrasts rather than picking them after seeing the data.

The honesty about the analysis population is also worth noting. The paper states plainly that an intention-to-treat analysis could not be performed and that the results describe a per-protocol, complete-case population. That is a real constraint, stated rather than buried.

The Jump Result Was the Clearest Signal

Countermovement jump produced the strongest omnibus effect of the four outcomes: F(2,67) = 29.162, p < 0.001, with a partial eta squared of 0.4654. In plain terms, group assignment explained a large share of the variation in how high these boys jumped after the programme.

The suspension training arm gained an adjusted 4.822 cm over control, with a familywise simultaneous confidence interval from 3.198 to 6.445 cm. Even the bottom of that interval, a little over 3 cm, is a change an athlete would notice. The bodyweight resistance arm gained an adjusted 3.710 cm over control, with an interval from 2.072 to 5.348.

Both figures are consistent with what the wider youth literature reports. Our summary of resistance training and muscle growth across 1,443 young people found small size changes alongside larger performance changes, which is the usual pattern at this age: neural and coordinative gains arrive well before tissue does.

Sprint Times Improved, Agility Did Not

The 30-metre sprint followed the jump result. The omnibus test was significant at F(2,67) = 12.090, p < 0.001, partial eta squared 0.2652. Suspension training cut an adjusted 0.200 seconds off the sprint relative to control, with an interval from -0.302 to -0.099. The bodyweight arm cut 0.127 seconds, with an interval from -0.230 to -0.024 and a p value of 0.007.

Agility told a different story. The T-Agility omnibus test was not significant (F(2,67) = 1.304, p = 0.278, partial eta squared 0.0375), and no agility contrast reached significance. The medicine ball throw omnibus test was significant at p = 0.034, but none of its three pairwise contrasts survived Holm correction.

That split is instructive rather than disappointing. Vertical jump and straight-line acceleration both load the same extension pattern through hip, knee and ankle. A change-of-direction test asks for braking, re-orientation and re-acceleration, a skill set that a general strength stimulus does not automatically transfer to.

Suspension Training and Bodyweight Work Were Statistically Level

The comparison most readers will care about produced no winner. Suspension training and bodyweight resistance training did not differ significantly on any outcome. Both beat the control condition on jump and sprint; neither beat the other.

That is a genuinely useful null result. Suspension straps are a piece of equipment with a cost attached, and this trial gives no evidence that untrained adolescents need them to get the jump and sprint gains on offer. The adjusted point estimates favoured suspension training slightly, but the confidence intervals overlap and the formal test did not separate them.

A similar pattern has shown up elsewhere in the bodyweight literature. Our report on blood flow restriction added to remote bodyweight training found that the added technique produced no extra benefit over the bodyweight programme alone.

Why the Statistics Deserve a Closer Look

Seventy-one adolescents split three ways leaves roughly 24 per arm, which is adequate for a large effect and thin for a modest one. The jump and sprint effects were large enough to clear that bar. The agility and throw results may be genuine nulls, or they may be a sample too small to detect smaller differences. The trial cannot distinguish between those possibilities.

The absence of an intention-to-treat analysis is the other constraint the authors flag. With a single dropout it is unlikely to have changed the direction of anything, but per-protocol analysis is the more permissive of the two approaches and it is right that the paper says so.

Generalisability is bounded by design. These were untrained male adolescents of a narrow age band, and untrained populations improve on almost any structured stimulus. Trained athletes of the kind covered in our piece on sprint interval training and explosive performance start from a very different baseline.

What Suspension Training Offers a School or Home Setup

This is a trial in untrained teenagers, run under supervision, and it is not a programme template for adults or for trained young athletes. It does speak to a practical question that school and community coaches face constantly: what is worth buying.

A few reasonable readings for a general audience:

• Both suspension training and plain bodyweight work produced meaningful jump and sprint gains against doing nothing extra.
• Neither approach improved change-of-direction agility, so that quality likely needs to be trained directly.
• Equipment was not the deciding variable in this trial, which is encouraging for programmes without a budget.
• Young athletes training for the first time should start under qualified supervision, whatever the modality.

For context on how other structured programmes have performed with adolescent athletes, our coverage of a multimodal HIIT trial in adolescent basketball players tracks a different outcome in a comparable age group.