For most readers this resolves to a single decision: don't buy it. The over-the-counter form has no fracture-prevention evidence in the bottle on the shelf, the prescription form was withdrawn for a cardiovascular safety signal, and the bone scan you'd use to check whether it's working is the one signal strontium specifically corrupts. If a specialist has put you on it because everything else failed, that's a different conversation; this entry is about the bottle in the supplement aisle.
The drug that ran the trials was called strontium ranelate, sold in Europe as Protelos. Two large studies in postmenopausal women with established osteoporosis showed real fracture reductions over three years — about 41% fewer spinal fractures in the higher-risk trial (Meunier 2004), about 16% fewer non-spinal fractures in the broader one (Reginster 2005). Hip-fracture reduction only showed up in a high-risk subgroup of the broader trial, not in the trial population overall. For comparison, the bisphosphonates already on pharmacy shelves cut spinal fractures by 40 to 70%, with a much longer safety record.
Now the catch on the bone-scan numbers. Strontium's atomic number is 38; calcium's is 20. The DXA scanner measures how much X-ray your bones absorb, and the heavier atom absorbs more per unit mass. A bit of strontium in newly forming bone reads as roughly 2.5 times as much bone density as the same mass of calcium would (Blake and Fogelman 2007). The trials reported a 12.7% lumbar-spine BMD gain over three years; corrected for the artefact, the actual bone-mineral gain is roughly half that (Pors-Nielsen 1999). The fracture-rate reductions are real — those are clinical events, not surrogate readings. But the bone-scan number is partly the X-ray catching a heavier ion, and most clinical scanners and most clinicians do not apply the correction.
And the gap nobody flags in the supplement aisle: the drug, strontium ranelate, was tested. The supplement, strontium citrate, was not — not in a phase III fracture trial, not at any dose. The pivotal trials used ranelic acid as the carrier; the supplement world transplants the efficacy onto a different salt and a different regulatory regime as if the chemistry made no difference. Maybe it does, maybe it doesn't — nobody has run the trial. What's certain: the over-the-counter pill marketed as "the strontium that works" has no broken-bone evidence in the form on the shelf.
What strontium does in bone
Strontium and calcium are chemical cousins — same column of the periodic table, both divalent, similar size. Your body already carries trace strontium in bone, a few hundredths of a percent of skeletal mineral, picked up from food and water (Nielsen 2004). At supplement-level doses, more of it gets absorbed and slots into the hydroxyapatite crystal that makes up the mineral of bone — the same lattice that calcium normally fills.
What it does, mechanistically, is uncouple the two halves of bone remodelling: lab and animal data show a small push on the cells that build bone (osteoblasts) and a small brake on the cells that break it down (osteoclasts), through a receptor on both that normally senses calcium (Marie 2001)(Saidak and Marie 2012). That dual action was the original marketing story — build-up and slow-down, mechanistically novel compared to the pure brakes (bisphosphonates) and pure throttles (teriparatide) on the rest of the osteoporosis shelf.
In actual humans, on actual doses, the effect is modest. Bone biopsies from women on strontium ranelate showed only small changes in resorption and only minor formation increases (Saidak and Marie 2012). Strontium-substituted bone is not mechanically identical to calcium-only bone, and the substitution concentrates in newly formed bone rather than spreading evenly through the skeleton. None of that is catastrophic at the doses studied. But the gap between "uncouples remodelling in a petri dish" and "produces meaningfully more functional bone in a 70-year-old woman" is wider than the marketing suggests.
Three traps the marketing relies on
"My bone scan went up — my bones are stronger." They might be, a little, but not by what the scan says. About half the apparent gain on strontium is the X-ray attenuation artefact, not new bone (Blake and Fogelman 2007)(Pors-Nielsen 1999). If you've been buying based on watching the number climb, you're watching the wrong number. And once strontium is in your bones it stays for years, so the next scan after you stop is still partly reading the residue.
"Strontium citrate is the same as the drug, minus the patent." No phase III fracture trial has ever been run on strontium citrate at any dose. The pivotal trials studied a different salt — ranelic acid as the carrier. The supplement industry markets the trial data as if the salt didn't matter. Maybe it doesn't, but nobody has tested the form on the shelf, and the regulatory bar for supplements doesn't require them to.
