Twelve Weeks of Strength Training Reshapes Aging Bodies Unevenly, Study Finds

Resistance training has long been prescribed as one of the most reliable non-pharmacological tools for preserving muscle and mobility in later life, yet a new study suggests that the way older adults respond to it is far more complicated than a simple before-and-after measurement can capture. In research published in European Geriatric Medicine, a team led by Špela Matko and Vincent Grote of the Ludwig Boltzmann Institute for Rehabilitation Research followed 126 healthy older adults through a 12-week supervised, resistance-based exercise programme and then examined their progress across four distinct clinical domains rather than a single endpoint. The results reveal a strikingly asynchronous pattern of adaptation: functional performance surged, muscle capacity and body composition barely moved, and pain and health-related quality of life followed wildly divergent trajectories from one participant to the next.

The study was conducted within the Center of Active Ageing Project in St. Pölten, Austria, and recruited community-dwelling volunteers aged 60 to 80 who could walk with or without aids. After exclusions for missing assessments and withdrawals, the final analytic sample comprised 126 participants with a mean age of 67.7 years, roughly 63 percent of them women, and a mean body mass index of 28.1 kilograms per square metre. Importantly, this was a generally healthy and well-functioning cohort: nutritional screening placed 98.4 percent of participants in the normal range, and only 4.8 percent were classified as at risk for sarcopenia based on sex-specific phase angle cut-offs derived from bioelectrical impedance analysis. The participants completed two supervised 60-minute sessions each week, one full-body resistance-training session and one neuromotor session focused on coordination, mobility, and balance.

The resistance training followed a carefully periodised progression referenced to each participant’s estimated one repetition maximum. During the first three weeks, participants performed two sets of 20 repetitions at 75 percent of their one repetition maximum; from weeks four to six, the load increased to 80 percent for two sets of 12 to 15 repetitions; and from weeks seven to twelve, they progressed to three sets of 8 to 12 repetitions at 85 percent. The programme targeted all major upper-limb, lower-limb, and trunk muscle groups with individualised load progression, and the neuromotor sessions became progressively more demanding. Sessions were delivered by a multidisciplinary team of physiotherapists and sport scientists, who, together with dietitians, also provided brief education on healthy lifestyle, safe exercise, nutrition, and ageing-related risks.

What sets this study apart methodologically is how the researchers handled the sheer volume of measurement data. At baseline and after 12 weeks, participants underwent a comprehensive battery covering anthropometry, body composition, resting hemodynamics, upper and lower extremity function, walking capacity, static and dynamic balance, joint mobility, muscle strength, health-related quality of life, nutritional risk, physical activity level, and medical history. Starting from 43 outcome measures, the team applied a multistep dimensionality reduction procedure based on principal component analysis to distil the data into 14 key indicators, which were then aggregated into four latent domains: body composition, muscle capacity, functional performance, and pain and health-related quality of life. The four-domain solution explained 72.7 percent of the variance in the retained indicators, with functional performance and body composition each contributing roughly a quarter of that explained variance.

To capture overall responsiveness, the researchers combined the four domain scores into a composite Multidomain Quality Outcome Index, calculated as the unweighted mean of the standardised factor scores at each time point. Individual change was quantified as the difference between post-intervention and baseline z-scores, with higher values indicating improvement. Participants were classified as multidomain responders if their composite index change reached at least 0.33 standardised mean difference units, a threshold described in the multidimensional rehabilitation literature as both perceptible and clinically meaningful. The authors emphasise that this cut-off was selected a priori as a conservative, descriptive candidate threshold and should not be interpreted as a validated clinical standard, particularly since it was derived and applied within the same sample.

The headline finding was the dramatic and consistent improvement in functional performance, which registered a standardised mean difference of 0.67, by far the largest effect in the study. Eighty percent of participants improved in this domain, encompassing gains in the 30-second Chair Stand test, the Timed Up and Go test, and both usual and fast-paced 10-metre walking speed. The composite index improved by a more moderate 0.30 standardised mean difference, with 59.5 percent of participants showing overall improvement. By contrast, muscle capacity remained essentially stable, body composition showed a slight negative shift, and changes in pain and health-related quality of life were strikingly heterogeneous, with nearly equal proportions of participants improving and worsening in that domain. The authors note that baseline performance was already high, with a mean Timed Up and Go time of 6.10 seconds, faster than pooled age-specific reference values, making recovery from impairment an implausible explanation for the gains.

