Benefits regardless of diabetes and urine protein
Sodium-glucose cotransporter 2 inhibitors reduced kidney worsening and hospitalizations across diabetes status and urine protein levels.
*Meta-analysis of randomized trials; Level 1a (OCEBM).

Citation

Staplin N, Roddick AJ, Neuen BL, et al; SMART-C. Effects of Sodium Glucose Cotransporter 2 Inhibitors by Diabetes Status and Level of Albuminuria: A Meta-Analysis. JAMA. 2026;335(3):220-232. doi:10.1001/jama.2025.20835

Background

These medicines are recommended for chronic kidney disease, but some guidelines give weaker recommendations for people without diabetes or with lower urine protein. This study pooled large trials to clarify benefits and serious harms across these groups.

Patients

58,816 adults from 8 randomized trials; 48,946 with diabetes and 9,870 without diabetes; grouped by urine albumin-to-creatinine ratio (200 mg/g or more vs less than 200 mg/g).

Intervention

Sodium-glucose cotransporter 2 inhibitor added to usual care.

Control

Placebo added to usual care.

Outcome

Kidney disease progression, acute kidney injury, any hospitalization, death, and key serious harms.

Follow-up Period

Approximately 1.6 to 3.3 years (varied by trial and outcome).

Results

Group Outcome Hazard ratio (95% confidence interval) Absolute change (events per 1000 patient-years) Number needed to treat or harm (1 year)
Diabetes Kidney disease progression 0.65 (0.60 to 0.70) 15 fewer Treat 67
No diabetes Kidney disease progression 0.74 (0.63 to 0.85) 14 fewer Treat 72
Diabetes Acute kidney injury 0.77 (0.69 to 0.87) 4 fewer Treat 250
No diabetes Acute kidney injury 0.72 (0.56 to 0.92) 5 fewer Treat 200
Diabetes Any hospitalization 0.90 (0.87 to 0.92) 29 fewer Treat 35
No diabetes Any hospitalization 0.89 (0.83 to 0.95) 34 fewer Treat 30
Diabetes Death from any cause 0.86 (0.80 to 0.91) 5 fewer Treat 200
Diabetes Ketoacidosis (serious harm) 2.29 (1.42 to 3.71) 0.5 more Harm 2000
Ketoacidosis is a dangerous buildup of acids in the blood.
Analyses were intention-to-treat. Relative benefits were similar in higher vs lower urine protein groups; kidney benefits were larger in absolute terms when urine protein was 200 mg/g or more.

Limitations

Not all eligible trials had urine protein data; subgroup tests may miss real differences; absolute benefit estimates reflect trial populations and were not adjusted for competing risks.

Funding

No industry funding for meta-analysis; included trials were industry-funded.

Clinical Application

For chronic kidney disease, consider these medicines even without diabetes or with lower urine protein, while counseling people with diabetes about rare ketoacidosis risk.