Written by Dr Zeeshan Afzal (MBBS) — Medical Doctor & Medical Content Lead, Welzo
Medically reviewed · Evidence-based · Aligned with UK clinical guidance
Last updated: July 2026 · Reading time: ~14 minutes
Everything attributed to Sam Altman comes from his own recorded interviews and public statements, or from reporting of them. Where a claim is widely repeated online but has no primary source, we say so rather than passing it on.
Important: This article discusses metformin, a prescription-only medicine in the UK. It is licensed to treat type 2 diabetes. Taking it for anti-ageing purposes is off-label, unlicensed use. Nothing here is a recommendation to obtain or take it, and buying prescription medicines without a prescription is both unsafe and unlawful.
Sam Altman is the first figure in this series to have actually disclosed part of his routine — and the most interesting thing about it is not the drug everyone focuses on. It is the method underneath: test first, then correct what is actually low. The drug itself has better evidence against it than most coverage admits.
Sam Altman, born in 1985, is the chief executive of OpenAI, the company behind ChatGPT, and formerly president of the startup accelerator Y Combinator. He is one of the most prominent figures in technology, and in his early forties is considerably younger than the other executives whose health routines attract this level of attention.
Most technology executives who invest in longevity science say nothing about their personal habits. Altman has been comparatively open. He has discussed his use of metformin, his approach to blood testing, his vegetarianism, his fasting window and his morning light exposure.
That candour makes this the only article in our celebrity series where a genuine, sourced routine exists to examine. It also means there is something concrete to assess against the evidence — and on the central element, the evidence is not favourable.
Unusually, there is a real answer to this — though not the one most articles give.
Altman has described undergoing quarterly blood tests across a year, which identified that he was low in four things: methylated vitamin B12, omega-3, iron and vitamin D3. He then supplemented to correct those specific deficiencies, and has said he now retests roughly every eighteen months.
That is the whole disclosed list. Note what it is: not a curated performance stack, but four corrections to measured deficiencies in one individual.
Doses, brands, forms and timings. Anyone publishing "Sam Altman's supplement stack" with milligram figures is inventing them, exactly as with the other figures in this series.
The important point, and it is easy to miss: Altman's four supplements are not a template. They are the answer to his blood results. A vegetarian is at elevated risk of low B12, iron and long-chain omega-3, which is why his list looks the way it does. Your list would only look the same if your bloods did. The transferable part of his routine is the testing, not the products.
| Claim | Status | Notes |
|---|---|---|
| Takes metformin off-label for anti-ageing purposes | Confirmed | Reported by MIT Technology Review; Altman has acknowledged it publicly |
| Supplements methyl B12, omega-3, iron and vitamin D3 | Confirmed | Identified via a year of quarterly blood tests; doses never stated |
| Retests bloods roughly every 18 months | Confirmed | His own account |
| Vegetarian since childhood | Confirmed | His own statement |
| Fasts around 15 hours a day | Confirmed | Reported consistently, including by BBC Science Focus |
| Uses bright light exposure in the morning while working | Confirmed | Reported; consistent with circadian light practice |
| Invested about $180 million in Retro Biosciences | Confirmed | Announced 2021–2022 |
| Specific doses of anything | No evidence | Never disclosed |
| Takes NMN, resveratrol or rapamycin | No evidence | Frequently assumed; never stated |
Metformin is a biguanide, the most widely prescribed medicine for type 2 diabetes worldwide. It lowers hepatic glucose production, improves insulin sensitivity and increases muscle glucose uptake. It is cheap, generic and has decades of safety data in the population it is licensed for.
The excitement traces largely to a 2014 analysis suggesting that people with diabetes taking metformin outlived matched non-diabetic controls. If true, that is a remarkable finding: a drug so beneficial it more than cancels out the disease it treats.
That result launched a decade of interest, a proposed clinical trial, and a great many supplement-adjacent businesses.
Metformin is prescription-only. It is licensed for type 2 diabetes and, in some cases, conditions such as polycystic ovary syndrome. It is not licensed or approved anywhere as an anti-ageing treatment. A UK GP will not prescribe it to a healthy adult for longevity, and obtaining it without a prescription is unsafe and unlawful.
This is where the story turns, and where most coverage stops short.
The 2014 analysis that started the enthusiasm was re-examined, and a 2022 paper was unable to reproduce its findings, identifying methodological problems in the original. The replication failure circulates far less widely than the original claim.
