Do not give vitamin D to people who are not deficient simply to extend their lives
Supplementing for five years did not clearly reduce cancer, cardiovascular disease, or deaths in trials that did not specifically recruit vitamin-D-deficient people. Confirmed deficiency, osteoporosis, and prolonged lack of sunlight are different situations; follow medical advice.
What it takes
About RMB 0.2–0.5 a day, or 100–200 a year. The larger cost is the false reassurance that you have taken an effective preventive measure.
What you may gain
VITAL enrolled 25,871 people, used 2,000 IU daily, and followed participants for a median of 5.3 years. It found no clear differences in invasive cancer, HR 0.96 (95% CI 0.88–1.06); major cardiovascular events, HR 0.97 (0.85–1.12); or all-cause mortality, HR 0.99 (0.87–1.12). An HR of 1 means no difference, and the confidence intervals describe uncertainty around the estimates. D-Health enrolled 21,315 Australians aged 60 or older, gave 60,000 IU monthly, and followed them for five years. All-cause mortality was also unchanged: HR 1.04 (0.93–1.18), with deaths in 5.3% versus 5.1%. The intuitive appeal comes from observational studies linking low vitamin D to higher mortality. That association does not demonstrate that supplementation fixes the risk.
Context & considerations
Neither trial specifically recruited people with genuine vitamin D deficiency, so this conclusion should not be extended to treating deficiency. People with confirmed deficiency or osteoporosis, people who get very little sunlight, and infants should follow clinical supplementation advice. The outcomes considered here are overall mortality, cancer, and cardiovascular disease, not every possible use of vitamin D.
Research & references
Manson JE et al. (2019). Vitamin D Supplements and Prevention of Cancer and Cardiovascular Disease. NEJM. https://doi.org/10.1056/NEJMoa1809944; Neale RE et al. (2022). The D-Health Trial: a randomised controlled trial of the effect of vitamin D on mortality. Lancet Diabetes Endocrinol. https://doi.org/10.1016/S2213-8587(21)00345-4.