Bone density is one of the few areas of health where the popular advice and the trial evidence point in genuinely different directions. Most people reach for a calcium supplement. The research keeps pointing somewhere else entirely: mechanical force.
The clearest natural experiment on this question was not run in a clinic. It was run in orbit, over three decades, by an agency with an unlimited motivation to solve it and a series of expensive failures along the way.
Table of Contents
- What Actually Signals Bone To Grow?
- Do Calcium Supplements Increase Bone Density?
- Are Calcium Supplements Bad For The Heart?
- How Much Calcium And Vitamin D Is Needed?
- Does Exercise Increase Bone Density?
- What The Orbital Evidence Does And Does Not Prove
- Is Hopping Enough To Build Bone?
- How Heavy Does Resistance Training Need To Be?
- When Are Bone Medications Considered?
- The Evidence, Ranked
- References
What Actually Signals Bone To Grow?
Skeleton is not inert scaffolding. It is turned over continuously by two opposing cell populations: osteoclasts, which strip away existing bone, and osteoblasts, which lay down replacement. Whether density rises or falls comes down to which of those two crews is winning.
That balance is dynamic rather than fixed. Resorption and formation run continuously and in parallel throughout adult life, which is why density is never simply banked and held. It is the net outcome of a process that can be tipped in either direction, and identifying which input tips it is the entire practical question.
The intuitive assumption is that the deciding factor is raw material, since calcium is the mineral bone is built from. Spaceflight data dismantled that assumption in the most direct way available.
Across four- to six-month missions, crew hip bone mineral density fell 1.4 to 1.5% every month [1]. The terrestrial comparison is stark: at the same skeletal site, a woman passing through menopause loses in the order of 1% across an entire year [2]. Crews were shedding in four weeks what menopause takes more than twelve months to do.
Critically, those crews were not sedentary. Treadmill and cycle sessions were already mandatory to defend cardiovascular fitness and muscle mass. Aerobic training, performed diligently, did not slow the decline [1]. Movement alone was not the variable that mattered.
Do Calcium Supplements Increase Bone Density?
Supplemental calcium produces a small, one-off density increase of roughly 1%, which does not continue and does not translate into the outcome that actually matters, which is a broken bone avoided.
The definitive test pooled data from over 51,000 community-dwelling older adults. Calcium, vitamin D, or the two combined showed no association with lower risk of hip fracture, vertebral fracture, or total fractures [3].
The distinction between those two endpoints is worth pausing on, because it explains how a supplement can look effective on paper and still fail in practice. Density is a surrogate: a number produced by a scanner. Fracture is the event a patient actually experiences. A small upward nudge in the surrogate that never shows up as fewer broken bones has not delivered anything a person would notice, and fracture is the endpoint on which this class of supplement has repeatedly come up empty [3].
Orbital testing had already produced the same verdict. Supplying additional calcium and vitamin D to crews failed to arrest their bone loss, which pointed to a shortage of something other than mineral supply.
Are Calcium Supplements Bad For The Heart?
A second and less expected signal emerged from Auckland, New Zealand. Ian Reid and Mark Bolland designed a randomised trial to demonstrate that calcium would strengthen bone and reduce fractures in older women. The fracture benefit did not appear. What did appear was an excess of myocardial infarction in the calcium arm: 31 events against 14 on placebo, a relative risk of 2.24 [4].
The finding attracted sustained opposition from segments of the supplement industry and from parts of the bone-medicine community. The Auckland group did not retract it. New Zealand's Prime Minister's Science Prize was awarded to Reid, Bolland and Andrew Grey in recognition of the work [5].
Guidance followed the data. The US Preventive Services Task Force now advises against routine daily supplementation at or below 1000 mg of calcium plus 400 IU of vitamin D for primary fracture prevention in community-dwelling postmenopausal women [6].
Two features of that sequence deserve attention. The cardiovascular signal was not the outcome the trial set out to measure, which is exactly the circumstance in which a single finding warrants caution rather than certainty. Equally, the direction of travel since then has been consistent: the fracture benefit that justified widespread supplementation in the first place has not held up, so the risk-benefit calculation no longer rests on a demonstrated upside [3][6].
How Much Calcium And Vitamin D Is Needed?
None of the above means the mineral is unimportant. Calcium remains structurally essential; the evidence simply favours dietary sources over pills. Dairy, tinned fish with edible bones, and leafy green vegetables all deliver it within a food matrix.
Practically, this reframes the goal. Rather than treating a daily tablet as the mechanism of protection, the more defensible approach is to cover the mineral adequately through ordinary meals and to direct real effort towards the input that the trial evidence actually rewards, which is mechanical loading.
For vitamin D, the recommended dietary allowance sits at 600 IU daily for adults through age 70, rising to 800 IU from 71 onward. Both figures are set on the assumption of minimal sun exposure [17].
