Resistance training and bone density: what the evidence supports

The short answer: resistance training has been shown to preserve bone mineral density and produce small increases at the hip and spine, and it clearly beats doing nothing. But the effect is modest — typically 1–3% over 6–12 months in trial conditions, not the double-digit gains sometimes advertised. The load has to be meaningful and the training has to continue, because bone reverts when loading stops. Its larger contribution is probably indirect: stronger, better balanced people fall less, and falls are what break bones.

What it is

Bone is living tissue that rebuilds itself continuously. Cells called osteoclasts remove old bone; osteoblasts lay down new bone. Through your thirties those processes roughly balance. After that, removal gradually outpaces replacement, and bone mineral density drifts downward.

The drift is not uniform. In women, the years around menopause bring the sharpest losses — commonly cited figures run to around 2% per year in the first several years after the final period, driven by falling estrogen, before settling to a slower rate. Men lose bone more gradually and start from a higher peak, which is why their fracture risk arrives about a decade later rather than not at all.

Bone responds to mechanical strain. When a load deforms bone slightly, the cells within it register that strain and the remodeling balance shifts toward building. The key features appear to be strain magnitude, rate, and novelty — bone responds better to loads that are heavy, applied quickly, and unlike its routine. It adapts quickly to the familiar and stops responding, which is why walking, whatever else it does for you, does relatively little for bone in people already walking daily.

Density is measured by DXA scan and reported as a T-score comparing you to a healthy young adult. A T-score of −2.5 or lower defines osteoporosis; between −1 and −2.5 is labelled osteopenia. Those thresholds are useful and also somewhat arbitrary — fracture risk is continuous, and most fractures actually occur in people whose scores sit above the osteoporosis line. Bone density is a strong risk factor, not a diagnosis of what will happen.

What the evidence shows, graded

Has been shown to

• Preserve bone density compared with not training. This is the most robust finding in the area. Across many trials in postmenopausal women and older men, resistance training groups maintain or slightly increase density at the spine and hip while control groups lose it. The gap between groups is often larger than the gain within the training group — preventing loss is most of the benefit.
• Produce small gains at loaded sites. Increases of roughly 1–3% at the lumbar spine and femoral neck over 6–12 months are the typical trial result with progressive, meaningful loads. Real, measurable, and modest.
• Be site-specific. Bone adapts where it is loaded. Squats and deadlifts affect the hip and spine; upper-body work does not. There is no systemic bone-building effect that spreads from one region to another.
• Reduce falls when combined with balance training. Multi-component exercise programmes including strength and balance work have been shown to reduce fall rates in older adults. Given that the large majority of fragility fractures follow a fall, this may be the most important bone-relevant thing exercise does.
• Improve strength, muscle mass, and function at every age studied. Not a bone outcome, but it’s the reliable one, and it arrives regardless of what the scan says.
• Be safe in most people with low bone density when supervised and progressed sensibly. Trials including women with osteoporosis have reported low adverse event rates with appropriately prescribed resistance training.

Research suggests

• Higher loads work better than lighter ones. Trials using heavier resistance — in the range of 80–85% of a one-repetition maximum, progressed over time — tend to report larger density effects than light-load protocols. Some supervised high-intensity programmes in postmenopausal women with low bone mass have reported gains at the upper end of that 1–3% range with good safety records. The heavier-is-better direction is reasonably consistent; the optimal prescription is not settled.
• Impact adds something resistance training alone doesn’t. Jumping, hopping, and bounding generate high strain rates, and programmes combining impact with resistance work tend to outperform either alone at the hip. Suitability depends heavily on joint health and existing bone status.
• Effects require maintenance. When training stops, density gains fade over the following year or two. Bone follows the same use-it-or-lose-it logic as muscle, on a slower clock.
• Adequate calcium and vitamin D are permissive. Training against a background of insufficient calcium or vitamin D appears to blunt the response. Sufficiency helps; excess doesn’t add more.
• Benefits extend to men. Most trials have studied postmenopausal women, where the signal is strongest. The smaller body of work in older men points the same direction.

