Association of D-dimer levels with vascular calcification in patients undergoing maintenance hemodialysis: a cross-sectional study

Background Vascular calcification (VC) is common in patients undergoing maintenance hemodialysis (MHD) and is associated with adverse cardiovascular outcomes. D-dimer, a marker of coagulation and fibrinolysis, may reflect pathophysiological processes linked to VC, but its relationship with VC in MHD patients remains unclear. Methods In this single-center cross-sectional study, 346 MHD patients were enrolled. VC was assessed by chest x—ray-based aortic arch calcification grading and defined as grade 0 vs. grades 1–3 for regression analyses. Multivariable logistic regression was used to evaluate the association between D-dimer and VC. A clinically applicable risk assessment model was developed. Internal validation used 1,000 bootstrap resamples in which the entire model-development process, including variable screening, was repeated, and the incremental value of D-dimer was evaluated by comparing models with and without D-dimer using optimism-corrected discrimination, calibration, overall performance, and decision curve analysis. Model performance was assessed by receiver operating characteristic analysis, calibration, and decision curve analysis. Results D-dimer levels increased across VC grades ( P < 0.001). After multivariable adjustment, D-dimer remained independently associated with VC; each 100 μg/L increase in D-dimer was associated with a 9.5% higher odds of VC (OR 1.095, 95% CI 1.014–1.182; P = 0.020). Age, hemoglobin, and calcium-based phosphate binder use were also independently associated with VC. The full model showed acceptable discrimination (apparent AUC 0.802, 95% CI 0.754–0.849; optimism-corrected AUC 0.770). Adding D-dimer to the multivariable model yielded only a modest, non-significant increase in AUC over the model without D-dimer (apparent AUC 0.802 vs. 0.787; optimism-corrected 0.770 vs. 0.756). Conclusions In MHD patients, higher D-dimer levels were independently associated with VC. A model based on routinely available variables showed acceptable performance for individualized VC risk assessment, although the incremental discriminative value of D-dimer was modest.