The Ozempic Aging Signal Came From an HIV Trial

A 32-week analysis of 84 adults with HIV-associated lipohypertrophy found weekly semaglutide slowed several DNA methylation aging clocks versus placebo. The paper appeared in Nature Communications on May 19, 2026.

It is the first randomized, placebo-controlled signal that a GLP-1 drug can move those clocks in people. The volunteers were not wellness clients. They were enrolled in a fat-distribution trial because HIV and its treatments had packed visceral fat around their organs.

The First Clock Signal Came From an HIV Fat Trial

People with HIV often show a faster biological-age pattern even when the virus is controlled on antiretroviral therapy. Prior work cited in a related Corley analysis puts that gap at about 3 to 7 years above chronological age. HIV-associated lipohypertrophy adds another load: extra visceral fat, chronic immune activation, and a higher cardiometabolic risk.

The parent study was phase 2b trial NCT04019197, a single-centre, double-blind comparison run from June 10, 2019, to July 28, 2022. Investigators randomized 108 adults 1:1 to weekly shots or placebo. Eight people (15%) in each arm left early. No one had type 1 or type 2 diabetes. Body-mass index had to be at least 25, and HIV-1 RNA had to stay under 400 copies per millilitre for six months.

Epigenetic age was never the original endpoint. Stored blood cells from week 0 and week 32 were pulled later. Of 92 people who finished both visits, 84 had paired methylation data: 45 on semaglutide and 39 on placebo. Mean age was 49 (standard deviation 12). Median body-mass index was 32.9. About 42% were women, though men made up 67% of the drug arm and 49% of the placebo arm.

THE RESEARCH PATH

1. June 10, 2019: Enrollment opens in Cleveland for a visceral-fat trial of weekly semaglutide in people with HIV-associated lipohypertrophy.
2. July 28, 2022: Enrollment closes after 108 people are randomized, 54 to each arm.
3. July 1, 2024: The parent body-composition results appear in Lancet Diabetes and Endocrinology.
4. March 9, 2025: Michael J. Corley presents the first epigenetic-clock readout at the Conference on Retroviruses and Opportunistic Infections.
5. May 19, 2026: Nature Communications publishes the full post hoc clock analysis.

Lead author Michael J. Corley, an associate professor of medicine at the UC San Diego School of Medicine and the Stein Institute for Research on Aging, has said this group can surface aging biology earlier because the same inflammatory and metabolic processes show up sooner than they do in the wider public. That is a research rationale. It is also how a special clinic population became the opening exhibit for a general-audience aging claim.

What the 32-Week Analysis Measured

The team ran a battery of DNA methylation clocks on peripheral-blood cells, then compared annualized change on drug versus placebo. Models adjusted for sex, baseline body-mass index, high-sensitivity C-reactive protein, and soluble CD163. The dose was 1.0 mg weekly after an 8-week titration, given as a subcutaneous shot, not the 2.4 mg Wegovy dose used in large heart trials.

The headline number is DunedinPACE, a third-generation pace-of-aging measure. Semaglutide was tied to a 0.09 lower pace versus placebo (95% CI -0.17 to -0.02, p=0.01), which the authors read as a 9 percent slower pace of biological aging. Second-generation clocks that track morbidity and death risk moved in the same direction. So did several organ-system clocks built from the same blood assay.

CLOCKS THAT MOVED, AND ONE THAT DID NOT

The post hoc analysis of 84 adults also reported slower readings on OMICmAge and RetroAge (both -2.2 years per year). Liver and kidney system clocks improved. Lung, musculoskeletal, immune, and hormone system clocks did not reach p<0.05. DNAmFitAge did not either (p=0.323). P-values were two-sided and unadjusted for the many clocks tested, which the paper flags as exploratory.

A related single-arm pilot in people with HIV and fatty liver disease (SLIM LIVER, 24 weeks, n=41, same 1.0 mg dose) was messier. Median DunedinPACE rose slightly (+0.018). Seventeen of 41 people (41.5%) still showed a drop in pace, and those people lost more liver fat. Twenty of 41 had a longer methylation-based telomere reading. That is a split cohort, not a clean replication.

