Resistance Training Produced Small Muscle Growth in 1,443 Youth, 23-Study Review Finds

A new meta-analysis pooled 23 controlled trials in 1,443 people under 18 and found that resistance training produced small but measurable muscle growth against inactive comparison groups. Against active comparison groups the effect was substantially larger, and postpubertal teenagers gained the most.

Muscle Growth in young lifters has been argued over for decades, with the long-standing claim that children gain strength mainly through neural adaptation rather than added tissue. A systematic review published on September 17, 2026, in Sports Medicine puts numbers on the question across 23 trials and 1,443 participants younger than 18.

The review was led by Yannik Komm and Linda Zeiler at the Institute for Sport Science, University of Augsburg, with Claudia Augste as senior author. The co-author list includes Steffen Held, Paul Comfort and Avery Faigenbaum, names long associated with youth strength research.

The team set out to quantify how much resistance training produces muscle growth in young people, and whether the answer depends on maturational stage, sex, or whether the programme was actually designed to build tissue. They compared training groups both with passive controls, who did nothing extra, and with active controls, who did some other form of training.

That second comparison is where the result turns interesting, because it reverses the usual expectation about which contrast should produce the bigger number.

Why the Comparison Group Changed the Answer

Against a passive control, the pooled effect was small: a Cohen d of 0.22, with the confidence interval running from 0.06 to 0.37. That is a real but modest signal, and it sits comfortably with the conventional view that young lifters do not achieve large amounts of muscle growth quickly.

Against an active control, the same training produced a Cohen d of 0.61. The interval, 0.03 to 1.20, only barely clears zero, so the precision is poor. But the point estimate is nearly three times larger.

That ordering is counterintuitive. Comparing against people who are also training should normally shrink an effect, not inflate it. The finding suggests the active comparison groups in these trials may have been doing something that worked against muscle growth, or that the trials using active controls differed systematically from those using passive ones.

Muscle Growth by Stage of Puberty

The maturational breakdown is the part most likely to change practice. Against passive controls, peripubertal participants showed the clearest result, a Cohen d of 0.30 with an interval of 0.01 to 0.58. Prepubertal participants came in at 0.16, with an interval from minus 0.04 to 0.36 that crosses zero. Postpubertal participants registered 0.01, essentially nothing.

Against active controls the picture inverted. Subgroup differences by maturational stage reached statistical significance at p equal to 0.0227, and postpubertal participants showed a large effect, a Cohen d of 1.50 with an interval of 0.46 to 2.54.

The two comparisons therefore point in opposite directions for the oldest group. The authors conclude that hypertrophic adaptations occur across various maturational stages, but that postpubertal youth show the greatest increases when contrasted with active comparators.

What the Sex Comparison Did and Did Not Show

Against active controls, male participants showed a Cohen d of 1.26, with an interval of 0.29 to 2.22. Female participants showed minus 0.12, with an interval from minus 1.05 to 0.81 that spans zero in both directions.

It would be easy to read that as evidence that young women do not achieve muscle growth. The authors do not, and neither should readers. Their stated conclusion is that sex did not consistently influence outcomes, and the female interval is so wide that it is compatible with a meaningful gain, no change, or a small loss. Wide intervals usually mean too few participants, not a settled answer.

How Muscle Growth Was Measured

Methodology is doing real work in this review. The authors excluded studies that relied only on anthropometric proxies such as skinfolds or limb circumferences, on the grounds that these methods do not provide sufficient granularity. Included trials had to use direct imaging or instrumented methods: MRI, CT, ultrasound, DXA, bioelectrical impedance or air displacement plethysmography.

That screen raises confidence that the changes reported reflect real muscle growth rather than tape-measure noise. It also narrows the pool, which is part of why 23 trials and 1,443 participants is the whole available evidence base rather than a selection from a larger one.

Where the Evidence Is Thin

Study quality ranged from 3 to 7 points on the PEDro scale, a spread that runs from weak to reasonably solid. The authors note that blinding of participants as well as the personnel delivering and supervising the intervention is generally not feasible in training research, which caps the score any such trial can reach.

Beyond that, the intervals tell their own story. A confidence interval of 0.03 to 1.20, or 0.46 to 2.54, is not a precise estimate. It is a direction with a great deal of uncertainty attached.

What This Means for Coaches and Parents

The safe reading is that supervised resistance training produces muscle growth in young people by a small amount, and that the size of the change depends on what the comparison group was doing. None of this speaks to whether any individual young person should start training, which is a decision for them and their clinician.

General points that follow from the evidence rather than from any individual case:

• Expect modest muscle growth. The most defensible number in this review is a Cohen d of 0.22 against doing nothing extra.
• Maturational stage matters more than birthday. The review sorted participants by stage for a reason.
• Do not read the female estimate as a ceiling; it is an underpowered estimate, not a finding about capacity.
• Strength gains and muscle growth are different outcomes, and this review measured only the second.

That last point connects to work we have covered before. Our report on how type II fibres showed the biggest protein changes in adult lifters looked at the molecular layer beneath tissue change, while a six-week trial in which isometric training raised strength and rapid force is a reminder that strength can climb without much size following it.

Why Muscle Growth in Youth Was Doubted for So Long

The prevailing view for decades held that children who get stronger do so mainly by learning to recruit the muscle they already have, and that real muscle growth waits for the hormonal changes of puberty. That belief shaped curriculum, policy and a good deal of coaching caution about putting barbells in front of young people.

This review does not overturn that view so much as complicate it. A Cohen d of 0.22 against passive controls is small, which is broadly consistent with the neural-adaptation account. But it is not zero, and the peripubertal estimate of 0.30 sits above the pooled average, which argues that muscle growth is available earlier than the older framing allowed.

What the review cannot do is say how much muscle growth is possible under optimal programming. The trials it pooled were not designed to answer that question. They were designed to test whether training helped at all, which is a lower bar and a different one.

The Youth Measurement Problem Again

Sorting young athletes by maturational stage rather than age is not a statistical nicety. Two 13-year-olds can be years apart biologically, which scrambles any comparison built on birthdays. We made the same point covering why early talent selection may miss future elite athletes.

The same measurement problem runs through the other youth story we publish today, in which physical fitness predicted arithmetic fluency in 1,166 adolescents while activity minutes predicted nothing. In both cases, what gets measured determines what gets found, and capacity measures behave differently from participation measures.