"Take it with calcium — they're both bone minerals." They compete for absorption in the gut. Taking them together cuts strontium uptake and may modestly cut calcium uptake. Even the supplement labels recommend separating them by at least two hours (Reginster 2003); in real life, compliance is mixed. If you're going to take strontium, you can't dose it with your calcium-fortified breakfast.
The cardiovascular signal that pulled the drug
In 2014, European regulators reviewed pooled data from the strontium ranelate trials and found an increased rate of heart attacks in treated patients — a relative risk around 1.6, with an absolute increase of about half a percentage point over five years (EMA 2014). Venous blood clots ran somewhat higher too. The agency restricted prescribing to severe osteoporosis without cardiovascular risk factors; in August 2017, the manufacturer withdrew the drug entirely (Servier 2017). It had never been approved in the United States.
Routine-practice safety data muddied the picture. A Danish registry found that strontium ranelate had been preferentially prescribed to patients already at higher cardiovascular baseline — the higher event rates reflected the population at least in part (Abrahamsen and colleagues 2014). A UK case-control analysis didn't find a statistically significant heart-attack elevation in everyday clinical use (Cooper 2014). The trial-level signal is what survived regulatory review, though: randomization handles the confounders the registries can't.
Whether the over-the-counter citrate carries the same risk, nobody knows. The signal may be specific to ranelic acid, or it may be intrinsic to the strontium cation. No cardiovascular outcome trial has been run on citrate, in anyone, at any dose. The conservative read is that if the cation contributes at all, the OTC supplement does too — at the long timescales people actually take it, in a population that has never been studied.
What actually moves the needle on bones
For someone with osteoporosis on a real bone-density scan, the first-line drugs are oral bisphosphonates — alendronate, risedronate — both generic, both with decades of safety data, both with spinal fracture reductions of 40 to 70% and meaningful non-spinal effects. Denosumab is a twice-yearly injection with similar efficacy; teriparatide and abaloparatide are anabolic agents for severe cases. None of these carries the strontium cardiovascular signal, and all of them have larger fracture-prevention effect sizes than what strontium produced in its best trials.
For someone without osteoporosis trying to keep what they have: mechanical load is the strongest signal bone has access to — resistance training, hopping, jumping, anything that lands hard. Enough protein. Calcium and vitamin D sufficiency, not megadoses. Don't smoke. Don't overdrink. None of this is exciting and none of it costs forty dollars a month at the supplement shop, which is the honest reason the strontium aisle exists.
What happens if you keep taking it
The common failure isn't dramatic. It's the woman who has been on strontium citrate for two years, watches her bone scan climb, reads the climb as her bones improving, and tells her doctor she'd rather skip the bisphosphonate. The climb is partly the heavier-atom artefact (Blake and Fogelman 2007). Her bones, under the strontium reading, are tracking with her untreated osteoporosis. The first signal she gets that something is wrong is a wrist fracture from a low fall, or worse, a hip on the kitchen floor. The supplement bought her the feeling of having done something; it didn't buy her bone. A bisphosphonate would have cut her hip-fracture rate roughly in half over five years.
The quieter failure is the cardiovascular one. The trial-level absolute increase on the drug was small — about half a percentage point of extra heart attacks over five years (EMA 2014). That's the kind of risk you can carry for a decade without ever feeling it. If the strontium cation contributes to the signal — and nobody has ruled it out — every month on the supplement is a month adding the unstudied piece to whatever cardiovascular baseline you walked in with. People at low baseline risk may never know one way or the other. People at high baseline risk — a stent in the past, a stroke, controlled hypertension, a strong family history — should not be running that experiment on themselves to find out.
Adjacent topics, if you came here through the bone-health door: osteoporosis itself (who needs which drug, when to start); DXA scans (what the number means and what corrupts it); calcium and vitamin D (sufficiency versus the megadose habit); resistance training for bone density; bisphosphonates and denosumab (the actual first-line drugs).
Substance and claimed effects
Strontium is a divalent trace mineral chemically similar to calcium (group 2, atomic numbers 38 vs 20). It is naturally present at trace levels in human bone — roughly 0.035% of skeletal mineral mass — substituting for calcium in the hydroxyapatite lattice (Nielsen 2004). At supraphysiological doses it concentrates in newly forming bone, where it slightly uncouples bone remodelling: in vitro and animal data show stimulation of osteoblast differentiation and inhibition of osteoclast resorption (Marie et al. 2001), (Saidak and Marie 2012).
