It’s not too late to get stronger.

In 1990, ten residents of a Boston nursing home, average age 90, spent eight weeks training their legs with weights. Several had arrived using walkers. Their strength nearly tripled. More than three decades of research since has said the same thing in larger and more careful ways: muscle and bone answer to training at 50, at 70 and at 90, in women and men alike. What changes with age is not whether you respond. It is how much it costs to wait.

Strength leaves before muscle does

The usual picture of aging muscle is shrinkage. The bigger problem is weakness. In adults over 70, muscle mass falls by about 0.5 to 1 percent a year, but strength falls by 2 to 4 percent, two to five times faster (Mitchell, Frontiers in Physiology 2012). In one study that followed 1,678 older adults for five years, men lost about 16 percent of their thigh strength but only 5 percent of their thigh muscle. Women lost 13 percent of their strength and 3 percent of their muscle.

Three details explain why it feels the way it does.

• Legs go first. Lower-body muscle is lost more than twice as fast as upper-body muscle, which is why stairs, low chairs and curbs get harder before anything else does.
• Power fades fastest. Power is strength applied quickly. It is what catches you when you trip, and it declines even faster than strength.
• Weakness predicts what happens next. Older adults with below-average knee strength had about 2.5 times the odds of dying within four years. In the PURE study of nearly 140,000 adults in 17 countries, each 5 kg (11 lb) of lower grip strength meant a 16 percent higher risk of death, and grip predicted death better than systolic blood pressure did (Leong, Lancet 2015).

Most of this is not written in your genes. In older adults, genetics explains only a small share of the differences in muscle and strength. Activity and protein intake explain most of it, and both are things you can change.

The trials keep finding the same thing

Every major study of older beginners points the same way. Muscle responds to lifting at any age studied, and the gains are not small.

Study	What happened
Fiatarone, JAMA 199010 nursing-home residents, average age 90 · 8 weeks of leg training	Strength up 174%, thigh muscle up 9%, walking speed up 48%.
Fiatarone, NEJM 1994100 frail residents, average age 87 · 10 weeks at about 80% of max	Strength up 113% against 3% without exercise; walking speed up 11.8% against a 1.0% fall.
Peterson, 201047 studies, 1,079 adults aged 50 to 92 · 6 to 52 weeks	Strength up 24% to 33% across the major lifts. Age and sex did not limit the gains.
Peterson, 201149 studies, 1,328 adults over 50 · about 20 weeks	About 1.1 kg (2.4 lb) of lean mass gained, and more training volume brought more gain.
Churchward-Venne, 2015110 men and women over 65 · 12 to 24 weeks, supervised	Every participant improved on at least one measure. There were no true non-responders.

The benefits reach well past the muscle itself. In randomized trials:

• Blood sugar. In type 2 diabetes, supervised resistance training lowered HbA1c by 0.57 points. Advice to exercise, on its own, did not lower it at all (Umpierre, JAMA 2011).
• Blood pressure. Lifting lowered resting pressure modestly, by 1.8/3.2 mmHg. Isometric holds such as wall sits and handgrip squeezes lowered it by 10.9/6.2 mmHg (Cornelissen & Smart, J Am Heart Assoc 2013).
• Mood. Across 33 trials and 1,877 adults, resistance training clearly reduced depressive symptoms, regardless of age or health status (Gordon, JAMA Psychiatry 2018).
• Thinking. Women aged 65 to 75 who lifted once or twice a week for a year improved executive function by 11 to 13 percent (Liu-Ambrose, Arch Intern Med 2010). In mild cognitive impairment, six months of heavy training improved overall cognition, and the benefit was still there a year after training stopped (SMART; Broadhouse 2020).
• Knee arthritis. Exercise cut pain by about 12 points on a 100-point scale, and one in four people got a meaningful benefit they would not otherwise have had (Fransen, Cochrane 2015).

The longer-term picture comes from observational studies rather than trials, so it shows association rather than proof. Adults who do muscle-strengthening work have a 10 to 17 percent lower risk of early death, heart disease, cancer and diabetes, with most of the benefit at 30 to 60 minutes a week (Momma, Br J Sports Med 2022).

What it takes

The dose is smaller than most people assume. Two or three sessions a week, building gradually toward fairly heavy loads, is what the trials used, and in long-term studies most of the health benefit appears by 30 to 60 minutes a week.

The training. The National Strength and Conditioning Association’s position statement for older adults, and the largest dose-response analysis behind it, point to a consistent shape (Fragala, NSCA 2019; Borde, Sports Med 2015):

• Two to three sessions a week, with one or two multi-joint exercises for each major muscle group and the emphasis on the legs.
• Two to three sets of seven to nine repetitions, which was the most effective range for both strength and size.
• Loads that build from light toward 70 to 85 percent of your maximum. Heavier training produced bigger strength gains, about 5 percent more for each step up in intensity in the Peterson analysis above.
• Power work once a base is built: lift the weight quickly, lower it under control. It builds the fast strength that catches a stumble better than slow lifting does (El Hadouchi 2022).
• Balance and functional work alongside every program, for reasons the next section makes clear.

In practice the first month is for learning the movements (sit-to-stands, step-ups, rows, push-ups against a bench, hip hinges, carries) with two or three repetitions left in reserve on every set. The load rises in small steps from there. Heavier work comes after about three months, with coaching. Strength usually climbs within the first few weeks, largely because the nervous system learns to use the muscle you already have. Muscle size follows over months.

