m8kThe Librarian's Notebook

GLP-1 Drugs: What the Evidence Shows, Organ by Organ

SCIENCE & TECHNOLOGY · OCTOBER 7, 2026 · updated October 7, 2026

Four semaglutide (Wegovy) injection pens of different dosage strengths, laid out side by side in their packaging
Nelson R. de Lima Filho, CC BY 4.0, via Wikimedia Commons.

This is a living page. GLP-1 drugs, the semaglutide and tirzepatide family, keep getting credited with benefits that have nothing to do with the scale, and a new claim lands every few weeks. Rather than write a fresh post each time, I'll keep this scorecard current: every claimed benefit, what kind of study stands behind it, and how much weight it can bear. The two longer pieces on the Regimen are the detail behind it: the benefit list and what happens when people stop. This is reporting on published research, not medical advice. Last updated October 7, 2026.

The scorecard at a glance

Standing page · updated Oct 7, 2026
Four benefits are backed by large randomized trials. The newest claim, a younger-heart finding, is a preprint.
The September 22 preprint reports that semaglutide lowered the proteomic “biological age” of the heart by roughly 2 to 4 years against placebo, across 10,052 trial participants. That is a blood-protein score, not a clinical outcome, and it has not been peer reviewed.
Kidney, heart, liver, airway: randomized Mechanism and clocks: early Addiction, dementia: observational
24%lower risk of kidney failure, kidney or cardiovascular death in FLOW (3,533 people, semaglutide vs placebo)
20%fewer heart attacks, strokes and cardiovascular deaths in SELECT (17,604 adults without diabetes)
62.9% vs 34.3%liver inflammation (MASH) resolved without worse scarring, semaglutide vs placebo, ESSENCE
−29.3breathing interruptions per hour on tirzepatide (vs −5.5 placebo), SURMOUNT-OSA with CPAP
10,052trial participants in the organ-age preprint, pooled from five semaglutide trials
2–4 yrsyounger heart proteomic age vs placebo, by the preprint’s clock (not peer reviewed)
The newest claim: organ-age clocks
Heartabout 2 to 4 years younger
Kidney1 to 4 years
Brain1 to 4 years
Pancreas1 to 4 years
Lung1 to 4 years
Multi-organ1 to 4 years
Each bar spans the reported range of years younger than placebo, on an axis from 0 to 4 years (the right edge is 4); the preprint gives ranges, not a single number per organ. The clocks are blood-protein patterns trained elsewhere to predict mortality. The authors themselves say the work does not establish “biological rejuvenation, causal mediation or surrogate validity.” Cardiologist Eric Topol’s October 7 post summarising it listed the heart, artery and kidney clocks; the abstract I could read names heart, kidney, brain, pancreas, lung and multi-organ.
Timeline
2023
SELECT: 20% fewer major cardiovascular events in adults with obesity and heart disease but no diabetesNEJM. Randomized, 17,604 people.
2024
FLOW: kidney outcomes trial stopped early for benefit; SURMOUNT-OSA: tirzepatide cuts sleep-apnea severityNEJM. Randomized, hard or measured endpoints.
Apr 2025
ESSENCE part 1: semaglutide reverses liver inflammation in biopsy-confirmed MASHNEJM. Randomized, 800 people.
2026
REMODEL (kidney biopsies, 33-person subgroup) shows less inflammation inside kidney tissue; VA cohorts link GLP-1 use to less addiction and, with stopping, to more cardiovascular riskConference presentation and observational studies; none is a randomized outcome result.
