Key Takeaways
  • Muscle loss after 40 is real, slow, and mostly in the legs — and strength falls two to five times faster than size. Strength, not size, is what predicts disability and death
  • Training works at every age studied. Over-50s gained an average 1.1 kg of lean mass across 49 studies; 90-year-olds raised their strength 174% in eight weeks. The response shrinks a little with age — a reason to start, not to wait
  • 30–60 minutes of strength training a week was associated with 10–20% lower mortality — and combined with aerobic exercise, 40% lower
  • The dials: two sessions a week per muscle beats one; more sets help up to roughly 10 a week; you don't have to go to failure; light loads build size but heavy loads build strength and bone; cardio doesn't blunt your gains
  • Protein: about 1.6 g/kg a day, and more per meal than a young person needs — around 0.4 g/kg — spread over three or four meals
  • Creatine is the one supplement with consistent evidence at this age: +1.37 kg lean mass in 22 studies. Testosterone is a treatment for a diagnosis, not a training aid
  • The plan at the end is two or three 45-minute sessions a week. It's enough

Somewhere in your forties the maths changes. The recovery from a heavy weekend takes longer, the jeans that fit at 35 fit differently, and the conventional wisdom arrives with a shrug: you lose muscle as you age, that's just how it is. The first half of that sentence is true. The second half is the part the research disagrees with — loudly, for decades, in people from their forties to their nineties.

This is the long version: what actually happens to muscle as you age, how much of it training reverses, and what the evidence says about each of the dials you can turn — how many sets, how heavy, how often, whether to go to failure, how much protein, which supplement, and what to make of hormones. It's written for men and women, with the places the evidence differs by sex called out. Every figure is named to the study it came from, and where the study was done in young people rather than people over 40, that's said too, because it happens more often than the fitness industry admits. At the end there's a plan — two versions — built from all of it.

174%
Average strength gain in eight weeks of high-intensity training among frail nursing-home residents averaging 90 years old — mid-thigh muscle area rose 9% (Fiatarone, JAMA 1990; 9 completers, no control group)
Part one · What actually happens

The Decline Is Real, Slow, and Mostly Strength

Evidence: Strong — MRI cohorts and longitudinal studies

The best picture of how muscle changes across a lifetime comes from Janssen and colleagues, who measured skeletal muscle by whole-body MRI in 468 men and women aged 18 to 88 (Journal of Applied Physiology, 2000). Muscle as a share of body weight starts falling in the twenties — largely because body weight rises around it, not because muscle is being lost yet. Absolute muscle mass, though, held steady for decades and only showed a noticeable decline from the end of the fifth decade — the late forties — and that decline was mainly in the lower body. The legs go first. That single finding explains a lot of what people notice in their fifties: the stairs, the getting up off the floor, the first time a hike hurts for three days.

How fast is it? Mitchell and colleagues pooled the longitudinal studies in Frontiers in Physiology (2012). At age 75, muscle mass was being lost at about 0.64–0.70% a year in women and 0.80–0.98% a year in men. Strength, at the same age, was falling at 2.5–3% a year in women and 3–4% a year in men. Two findings in that review matter more than the rates themselves. Strength was lost two to five times faster than muscle mass in studies that measured both. And loss of strength was a more consistent risk factor for disability and death than loss of mass. This is why the rest of this article keeps saying "strength" where you might expect "muscle": the size is the visible part, but the function is what you're protecting.

For women, there's a transition inside the trend. The SWAN study followed 1,246 women through menopause (Greendale, JCI Insight 2019). At the start of the menopause transition, the rate of fat gain doubled and lean mass began to decline — by about 0.2% a year in the full-text figures — continuing until about two years after the final period, after which both levelled off. The lean-mass loss is modest; the shift in the fat-to-muscle balance is what women tend to notice, and it's the window where training has the most to protect. The women's longevity pyramid covers the bone side of the same years.

