The Effectiveness of Sodium–glucose Cotransporter-2 Inhibitor Therapy in Slowing the Progression of Chronic Kidney Disease: A Systematic Review and Meta-Analysis of Randomized Controlled Trials and Primary Studies

Introduction: Chronic kidney disease (CKD) affects an estimated 788 million adults and is now the ninth leading cause of death worldwide, yet for decades renin–angiotensin system (RAS) blockade was the only pharmacological strategy proven to slow its progression. Sodium–glucose cotransporter-2 inhibitors (SGLT2i) have since been evaluated in dedicated kidney trials, cardiovascular outcome trials (CVOTs), heart failure trials and real-world cohorts, but these sources use heterogeneous definitions of kidney progression and enrol different populations. This review aimed to quantify the effectiveness of SGLT2i in slowing CKD progression by integrating randomised and real-world evidence, and to establish whether the effect is modified by diabetes status, baseline kidney function or trial population. Methods: The review followed the PRISMA 2020 statement. PubMed, Europe PMC, Semantic Scholar and ClinicalTrials.gov were searched from January 2011 to November 2026 (≤15 years) and supplemented by citation searching. Randomised controlled trials (RCTs), comparative studies, cohort studies and case series comparing an SGLT2i with placebo or non-SGLT2i care and reporting a kidney progression outcome were eligible; systematic reviews and meta-analyses were excluded from the characteristics tables but retained for the Discussion. Primary outcomes were a kidney-specific composite outcome and the annualised estimated glomerular filtration rate (eGFR) slope; secondary outcomes were the primary cardiorenal composite, kidney failure, all-cause mortality, cardiovascular death or hospitalisation for heart failure (HHF), serious acute kidney injury (AKI) and albuminuria. Risk of bias was assessed with RoB 2 and ROBINS-I. Effects were pooled with DerSimonian–Laird random-effects models with Hartung–Knapp–Sidik–Jonkman (HKSJ) adjustment and 95% prediction intervals (PI), and certainty was graded with GRADE. Results: Twenty-three studies were included (14 RCTs, n = 89,705; nine cohort studies, n = 206,423; 39 reports). SGLT2i reduced the kidney-specific composite outcome by 34% (hazard ratio [HR] 0.66; 95% CI 0.59–0.75; 12 RCTs; I² = 61%; PI 0.45–0.97; high certainty), with consistent effects in CKD-dedicated trials (HR 0.66; 0.59–0.74; I² = 18%) and type 2 diabetes CVOTs (HR 0.58; 0.51–0.65) and an attenuated effect in heart failure trials (HR 0.81; 0.59–1.10; P for subgroup difference = 0.09). In CKD-dedicated trials, SGLT2i reduced the primary cardiorenal composite (HR 0.70; 0.64–0.77), all-cause mortality (HR 0.85; 0.74–0.98), cardiovascular death or HHF (HR 0.75; 0.68–0.83) and serious AKI (HR 0.80; 0.68–0.95); kidney failure was reduced in RCTs (HR 0.63; 0.54–0.74). SGLT2i slowed eGFR decline by 1.61 mL/min/1.73 m² per year (95% CI 1.32–1.91; 10 RCT estimates), above the 0.75 mL/min/1.73 m² per year threshold in every trial, and lowered albuminuria by 19% (ratio 0.81; 0.76–0.86). Benefits were similar with and without diabetes (HR 0.64 vs 0.66; P for interaction = 0.92) and across baseline eGFR strata, including eGFR <30 mL/min/1.73 m² (HR 0.73; 0.63–0.85; P for interaction = 0.85). Real-world cohorts showed a concordant reduction in kidney events (HR 0.62; 0.54–0.73; low certainty). There was no evidence of small-study effects for the kidney composite (Egger P = 0.97). Discussion: The kidney-protective effect of SGLT2i is large, reproducible across agents, designs and continents, and independent of diabetes, baseline eGFR down to 20 mL/min/1.73 m² and RAS blockade. The weaker signal in heart failure trials reflects low kidney event rates, short follow-up and composite definitions sensitive to eGFR thresholds rather than an absence of nephroprotection, because eGFR slope benefits were preserved in those trials. The reductions in kidney failure, AKI and mortality indicate that the early haemodynamic eGFR dip should not deter initiation. Conclusion: SGLT2i constitute a foundational, disease-modifying therapy that slows CKD progression by approximately one-third, preserves roughly 1.6 mL/min/1.73 m² of eGFR per year, and reduces kidney failure, AKI and death across the spectrum of diabetic and non-diabetic CKD. Implementation should be prioritised for all eligible patients, while adequately powered trials are needed in eGFR <20 mL/min/1.73 m², dialysis, kidney transplantation and paediatric populations.