Two forms exist in clinical reality, and conflating them is the central editorial problem:
- Strontium ranelate — pharmaceutical form (Protelos/Osseor), 2 g/day, approved in the EU from 2004, restricted by EMA in 2014 on cardiovascular grounds, and commercially withdrawn by Servier in 2017 (EMA 2014), (Servier 2017). Never FDA-approved. The only form of strontium with phase III RCT data on fracture endpoints.
- Strontium citrate (or chloride / gluconate) — supplement form, OTC in the US, Canada, parts of Asia and Europe, typically 340–680 mg elemental strontium per capsule. No phase III fracture-endpoint RCTs. Marketing routinely cites the ranelate trials.
Consequences this entry covers holistically, per the input brief: DXA-measured bone density, fracture risk, cardiovascular risk, and calcium metabolism. Each is treated below.
Evidence by addressing question
mechanism
Strontium ions (Sr2+) substitute for calcium in hydroxyapatite during bone formation. In rodent and cell-culture models, strontium ranelate increases pre-osteoblast replication and collagen synthesis while reducing osteoclast differentiation and resorption — a dual effect distinct from bisphosphonates (pure anti-resorptive) and teriparatide (pure anabolic) (Marie et al. 2001), (Saidak and Marie 2012). The mechanism is partially mediated through the calcium-sensing receptor (CaSR) on osteoblasts and osteoclast precursors.
Caveat: in vivo human histomorphometry showed only modest changes in resorption markers and only minor bone-formation increases at the clinical 2 g/day ranelate dose (Saidak and Marie 2012). The mechanism is dose-dependent in animal models, with high doses producing mineralization defects (osteomalacia-like changes). Strontium-substituted apatite is mechanically slightly different from pure calcium apatite, and substitution is heterogeneous across the skeleton, concentrated in newly remodelled bone (Nielsen 2004).
evidence
Ranelate fracture endpoints. Two phase III trials. SOTI (n=1,649 postmenopausal women with ≥1 prior vertebral fracture, mean age 70): 41% relative risk reduction in new vertebral fractures at 3 years (absolute 20.9% vs 32.8% in placebo) (Meunier et al. 2004). TROPOS (n=5,091, broader inclusion): 16% relative reduction in non-vertebral fractures at 3 years; a 36% hip-fracture reduction in a post-hoc high-risk subgroup (women ≥74 years with femoral neck T-score ≤ −3), but no significant hip-fracture reduction in the overall trial population (Reginster et al. 2005). The 5- and 8-year open-label extensions reported maintained antifracture efficacy versus historical placebo, though the uncontrolled design weakens that inference (Reginster et al. 2008). Pooled meta-analysis: ~37% relative reduction in vertebral fracture and ~14% in non-vertebral fracture (Kanis et al. 2008). For context, oral bisphosphonates achieve 40–70% vertebral reductions and similar or larger non-vertebral effects with a much stronger long-term safety record.
The BMD effect is partly artefactual. Strontium has an atomic number of 38 versus calcium's 20; it attenuates the DXA X-ray beam more strongly per unit mass. A given mass of bone-incorporated strontium reads as roughly 2.5× the mass of calcium would on a DXA scan (Blake and Fogelman 2007), (Pors-Nielsen et al. 1999). The SOTI trial reported a 12.7% lumbar BMD gain at 3 years; correcting for X-ray attenuation, the true bone-mineral gain is roughly 5–7% — half what the raw number says. Most clinical DXA scanners do not apply a strontium correction; most clinicians do not request one. The implication: a patient on strontium who tracks BMD is reading a number whose meaning has changed and whose comparability with their pre-strontium baseline is broken.
Citrate fracture endpoints — none. No phase III RCT has been published on strontium citrate at any dose, with fracture as the endpoint. A handful of small open-label or observational studies (typically n<50, often without controls) report BMD increases — increases which, as above, are partly artefactual. The marketed claim "strontium reduces fracture risk" derives from the ranelate trials; transferring it to citrate assumes pharmacokinetic and safety equivalence that has not been demonstrated.
contraindications
The EMA's 2014 restriction excluded ranelate from use in: established ischaemic heart disease, peripheral arterial disease, cerebrovascular disease, uncontrolled hypertension, history of venous thromboembolism, and immobilization (EMA 2014). The restriction followed a pooled re-analysis of pooled clinical trial data showing increased myocardial infarction (relative risk ~1.6; absolute event-rate increase ~0.4 percentage points over 5 years in the pooled placebo-controlled population) and VTE (relative risk ~1.4). Post-marketing surveillance documented rare but serious cutaneous reactions including DRESS (Drug Reaction with Eosinophilia and Systemic Symptoms) and Stevens-Johnson syndrome.
