Lifting rewires the brain's output circuit, not just the muscle
In short
A meta-analysis of 11 randomised trials covering 244 healthy adults found that resistance training significantly increased corticospinal excitability measured by transcranial magnetic stimulation (standardised mean difference 0.59, 95% CI 0.10-1.08) and significantly improved motor performance (SMD 0.40, 95% CI 0.02-0.79). It quantifies what lifters see early on: the adaptation is not only in muscle tissue but in central nervous system plasticity.
In the first weeks of training, strength climbs while the mirror stays the same. That gap usually gets labelled neural adaptation. This meta-analysis measured the pathway from brain to muscle directly and put a number on it.
What was measured?
Transcranial magnetic stimulation (TMS) uses a magnetic field applied over the scalp to stimulate motor cortex and non-invasively gauge how readily the corticospinal tract — brain to spinal cord to muscle — responds. That readiness is excitability.
Following PRISMA 2020, the team searched Web of Science, PubMed, Embase, Cochrane Library, SPORTDiscus, Scopus, and CNKI to a cutoff of 27 February 2026, including only randomised trials in healthy adults where the experimental arm did resistance training and the control arm did sedentary activity or no exercise. That left 11 studies and 244 adults.
How large was the effect?
Under a random-effects model, resistance training significantly increased corticospinal excitability (SMD 0.59, 95% CI 0.10-1.08) and significantly improved motor performance (SMD 0.40, 95% CI 0.02-0.79). Methodological quality was high — mean Cochrane score 6.4 — and funnel plots indicated low publication-bias risk.
Heterogeneity was not small, though. I² was 69.3% for excitability and 64.8% for performance, meaning protocols and populations varied enough to scatter the effect widely.
How do you train on this?
The practical reading is lifting as skill practice. If adaptation happens centrally too, then repeating heavy squats, benches, and deadlifts often enough to learn the pattern is one axis of getting stronger. Adding volume and rehearsing a pattern are not the same job.
Eleven studies and 244 participants is a small evidence base with high heterogeneity. The authors state plainly that dose-response relationships between specific training parameters and neural adaptations remain to be worked out.
Reading it against your Muscle Index curve
Early on, Muscle Index can climb while bodyweight and measurements do not move. This result is evidence that stretch is real adaptation, not a measurement artefact. Neural adaptation does have a ceiling, so the rate will not hold.
Read alongside strength gains outrun muscle mass and squat strength rises before technique changes, the early surge stops looking mysterious.
Frequently asked questions
What is the evidence that resistance training changes the brain?
A meta-analysis of 11 randomised trials covering 244 healthy adults found that resistance training significantly increased corticospinal excitability measured by transcranial magnetic stimulation, with a standardised mean difference of 0.59 (95% CI 0.10-1.08).
What is corticospinal excitability?
It is how readily the neural pathway from the motor cortex through the spinal cord to the muscle responds to stimulation. It is measured non-invasively with transcranial magnetic stimulation applied over the scalp.
Did the neural changes translate into performance?
Yes. The same meta-analysis found a significant improvement in motor performance, with a standardised mean difference of 0.40 (95% CI 0.02-0.79).
How much training is needed to produce these adaptations?
This meta-analysis cannot say. The authors identify the dose-response relationship between specific training parameters and neural adaptations as an open question for future research.
What are the limitations of this meta-analysis?
The evidence base is small — 11 studies, 244 adults — with high heterogeneity: I² was 69.3% for corticospinal excitability and 64.8% for motor performance. Methodological quality was high (mean Cochrane score 6.4) and publication-bias risk was low.
Source: PubMed