The relationship between three eGFR formulas and hospitalization for heart failure in 54 486 individuals with type 2 diabetes

Abstract Objective This study aims to study the association between renal function and hospitalization for heart failure (HF) in individuals with type 2 diabetes. Methods Renal function was determined according to three formulas used to estimate glomerular filtration rate (eGFR): Cockcroft–Gault, modified diet in renal disease (MDRD) and chronic kidney disease epidemiology (CKD‐EPI). Proportional hazards regression models adjusted for age, sex, HbA 1c , blood pressure, smoking and cardiovascular comorbidities were constructed for each eGFR formula to estimate risk of hospitalization for heart failure. Systematic pairwise likelihood ratio tests of nested models were used to compare the predictive power of each eGFR formula. Results In 54 486 patients, evaluated over a median follow‐up of 7.0 years, a total of 5936 (10.9%) developed heart failure, with an excess risk in all eGFR categories below 60 mL/min/1.73 m 2 (reference: eGFR >90 mL/min/1.73 m 2 ). Hazard ratios ranged from 1.25 to 1.35 for eGFR 45–60 mL/min/1.73 m 2 , 1.62 to 1.66 for eGFR 30–45 mL/min/1.73 m 2 and 2.18 to 2.52 for eGFR <30 mL/min/1.73 m 2 in the three eGFR formulas. In pairwise comparisons, the model with the MDRD variable added significantly more information than the Cockcroft–Gault variable. For the model with the CKD‐EPI variable, no clear differences in predictive power for HF hospitalization existed in relation to the other eGFR formulas. Conclusion Patients with type 2 diabetes, with eGFR 45 to 60 mL/min/1.73 m 2 , have approximately 25–35% increased risk of hospitalization for HF, increasing with lower eGFR, to 2–2.5 times in those with eGFR <30 mL/min/1.73 m 2 . The MDRD formula for calculating eGFR is more predictive of hospitalization for heart failure than the Cockcroft–Gault formula. Copyright © 2016 John Wiley & Sons, Ltd.