The post-marketing CV picture is mixed. A Danish nationwide registry analysis found that strontium ranelate was preferentially prescribed to higher-CV-risk patients (channelling bias) and that adjusted CV event rates were elevated (Abrahamsen et al. 2014). The UK CPRD nested case-control analysis by Cooper et al. did not find a significant increase in MI in routine clinical use after adjusting for confounders (Cooper et al. 2014). The trial-level signal is the firmer estimate (randomization handles confounding); the regulatory restriction held.
Strontium citrate has not been studied for cardiovascular outcomes at any meaningful scale. Whether the CV signal is specific to ranelic acid (the organic carrier) or intrinsic to the strontium cation remains unresolved. The conservative position: extend the cardiovascular contraindications to citrate until citrate-specific safety data exist.
misconceptions
Three biases worth naming.
- "My DXA went up X%, so my bones are X% stronger." They are not. X-ray attenuation by the heavier atom inflates the reading; without strontium correction, the apparent BMD increase overstates the true mineral gain by roughly 2× (Blake and Fogelman 2007). The fracture-rate evidence (which is a real clinical endpoint, not a surrogate) is the firmer ground; the DXA number is misleading.
- "Strontium citrate is the same as the trial drug minus the patent." The trials studied ranelic acid as the carrier; no phase III fracture trial exists on the citrate salt. Marketing that points to SOTI/TROPOS as efficacy support for OTC citrate is making an unjustified transfer of evidence across formulations.
- "Strontium and calcium are both bone minerals — take them together." They compete for intestinal absorption. Co-administration reduces strontium uptake and may modestly reduce calcium uptake. Even the supplement protocols recommend at least 2 hours of separation; compliance is mixed.
alternatives
For osteoporosis prevention or treatment, alternatives with stronger evidence and cleaner safety profiles dominate: oral bisphosphonates (alendronate, risedronate — first-line, generic, decades of safety data), denosumab (twice-yearly subcutaneous injection, similar fracture-reduction efficacy, requires uninterrupted continuation to avoid rebound), zoledronic acid (annual IV bisphosphonate), and the anabolic agents teriparatide and abaloparatide for severe disease. None of these carries the strontium cardiovascular signal; all have larger fracture-prevention effect sizes than strontium ranelate's pooled estimate. Non-pharmacologic anchors: weight-bearing and resistance exercise (mechanical load is the strongest osteogenic signal), adequate dietary protein, calcium and vitamin D sufficiency, smoking cessation, moderate alcohol.
practicalities
Strontium ranelate is no longer commercially available in the EU (Servier withdrew Protelos/Osseor in August 2017) (Servier 2017). Never approved in the United States. OTC strontium citrate is widely sold in the US, Canada, and online at roughly $20–50 per month for 340–680 mg/day. Classified as a dietary supplement; not subject to FDA fracture-outcome efficacy review.
If used, dose timing matters: away from calcium-containing meals, separated by at least 2 hours (Reginster et al. 2003). Bone strontium accumulates and persists for years after discontinuation, so DXA readings remain contaminated long after stopping.
stakes
Two failure modes. First: the postmenopausal woman who reads her inflated DXA as evidence the supplement is working, declines or delays an evidence-backed therapy (bisphosphonate or denosumab), and fractures a hip or vertebra at the rate her untreated disease predicts. The artefactual reading provided false reassurance and crowded out a real intervention. Second: the patient with subclinical cardiovascular disease whose risk is elevated by chronic strontium intake without their knowing — the trial-level absolute MI increase is small but real (EMA 2014), and the OTC citrate population has been entirely unstudied.
payoff
For the avoid framing, the payoff is what the reader gets back. Money: $250–600/year not spent on a supplement with no fracture-endpoint evidence in its sold form. Information: an uncontaminated DXA reading that actually tracks their bone health rather than the X-ray attenuation of an extra ion. Risk reduction: zero contribution to a small-but-real cardiovascular signal that the EMA found worth restricting an entire drug for. And, when paired with the evidence-backed alternatives, real fracture prevention rather than the suggestive ~14% non-vertebral / ~37% vertebral effect that ranelate produced at the cost of CV exclusion criteria.
