Progression of coronary artery calcification in chronic kidney disease: nonlinear associations with serum magnesium and parathyroid hormone

Introduction Coronary artery calcification (CAC) progresses rapidly in chronic kidney disease (CKD), but its trajectories across stages and nonlinear relationships with serum magnesium and parathyroid hormone (PTH) remain unclear. Methods We evaluated 800 CKD patients stratified into four groups (stages 1–2, 3, 4–5, and dialysis) over a median follow-up of 20.5 months using serial chest CT scans. Linear mixed-effects models, restricted cubic splines, and LASSO regression were applied to analyze trajectories, nonlinear relationships, and construct a predictive model. Results The annualized rate of CAC progression increased significantly with advancing CKD stages (8.5%, 13.2%, 22.8%, and 29.5%/year, respectively; P < 0.001). Serum magnesium exhibited an L-shaped nonlinear relationship with rapid CAC progression (inflection point: 0.82 mmol/L), while PTH showed a J-shaped relationship (inflection point: 120 pg/mL). Uric acid was linearly and positively correlated. A 7-factor LASSO-derived prognostic model demonstrated robust performance, yielding a C-statistic of 0.802 (95% CI: 0.772-0.832) and good calibration. Conclusion In conclusion, CAC progression accelerates as renal function declines, peaking in dialysis patients. Low magnesium (<0.82 mmol/L) and high PTH (120 pg/mL) are key nonlinear risk factors. The resulting predictive model provides clinical utility for stratified cardiovascular risk management.