People with low lean mass had roughly a 30% higher risk of dying from any cause over the follow-up period than people with normal lean mass, according to a 2025 dose-response meta-analysis of more than 130,000 middle-aged and older adults published in Frontiers in Medicine by Li and colleagues.1 The same paper reported that every additional kilogram of lean mass was associated with about a 1% lower mortality risk, an effect that held after adjustment for age, sex, and a long list of usual suspects.
That is a striking finding, and it is one in a growing pile. Earlier work on muscular strength, summarized across roughly two million people in 2018, pointed in the same direction: stronger adults outlive weaker ones, even after researchers control for body weight and disease.2 Muscle, in other words, is not just a vanity tissue.
What did the new meta-analysis actually find?
The Li paper pooled prospective cohort studies that measured lean body mass (the non-fat part of body weight, dominated by skeletal muscle) and tracked participants for years. People in the lowest lean-mass categories had a hazard ratio for all-cause mortality of around 1.30 compared with the reference group.1 Translating that into plain English: for every 100 people with normal lean mass who died during follow-up, roughly 130 people with the lowest lean mass died over the same window.
The dose-response part is the interesting bit. Rather than only comparing low versus high categories, the authors fit a curve. The relationship was close to linear in the middle range, which is why a per-kilogram estimate is even meaningful. A 1% reduction in mortality risk per extra kilogram of lean mass is small at the level of an individual kilogram and large at the level of, say, a 10 kg difference between two people of similar height.1
Two caveats matter. First, this is observational data. The authors cannot prove that adding muscle causes a longer life; only that the two travel together in large populations. Second, “lean mass” in these studies is usually measured by DXA or bioelectrical impedance, both of which capture more than just muscle. Bone, organs, and fluid all count. The signal is muscle-dominated, but it is not pure.
Why would muscle protect you?
The mechanistic story is unglamorous and convincing. Skeletal muscle is the body’s largest reservoir of glucose and amino acids. After a meal, healthy muscle hoovers up roughly 80% of the glucose dose under insulin’s direction. When muscle mass shrinks or its insulin signaling dulls, glucose lingers in the blood, and the pancreas works harder. Over years, that pattern is one of the engines of type 2 diabetes.
Muscle also contracts. That sounds obvious until you notice what contraction does at the system level. It pumps blood back from the legs, supports the spine, catches you when you trip. The Li authors argue, plausibly, that some of the mortality signal flows through falls and fractures, which kill far more older adults than people realize.1 A hip fracture at 80 carries a one-year mortality of roughly one in four. Stronger legs make hip fractures less likely, both because falls happen less often and because the falls that do happen are softer.
Then there is the endocrine angle. Muscle behaves like an organ that releases its own signaling molecules, called myokines, during exercise. Some of these, including interleukin-6 in its acute exercise form and irisin, appear to have anti-inflammatory and metabolic effects. The picture is still being worked out, and not every claim about myokines holds up. But the general direction, that contracting muscle talks to the rest of the body in helpful ways, is well established.
It is not just one study
If the Li meta-analysis stood alone, it would be a curiosity. It does not. García-Hermoso and colleagues, working with data from roughly two million adults, found that low muscular strength predicted all-cause mortality even in apparently healthy populations.2 Their estimate for the lowest-strength versus highest-strength groups was a hazard ratio of about 1.51, meaning roughly a 50% higher risk of death over follow-up, after adjustment for the usual confounders.
That paper relied heavily on grip strength and leg strength as proxies. Grip strength is a humble bedside measurement, a squeeze on a hand dynamometer, and it has turned out to be one of the more durable predictors in the modern literature. It tracks roughly with whole-body strength, and it picks up the same signal as more elaborate tests. If your grip is weak for your age, that is information.
The convergence between the two papers matters. Mass and strength are related but not identical. You can have decent muscle bulk and weak muscle (sometimes called dynapenia), or modest bulk and surprising strength. Both meta-analyses point at the same outcome: less of either is worse for survival, more of either is better.1,2
Does training actually help, or only correlate?
This is the question that decides whether any of this is actionable. Two large reviews say yes, training itself, not just being naturally muscular, is associated with lower mortality.
Shailendra and colleagues pooled studies of resistance training and mortality in 2022, covering hundreds of thousands of participants, and found that adults who lifted weights regularly had roughly a 15% lower all-cause mortality risk than those who did not, with a sweet spot around 30 to 60 minutes of resistance training per week.3 Past about an hour a week, the curve flattened. More was not obviously better in this dataset.
Saeidifard and colleagues, in 2019, ran a similar exercise on a slightly different study set and found that any reported resistance training was associated with about a 21% lower all-cause mortality risk compared with no resistance training, and that combining resistance training with aerobic activity produced the lowest risk of all.4 Aerobic and resistance work appear to be complements, not substitutes.
Neither paper is a randomized trial of “lift weights for 20 years and we will see who dies.” Such a trial will never exist. But the dose-response shape, the consistency across cohorts, and the biological plausibility of the underlying mechanism make this one of the more credible exercise findings in epidemiology.
How much muscle do you actually need?
There is no single threshold. The Li paper used relative cutoffs based on appendicular lean mass divided by height squared, the same approach used in clinical sarcopenia definitions.1 A common cutoff for low lean mass is below about 7.0 kg/m² in men and below about 5.5 kg/m² in women, though the exact numbers vary by population and equipment.
For most readers, that is more precision than is useful. A simpler rule of thumb works: if you can stand up from a chair without using your arms, climb a flight of stairs without holding the rail, and carry your own groceries, you are probably above the danger zone for your age. If any of those is getting harder year over year, that is the signal to act, not the moment to wait for a clinical diagnosis.
