The Ultimate Longevity Stack 2026: Evidence-Tiered Recommendations

Ask ten people for their longevity stack and you get ten shopping lists — twenty compounds deep, no hierarchy, mouse data and phase-3 trials sitting side by side as though they carry the same weight. The list format is the problem. It implies everything on it earned its place the same way.

This one is sorted by how much human evidence actually exists. The result is uncomfortable if you enjoy long stacks: only two compounds survive to Tier 1. The rest are bets of varying honesty — which is fine, as long as you know that's what you're holding.

The Stack at a Glance

* Grade A for the metabolic outcomes it was tested on — not for longevity.

Tier 1 — The Two That Earn It

A Tier 1 compound has to clear a bar the rest don't: large randomised trials, clinical endpoints, and a safety record measured in decades. Two compounds clear it.

Creatine (Grade A)

Creatine is the most studied supplement in existence, and the longevity case is stronger than the gym reputation suggests. The ISSN position stand — a review of hundreds of trials — found no evidence of renal harm in healthy people and confirmed efficacy for muscle and performance.

^[1]

That matters here because muscle is the organ of ageing: sarcopenia predicts frailty, falls, and mortality better than most blood markers. But creatine also feeds cerebral energy metabolism, and a systematic review of RCTs found measurable cognitive benefits in healthy adults.

^[2]

• Dose: 5 g/day of plain monohydrate. No loading, no cycling, no exotic form is better.
• Why it's Tier 1: two organ systems, hundreds of trials, an exceptional safety profile, and roughly the cost of a coffee per month. Full detail on the creatine profile.

Omega-3, EPA + DHA (Grade A)

Omega-3 is where you must read the trials rather than the headline. VITAL — over 25,000 participants — missed its primary endpoint: 1 g/day of fish oil did not significantly reduce major cardiovascular events or cancer overall.

^[3]

Then DO-HEALTH tested 1 g/day in older adults across five years and found benefits on several clinical outcomes, including a lower infection rate.

^[4]

Read together, the honest conclusion is: omega-3 is not a cardiac drug, and it is a well-evidenced structural nutrient with a real, modest effect footprint. Most people are genuinely under-supplied in EPA and DHA, and unlike almost everything else here, the deficiency case is measurable.

• Dose: 2–3 g/day combined EPA+DHA — meaningfully above the 1 g used in the trials that underwhelmed. Take with a fatty meal.
• Quality caveat: oxidised fish oil is common and worthless. Buy on third-party peroxide-value testing. See the omega-3 profile.

What about vitamin D? It belongs in this conversation and doesn't yet have a profile on this site — we're building one. The short version: correct a measured deficiency, don't megadose on faith. DO-HEALTH tested 2000 IU/day and found no benefit on its primary outcomes in already-replete older adults.

^[4]

That is the whole lesson of Tier 1 in one line: supplementation fixes deficits; it doesn't add superpowers to people who aren't deficient.

Tier 2 — Reasonable Bets

Tier 2 compounds have real human evidence that is incomplete — a good trial in one population, or a strong mechanism with the outcome trial still pending. Take them knowing you're paying to be early.

Urolithin A (Grade B)

The best-evidenced compound in Tier 2, and the one with the most interesting mechanism: it restores mitophagy, the recycling of damaged mitochondria that declines with age. Most longevity compounds only claim to do something like this. Urolithin A was measured doing it in human muscle.

Two randomised trials: one in older adults improved muscle endurance and mitochondrial biomarkers,

^[5]

and one in middle-aged adults improved muscle strength and exercise performance.

^[6]

• Dose: 500 mg/day, morning, food optional.
• Why not Tier 1: the trials are modest in size, the endpoints are muscle-functional rather than clinical, and both were conducted with industry involvement. Real, but not settled. See the urolithin A profile.

Taurine (Grade B)

Taurine got its 2023 moment when a Science paper showed taurine declines with age across species and that restoring it extended median lifespan in mice by roughly 10%, with healthspan improvements in monkeys.

^[7]

That is a genuinely impressive animal result. It is also, as of today, an animal result. The human outcome trial hasn't reported. What taurine has going for it is an unusually forgiving risk profile: it's a conditionally essential amino acid, cheap, and well tolerated at these doses.

• Dose: 3 g/day. Evening suits most people.
• Honest grade: B on safety and mechanism, not on human longevity outcomes. Details on the taurine profile.

Spermidine (Grade B) — and an inconvenient trial

Spermidine is the autophagy story: dietary intake tracks with lower cardiovascular mortality in cohort studies, and it extends lifespan across model organisms.

^[9]

Then it was actually tested. SmartAge — a 12-month RCT in older adults with subjective cognitive decline — found no significant cognitive benefit over placebo.

^[8]

We keep spermidine in Tier 2 rather than dropping it, because the trial tested cognition in a specific population, not the cardiovascular and autophagy endpoints where the epidemiology is strongest. But that null result is exactly the sort of thing a shopping-list stack quietly omits. It belongs here.

• Dose: 5 mg/day with breakfast. See the spermidine profile.

Berberine (Grade A — with an asterisk)

Berberine has meta-analysis-grade evidence for improving the components of metabolic syndrome: glucose, lipids, blood pressure.