When the researchers applied the responder threshold, 46 percent of participants qualified as responders, while the remaining 54 percent did not. Crucially, the non-responders were not simply people who failed to benefit. They still exhibited measurable improvements in functional performance, but their profiles were characterised by worsening in the pain and quality-of-life domain and small deteriorations in muscle capacity and body composition. The only baseline difference between the groups emerged in the pain and quality-of-life domain, where responders reported poorer subjective health at the outset. This suggests that responder status was driven largely by baseline symptom burden and the potential for subjective improvement rather than by any objective functional or physiological impairment, a finding with significant implications for how clinicians interpret short-term exercise outcomes in healthy ageing populations.

The pattern of early functional gains without parallel structural change fits well with established exercise physiology. Neural adaptations are known to dominate the early phases of resistance training, while hypertrophic responses typically require longer intervention periods or higher training intensities. In this framework, resistance-based exercise primarily enhances neuromuscular efficiency and coordination, producing functional improvements even in the absence of measurable changes in muscle mass. The study’s regression analyses reinforced the complexity of the picture: baseline characteristics such as sex, age, body mass index, physical activity, and nutritional status explained a strong proportion of variance in baseline muscle capacity but had limited explanatory value for short-term changes. Poorer baseline pain and quality-of-life scores and lower composite index values were associated with greater overall improvement, consistent with baseline dependency and regression-to-the-mean phenomena well documented in rehabilitation science.

The authors are careful to acknowledge the limitations of their design. As a single-arm study without a non-exercising control group, the observed changes cannot be categorically attributed to the intervention alone; learning effects, test familiarisation, spontaneous change, and regression to the mean may all have contributed. The self-selected, healthy cohort with preserved baseline function may also have introduced ceiling effects, particularly for subjective outcomes, and body composition was assessed with bioelectrical impedance rather than reference imaging. No formal adjustment for multiple comparisons was applied, and the four-domain structure, composite index, and responder threshold were all developed within the same sample, meaning external validation in independent cohorts is required before clinical application.

Nevertheless, the study’s central message carries real weight for geriatric medicine and preventive health. With the global population aged 60 and older projected to reach approximately 2.1 billion by 2050, and with healthy life expectancy failing to keep pace with lifespan gains, identifying who truly benefits from exercise interventions, and who does not, is becoming a pressing clinical question. By showing that nearly half of healthy older adults can display meaningful functional gains while their subjective wellbeing or structural metrics quietly deteriorate, the study makes a compelling case that multidomain outcome monitoring, rather than reliance on performance tests alone, may be essential for catching unfavourable trajectories early and adjusting exercise prescriptions before small setbacks become lasting ones.

Subject of Research: Multidomain clinical outcome changes and non-responder identification in healthy older adults after 12 weeks of resistance-based exercise training.

Article Title: Multidomain changes in clinical outcomes after 12 weeks of resistance-based exercise training in healthy older adults: a four‑domain framework for identifying non‑responders

Article References: Matko, Š., Strasser, B., Steinecker-Frohnwieser, B., Stamm, T., Kern, H., Löfler, S., Hochenwarter, S., Prüfer, F., Unger, A., Tirpáková, V., Cvecka, J., Šarabon, N., & Grote, V. (2026). Multidomain changes in clinical outcomes after 12 weeks of resistance-based exercise training in healthy older adults: a four‑domain framework for identifying non‑responders. European Geriatric Medicine. https://doi.org/10.1007/s41999-026-01614-8

Image Credits: AI Generated

DOI: 10.1007/s41999-026-01614-8

Keywords: healthy ageing, resistance training, sarcopenia, functional performance, multidomain assessment, rehabilitation outcomes, non-responders, health-related quality of life, body composition, muscle capacity, older adults, exercise intervention