The Company That Sells the Clocks Helped Write the Paper

Two co-authors, Varun B. Dwaraka and Ryan Smith, are listed as employees of TruDiagnostic, a Lexington, Kentucky, lab that sells consumer biological-age tests built on DNA methylation. Corley serves as a scientific advisor for the same company. The competing-interests statement in the paper records all three ties. TruDiagnostic software was used for some of the metrics.

That arrangement does not invent the statistics. It does put a vendor that needs aging clocks to stay clinically interesting inside the author list of the first randomized GLP-1 clock paper. After the preprint circulated in 2025, Smith described an average 3.1-year reversal of biological age, language that tracks the PCGrimAge point estimate and goes further than the paper’s own claim.

We are not saying that semaglutide reverses aging or makes people younger. What we are seeing is a signal that it may slow some of the biological processes associated with aging.

Michael J. Corley, PhD, associate professor of medicine, UC San Diego School of Medicine, university release

Physician-scientist Eric Topol (@EricTopol) called the preprint the first randomized-trial clock shift of its kind for any drug, then added the limit that still sits on the result: the participants had a body-mass index around 30, and the finding needs independent replication in people who are not overweight. Longevity clinics do not wait for that sentence. A 9% DunedinPACE move is already easy to package next to a mail-in methylation kit.

Visceral Fat Fell, and So Did Lean Mass

The clock paper keeps pointing at fat and inflammation as the likely path. The parent trial, published July 1, 2024, was designed to measure adipose tissue, not aging, and it did move the fat that wraps organs.

WHAT THE FAT TRIAL CHANGED IN 32 WEEKS

• Visceral fat: Abdominal visceral adipose tissue showed a 30.6% drop in visceral fat on semaglutide versus placebo (β -30.82 cm², 95% CI -50.13 to -11.51).
• Subcutaneous fat: Abdominal subcutaneous adipose tissue fell 11.2% (β -42.01 cm², 95% CI -75.49 to -8.52).
• Total body fat: Whole-body fat declined 18.9% on the log-transformed analysis.
• Lean mass: Investigators later wrote that about one third of the weight loss was lean mass, a change they said deserves more scrutiny in people who already develop sarcopenia.

Allison Ross Eckard of the Medical University of South Carolina, who led the parent trial, also reported drops in inflammatory markers over the same 32 weeks, including high-sensitivity C-reactive protein, interleukin-6, and soluble CD163. The clock models still showed a drug effect after those inflammation markers and body-mass index were put in the regression, which is why the authors argue they are not only watching weight come off.

The musculoskeletal system clock, the one that should care about muscle, did not move with statistical confidence (p=0.15). A drug can slow an inflammation clock in blood while the body still sheds lean tissue. In a group that already fights early sarcopenia, that is not a side note. It is the clinical trade the methylation printout does not price.

Heart Trials Already Changed How Long People Stay Alive

If the question is whether semaglutide can change the odds of dying, the aging clocks are late to the file. The SELECT trial assigned 17,604 adults with overweight or obesity and established heart disease, and without diabetes, to 2.4 mg weekly semaglutide or placebo. Mean follow-up was 39.8 months.

TWO TRIALS, TWO QUESTIONS

A primary heart event occurred in 569 of 8,803 people on drug (6.5%) and 701 of 8,801 on placebo (8.0%), a 20% lower rate of major heart events (hazard ratio 0.80, 95% CI 0.72 to 0.90). All-cause death was 4.3% versus 5.2% (hazard ratio 0.81, 95% CI 0.71 to 0.93). Cardiovascular death itself missed the trial’s strict significance bar (hazard ratio 0.85, p=0.07). Permanent stops for adverse events were 16.6% on drug and 8.2% on placebo.

Those are event counts, not a chemical tag on DNA. The Food and Drug Administration added a Wegovy heart-risk indication in 2024 on the back of that trial, for adults who already have cardiovascular disease and obesity or overweight. The HIV clock paper is being asked to do a different job: to recast the same molecule as a gerotherapy, a drug aimed at aging itself, on a 1.0 mg dose, in 84 people, over 32 weeks.

