GLP-1 Beyond Weight: Sleep Apnea, Kidneys, Liver and Heart Failure — What the Trials Measured

A sleep clinic suggests a weight-loss injection. A nephrologist mentions one. A hepatologist writes one into a plan for a fatty liver. If that has happened to you, the useful question is not whether these drugs work but a narrower one: which drug, at which dose, was tested in which patients, against which endpoint. The four answers differ, and the queries people type — "Ozempic for sleep apnea", "Mounjaro for heart failure" — mostly name the wrong pairing.

This is informational material. It does not replace a consultation with a doctor. Doses appear below only to identify what a trial tested; they are not a dosing recommendation.

Sleep apnea: what was measured, and in whom?

Tirzepatide, not semaglutide, is the drug with a sleep-apnea indication: the FDA approved Zepbound on 20 December 2024 for moderate to severe obstructive sleep apnea in adults with obesity, alongside a reduced-calorie diet and more activity. It was the first drug approved for that use.

The evidence is SURMOUNT-OSA: two randomized, double-blind, placebo-controlled phase 3 trials, 52 weeks, maximum tolerated tirzepatide of 10 or 15 mg, in 469 adults with obesity and moderate-to-severe apnea and without type 2 diabetes — 234 not using positive airway pressure (PAP), 235 using it. Baseline apnea–hypopnea index (AHI, breathing pauses per hour of sleep) was 51.5 and 49.5; baseline BMI 39.1 and 38.7. At week 52, AHI fell by 25.3 events/hour on tirzepatide versus 5.3 on placebo off PAP (difference −20.0; 95% CI −25.8 to −14.2), and by 29.3 versus 5.5 on PAP (difference −23.8; −29.6 to −17.9).

The widely quoted figures beyond AHI — about 18% and 20% weight loss, and 42% and 50% reaching remission or mild apnea without symptoms against 16% and 14% on placebo — come from the manufacturer's approval release, not the journal abstract. The FDA's own sentence is narrower: both studies showed a significant AHI reduction, and more people on tirzepatide reached remission or mild disease.

What this does not show. No result on crashes, heart attacks or death. People with type 2 diabetes were excluded, and people with apnea who do not have obesity were not the population. AHI fell in those who stayed on PAP too, which is not evidence that PAP should be stopped. Follow-up is one year.

Kidneys: what did FLOW actually test?

FLOW tested semaglutide 1 mg weekly — the diabetes dose, not the 2.4 mg weight-management dose — in 3,533 adults with type 2 diabetes and chronic kidney disease, on top of usual care, with median follow-up 3.4 years. Enrolment required diabetes plus either eGFR 50–75 with high albuminuria, or eGFR 25 to under 50 with albuminuria above 100 mg/g.

The primary endpoint was time to the first of: kidney failure (dialysis, transplant, or eGFR below 15), a sustained drop in eGFR of at least 50%, or death from kidney or cardiovascular causes. First events occurred in 331 people on semaglutide and 410 on placebo; hazard ratio 0.76 (95% CI 0.66–0.88). The kidney-only components gave HR 0.79 (0.66–0.94) and cardiovascular death HR 0.71 (0.56–0.89); eGFR declined 1.16 mL/min/1.73 m² per year more slowly, major cardiovascular events HR 0.82 (0.68–0.98), death from any cause HR 0.80 (0.67–0.95). Serious adverse events were 49.6% versus 53.8%. The company's statement of a 4.9% absolute risk reduction at three years is the sponsor's figure, not a number in the journal abstract. The company also announced FDA approval of this use on 28 January 2025; the agency's own release was not opened for this article.

What this does not show. FLOW is not a study of kidney disease without diabetes, and not a study of the weight-loss dose. Among the 550 participants already taking an SGLT2 inhibitor at baseline, the primary-outcome hazard ratio was 1.07 (95% CI 0.69–1.67), against 0.73 (0.63–0.85) in those not taking one, with an interaction p-value of 0.109 — a subgroup too small to claim either extra benefit or harm on top of an SGLT2 inhibitor. And the trial was stopped early for efficacy at a planned interim look, so the published effect is the one that crossed the stopping line.

