Estimated Glomerular Filtration Rate Variability in Patients with Diabetes Receiving SGLT2 Inhibitors Versus DPP4 Inhibitors

Background: Major clinical trials regarding sodium–glucose cotransporter 2 inhibitors (SGLT2is) have focused on estimated glomerular filtration rate (eGFR) slope and composite kidney outcomes, with limited evaluation of eGFR variability as a kidney outcome or its long-term implications in patients receiving SGLT2i versus placebo. Methods: This retrospective study analyzed 3777 propensity score-matched patients with type 2 diabetes receiving either SGLT2i or dipeptidyl peptidase-4 inhibitor (DPP4i) between June 2016 and December 2021. Each patient had eGFR data at three time points before (−15, −9, and −3 months) and after (3, 9, and 15 months) the drug-index date. eGFR variability was assessed using the coefficient of variation (COV) and standard deviation (SD) of eGFR values. Results: SGLT2i therapy was associated with a significant reduction in eGFR variability by both COV (from 0.072 (0.001) to 0.069 (0.001); p = 0.014) and SD (mL/min/1.73 m²) (from 5.34 (0.07) to 4.82 (0.07); p < 0.001). In contrast, DPP4i therapy resulted in increased COV (from 0.072 (0.001) to 0.080 (0.001); p < 0.001) and no significant improvement in SD (from 5.06 (0.07) to 5.22 (0.07); p = 0.082). Greater reduction in eGFR variability was observed in SGLT2i relative to DPP4i with high pre-treatment eGFR variability, pre-existing chronic kidney disease, or rapid pre-treatment eGFR decline. Greater pre-treatment eGFR variability predicted risk of major adverse kidney events (MAKEs) and abrupt kidney decline in DPP4i-treated patients, but not in those on SGLT2is. SGLT2i consistently reduced the risk of MAKE and abrupt kidney decline across the spectrum of pre-treatment eGFR variability, with a greater risk reduction on the MAKE for SGLT2i versus DPP4i therapy with a higher pre-treatment eGFR variability estimated by SD (p interaction = 0.014). Conclusions: SGLT2i therapy improved eGFR variability and reduced the risk of adverse kidney outcomes compared to DPP4i, particularly in patients with higher pre-treatment eGFR variability.