Abstract

RATIONALE AND OBJECTIVE: Serum magnesium levels have been inversely yet inconsistently associated with cardiovascular (CV) outcomes. In this study, we examined the association of serum magnesium levels with CV outcomes in the Systolic Blood Pressure Intervention Trial (SPRINT) participants. STUDY DESIGN: Case-control post hoc analysis of SPRINT. SETTING & PARTICIPANTS: A total of 2,040 SPRINT participants with available serum samples at baseline level were included in this study. Case participants (n = 510) who experienced a CV event during the SPRINT observation period (median follow-up of 3.2 years) and control participants (n = 1,530) without CV events were sampled in a 1:3 ratio for measurements of serum magnesium level at baseline and 2-year follow-up. PREDICTORS: Baseline serum magnesium levels and 2-year percentage change in serum magnesium levels (ΔSMg). OUTCOME: SPRINT primary composite CV outcome. ANALYTICAL APPROACH: Multivariable conditional logistic regression analysis, accounting for matching factors, was used to evaluate the association of baseline and ΔSMg with CV outcomes. Individual matching of cases and controls was based on the SPRINT treatment arm allocation (standard vs intensive) and prevalence of chronic kidney disease (CKD). RESULTS: The median serum magnesium level at baseline was similar among the case and control groups. In a fully adjusted model, each standard deviation (SD) (0.18 mg/dL) higher of the baseline serum magnesium level was independently associated with a lower risk for composite CV outcomes in all study participants (adjusted odds ratio 95% CI, 0.79 [0.70-0.89]). This association was similar when serum magnesium levels were analyzed in quartiles but dissipated in the standard (vs intensive) arm of SPRINT (0.88 [0.76-1.02] vs 0.65 [0.53-0.79], respectively; <i>P</i>_interaction = 0.06). The presence or absence of CKD at baseline did not modify this association. ΔSMg was not independently associated with CV outcomes occurring after 2 years. LIMITATIONS: ΔSMg was small in magnitude, limiting effect size. CONCLUSIONS: Higher baseline serum magnesium levels were independently associated with reduced risk for CV outcomes in all study participants, but ΔSMg was not associated with CV outcomes.