The drop in glomerular filtration rate often observed in the first weeks of nephroprotective treatment, whether because of protein restriction or renin-angiotensin system blockade, can be interpreted as a marker of the relationship between the intensity of baseline glomerular anomalies and the potential treatment benefit. In this issue, Oshima et al. reported a post hoc analysis of the Canagliflozin and Renal Events in Diabetes With Established Nephropathy Clinical Evaluation (CREDENCE) study, the results of which confirm that such a relationship is also observed under sodium-glucose cotransporter 2 inhibition in diabetic patients with chronic kidney disease.