Factors associated with increasing vascular stiffness in PD

Sir, We read with interest the article by Jung et al. [1] investigating factors associated with aortic stiffness in peritoneal dialysis (PD) patients. The authors report significant association between heart-to-femoral pulse wave velocity (hfPWV), vascular calcification score and serum fetuin-A, but no significant association between fetuin-A concentration and change in hfPWV >1 year. This cohort of PD patients contains 35% diabetic subjects. Arterial stiffening occurs in diabetes at a faster rate and is proportional to glycated haemoglobin levels [2]. In addition, PD-related factors such as dialysis prescription, glucose absorption and advanced glycation end-products have the potential to contribute to the stiffening process. The authors suggest that these results are discrepant to our published findings [3] because they relate to our use of singleton measurements of variables and not ‘time-averaged parameters’, thereby overestimating the effect of fetuin-A in our study. Merely on theoretical grounds, it is more likely that this would underestimate the effect of fetuin-A due to random error and biological noise. Furthermore, their own data does not support their assertion, as baseline and time-averaged values have very similar means and SDs. We believe the use of time-averaged values could potentially suppress important changes in these variables, where the directionality and magnitude of change may provide more relevant information. The authors also discuss the apparent differences in fetuin-A concentrations reported in different studies and attribute this to variation in study populations and dialysis modality. While this undoubtedly contributes to the differences observed, a potentially far greater, yet relatively underappreciated problem, relates to the poor agreement between different fetuin-A assays [4]. We also note that the authors show a relationship between hfPWV and triglycerides (TG) and hypothesize that this may be mediated via asymmetric dimethylarginine (ADMA). We thought it might be of interest to show the relevant results from our stage 3 & 4 chronic kidney disease (CKD) cohort. We found no significant correlation between change in aortic pulse wave velocity (aPWV) >1 year and serum free fatty acid (Half-Micro test; Roche Diagnostics, Burgess Hill, UK) (r = 0.066, P = 0.576) or TG concentration (r = 0.128, P = 0.282). However, we note that a link between vascular function and TG may only occur when patients have significant hypertriglyceridaemia. The authors cite the relationship between change in brachial artery flow (a measure of endothelial function rather than arterial stiffness) and TG in two cohorts (mean TG 2.85 and 4.03 mmol/L) [5, 6]. In the present study [1], the mean random TG level was only 1.52 mmol/L. A further study found that patients with lipoprotein lipase deficiency, who have severe hypertriglyceridaemia, have preserved endothelial function [7]. Finally, we also measured ADMA by tandem mass spectrometry and found no significant correlation between ADMA in the non-diabetic group for either baseline aPWV (r = 0.10, P = 0.239) or change in aPWV >1 year (r = 0.059, P = 0.617). Thus, while we agree that vascular calcification is accelerated in dialysis patients, we feel that it is difficult to draw firm conclusions from this study with respect to the role of either fetuin-A, TGs or ADMA. Conflict of interest statement. SGH has previously been a member of Baxter and Amgen advisory boards, and has received honoraria for these services.