Do Calcium and Vitamin D Supplements Actually Prevent Fractures and Falls?
SEPTEMBER 23, 2026
Do calcium and vitamin D supplements actually prevent broken bones and falls in older adults? A systematic review published in The BMJ on 20 May 2026 — and back in wire coverage this month as ScienceDaily re-ran its summary on 20 September — pooled 69 randomised controlled trials covering 153,902 adults and found calcium alone, vitamin D alone, and the two taken together each produced little to no meaningful reduction in fractures or falls for most people. One combination technically cleared statistical significance. It still missed the line the reviewers had drawn, in advance, for a benefit worth caring about — and when they pulled out the single most influential trial behind it, that signal mostly evaporated too.
A Canadian Pharmacy Team Re-Pools Three Decades of Trials
The review was led by Olivier Massé and colleagues at the Department of Pharmacy, CIUSSS du Nord-de-l'Île-de-Montréal, in Québec — a hospital pharmacy research group, not a supplement manufacturer or a specialty osteoporosis society — and was registered in advance on PROSPERO (CRD42023483915). They searched Medline, Embase and the Cochrane trials register through 19 February 2025, plus the trial lists from earlier systematic reviews back to 2014, clinical trial registries, and conference abstracts, then pulled in every randomised trial that tested calcium, vitamin D, or both against placebo or no treatment in adults 18 and older. One eligibility rule matters more than it looks: trials of people already taking a prescription osteoporosis drug (a bisphosphonate, for instance) were excluded, because those drugs' own trials use calcium and vitamin D as routine background co-therapy — this review is about supplementation on its own, not as an add-on to a drug that's already doing the main work.
Two pairs of reviewers independently screened trials, extracted data and rated risk of bias with the Cochrane RoB2 tool; the certainty of every pooled result was then graded high, moderate, low or very low using GRADE. Across the 69 trials, participants were overwhelmingly ordinary: 87% lived in their own homes rather than a care facility, and 73% were not classified as being at high risk of a fracture or a fall to begin with. That matters for reading everything that follows — this is mostly a review of healthy-ish older adults taking a supplement as a precaution, not a review of frail, already-fracturing patients.
The Numbers, Three Ways
The primary outcome was risk of any fracture, and for that outcome all three approaches landed in the same place — a risk ratio close enough to 1.0 that none cleared the bar for a real effect:
- Calcium alone — 11 trials, 9,067 participants; risk ratio 0.91 (95% CI 0.81 to 1.01; moderate certainty). The confidence interval brushes but does not cross below 1.01, so this narrowly misses even statistical significance.
- Vitamin D alone — 36 trials, 92,045 participants; risk ratio 1.00 (0.95 to 1.06; high certainty). As flat a null result as a meta-analysis produces.
- Calcium and vitamin D combined — 15 trials, 51,126 participants; risk ratio 0.91 (0.84 to 0.99; high certainty). This one does cross the line into statistical significance — the upper bound of the interval sits at 0.99, just under 1.0.
The secondary outcomes tell a more textured story. For hip fracture specifically, calcium monotherapy showed an unexpected — and almost certainly spurious — increase in risk (6 trials, 6,703 participants; risk ratio 1.63, 0.86 to 3.03; very low certainty), a finding so imprecise and so biologically implausible that the authors themselves describe it exactly that way in their discussion, rather than treating it as a real signal. Vitamin D monotherapy's hip-fracture result sits right at the edge of concerning: 15 trials, 72,344 participants; risk ratio 1.13, and the lower bound of that interval is exactly 1.00 (1.13, 1.00 to 1.27; high certainty) — a result that just barely avoids statistical significance in the direction of more hip fractures, not fewer. Combined supplementation, by contrast, showed its clearest secondary benefit here: 9 trials, 49,909 participants; risk ratio 0.84 (0.74 to 0.96; high certainty), alongside a similar reduction in non-vertebral fracture (7 trials, 10,324 participants; 0.87, 0.78 to 0.96; moderate certainty). None of the three approaches meaningfully moved the risk of falling or the total count of falls.
Statistically Significant Is Not the Same as Worth Taking
This is the part of the paper that actually earns the headline "little to no benefit" despite those three statistically significant results for combined supplementation. Because no formally established minimal clinically important difference exists for fracture or fall prevention, the review team set their own thresholds in advance, by consensus, informed by a review of what clinicians and patients tend to value: an absolute risk reduction of at least 0.7% for hip fracture, 2% for any fracture, non-vertebral fracture or vertebral fracture, 3% for risk of falling, and five fewer falls per 100 person-years for the total-falls outcome. A result could pass a p-value test and still fail this one.
