What Your Grip Test Can’t Tell You About Exercise

The grip test said exercise barely helped.

Across seventeen studies and 646 older adults with diagnosed sarcopenia, resistance training improved handgrip strength by 2.30 kilograms. That sounds like progress until you learn the clinical threshold — the amount doctors consider meaningful — is 5.0 kilograms. The grip test reached 46% of the bar that matters.

Then the researchers measured their legs.

Knee extension strength — the force that gets you out of a chair, up a flight of stairs, across a parking lot before your knees give — showed a large effect. Eleven studies, 436 people, an improvement that cleared every significance bar by a wide margin. Two tests, pooled from different sets of trials. Two very different answers to whether exercise is working.

That split sits at the center of a 2025 meta-analysis by Yan and colleagues in Aging Clinical and Experimental Research.

They analyzed 24 randomized controlled trials with 951 sarcopenic adults — mean age 73, mostly women. They asked a question that sounds simple until the data fractures: does resistance training work for age-related muscle loss? The answer depends entirely on what you measure.

Two Jobs, One Tool

Grip strength deserves its reputation. It predicts mortality, frailty, and functional decline in older adults — people with weak grip have a 79% higher risk of dying earlier, according to large-scale prognostic data. As a screening tool for sarcopenia, grip is fast, cheap, and remarkably reliable.

The problem is that finding a disease and tracking whether treatment works are two different jobs — and the grip test is being asked to do both.

The pattern held across the other measures. Grip, gait speed, the timed up-and-go and the five-times sit-to-stand all improved statistically, while the short physical performance battery did not — and none crossed the threshold that counts as clinically meaningful. Muscle mass, overall, did not rise significantly. The diagnostic measures that doctors check showed modest changes at best.

Then there was knee extension. The force that determines independence — stairs, chairs, parking lots. An effect size of 1.04, a large one. The researchers themselves noted that lower-limb measures showed "larger and more consistent improvements" and may be better markers of training progress.

The test your doctor uses to find the problem showed only limited response to the treatment. The grip dynamometer excels at its first job. It is limited at its second. A disappointing grip result after months of training does not mean the exercise failed. The authors' explanation is that most of the programmes did not train grip directly.

The Upgrade Nobody Can See

The mirror told a quieter story than the leg test. Pooled across ten studies, the muscle on arms and legs — the tissue that doctors measure with body scans — came out 0.02 kilograms ahead of usual care, a result the researchers reported as no effect. The muscle mass index, pooled across seven studies, was 0.26 kg/m² higher, short of significance, with results that varied widely from trial to trial.

One subgroup broke the pattern. In the trials that used constant resistance, the muscle mass index rose significantly. The authors' clinical advice follows from it: constant resistance "may be more suitable when the focus is on gains in muscle strength and mass."

Yet the legs got much stronger.

The paper does not say why. It measured no nerves, motor units or coordination, and it discusses those only through other studies. Its authors also flag the weak spot themselves: the number of studies that measured muscle mass was limited.

This distinction — muscle quality versus muscle quantity — aligns with a shift in how researchers think about treatment success. The traditional measure asked: did muscle mass increase? The newer question asks: did muscle function improve? For the older adults in these trials, the pooled answer to the first question was no, with one subgroup as the exception. The answer to the second was yes, though mostly by less than a clinically meaningful amount.

Most of the trials ran for 26 weeks or less, which the authors say limits how far their dose findings generalise. The paper does not report protein intake.

A meta-analysis of 20 trials in adults over 50 identified a specific daily threshold — 1.0 grams of protein per kilogram of body weight — above which dieters preserved significantly more lean mass, with the benefit 38% larger in older adults than younger ones.

What is clear is that strength moved further than size in these trials, and that the reason is still open.

Add Saturday

Of the variables the researchers tested — load, duration, type of resistance, training setting, frequency — they single out two as key: the type of resistance, and how often. Heavier loads did no better than lighter ones.

In trials where older adults trained three times per week, grip strength improved by 3.18 kilograms more than with usual care. In trials with two sessions a week, it improved by 1.42 kilograms. These were different trials, not the same people adding a day, so the gap is a comparison between programmes: the three-session trials gained more than twice as much grip strength — a ratio of 2.24 to 1.

The gap between frequency groups was significant. A regression across the trials linked more frequent training to bigger gains in grip and leg strength, and for grip, frequency accounted for 60% of the differences between trials' results. Age, session length and programme length showed no such link. The abstract is more cautious, saying the meta-regression found no significant sources of heterogeneity.

Current guidelines recommend resistance training "at least two days per week" for older adults. That recommendation was built from studies of healthy populations. In those populations, one, two, and three sessions per week often produce comparable results.

For older adults with muscle loss — the population that needs resistance training most — trials at the guideline minimum gained less than half the grip strength of three-session trials. The prescription for healthy older adults and the prescription for those actively losing muscle may not be the same prescription.

The framing matters here. This is not a finding about what someone did wrong by training twice a week. The existing effort has value — the two-session trials produced measurable improvement. It is a finding about what went with one more session: programmes with a third weekly session gained more than twice the grip strength of two-session ones.

Forty Minutes, Three Sessions

The researchers went beyond frequency. They modelled the relationship between weekly training dose and outcomes — and the curve was not a straight line.

In that model, handgrip strength gains climbed with weekly dose up to a point, then began to decline. Past that point, more was not better. For walking speed, the modelled benefits kept climbing with no plateau. Clinically meaningful gains in walking speed may come above about 600 MET-minutes a week.

From that curve the authors suggest a place to start rather than a tested programme: about 120 minutes a week of moderate-intensity training in three sessions, roughly 40 minutes each. Not a ceiling — a floor, increased gradually as tolerated.

The type of resistance training mattered too, but differently depending on the outcome. Trials using constant resistance showed greater knee extension strength gains than trials combining constant and variable resistance. That combination, in turn, was best for walking speed.

The authors' own advice follows the same split: mixed training when the goal is function, constant resistance when it is strength and muscle mass. Both are valid starting points.

What the Safety Reports Showed

Four of the 24 studies reported adverse events. All were mild and reversible. Most were transient joint discomfort, muscle soreness or fatigue at the start of training, which typically resolved once intensity or technique was adjusted.

None of those events led anyone to withdraw. The paper's safety summary covers those four trials; it does not describe side effects in the other 20.

The authors read this as support for the safety and tolerability of resistance training in older adults with sarcopenia. The paper does not test whether more training, more often, changes that.

A separate 2025 review by Chaabène and colleagues examined whether the type of contraction matters — eccentric versus traditional resistance training in older adults.

Across 11 trials and 451 participants, eccentric training showed a small edge in raw strength. But there was no meaningful difference in functional capacity, power, or muscle growth. The distinction was modest enough that both approaches produced significant within-group improvements.

The finding reinforces a pattern: the type of contraction matters less than showing up and steadily challenging the muscles. [1]

The question that remains is not whether resistance training works for sarcopenia — the authors call the evidence robust that it probably improves strength and physical function, though mostly by less than a clinically meaningful amount. The question is how much to do in each session. This meta-analysis singled out frequency and resistance type as key variables and suggested a starting dose.

But the optimal number of sets per workout — the volume within each session — draws on a different body of evidence. That evidence comes from a separate meta-analysis, with a finding that challenges the assumption that more sets always mean more growth.