It is also worth knowing that lean mass declines naturally from roughly age 30 onward, at about 3 to 8% per decade in sedentary adults, and faster after 60. That decline is not destiny. It is the average outcome in a population that mostly sits.
What the evidence says about training
The strongest practical takeaway from the resistance-training meta-analyses is that the dose is small. Two sessions a week, of perhaps 30 to 45 minutes each, covering the major movement patterns (a squat or sit-to-stand variation, a hinge or deadlift variation, a push, a pull, and a carry), captures most of the benefit Shailendra and colleagues observed.3 You do not need a gym membership; bands and bodyweight can carry beginners a long way.
Protein matters too, though it is the support cast, not the lead. Most older adults under-eat protein relative to their needs, and the literature converges on roughly 1.0 to 1.2 grams per kilogram of body weight per day for healthy older adults, with higher amounts (1.2 to 1.6 g/kg) discussed for those who are training hard or recovering from illness. Spreading that across three meals appears to help, because the muscle’s ability to build new protein in response to a single feeding is capped.
Sleep, stress, and basic medical care are the silent partners. A person sleeping six broken hours a night, drinking heavily, and skipping meals will struggle to build muscle no matter how good the program looks on paper. None of those facts are exciting, and none of them sell supplements, which is roughly why they show up at the bottom of articles instead of the top.
Who has the most to gain?
The mortality curves in the Li paper are flatter for younger adults and steeper for older ones.1 A 35-year-old with somewhat low lean mass has time to course-correct without much drama. A 70-year-old in the same percentile is closer to clinical sarcopenia and the cliff that follows it.
That said, two groups are quietly under-served by the standard “go to the gym” advice. The first is women in midlife. Estrogen loss around menopause accelerates the loss of lean mass and bone density, and the women who pick up resistance training in their fifties tend to do unusually well on follow-up measures. The second is men over 65 who have always been “thin and active” and assume they are fine. Walking and cycling, while excellent in their own right, do little for upper-body strength or lean mass. The pattern of low body weight plus low muscle (sometimes called sarcopenic obesity’s mirror image) is its own risk profile.
The original Facebook post that prompted this article put it cleanly: “Protect your muscle. Your future self may thank you.” That is a fair summary of what the meta-analyses suggest, with one footnote. The future self most likely to thank you is the one in their seventies and eighties, doing things they assumed they would have given up by then.
Common questions about muscle mass and longevity
How accurate is the 30% number?
That figure comes from the pooled hazard ratio in the Li 2025 meta-analysis comparing low to normal lean mass, after adjustment for common confounders.1 It is a population-level estimate, not a personal forecast. The confidence intervals are reasonably tight, but individual outcomes vary widely.
Can older adults still build muscle?
Yes, and the response to training is larger than people often expect. Untrained adults in their seventies routinely add measurable lean mass and considerable strength in 12-week programs. The first eight weeks of any program also produce neural gains, where existing muscle gets better at firing, before visible size changes appear.
Is cardio bad for muscle?
No. The Saeidifard analysis found that combining aerobic activity with resistance training had the lowest mortality risk of all the patterns examined.4 Cardio and lifting are complements. Only at extreme volumes of endurance work does interference with muscle growth become a real concern.
Do supplements help?
Creatine monohydrate has the strongest evidence for adding small amounts of muscle and strength on top of training, and protein powder is a convenience food, not a magic substance. Most other muscle-building supplements have weak or contested evidence.
What if I have a chronic illness?
The risk-benefit math usually still favors some form of resistance work, but the program needs adapting. A primary care visit and, where available, a session with a physical therapist or qualified trainer who has worked with your condition is worth its weight in months of trial and error.
The honest summary
The new meta-analysis does not say that lifting weights guarantees a longer life, and a careful reading of the literature does not support that claim either. What it says is narrower and more durable: across hundreds of thousands of adults, having less muscle, or being weaker for your age, tracks with dying earlier, and the relationship looks roughly linear in the middle of the distribution.1,2
The interventional evidence, while observational rather than randomized, points in the same direction and suggests that the dose required to capture most of the benefit is modest.3,4 Two short sessions a week, enough protein, enough sleep, and a long enough horizon to let the adaptations compound. That is not a slogan. It is, more or less, what the data look like.
Sources
- Li J, Liu X, Yang Q, Huang W, Nie Z, Wang Y. Low lean mass and all-cause mortality risk in the middle-aged and older population: a dose-response meta-analysis of prospective cohort studies. Frontiers in Medicine. 2025. PubMed: 40636391
- García-Hermoso A, Cavero-Redondo I, Ramírez-Vélez R, Ruiz JR, Ortega FB, Lee DC, Martínez-Vizcaíno V. Muscular Strength as a Predictor of All-Cause Mortality in an Apparently Healthy Population: A Systematic Review and Meta-Analysis. Archives of Physical Medicine and Rehabilitation. 2018. PubMed: 29425700
- Shailendra P, Baldock KL, Li LSK, Bennie JA, Boyle T. Resistance Training and Mortality Risk: A Systematic Review and Meta-Analysis. American Journal of Preventive Medicine. 2022. PubMed: 35599175
- Saeidifard F, Medina-Inojosa JR, West CP, Olson TP, Somers VK, Bonikowske AR, Prokop LJ, Vinciguerra M, Lopez-Jimenez F. The association of resistance training with mortality: A systematic review and meta-analysis. European Journal of Preventive Cardiology. 2019. PubMed: 31104484
