Two claims about strength usually arrive bundled together, and only one of them holds. The first is that strength is worth measuring on its own merits: it predicts hard outcomes, it is free, and it is available every session. That claim is well supported. The second is that strength tells you whether you are holding on to muscle during weight loss. That claim fails against direct evidence — in a randomised trial, dieting reduced lean mass measurably while measured strength did not move at all. This article separates the two, because the second one is what sends people into a maintenance phase believing they protected muscle they had already lost.
Proxy measure: something easy to measure, used to stand in for something that is hard to measure. A proxy is only worth as much as the link between the two, and this particular link is weaker than almost anyone assumes.
Why strength is worth measuring in its own right
Because low strength predicts the outcomes people actually care about.
The largest study of it followed 139,691 adults across 17 countries. Every 5 kg of lower grip strength was associated with a 16% higher relative risk of all-cause mortality (HR 1.16, 95% CI 1.13–1.20), a 17% higher risk of cardiovascular mortality and a 9% higher risk of stroke. In that cohort grip strength predicted all-cause and cardiovascular mortality more strongly than systolic blood pressure did (Leong 2015). The design is observational, so it shows association rather than cause, and squeezing a dynamometer harder is not a life-extension intervention.
The ageing literature points the same way and adds a comparison. Across cohorts, strength is lost two to five times faster than muscle mass, and — the sentence that matters — "loss of strength is a more consistent risk for disability and death than is loss of muscle mass" (Mitchell 2012). Both European and Asian sarcopenia consensus statements are built on that ordering: sarcopenia is defined by low strength first, and confirmed by low muscle quantity afterwards (Cruz-Jentoft 2019).
So strength earns its place as something to track. What it has not earned is the second job people give it.
Why strength cannot tell you whether you are keeping muscle
Because the two have been measured in the same people at the same time, and they came apart.
The cleanest demonstration is a randomised trial in 52 overweight, sedentary adults aged 45 to 65, who lost about 7% of body weight over roughly 17 weeks. In the diet-only arm, whole-body lean mass on DXA fell about 2% (P = 0.003) and lower-extremity lean mass about 4% (P < 0.0001). And in the same participants, over the same weeks: "No changes in muscle strength or bone were observed" (Weiss 2017). Lean mass fell measurably. Strength sat still. Anyone in that trial using strength as their muscle check would have concluded, wrongly, that nothing had been lost.
The observational evidence is larger and says the same thing from the other direction. In 1,880 older adults followed for three years, annualised leg-strength decline ran between 2.6% and 4.1% depending on sex and ethnic group — "about three times greater than the rates of loss of leg lean mass (approximately 1% per year)". More pointedly still, the participants who gained lean mass got no strength benefit for it (β coefficients: men −0.48 ± 4.61, p = .92; women −1.68 ± 3.57, p = .64) (Goodpaster 2006). Over five years in a related analysis of 1,678 adults, leg torque fell two to five times faster than leg muscle cross-sectional area — including in those who were losing weight — and weight gain did not prevent the torque loss despite a small increase in muscle size (Delmonico 2009).
Those are ageing cohorts rather than treated patients, and the Weiss trial used endurance exercise rather than resistance training, so none of them is a description of a person on a weight-management programme. What they establish together is narrower and firmer: strength and muscle mass are correlated when you compare different people, and dissociable within the same person over months. One cannot stand in for the other.
The practical rule follows directly. Track strength because losing it matters. Never present it, to yourself or to anyone else, as evidence that muscle is being preserved — and never let it replace a body-composition measurement.
Why the two come apart
Three separate mechanisms, and all of them are relevant to someone losing weight.
Muscle quality changes independently of muscle quantity. That is the conclusion the Health ABC investigators drew from their own data: the strength decline was "much more rapid than the concomitant loss of muscle mass, suggesting a decline in muscle quality" (Goodpaster 2006). The five-year analysis added a visible correlate — fat inside the muscle increased with age in everyone, in those who lost weight, gained weight or stayed the same (Delmonico 2009).
Early strength gains are largely nervous-system gains. In previously untrained people, neural factors account for the larger share of the initial strength increment, with muscle growth becoming the dominant contributor only after about three to five weeks (Moritani 1979), and modern reviewers still describe the first gains as primarily neural (Škarabot 2021). A beginner whose loads are climbing in month one is watching the nervous system learn the movement. That is a genuinely good sign about the training. It is not information about tissue.
Strength can rise while lean mass falls, in the same person, on treatment. In a single-arm cohort of 106 people with grade 3 obesity followed for a year on semaglutide, lean mass fell about 3.0 kg by month seven and then stabilised while fat loss continued — and mean handgrip strength rose 4.5 kg over the twelve months (Alissou 2026). With no comparison group that study cannot attribute anything to the medication, and grip is a strength measure rather than a mass measure. As an illustration of the dissociation it is hard to beat: the two numbers moved in opposite directions at the same time.
