Benefits
Bone health: essential when deficient, no fracture benefit when replete
Severe vitamin D deficiency causes rickets in children and osteomalacia in adults — at this level, supplementation matters absolutely. Older trials that found fewer hip fractures with calcium plus vitamin D were done in very old people who often had low vitamin D levels and low calcium intake, such as a French trial in 3,270 women with an average age of 84. In people over 50 living in the community, a meta-analysis of 33 trials (51,145 adults) found that calcium, vitamin D or both did not reduce hip or total fractures, and the VITAL fracture study (25,871 adults, 2,000 IU/day for a median 5.3 years) found no effect on total, nonvertebral or hip fractures, with no difference by baseline vitamin D level. If your level is adequate, extra vitamin D does not protect your bones; in one 3-year trial, 4,000 and 10,000 IU/day produced lower forearm bone density than 400 IU/day.
Cardiovascular events — supplementation does not prevent CVD
Despite strong observational links between low vitamin D and heart disease, the definitive randomized trials are negative. The largest trial (~26,000 adults, 2,000 IU/day for over 5 years) showed no reduction in heart attacks, strokes, or cardiovascular mortality. A second large trial replicated the null result. Vitamin D supplementation does not prevent cardiovascular disease in unselected adults — the observational link most likely reflects confounding, since people with low vitamin D tend to be less healthy in other ways.
Cancer: no prevention, and mixed signals on cancer death
Vitamin D supplementation does not prevent cancer: VITAL found no reduction in invasive cancer (HR 0.96). For cancer death, VITAL's estimate was 17% lower but not statistically significant (HR 0.83, 95% CI 0.67 to 1.02), and a 2022 meta-analysis found about 13% fewer cancer deaths only in trials of daily dosing (SRR 0.87), not in trials of large intermittent doses. The D-Health trial (21,315 older Australians, 60,000 IU monthly) pointed the other way: cancer deaths were not significantly different overall (HR 1.15, 95% CI 0.96 to 1.39), but exploratory analyses excluding the first 2 years suggested more cancer deaths (HR 1.24), and the investigators advised that this regimen might not be appropriate for people who are already vitamin D replete. Do not take vitamin D to prevent or treat cancer.
Autoimmune diagnoses in VITAL: an early signal that faded
In an ancillary analysis of VITAL, people taking 2,000 IU/day of vitamin D3 (with or without omega-3) for a median 5.3 years had 22% fewer confirmed autoimmune diagnoses, including rheumatoid arthritis, polymyalgia rheumatica, autoimmune thyroid disease and psoriasis (123 vs 155 cases; HR 0.78, 95% CI 0.61 to 0.99, P=0.05). The finding has not held up: when later-confirmed cases that began during the trial were added, the estimate was no longer statistically significant (HR 0.85, 95% CI 0.70 to 1.04), and over 7 years of follow-up, including 2 years after the capsules stopped, there was no difference (HR 0.98). It is a single, unreplicated result about disease incidence and is not a reason for healthy adults to supplement.
Falls in older adults: not reduced in large trials
A 2009 meta-analysis found 19% fewer falls in 7 trials (1,921 older adults) that used 700 to 1,000 IU/day. Larger, later trials did not confirm this: VITAL (2,000 IU/day, 25,871 adults) found no reduction in falls, with no difference by baseline vitamin D level, and D-Health (60,000 IU monthly, 21,315 adults) found no reduction overall, with more falls in the subgroup of people of normal body weight. A 2018 meta-analysis of 37 trials (34,144 people) found no effect on falls (RR 0.97). A single yearly dose of 500,000 IU increased falls by 15% in older women. In 2018 the US Preventive Services Task Force recommended against vitamin D to prevent falls in community-dwelling adults 65 and older not known to have osteoporosis or vitamin D deficiency.
Respiratory infections: no clear protection in updated evidence
A 2017 pooled analysis of 25 trials found 12% lower odds of acute respiratory infection (OR 0.88), with the largest effect in people with very low starting levels. The 2025 update, covering 40 trials and 61,589 participants, found a similar point estimate but no statistically significant protection (OR 0.94, 95% CI 0.88 to 1.00), and the effect did not differ by baseline vitamin D status, dose or dosing frequency. The CORONAVIT trial (6,200 UK adults, test-and-treat with 800 or 3,200 IU/day) did not reduce respiratory infections or COVID-19. Do not rely on vitamin D to prevent colds or flu.
COVID-19: vitamin D3 trials were null
Low vitamin D was associated with worse COVID-19 outcomes in observational data, but randomized trials of vitamin D3 were null: a single 200,000 IU dose did not shorten hospital stay in 240 patients with moderate to severe COVID-19, and the CORONAVIT test-and-treat trial in 6,200 UK adults did not reduce COVID-19 infections. Small studies of calcifediol (25-hydroxyvitamin D), a different chemical form sold as a prescription medicine in some countries, do not apply to ordinary vitamin D3 supplements. Vitamin D does not treat or prevent COVID-19 and does not replace vaccination or medical care.
Depression and mood: no benefit in a large prevention trial
Vitamin D deficiency consistently tracks with depression in observational studies. In VITAL-DEP (18,353 adults aged 50 and older, 2,000 IU/day for 5.3 years), vitamin D did not reduce new or recurrent depression or change mood scores. Whether correcting a true deficiency improves mood has not been settled by large trials. Vitamin D is not an antidepressant, and persistent low mood needs medical care.