The credibility range
Optimist case
Two large phase III RCTs (SOTI, TROPOS) showed real fracture-endpoint reductions on vertebral and (in a high-risk subgroup) hip endpoints. Mechanism is plausible and supported by molecular biology (CaSR-mediated uncoupling). Eight-year open-label data continued to show effect. Strontium citrate at supplement doses has no acute toxicity signal. For a patient who genuinely cannot tolerate bisphosphonates or denosumab and who has clean cardiovascular risk, strontium remains a defensible third-line anti-fracture option. The DXA artefact is real but doesn't erase the fracture benefit — that was measured against clinical events, not surrogates.
Skeptic case
Once the X-ray attenuation artefact is corrected, the ranelate trials' BMD gains shrink by half; the fracture reductions are smaller than bisphosphonates' and were achieved with a cardiovascular safety signal strong enough that the EMA restricted the drug to severe disease without CV risk factors, after which the manufacturer voluntarily withdrew it. No fracture-endpoint trial has ever been run on strontium citrate at any dose; transferring efficacy from ranelate to citrate is a marketing leap unsupported by the salt-specific pharmacokinetic or safety data. The supplement market exists because the science was just suggestive enough to anchor a story, the regulatory bar for supplements doesn't require fracture trials, and inflated DXA readings give buyers a feedback signal that confirms the purchase without reflecting biology. This is a textbook biomarker-not-outcome trap.
Author's call
For nearly every reader: skip it. The pharmaceutical form was withdrawn for genuine safety reasons. The supplement form has no fracture-endpoint evidence at all in the form sold and probably carries an unknown share of the same CV risk. Where bisphosphonates, denosumab, or anabolic therapy are options, strontium is dominated on both efficacy and safety. The DXA inflation alone is reason enough to avoid quietly degrading the reader's ability to track their own bone health. The narrow exception — severe established osteoporosis, documented intolerance of all first- and second-line agents, clean cardiovascular risk profile, and clinician supervision — is a decision for an endocrinologist, not a self-prescription from an online shop. Action class for the entry: avoid (with a small decide caveat for the narrow exception).
Stakeholder and incentive map
- Supplement manufacturers — strontium citrate is a high-margin OTC product; marketing routinely cites the ranelate trials without flagging the salt difference. Low regulatory bar in the US.
- Servier (originator of strontium ranelate) — funded SOTI/TROPOS, restricted-then-withdrew after the EMA review; commercial pressure and safety regulation converged.
- EMA, FDA, MHRA — EMA restricted then effectively removed ranelate; FDA never approved it; MHRA aligned with EMA. The regulatory verdict on the pharmaceutical was negative.
- Endocrinology/rheumatology societies (AACE, NOF/BHOF, ESCEO) — guidelines uniformly recommend bisphosphonates or denosumab as first-line; strontium does not appear as a recommended option in current US-facing osteoporosis pathways and has dropped out of EU pathways since 2017.
- Alternative-medicine / wellness ecosystem — strontium citrate has become a "what they don't want you to know" supplement in some subcultures; the FDA's non-approval is framed as suppression of a working bone therapy rather than as absence of US trials. This framing drives a meaningful share of consumer demand.
- DXA equipment vendors — strontium correction is rarely implemented at the scanner-software level; the field has not standardized a workflow for it.
Population variability
Postmenopausal women with established osteoporosis (mean age ~70 in SOTI) were the only population studied at fracture-endpoint scale. The hip-fracture effect was confined to the very high risk subgroup of TROPOS (women ≥74, femoral neck T-score ≤ −3); the overall TROPOS hip endpoint was null (Reginster et al. 2005). No adequate fracture-endpoint data on men, premenopausal women, glucocorticoid-induced osteoporosis, or younger populations using strontium "for longevity" or "for prevention" — these uses are extrapolation.
Cardiovascular risk concentrates in the patients already at baseline risk; the EMA restriction was effectively a higher-baseline-risk exclusion. Strontium absorption pharmacokinetics resemble calcium's — reduced by low gastric acid and competing dietary calcium intake. Strontium accumulates in bone and is cleared slowly; bone strontium remains detectable for years after stopping, so DXA tracking remains contaminated well past the discontinuation date (Nielsen 2004).
Knowledge gaps
- No fracture-endpoint RCT has ever been run on strontium citrate at any dose. The supplement industry's headline claim — strontium reduces fractures — has never been tested in the form it is sold in.
