Magnesium, vitamin D status and mortality: results from US National Health and Nutrition Examination Survey (NHANES) 2001 to 2006 and NHANES III

Population-based NHANES 2001–2006 (cross-sectional) and NHANES III (prospective cohort); adults; vitamin D status defined by serum 25(OH)D; mortality outcomes in NHANES III via National Death Index; nutrient intakes assessed from 24-h dietary recalls and 30-day supplement questionnaire; 25(OH)D measured by radioimmunoassay; PTH measured by electrochemiluminescent; analyses included logistic regression for vitamin D deficiency/insufficiency and Cox proportional hazards for mortality; interactions between magnesium and vitamin D intake tested; covariates included demographics, lifestyle factors, energy and calcium intake; analyses weighted by NHANES sampling weights.

Magnesium intake (total, dietary, and supplemental) is independently linked to lower risk of vitamin D deficiency and insufficiency; magnesium intake significantly interacts with vitamin D intake for both deficiency and insufficiency. The inverse magnesium–deficiency association is strongest among those >50 years or with extreme PTH levels; the inverse magnesium–insufficiency association is strongest among high-risk groups (winter, low vitamin D intake, women, non‑Hispanic Blacks, obesity, high PTH). In NHANES III, higher serum 25(OH)D is associated with lower risk of total mortality and mortality from cardiovascular disease (CVD) and colorectal cancer, mainly among those with magnesium intake above the median; the interaction is statistically significant for CVD mortality (P = 0.03); colorectal cancer mortality results are imprecise due to small sample size. Conclusion: Magnesium intake alone or in combination with vitamin D intake may contribute to vitamin D status, and the link between 25(OH)D and mortality may be modified by magnesium intake; further cohort studies and clinical trials are needed to confirm.

Observational design; cannot establish causality; one-time 24-h dietary recall may misclassify long-term intake; small sample size for colorectal cancer mortality; potential residual confounding and measurement errors; lack of data on other vitamin D/PTH axis parameters.