Why it matters: strength is a mortality variable

This isn't vanity. Momma and colleagues' meta-analysis of 16 cohort studies in the British Journal of Sports Medicine (2022) found muscle-strengthening activity associated with 10–17% lower risk of all-cause mortality, cardiovascular disease, total cancer, diabetes and lung cancer. The largest reduction in mortality — about 10–20% — came at roughly 30 to 60 minutes a week. Combined with aerobic activity, the association was far larger: a relative risk of 0.60 for all-cause mortality, 0.54 for cardiovascular death. Two caveats the authors are explicit about: these are observational associations, and the curve was J-shaped — beyond about an hour a week, the benefit for mortality didn't clearly keep growing. That's good news for anyone short of time. The dose that matters is two or three sessions, not a second career.

Grip strength, the cheapest proxy for total strength, tells the same story in larger numbers — in UK Biobank's half-million participants, every 5 kg of lost grip carried 16–20% higher mortality. The grip strength piece has the test.

Part two · Does training still work?

Yes — at 50, at 65, and at 90

Evidence: Strong — meta-analyses of randomised trials in older adults

Peterson and colleagues pooled 49 studies of resistance training in adults over 50 — 1,328 participants — in Medicine & Science in Sports & Exercise (2011). The average gain in lean body mass was 1.1 kg. Two moderators came out significant: higher-volume programmes produced larger gains, and older participants gained somewhat less. The authors' own summary is the one to remember — resistance exercise is effective for building lean mass in ageing adults, and starting earlier in life may make it more effective. It isn't too late at 50. It is later than 45.

For people over 65, Borde and colleagues' meta-analysis of 25 randomised trials in Sports Medicine (2015) found large effects on strength and smaller but real effects on muscle size, and — unusually — tried to work out which training parameters did the most. The largest strength gains came with programmes lasting 50–53 weeks, at 70–79% of one-rep maximum, two to three sets per exercise, seven to nine repetitions, twice a week. The authors call their findings preliminary (the included trials were small and varied), but the prescription matches almost everything else in this article: moderate-to-heavy, a few sets, twice a week, for a year rather than a month.

And at the far end of the age range, the paper that changed the field: Fiatarone and colleagues took ten frail nursing-home residents averaging 90 years old and put them through eight weeks of high-intensity leg training (JAMA, 1990). In the nine who finished, strength rose by an average of 174%, mid-thigh muscle area by 9%, and tandem walking speed by 48%. It's a tiny study with no control group, and its job in this article isn't to give you a number — it's to retire the idea that there's an age past which muscle stops responding. There isn't one in the data.

Part three · The training dials

Volume, Load, Frequency, Failure — One at a Time

Evidence: Moderate — good meta-analyses, but many in young adults

The rest of the fitness internet argues about these dials endlessly. The meta-analyses are calmer. Here's what each one says, with a note wherever the evidence comes from twenty-somethings rather than people your age.

Frequency2× a week per muscleBeats once a week at the same total volume.
Volume~10 sets/week per muscleMore sets, more growth — with diminishing returns.
Load70–85% 1RM for main liftsLight works for size if near failure; heavy builds strength and bone.
EffortStop 1–3 reps shortFailure isn't required for growth.

Volume — how many sets. Schoenfeld, Ogborn and Krieger's dose-response meta-analysis in the Journal of Sports Sciences (2017) found that each additional weekly set per muscle was associated with an extra 0.37% gain in muscle size, and that higher-volume conditions beat lower-volume ones within the same studies. Grouped by band, the average gains were 5.4% with fewer than 5 sets a week, 6.6% with 5–9, and 9.8% with 10 or more — though that three-way comparison was only a trend, not statistically significant. The practical reading: aim for roughly 10 hard sets per muscle group per week, build towards it, and don't expect doubling that to double the result. Peterson's over-50s showed the same volume effect.

Load — how heavy. The same group's meta-analysis of 21 studies in the Journal of Strength and Conditioning Research (2017) compared low loads (60% of one-rep max or less) with high loads, all taken to failure. Muscle growth was similar. Strength gains were significantly greater with heavy loads. For anyone over 40 with a cranky shoulder or knee, that's liberating — lighter weights, taken close to failure, still build muscle. But there are two reasons not to live there entirely. Strength is the variable that predicts outcomes, and heavy loading is what builds bone: in the LIFTMOR trial, postmenopausal women with low bone mass lifting at over 85% of their maximum, twice a week under supervision, raised their spine bone density 2.9% in eight months while the control group lost 1.2%. The National Strength and Conditioning Association's position for older adults lands at 70–85% of one-rep max for the main lifts, which is where this plan lands too.

