Ozempic for Kidney Disease: What the FLOW Trial Changed

If you have type 2 diabetes and chronic kidney disease, the GLP-1 conversation changed in 2024. Novo Nordisk's FLOW trial showed semaglutide cut the combined risk of kidney failure, sustained 50% loss of kidney function, kidney death, and cardiovascular death by 24% in 3,533 adults followed for about 3.4 years. The trial was stopped early for benefit in October 2023, published in the New England Journal of Medicine in May 2024, and led to a first-in-class FDA approval on January 28, 2025 that added a kidney indication to Ozempic.

That approval sits next to angiotensin blockers, SGLT2 inhibitors, and finerenone as the fourth pillar of treatment for diabetic kidney disease, with semaglutide the only GLP-1 drug in the group. Wegovy was not part of the approval. The big open questions are how much the benefit adds on top of SGLT2 inhibitors, how it generalizes to people without diabetes, and how it compares with tirzepatide, where the evidence is newer and the story is similar.

FLOW enrolled 3,533 adults with type 2 diabetes and chronic kidney disease who were already on standard kidney protection, meaning a maximum tolerated dose of a RAAS inhibitor. Half got once-weekly semaglutide 1.0 mg, the dose used in Ozempic, and half got placebo. Mean baseline kidney function was an eGFR of 47, and most participants had heavy albuminuria. The primary endpoint was a composite of four things that matter for kidney patients: kidney failure (needing dialysis or transplant), sustained 50% loss of eGFR, kidney death, and cardiovascular death.

The primary result was a hazard ratio of 0.76 with a confidence interval of 0.66 to 0.88, a 24% relative reduction in those events. Cardiovascular events alone fell 18%, and all-cause mortality dropped 20%. The kidney-specific composite, stripping out cardiovascular death, fell 21%. eGFR declined about 1.16 mL/min/1.73 m² per year more slowly on the drug, and urinary albumin (a measure of kidney protein leak) was about 38% lower. The trial was stopped early in October 2023 because the data monitoring committee judged that the benefit was clear before the planned follow-up was complete.

The FDA approved Ozempic for the new indication on January 28, 2025, with label language that lets prescribers use it to reduce the risk of kidney disease worsening, kidney failure, and death from cardiovascular disease in adults with type 2 diabetes and chronic kidney disease. That is a meaningful expansion: Ozempic was already approved for type 2 diabetes management and cardiovascular risk reduction, and the kidney indication makes it the only GLP-1 cleared specifically for CKD outcomes.

The approval is dose-specific (0.5, 1, and 2 mg of Ozempic) and brand-specific. Wegovy, the higher-dose semaglutide product for weight management, did not get a CKD indication, even though it shares the same active drug. For Medicare patients, this matters: Part D cannot cover GLP-1 drugs used for weight loss alone, but Ozempic for diabetes-plus-kidney disease is on label and is covered, typically with a prior authorization that documents the type 2 diabetes diagnosis.

The kidney effect is partly mechanical and partly something else, and the mechanism is the interesting part. The mechanical part is the things you would expect: lower body weight, better glucose, lower blood pressure (about 2 to 5 mmHg systolic), all of which take strain off the kidneys.

The "something else" is more direct. GLP-1 receptors sit on cells in the proximal kidney tubule and the small vessels feeding the glomerulus. Activating them reduces intraglomerular pressure, lowers the kind of inflammation that drives diabetic kidney injury, and shifts how sodium is handled, producing a mild natriuretic effect that reduces blood pressure even before weight loss kicks in. The REMODEL mechanistic trial, with MRI and paired kidney biopsies before and after a year of treatment, found semaglutide cut urinary albumin by about 40%, lowered renal vascular resistance, reduced fat in the kidney itself, and most-changed gene expression in glomerular endothelial cells. The drug seems to do something the kidney recognizes as kidney-protective beyond the metabolic effects alone.

This is the most important open question in everyday practice, because SGLT2 inhibitors (Farxiga, Jardiance) are already first-line for diabetic kidney disease. The honest answer is that FLOW was not designed to test the combination. Only 15.6% of patients were on an SGLT2 inhibitor at baseline, which leaves the analysis underpowered.

