Résumé
In nephrology, chronic kidney disease is defined by both proteinuria and measurement of glomerular filtration rate (GFR). This article focuses on GFR and different ways to define its normal reference values. In this context, we compare two perspectives: first the reference values defined by measuring GFR in normal individuals (the oclassical way') and secondly a fixed cut-off value at 60 mL/min/1.73 m(2) according to the associated mortality risk (the oprognostic way'). Following the classical way, we can assert that normal GFR values are largely over 60 mL/min/1.73 m(2) in healthy subjects, at least before the age of 70 years. However, we know that GFR physiologically decreases with age, and in adults older than 70 years, values below 60 mL/min/1.73 m(2) could be considered normal. Following the oprognostic way', the fixed cut-off of 60 mL/min/1.73 m(2) has been retained in the K-DIGO guidelines. However, we challenge this concept and the fact that the variable oage' is poorly taken into account in these data. There is an obvious discrepancy between the reference values defined either by the oclassical way' or by the oprognostic way' which we think could be largely reduced, if age was better taken into consideration in these definitions.