Does vitamin D prevent fractures?

Disclosure up front: I run a supplement company, so I have a general stake in how people think about this whole category. The full disclosure is at the bottom. This report is about a review of other people's trials of vitamin D and calcium.

The paper is by Olivier Massé and colleagues, clinical pharmacists and pharmacy researchers in Montréal, published in The BMJ on 20 May 2026. It's open access, the protocol was registered in advance (PROSPERO CRD42023483915), and the authors report no funding for the work. Here is what they did and what they found.

What they pooled

They gathered randomized controlled trials comparing calcium, vitamin D, or both against placebo or no treatment in adults who weren't on osteoporosis drugs. That gave them 69 trials with 153,902 participants. The median trial ran 2 years, and the median participant age across trials was about 71.

Keep the population in mind, because it decides who this answer applies to. In 87% of the trials, everyone lived in the community rather than in residential care. Only 19 of the 69 trials studied high-risk groups, which the authors defined as an average age of 80 or more, living in residential care, a past fracture or fall, diagnosed osteoporosis, or average vitamin D below 25 nmol/L at the start. Only 2 trials started with an average vitamin D level below that deficiency cut-off.

The three answers

Vitamin D alone: no effect. Across 36 trials and about 92,000 people, the risk ratio for any fracture was 1.00 (95% CI 0.95 to 1.06). In plain terms, 6.5% of people had a fracture whether they took vitamin D or placebo. Falls were the same story: a risk ratio of 1.01 across 32 trials. The authors rated both findings high certainty.

Calcium alone: likely little to no effect. Risk ratio 0.91 for any fracture (0.81 to 1.01) across 11 trials and 9,067 people, moderate certainty. The hip-fracture estimate was very uncertain and, if anything, leaned the wrong way, which the authors call "imprecise and uncertain and biologically implausible."

Vitamin D plus calcium: this is the one to look at closely. Risk ratio 0.91 for any fracture (0.84 to 0.99) across 15 trials and 51,126 people, high certainty. That confidence interval stays below 1.00. By the usual test, it worked.

Why "significant" still lost

Turn the risk ratio into people. In the control groups, 11.3% had a fracture. With combined supplements, 10.3% did. That's 1 fewer fracture per 100 people treated over a couple of years, a number needed to treat of 100. For hip fractures specifically, it was 0.3 fewer per 100 (treat 333 people to prevent one). For non-vertebral fractures, 1.6 fewer per 100.

Before looking at the data, the team wrote down how big a benefit had to be before they would call it clinically important: a 2% absolute drop for any fracture, 0.7% for hip fracture, 3% for the risk of falling. Every combined result came in under its bar. So the verdict is "little to no effect," even with p-values that would make a press release.

That threshold is a judgment call, and the authors say so. They set it by consensus after reviewing the literature on what clinicians and patients value, and they write that they "defined arbitrary absolute effect thresholds," and that other teams "may have alternative definitions of what is clinically meaningful." Fair enough. You can decide for yourself whether 1 fewer fracture per 100 people is worth a daily pill. What you can't honestly do is call it "prevents fractures" without the size attached.

Most of the benefit came from one 1992 trial

When the team removed trials one at a time to see whether any single study was steering the pool, one was. Without it, the combined risk ratio for any fracture moved from 0.91 to 0.96 (0.91 to 1.01) and was no longer statistically significant. The same thing happened to hip and non-vertebral fractures.

That trial is the original French study published in 1992 in the New England Journal of Medicine, the one that launched much of the enthusiasm. The review describes who was in it: women with an average age of 84, living in residential care, with a high risk of hip fracture, very low vitamin D (about 20 nmol/L by modern assays) and low calcium intake (about 513 mg a day). The authors note many "may have had unrecognised osteomalacia," a bone-softening condition from severe deficiency. That's not the 65-year-old buying a bottle at the pharmacy.

The pattern I see here is common in supplement science. A striking early result in a very specific, very depleted population becomes a general recommendation, and later trials in healthier people don't reproduce it.

What this review doesn't settle

Treating a real deficiency. Only 2 of 69 trials started with average vitamin D below 25 nmol/L. The authors found no subgroup or dose effect and no link between blood levels and benefit, but the evidence in people who are truly deficient is thin. This review answers "does routine supplementation help?" It isn't built to answer "should a person with a diagnosed deficiency be treated?" That's a question for a clinician with the lab result in hand.

People on osteoporosis drugs. The trials of those drugs almost always give calcium and vitamin D as background treatment, which is why guidelines recommend them alongside. The review excluded people on those drugs and says its findings don't apply to them, or to people with specific bone disorders or on long-term corticosteroids.

High-risk and residential-care groups. The authors write that for combined supplements in those groups, their absolute numbers "may somewhat underestimate benefits," because the evidence there is limited.

Trial quality. Most of the individual trials were judged at high risk of bias or with some concerns. Almost all let participants keep taking their own supplements or didn't clearly forbid it, which would push results toward no difference. The authors flag both. The GRADE certainty ratings still came out moderate to high for most findings, largely because the big recent trials agree.

This isn't a new conclusion

The 2018 evidence review for the U.S. Preventive Services Task Force reached a similar answer for community-dwelling adults without deficiency, osteoporosis or a past fracture, and also reported a small increase in kidney stones with combined vitamin D and calcium (I read that abstract only). What the 2026 review adds is size: several large trials published since then, including VITAL, D-Health and DO-HEALTH, more than 54,000 new participants, and an explicit test of whether the "significant" combined result is big enough to matter.

How I'd read it

For a generally healthy adult not on osteoporosis treatment, taking vitamin D to prevent fractures or falls is not supported by the best trial evidence we have. Vitamin D alone did nothing measurable. Vitamin D with calcium did something statistically, about 1 fracture per 100 people, mostly carried by one old trial in severely deficient nursing-home residents. The authors' bottom line is that these findings "do not support routine supplementation" for fracture or fall prevention.

Their review also points at what does have evidence for falls: the Cochrane work they cite finds exercise and home environmental changes "appear promising." If bones are the goal, that's where I'd start the conversation with a clinician, not in the supplement aisle.

And the reusable lesson: when a supplement study says "significant," ask for the absolute number. "Reduced risk by 9%" and "1 fewer case per 100 people" can describe the same result.

Not medical advice. This is educational analysis, not a recommendation — a study is not a prescription. Talk to a qualified clinician before acting on anything you read here. Full disclaimer →

Disclosure, plainly: I founded and run Shroombiosis (a company I run), which formulates and sells supplements. That's a stake in how people view supplements generally. I have no relationship with the review's authors or with any vitamin D or calcium maker named in the trials. Nothing here is sponsored and no link earns a commission; here's the full policy. A recommendation with no stake at all: for performance nutrition, Die Tryin Co. is a fellow combat-veteran-owned brand I'm glad to point people to. I don't own it and earn nothing from the link.