The TAME trial — Targeting Aging with Metformin — was designed specifically to test the drug in non-diabetic adults. It has never been funded or launched, and has been stalled for roughly a decade. As BBC Science Focus put it, the obstacle is not the cost of the pills, which are pennies, but the $45–70 million needed to run the trial.
The practical consequence: there is currently no good evidence that metformin extends lifespan or healthspan in healthy people. Not negative evidence — absent evidence.
Two randomised controlled trials in older adults found that metformin interferes with the benefits of exercise. This is the most important section of this article.
Konopka and colleagues (2019), published in Aging Cell, found that metformin inhibited the mitochondrial adaptations to aerobic exercise training in older adults, blunting the expected improvements in cardiorespiratory fitness and whole-body insulin sensitivity. Metformin and exercise each improve insulin sensitivity alone; combined, they worked against each other rather than adding up.
The MASTERS trial (Walton and colleagues, 2019, also in Aging Cell) was a randomised, double-blind, placebo-controlled, multicentre study. Ninety-four adults aged 65 and over received either 1,700 mg of metformin daily or placebo, alongside fourteen weeks of supervised progressive resistance training. The researchers hypothesised metformin would enhance the muscle response, because it reduces inflammation. It did the opposite: metformin blunted the training-induced muscle growth.
The awkward conclusion. Resistance and aerobic training are among the best-evidenced interventions for healthy ageing that exist. Metformin's longevity benefit in healthy people is unproven. Taking an unproven drug that measurably reduces the benefit of a proven intervention is a difficult trade to justify.
Altman has said he has been vegetarian since childhood. This is directly relevant to his supplement list: vegetarian and vegan diets carry a well-established elevated risk of low vitamin B12, low iron and low long-chain omega-3 (EPA and DHA). Three of his four corrected deficiencies are the classic vegetarian pattern. The diet explains the stack.
He is reported to fast around fifteen hours daily, leaving roughly a nine-hour eating window. This is time-restricted eating rather than extended fasting.
Time-restricted eating is reasonably well tolerated and can help some people control total intake, which is where most of its measured benefit appears to come from. Evidence that it extends human lifespan independently of calorie intake is not established. It is a plausible, low-risk habit — not a proven longevity intervention.
One caution specific to his combination: a long daily fast alongside vegetarianism narrows the window in which to hit protein and micronutrient targets. That may be part of why his bloods showed what they did.
Altman is reported to use bright light exposure in the morning while working through email. Morning light is the strongest signal for setting circadian rhythm, and this is a sensible, evidence-aligned habit — particularly through a British winter, when natural morning light is limited.
Altman committed around $180 million to Retro Biosciences, a company whose stated mission is to add ten years to the healthy human lifespan, working on autophagy and cellular-reprogramming-inspired therapeutics.
As with Bezos and Altos Labs, the relevant consumer point is the gap: none of this is purchasable. These are early-stage therapeutic programmes, not products. Supplements marketed on the back of "the science billionaires are funding" are trading on an association that does not exist chemically.
Strip away the drug and the fasting window, and Altman does one thing that is straightforwardly good practice: he measured before he supplemented.
He did not assume he needed B12 because he is vegetarian. He tested, found four specific deficiencies, corrected those, and retests periodically to check the correction is holding. That is how supplementation is supposed to work, and it is the opposite of buying a celebrity stack.
Testing is most valuable when you have a reason to suspect a gap: a vegetarian or vegan diet, persistent fatigue, heavy menstrual bleeding, malabsorption conditions, older age, limited sun exposure, or long-term use of medicines that deplete nutrients. Vitamin D is the notable exception where UK guidance recommends supplementation for the general population regardless of testing, through autumn and winter.
If you want to follow the useful half of Altman's routine, a nutrition and vitamin blood test or a broader vitamin deficiency test is where it starts.
This is the clearest safety message in this series. Metformin is a prescription medicine, its longevity benefit in healthy people is unproven, it has randomised evidence of blunting exercise adaptations, it depletes B12, and it carries a rare but serious risk of lactic acidosis that rises with kidney impairment. If you have concerns about metabolic health, the appropriate step is a conversation with your GP, not an online pharmacy.
Of Altman's four supplements, iron is the one that carries real risk taken blind. Excess iron accumulates and can cause organ damage, and some people carry genetic conditions such as haemochromatosis that make supplementation actively dangerous. Iron should be taken on the basis of a blood test, not a diet type.
Time-restricted eating is inappropriate for anyone with a history of disordered eating, for people who are pregnant or breastfeeding, for those with type 1 diabetes, and for anyone on medication requiring food. If restricting your eating window feels like something you cannot stop tightening, that is a signal to speak to someone rather than optimise further.