All three of combined therapy's statistically significant results did exactly that — passed the p-value test, failed the meaningfulness test. Converted into absolute terms, combined supplementation worked out to 1% fewer fractures at any site (a number needed to treat, or NNT, of 100 — meaning 100 people would need to take both supplements for one of them to avoid one extra fracture), 0.3% fewer hip fractures specifically (NNT 333), and 1.6% fewer non-vertebral fractures (NNT 63). Every one of those falls short of the pre-registered bar. The authors' own framing: "Calcium and vitamin D monotherapy did not significantly reduce any outcome, either statistically or clinically... [combined therapy's] absolute benefits for fracture prevention were, however, lower than the minimal clinically important differences established."
One 1992 nursing-home trial is still doing most of the work. A leave-one-out sensitivity analysis found that a single trial had a disproportionate influence on every one of combined therapy's positive results. Removing it moved the risk ratio for any fracture from 0.91 (0.84–0.99) to 0.96 (0.91–1.01), for hip fracture from 0.84 (0.74–0.96) to 0.88 (0.75–1.05), and for non-vertebral fracture from 0.87 (0.78–0.96) to 0.90 (0.78–1.05) — losing statistical significance on all three. That trial is Marie-Claude Chapuy's 1992 study in the New England Journal of Medicine, which gave 1,634 French women (mean age 84, living in residential care) daily tricalcium phosphate (1.2 g elemental calcium) plus 800 IU of vitamin D3, against 1,636 women on a double placebo, over 18 months. Those women weren't a typical supplement-taking population: their baseline vitamin D averaged just 20 nmol/L by modern lab standards (genuinely deficient), their dietary calcium intake averaged 513 mg a day (roughly half a typical adult target), and their three-year baseline hip-fracture risk ran around 11% — high enough that, per the new review's authors, "many of whom may have had unrecognised osteomalacia," a softening of the bones distinct from ordinary age-related bone loss. That 1992 trial is, in a real sense, still the whole modern case for calcium-and-D supplementation. Thirty-four years and 68 subsequent trials later, it has not been replicated at anything like the same size of effect.
What About People Who Are Actually Deficient?
The obvious objection to a flat null result is that it's being diluted by people who didn't need the supplement in the first place — the 73% not at high risk, people with normal vitamin D levels to start with. The review tested this directly, running subgroup analyses across baseline vitamin D level (below versus above 25 nmol/L, and separately below versus above 50 nmol/L), age 80 and older versus younger, residential care versus living independently, prior fracture or fall history, dose (under 1,000 versus 1,000–2,000 versus over 2,000 IU daily), and vitamin D2 versus D3. Across all of it, for all three interventions, the authors report no credible subgroup effects — the treatment didn't work meaningfully better in the people who, on paper, had the most to gain from it. The caveat is real and stated plainly: many of these subgroups drew on a small number of trials, and the review's own limitations section flags that finding as one to interpret cautiously rather than treat as final.
Where the Evidence Is Weak
Three limitations are worth naming honestly, because they cut in different directions. First, the minimal clinically important difference thresholds were the authors' own consensus judgment, not an externally validated standard — they say so themselves, noting they did not involve patient partners in setting them and that another team could reasonably draw the line somewhere else. Second, and pulling the opposite way, "almost all studies either allowed participants to take non-trial supplements or did not clearly report instructions for avoiding non-trial supplementation" — meaning some people in the placebo arms were likely taking calcium or vitamin D on their own, which would tend to shrink any true difference between groups rather than manufacture a fake one. Third, evidence specifically for people in residential care or otherwise at genuinely high risk remained thin for calcium monotherapy and for combined supplementation across several outcomes — this is a review that's confident about ordinary older adults and much less confident about the frailest ones, who are exactly the population the original 1992 trial targeted.
There are also real, if modest, downsides on the other side of the ledger that this review didn't measure directly but is worth knowing. Calcium supplements commonly cause gastrointestinal side effects — constipation, bloating, cramping — that make them poorly tolerated in older adults, and one study has suggested a possible increase in GI-related hospital admissions. In the Women's Health Initiative trial, combined calcium and vitamin D was linked to a higher rate of kidney stones, though the absolute increase was small (0.05%). Some data suggest a 10–20% relative (roughly 0.5–1% absolute) increase in heart attack risk associated with calcium supplements specifically, though other meta-analyses have found conflicting results on that point.
