Muscle Power Leaves Before Strength Does

There is a dynamometer in your kitchen, and you have been sitting in it for years.

The test is unglamorous enough to be insulting. Sit in an ordinary chair, cross your arms over your chest, and stand up and sit down as many times as you can in thirty seconds. No equipment, no lab, no subscription. Researchers in Toledo, Spain have been running a version of this on thousands of older adults, and when they turned the repetition counts into watts and looked at what happened to people afterwards, the chair turned out to know things.

What interests me is not that the test predicts something. Plenty of cheap tests do. It is that the quantity it measures — power — is the one the fitness culture almost never talks about, while the quantity we all talk about, strength, is the one that holds up longest.

Power falls at roughly twice the rate of force

Muscle strength is how much force you can produce. Muscle power is force multiplied by the speed you produce it at — the physics of catching yourself on a stair, not the physics of a heavy shrug. They sound like the same virtue described twice. They age completely differently.

A Belgian group followed 489 adults for a decade, testing knee extension at baseline and again ten years later, and separated the force component from the velocity component instead of collapsing both into one score. In the older adults, maximum power fell by about 2.2–2.4% per year. Force, in the same people, fell by about 1.09–1.27%. Power was draining out at roughly double the rate of the thing most of us train.

The composition of the loss also shifted. In middle age, the early dent in power came mostly from force. Later, velocity joined in — a progressively blunted ability to produce force quickly. Which means two people can post the same one-rep max and be in genuinely different condition, because one of them can still express it in under a second and the other cannot.

Why the grip test misses it

We have written admiringly about grip strength, and I still think it earns its place as a five-second proxy for how the whole body is holding up. But it is worth reporting the study that argues against us.

Researchers looked at older adults around the age of ninety and asked which measurement best explained who was still physically independent. They put sit-to-stand power, handgrip strength, physical activity and sedentary time into the same model. Sit-to-stand power carried a standardised coefficient of 0.45. Handgrip came in at −0.02 — statistically indistinguishable from contributing nothing once power was in the room. Sedentary time (−0.28) mattered more than grip did.

That is a cross-sectional analysis in a cohort near ninety, so it describes a snapshot rather than a sequence, and grip may well be measuring something different and still useful earlier in life. But the direction is uncomfortable and worth sitting with: a squeeze is a slow, isometric act. It samples force with the velocity term set to zero. It is, almost by design, blind to the thing that erodes first.

What low power travels with — in three honest tiers

The evidence here is real but stacked, and the tiers are not equally strong. Worth separating them rather than quoting the scariest number.

Cross-sectional, so the largest numbers. Working from 1,475 adults aged 65–98 in Toledo, researchers derived cut-offs for low relative sit-to-stand power: below 2.53 W/kg for men, below 2.01 W/kg for women. People under those lines were more likely to be frail (odds ratio 4.4 in men, 5.2 in women) and to have slow habitual gait speed (3.4 and 6.1). Large odds — but measured at the same moment, so they cannot tell you which came first.

Prospective, so smaller and more believable. Following 839 of the same adults for about five years, low power was associated with later development of frailty (OR 2.51), difficulty with basic daily activities (1.70), difficulty with instrumental ones (1.79), and increased medication use (1.51). Note what did not reach significance: cognitive impairment (1.38), depression (1.29), and frailty measured by a second definition. The signal is about physical function specifically. It is not a general index of everything.

Mortality. In the 1,876-person analysis, men below the power threshold had a 57% higher risk of death from any cause and women more than double the risk over roughly nine years. Falls and fractures also clustered below the line — but those were history, recorded at the same visit as the test, which is a different and weaker kind of finding than the prospective ones. More than half the cohort sat below the cut-off for their sex.

All of it is association. The most obvious rival explanation is that early illness makes people slower out of a chair, rather than slowness making them ill, and no observational study of this design can rule that out. The five-year study also lost its frailer participants to attrition, which biases toward the healthy. Read these numbers as a signal about the body's current condition, not as a forecast.

The velocity term is trainable, which is the whole point

A risk marker you can't move is trivia. This one moves.

A meta-analysis pooled fourteen trials of high-speed resistance training in independent older adults aged 69–82 — 408 participants randomised. The protocols were remarkably consistent: loads of 40–80% of one-rep max, the lifting phase performed as fast as possible (under a second), the lowering phase controlled over two to three seconds, three sessions a week for eight to sixteen weeks. Pooled effect on neuromuscular function was moderate (SMD 0.70, 95% CI 0.24–1.15); on physical function, 0.55. The authors concluded that ten weeks or more, three times a week, is where the reliable gains showed up.

Honesty about that evidence: twelve of the fourteen studies were rated low methodological quality, only one was pre-registered, the samples were small, heterogeneity was moderate to substantial, and nobody followed participants long enough to know whether the gains stick. This is a promising and coherent literature, not a settled one.

Intent is what you change, not the weight

Here is the part that surprised me when I first understood it, and the part worth taking into this week's training.

Power training is not a different exercise list. It is the same squat, press and row performed with a different intention: move the load fast on the way up, control it on the way down. The load itself sits in a moderate range — 40–80% of your maximum, not the grinding heavy end, because you cannot accelerate something you can barely lift. If your gym time is a slow, controlled, respectable grind through every repetition, you are training force and leaving velocity untouched.

The everyday version needs no gym at all. Stand up out of the chair fast; sit back down slowly. Take the stairs two at a time where it is safe. Carry the shopping and set it down deliberately. These sound like nothing, and they are structurally the same instruction the trials gave: brief, deliberate, fast concentric effort, repeated often enough to matter.

One caution the enthusiasm tends to skip. Fast movement is a skill before it is a stimulus, and connective tissue adapts on a slower clock than muscle does. If you are coming back from a long gap, add speed gradually and to movements you already own, and if you have joint problems, a balance issue, or any cardiac history, get advice from a clinician before you start accelerating loads.

Measure the trend, not the digit

The reason this test appeals to me is the same reason we keep arguing that a sensor is only as good as the question you ask it. A repetition count is crude. It is also yours, repeatable, free, and sensitive to something that is genuinely changing underneath.

Do it once and you have a number of no particular meaning. Do it in February, May and September under the same conditions and you have a trend line for a physical capacity that most people never look at until it announces itself in a stumble. That is the same logic behind reading your resting heart rate and HRV as drift against your own baseline rather than against a population chart, and behind treating a longevity protocol as a loop — measure, act, re-measure — instead of a list.

Use numbers to correct fantasy, not to replace experience. The fantasy here is that strength is the whole story and that a respectable deadlift means the machinery is intact. The experience is standing up from a low sofa and noticing, for the first time, that you thought about it first.

The bottom line

Power — force expressed quickly — declines about twice as fast as force does, starts earlier than most people assume, and appears to track physical independence better than the grip test we all reach for. Low sit-to-stand power is associated with frailty, functional difficulty and higher mortality in adults over 65, though the evidence is observational and the opposite explanation cannot be excluded. The trainable variable is not the weight but the intent: moderate loads moved fast on the way up, controlled on the way down, three times a week, judged over months. Test yourself in a chair a few times a year, and read the direction rather than the digit. If you are tracking training load and recovery alongside it, the Agen Band and the physical performance range are there for the parts a chair can't see — but the chair remains the cheapest honest instrument in the house.