46% of 288 Older Adults Were High Responders for Knee Strength in 12-Month Trial

A 12-month training trial in Finland followed 288 older adults and found that training responsiveness differed sharply from one physical outcome to the next. Knee extension strength produced high responders in 46 percent of participants, while lean tissue barely moved at all.

Training responsiveness in later life is almost always reported as a single group average, and a study published in European Journal of Applied Physiology on October 24, 2026 set out to describe what those averages conceal in people past 70.

The analysis draws on the PASSWORD trial, registered as ISRCTN52388040 and run by researchers at the Faculty of Sport and Health Sciences at the University of Jyvaskyla in Finland. Community-dwelling adults aged 70 to 85 who were not meeting physical activity recommendations were recruited and followed over 12 months during 2017 to 2019.

A total of 314 participants entered the trial, with a mean age of 74.5 years. They were randomized to physical training alone or to physical training combined with cognitive training. The responder analysis reported here pooled 288 participants regardless of which arm they had been assigned to.

Grip strength, knee extension strength, maximal walking speed and six-minute walking distance all improved significantly across the year, each at a p value below 0.01. Appendicular lean mass did not change significantly. Beneath those group averages, the share of participants who cleared the bar for a meaningful individual gain ranged from 17 percent to 46 percent depending on the outcome examined.

How the Study Measured Training Responsiveness

Most exercise trials in older adults answer one question: did the group improve on average. This analysis asked a different one, and in doing so it treated training responsiveness as something to be counted rather than assumed. For each outcome, the researchers set a threshold called the smallest worthwhile change, defined as 0.2 times the between-subject standard deviation of the baseline scores, and then counted how many individuals cleared it.

That approach turns a single average into a distribution of training responsiveness. It is also the reason the paper can report training responsiveness as a percentage of people rather than a mean change in newtons or metres, which is closer to how a reader might think about their own odds.

Where Training Responsiveness Was Highest

Knee extension strength carried the highest training responsiveness in the trial. Adjusted for age and sex, the mean change was 40.1 newtons, with a 95 percent confidence interval of 34.5 to 45.7, and 46 percent of participants qualified as high responders.

Six-minute walking distance came next. The adjusted mean change was 36.7 metres, with a confidence interval of 31.9 to 41.5, and 44 percent of participants cleared the worthwhile-change threshold. For adults in their mid-seventies, both of these are the kinds of measures that track with getting out of a chair and crossing a street inside the signal, which is part of why training responsiveness in these two outcomes carries practical weight.

Grip strength and maximal walking speed improved significantly as a group but produced fewer individual high responders, at 27 percent and 30 percent respectively. The pattern suggests that training responsiveness is specific to what is being trained and measured, not a single trait a person either has or lacks.

The Measure That Barely Moved

Appendicular lean mass was the exception. It showed no significant change across the trial, and only 17 percent of participants met the high-response threshold, the lowest training responsiveness of any measure in the study. The authors note that even the association they examined between resting heart rate and lean mass responsiveness was weak.

This matters for how older adults read strength results and judge their own training responsiveness. Gains in what a muscle can do did not arrive alongside measurable gains in how much muscle tissue the scan detected, at least not over a year in this group. Function improved; tissue volume largely did not.

What Predicted a Strong Response

Four factors were associated with higher training responsiveness across two to five outcomes. Multimorbidity had the largest association, with an odds ratio of 2.60 and a confidence interval of 1.47 to 4.59. Younger age within the 70-to-85 band came in at 0.90, with an interval of 0.84 to 0.96, meaning each additional year lowered the odds slightly.

Lower body mass index was associated with an odds ratio of 0.94, interval 0.88 to 0.99, a modest shift in the odds. Higher adherence to the resistance training sessions gave an odds ratio of 1.03, interval 1.01 to 1.04 — a small per-unit effect, but the only predictor on the list that a participant can actually change.

Why Multimorbidity Pointing Upward Is Not a Paradox

At first reading, it seems odd that having more chronic conditions would raise the odds of a strong training response. The likeliest explanation is statistical rather than biological, and it has little to do with illness improving training responsiveness: participants starting from a lower baseline have more room to move before they hit a ceiling, and a fixed worthwhile-change threshold is easier to clear from further down.

The authors are careful here. They describe the participants as mainly having common chronic diseases that were well controlled, which is a different population from one with advanced or unstable disease. Reading this result as evidence that illness improves training responsiveness would overstate what a secondary analysis can support.

How It Sits Alongside Other Research on Older Adults

The direction of travel matches several recent findings. A 12-week trial in older women reported that elastic band training improved gait and strength without requiring gym machinery, another case of function responding before body composition does.

Population work points the same way. An analysis across five ageing cohorts found that becoming active after 50 was linked to markedly less functional decline, and a separate trial found that resistance training raised respiratory muscle strength in older adults. Each study measures something different, which is precisely the point this new analysis makes about training responsiveness in later life.

Comparisons with combined-modality work are harder. A network meta-analysis of exercise for metabolic syndrome ranked cardio plus lifting at the top for blood pressure, but it pooled group averages rather than counting individual responders, so the two bodies of evidence are not directly comparable.

What This Does and Does Not Mean for Training After 70

The useful reading is about expectations, not prescriptions. A trial reporting a significant group improvement does not promise that improvement to every participant, and this analysis puts rough numbers on that gap in training responsiveness. Depending on the outcome, between roughly one in six and nearly one in two participants saw a change large enough to count as meaningful for them individually.

It also suggests which measures are worth watching as markers of training responsiveness. Lower-body strength and walking distance moved most, so those are reasonable things to track over a year of training. Scale weight and body composition are poorer early signals of training responsiveness in this age group.

Some general points that follow from the design, none of them individualized advice:

• Attendance mattered measurably for training responsiveness, so consistency over a full year is more informative than intensity in any single month.
• Different outcomes improve on different timelines, so one flat measure is not proof of low training responsiveness overall.
• Anyone over 70 with chronic conditions or new symptoms should discuss an exercise plan with their own clinician before starting or changing it.