How this list was made
Three tests. Is there a randomised trial, or a meta-analysis of them, with an outcome you would notice — strength, a fall, a fracture, a heart attack — rather than a blood marker? Is the effect large enough to matter at a dose a person would take? And is there a lab test that says whether you, specifically, need it? Most of the longevity supplement market fails the first test; nearly all of it fails the third.
The short list
| Supplement | Evidence | Dose the trials used | Who it is for |
|---|---|---|---|
| Creatine monohydrate | Strong — hundreds of RCTs; meta-analysis of 22 trials in adults over 55 (Chilibeck 2017): +1.4 kg lean mass and more strength with resistance training | 3–5 g daily, every day; no loading, no cycling | Anyone who lifts, and especially anyone over fifty. Measure kidney function as cystatin C, because it raises creatinine harmlessly. |
| Omega-3 (EPA/DHA) | Mixed — VITAL (25,871 adults, 1 g/day): no overall reduction in cardiovascular events or cancer, though fewer heart attacks in those eating little fish; REDUCE-IT (8,179 high-risk adults, 4 g/day icosapent ethyl): 25% fewer events | 1–2 g EPA+DHA daily for most; 4 g of EPA is a prescription decision | People who eat little oily fish, or with an omega-3 index under 8%. The index is the test. |
| Vitamin D3 | Mixed — VITAL (2,000 IU) and D-Health (60,000 IU monthly): no reduction in cancer, cardiovascular events or fractures in people who were mostly replete; meta-analyses show fewer falls and respiratory infections in the deficient | 1,000–4,000 IU daily, to a 25-OH level of 75–125 nmol/L | Anyone below 50 nmol/L — common in northern winters. Not a longevity drug in the replete. |
| Magnesium | Moderate — small trials in sleep and blood pressure; cohorts tie low intake to diabetes and arrhythmia | 200–400 mg elemental, as glycinate or citrate, evening | People with low dietary intake, cramps, poor sleep, or on diuretics or PPIs. Serum magnesium is a poor test; red-cell magnesium is better. |
| Protein powder | Strong, as a food | Whatever closes the gap to 1.6 g/kg | Anyone who cannot reach the protein floor from meals. It is food, not a supplement. |
| Taurine | Animal — Singh 2023, Science: taurine falls ~80% with age in mice, monkeys and people; supplementation extended median lifespan 10–12% in mice and improved health markers in monkeys | 1–3 g daily in the small human trials of blood pressure and exercise | Interesting, cheap and safe; no human outcome trial. An honest "early". |
| Fibre (psyllium) | Strong — lowers LDL ~7%, improves glycaemic control, meta-analyses | 10 g daily | Anyone under 30 g of dietary fibre a day, which is most adults. |
The honest middle
Compounds with a reasonable mechanism, some human data and no outcome trial: glycine (3 g before bed for sleep onset, small trials), CoQ10 (useful in statin myalgia and heart failure, not for ageing), NAD+ precursors (raise a blood marker, little else — see the mitochondria article), urolithin A (two small trials with modest muscle-endurance gains), berberine (lowers glucose and LDL in short trials; GI effects; interactions), and collagen peptides (modest, industry-funded skin and joint data). None of these is wrong to take. None of them is a lever, and a protocol that includes one says why.
What we do not recommend
- "Organ-support" and "cycle-support" stacks — TUDCA, NAC, milk thistle, "liver protect", "kidney support" — taken without labs, on the theory that a compound protocol needs a counterweight. If the liver enzymes and cystatin C are normal there is nothing to support; if they are abnormal, a supplement is not the answer and a physician is. A stack that exists to make a plan feel safer is a sign the plan was not safe.
- Mega-dose vitamins. High-dose vitamin E raised mortality in meta-analyses; high-dose beta-carotene raised lung cancer in smokers; vitamin A above 3,000 µg thins bone. More is not more.
