Glycated hemoglobin and risk of mortality in dialysis patients with diabetes mellitus: a systematic review and dose-response meta-analysis

Article type
Authors
Wang W1, He Q1, Sun X1
1Chinese Evidence-based Medicine Centre and CREAT Group, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University and Collaborative Innovation Centre, Chengdu, China
Abstract
Background: Evidence suggested that either excessively high or low HbA1c level could result in adverse outcomes. The non-linear relationship between HbA1c level and adverse outcomes among patients receiving dialyssi, however, has not been established.

Objectives: We conducted a dose-response meta-analysis aiming to explore if there is a non-linear relation between HbA1c level and mortality in diabetic patients receiving dialysis

Methods: We searched PubMed, EMBASE and Cochrane Central Register of Controlled Trials (CENTRAL) from inception to November 2018. Eligible studies should randomized controlled trials (RCTs), non-RCT and observational studies that assessed the relationship between HbA1c level and mortality in in diabetic patients receiving dialysis. A modified version of the Cochrane Collaboration’s tool and Newcastle-Ottawa Scale (NOS) were used to assess risk of bias for RCTs and observational studies. We performed a dose-response meta-analysis to investigate the possible relationship between HbA1c level and mortality in dialysis patients with DM. We used adjusted HR as the effect measure and a one-stage robust error meta-regression (REMR) model to fit the potential non-linear trend.

Results: A total of 19 studies involving 113,119 participants were included in the data analysis, of which, 13 were prospective cohort studies, 5 were retrospective cohort studies, and the other one was RCT. All the 18 cohort studies selected participants from the same population, had confident in ascertaining exposure and control, and well adjusted for prognostic factors. Fifteen studies had adequately followed up and only five studies reported similar co-intervention between groups. The one RCT adequately generated their randomization sequence, concealed treatment allocation, blinded participants, caregivers and outcome assessors, and free from selective reporting.The dose-response meta-analysis analysis showed J-shaped association between HbA1c and mortality (P < 0.05 for non-linear test). The HbA1c level at about 7% had the lowest all-cause motality. Both low and high HbA1c level were associated with increased risk of all-cause mortality compared with the reference group of 7%: (HbA1c at 5%: HR=1.03, 95% CI 0.99-1.07; HbA1c at 6%: HR=1.01, 95% CI 0.99-1.02; HbA1c at 8%: HR=1.02, 95% CI 1.01-1.02; HbA1c at 9%: HR= 1.05, 95% CI 1.04-1.06; HbA1c at 10%: HR=1.10, 95% CI 1.08-1.13; HbA1c at 11%: HR=1.15, 95% CI 1.09-1.20).Fourteen studies investigated the association between the HbA1c level and all-cause mortality in hemodialysis patients, and the dose-response analysis also showed a J-shape association between HbA1c level and mortality, and HbA1c at 7% had the lowest mortality.

Conclusions: Current evidence suggested J-shaped relationship between HbA1c level and mortality in diabetic patient with dialysis, and patients with HbA1c at about 7% had the lowest mortality.