Sep 22
Preprint: proteomic organ-age clocks improve on semaglutide across five placebo-controlled trialsResearch Square, posted Sep 22, not peer reviewed. Trials named in the keywords: SELECT, STEP-HFpEF, STEP-HFpEF DM, FLOW, SOUL.
Oct 7
Eric Topol shares the preprint on X: “consistently reduced proteomic organ age 2-4 years”His post gives the heart, artery and kidney clocks; the preprint abstract is broader.
Claim check
Randomized trialEarly or mechanisticObservationalNot established
GLP-1 protects the kidneys in people with diabetes and kidney diseaseFLOW, NEJM 2024: 24% lower risk of the composite outcome over a median 3.4 years; also 18% fewer major cardiovascular events and 20% lower all-cause death.
Randomized
Semaglutide lowers heart-attack and stroke risk even without diabetesSELECT, NEJM 2023: 20% lower risk over about 40 months; the benefit curve separated before most weight was lost.
Randomized
Semaglutide reverses fatty-liver inflammation (MASH)ESSENCE part 1, NEJM 2025: 62.9% vs 34.3% resolution without worse fibrosis; 36.8% vs 22.4% improved fibrosis stage.
Randomized
Tirzepatide eases obstructive sleep apneaSURMOUNT-OSA, NEJM 2024: with CPAP, apnea-hypopnea index fell 29.3 events/hour vs 5.5 on placebo. Hard to separate fully from weight loss.
Randomized
It calms inflammation inside the kidneyREMODEL: 40% lower albuminuria; paired biopsies in 33 people show fewer immune cells around the filters. Conference presentation, not yet peer reviewed.
Early
It makes organs biologically younger (proteomic clocks)Preprint, 10,052 participants: heart about 2 to 4 years younger, other organs 1 to 4. Authors disclaim rejuvenation and surrogate validity.
Preprint
Heart-age change explains a third to a half of the benefitThe preprint calls its 34% to 49% mediation figures exploratory.
Exploratory
It cuts addictionVA cohort of 606,434 veterans: hazard ratios 0.75 to 0.86 across substances. Association, not proof of cause.
Observational
It lowers dementia riskClaims-data cohort: 10% lower risk vs DPP-4 inhibitors; not distinguishable from SGLT-2 inhibitors. Novo’s EVOKE trial is the real test.
Observational
Stopping the drug erodes the heart benefitVA study of 333,687 people: risk rose the longer people stayed off (4% higher at six months off, 14% at a year, 22% at two years). Observational.
Observational
GLP-1s “reverse aging”No study measures aging itself. A proteomic clock is a predictor of mortality, not a stopwatch on a person’s biology.
Not established
What I am watching for
Peer review of the organ-age preprint
Whether a journal accepts it, and whether the reviewers press on how the externally trained clocks were applied.
Tirzepatide and the clocks
The preprint is semaglutide only. Nothing yet says whether tirzepatide or newer drugs move the same scores.
EVOKE (Alzheimer’s)
Novo Nordisk’s randomized dementia trial was due to report in 2026; the dementia row stays observational until it does.
Anything after stopping
Whether kidney, liver and clock benefits fade off treatment the way the VA data suggest the heart benefit does.
Grades describe the study design behind each claim, not whether I think the result is true. A randomized trial can still be wrong and an observational signal can still be right; the grade only says how much each can carry. Numbers are from the papers and abstracts cited below; the organ-age figures are from the preprint abstract as posted September 22, 2026.

