Longevity Myths: What If the Headline Was Never in the Study?

In August 2014, a university press release carried a headline almost anyone would want to believe: "Type 2 diabetics can live longer than people without the disease."

The study behind it was real. British researchers followed 78,241 people who started metformin, the most common diabetes pill, and matched each to people without diabetes. The metformin group lived slightly longer. Within a few years, a diabetes pill that had been on the market for decades was being called "the anti-aging drug" on podcasts, forums and supplement sites.

In 2022, Danish researchers ran the same comparison on national health registers. People who started metformin died at a higher rate than people without diabetes, not a lower one. The trial that would settle the question has still not enrolled a single person.

Nobody in this story lied.

So what if most longevity myths start with a true study?

How a diabetes pill became "the longevity drug"

It started in the lab, where the results were real and modest. Metformin lengthened the lives of worms by up to 36% and of mice by 6% to 14%, depending on the strain, according to a 2026 review in Aging. In 2016, a program that tests drugs in mice at three laboratories at once found that metformin alone "did not significantly extend lifespan". Combined with another drug, rapamycin, it did.

Then came the 2014 records study. The researchers compared 78,241 people starting metformin with 90,463 people without diabetes. Before adjustment, the metformin group had 14.4 deaths per 1,000 person-years and the comparison group 15.2. After adjustment, the metformin group's median survival came out about 15% longer.

The same study had a second finding that travelled less well. People with diabetes who started a different pill, a sulfonylurea, had a median survival 38% shorter than people without diabetes.

What a records study cannot do. Nobody was assigned to metformin at random. Doctors choose who gets it first, and it is not given to people whose kidneys work very poorly, for example. The researchers adjusted for what the records held. No adjustment reaches what the records leave out.

The press release went one step further. It said metformin "may have beneficial effects not only on patients with diabetes but also for people without". The study had no healthy people taking metformin. From there, "diabetics on metformin outlived matched non-diabetics" became "metformin makes you live longer".

The retest. In 2022, a Danish team compared metformin users with people without diabetes, using registers for the whole country. After adjustment for other illnesses and social factors, metformin users died at a rate 32% higher. Among twins where one had diabetes and one did not, the rate was 64% higher.

The trial that would settle it has not started. TAME was designed with an agreement from the FDA in 2015: more than 3,000 adults aged 65 to 79 without diabetes, metformin or placebo, about six years, at 14 US sites. In September 2026 it was reported as still awaiting funding, with no participants enrolled. Estimates of its cost run from $45 million to $75 million.

None of this means metformin does nothing for aging. It means nobody knows yet. Our article Metformin for Longevity: What If It Was Never Meant to Work Alone? covers what is known, and a case for using it alongside other treatments.

Five questions to ask of any longevity claim

Each question catches one way a true finding grows on its way to you. You do not need a science degree to ask them, only the habit.

1. In what: cells, worms, mice or people?

Animal results are a start, not an answer. Researchers who traced treatments first tested in animals at least ten years earlier found that half reached any study in people, 40% reached a randomized trial and 5% were approved. The median wait for approval was ten years. In the press release study, 36% of releases stretched animal or cell work to people.

2. Compared with whom?

A study that watches who happened to take something cannot tell you the something caused the result. Only a trial that assigns people at random can. One in three of those press releases turned a link into a cause. Metformin's 2014 headline came from watching, and the watching did not repeat.

3. Measured what: years, illness or a marker?

Many longevity studies measure something that moves easily: a blood level, a "biological age" score, a hormone. A rise in NAD+ in the blood is not the same as more healthy years, and a younger score on an aging clock is not a longer life. Ask whether anyone counted heart attacks, dementia, frailty or deaths.

4. How much, and how much of it?

Two questions in one. The size of the effect: in the 2014 study, deaths ran 14.4 against 15.2 per 1,000 person-years before adjustment, a gap most people would not call dramatic. And the amount used: a laboratory study that showed metformin draining NAD+ from cancer cells used concentrations hundreds of times higher than those found in the blood of people who take it.

5. Who is telling you, and what do they sell?

A podcast host with a supplement line, a clinic with an injection, a news site paid by clicks, and us. Selling something does not make a claim false. It means the claim deserves the four questions above.

For anything injected, add one more: who made it? In 2025, the FDA placed a recall of injectable NAD+ in its most serious class, Class I, for bacterial toxins (endotoxins). In 2019 it warned pharmacies after patients became ill from glutathione injections made with ingredients meant for supplements.

