Stringent glycaemic control prolongs survival in diabetic patients with end‐stage renal disease on haemodialysis

Abstract

No suitable index or optimal target for diabetic control has been established for diabetic patients with end-stage renal disease (ESRD) undergoing haemodialysis. To address these issues, the single-centre observational study was conducted. Two hundred and forty-five diabetic ESRD patients (23.3% female; age at initiation of haemodialysis 61.7 ± 10.7 years) at start of haemodialysis between 1 January 1995 and 31 December 2004 were enrolled. Subjects were grouped according to glycaemic control level throughout the observational period as follows: mean postprandial plasma glucose (PPG) <8.9 mmol/L, 8.9 mmol/L ≤ PPG < 10.0 mmol/L, 10.0 mmol/L ≤ PPG < 11.1 mmol/L, 11.1 mmol/L ≤ PPG < 12.2 mmol/L and PPG ≥ 12.2 mmol/L; and HbA1c < 6.0%, 6.0-6.4%, 6.5-6.9% and ≥ 7.0%. Survival was then followed until 31 December 2005. Cumulative survival of groups of 10.0 mmol/L ≤ PPG < 11.1 mmol/L, 11.1 ≤ PPG < 12.2 and PPG ≥ 12.2 mmol/L was significantly lower than that for PPG < 8.9 mmol/L as determined by Kaplan-Meier estimation (P = 0.016, 0.009 and 0.031, respectively; log-rank test). In both uni- and multivariate Cox proportional hazard models, mortality hazard ratios were significantly higher for PPG ≥ 10.0 mmol/L than for PPG < 8.9 mmol/L (P = 0.002-0.021). Kaplan-Meier survival curves grouped by HbA1c levels showed no correlation between HbA1c and survival during the observational period. No significant difference in mortality hazard ratios was seen for any HbA1c groups evaluated by Cox proportional hazard model. Intensive management of diabetic control at a stringent mean on-study PPG < 10.0 mmol/L will improve the life expectancy in diabetic dialysis patients. However, no range of HbA1c values obtained in this study showed any clear difference in clinical outcomes.