Skip to content
What Causes What
MENDELIAN~ NON-LINEAR

25-hydroxyvitamin D → All-cause mortality

Higher 25-hydroxyvitamin D moves All-cause mortality in a non-linear way. Dose or timing can flip the sign.

Sutherland 2022 found an L-shaped Mendelian randomization link between genetically predicted 25(OH)D and all-cause death in 307,601 white-European UK Biobank adults, with risk falling until about 50 nmol/L, while Trummer 2013 and Ong 2019 found no linear genetic effect.

Mendelian randomization · Contested · N=307601

Mendelian randomization. It's a great tool for ruling out environmental factors, though it has its own genetic limitations.

Mechanism

People with low 25(OH)D die sooner in observational catalogs. Time outdoors, body fat, illness, and money sit on that backdoor. Illness can also pull D down. Researchers use a genetic shortcut called Mendelian randomization. It is good at ruling out diet, income, and who still goes outside. It can still be fooled when one gene affects two traits. A linear instrument can miss a threshold. Trummer, Pilz, März, and colleagues (2013) typed three 25(OH)D SNPs (GC, DHCR7, CYP2R1) in 3,316 people sent for coronary angiography. The genotypes predicted measured D (r-squared 0.175). Over a median 9.9 years, 955 people died, 619 from heart disease. The genotypes did not predict all-cause, cardiovascular, or other death. Ong, Gharahkhani, Neale, and MacGregor (2019) used five 25(OH)D markers in 438,870 white-British UK Biobank adults. They counted 6,998 cancer deaths. A 20 nmol/L genetically predicted rise did not move cancer death or overall cancer risk. Sutherland, Zhou, and Hyppönen (2022) stayed in UK Biobank but allowed a curve. They used 35 variants in 307,601 unrelated white-European adults and counted 18,700 deaths through June 2020. Genetically predicted 25(OH)D against all-cause death was L-shaped (P for nonlinearity under 0.001). Risk fell steeply until 50 nmol/L. At a measured 25 nmol/L versus 50, the genetic analysis put the odds of death at 1.25 (1.16 to 1.35). Cancer, heart, and lung death showed the same shape (P of 0.033 or lower). The main paper is a deficiency threshold, not a linear lifelong bonus. The earlier linear papers stay on the page because they said the opposite.

Caveats

This page is circulating 25(OH)D and later death. It is not a job, not a cognitive score, and not a heart attack someone survived. Sutherland and Ong are white-European or white-British UK Biobank. A volunteer cohort is healthier than the country. Trummer's sample was already in the cath lab. Mendelian randomization assumes no pleiotropy. Nonlinear Mendelian randomization still uses measured D to place people on the curve. The authors say the size of the association is approximate. A linear instrument can miss a threshold if most of the sample sits above it. Each paper is Mendelian randomization, a genetic shortcut. Trummer 2016 (biochemical recurrence, metastases, and overall survival in 702 prostate cancer patients) is a prognosis file, not this page. This is not medical advice and not a recommendation to take vitamin D, to skip it, or to treat a number.

Effect

odds_ratio = 1.25. Sutherland 2022 nonlinear MR: all-cause OR 1.25 (1.16-1.35) at 25 vs 50 nmol/L. L-shaped, P nonlinearity <0.001. Trummer 2013 and Ong 2019 linear MR were null.

Nodes

Biology & endocrinology25-hydroxyvitamin DBiology & endocrinologyAll-cause mortality

Named confounders

Cite this page

What Causes What. “25-hydroxyvitamin D → All-cause mortality.” https://whatcauseswhat.org/edges/e-vitd-mortality (atlas updated 2026-09-02).

https://whatcauseswhat.org/edges/e-vitd-mortality

Papers