Kidney effects of tirzepatide across albuminuria, filtration markers, and clinical outcomes: a parent-trial-level systematic review and meta-analysis
In brief
Tirzepatide lowers albuminuria 27% versus insulin, but kidney protection remains unproven
Across randomized trials, tirzepatide reduced urinary albumin by 27% versus insulin, but placebo-controlled evidence was uncertain and filtration markers showed no clear benefit. A post hoc analysis suggested fewer hard kidney outcomes in a high-risk group, but the result was not a formal subgroup test, end-stage kidney disease was not clearly reduced, and dedicated kidney-outcome trials are still needed.
- Journal
- Diabetes research and clinical practice (Q1)
- Published
- 13 November 2026
- Study design
- Systematic review of cohort studies
- Evidence level
- Level 2, Moderate (CEBM 2a)
- Authors
- Shuangwen Li, Shilong Li, Pengcheng Xu, Junya Jia
- PMID
- 43025465
- DOI
- 10.1016/j.diabres.2026.113668
Why clinicians should know about it
- Picked for Internal Medicine (paper of the day, 18 November 2026): Systematic review evaluates tirzepatide kidney effects and clinical outcomes
Abstract
Whether tirzepatide's kidney effects extend from albuminuria lowering to filtration preservation and clinical outcomes remains uncertain. We searched PubMed, Embase, and the Cochrane Library through August 2026 and mapped randomized reports to independent parent trials. Ten parent trials provided kidney data; eight informed each principal synthesis. Versus insulin, tirzepatide reduced urinary albumin-to-creatinine ratio by 27.0% (95% CI, 16.0%-36.6%; I2 = 37.8%; moderate certainty), whereas placebo-controlled evidence was uncertain (-18.1%; 95% CI, -43.2% to 18.3%; very low certainty). Fixed-time creatinine-based estimated glomerular filtration rate showed no clear benefit (0.36 mL/min/1.73 m2; 95% CI, -0.34 to 1.07; low certainty). In SURPASS-CVOT, approximate post hoc aggregate contrasts suggested larger high-risk effects for filtration and a hard-only composite, but were multiplicity-unadjusted and not formal individual-participant interaction tests. In that high-risk stratum, the hard-only hazard ratio versus dulaglutide was 0.70 (0.57-0.86), while end-stage kidney disease was not clearly reduced (1.04; 0.70-1.55). Albuminuria lowering versus insulin was reproducible; hard-outcome benefit was population- and endpoint-dependent, and dedicated kidney-outcome trials remain needed.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.