An important exception applies throughout. Anyone prescribed calcium by a clinician, whether for a documented deficiency or a diet excluding dairy, should not reverse that on the basis of population-level trial averages. That decision belongs with the prescribing doctor.
Does Exercise Increase Bone Density?
Bone deposits itself where it is mechanically stressed, and withdraws where it is not. Racket sports provide an unusually clean demonstration, because a single athlete carries both the loaded limb and the unloaded control. Cortical thickness on the playing side exceeded the non-playing side by 34.9% in male professionals and 28.4% in females [8].
That architecture is remarkably durable. Imaging of a former professional pitcher aged 94, fully 55 years after retirement, still found 14.3% greater bone cross-sectional area in the throwing arm [9]. Adaptation acquired before puberty was two to four times larger than adaptation acquired in adulthood, which makes childhood and adolescence the highest-yield window for building skeletal reserve.
These within-person comparisons carry unusual evidential weight. Nutrition studies must contend with the possibility that people who supplement differ systematically from people who do not, in diet, activity, income or health awareness. Comparing one limb against the other in a single body removes that entire category of confounding at a stroke. Genetics, hormones, diet and mineral intake are held identical by definition, and loading is the only variable left standing [8][9].
The orbital programme converged on the same principle by elimination. An early resistive device capped around 300 pounds of load delivered no more skeletal protection than aerobic training did [7].
Replacing it in 2008 with the Advanced Resistive Exercise Device, capable of roughly 600 pounds, changed the result completely. Only crews training on the heavier unit returned with preflight density intact across pelvis, hip and lumbar spine [7]. Site by site, the contrast is unambiguous: total hip fell about 8% on the lighter device against roughly 2% on the heavier one; lumbar spine moved from −4% to zero; whole body from −3% to zero.
One qualification belongs alongside that result. Resistive training preserved cortical bone at the femoral neck but did not halt losses in hip trabecular bone, the spongy interior compartment [10]. Loading is powerful, though not complete on its own.
What The Orbital Evidence Does And Does Not Prove
Weightlessness is an extreme condition, and crew numbers in these studies are small, so the space findings are best read as a magnified version of an ordinary process rather than a direct model of ageing. What transfers is the principle, not the rate.
The transferable point is that the programme functioned as an elimination experiment. Aerobic training was supplied and failed. Mineral and vitamin supply was supplied and failed. Light resistance was supplied and failed. Only when the load was roughly doubled did the outcome change [1][7]. Few terrestrial trials get to rule out that many candidate explanations in the same population, which is precisely what makes the conclusion so difficult to argue around.
Is Hopping Enough To Build Bone?
Unilateral hopping trials use the same within-person design as the tennis studies, with the non-hopping leg serving as control. Fifty hops daily across seven days a week raised femoral-neck density by 1.8% on the trained side. Frequencies below daily achieved nothing measurable [11].
The result carries past menopause. In postmenopausal participants, femoral-neck density climbed 0.81% in the hopping limb while the control limb declined 0.57% [12].
The frequency requirement is the detail most often lost in translation, and it changes how the habit should be built. This was not a few sessions a week producing a smaller version of the same benefit. Below daily, the measured effect disappeared entirely [11]. Bone appears to respond to a stimulus that recurs before the adaptation window closes, which makes consistency the active ingredient rather than volume. Fifty hops attached to an existing daily anchor, such as getting up from a desk or waiting for the kettle, is a more realistic design than a scheduled workout that will inevitably be missed on some days.
Two limits deserve emphasis before anyone treats hopping as sufficient. The magnitude is modest, and the benefit is regional: pooled analysis of impact exercise detected no significant lumbar-spine effect in any subgroup examined [13]. As a zero-cost daily habit it earns its place. As a standalone strategy it does not.
How Heavy Does Resistance Training Need To Be?
Meaningful gains require intensity, and the trial that established this recruited precisely the population usually steered away from heavy loading: postmenopausal women with osteopenia and low bone mass.
High-intensity resistance and impact training lifted lumbar-spine density by 2.9%. The low-intensity comparison arm, performing the cautious protocol conventionally recommended for fragile bone, lost 1.2% [14]. The gentle option was not merely less effective; participants following it went backwards at both spine and hip.
The prescription itself was compact: eight months, two sessions weekly, 30 minutes each. Deadlift, overhead press and back squat at 5 sets of 5 repetitions, loaded above 80 to 85% of one-repetition maximum, with optional lighter deadlift warm-up sets at 50 to 70%. Impact loading came from jumping chin-ups with drop landings, taking an underhand shoulder-width grip with shoulders and elbows at 90 degrees, jumping and pulling simultaneously, then dropping to land as heavily as comfort allows [14].