What the prespecified subgroup did show was no signal that semaglutide's effect was blunted in the SGLT2 group: hazard ratios were directionally similar in both, with overlapping confidence intervals and a p-value for interaction of 0.109. KDIGO, the international kidney guideline body, read that as enough to recommend GLP-1 as an add-on for type 2 diabetes patients with CKD who are already on metformin and an SGLT2 inhibitor, and the American Diabetes Association did the same in its 2026 Standards of Care. But neither group claims the additive size of the benefit is established. A dedicated combination trial would settle it; one does not yet exist.

Tirzepatide does not have a CKD indication, but the evidence around kidney protection has caught up fast. The SURPASS-CVOT trial, published in NEJM in December 2025, compared tirzepatide directly against dulaglutide (another GLP-1) in 13,299 adults with type 2 diabetes and established cardiovascular disease over a median four years. Tirzepatide was non-inferior on the primary cardiovascular endpoint and pulled ahead on weight loss and A1C.

A prespecified kidney analysis published in Lancet Diabetes and Endocrinology in May 2026 added that tirzepatide cut a four-component kidney composite versus dulaglutide with a hazard ratio of 0.77 (confidence interval 0.68 to 0.88), driven by less new-onset macroalbuminuria and slower eGFR decline. A pooled SURPASS analysis across earlier trials showed similar albuminuria reductions in patients with type 2 diabetes. None of this is a head-to-head against semaglutide on hard kidney outcomes, and Lilly's dedicated SURMOUNT-MMO trial does not read out until October 2027, but the direction is consistent.

FLOW was a diabetic CKD trial, but the kidney story may extend further. The SELECT trial enrolled 17,604 adults with obesity and established cardiovascular disease but no diabetes, and a secondary analysis published in Nature Medicine in May 2024 found semaglutide 2.4 mg cut a five-component kidney composite with a hazard ratio of 0.78. About a fifth of SELECT participants had baseline markers of kidney disease, and the eGFR benefit was larger in those with lower starting kidney function.

A smaller 2024 Nature Medicine trial in non-diabetic adults with overweight or obesity and CKD reported semaglutide 2.4 mg cut urinary albumin by about 52% over 24 weeks. These are encouraging signals, but they are secondary analyses and short trials rather than dedicated kidney-outcomes work, and no GLP-1 carries a non-diabetic CKD indication. That is a genuine evidence gap, not a regulatory oversight.

A few things worth knowing if you have CKD and are considering or already taking a GLP-1. First, semaglutide does not need a dose adjustment for kidney function. The drug is almost entirely bound to albumin in the blood and barely cleared by the kidneys, so even severe renal impairment increases exposure by only about 22%, and dialysis does not change the pharmacokinetics. Second, the typical GI side effects (nausea, occasional vomiting) matter more when kidney function is borderline, because dehydration can cause acute kidney injury; staying hydrated and titrating slowly are the standard cautions. Third, semaglutide does not seem to worsen kidney function during treatment; eGFR trajectory was better on drug than placebo in FLOW.

Dialysis and transplant are the harder gaps. A pooled safety analysis published in Diabetes Care in June 2026 followed 165 patients who started dialysis on a semaglutide trial and found no excess adverse events, but efficacy data on dialysis are observational rather than randomized. Kidney transplant recipients have small cohort data showing weight and glucose benefits without obvious interactions with calcineurin inhibitors, but again, no proper trial. For those settings, the decision belongs to a nephrologist who knows the case.

FLOW moved semaglutide from a diabetes-and-weight drug to a drug with a real, measured kidney effect: a 24% reduction in major kidney and cardiovascular events, slower eGFR decline, less albumin in the urine, and an FDA indication to back it up. KDIGO and the ADA have added GLP-1 receptor agonists to the standard sequence for diabetic CKD, and the evidence around tirzepatide has caught up enough to make the class effect look real even if the labels lag. For someone with type 2 diabetes and kidney disease, the conversation with a clinician now includes whether to add a GLP-1 to existing therapy.

The limits are honest. The combination with SGLT2 inhibitors is recommended but not formally proven. The non-diabetic kidney case rests on secondary analyses. Tirzepatide does not yet carry a CKD label. And the heaviest kidney patients (eGFR under 25, dialysis, transplant) are still on a thinner evidence base. Within those boundaries, GLP-1s are now part of the standard kidney-care toolkit, which is not where the field was eighteen months ago.