Is it safe to lift weights with arthritis?

Usually, yes. For most people with arthritis, strength training is not just permitted. It is part of the standard care. The American Academy of Orthopaedic Surgeons recommends exercise for knee arthritis. The largest review of the evidence found that exercise lowered pain and improved function, with no serious harm reported in any trial.

The instinct to rest a worn joint feels protective, and it usually runs backwards. The muscle around an arthritic joint is what shares its load. Let that muscle shrink and the joint absorbs more force, not less. For most people the honest questions are about how to load the joint and how fast to progress, not whether to start. There is one exception, and it is the one this page takes most seriously: a joint that has stopped moving. This page walks through the evidence, including the parts that cut against heavy lifting, and the point at which strengthening runs out of room.

What does exercise actually do for an arthritic joint?

A Cochrane review from 2015 pooled 54 studies of exercise for knee arthritis. In 44 trials covering about 3,500 people, exercise reduced pain by roughly 12 points on a 0 to 100 scale, in groups whose pain averaged 44 points without it. Read plainly: about a quarter of the pain, gone. The reviewers rated that evidence high quality. The size of the relief was in the same range as what anti-inflammatory drugs achieve.

Physical function improved by about 10 points on the same kind of scale, across 44 trials of about 3,900 people. That evidence was rated moderate quality.

The safety record matters as much as the benefit. A handful of studies reported side effects, and all of them were flare-ups: more knee pain or low back pain for a while. No study reported a serious harm from the exercise itself.

One honest limit: the benefit fades when the training stops. Two to six months after formal programs ended, the pain benefit had shrunk to about 6 points. Exercise for arthritis works like exercise for anything else. It keeps working as long as you keep doing it.

Can lifting weights really rebuild bone?

This is where the strongest single trial comes in, and it deserves to be described exactly.

The LIFTMOR trial, published in the Journal of Bone and Mineral Research in 2018, enrolled 101 women past menopause with low bone density, average age 65. It split them at random into two programs for eight months. One group did a home program of light exercise. The other lifted heavy, twice a week, thirty minutes at a time: five sets of five repetitions at more than 85 percent of the most they could lift, always under close supervision.

The results ran in opposite directions. Spine bone density rose 2.9 percent in the lifting group while the light-exercise group lost 1.2 percent. At the femoral neck, where hip fractures happen, the lifters held roughly steady while the light-exercise group kept losing bone. The lifters also improved on every functional test in the study: rising from a chair, a timed get-up-and-go, back and leg strength, balance reach. Those are the plain measures that track with staying independent.

Across eight months of heavy lifting in women who averaged 65 years old with fragile bones, the trial recorded one adverse event: a minor lower back spasm that cost two training sessions.

Now the honest bridge. These women had low bone mass, not arthritis. They were screened and otherwise healthy. Every session was watched by a professional. So LIFTMOR cannot promise an arthritic knee the same experience. What it did was retire an assumption: that older people with fragile bones are too fragile to train heavy. Under supervision, they were not. Bone responded to serious load, at an age when most people have been told to be careful with themselves.

Do I have to lift heavy?

No, and the best trial on this question cuts both ways. It is worth hearing both.

The START trial, published in JAMA in 2021, followed 377 adults age 50 and older with knee arthritis. For eighteen months, each was assigned at random to heavy strength training, light strength training, or a comparison group that did no strength training. Heavy training did not beat light training for knee pain, and it did not beat the comparison group either. The forces passing through the knee during walking were no different across groups. Minor problems were more common in both training groups. The serious medical events that occurred were unrelated to the study.

Two findings sit inside that result. First, on the measures the trial took, eighteen months of heavy strength training showed no sign of harming arthritic knees. The fear that loading a worn joint grinds it down faster did not show up in the measurements. Second, heavy earned no extra pain relief. An arthritic knee does not require maximal loads to feel better.

Put LIFTMOR and START side by side and the picture gets practical. Bone responds to heavy load. Arthritis pain responds to consistent load. If your main concern is bone density, there is a case for training heavy with supervision. If your main concern is an aching knee, lighter progressive work does the job, and consistency beats intensity.

Why does the lowering half of a lift matter so much?

Every repetition has two halves. Lifting the weight shortens the muscle. Lowering it lengthens the muscle while it is still working. Trainers call these concentric and eccentric, and for an arthritic joint the second half is the more interesting one.

Two things are true about the lowering half. You are roughly 20 to 50 percent stronger lowering a weight than lifting it, so the same muscle can be worked harder on the way down without adding load. And lowering is the half that protects a joint in daily life. Going down stairs, sitting into a chair, catching your balance: those are all a muscle lengthening under control. When that control is weak, the joint takes the drop.

