The honest starting point on longevity supplements: none has been shown to extend human lifespan. A small number correct deficiencies. A smaller number support a specific function with reasonable evidence behind them.
The rest are sold on mechanism, animal data, or nothing at all. Every entry below is graded on evidence strength rather than popularity, and several conclude against the compound entirely.
The most-studied supplement in sports science, and one of the few where the evidence is genuinely settled. The mechanism, the dosing protocols that hold up, and an honest account of what it doesn’t do despite the claims.
Deficiency turns out to be common even in sunny countries, which surprises people every time. What the vitamin actually does, what the mortality evidence supports, and the dose that reliably corrects a genuinely low level.
Fish oil has strong mechanistic backing and a considerably messier trial record than the supplement aisle suggests. The cardiovascular and inflammatory evidence, and the REDUCE-IT result that muddied the conversation afterwards.
A protein shake is food in a convenient format, not a separate category of nutrition. Where whey genuinely helps, what leucine content and fast digestion contribute, and when whole food does the same job just as well.
The form printed on the label changes what the supplement is good for, and oxide is the cheapest for a reason. How magnesium works in the body, the sleep and blood pressure evidence, and which form suits which purpose.
Zinc matters for immune function and sits unusually close to a ceiling where more becomes counterproductive. The deficiency-correction evidence, the immunosenescence connection, and where the sensible upper limit falls.
B12 deficiency is serious, genuinely common in some groups, and irrelevant for most of the people supplementing it. Who is actually at risk, what the neurological stakes are, and what the trials found in everybody else.
CoQ10 is most defensible for one specific group: people taking statins who report muscle complaints. How it works, the oxidative stress evidence, and why that statin connection is more nuanced than it’s usually presented.
Turmeric’s active compound is poorly absorbed, which makes bioavailability the entire story rather than a footnote. How curcumin works, the anti-inflammatory and LDL evidence, and what formulation genuinely changes.
The animal data on NAD+ precursors is genuinely interesting. The human data measures surrogate markers rather than outcomes. What the trials have shown so far, and the gap between raising a level and changing anything you’d feel.
Most melatonin doses on sale are many times what the research actually uses, and more isn’t better here. What it does to your body clock, what the evidence supports, and how the dosing convention drifted so far off course.
The compound behind the French Paradox headlines, and a useful case study in a hypothesis outrunning its data. The sirtuin theory, what more than 150 human trials found, and the bioavailability problem sitting underneath.
Spermidine is tied to autophagy, a genuinely important process and a heavily overused marketing word. What the human data actually shows, and the considerable distance between an observed association and a tested intervention.
A senolytic candidate with striking mouse results and human trials still in progress. What the animal work showed, how the estimated human doses were arrived at, and why that extrapolation deserves caution.
Called “nature’s metformin” often enough that the comparison deserves examining rather than repeating. AMPK activation, what the trials actually found, and how the effect size holds up next to GLP-1 medications.
Read the full entry →A 2023 Science paper put taurine back into the longevity conversation, alongside a separate and much older cardiovascular literature. This entry keeps the two evidence bases apart, because they support noticeably different claims.
The clearest illustration of why more antioxidants isn’t automatically better, and can occasionally be worse. The 2025 Cochrane review, where the harm signal appears, and why below-RDA and above-RDA are entirely different situations.
A ranked triage across every supplement tracked here: worth it, conditional, promising but unproven, and the two with trial evidence of harm. The honest sorting eliminates most of the list.
Start from what a supplement can actually do. None has been shown to extend human lifespan. What some can do is correct a deficiency, or support one function with reasonable evidence behind it. Those are much smaller claims than the packaging makes, and they are the only ones the research supports.
Evidence strength and popularity are unrelated. The best-selling products in this category include several that the trials do not support, and two where the trial evidence points at harm rather than benefit. Grading by evidence rather than by sales volume reorders the list almost completely.
The strongest cases are unglamorous. The compounds with settled human evidence tend to be cheap, old and boring. What they have in common is decades of trials with outcomes rather than surrogate markers, which is a much harder bar than a plausible mechanism.
Mechanism is not evidence. Many frontier compounds target real biology, including several hallmarks of aging, and still have thin human outcome data. A compelling pathway explains why something might work. It does not establish that it does, and the gap between the two is where most of this category lives.
Animal results rarely transfer cleanly. Striking mouse findings are common here, and the dose extrapolations from them are shakier than they look. This is the same pattern that runs through senolytics and much of the experimental field.
Food and training outrank almost all of it. A supplement is a small adjustment on top of a much larger base. Protein intake and aerobic capacity both carry more evidence than anything in this category, and no capsule substitutes for either.
None has been shown to extend human lifespan. A small number correct deficiencies, and a smaller number support a specific function with reasonable evidence. Those are real but modest claims. Most of what is sold in this category rests on mechanism, animal data, or nothing at all.
A supplement is regulated as food and needs no proof of effect before sale. A drug must show efficacy and safety for a stated condition before approval. That difference explains why the evidence bar for what reaches a supplement shelf is so much lower than for anything in a pharmacy.
The unglamorous ones. Creatine monohydrate is the most studied supplement in sports science and the evidence is genuinely settled. Vitamin D3 matters where a deficiency exists, which is common even in sunny countries. Omega-3s have strong mechanistic backing and a messier trial record than the aisle suggests.
Some can. Two compounds tracked here have trial evidence of harm rather than benefit, and antioxidants illustrate the pattern: correcting a deficiency helps, while pushing intake well above it can do damage. More is not a safe default in this category.
Mouse lifespan studies are shorter, cleaner and use doses that rarely scale sensibly to a human body. Striking rodent findings are common in this field and the extrapolations from them are shakier than they look. Human outcome trials are the bar, and most frontier compounds have not cleared it.
Price tracks marketing more than evidence. Several of the best-supported compounds are among the cheapest available, while some of the most expensive rest on mechanism alone. Formulation occasionally matters, notably where absorption is the whole problem, but cost is a poor guide on its own.
Usually not. The clearest case for supplementing is correcting a measured deficiency. If a marker is already in range, adding more rarely moves an outcome and occasionally moves it the wrong way. Testing first turns a guess into a decision.
No, and the gap is not close. Protein intake and aerobic capacity each carry more evidence than anything in this category. A supplement is a small adjustment on top of a much larger base, and it cannot substitute for the base being absent.
Published personal protocols vary enormously and agree on very little, which is itself the useful signal. Where they overlap it tends to be on the compounds with settled evidence. Where they diverge, you are looking at one person’s judgement about unsettled science, not a consensus. A protocol is not evidence, however well known its author.
Not as a package. Stacks are a marketing structure rather than a research finding, and no trial has tested one as a combination. The evidence attaches to individual compounds for individual purposes, so the sensible approach is to work out which ones apply to you and skip the rest.
It depends entirely on what “work” means for that compound. Correcting a measured deficiency shows up in bloodwork over weeks to a few months. Anything claiming to slow ageing has no timeframe, because there is no marker that would confirm it happened.
More than the price does. Supplements are regulated loosely in most countries, and what is on the label is not always what is in the bottle. Independent third-party testing is the thing worth looking for. Beyond that, a more expensive product is usually a better-marketed one rather than a better-made one.
Every marker on this page, benchmarked against your own age and sex.
Start your 7-day free trial