The D-Health Trial: a randomised controlled trial of the effect of vitamin D on mortality

Background: The effect of supplementing unscreened adults with vitamin D3 on mortality is unclear. We aimed to determine whether monthly doses of vitamin D3 influenced mortality in older Australians.

Methods: We did a randomised, double-blind, placebo-controlled trial of oral vitamin D3 supplementation (60 000 IU per month) in Australians 60 years or older who were recruited across the country via the Commonwealth electoral roll. Participants were randomly assigned (1:1), using automated computer-generated permuted block randomisation, to receive one oral gel capsule of either 60 000 IU vitamin D3 or placebo once a month for 5 years. Participants, staff, and investigators were blinded to study group allocation. The primary endpoint was all-cause mortality assessed in all participants who were randomly assigned. We also analysed mortality from cancer, cardiovascular disease, and other causes. Hazard ratios (HRs) and 95% CIs were generated using flexible parametric survival models. This trial is registered with the Australian New Zealand Clinical Trials Registry, ACTRN12613000743763.

Findings: Between Feb 14, 2014, and June 17, 2015, we randomly assigned 21 315 participants, including 10 662 to the vitamin D group and 10 653 to the placebo group. In 4441 blood samples collected from randomly sampled participants (N=3943) during follow-up, mean serum 25-hydroxy-vitamin D concentrations were 77 (SD 25) in the placebo group and 115 (SD 30) nmol/L in the vitamin D group. Following 5 years of intervention (median follow-up 5·7 years [IQR 5·4-6·7]), 1100 deaths were recorded (placebo 538 [5·1%]; vitamin D 562 [5·3%]). 10 661 participants in the vitamin D group and 10 649 participants in the placebo group were included in the primary analysis. Five participants (one in the vitamin D group and four in the placebo group) were not included as they requested to be withdrawn and their data to be destroyed. The HR of vitamin D3 effect on all-cause mortality was 1.04 [95% CI 0·93 to 1·18]; p=0·47)and the HR of vitamin D3 effect on cardiovascular disease mortality was 0·96 (95% CI 0·72 to 1·28; p=0·77). The HR for cancer mortality was 1·15 (95% CI 0·96 to 1·39; p=0·13) and for mortality from other causes it was 0·83 (95% CI 0·65 to 1·07; p=0·15). The odds ratio for the per-protocol analysis was OR 1·18 (95% CI 1·00 to 1·40; p=0·06). In exploratory analyses excluding the first 2 years of follow-up, those randomly assigned to receive vitamin D had a numerically higher hazard of cancer mortality than those in the placebo group (HR 1·24 [95% CI 1·01-1·54]; p=0·05).

Interpretation: Administering vitamin D3 monthly to unscreened older people did not reduce all-cause mortality. Point estimates and exploratory analyses excluding the early follow-up period were consistent with an increased risk of death from cancer. Pending further evidence, the precautionary principle would suggest that this dosing regimen might not be appropriate in people who are vitamin D-replete.