Why Muscles Matter

Why a doctor cares about your muscle

Most people think of muscle as an aesthetic project with a health benefit attached. Clinically it is closer to the opposite.

Skeletal muscle is the largest organ in most bodies by mass. It is where the majority of glucose goes after a meal. It is the reserve the body draws on during illness, surgery and injury. And it is one of the few organs whose function you can measure in a clinic in under a minute, with a grip dynamometer, which is why researchers keep using it.

That measurability is the reason the evidence below exists at scale. You cannot easily follow half a million people's liver function. You can follow half a million people's grip.

So this piece is about what those measurements predicted. Not about how muscle looks.

The stroke study

This is the one that should change how you think about the word sarcopenia.

A prospective study examined 482,699 UK Biobank participants, testing whether sarcopenia status, grip strength and walking pace were associated with incident stroke and with death after stroke. Sarcopenia was assessed using the EWGSOP2 criteria, strokes were ascertained from health records, and risk was estimated with multivariable-adjusted Cox models.

Almost half a million people. Mean age 56, so not a study of the very old.

Participants with probable sarcopenia had a higher risk of any stroke, with an adjusted hazard ratio of 1.30 and a confidence interval from 1.21 to 1.39. For ischemic stroke it was 1.31, and for hemorrhagic stroke 1.41 with an interval from 1.20 to 1.67.

Three stroke types, three elevated risks, and the confidence intervals do not touch 1.

Grip strength tracked the same way: a hazard ratio of 1.07 per 5 kg lower absolute grip, and 1.36 per unit of lower relative grip. Slow walking pace carried a hazard ratio of 1.64 against a brisk pace.

A slow walk was associated with a 64 percent higher relative risk of stroke. That is a bigger number than most people expect from something so ordinary.

Among those who did have a stroke, probable or confirmed sarcopenia was associated with increased all-cause mortality afterwards. The authors also ran a two-sample Mendelian randomisation analysis using genetic instruments, which is the technique used to probe whether an association might be causal rather than confounded.

This is an observational cohort. The UK Biobank is healthier and less deprived than the general UK population, so absolute rates do not transfer. Walking pace was self-reported, and slow walking can be a marker of illness you already have rather than a cause of illness to come. Stroke risk is a matter for your doctor.

Muscle mass and dying of anything

The stroke finding is specific. This one is not.

A population-based cohort study analysed three cycles of NHANES, from 1999 to 2004, a nationally representative sample of US adults, using a predicted skeletal muscle mass index derived from the serum creatinine-to-cystatin C ratio. Analysis used multivariable Cox regression, restricted cubic splines and Kaplan-Meier survival curves.

Over a median follow-up of 193.2 months, with 2,217 deaths, higher predicted muscle mass was significantly associated with reduced all-cause mortality, with a hazard ratio of 0.76 and a confidence interval from 0.72 to 0.80.

Sixteen years of follow-up. And the relationship held for death from any cause, not one disease.

The shape matters as much as the size. The authors describe an L-shaped, nonlinear association with a distinct threshold effect, consistent across both sexes though with different inflection points.

L-shaped means the benefit is concentrated at the bottom. Going from very low muscle to moderate muscle is where the curve moves; going from a lot to more does not buy the same thing. That is a genuinely useful piece of information, and it argues against the idea that this is about being big.

This used a blood-marker estimate of muscle mass rather than a scan, which is a proxy and not a direct measurement. It is observational, so low muscle mass may be marking underlying illness rather than causing death. NHANES is a US sample.

Strength and weight are separate questions

There is a persistent assumption that body weight is the health variable and strength is a bonus. A large cohort tested them together.

The study used UK Biobank data from 2006 to 2010 to build a dynamic cohort of 200,405 adults aged 60 and over, examining weight status, handgrip strength, and their combined effect on premature mortality with multivariable Cox models, stratified by age, sex and age at death.

Over a median follow-up of 5,311 days there were 28,094 deaths from any cause and 5,172 premature deaths.

After adjustment, both weight category and muscular function were independently associated with premature mortality.

Independently. That word is the finding. Strength was not simply a proxy for being lean, and weight was not simply a proxy for being weak. They each carried information the other did not.

The weight numbers themselves are worth seeing, because they are not the ones people expect. Compared with normal weight, underweight individuals had a hazard ratio of 2.40, with a confidence interval from 1.79 to 3.23. Obese individuals had a hazard ratio of 1.07, interval 1.00 to 1.14.

The study also noted that obesity prevalence was 62.5 percent under the newer EASO classification framework against 24.9 percent by BMI, which tells you how much these conclusions depend on the definition being used.

This is observational, in adults aged 60 and over, in the UK Biobank, and the comparison between weight categories is sensitive to which classification framework is applied. It is not a statement about any individual, and body weight is a conversation for you and your clinician, not an article.

What actually builds it back

Association studies tell you muscle matters. They do not tell you that adding muscle helps. For that you need trials.

A systematic review and meta-analysis synthesised randomised controlled trials of physical exercise in older adults with sarcopenia, searching PubMed, Embase, CENTRAL, Web of Science, CINAHL, PEDro, SPORTDiscus and LILACS with no language or date restriction up to May 2025. Methodological quality was assessed with the PEDro scale and certainty of evidence graded with GRADE.

Eight databases, no language restriction, GRADE certainty ratings. This is a careful piece of work.

Seventy-two studies were included, of which 36 were classified as low risk of bias.

Exercise was superior to control for skeletal muscle mass, with a small effect size, and for muscle strength and physical performance, with moderate effect sizes. There was no significant effect on quality of life.

The honest reading of that: exercise reliably improved what it was measured on physically, and did not move a self-reported quality of life score. Reporting the null result is the point.

And the comparison that matters for anybody choosing what to do: resistance exercise was more effective than aerobic exercise for increasing skeletal muscle mass, with a large effect size.

That is the sentence behind every strength session in your Program. Walking is excellent and it is not the thing that rebuilds muscle mass.

This meta-analysis studied older adults who already had diagnosed sarcopenia, so effect sizes in a healthy 30-year-old are a different question. The literature was described by the authors as heterogeneous, and half the included studies were not low risk of bias.

What to take from this

Treat strength as a health measurement, not a vanity one. Grip strength and walking pace are being used as risk markers in cohorts of half a million people because they carry information.

Understand where the benefit sits. The mortality association was L-shaped, concentrated at the low end. This is an argument for not being weak, which is a different and far more achievable goal than being big.

Do resistance work specifically. In the pooled trials, resistance exercise beat aerobic exercise for muscle mass with a large effect size. Cardio is not a substitute for it, and this is why a Program has both.

Do not read a mortality hazard ratio as a personal prediction. These are population associations, mostly observational, and they cannot tell you anything about you.

These are my conclusions from the material. Nothing in this article is medical advice, and anything concerning stroke risk, body weight or an existing condition belongs with your doctor.