Resistance training missed its brain-lesion target, but helped women

White-matter hyperintensities — the bright patches that show up on brain MRI as small vessels age and leak — are a hallmark of subcortical vascular cognitive impairment and a recognised risk factor for dementia [s1]. A new randomised trial asked whether a year of progressive strength training could slow their growth. The honest answer is: not on the measure the trial was built around, and selectively on others [s1].

What the trial did

The study enrolled 89 community-dwelling adults aged 55 or older living with subcortical vascular cognitive impairment, and randomly assigned them 1:1 to one of two 12-month programmes [s1]. One group did progressive resistance training twice a week, 60 minutes a session (n=44); the control group did a matched schedule of balance-and-tone exercises (n=45) [s1]. Participants were mostly women — 58 of 89, or 65.2% — with a mean age of 74.5 years [s1]. The trial is registered as NCT02669394 [s2].

White-matter lesion volume was measured across the whole brain and in specific regions using MRI, and the main comparison was the change over 12 months, adjusted for baseline lesion load, head size, age, sex, and whether training was done remotely or in person [s1]. Calling it a "proof-of-concept" trial is the authors' own framing, and it sets expectations correctly: this was a small study designed to see whether an effect is plausible, not to settle the question [s1].

The result, starting with the one that counts

On the primary measure — change in whole-brain white-matter lesion volume at 12 months — there was no difference between the groups. The estimated mean difference was −0.016 (95% confidence interval −0.052 to 0.019) on a log-transformed scale, with a p-value of 0.364 [s1]. That interval straddles zero comfortably; by the trial's own front-line test, strength training did not slow lesion progression across the brain as a whole.

What makes the paper more than a null result is where the signal did appear. In two prespecified regions — the temporal lobe (−0.115; 95% CI −0.207 to −0.023; p=0.015) and the splenium of the corpus callosum, a bundle at the back of the brain's main connecting tract (−0.120; 95% CI −0.215 to −0.025; p=0.014) — lesion volume favoured the training group [s1].

The sharper finding was by sex. There was a significant group-by-sex interaction (p=0.019), and when the authors looked separately, resistance training reduced whole-brain lesion volume in women (−0.049; 95% CI −0.093 to −0.005; p=0.029) but not in men (0.044; 95% CI −0.017 to 0.105; p=0.152) [s1]. An exploratory analysis linked rising levels of vascular endothelial growth factor, a signalling protein involved in blood-vessel health, to lesion reduction (standardised β −0.503; 95% CI −0.838 to −0.167; p=0.004), while a second growth factor, IGF-1, showed no association [s1]. White-matter lesions normally grow slowly, so detecting any difference over a single year in a group this size is hard — part of why a mixed result is not surprising [s1].

How much weight the hopeful parts can bear

Not as much as a headline would like. The whole-brain result was the question the trial was powered to answer, and it was negative [s1]. The regional and sex-specific effects are the kind of secondary and subgroup findings that are easy to over-read: with a sample of 89 split further by sex and region, these analyses are underpowered and prone to chance, and the authors present the sex breakdown as a post-hoc contrast [s1]. The growth-factor link is explicitly exploratory and describes an association, not a cause [s1].

So the result is genuinely two-sided. It does not support telling people with vascular cognitive impairment that lifting weights will clear white-matter lesions — the primary endpoint says the opposite [s1]. But it does leave a specific, testable hypothesis on the table: that the benefit, if real, is concentrated in women and in particular tracts, possibly through a vascular pathway. That is a lead for a larger, sex-stratified trial, not a conclusion.

What to watch

The wider case for exercise in ageing brains does not rest on this study, and nothing here weakens the established reasons to stay physically active. What this trial adds is narrow: a hint, in a small proof-of-concept study, that the effect of strength training on vascular brain lesions may not be uniform across patients [s1]. The next question is whether a trial built from the start to test the women-specific signal — adequately powered, with the regional measures prespecified — reproduces it. Until then, the primary result is the one to hold onto: no overall effect [s1].

This article is informational and does not constitute medical advice.

Sources

• [s1] Exercise and white matter hyperintensities: a randomized clinical trial. Brain Communications, 8 October 2026. https://doi.org/10.1093/braincomms/fcag381
• [s2] Reshaping the Path of Vascular Cognitive Impairment With Resistance Training. ClinicalTrials.gov, registration NCT02669394. https://clinicaltrials.gov/study/NCT02669394