Supervision was not incidental to that safety record. Sessions ran at a maximum of eight participants per instructor, and every instructor held qualifications as both exercise scientist and physiotherapist. Across the intervention a single minor adverse event was recorded, a lower back spasm. Anyone adopting this protocol should have the lifts coached properly before loading them heavily, and anyone with established osteoporosis or a previous fracture should clear the plan with their doctor first.
Note what the intensity figure implies. Loading above 80% of a one-repetition maximum means a weight that permits only about five controlled repetitions, and as strength improves the absolute load has to keep climbing to stay in that zone. A fixed weight that felt demanding in month one is no longer a comparable stimulus by month six. That progression requirement is the practical bridge between a laboratory protocol and the heavier orbital device: in both cases the variable that produced results was the magnitude of force, deliberately kept high rather than allowed to drift down [7][14].
When Are Bone Medications Considered?
Where osteoporosis is already established, or loading alone proves insufficient, pharmacological options exist. The drug class NASA layered on top of heavy resistance training for its crews has an origin outside medicine altogether: the earliest applications of these compounds were industrial, used to stop scale and lime deposits forming in water pipes, boilers and detergents [15].
Chemistry that prevents mineral crusting inside plumbing turned out to bind avidly to skeletal mineral and blunt osteoclast activity. The therapeutic agents are not the identical molecules used industrially but tuned relatives of them, and the lineage runs directly from descaling chemistry to bisphosphonates.
Effect sizes here dwarf anything nutritional. Alendronate cut vertebral fracture risk by approximately 47% and roughly halved hip fractures among women with prior fractures [16]. Set against the roughly 1% density shift attributable to a calcium supplement, the difference is not marginal.
It is also worth being clear that medication and loading are not competing options. NASA did not swap the heavy resistance device for a drug; the drug was layered on top of it, specifically because even doubled mechanical loading left the trabecular compartment of the hip incompletely protected [7][10]. The same logic applies on the ground. Pharmacological treatment addresses the resorption side of the equation, while loading remains the only input shown to drive the formation side and to build the surrounding muscle that determines whether a fall happens in the first place.
Side effects require discussion with a prescriber. The most consequential to understand is osteonecrosis of the jaw, an uncommon condition featuring exposed non-healing jawbone, pain and infection, which is why dental health is assessed and any outstanding treatment completed beforehand.
The Evidence, Ranked
Ordered by demonstrated effect on bone, from weakest to strongest, the picture looks like this. Each rung is measured against the evidence cited above, and the gaps between them are substantial rather than incremental.
- Calcium and vitamin D supplements — no demonstrated reduction in fracture rates across community-dwelling populations [3][6]. Food sources preferred; supplementation reserved for documented deficiency or clinician direction.
- Daily hopping — modest and hip-specific, around 1 to 2% at the femoral neck, and only when performed every single day [11][12][13].
- Heavy progressive resistance training — the strongest non-pharmacological lever, roughly +2.9% at the lumbar spine, and requires competent supervision [14].
- Bisphosphonates — largest effect on fracture outcomes, around 47% fewer vertebral fractures, and a prescription decision made with a doctor [16].
Sequencing matters as much as the ranking itself. Adequate mineral intake through food is the floor, not the intervention; it permits bone to be built without being the thing that builds it. Loading is the intervention, and it is where effort returns the most. Medication occupies the top of the ladder for those whose bone is already fragile, and belongs in a conversation with a doctor rather than at the end of an article.
Three decades and a very large budget produced a conclusion that is simple to state and difficult to shortcut. Skeletal tissue answers to the forces placed through it, not to the mineral content of a capsule. Building bone and building muscle turn out to be the same assignment.
References
1. https://pubmed.ncbi.nlm.nih.gov/15125798/
2. https://pubmed.ncbi.nlm.nih.gov/18160467/
3. https://pubmed.ncbi.nlm.nih.gov/29279934/
4. https://pubmed.ncbi.nlm.nih.gov/18198394/
5. https://www.rnz.co.nz/news/national/289423/myth-breaking-bone-scientists-take-top-prize
6. https://pubmed.ncbi.nlm.nih.gov/29677309/
7. https://onlinelibrary.wiley.com/doi/10.1002/jbmr.1647
8. https://pubmed.ncbi.nlm.nih.gov/845205/
9. https://pmc.ncbi.nlm.nih.gov/articles/PMC3986122/
10. https://pubmed.ncbi.nlm.nih.gov/23334732/
11. https://pubmed.ncbi.nlm.nih.gov/20004758/
12. https://onlinelibrary.wiley.com/doi/10.1002/jbmr.3867
13. https://pubmed.ncbi.nlm.nih.gov/37555459/
14. https://pubmed.ncbi.nlm.nih.gov/28975661/
15. https://www.sciencedirect.com/science/article/abs/pii/S8756328211000652
16. https://pubmed.ncbi.nlm.nih.gov/8950879/
17. https://ods.od.nih.gov/factsheets/VitaminD-HealthProfessional/