The research leans the same way without overclaiming. A 2017 meta-analysis of trials comparing the two halves found that eccentric training produced somewhat more muscle growth, about 10 percent versus 7, though the gap did not reach statistical significance, and the authors’ conclusion was that a good program needs both. In adults over 60 specifically, a 2025 review of eleven randomized trials found eccentric-focused training produced slightly larger strength gains than conventional lifting, with the two about equal on function and muscle size.

The practical version costs nothing. Lower every repetition slowly, three or four seconds, and lift it at normal speed. For an aching knee this is a better lever than adding plates, which is exactly what START found: more load did not buy more relief.

What if the joint will not move?

This is the caveat that matters most, and it is usually left out.

A joint is a lever. The muscle attaches close to the pivot, the weight sits far from it, so the muscle has to produce several times the force of whatever you are holding, and that force passes through the joint surface. Two consequences follow. First, how you load the joint matters more than how much. Second, the lever changes with the angle. A knee that no longer straightens fully, or a hip that no longer opens, is being loaded at angles where the muscle has poor leverage and the joint surface takes more of the force.

There is a bigger problem underneath that one. A muscle gets strong mostly in the range you train it. If arthritis has taken the last twenty degrees of bend from a knee, the quadriceps cannot be trained where it does its most important work, which is the deep part of a squat or the bottom of a stair. Strength work still helps in the range that is left. But it cannot reach the range that was lost, and the range that was lost is often the one causing the trouble.

This is why motion is the first thing to check, before load and before frequency. If the joint moves through most of its range with tolerable pain, the evidence above applies to you and the question is only how to progress. If motion is the thing arthritis has taken, strength training alone has a ceiling, and that ceiling is a mechanical fact rather than a failure of effort. It is also the clearest signal that the conversation in the last section of this page has come due.

Does muscle still grow at my age?

Yes. Older muscle grows a little slower and needs a little longer, but it grows.

The honest version first. When researchers train younger and older adults side by side, the younger group usually adds muscle fiber size faster. In one 21-week trial, fiber growth was larger in men in their twenties than in men in their sixties. Strength was a different story: the older men lagged at ten weeks and had caught up by twenty-one. Age slows the rate. It does not close the door.

The most extreme test anyone has run makes the point. In 1990, researchers put ten frail nursing-home residents, average age 90, through eight weeks of heavy leg training. Strength rose an average of 174 percent. Mid-thigh muscle area grew 9 percent. Walking speed improved 48 percent. These were people in their nineties, some of them using walkers, and their muscle answered in two months.

A 2024 analysis of trials in adults 65 and older found that how long people trained predicted muscle growth better than how old they were. That is the variable you control. The muscle does not check your birth certificate. It checks whether you showed up last week and the week before.

How should I start without flaring the joint?

• Start below what you think you can do, and add load slowly. The trials that produced the safety record above were progressive, not sudden.
• Two or three sessions a week is what the evidence used. More is not required.
• Check motion first. If the joint moves through most of its range, progress the load. If it does not, get it examined before you build a program around what it will not do.
• Lower slowly, lift at normal speed. The eccentric half is the cheapest gain available and the one that protects the joint on stairs.
• Expect some soreness. A flare that settles within a day or two is common and acceptable. Pain that escalates session over session, or lingers, means the load or the exercise needs to change, not that training is over.
• Get supervision early. The LIFTMOR safety record was earned with a professional watching every session. The American Association of Hip and Knee Surgeons notes that in advanced arthritis, exercise can sometimes increase knee pain, which is exactly when a physical therapist or qualified trainer earns their fee: they change the exercise instead of letting you quit.
• Let balance work ride along. Single-leg practice and step work cost minutes and feed the same functional measures LIFTMOR improved.

This is the ground the RISE approach is built on: strength and capability built before, after, and sometimes instead of an operation.

When strengthening alone is not enough

Some joints reach a point where training cannot get traction. The signal is usually not an X-ray. Often it is motion: the range the joint has lost is the range the muscle can no longer be trained in. Beyond that it is the pattern: the strength work that was helping has quietly stopped happening because the joint will not allow it, simpler measures that used to hold for months now hold for weeks, and the joint gets a vote in more of your plans.

That pattern is worth a conversation, and the conversation is not a commitment. A consultation means an examination, an honest read of where the arthritis is, and a plan, which for many people is a better version of the non-surgical care they were already doing. The bone-on-bone article covers why even an advanced X-ray does not write the plan by itself.

And if an operation does eventually make sense, nothing built in the gym is wasted. The evidence on muscle loss and joint replacement points the same direction: the strength you bring to a surgery shapes the recovery you get from it.

Worth asking at the visit

• Is there any part of my joint problem that makes strength training unwise for me specifically?
• Should I be training with a physical therapist first, or is a gym program reasonable now?
• How much motion has this joint actually lost, and is that what is limiting my strength work?
• I want to protect my bone density too. Does heavier training make sense for me, and with what supervision?
• What would tell us that strengthening has stopped being enough?

More on how arthritis behaves and what the options look like: knee arthritis.