Abstract
BACKGROUND: Cancer survivors face significant mortality risks, including from cardiovascular disease (CVD) and cancer. Magnesium depletion score (MDS) is a prognostic biomarker, but its prognostic value in cancer survivors is unknown. METHODS: This prospective study utilizes data from two distinct cohorts: the National Health and Nutrition Examination Survey (NHANES) and the Shaanxi Provincial People's Hospital. We analyzed 3,528 cancer survivors from NHANES (1999-2018) and 473 patients from the hospital cohort. Associations between MDS (0, 1, 2, ≥3) and all-cause, cancer-specific, and CVD mortality were assessed using weighted multivariate Cox regression, Kaplan-Meier analysis and Fine-Gray competing risk model. Predictive accuracy was evaluated using time-dependent ROC curves. RESULTS: During follow-up (median 165 months for NHANES), higher MDS (≥3) was significantly associated with increased mortality risk after full adjustment: all-cause (HR = 2.48, 95% CI: 1.86-3.29), cancer (HR = 2.17, 95% CI: 1.44-3.26), and CVD (HR = 3.89, 95% CI: 1.91-7.91) compared to MDS = 0 (all p for trend<0.001). Kaplan-Meier curves and competing risk model confirmed worse survival with higher MDS. Time-dependent ROC demonstrated good predictive accuracy for all mortality risk. Results remained robust in sensitivity analyses and the single-center cohort. CONCLUSION: Higher MDS is a strong, independent predictor of increased all-cause, cancer-specific, and CVD mortality risk among cancer survivors. MDS provides valuable prognostic information and could aid in risk stratification for personalized survivorship care.