Science Corner 74 | Calcium just did a U-turn
On SuppCo, calcium is stacked over 219,000 times across almost 38,000 products. But here are a few numbers that might be even more important...
153,902 adults. 69 randomized trials… And no drop in fractures big enough to matter.
That is the plain reading of the largest analysis ever conducted on calcium and vitamin D supplementation, published in The BMJ this past June. A Canadian research team led by Olivier Massé pooled six decades of trial data and found that calcium alone, vitamin D alone, and the two combined all failed to clear a pre-set bar for clinically meaningful protection against fracture or falls in adults over 60 who live independently.
None of this should surprise the U.S. Preventive Services Task Force. Back in 2018, the agency already told postmenopausal women that low-dose calcium and vitamin D did not prevent fractures. Men, and anyone taking higher doses, were the open question and a December 2024 draft closed that question, extending the same verdict to any dose or sex.
Put together: the agency has been skeptical of low-dose supplementation for eight years now, and the doses and populations it once called a toss-up (higher amounts, men) aren't a toss-up anymore. And now, in 2026, the largest dataset ever assembled on the question confirmed the same thing.
The Trial Data, Sorted By What Actually Matters
Most prior reviews of calcium and vitamin D leaned on relative risk ratios: how much better one group did compared to another, proportionally. A supplement that drops fracture rates from 2% to 1.82% produces a relative risk ratio of 0.91, a number that reads as meaningful. The actual drop is 0.18 percentage points.
Massé's team threw that framing out. They set a pre-agreed threshold for absolute risk reduction, then measured every trial arm against it:
Calcium alone: 11 trials, 9,067 participants, moderate-certainty evidence, risk ratio 0.91. No clinically meaningful effect.
Vitamin D alone: 36 trials, 92,045 participants, high-certainty evidence, risk ratio 1.00. No directional signal at all.
Calcium and vitamin D combined: 15 trials, 51,126 participants, high-certainty evidence, risk ratio 0.91, absolute risk reduction of roughly 1 percentage point. This was the closest any arm came to clearing the bar, and it still fell short.
For hip fractures specifically, the researchers required a full 0.7 percentage point drop in risk before it would count as meaningful. That's a high bar, since hip fractures are rare in these trials to begin with. No supplement came anywhere close.
Now it is worth noting that this threshold is itself, a judgment call. In the United States alone, more than 300,000 people suffer a hip fracture every year, at a cost to the healthcare system in excess of $17 billion annually. And there are arguments being made that even a smaller reduction in hip fractures would still be worth having at that scale.
In my opinion, that's a fair methodological critique. But it doesn't change what the data shows under the framework the researchers actually used: measured the same way across every arm, none of the three approaches cleared it.
The result held up when the researchers split the data by age, sex, prior fracture history, prior fall history, and how much dietary calcium people were already getting. Consistency across all five factors is what gives the finding so much weight.
Why the Advice Outlasted the Evidence
The "take calcium for your bones" era has a start date. In 2001, a National Institutes of Health consensus panel concluded that older adults needed 1,000 to 1,500 milligrams of calcium a day to maintain bone mass, and that recommendation spread into clinical practice, product labels, and doctor's-office advice almost immediately. Repeated for decades, it stopped sounding like a claim and started sounding like a fact.
The USPSTF started chipping away at that assumption in 2018, when it recommended against low-dose supplementation for fracture prevention in postmenopausal women, but the correction didn't reach higher doses or men for another six years, and the update addressing that gap still hasn't been finalized.
Guidance revisions move on a committee timeline, not on how quickly a habit needs correcting, and a daily calcium pill carries no obvious downside that would make someone stop and ask why they're still taking it. No single study or agency spent two decades defending outdated advice… It has just taken us this long to catch up.
Exceptions to the Rule
Very few of the 153,902 people in the BMJ analysis lived in a nursing home. Why is that important? Simply put, institutionalized older adults are often at the highest risk of fracture, and this study didn't include many of them.
Roughly 87% of the 69 trials enrolled only community-dwelling people, and about 73% of trials were not focused on people at high risk of fracture or falls to begin with. The researchers said evidence was more limited for higher-risk groups, like institutionalized older adults and people needing residential care, especially for calcium-only and combined approaches. Less evidence is not the same as evidence of no effect. It also means the question wasn't tested with the same rigor in the population most likely to need an answer.
The people most likely to actually benefit from these supplements were the people this study tested the least.
That's what keeps this from being a flat "calcium supplements don't work" story. Confirmed deficiency, diagnosed osteoporosis under active treatment, and malabsorption are still situations where supplementation may make biological sense, and the trial evidence, while sparser, hasn't contradicted that yet.
What the Evidence Supports
The right move depends on which category you're in, not on a blanket rule.
Healthy, independently living, no diagnosed deficiency: the evidence no longer supports calcium or vitamin D supplementation as a fracture-prevention strategy. Prioritize dietary calcium (dairy, leafy greens, fortified foods) and resistance or weight-bearing exercise (balance too!), which has demonstrated fall and bone-density benefits that supplementation did not show here.
Diagnosed osteoporosis on active pharmacological treatment: calcium and vitamin D are usually still indicated alongside medications. Follow your prescriber's dosing rather than a general bone-health assumption.
None of this is a case for abandoning a supplement a doctor prescribed for a specific, diagnosed reason. It is a case for checking whether the reason was ever specific to begin with.
Where This Leaves You
The Massé analysis is the largest test this question has ever gotten: 69 trials, 153,902 people, measured against a threshold set before the results came in rather than adjusted after. Calcium alone, vitamin D alone, and the two combined all landed in the same place. None of them produced a reduction in fracture or fall risk large enough to count as clinically meaningful, and the result held up across age, sex, fracture history, fall history, and dietary calcium intake.
But remember, thirty years of consensus didn't collapse because new biology emerged overnight. It collapsed because a research team finally took the time to measure the numbers the way patients actually live them: as an absolute chance of a broken hip, not a relative ratio in a table.
So does this finding mean calcium is irrelevant?
No. It means the specific claim behind decades of standard advice, that a daily supplement meaningfully lowers fracture risk for the average older adult, does not hold up against the best evidence available.
For someone with a confirmed deficiency, a diagnosis, or an active treatment plan, this study changes nothing. But for everyone else taking a supplement on the assumption that it protects their bones, it gives a good reason to ask the question again.
--
Personal note from Jordan
I've spent close to fifteen years working in supplement research, and calcium was one of the original things I was taught mattered for bone health. During that time, it wasn't presented as one option among several. It was the given, the kind of recommendation nobody in the field questioned because it had been repeated so consistently for so long.
Reading through the Massé data, that “given” just… stopped being so. It's important to point out that providers weren't being careless 15 years ago; the evidence available then supported the recommendations. The evidence available now mostly doesn't, and pretending otherwise doesn't help anyone.
I don't think being wrong is a problem. But I do think that holding a position after the evidence has moved on is. Changing your mind can feel like admitting you were careless the first time, but really it usually just means you were working with what evidence was available at the time.
Fifteen years in, that's the discipline I'm trying to practice: hold conclusions loosely, and let newer good evidence change them when relevant… even when the conclusion is in opposition to one I used to teach.