Frequency — how often. Schoenfeld's 2016 meta-analysis in Sports Medicine compared training each muscle once versus twice a week with total volume held equal: twice a week produced a significantly larger effect on muscle size (effect size 0.49 against 0.30). So ten sets of legs a week are better as two sessions of five than one of ten — which is also kinder to a 45-year-old's recovery.

Failure — how hard each set. Grgic and colleagues' 2022 meta-analysis in the Journal of Sport and Health Science found no significant difference in strength or muscle growth between sets taken to failure and sets stopped short of it — and when volume wasn't matched, non-failure training actually produced more strength. There was one exception, in already-trained people, where failure helped muscle size a little. And one major caveat for this article: every study included was in young adults. For older lifters, the rule that follows is the sensible one — finish most sets with one to three good reps still in the tank, and save true failure, if you use it at all, for the last set of a safe exercise on a machine.

Power — the dial most programmes forget. Explosive strength — the ability to produce force quickly — is what catches you when you trip, and ordinary slow lifting doesn't fully train it. The NSCA's 2019 position statement for older adults (Fragala, Journal of Strength and Conditioning Research) recommends 2–3 sets of one or two multi-joint exercises per major muscle group at 70–85% of one-rep max, two to three times a week — and explicitly includes power work: lifting moderate loads (40–60% of maximum) as fast as you safely can on the way up. It also states that properly designed, supervised resistance training is safe for healthy older adults.

Does cardio interfere? The old fear was that running or cycling would blunt strength gains. Schumann and colleagues' 2022 meta-analysis of 43 studies in Sports Medicine found it doesn't — concurrent aerobic training did not compromise muscle growth or maximal strength. It did modestly blunt explosive strength, especially when both were done in the same session. So keep your cardio — the Norwegian 4x4 and your daily steps are on the base of the longevity pyramid for a reason — and, if you can, put the hard cardio on a different day or at least a few hours from the lifting.

Part four · Protein

More Than You Think, and More Per Meal

Evidence: Strong for the daily total; moderate for the per-meal dose

The largest meta-analysis of protein and training is Morton and colleagues' in the British Journal of Sports Medicine (2018): 49 trials, 1,863 people. Protein supplementation during resistance training increased gains in fat-free mass by 0.30 kg and one-rep-max strength by 2.49 kg — modest numbers, worth being honest about. The most-cited finding is the ceiling: beyond a total intake of 1.62 g of protein per kg of body weight per day, extra protein produced no further gains (the confidence interval runs from 1.03 to 2.20, which is why you'll see "up to 2.2 g/kg" quoted). And the effect of protein supplementation shrank with age — which sounds like bad news for over-40s, and is really a clue about the next finding.

Nunes and colleagues' 2022 review of 74 randomised trials in the Journal of Cachexia, Sarcopenia and Muscle found that more protein enhanced lean mass gains in people doing resistance training — the benefit was in the lifters — with significant effects at 1.6 g/kg or more in people under 65, and at 1.2–1.59 g/kg in people 65 and over. The additional gains were small. Protein doesn't build muscle; training does, and protein lets it.

Why older muscle needs more at each meal. Moore and colleagues pooled protein-dose studies in older and younger men in the Journals of Gerontology (2015). Muscle protein synthesis plateaued at about 0.40 g/kg of body weight per meal in men around 71, against 0.24 g/kg in men around 22 — a difference that was clearly significant when expressed per kg of lean mass, and borderline (p = 0.055) per kg of body weight. This is the "anabolic resistance" of ageing muscle, and the fix is dosing: a 40-year-old should behave more like the 71-year-olds than the 22-year-olds. For an 80 kg person that's about 30–35 g of protein at each of three or four meals, not 10 g at breakfast and 80 g at dinner. (Protein before bed has a trial behind it — 27.5 g nightly improved gains over 12 weeks — but it was in young men and against a calorie-free placebo, so it's an option, not a rule.)

