Background: In patients with chronic kidney disease (CKD), although strong associations have been observed between malnutrition and atherosclerosis, the relationship between serum albumin concentration and angiographic changes of coronary artery disease (CAD) remains poorly explored. The goal of the present study was, in patients with CKD, to clarify the relationship between the angiographic severity of CAD and serum albumin concentration reflecting either inflammation or nutrition or both.
Methods: In this study, 100 end-stage renal disease (ESRD) patients were enrolled, who commenced long-term dialysis therapy at our hospital and underwent coronary angiography within 3 months of the first haemodialysis (HD) session. Mean age was 63+/-11 years, 20% of the subjects were female and 62% had diabetes. Severity of CAD was evaluated in terms of (i) number of vessels exhibiting CAD (>or=75% stenosis) and (ii) Gensini score (GS). Clinical characteristics and laboratory findings were recorded at initiation of long-term HD therapy. We then evaluated a possible association with the presence and degree of CAD.
Results: Sixty-four patients exhibited signs of CAD. Forty-one among them (64%) had multivessel disease. On univariate logistic regression analysis, age, diabetes and hypoalbuminaemia were significantly associated with multivessel CAD. Univariate linear regression analysis demonstrated a positive correlation of age and diabetes with GS, and an inverse correlation of BMI and serum albumin level with GS. Stepwise regression analysis showed age and serum albumin level to be independently associated with multivessel CAD and GS. The ROC curves demonstrated best cut-off levels of age and albumin for predicting multivessel CAD to be 70 years and 3.15 g/dl, respectively.
Conclusion: Hypoalbuminaemia at the initiation of dialysis is an important predictor of advanced CAD, particularly in male and in diabetic patients. It may reflect mainly a state of inflammation. However, malnutrition as a confounding factor cannot be entirely excluded.