The impact of diabetes on patients' survival in dialysis patients with non-diabetic renal disease and in patients who develop diabetes during chronic dialysis.

BACKGROUND It is well known that dialysis patients with diabetic nephropathy have a poor prognosis, but data concerning the survival of dialysis patients with diabetes plus a non-diabetic primary nephropathy or the survival of patients who develop diabetes after the start of regular dialysis are scarce. AIM AND METHODS We reviewed the survival of two cohorts of dialysis patients in whom diabetes mellitus was associated with non-diabetic primary nephropathy. In the first cohort (18 patients with a primary diagnosis of APKD) diabetes mellitus precede hyperazotaemia, whilst the second cohort of 34 patients developed diabetes after the start of regular dialysis. We compared the survival of each group of patients to the survival of a group of dialysis patients with a primary diagnosis of diabetic nephropathy, and to the survival of each control group of non-diabetic dialysis patients. Within each case series, groups were similar according to age at start of RRT, and place of treatment. All patients were selected among those alive in treatment at 31 December 1986 and were followed up to 31 December 1991. RESULTS In both case series the survival of patients with diabetes was similar irrespective of the primary diagnosis (Lee-Desu statistics: first cohort P=0.43; second cohort, P=0.08). Moreover, the survival of patients either with diabetic nephropathy or with diabetes in association with non-diabetic primary nephropathy was significantly worse compared to the survival of the non-diabetic patients (Lee-Desu statistics: first case series P=0.02 and P<0.01; second case series P<0.05 and P<0.01). Logistic regression showed that survival was negatively associated to diabetes and age but not to sex, duration of diabetes and diagnosis of diabetic nephropathy. CONCLUSIONS Our limited data show that the survival of diabetic patients on regular dialysis is poor, irrespective of the primary cause of renal failure and of the duration of diabetes. These data need confirmation and further study.