- Whether the cardiovascular signal is specific to ranelic acid or intrinsic to the strontium cation. Mechanistic studies are limited; the conservative interpretation extends the warning to citrate.
- The strontium-corrected DXA factor is not implemented at the scanner level in routine practice. Patients on strontium are getting uninterpretable readings, and the manufacturers do not surface this in their marketing.
- Long-term bone quality (beyond ~8 years of exposure) is unknown. Animal data at high doses show osteomalacia-like changes; whether chronic low-dose supplement exposure has any similar mechanical disadvantage has not been characterized.
- Whether any benefit holds for users with normal BMD ("for longevity" or "for prevention" use). The trials studied established osteoporosis only; transfer to healthy adults is not evidence-based.
Scope vs brief. The brief named four consequences (bone-density readings, fracture risk, cardiovascular risk, calcium metabolism). All four are covered: the DXA-artefact picture in evidence and misconceptions; the fracture-endpoint trial picture in evidence; the EMA cardiovascular signal in contraindications and stakes; the calcium-absorption interaction in misconceptions and (briefly) practicalities-equivalent material in the evidence section. No silent narrowing.
The citrate vs ranelate scoping decision. The article treats them as one substance with two forms, because the supplement consumer's marketing-and-decision context is ranelate trials being used to sell citrate. Splitting into two entries was considered and rejected — the central editorial message (the trials are on the wrong salt, and the right-salt trials were pulled for safety) needs both forms in one frame.
Evidence score = 2 — the hard call. In isolation, the ranelate evidence is closer to a 4 (two large phase III RCTs, positive fracture endpoints, regulatory restriction). The citrate evidence is closer to a 1 (no fracture-endpoint trials at any dose). The entry's evidence dimension reflects the substance the reader actually buys, which is citrate; the ranelate evidence is mechanistically suggestive only when transferred. Scored at 2: sparse-or-contested literature with plausible mechanism. A reviewer who disagrees and wants this scored on the ranelate body of work would push it to 3.
Action = avoid (not decide). The narrow case where strontium might be a clinician-supervised third-line option (severe osteoporosis, intolerance to bisphosphonates/denosumab/anabolics, clean CV risk) is small enough to call out in the body rather than make it the default action class. For the reader walking into a supplement aisle, the answer is avoid, full stop.
Cardiovascular signal — settled enough vs contested? The trial-derived MI signal (Cooper 2014's null finding notwithstanding) was strong enough for EMA restriction and the manufacturer's withdrawal. The article treats this as the firmer evidence, with the contested observational picture noted. A reviewer who reads Cooper 2014 as the stronger signal will want softer language; I do not.
The DXA-artefact magnitude. The "roughly 50%" figure is widely cited in densitometry literature and traces to Blake/Fogelman and Pors-Nielsen. Exact correction factor varies by scanner and bone-strontium concentration; the article uses "roughly half" to avoid false precision. Reviewer note: if a more precise scanner-specific correction has been published since 2007 it should replace the round number.
Applicability scored at 3, not 2. Postmenopausal women are a whole sex/age band (over half have low bone mass), and the decision audience extends to anyone shopping bone supplements after a DXA, family of fracture survivors, and women approaching menopause. The avoid-entry stakes-break-ties guardrail was considered but not used to push higher — the wider audience is real, not inflated by smoking-style decision-audience logic.
Future-link candidates (none exist in catalogue yet). Osteoporosis as a condition entry; DXA scans (what the number means, what corrupts it); bisphosphonates (the actual first-line); denosumab; resistance training for bone density; calcium and vitamin D sufficiency. The out-of-scope closing already signposts these as adjacent topics.
Dream narrative — written though score is well below 40. The relief lever fits an avoid/myth-busting entry; writing it gave the dek and tagline a cleaner aim ("the number means something") than a straight-write would have. Optional, not obligatory at this score.
No men-specific audience scoping. Trials studied postmenopausal women only, but the decision (whether to buy the supplement) faces all adults reading bone-health marketing. Audience left unscoped on the meta; the population-specificity is in the article body.
Strontium for Bone Density
One capsule a day, taken at least two hours away from anything with calcium in it.
Around $20–50 a month for a supplement that has never been tested for whether it actually prevents broken bones.
The prescription version had real trials and got pulled for heart-attack risk. The over-the-counter version has no real trials at all.
Maybe trims a sliver of fracture risk in older women, on the drug version that got pulled from the market. The pill version sold over the counter has never been tested for fracture prevention.