Your protein numbers

Daily range from Morton 2018 and Nunes 2022; per-meal figure from Moore 2015's older-adult plateau (0.4 g/kg). If you have kidney disease, check with your doctor before raising protein. The site's protein guide has food sources.

Part five · Supplements

Creatine, and Not Much Else

Evidence: Moderate — consistent meta-analytic effect in older adults

One supplement has consistent evidence in this age group. Chilibeck and colleagues' meta-analysis in the Open Access Journal of Sports Medicine (2017) pooled 22 studies of 721 men and women whose average ages ran from 57 to 70, all doing resistance training two or three days a week for 7 to 52 weeks. Adding creatine produced 1.37 kg more lean tissue than placebo, and small additional gains in chest press and leg press strength. Two honest notes: some of that "lean tissue" is water drawn into muscle, which the scans count as lean mass; and creatine works with training, not instead of it. At 3–5 g a day it's cheap, among the most-studied supplements in existence, and the site's full creatine article (and the version for women) covers dosing, safety and the brain evidence. Beyond protein and creatine, no supplement on the shelf has comparable evidence for building muscle at 40-plus.

Part six · Hormones

Testosterone, Menopause, and What They Don't Replace

Evidence: Strong that testosterone builds muscle; strong that it's not a training aid for normal men

Men. Testosterone builds muscle; that was never in doubt. Bhasin's 1996 trial in the New England Journal of Medicine gave 43 healthy young men 600 mg of testosterone a week — several times a replacement dose — for ten weeks, and the group that also trained gained 6.1 kg of fat-free mass. Even the group that didn't train gained muscle. That's exactly why it's not the answer: it's a supraphysiological dose, in young men, and the result is the definition of what a drug does, not what a 45-year-old with normal levels should do. For men who genuinely have low testosterone and symptoms, replacement is a treatment, and the TRAVERSE trial (NEJM 2023; 5,246 men aged 45 to 80 with levels below 300 ng/dL) found it didn't raise major cardiovascular events (hazard ratio 0.96) — while also finding more atrial fibrillation, acute kidney injury and pulmonary embolism. TRAVERSE measured heart safety, not muscle, and it said nothing about men with normal levels. If you're tired and losing strength, get a blood test through a doctor. If it's normal, the answer is on this page, not in a clinic's subscription plan. The men's longevity pyramid puts it at the tip for that reason.

Women. The menopause transition shifts body composition, as SWAN showed, and hormone therapy is a legitimate treatment for bothersome symptoms and bone protection in women under 60 or within ten years of menopause — the women's pyramid covers the current evidence, including the 18-year Women's Health Initiative mortality follow-up. What no hormone replaces is load. The single best-evidenced bone intervention on this site is still LIFTMOR's heavy, supervised lifting, and the strength and muscle that come with it are the same ones that predict the outcomes above.

Part seven · The plan

Two Versions, Built From the Evidence

Everything above, turned into a week. The structure follows the NSCA's recommendations for older adults and the dose-response data: each major muscle group trained twice a week, main lifts at roughly 70–85% of maximum, 2–3 working sets, most sets stopped one to three reps short of failure, a power element, and progression over months. Choose the version that fits your week; both get each muscle to roughly 8–12 hard sets a week.

Version A — two full-body sessions a week (≈45 min)
ExerciseSession 1Session 2Sets × reps
Legs — squat patternGoblet or back squatLeg press or split squat3 × 6–10
Legs — hingeRomanian deadliftTrap-bar or conventional deadlift3 × 5–8
PushBench or dumbbell pressOverhead press3 × 6–10
PullRow (cable or dumbbell)Lat pulldown or assisted pull-up3 × 8–12
PowerBox step-up, fast upMedicine-ball chest throw3 × 5, 40–60% effort, fast
Carry / coreFarmer's carrySuitcase carry or plank3 × 30–40 m / 30 s
Version B — three sessions a week (≈40 min each)
DayFocusMain liftsSets × reps
MondayLower + pullSquat · Romanian deadlift · row · farmer's carry3 × 6–10
WednesdayUpper + powerBench or push-up · overhead press · pulldown · med-ball throw3 × 6–10; power 3 × 5
FridayFull bodyDeadlift variation · split squat · dumbbell press · row · step-up (fast)2–3 × 5–10
01

