Associations between vascular calcification, arterial stiffness and bone mineral density in chronic kidney disease

Background. Vascular calcification (VC) and arterial stiffness are major contributors to cardiovascular (CV) disease in chronic kidney disease (CKD). Both are independent predictors of CV mortality and are inversely correlated with bone mineral density (BMD). Few studies have addressed the extent of VC in the pre-dialysis CKD population, with associated measurements of BMD and arterial compliance. Methods. We report cross-sectional data on 48 patients with CKD (GFR 1755 mlmin) assessing the prevalence of VC and its associations. All patients had computed tomography (CT) scans through abdominal aorta and superficial femoral arteries (SFAs) to determine VC, pulse wave velocity (PWV) using SphygmoCor device (AtCor PWV Inc., Westmead, Australia) measuring arterial stiffness, and dual-energy X-ray absorptiometry (DEXA) scans to determine BMD, as well as serum markers of renal function and mineral metabolism. Results. Patients, 71 male, 54 diabetic, had a median age 64.5 years. Mean estimated GFR was 35.1 +/- 10 mlmin. Mean PWV was 10.0 +/- 4.5 ms and mean aortic VC score was 421.5 +/- 244 Hounsfield units, with 90 of subjects having some aortic VC present. In univariate linear regression analysis, aortic VC correlated positively with age (r 0.50, P 0.001), triglycerides (r 0.47, P < 0.002) and PWV (r 0.33, P = 0.03). There was also greater VC with declining renal function (r -0.28, P = 0.05). There was no significant association between VC and serum markers of mineral metabolism, however phosphate and Ca P correlated positively with PWV (r 0.35, P = 0.02, r 0.36, P = 0.02, respectively). There was also a positive association between PWV and triglycerides (P = 0.008), and a trend towards greater PWV with increasing age (P = 0.09). In multivariate regression analysis only increasing age and triglyceride levels were significantly associated with aortic VC and PWV. Mean spine and femoral T-scores on DEXA were 0.48 and -1.31 respectively, with 13 of subjects having femoral T-score < -2.5 (osteoporotic range). SFA VC inversely correlated with femoral T-scores (r -0.43, P = 0.004); however, there was a positive (likely false) association between spine T-scores and aortic VC (r 0.37, P = 0.01), related to the limitation of vertebral DEXA in CKD. Conclusion. There is a high prevalence of VC in pre-dialysis CKD patients, worse with increasing age, triglycerides and reducing renal function. Correlation exists between VC and PWV and determination of one or both may be useful for CKD patient CV risk assessment. Femoral BMD is inversely associated with SFA VC, but measurement of vertebral BMD by DEXA is unreliable in CKD patients with aortic VC.