The full write-up

The old way to talk about these drugs was simple: they treat diabetes and obesity, and people lose weight. The list since then has grown in a way that is both impressive and easy to oversell. The honest way to read it is by rung. At the top are trials that randomized thousands of people and counted events that matter, such as kidney failure or a heart attack. In the middle are studies that show a plausible mechanism in a small number of people. At the bottom are cohort studies, which compare people who happened to take a drug with people who happened not to.

The newest claim: younger organs, by a blood test

On October 7 Eric Topol posted that semaglutide “consistently reduced proteomic organ age 2-4 years for the heart, artery, and kidney clocks across 5 placebo-controlled trials.” The source is a preprint, Semaglutide reduces proteomic organ-specific biological age across randomized clinical trials, posted on Research Square on September 22 and not yet peer reviewed.

What the abstract says: the authors analysed blood proteomics (levels of thousands of circulating proteins) from 10,052 participants across five placebo-controlled semaglutide trials, naming SELECT, STEP-HFpEF, STEP-HFpEF DM, FLOW and SOUL. They applied organ-specific mortality clocks that were trained elsewhere on other cohorts; each is a pattern of proteins that predicts who is likelier to die, read off as an “age.” Semaglutide lowered the heart’s biological age by about 2 to 4 years against placebo, and the kidney, brain, pancreas, lung and multi-organ clocks by 1 to 4 years. Frailty and cardiometabolic risk scores also improved over up to 156 weeks. The effects were not fully explained by weight loss, blood sugar or C-reactive protein, and an exploratory analysis suggested heart-age change could account for 34% to 49% of effects on some outcomes.

What it does not say, in the authors’ own words: the results support using these scores as a treatment-response readout “without establishing biological rejuvenation, causal mediation or surrogate validity.” A clock that moves is not the same as a person who is 3 years younger. It is a reason to expect the already-proven cardiovascular benefit to show up in the blood, and it fits the SELECT and FLOW results, which are the outcomes trials two of the five datasets come from. Because the datasets are those same trials, this is also less independent than “five trials” sounds. I have only read the abstract and Topol’s post, not the full methods, so I cannot say how the clocks were calibrated for these populations.

Update log

Where I could be wrong

Sources

  1. Dermit M, Colhoun HM, Goeminne LJE, Ryan D, Adelborg K, Belmont-Rausch D, et al. Semaglutide reduces proteomic organ-specific biological age across randomized clinical trials. Research Square preprint, posted September 22, 2026, not peer reviewed. rs-11082091
  2. Eric Topol (@EricTopol). Post on X, October 7, 2026. x.com
  3. Perkovic V, Tuttle KR, Rossing P, et al. Effects of Semaglutide on Chronic Kidney Disease in Patients with Type 2 Diabetes (FLOW). N Engl J Med, 391:109-121, 2024. doi:10.1056/NEJMoa2403347
  4. Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes (SELECT). N Engl J Med, 389:2221-2232, 2023. doi:10.1056/NEJMoa2307563
  5. Sanyal AJ, Newsome PN, Kliers I, et al. Phase 3 Trial of Semaglutide in Metabolic Dysfunction-Associated Steatohepatitis (ESSENCE). N Engl J Med, 392:2089-2099, 2025. doi:10.1056/NEJMoa2413258
  6. Malhotra A, Grunstein RR, Fietze I, et al. Tirzepatide for the Treatment of Obstructive Sleep Apnea and Obesity (SURMOUNT-OSA). N Engl J Med, 391:1193-1205, 2024. doi:10.1056/NEJMoa2404881
  7. Bjornstad P, Tuttle KR, et al. REMODEL, presented at the World Congress of Nephrology, Yokohama, 2026. PMC12559791
  8. The Regimen. The GLP-1 Benefit List Keeps Growing (October 2, 2026) and Does Stopping a GLP-1 Drug Raise Your Heart Risk? (September 22, 2026), which cite the VA addiction, VA discontinuation and dementia cohort studies in full.
GLP-1semaglutidetirzepatideOzempicWegovyclinical trialskidney diseaseheart diseasebiological ageproteomics
💬 Comment on X🔍 See what others said

Keep reading

TECHNOLOGY · SEPTEMBER 22, 2026 CagriSema vs. Tirzepatide: What the New Diabetes Trial Actually Shows CagriSema beat Lilly's tirzepatide on weight loss in a diabetes trial — but only against a 5mg dose, not tirzepatide's strongest one. TECHNOLOGY · OCTOBER 11, 2026 Atomic Machines came out of stealth with a machine that builds from code. Its first product is a power relay. Jeff Holden's Atomic Machines left six years of stealth on October 7, 2026, with $250 million and a claim that code alone can drive the manufacture of… TECHNOLOGY · OCTOBER 8, 2026 The Super Intelligence Scorecard: Which Model Is Best, by Whose Test Which Super Intelligence model is best right now? A living scorecard from LMArena, Artificial Analysis, ARC Prize, METR and Epoch AI, plus costs. TECHNOLOGY · OCTOBER 8, 2026 Gemini 4 Argon: Google's Benchmarks, Checked Gemini 4 Argon wins 13 of 19 of Google's benchmarks. Which rows Google ran itself, what the independent leaderboards say, and what it costs.

← Back to the Notebook