10/20/2026
Most people think of vitamin D as a bone nutrient, but in a new study of health records, supplement buyers whose vitamin D was estimated to rise by more than 10 ng/mL tracked with a 28% lower risk of death, and severe deficiency tracked with a 1.6 to 2.6 times higher risk over up to 10 years.
A team from Leumit Health Services in Israel reported the work in Aging Cell. In Israel, they matched 12,352 adults whose vitamin D was severely low, under 10 ng/mL, with 12,352 adults whose level was sufficient, over 30 ng/mL, on age, s*x, smoking, weight, blood pressure, ethnic sector, socioeconomic status and the year they enrolled. Within 8 years about 14% of the deficient group had died, compared with about 8% of the sufficient group, and after adjustment the risk of death over the follow-up was about 2.6 times higher. In a US research network with 223,175 matched pairs, the risk was about 1.6 times higher.
Severe deficiency also tracked with more strokes and more kidney dialysis in Israel, with risks about 1.5 times and 1.8 times higher, and with more dementia, about 1.4 times higher.
The team then followed 247,848 members of the same health system and used pharmacy purchases of vitamin D to estimate how far each person's blood level would have risen. People whose estimated rise was more than 10 ng/mL had a 28% lower risk of death than people with almost no purchases, after accounting for their starting level, and a stricter analysis that adjusted for who tends to buy supplements put the figure at 35% lower. Smaller estimated rises tracked with smaller drops, 19% lower for 0.2 to 5 ng/mL and 22% lower for 5 to 10 ng/mL.
This was not a trial, so it cannot show that vitamin D itself changed anyone's risk. Severe deficiency often goes along with poor health, little time outdoors and obesity, and the supplement estimate came from pharmacy purchases rather than from what people actually took. In the VITAL trial, 25,871 US men aged 50 and older and women aged 55 and older took 2,000 IU a day for a median of 5.3 years and had no lower risk of cancer or major cardiovascular events, so any benefit is more likely to sit with people who are truly low. A genetic study of 307,601 people in the UK Biobank, by Sutherland and colleagues, pointed the same way, finding that the genetically predicted risk of death fell steeply as vitamin D rose up to about 50 nmol/L, which is 20 ng/mL.
Vitamin D has come up on the podcast before, and for practical purposes the Endocrine Society guideline suggests taking it if you are 75 or older, while anyone who may be low because of little sun, a digestive condition or other risk factors can ask a doctor for a 25-hydroxyvitamin D blood test before choosing a dose.
Israel et al., Aging Cell 2026;25(10):e70752
Manson et al., N Engl J Med 2019;380(1):33-44
Sutherland et al., Ann Intern Med 2022;175(11):1552-1559
Demay et al., J Clin Endocrinol Metab 2024;109(8):1907-1947
