Dairy prevented fractures in one large trial and nothing in the cohort data. Both hold.

Whether dairy protects bones depends almost entirely on how little calcium and protein a person was getting to begin with. In frail Australian aged-care residents consuming about 700 mg of calcium and 58 g of protein a day, adding milk, yoghurt and cheese cut all fractures by 33% and hip fractures by 46% over two years [s1]. In the general population over 50, a systematic review of 44 cohort studies found that dietary calcium intake is not associated with fracture risk, and that there is no clinical trial evidence that increasing calcium from dietary sources prevents fractures [s2]. Both findings come from The BMJ, and they are not in conflict.

The trial that worked

Sixty accredited residential aged-care facilities in Australia, housing 7,195 permanent residents with a mean age of 86.0 years, 68% of them women, were stratified by location and organisation and randomised by facility [s1]. All residents were replete in vitamin D at baseline [s1]. Thirty facilities served additional milk, yoghurt and cheese providing 562 mg more calcium and 12 g more protein per day, bringing total intake to 1,142 mg of calcium and 69 g of protein — 1.1 g per kilogram of body weight [s1]. The 30 control facilities kept their usual menus, with residents consuming 700 mg of calcium and 58 g of protein a day, or 0.9 g/kg [s1].

Across 27 intervention and 29 control facilities analysed, there were 324 fractures including 135 hip fractures, 4,302 falls and 1,974 deaths [s1]. The intervention reduced all fractures by 33% (121 versus 203 events; hazard ratio 0.67, 95% CI 0.48 to 0.93, P = 0.02), hip fractures by 46% (42 versus 93; 0.54, 0.35 to 0.83, P = 0.005) and falls by 11% (1,879 versus 2,423; 0.89, 0.78 to 0.98, P = 0.04) [s1]. The hip-fracture separation reached significance at five months and the falls separation at three [s1]. Mortality was unchanged (900 versus 1,074; 1.01, 0.43 to 3.08) [s1].

The review that found nothing

The 2015 systematic review searched for the evidence underpinning advice to increase calcium intake in adults over 50 [s2]. It found only two eligible randomised trials of dietary calcium sources, covering 262 people between them, alongside 50 reports from 44 cohort studies [s2].

The cohort results were consistently null. Most studies reported no association between calcium intake and fracture: 14 of 22 for total fracture, 17 of 21 for hip, 7 of 8 for vertebral and 5 of 7 for forearm [s2]. Most milk studies (25 of 28) and most dairy studies (11 of 13) likewise reported no association [s2].

Calcium supplements fared only marginally better. Across 26 randomised trials, supplements reduced total fracture (20 studies, 58,573 participants; relative risk 0.89, 0.81 to 0.96) and vertebral fracture (12 studies, 48,967; 0.86, 0.74 to 1.00), but not hip fracture (13 studies, 56,648; 0.95, 0.76 to 1.18) or forearm fracture (8 studies, 51,775; 0.96, 0.85 to 1.09) [s2]. Funnel-plot inspection and Egger's regression suggested publication bias toward supplements [s2]. Restricted to the four trials at lowest risk of bias, covering 44,505 participants, there was no effect on fracture at any site [s2].

One exception stands out in the review's own summary, and it prefigures the 2021 trial: only one trial, in frail elderly women in residential care with low dietary calcium intake and low vitamin D, showed significant fracture reductions [s2].

The bone-density evidence explains why

A companion review pooled 59 randomised trials of calcium and bone mineral density — 15 of dietary sources (1,533 participants) and 51 of supplements (12,257) [s3]. Increasing dietary calcium raised bone mineral density by 0.6–1.0% at the total hip and total body at one year, and by 0.7–1.8% at those sites plus lumbar spine and femoral neck at two years, with no effect at the forearm [s3]. Supplements produced 0.7–1.8% increases at all five sites, but the increase at later time points was similar to the increase at one year — a one-off step, not a progressive gain [s3]. Increases were similar whether the dose was at least 1,000 mg or 500 mg a day, and whether or not vitamin D was co-administered [s3]. The authors concluded these small non-progressive increases are unlikely to produce a clinically significant reduction in fracture risk [s3].

A 1% density gain does not prevent hip fractures. What the 2021 trial changed was not bone density but falls — and in a population falling frequently from a low protein baseline, an 11% reduction in falls plus improved muscle-relevant protein intake is a plausible route to a 46% hip-fracture reduction that has nothing to do with calcium's effect on the skeleton [s1].

What this does not establish

The trial was conducted in residents with a mean age of 86 living in institutional care, replete in vitamin D, and starting from calcium and protein intakes below recommended levels [s1]. It says nothing about a well-nourished adult in their fifties adding a yoghurt. The cohort evidence addressing that person is null [s2], and the trial evidence that would settle it does not exist — two randomised trials of dietary calcium totalling 262 participants is not a literature [s2].

This article is informational and is not medical advice. Osteoporosis treatment and calcium or vitamin D supplementation are clinical decisions.