^[10]

Read that precisely. It's grade A for treating metabolic dysfunction, and grade nothing for longevity in metabolically healthy people. If your fasting glucose, HbA1c, and lipids are already good, berberine has no demonstrated job to do — and you'd be taking a compound that meaningfully affects the gut microbiome for no defined endpoint.

• Dose: 1 g/day split across your two largest meals, cycled 8–12 weeks on / 4 weeks off. Only with a metabolic indication. See the berberine profile.

Tier 3 — Honest Experimentation

Tier 3 isn't a warning label. It's a category: compounds where the case rests on one trial, one population, or animals only. Take them as experiments, not as foundations.

NMN and the NAD+ precursors (Grade C)

The NAD+ decline story is real, and NAD+ precursors do raise NAD+ — Martens showed nicotinamide riboside sustainably elevates it in middle-aged and older adults.

^[12]

The gap is between raising the marker and changing an outcome. The strongest NMN result to date: 250 mg/day improved muscle insulin sensitivity — in prediabetic postmenopausal women.

^[11]

One trial, one population, one surrogate endpoint. That's a legitimate grade C, and it's a long way from the "cellular fountain of youth" the category is sold as. If you take it, take it knowing which of those two sentences the evidence supports. Profiles: NMN and nicotinamide riboside.

Fisetin (Grade C)

Fisetin is the senolytic with the best animal data — it cleared senescent cells and extended lifespan in mice even when started late in life.

^[14]

Human senolytic trials are ongoing and small. Its saving grace is the pulse protocol: 500–1500 mg for two consecutive days per month rather than daily exposure, which limits the cost of being wrong. See the fisetin profile and the senolytic protocol.

What We Deliberately Left Out

A stack is defined as much by its exclusions as its contents.

Resveratrol — the cautionary tale

Resveratrol launched the modern longevity supplement industry on the back of spectacular mouse data. Then it was tested properly in humans: a well-run trial in nonobese women found it did not improve insulin sensitivity, glucose tolerance, or any other metabolic measure.

^[13]

We keep a resveratrol profile because people ask, and we keep it out of this stack because mechanism is not evidence. If you remember one thing from this page, make it that sentence — it's the filter that would have saved the category a decade.

Prescription drugs are not stack components

Rapamycin has the most impressive longevity data of any known molecule: it extended lifespan in mice even when started in old age,

^[16]

and low-dose mTOR inhibition improved vaccine response in older adults.

^[15]

The same goes for GLP-1 agonists, which are genuinely transforming metabolic medicine. Both are prescription medicines with real immunosuppressive and metabolic consequences, requiring diagnosis, monitoring, and a clinician who knows your history. They are a medical decision that happens outside this stack — not a tier you graduate into by reading an article. We cover them (rapamycin, the GLP-1 comparison) because informed patients have better conversations with their doctors.

Everything else

The compounds not mentioned aren't condemned — they're targeted, not foundational. Ashwagandha for stress, lutein for the eyes, collagen for skin: they belong to goals, not to a general longevity base. Those live in their own protocols, like stress resilience and vision longevity.

How to Actually Build It

Do not start eight compounds on a Monday. You'll learn nothing and attribute everything wrongly.

1. Months 1–3: Tier 1 only. Creatine and omega-3. If you never go further, you have captured most of the available benefit for about 5% of the cost of a maximal stack.
2. Month 4: add one Tier 2 compound — urolithin A if your interest is muscle and mitochondria, berberine only if your bloodwork shows a metabolic reason.
3. One new compound at a time, with 8+ weeks between additions. Anything faster and your stack is a guess wearing a lab coat.
4. Tier 3 last, and only if you accept the terms — you are the trial.
5. Re-test bloodwork before and after. A stack you never measure is a subscription, not an intervention.

The assembled version — timing, doses, and stacking notes — is on the Ultimate Longevity Stack protocol page. The reasoning behind the tiering is laid out in The Physician's Longevity Supplement Framework.

Buy on Testing, Not on Branding

One structural warning: the supplement market's quality floor is low. An analysis of products sold online found label inaccuracy was the norm rather than the exception in the category tested.

^[17]

Compounds like urolithin A and spermidine have almost no consumer-level way to verify identity by eye. Buy on third-party certificates of analysis, or accept that you may be running your careful, evidence-tiered protocol on rice flour.

Foundations First — the part that outranks this whole page

If the stack above is competing for attention with your sleep, the stack loses. Nothing on this page comes close to:

• Resistance training — the only intervention that reliably rebuilds the organ of ageing.
• Sleep — 7–9 hours; sleep deprivation degrades glucose control and cognition faster than any supplement improves them.
• Protein — roughly 1.6 g/kg/day; creatine has nothing to work with otherwise.
• Blood pressure and metabolic health — the two most treatable drivers of how you actually age.
• Not smoking, and alcohol at genuinely low levels.

Get those right and the stack adds an honest margin on top. Get them wrong and the most sophisticated tier list in the world is an expensive way to feel proactive. What extreme optimisers get right — and where they overspend — is the subject of our data-driven blueprint analysis.

This article is educational and not medical advice. Supplements interact with medications and medical conditions. Talk to a healthcare provider before starting any protocol, especially if you are pregnant, managing a chronic condition, or taking prescription drugs.