Estimating Glomerular Filtration Rate Calculated by Serum Creatinine Lacks Precision and Accuracy in Adults with Type 2 Diabetes with Preserved Renal Function

We evaluated the performance of creatinine-based estimating equations for glomerular filtration rate (GFR) in adults with type 2 diabetes without nephropathy compared to measured GFR by gold-standard inulin clearance methods. mGFR was performed 4 times over the course of a 1 month treatment period in the context of a small randomized placebo-controlled trial measuring the intrarenal hemodynamics function responses to the dipeptidyl peptidase 4 inhibitor sitagliptin. Thirty-two adults with type 2 diabetes (randomized 1:1 to placebo or sitagliptin) completed the trial, and all participants underwent mGFR by inulin clearance and GFR estimated by serum creatinine using the MDRD (eGFR_MDRD) and CKD-EPI (eGFR_CKD-EPI). Performance of eGFR equations was evaluated using measurements of bias (mean difference), precision (standard deviations [SD]), and inaccuracy (defined as the proportion of eGFR that differed by >20% of mGFR). The effects of treatment and time on bias were evaluated using linear mixed effects models. Mean±SD age, diabetes duration, HbA1c, and BMI was 60±8 years, 9±7 years, 7.2±0.8%, and 31.0±6.2 kg/m², respectively, and 12(38%) participants were female. At baseline, mean mGFR was 113±24, mean eGFR_MDRD was 93±12, and mean eGFR_CKD-EPI was 94±9, all in mL/min/1.73m². When all 128 observations (32 participants measured 4 times) were evaluated, both equations substantially underestimated mGFR. For the eGFR_MDRD, mean bias was -21.5 mL/min/1.73m², precision was 22.7 mL/min/1.73m², and 46% of observations differed by more than 20%. Results were similar for eGFR_CKD-EPI. No time or treatment effects on bias were observed. In adults with type 2 diabetes, eGFR equations underestimated mGFR, lacked precision and accuracy, and had lower performance at higher ranges of mGFR. More accurate measures of kidney function may improve clinical decision-making in adults with type 2 diabetes.