Mechanism of action
Metabolic Activation
Vitamin D (either D2 or D3) is first hydroxylated in the liver to form 25-hydroxyvitamin D [25(OH)D], the main circulating form. A second hydroxylation occurs primarily in the kidney, producing the active form, 1,25-dihydroxyvitamin D [1,25(OH)₂D, also called calcitriol]. These steps are catalyzed by cytochrome P450 enzymes (CYPs), such as CYP2R1 in the liver and CYP27B1 in the kidney
Genomic Actions
Calcitriol binds to the vitamin D receptor (VDR), a nuclear transcription factor present in many cell types. The VDR-calcitriol complex forms a heterodimer with the retinoid X receptor (RXR). This complex binds to vitamin D response elements (VDREs) in the DNA, regulating the transcription of hundreds of genes. These genes are involved in calcium and phosphate homeostasis, cell proliferation, differentiation, and immune function
Non-Genomic Actions
Some effects of vitamin D are too rapid to be explained by gene transcription, such as rapid calcium uptake in cells. These may be mediated by membrane-associated receptors and signaling pathways, including PDIA3
Clinical trials
Largest vitamin D primary prevention clinical trial to date — randomized double-blind placebo-controlled 2x2 factorial trial of vitamin D3 alongside omega-3 fatty acids. Published in NEJM. Participants were not selected for vitamin D deficiency; mean baseline 25(OH)D was about 77 nmol/L (31 ng/mL).
25,871 US adults (men ≥50, women ≥55). 5.3-year intervention with vitamin D3 2,000 IU/day or placebo.
Primary endpoints (invasive cancer, major cardiovascular events) were negative — no significant difference vs placebo. In adults not selected for vitamin D deficiency, vitamin D did not reduce invasive cancer (HR 0.96) or major cardiovascular events (HR 0.97); death from cancer, a secondary endpoint, was not significantly lower (HR 0.83, 95% CI 0.67 to 1.02). Established the strongest single counterpoint to widespread observational data linking low vitamin D to CV and cancer risk.
Ancillary study of the VITAL trial (median 5.3 years) with confirmed incident autoimmune disease as its endpoint; not one of VITAL's main cancer and cardiovascular endpoints. Published in BMJ (376:e066452). Autoimmune disease cases verified through medical record review and rheumatology consensus.
25,871 VITAL participants followed for new-onset autoimmune disease over 5 years.
Vitamin D 2,000 IU/day (with or without omega-3) was followed by fewer confirmed autoimmune diagnoses: 123 vs 155 (HR 0.78, 95% CI 0.61 to 0.99, P=0.05). Vitamin D alone and vitamin D plus omega-3 gave similar estimates versus double placebo (HR 0.68 and 0.69). Diseases included rheumatoid arthritis, polymyalgia rheumatica, autoimmune thyroid disease and psoriasis. In a 2024 follow-up that added later-confirmed cases, the in-trial effect was no longer significant (HR 0.85, 95% CI 0.70 to 1.04), and over 7 years of follow-up, including 2 years after supplements stopped, there was no difference (HR 0.98).
5-year randomized placebo-controlled trial in elderly Finnish adults comparing two vitamin D3 dose levels vs placebo. Published in American Journal of Clinical Nutrition. Northern latitude population with naturally lower sun exposure.
2,495 elderly Finnish adults (men ≥60, post-menopausal women ≥65). Three arms: placebo, 1,600 IU/day, or 3,200 IU/day for 5 years.
NO significant differences across the three arms for cardiovascular events, invasive cancer, or all-cause mortality. Reinforces the VITAL findings at higher doses. Mean baseline 25(OH)D in the measured subcohort was 75 nmol/L, so most participants were already replete, which the authors suggested may explain the null result.
Individual participant data pooled analysis pooling raw data from randomized trials of vitamin D supplementation for prevention of acute respiratory infections. Published in BMJ. The 2025 update of this analysis (40 trials, 61,589 participants) no longer found statistically significant protection.
10,933 participants across 25 clinical trials. Various supplementation regimens (daily, weekly, bolus).
Across all participants, vitamin D lowered the odds of acute respiratory infection by 12% (adjusted OR 0.88, 95% CI 0.81 to 0.96). Daily or weekly dosing gave OR 0.81, bolus dosing gave no protection (OR 0.97), and people starting below 25 nmol/L on daily or weekly dosing had the largest effect (OR 0.30). The 2025 update found no statistically significant overall protection (OR 0.94, 95% CI 0.88 to 1.00) and no difference by baseline vitamin D status.
Evidence review and pooled analysis of fracture prevention trials in community-dwelling older adults. Published in JAMA. Distinguished community-dwelling from institutionalized populations, which has been an important confounding variable in earlier analyses.
51,145 community-dwelling older adults across 33 clinical trials.
Calcium, vitamin D, or the combination was not associated with reduced fracture incidence in community-dwelling older adults — contradicting long-held assumptions about routine supplementation for fracture prevention. Vitamin D alone showed no reduction in hip fracture (RR 1.21, 95% CI 0.99 to 1.47), and results were generally consistent regardless of dose, sex, fracture history, dietary calcium and baseline 25(OH)D. The review included only community-dwelling adults, so it does not address nursing-home residents.