Weeks 1–4: learn the movements, leave reps in the tank

Start at a weight you could lift for 12–15 reps and do 8–10. Two working sets, not three. Your joints and tendons adapt more slowly than your muscles, and the injury that ends most over-40 lifting careers happens in the first month, not the first year. If you've never deadlifted or squatted with a bar, pay for three or four sessions with a qualified coach — the supervised trials that showed heavy lifting is safe for older adults were supervised for a reason.

02

Weeks 5–12: add the third set and the load

Move to three sets on the main lifts and work towards weights you can lift 6–10 times with one to three reps to spare — roughly 70–85% of your maximum. When you can do the top of the rep range on every set with good form, add the smallest weight increment available next session. Keep the power exercise light and fast. Track every session in a notes app; progression is the whole mechanism, and you can't progress what you don't record.

03

Months 4–12: keep going — the best results in the data took a year

Borde's older-adult trials that ran 50–53 weeks produced the largest strength gains. Rotate exercise variations every two or three months to keep your joints happy, take an easier week every six to eight weeks, and re-test your grip and your key lifts each quarter. Eat to the protein numbers above, take creatine if you want the extra 1–1.5 kg, and sleep — the sleep piece explains why a week of six-hour nights undoes more than you think.

04

Around it: walk, and keep the cardio

Concurrent training doesn't cost you muscle, and strength plus aerobic activity was associated with 40% lower mortality in Momma's data — far more than strength training alone. Get your daily steps, add one hard interval session a week on a non-lifting day if you can, and eat the way the Blue Zone meal plan eats — then add protein to it.

When to see a doctor first. If you have known heart disease, chest pain or breathlessness on exertion, uncontrolled blood pressure, a recent fracture or joint replacement, osteoporosis with previous fractures, or a hernia, get clearance and ideally a supervised start. Pain that is sharp, joint-line, or gets worse set by set is a signal to stop that exercise, not push through it.
Common Questions
Can you still build muscle after 40?
Yes, at every age that's been studied. Across 49 studies in over-50s, resistance training added an average 1.1 kg of lean mass, more with higher-volume programmes. Frail 90-year-olds raised their strength 174% and their thigh muscle area 9% in eight weeks. The response shrinks somewhat with age — an argument for starting now.
How much protein do you need to build muscle after 40?
About 1.6 g per kg of body weight a day while strength training — the point beyond which Morton's 49-trial meta-analysis found no further gains (confidence interval 1.03–2.20). Older muscle needs more per meal: around 0.4 g/kg per meal in older men versus 0.24 g/kg in young men, so spread it across three or four meals of 30–35 g for an 80 kg person.
Do you need to lift heavy weights after 40?
Not for muscle size — light and heavy loads grew similar muscle across 21 studies when sets went close to failure. Heavy loads build more strength, and heavy supervised lifting is what raised bone density in postmenopausal women. The NSCA recommends 70–85% of one-rep max for older adults' main lifts.
Should people over 40 take creatine?
It's the one supplement with consistent evidence at this age: +1.37 kg lean tissue versus placebo across 22 studies of 721 adults averaging 57–70, with small extra strength gains. Some of that is water in muscle. It works alongside training, not instead of it.
The Bottom Line

The decline is slow. The fix is two sessions a week.

You do lose muscle after 40 — slowly, mostly in the legs, and strength faster than size. But the same research that measured the decline measured the reversal, in people from their fifties to their nineties, and it works: about a kilogram of lean mass on average, strength gains measured in tens of percent, and an association with 10–20% lower mortality from roughly an hour a week.

Train each muscle twice a week. Lift moderately heavy, stop a rep or two short, add weight when it gets easy, and include something fast. Eat about 1.6 g/kg of protein, more of it at breakfast than you do now. Take creatine if you like. Keep walking. And give it a year — that's how long the best results in the data took.