Vitamin D 1,600 or 3,200 IU for 5 years did not reduce fractures in Vitamin D–replete Finnish seniors - RCT
The effect of vitamin D supplementation on fracture incidence in an aging population — The Finnish vitamin D trial
Bone, December 2026, https://doi.org/10.1016/j.bone.2026.118081
Toni Rikkonen, Sari Hantunen, Heikki Kröger, Christel Lamberg-Allardt, JoAnn E. Manson, Tarja Nurmi, Marjo Tuppurainen, Ari Voutilainen, Tomi-Pekka Tuomainen, Jyrki K. Virtanen
Summary by Claude - September, 2026
Five years of daily vitamin D3, at either 1,600 IU or 3,200 IU, did not reduce fractures in older Finns who already had good vitamin D levels. This was an ancillary outcome of the FIND trial — a double-blind RCT of 2,495 community-dwelling adults (men 60+, women 65+, no history of cardiovascular disease or cancer) randomized 1:1:1 to placebo, 1,600 IU/day, or 3,200 IU/day. Fractures were captured from national health registers, so ascertainment was near-complete.
The starting point is the key to reading this trial. In a 550-person subgroup, mean baseline 25(OH)D was 29.9 ng/mL (74.8 nmol/L); only 9% were below 20 ng/mL, and half were already above 30 ng/mL. At 12 months the arms sat at 29.3, 40.0, and 48.3 ng/mL. This was "good vs. better," not "deficient vs. replete." Dietary calcium was also high, averaging 1,369 mg/day, and no calcium was given with the vitamin D.
Over a mean 4.1 supplemented years there were 49, 73, and 52 fractures in the placebo, 1,600 IU, and 3,200 IU arms. The 1,600 IU arm had a 52% higher fracture rate (HR 1.52, 95% CI 1.06–2.18), still elevated across 7.3 years of follow-up (HR 1.40). The 3,200 IU arm showed no increase (HR 1.07). The authors judge the 1,600 IU signal unlikely to be causal: there was no dose-response, and falls did not differ between arms.
What this does not show / limitations: nothing about people who are actually deficient — with only 9% under 20 ng/mL, that analysis could not be done. Fractures were an ancillary endpoint, so the trial was not powered for them, and many subgroup and site-specific comparisons were run without interaction significance. No co-administered calcium. Participants were all white, elderly, and living in a country with successful vitamin D food fortification.
Related in VitaminDWiki
- Both Fracture and Falls were reduced by more than 40 ng of Vitamin D
- Risk of falling not changed by 1,600 and 3,200 IU Vitamin D - RCT
- Hip fractures 60% more likely in people who drank a lot of milk - Greger
- Hip fractures are predicted by 10 factors – low Vitamin D was the biggest risk
- Fractures reduced by monthly 30,000 IU of Vitamin D, perhaps 100,000 IU would be better