Dongze Wang, Xiaohan Wang, Wenyuan Wang, Yanyong Wang
Magnesium participates in neurotransmission, stress regulation, and inflammatory signaling, but epidemiological evidence linking circulating magnesium to depression is inconsistent. Serum magnesium remains clinically accessible and objectively measured, yet recent nationally representative evidence is limited. We analyzed 4061 adults aged ≥20 years from National Health and Nutrition Examination Survey August 2021 to August 2023. Moderate-to-severe depressive symptoms were defined as Patient Health Questionnaire-9 (PHQ-9) score ≥10. Serum magnesium was modeled as survey-weighted quartiles and per 1-standard-deviation increase. Analyses used the National Health and Nutrition Examination Survey phlebotomy weight (WTPH2YR), strata, and primary sampling units. Survey-weighted logistic regression, 3-knot restricted cubic splines, sex-stratified analyses, formal interaction testing, and sensitivity analyses were performed. Serum magnesium was not independently associated with moderate-to-severe depressive symptoms in the overall U.S. adult population. A male-specific inverse signal was observed but was not supported by a statistically significant sex interaction and should be considered exploratory. Serum magnesium did not emerge as a sensitive population-level indicator of depressive symptom status. The weighted prevalence of PHQ-9 ≥10 was 11.7%. In fully adjusted models, serum magnesium was not associated with moderate-to-severe depressive symptoms. Compared with Q1, odds ratios (ORs) for Q2 to Q4 were 0.82 (95% confidence interval [CI], 0.59-1.14), 0.90 (0.67-1.20), and 0.92 (0.59-1.44), respectively. The OR per 1-standard deviation increase was 0.92 (0.83-1.03; P = .134). Overall restricted cubic spline tests were nonsignificant (P-overall = .273; P-nonlinear = .679). An inverse association was observed in men (OR, 0.78; 95% CI, 0.64-0.96; P = .021) but not women (OR, 1.04; 95% CI, 0.91-1.19; P = .544); however, the magnesium-by-sex interaction was not significant (P-interaction = .113). Findings were unchanged using PHQ-9 ≥5, excluding estimated glomerular filtration rate <30 mL/min/1.73 m2, adjusting for dietary-magnesium, or additionally adjusting for serum calcium and 25-hydroxyvitamin D.