Objective

Hemoglobin A1c (HbA(1c)) has been widely used as a clinically important assessment tool for outcome analyses related to glycemic control. However, because of special conditions in dialysis patients, including the uremic milieu, there is no HbA(1c) blood glucose (BG) equation specific for patients on dialysis. In this study, we sought to develop HbA(1c)-BG equation models for hemodialysis patients.

Research design and methods

We examined associations between HbA(1c) and random serum BG over time in a contemporary cohort of diabetic patients with hemodialysis treated in DaVita dialysis clinics. We identified 11,986 patients (63 ± 12 years old and 49% male) with 69,764 paired measurements of HbA(1c) and BG over the course of 5 years (2001-2006). Bootstrapping method was used to estimate average BG and corresponding HbA(1c) levels. The association was adjusted by patient factors using linear regression.

Results

Linear regression analyses yielded the following three regression equations: BG = 59.2 + 29.4 × HbA(1c) - 20.8 × Alb (R(2) = 0.483); BG = 104.8 + 29.7 × HbA(1c) - 18.4 × Alb - 4.7 × Hb (R(2) = 0.486); and BG = 82.9 + 30.7 × HbA(1c) - 16.5 × Alb - 5.4 × Hb + 0.3 × age + race (R(2) = 0.491). All our models showed stronger association than previous equation models (R(2) = 0.468 in the Diabetes Control and Complications Trial and A1c-Derived Average Glucose equations).

Conclusions

The association between HbA(1c) and BG in hemodialysis patients is different than that of patients with normal kidney function. Our analysis suggests that equations including serum albumin or hemoglobin are better for hemodialysis patients.

Background

Although hemoglobin A1c (HbA1c) has been widely used as a clinical assessment tool for outcome analyses related to glycemic control, the relationship between HbA1c and average blood glucose (BG) specific to peritoneal dialysis (PD) patients with diabetes has not been characterized. We sought to develop HbA1c-BG equation models for PD patients.

Methods

We examined associations between HbA1c and random serum BG values over time in a contemporary 5-year (2001-2006) cohort of DaVita PD patients with diabetes. We identified 850 patients (mean age: 58 ± 13 years, 56% male) with 4,566 paired measurements of HbA1c and BG. The bootstrapping method was used to estimate average BG and corresponding HbA1c.

Results

Linear regression analyses yielded the following HbA1c-BG equations: (1) BG (mg/dl) = 24.1 + 28.6 × HbA1c - 12.2 × albumin [adjusted R(2) (R(2)adj = 0.454)], (2) BG = 55.3 + 28.8 × HbA1c - 10.2 × albumin - 3.3 × Hb (R(2)adj = 0.457), and (3) BG = 69.5 + 28.7 × HbA1c - 10.1 × albumin - 3.7 × Hb - 0.1 × age + race/ethnicity (-10.1 African Americans, -5.4 other race/ethnicities; R(2)adj = 0.457). All models showed greater explanatory power of BG variation than previously established HbA1c-BG equation models defined within non-PD cohorts [R(2)adj = 0.446 for both the Diabetes Control and Complications Trial (DCCT) and the A1c-Derived Average Glucose (ADAG) equations].

Conclusions

The association between HbA1c and BG in PD patients is different than that of patients with normal kidney function. Our analysis suggests that equations incorporating serum albumin and/or Hb values better estimate the HbA1c-BG relationship in PD patients compared to equations using HbA1c alone.