The protein. Older muscle needs a bigger dose to respond, which researchers call anabolic resistance. In men around 71, muscle building after a meal plateaued at about 0.4 g of protein per kg of body weight, compared with 0.24 g/kg in men around 22 (Moore, J Gerontol 2015). For most older adults that is roughly 30 to 35 g of protein at each meal. Across the day, the PROT-AGE group recommends at least 1.0 to 1.2 g/kg for adults over 65, and 1.2 g/kg or more for those who exercise (Bauer, JAMDA 2013). Kidney disease changes these numbers, so that conversation comes first.

The extras, ranked honestly. Creatine taken alongside lifting added about 1.2 to 1.4 kg of lean mass over lifting alone in older adults (Chilibeck 2017). Vitamin D improves strength only slightly, mainly in people who are deficient, so the sensible approach is to test and correct a deficiency rather than supplement everyone (Beaudart, JCEM 2014). Neither replaces the training.

Check in first if you have heart, lung, kidney or metabolic disease; chest pain, unusual breathlessness, dizziness or fainting with effort; a recent fracture, surgery or joint replacement; or known osteoporosis. None of these rule out lifting. They change how you start and who supervises you.

Why now rather than later. Older muscle disappears quickly when it isn’t used. Ten days of bed rest cost healthy adults around 67 almost a kilogram of leg muscle, more than younger adults lost in 28 days (Kortebein, JAMA 2007). Two weeks of simply walking less, down to about 1,400 steps a day, cut leg lean mass by 3.9 percent in adults around 72 (Breen, JCEM 2013). Strength built now is reserve for the surgery, illness or hospital stay that hasn’t happened yet.

Bone, balance and the fall you don’t take

Bone responds to heavy, fast loading, which is exactly what older women have usually been told to avoid. The LIFTMOR trial tested it directly. 101 postmenopausal women with low bone mass, average age 65, trained in supervised 30-minute sessions twice a week for eight months, working up to five sets of five repetitions above 85 percent of their maximum, plus jumping-style impact. Spine bone density rose 2.9 percent in the lifting group while the light-exercise group lost 1.2 percent. The lifters also stood from a chair faster, balanced better and gained a little height (Watson, J Bone Miner Res 2018).

The safety record is the part that surprises people. Across the 49 women in the lifting group there was one minor back spasm, and a follow-up found the training caused no vertebral fractures and improved the forward rounding of the upper back (Osteoporos Int 2019). A companion trial in 93 men, average age 67, found better spine and hip bone and better physical performance than in the comparison group (LIFTMOR-M, J Bone Miner Res 2020).

Falls are the other half of the story. At least one in three adults over 65 falls each year. The 2019 Cochrane review pooled 108 trials and 23,407 people: exercise cut the rate of falls by 23 percent overall, and by 34 percent in programs that combined balance, functional and resistance training. Resistance training on its own had an uncertain effect (Sherrington, Br J Sports Med 2019). Put plainly, if 1,000 inactive older adults would have 850 falls a year, exercise would prevent about 195 of them.

Walking speed belongs in the same conversation. In 34,485 adults 65 and older, each 0.1 meter per second of faster walking was linked to a 12 percent lower risk of death (Studenski, JAMA 2011). Progressive resistance training raises walking speed by about 0.08 meters per second on average (Liu & Latham, Cochrane 2009), and the residents in that 1990 study walked 48 percent faster after eight weeks.

What the research does not show

The case for lifting later in life is strong, but its pieces rest on different kinds of evidence, and it is worth knowing which is which.

• Lifespan findings are associations. The links to longer life, fewer heart attacks and less cancer come from observational cohorts. People who lift tend to be younger, more educated and healthier, and studies adjust for that only partly. No trial has tested whether starting to lift late in life extends it.
• Bone density is a stand-in for fractures. LIFTMOR measured bone, not breaks, and no trial yet shows that lifting reduces fractures directly. The men’s trial was not fully randomized, and one LIFTMOR investigator owns a clinic that licenses a program based on the protocol.
• Many trials are small and short. In the Cochrane review on physical function, many studies had fewer than 40 people and measured results after three or four months, with little longer follow-up.
• Most participants were healthy and under 80. Women over 80 showed little muscle growth after 12 weeks in one study, and people living with several chronic conditions may respond differently.
• Strength alone does not prevent falls. Programs that add balance work carry the strongest evidence.

None of this weakens the central finding, which comes from randomized trials: strength, muscle, physical function, bone density, blood sugar, mood and cognition all improve with training at every age that has been studied.

Where that leaves you

The best time to start was years ago. The second-best time is now, and the research says the window does not close. People in their sixties, seventies and nineties have been measured getting stronger in a matter of weeks, and the gains show up where life is actually lived: getting up from a chair, climbing stairs, staying upright, thinking, mood.

What makes it work after 50 is what makes everything else here work: a measured starting point, a plan that progresses, and someone watching the numbers. At Athlos Health that starts with a baseline. Grip strength, how many times you can stand from a chair in 30 seconds, walking speed and balance give you something to beat. A body-composition scan separates muscle from fat, and the labs show whether anything needs attention before you load a bar. You can see what a baseline includes, or take the Longevity Assessment first.

For people who want the whole plan, the Athlos Transformation Program pairs physician oversight with three coached sessions a week at our partner training facility, Athlos Iron Lair, including a Foundations track for anyone starting from a long layoff.

Written by Athlos Health, Torrance, California, and medically reviewed by Christopher Traughber, M.D. This article is general education and is not medical advice, a diagnosis, or a treatment recommendation. Talk with a physician before starting a training program if you have a heart, lung, kidney or metabolic condition, a recent fracture or surgery, or osteoporosis. Every figure above is linked to its source below.