Several Clocks Moved, and One Resilience Measure Did Not

The authors are careful in the limitations section in a way the wellness rewrite is not. The analysis is post hoc. The sample is small. Follow-up is short. The cohort is specific to HIV-associated lipohypertrophy. Methylation was read in blood cells, so organ clocks are indirect summaries, not biopsies of heart or brain. Residual shifts in white-cell mix can still color a blood clock. And, in the paper’s own words, epigenetic clocks are statistical predictors rather than mechanistic measures of aging, so a lower reading is not proof of a longer life.

The Intrinsic Capacity clock, built to track physical and cognitive resilience, did not budge (p=0.312). Clocks aimed at causal aging or adaptive aging (CausAge, DamAge, AdaptAge) were less responsive than the mortality-trained set. That split is useful. It shows the result is not “every aging number fell.” It shows that the clocks most tightly tied to inflammation, fat, and death risk were the ones that moved after a metabolic drug.

Corley has said emerging data suggest GLP-1 drugs may reprogram cells in different organs, which could explain why several system clocks shifted together. That remains a hypothesis. The Stein Institute has talked about turning these readouts into individualized aging dashboards for clinic use. A dashboard built on a post hoc HIV sample will look more certain than the underlying evidence.

A Mouse Lifespan Study Lands After the Human Paper

By early September 2026, the live argument on the drug and aging had already left the 84-person blood test. A Nature paper from Danica Chen’s lab at the University of California, Berkeley, reported that late-life semaglutide extended lifespan in older female mice and improved several aging-related functions. That is an actual survival result, in a different species, on a different protocol. It is also the kind of finding the human clocks cannot supply.

The human file is still this: a fat-loss shot, at 1.0 mg, in people with HIV-associated lipohypertrophy, moved several blood methylation clocks over 32 weeks, including a 9% slower DunedinPACE reading, while lean mass came off with the visceral fat. Two employees of a biological-age testing company are on the paper. The lead academic author advises that company. Larger, longer trials in people without that HIV phenotype are not optional if anyone wants to call semaglutide an aging drug. They have not been done.

Frequently Asked Questions

What is HIV-associated lipohypertrophy?

It is a pattern of extra visceral and ectopic fat, often around the abdomen, that can appear in people with HIV after antiretroviral therapy even when the virus is suppressed. Tesamorelin (sold as Egrifta) is already an FDA-approved option for this fat pattern, and the semaglutide fat trial was run because other tools remain limited and cardiometabolic risk stays high.

What dose of semaglutide did the aging analysis use?

Participants titrated for 8 weeks and then stayed at 1.0 mg once weekly for 24 weeks, with every injection given in clinic by a nurse, a design that removes home-use missed doses from the comparison. That 1.0 mg level sits in the Ozempic diabetes range and is lower than the 2.4 mg Wegovy dose in SELECT, and the aging cohort also excluded known cardiovascular disease at entry.

What is the DunedinPACE clock?

DunedinPACE estimates how many years of biological aging accrue per calendar year, with 1.00 meaning a year of aging per year of life, and it was trained on repeated health measures from the Dunedin longitudinal study in New Zealand. The Corley team computed it with the public PACEProjector tool (DunedinPACE R package v0.99.0), then reported a 0.09 lower pace on semaglutide, which they translated as about 9% slower.

Has Wegovy been approved to cut heart attacks?

On March 8, 2024, the FDA cleared Wegovy (semaglutide 2.4 mg) to reduce cardiovascular death, heart attack, and stroke in adults who already have cardiovascular disease and obesity or overweight, used with diet and activity. The label still carries a boxed warning about thyroid C-cell tumors, a risk class inherited from rodent findings and printed for the whole GLP-1 family.

Disclaimer: This article is news reporting and analysis of published trials and is for information only. It is not medical advice, a recommendation to start or stop semaglutide (Ozempic, Wegovy, or any GLP-1 drug), or a judgment that any person should use these medicines for aging, weight, HIV-associated lipohypertrophy, or heart-risk reduction. Readers should talk with a licensed physician who knows their history, current medicines, and lab results before changing treatment. Figures, trial statuses, and regulatory indications reflect the papers and agency materials cited here and can change as new studies and label updates appear.