Liver: what was approved, and on what endpoint?

Semaglutide 2.4 mg has an accelerated approval for noncirrhotic MASH (metabolic dysfunction-associated steatohepatitis) with moderate to advanced fibrosis, stages F2–F3. The manufacturer dates it 15 August 2025; the FDA notice confirms the accelerated pathway and the biopsy endpoint but does not print that date.

ESSENCE is an ongoing phase 3 trial: 1,197 people with biopsy-proven MASH and F2 or F3 fibrosis, randomized 2:1 to semaglutide 2.4 mg weekly or placebo, planned to run 240 weeks. The 72-week biopsy analysis used 534 versus 266 participants. Resolution of steatohepatitis without worsening fibrosis: 62.9% versus 34.3% (difference 28.7 points; 95% CI 21.1–36.2). Fibrosis improvement of at least one stage without worsening steatohepatitis: 36.8% versus 22.4% (14.4 points; 7.5–21.3). Gastrointestinal events were more common on semaglutide.

What this does not show. The approved endpoint is a biopsy — a surrogate, as the FDA's own notice says. Part 2 of the same trial, at 240 weeks, is the test of death, transplant and liver events, and it was not the basis of the approval. The label population is F2–F3 without cirrhosis; simple fatty liver without steatohepatitis and fibrosis was not the enrolled disease. A separate phase 2 trial in MASH with compensated cirrhosis (71 people, 48 weeks) found no significant fibrosis improvement on semaglutide 2.4 mg: 11% versus 29% on placebo (odds ratio 0.28; 95% CI 0.06–1.24; p=0.087), and no difference in MASH resolution. The group is small, so this is not evidence of harm; it is an absence of demonstrated benefit where there is no approval either.

Tirzepatide in MASH is phase 2 and not an approved use. SYNERGY-NASH randomized 190 people with F2–F3 for 52 weeks: resolution without worsening fibrosis was 10% on placebo versus 44%, 56% and 62% at 5, 10 and 15 mg. Fibrosis improvement was 30% on placebo versus 55%, 51% and 51%, and the intervals for the two higher doses barely clear zero. The authors wrote that larger and longer trials are needed. A phase 2 result is not a treatment option.

Heart failure: which endpoint moved?

All three trials enrolled heart failure with preserved ejection fraction (HFpEF) plus obesity — not reduced ejection fraction. No FDA heart-failure indication was verified for this article.

• STEP-HFpEF: 529 adults, ejection fraction ≥45%, BMI ≥30, no diabetes, semaglutide 2.4 mg, 52 weeks. The dual primary endpoints were a symptom score (KCCQ clinical summary, 0–100, higher is better) and percent weight change: KCCQ rose 16.6 versus 8.7 (difference 7.8; 95% CI 4.8–10.9), weight −13.3% versus −2.6%. Hospitalization and death were not primary endpoints.

• STEP-HFpEF DM: 616 adults with HFpEF, ejection fraction ≥45%, BMI ≥30 and type 2 diabetes, same dose and duration. KCCQ 13.7 versus 6.4 (difference 7.3; 4.1–10.4).

• SUMMIT: 731 adults, NYHA class II–IV, ejection fraction ≥50%, BMI ≥30, tirzepatide up to 15 mg, median follow-up 104 weeks. Cardiovascular death or a worsening heart-failure event occurred in 36 of 364 (9.9%) versus 56 of 367 (15.3%), hazard ratio 0.62 (95% CI 0.41–0.95); KCCQ at 52 weeks 19.5 versus 12.7 (difference 6.9; 3.3–10.6).

What this does not show. SUMMIT's composite splits apart on inspection: the worsening-heart-failure part moved (29 events, 8.0%, versus 52, 14.2%; HR 0.54, 0.34–0.85), while cardiovascular death did not (8, 2.2%, versus 5, 1.4%; HR 1.58, 0.52–4.83) and all-cause death was 19 versus 15 (HR 1.25, 0.63–2.45). A composite carried by one component is not a death benefit, and eight events against five cannot settle that question either way. Treatment was stopped for adverse events, mostly gastrointestinal, in 6.3% versus 1.4%.

