### Viewpoint on:

**Domislovic M., Domislovic V., Fucek M., Jelakovic A., Gellineo L., Dika Z., Jelakovic B.: Should the CKD EPI Equation Be Used for Estimation of the Glomerular Filtration Rate in Obese Subjects?. ****Kidney and Blood Pressure Research DOI 10.1159/000526115**

The definition of chronic kidney disease (CKD) using estimated glomerular filtration (eGFR) equations (indexed to a standard body surface area [BSA] of 1.73 m^{2}) has been recognized as problematic in obese subjects for many years. The magnitude of misclassification of CKD in such subjects using indexing by BSA can be quite substantial.

Domislovic and co-workers examined this issue in a random sample of rural-living subjects in Croatia, in which obesity (body mass index [BMI] >30 kg/m2) was rather common (30%). They compared the prevalence of CKD (defined by an absolute single threshold of eGFR-creatinine <60 mL/min/1.73 m2 kg, using the CKD-EPI equation of 2012) in indexed and de-indexed eGFR (applying two different formulas for BSA) in obese and non-obese subjects. In addition, they compared CKD prevalence using an eGFR formula incorporating height and weight (Salazar-Corcoran equation). They also examined the influence of obesity and leanness (BMI 24 kg/m2) on the prevalence of “hyperfiltration,” defined as an eGFR at or greater than the 95th percentile by age decade, and sex. The bias, precision, and accuracy of each eGFR equation according to BMI categories were investigated. Importantly, no measures of true GFR were included in the study.

In confirmation of earlier studies, they found that CKD-EPI eGFR equations indexed to BSA overstated the prevalence of CKD in obese subjects compared to a de-indexed eGFR formula or one using height and weight as variables. Simultaneously, such a BSA-indexed formula understated the prevalence of glomerular hyperfiltration. Opposite results were found for lean subjects. Thus, an indexed eGFR formula is unreliable in about 1/3 of the population. It is recommended that de-indexed eGFR formulas, or ones incorporating height and weight, be used for estimating CKD prevalence in general populations, containing subjects who are obese (BMI >30 kg/m^{2}) or lean (<25 kg/m^{2}). These confirmatory findings in a cohort of a rural-living general population raise questions concerning the reported epidemiology of CKD in the general population using eGFR formulas indexed for BSA, especially when obesity or leanness is common.

Richard Glassock