- Multi-ingredient "longevity blends" at proprietary doses. If the label does not state the milligrams of each ingredient, nobody can check it against a trial.
- Resveratrol, quercetin, fisetin, spermidine as lifespan agents. Resveratrol failed in the ITP; the senolytic pairs are in early human trials in specific diseases; spermidine has cohort data and one small trial. Not harmful; not evidence-based for ageing.
- Ashwagandha and kava as daily habits. Both have accumulating liver-injury case reports.
- Iron without a ferritin result. One of the few supplements that reliably harms when it is not needed, and high ferritin is common in men over forty.
- Testosterone "boosters", "GH boosters", "NO boosters". Marketing categories, not pharmacology.
What to measure
The point of the labs is that most supplements are either unnecessary or necessary for a specific, testable reason. Before anything is added:
- Vitamin D (25-OH): supplement to 75–125 nmol/L, re-test after three months, then yearly.
- Ferritin with hsCRP: iron only if ferritin is low; investigate if it is high.
- Omega-3 index (a finger-prick test): under 8% is the case for EPA/DHA.
- Red-cell magnesium, if the symptoms suggest it.
- B12 and folate, especially on metformin, a PPI, or a plant-based diet.
- Cystatin C before creatine and before any protocol; liver enzymes once a year.
- Not a hair mineral analysis, a 100-item "micronutrient panel" or a food IgG test.
The baseline-labs guide lists the full house panel, which already contains most of this.
Where peptides and therapies fit
A supplement is a different category from a research compound, and the two are often confused. The items above are foods and micronutrients with safety data over decades; the compounds in the research library are studied in research settings and gated on labs, and the library's NAD+ page is the clearest example of something that is sold as a supplement, infused as a therapy and studied as a compound without the evidence catching up with any of the three. In a protocol, the supplement layer is written first and is short: creatine for anyone lifting, vitamin D if low, omega-3 if the index is low, magnesium if indicated, protein powder to the floor. The longevity goal page shows where that short list sits in the full plan.
Common mistakes
- Twelve capsules a day and no DEXA. The money is in the wrong place.
- Vitamin D at 10,000 IU without a level. Toxicity is real above 250 nmol/L.
- Fish oil that has oxidised — if it smells, it has.
- Creatine "cycles". It does not need them, and the lean-mass benefit needs continuity.
- Stopping a supplement that was correcting a deficiency because the level came back normal.
- A liver-support stack instead of a liver panel.
Frequently asked questions
Which supplements are actually worth taking for longevity?
Creatine for anyone who lifts, vitamin D if your level is under 50 nmol/L, omega-3 if your omega-3 index is under 8% or you eat little fish, magnesium if intake or symptoms suggest it, and protein powder to reach 1.6 g/kg. That is most of the list with outcome evidence.
Does taurine extend lifespan?
In mice, median lifespan rose 10–12% in a 2023 Science study, and health markers improved in monkeys. Taurine falls sharply with age in people too. There is no human outcome trial; it is cheap and safe at 1–3 g, and it is honestly "early".
Should I take vitamin D if my level is normal?
The large trials — VITAL with 25,871 adults, D-Health — found no reduction in cancer, cardiovascular events or fractures in people who were mostly replete. The benefit is in correcting a deficiency, so test first and supplement to 75–125 nmol/L.
Why do you not recommend liver- and kidney-support stacks?
Because if the liver enzymes and cystatin C are normal there is nothing to support, and if they are abnormal a supplement is the wrong response. A stack that exists to make a protocol feel safer is a sign the protocol was not safe; the labs are the safety.
Is creatine safe for the kidneys?
Yes, in healthy kidneys, across hundreds of trials. It raises serum creatinine slightly because creatinine is a creatine by-product, which is why kidney function is measured as cystatin C before and during any protocol.
Want a plan built on your own numbers?
The protocol lists the few supplements your labs justify, at the dose the trial used, and leaves out the rest. €399, follow-up included.
Build my protocol Book a 15-minute call