What the strength numbers mean in an Asian population
The cut-points people quote are usually European, and Singapore is not in that reference group.
The Asian consensus sets low muscle strength at handgrip under 28 kg for men and under 18 kg for women. Low physical performance is a 6-metre walk under 1.0 m/s, a Short Physical Performance Battery score of 9 or less, or a five-time chair stand taking 12 seconds or more. Its height-adjusted muscle-mass thresholds are 7.0 kg/m² for men and 5.4 kg/m² for women on DXA, and 7.0 and 5.7 kg/m² on bioimpedance (Chen 2019). The author panel spans Taiwan, Hong Kong, Thailand, Japan, Korea, China and Singapore.
Three things have to be said about those numbers before anyone measures themselves against them. They are clinical diagnostic thresholds for sarcopenia in older adults, not fitness benchmarks and not targets. The Asian algorithm requires confirmation by a muscle-mass measurement — low strength or low performance alone is explicitly labelled possible sarcopenia, which is a prompt for assessment rather than a diagnosis. And they differ numerically from the European consensus values (Cruz-Jentoft 2019), so quoting a European figure to a Singaporean patient is quoting the wrong population.
For context rather than diagnosis, age- and sex-specific reference percentiles now exist for Asian populations, pooled from 20 national cohorts across 12 countries including Singapore — 277,921 people for grip strength alone. The regional finding is the useful one: grip strength values were higher in East and West Asia than in South and Southeast Asia, so "Asian" is not one reference population either (Grgic 2026). Country-level percentile tables sit in the paper's supplementary material. The paper declares three Abbott Nutrition employees among its authors, stated as having had no access to individual data.
What to track, and what each measure is for
Keep both, and keep them in separate columns.
Strength, every session. The loads and repetitions in a training log are the cheapest longitudinal data anyone will ever collect, and they answer a real question: is the capacity that predicts disability and death holding up. For a clinical reading of the same thing, grip strength, gait speed and a five-time chair stand are the measures the Asian consensus uses (Chen 2019).
Body composition, on one device, a few times a year. This is the measure that answers the muscle question, and nothing else does. Across fifteen bioimpedance devices tested against a four-compartment reference model, only five met a strict ±2% standard for a single reading — yet nine met a tighter ±1% standard for tracking change across a 12 to 16-week window (Siedler 2022). Which device matters less than using the same one every time. Our guide to InBody, DEXA and smart scales works through the interval, and why the scale misleads covers what body weight can and cannot report.
Training, regardless of what either number says. The reason to hold both measures at once is that the intervention is the same either way: across 114 trials covering 4,184 people, lean mass was statistically unchanged where resistance training accompanied caloric restriction (Lopez 2022).
At GetLean, our philosophy is that the medication is the catalyst and what you keep is the result — and a measure that cannot detect what you are keeping should not be asked to. Individual results vary, and clinical-trial figures describe the populations that were studied.
Common questions
Does keeping my strength mean I am keeping muscle?
No, and this is the single most common error in the area. In a randomised trial, about 7% weight loss by dieting alone reduced whole-body lean mass by around 2% and leg lean mass by around 4% on DXA, while measured muscle strength did not change at all (Weiss 2017). Strength holding steady is compatible with lean mass falling, so it cannot be used to check whether muscle is being preserved.
Is grip strength a good measure of muscle mass?
Grip strength is a strength measure, not a mass measure. It has a strong prognostic record of its own — in 139,691 adults across 17 countries, every 5 kg of lower grip strength was associated with a 16% higher relative risk of all-cause mortality (Leong 2015) — but that association is observational, and it tells you nothing about how much muscle a particular person is carrying.
What is a normal grip strength for an Asian adult?
There is no single normal figure, and the numbers that circulate are usually European. The Asian consensus on sarcopenia sets its low-strength threshold at under 28 kg for men and under 18 kg for women (Chen 2019), which is a clinical diagnostic threshold for older adults rather than a fitness target. Age- and sex-specific Asian reference percentiles drawn from 20 national cohorts including Singapore now exist, and they show grip strength differs across Asian regions rather than being uniform (Grgic 2026).
If strength is not a proxy for muscle, why track it at all?
Because losing it matters in its own right. Across ageing cohorts, strength is lost two to five times faster than muscle mass, and strength loss is the more consistent predictor of disability and death (Mitchell 2012). It is also the only measure available at every single session, at no cost.
What should I use instead to check whether I am losing muscle?
A body-composition measurement, repeated on the same device. Across fifteen bioimpedance devices tested against a four-compartment reference model, only five met a strict accuracy standard for a single reading, but nine tracked change acceptably over a 12 to 16-week window (Siedler 2022). Consistency of device matters more than which device it is, and no performance measure substitutes for the scan.