What to do with your numbers

The practical step is to check whether you are in the population that was studied — a question about four specific numbers rather than about the drug.

Which number defines the indication. Sleep apnea: the AHI from a sleep study, plus obesity, plus tirzepatide — semaglutide has no such indication. Kidneys: eGFR and the urine albumin-to-creatinine ratio, plus type 2 diabetes, plus semaglutide 1 mg. Liver: the fibrosis stage from a biopsy (F2–F3, no cirrhosis), plus semaglutide 2.4 mg. Heart failure: ejection fraction — ≥45% in the two semaglutide symptom trials, ≥50% in the tirzepatide event trial — plus a symptom score — and here there is evidence without a verified approval.

Which document your number comes from. An approval release, a label and a journal paper can quote different figures for the same trial; worth asking which is being read to you.

What not to change on your own. Nothing in the apnea data supports stopping PAP; AHI fell in the group that kept using it. FLOW looked at this and could not settle it: among the 550 participants already on an SGLT2 inhibitor the primary-outcome hazard ratio was 1.07 (0.69–1.67), against 0.73 (0.63–0.85) in the rest, interaction p 0.109 — the authors read that as benefit consistent across both groups, with power too low to measure its size in the smaller one. These are conversations with the clinician who prescribes, not conclusions from a text.

These values — AHI, eGFR, albumin-to-creatinine ratio, fibrosis stage, ejection fraction — usually sit in separate reports from separate years. Kept in one place, where the trend is visible, they make the eligibility question answerable in a ten-minute appointment.

Sources

• Loomba R et al. Semaglutide 2.4 mg once weekly in patients with non-alcoholic steatohepatitis-related cirrhosis: a randomised, placebo-controlled phase 2 trial. Lancet Gastroenterol Hepatol. 2023. PMID 36934740 · doi:10.1016/S2468-1253(23)00068-7 · NCT03987451

• Malhotra A et al. Tirzepatide for the Treatment of Obstructive Sleep Apnea and Obesity (SURMOUNT-OSA). NEJM, 21 June 2024. doi:10.1056/NEJMoa2404881 · NCT05412004

• Perkovic V et al. Effects of Semaglutide on Chronic Kidney Disease in Patients with Type 2 Diabetes (FLOW). NEJM, 2024. PMID 38785209 · doi:10.1056/NEJMoa2403347 · NCT03819153

• Mann JFE et al. FLOW, SGLT2-inhibitor subgroup analysis. Nature Medicine, 2024. doi:10.1038/s41591-024-03133-0

• Sanyal AJ et al. Phase 3 Trial of Semaglutide in Metabolic Dysfunction–Associated Steatohepatitis (ESSENCE). NEJM, 30 April 2025. PMID 40305708 · doi:10.1056/NEJMoa2413258 · NCT04822181

• Loomba R et al. Tirzepatide for MASH with Liver Fibrosis (SYNERGY-NASH), phase 2. NEJM, 8 June 2024. doi:10.1056/NEJMoa2401943 · NCT04166773

• Kosiborod MN et al. Semaglutide in HFpEF and Obesity (STEP-HFpEF). NEJM, 2023. doi:10.1056/NEJMoa2306963 · NCT04788511

• Kosiborod MN et al. Semaglutide in obesity-related HFpEF and type 2 diabetes (STEP-HFpEF DM). NEJM, 2024. PMID 38587233 · NCT04916470

• Packer M et al. Tirzepatide for HFpEF and Obesity (SUMMIT). NEJM, online 16 November 2024. PMID 39555826 · doi:10.1056/NEJMoa2410027 · NCT04847557

• Regulator and sponsor statements used for indications and dates: FDA news release on Zepbound for obstructive sleep apnea (20 December 2024); the FDA notice on accelerated approval of semaglutide for MASH on a biopsy endpoint with confirmation running to week 240; the manufacturer's announcements of FDA approval for chronic kidney disease (28 January 2025) and MASH (15 August 2025), whose agency releases were not opened here; and the manufacturer's SURMOUNT-OSA approval release for the weight and remission percentages.

Informational material. It does not replace a consultation with a doctor and is not a recommendation for or against treatment.