The five questions, applied to our own series

We sell most of the treatments in our series on aging, so here are the same questions turned on them. The strongest evidence in people belongs to the GLP-1 medications, and to B12 for people who run low. For NAD+ and glutathione, the evidence is earlier: promising in animals and small studies, not yet tested in the form we offer.

A thin row does not mean a treatment fails. It means the question is still open, and you should hear that from whoever sells it to you. That includes us.

What if the seller asked the questions first?

That is the rule we hold our own articles to. Each one in this series:

• Says what was studied, in what, and what was measured, with a link to the study.
• Keeps what is known apart from what we think, and labels our view as ours.
• Says when a question is still open, including for the treatments we sell.
• States the price, with the telehealth consultation included.

Keep reading: Longevity Medicine: What If You Started With What You Want to Change? · Metformin for Longevity: What If It Was Never Meant to Work Alone? · NAD+ Injections and Aging: What If You Never Ran Low? · NMN vs NAD+: What If You Skipped the Ingredient and Took the Real Thing? · Glutathione Injections: What If You Could Refill Your Master Antioxidant? · Sermorelin and Aging: What If Your Body Only Needed a Nudge? · Benefits of B12 Shots: What If You Are Not Tired, Just Running Low? · Do GLP-1 Medications Slow Aging? What the Research Shows

Questions people ask

Is metformin an anti-aging drug?

Not as far as anyone has shown. It extends life in worms and some mice, and one 2014 records study found people with diabetes on metformin outlived people without diabetes. A Danish study using national registers in 2022 did not find that. The trial in people without diabetes, TAME, has not enrolled anyone.

Why do health headlines exaggerate?

Often the stretch starts before the journalist. In a study of 462 university press releases, 40% exaggerated advice, 33% turned a link into a cause and 36% stretched animal results to people. News stories were far more likely to exaggerate when the press release did.

How can I tell whether a longevity study was done in people?

Look for the words "mice", "rats", "worms", "cells" or "in vitro" in the first lines of the study summary. If the headline does not say who was studied, assume it was not people until you check.

What is the difference between a records study and a randomized trial?

A records study watches people who chose, or were prescribed, a treatment. A randomized trial decides by chance who gets the treatment and who gets a placebo. Only the second can show that the treatment caused the result.

Are compounded injections FDA-approved?

No. Compounded medications are prepared by licensed pharmacies for individual prescriptions and are not reviewed by the FDA for safety or effectiveness. That is why it matters which pharmacy made yours.

Does Goprolean sell the treatments it writes about?

Yes. That is why each article states what is not known, and why this one exists.

Sources

• The association between exaggeration in health related science news and academic press releases, Sumner and others, BMJ, 2014.
• Can people with type 2 diabetes live longer than people without diabetes?, Bannister and others, Diabetes, Obesity and Metabolism, 2014.
• Type 2 diabetics can live longer than people without the disease, Cardiff University press release, via ScienceDaily, August 2014.
• Reassessing the evidence of a survival advantage in type 2 diabetes treated with metformin compared with controls without diabetes, Keys and others, International Journal of Epidemiology, 2022.
• Analysis of animal-to-human translation shows that only 5% of animal-tested therapeutic interventions obtain regulatory approval for human applications, Ineichen and others, PLOS Biology, 2024.
• Metformin at the convergence of aging and longevity, Manneh, Alasmar and El Hajj, Aging, 2026. The worm and mouse figures are read through this review.
• Longer lifespan in male mice treated with a weakly estrogenic agonist, an antioxidant, an alpha-glucosidase inhibitor or a Nrf2-inducer, Strong and others, Aging Cell, 2016.
• TAME: the aging trial with metformin that never enrolled a patient, Innovita Research, September 2026.
• Ability of metformin to deplete NAD+ contributes to cancer cell susceptibility to metformin cytotoxicity, Zhuang and others, Frontiers in Oncology, 2023.
• Cellular and Molecular Mechanisms of Metformin Action, LaMoia and Shulman, Endocrine Reviews, 2021.
• FDA issues Class I recall of NAD+ injection due to elevated endotoxin levels, HMP Global Learning Network, October 2025.
• FDA highlights concerns with using dietary ingredient glutathione to compound sterile injectables, US Food and Drug Administration, June 2019.
• The rows on GLP-1 medications, B12, sermorelin, glutathione, NAD+ and NMN summarise our articles on each, where their own sources are listed.