Superiority of glycated albumin over glycated haemoglobin as indicator of glycaemic control and predictor of all-cause mortality in patients with type 2 diabetes mellitus receiving peritoneal dialysis

Background Glycated albumin, in contrast to glycated haemoglobin, precisely reflects glycaemic control and predicts all-cause mortality in haemodialysis patients with diabetes mellitus. However, whether those associations exist in diabetes mellitus patients receiving peritoneal dialysis remains unclear.Methods This was a retrospective cross-sectional and longitudinal observational study. We measured glycated albumin, glycated haemoglobin and casual plasma glucose for two months in diabetes mellitus-peritoneal dialysis ( n  = 44) and diabetes mellitus-haemodialysis ( n  = 88) patients (age-, gender-matched). The diabetes mellitus-peritoneal dialysis patients were followed for three years to monitor occurrence of all-cause mortality.Results Glycated albumin and glycated albumin/casual plasma glucose ratios, but not casual plasma glucose, glycated haemoglobin, or glycated haemoglobin/casual plasma glucose, were significantly lower in the diabetes mellitus-peritoneal dialysis as compared with the diabetes mellitus-haemodialysis patients. The regression lines between casual plasma glucose and glycated albumin showed a significant parallel shift downwards in diabetes mellitus-peritoneal dialysis as compared with diabetes mellitus-haemodialysis patients, while the slope did not differ significantly between the groups, resulting in underestimation of glycaemic control by 4.5%. Kapan-Meier analysis of the diabetes mellitus-peritoneal dialysis patients revealed that higher glycated albumin (median >18.0%), but not glycated haemoglobin (median >6.6%), indicated significantly elevated risk for all-cause mortality, which occurred in 15 patients (34.1%), as compared with those with a lower glycated albumin concentration. Higher glycated albumin concentration was also significantly and independently associated with all-cause mortality in multivariate Cox proportional hazards analysis.Conclusions Glycated albumin, in contrast to glycated haemoglobin, more precisely reflects glycaemic control in diabetes mellitus-peritoneal dialysis patients, based on its significant association with all-cause mortality. Furthermore, adjustment of the true glycated albumin concentration by adding 4.5% might provide a more precise measurement for determining glycaemic control in such patients.