May help

• Changing fracture risk directly through density gains. This is the honest gap. Almost no exercise trial has been large enough or long enough to measure fractures as a primary outcome — that would take thousands of participants over many years, and nobody has funded it. We infer benefit from density changes, fall reduction, and consistency with drug trial data. It is a reasonable inference. It is not a demonstrated result.
• Whole-body vibration platforms. Studied repeatedly with inconsistent results. Some trials show small hip effects, others nothing. Not a substitute for loading.
• Bone quality improvements that DXA can’t see. Bone strength depends on architecture and geometry as well as mineral density, and training may improve those in ways a T-score doesn’t capture. Plausible, supported by some imaging work, not established.

Who it may suit

Nearly everyone over 40, with the case strongest for:

• Women approaching or in the first years after menopause. This is when loss is fastest and when there’s most to preserve. Starting before or during the transition means defending a higher baseline.
• Anyone with osteopenia on a scan. This is the group where training has the clearest role, since medication may not yet be indicated and the loss is ongoing.
• People with risk factors: family history of hip fracture, low body weight, smoking, long-term corticosteroid use, early menopause, or extended periods of immobility.
• People with diagnosed osteoporosis — but with guidance on which movements and loads are appropriate. Having osteoporosis is a reason to train differently, not a reason to avoid loading. Understanding it as a reason to be careful with the spine specifically, rather than with exercise generally, is the distinction that matters.

What a bone-relevant week looks like in practice: two to three sessions including loaded lower-body and spinal work — squat or leg press, hinge or deadlift variation, a carry, a step-up — progressed over months toward loads that are genuinely challenging for 5–8 repetitions. Add balance work. Add impact only if your joints and bone status allow it.

Who should ask a clinician first

• Anyone with a diagnosis of osteoporosis, particularly with a history of vertebral fracture. Loaded spinal flexion and twisting under load are the specific movements that warrant caution.
• Anyone who has had a fragility fracture — a break from a fall at standing height or less — at any age over 50.
• Anyone on long-term corticosteroids, aromatase inhibitors, or androgen deprivation therapy, all of which accelerate bone loss.
• Anyone with a history of eating disorders, extended amenorrhea, or bariatric surgery, which affect peak bone mass and absorption.
• Anyone over 65 who has never had a bone density scan and has risk factors.

What the marketing overstates

“Reverse osteoporosis with exercise.” No. Training can slow loss, produce small gains, and reduce falls. It does not restore a skeleton to its thirty-year-old state, and framing it that way sets people up to feel that a stable scan is a failure. A stable scan at 60 is a good result.

Percentage claims from short programmes. Headline figures of 10% or 15% gains generally come from small studies, unusual populations, specific skeletal sites measured favourably, or comparisons that include the control group’s losses in the number. The trial-level reality sits at 1–3%.

“Walking builds bone.” Walking is worth doing for many reasons. For bone specifically, in people who already walk regularly, the strain is too familiar and too low to drive much adaptation. Trials of walking alone show minimal density effects.

Supplement stacks marketed for bone. Calcium and vitamin D matter when you’re insufficient. Beyond sufficiency, more has not been shown to add bone, and high-dose calcium supplementation has raised safety questions of its own. Collagen, silica, and proprietary mineral blends are marketed well ahead of their evidence.

Vibration plates as a passive alternative. Sold as loading without effort. The evidence doesn’t support treating them as equivalent to resistance training.

The other direction — “exercise is dangerous with osteoporosis.” Also overstated, and it causes real harm. Avoiding all loading accelerates the problem it’s trying to prevent. The evidence supports careful, progressive loading with specific movement modifications, not avoidance.

When to talk to a clinician

Talk to a clinician before starting if you have diagnosed osteoporosis, a previous vertebral or fragility fracture, or you’re on a medication that affects bone. The advice you need is specific — which spinal positions to avoid under load, how fast to progress — and it takes one appointment to get.

Book an assessment if you’ve broken a bone from a minor fall after 50, if you’ve lost more than about 4 cm of height, if you’ve developed a rounded upper back, or if you have new persistent back pain, which can be the first sign of a vertebral compression fracture.

Ask about a DXA scan if you’re a woman over 65, a man over 70, or younger with risk factors — early menopause, long-term steroid use, low body weight, a parental hip fracture, or a history of smoking or heavy drinking. Screening guidance varies between countries, so this is a conversation rather than a rule.

And if you’re already on bone medication, keep training. Resistance training and pharmacological treatment address different parts of the same problem, and nothing about one argues against the other.