This is the entire lesson. A baseline panel covering B12, folate, ferritin and vitamin D tells you what you actually need. Retesting after a few months tells you whether the correction worked.
If you are vegetarian or vegan, the likely candidates are B12, iron, long-chain omega-3, iodine and vitamin D — but likely is not the same as confirmed. Our guide to vegan supplement gaps covers the reasoning in detail.
What to look for on a label
A single named ingredient rather than a proprietary blend. The dose stated in the form that matters — elemental magnesium, EPA and DHA rather than total oil, named forms such as cholecalciferol or creatine monohydrate. And third-party batch testing for identity and purity, given how variable this category is.
Whatever your blood results point to, the same label standard applies. The Welzo Ultra Purity range is built on single-ingredient formulations, disclosed dosages, no proprietary blends, and batch testing for identity and potency.
Single-ingredient formulations with disclosed dosages and published certificates of analysis. Shown: Magnesium L-Threonate.
Browse the range →UK guidance advises supplementing vitamin D through autumn and winter for the general population, because sunlight at British latitudes is insufficient to synthesise it during those months. This is the one case where supplementing without testing is the official recommendation.
Given randomised evidence that metformin blunts both aerobic and resistance training adaptations, the ordering is clear. Training is the proven intervention. Guard it.
Off-label prescription drugs sourced without a prescription, and any supplement sold on the implication that a technology executive takes it.
He has said he supplements methylated vitamin B12, omega-3, iron and vitamin D3. These were identified through a year of quarterly blood tests that showed he was low in each. He has never disclosed doses or brands, and he retests roughly every eighteen months.
Yes. He takes it off-label for anti-ageing purposes, as reported by MIT Technology Review and acknowledged publicly. Metformin is licensed for type 2 diabetes, not for ageing, and is prescription-only in the UK.
There is no good evidence that it does in healthy people. The 2014 analysis that generated the interest was not replicated in a later re-examination, and the TAME trial designed to settle the question has never been funded or run. Two randomised trials in older adults found metformin blunted the benefits of aerobic and resistance training.
No. It is a prescription-only medicine licensed for type 2 diabetes and related conditions. A GP will not prescribe it to a healthy adult for anti-ageing purposes, and obtaining prescription medicines without a prescription is unsafe and unlawful.
Gastrointestinal effects are common, particularly when starting. Long-term use impairs vitamin B12 absorption and is a recognised cause of deficiency. Lactic acidosis is rare but serious, and the risk increases with kidney impairment. Long-term effects in healthy non-diabetic users have not been studied.
Almost certainly because he has been vegetarian since childhood. Low B12, low iron and low long-chain omega-3 are the classic pattern in vegetarian and vegan diets, and his blood tests showed all three. His list reflects his diet and his results, not a universal protocol.
Around fifteen hours a day, leaving roughly a nine-hour eating window. This is time-restricted eating rather than extended fasting.
Not proven in humans. Time-restricted eating is generally well tolerated and can help some people manage total intake, which is where most of its measured benefit appears to originate. Evidence that it extends human lifespan independently of calorie intake is not established.
For most targeted supplements, yes — especially iron, which can cause harm if taken when not needed. Vitamin D is the main exception: UK guidance recommends supplementation for the general population through autumn and winter regardless of testing.
A longevity biotechnology company Altman backed with around $180 million, working on autophagy and cellular-reprogramming-inspired therapeutics with the stated aim of adding ten years to healthy human lifespan. Nothing it develops is available as a consumer product.
Medical disclaimer. This article is for general information and is not medical advice. It is not a substitute for consultation with a qualified healthcare professional. Sam Altman has no connection to Welzo and does not endorse Welzo or any product mentioned. Nothing here is a recommendation to take metformin or any prescription medicine, and nothing here should be taken as a claim that any supplement prevents, treats, cures or slows any disease or the ageing process.
Speak to your GP or pharmacist before starting a supplement, a fasting protocol or a new training programme if you take prescription medication, are pregnant or breastfeeding, have kidney, liver, heart or metabolic disease, have a history of disordered eating, or are being treated for any chronic condition. Iron supplements should not be taken without a blood test confirming they are needed.
Βαθμολογημένη εξαιρετική από 28,951+ κριτικές
100% εγγύηση επιστροφής χρημάτων
Η εγγύηση επιστροφής χρημάτων σας εξασφαλίζει ότι θα λάβετε τα στοιχεία σας ή θα λάβετε πλήρη επιστροφή χρημάτων.