What the evidence says. Across 69 randomised trials and 153,902 adults, calcium supplements alone (risk ratio 0.91, 0.81–1.01) and vitamin D supplements alone (1.00, 0.95–1.06) produced no statistically significant reduction in fracture risk, and neither meaningfully reduced falls. Combined calcium and vitamin D did reach statistical significance for any fracture (0.91, 0.84–0.99), hip fracture (0.84, 0.74–0.96) and non-vertebral fracture (0.87, 0.78–0.96) — but the absolute benefits (1%, 0.3% and 1.6% fewer fractures respectively) fell below the review's own pre-specified thresholds for a clinically meaningful effect, and largely traced back to one 1992 trial in a high-risk, vitamin D-deficient nursing-home population that has not been replicated since at a similar scale. Extensive subgroup analysis found no evidence that people who were actually deficient at baseline, or otherwise at high risk, benefited more. The authors' own conclusion: "Based on absolute risk reductions and thresholds considered clinically meaningful, this review found little to no benefits from use of calcium, vitamin D, or combined supplementation on the prevention of fractures and falls," and they call on "clinicians, guideline panels, and regulatory agencies" to revisit routine recommendations. This does not extend to people taking a prescription osteoporosis drug, whose own trials rely on calcium and vitamin D as background therapy, or to people with a diagnosed bone disorder.
What I'd do. One person's reading of a meta-analysis, not medical advice to anyone. What stands out to me isn't that these supplements "don't work" — it's how much weight a single 34-year-old trial in a very specific population (elderly, institutionalised, genuinely vitamin D-deficient, badly under-eating calcium) is still carrying for a recommendation that now reaches almost everyone over 50. The review's number-needed-to-treat figures — 100 people supplementing for years to prevent one extra fracture anywhere, 333 for a hip fracture specifically — are the kind of numbers that are easy to round up to "it can't hurt" and easy to lose track of once they're translated into a guideline. The question this leaves me with isn't "does calcium and vitamin D work," it's "for whom" — and on this evidence, the honest answer for a typical, not-yet-deficient, not-institutionalised older adult is that the case is much thinner than decades of default prescribing would suggest.
Where I Could Be Wrong
- The clinically-meaningful thresholds are a judgment call, not a law of nature. The authors set them by consensus, without patient input, and say plainly that another team could reasonably choose different cutoffs — a smaller threshold would turn some of these "little to no benefit" verdicts into "modest benefit."
- Contamination in the placebo arms likely biases the whole review toward the null. If people assigned to placebo were independently taking calcium or vitamin D — and the authors say most trials didn't clearly prevent or report this — the true effect of supplementation could be somewhat larger than what this review measured.
- The high-risk and institutionalised population is exactly the group with the thinnest evidence here. Both the group most similar to the 1992 trial that's still driving the positive signal, and the group current guidelines most specifically target, had limited data in this review — the null finding is strongest for people who probably needed the supplement least.
- This review can't and doesn't speak to people on osteoporosis drugs. Those trials use calcium and vitamin D as routine co-therapy, which is a different clinical question from "should a healthy older adult take these on their own."
- Subgroup analyses are underpowered by nature. A "no credible subgroup effect" finding, especially for baseline vitamin D deficiency, rests on fewer trials than the headline result and could look different with more data.
- This is one systematic review, however large. It supersedes some earlier positive reviews by including newer, larger vitamin D trials, but it is itself subject to revision by the next one.
Sources
- Massé O, Mercurio CM, Dupuis S, Al Sahwi M, Arruda A, Dallaire G, Desforges K, Dugré N, Williamson D. Calcium, vitamin D, or combined supplementation to prevent fractures and falls: systematic review and meta-analysis. BMJ, 393:e088050, 20 May 2026. doi:10.1136/bmj-2025-088050 · PMC13188451
- Pillay J, Reynolds DL, Thériault G. Calcium, vitamin D, or combined supplementation to prevent fractures and falls. BMJ editorial, 393:s913, 20 May 2026. doi:10.1136/bmj.s913
- BMJ Group. Calcium and vitamin D supplements offer little to no meaningful benefit on fracture and fall prevention. Press release, 20 May 2026. bmjgroup.com
- ScienceDaily. Millions take calcium and vitamin D for bone health. A major review finds little benefit. Re-run, 20 September 2026. sciencedaily.com
- Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 and calcium to prevent hip fractures in elderly women. New England Journal of Medicine, 327:1637-42, 3 December 1992. doi:10.1056/NEJM199212033272305
- Chapuy MC, Pamphile R, Paris E, et al. Combined calcium and vitamin D3 supplementation in elderly women: confirmation of reversal of secondary hyperparathyroidism and hip fracture risk: the Decalyos II study. Osteoporosis International, 13:257-64, 2002. doi:10.1007/s001980200023
This is one reader's reading of the research, not medical advice. If something here touches on your own health, take it to a clinician who knows you — and read how these entries are put together.
