No affiliate links. No “clinically proven” weasel words. Just an A-to-F report card on the supplements you keep seeing on the longevity podcasts — scored by what real human studies actually show. Only one of the 22 earned an A.
I'm 60, and I've bought most of these at some point. So I did the unglamorous thing: I read the human trials — not the press releases — and graded every one against the same honest bar. The pattern that fell out surprised even me.
Here's the bar. A supplement earns a high grade only when human studies — ideally randomized trials — show a real effect on something that actually matters: strength, a hard health outcome, a corrected deficiency. “It raised a biomarker” or “it worked in mice” does not cut it, because the longevity aisle is full of molecules that move a number in a tube without doing a thing you'd ever feel. Run any of these through the BS Detector and the same question keeps deciding the grade: did it work in actual people, on something real?
Sorted best evidence to weakest. The story tells itself: the boring, decades-old basics beat the exciting new molecules almost every time.
| Supplement | Grade | The one-line verdict |
|---|---|---|
| Creatine | A | The one real winner — with resistance training. |
| Vitamin D | B | Real — but only if you're actually low. |
| Vitamin B12 | B | A fix-a-deficiency win, common after 50. |
| HMB | B | Modest muscle help, second-tier after protein + creatine. |
| Berberine | B | Lowers blood sugar for real; “Nature's Ozempic” is nonsense. |
| Collagen | B | Real for skin, oversold for joints. |
| Urolithin A | C | Best-studied new molecule; modest, missed its primary endpoint. |
| Omega-3 | C | The textbook “it depends on dose and you.” |
| CoQ10 | C | Strong in heart failure, weak everywhere else. |
| Magnesium | C | Helps mostly if you're deficient. Food fixes most of it. |
| Calcium | C | Food first; pills carry an unsettled heart debate. |
| Curcumin | C | Modest, and barely absorbed without the right formula. |
| Glucosamine | C | Weak for pain, intriguing mortality association. |
| GlyNAC | C | Promising, but tiny single-lab trials. |
| Ashwagandha | C | Modest calm — with a real liver-injury flag. |
| NMN | C | Raises NAD+. Hasn't been shown to slow human aging. |
| Nicotinamide Riboside | C | Reliably raises NAD+; healthy-aging payoff unproven. |
| Resveratrol | D | Great science story, weak human supplement. |
| Spermidine | D | The dedicated memory trial came back null. |
| Fisetin | D | Exciting in mice; human results aren't in yet. |
| Quercetin | D | Rides on a prescription drug's coattails. |
| Taurine | D | 2025 human data pushed back hard on the 2023 hype. |
Now the detail — grouped by what people are actually trying to fix, with the single strongest human study for each.
This is the category where supplements actually earn their keep after 50 — because muscle is the tissue most tied to staying independent. It's also where the only A on the page lives.
The most-proven supplement on this entire list, and it isn't close. A 2025 meta-analysis (Sharifian et al., European Review of Aging and Physical Activity, 2025) pooled 20 trials in roughly 1,093 older adults: creatine plus resistance training added around 2 kg to one-rep-max strength versus training alone, with decades of safety behind it. The honest catch — it works with lifting, not instead of it. It's a strength amplifier, not a youth pill.
A muscle-preservation aid with real but smaller evidence. A 2025 meta-analysis (Li et al., Frontiers in Nutrition 2025, 21 RCTs, ~1,935 adults over 50) found about +1.5 kg appendicular muscle mass (with a wide confidence interval) and a modest grip-strength gain of roughly half a kilogram. Worth considering once you've already maxed protein and creatine — not before. It's a second-tier add-on, not a foundation.
The good evidence is for skin, not joints. Multiple RCTs show hydrolyzed collagen modestly improves skin elasticity and hydration over 8–12 weeks; the joint and tendon data is thinner and mixed. Reality check: collagen is just a specific protein, broken down into amino acids like any other. “Rebuilds your joints” is marketing; “modestly better skin” is fair.
The most legitimately studied of the new “longevity molecules.” Two 2022 human RCTs (Singh et al., Cell Reports Medicine, in middle-aged adults; and a trial in adults 65–90) showed improved mitochondrial biomarkers and roughly 10% better muscle endurance and strength. The honest caveat: the headline trial actually missed its primary endpoint (peak power), and the gains were modest. Promising, not proven.
An intriguing combo with thin evidence. Small trials from one Baylor lab (Kumar et al., Journals of Gerontology, 2023) reported improvements in oxidative stress, strength, and cognition in older adults across several aging “hallmarks.” The catch is the one that haunts everything early: tiny samples, a single research group, not yet independently replicated. Worth watching, not betting on.
These don't make a healthy, well-fed person younger. They matter when you're genuinely short on something — which, after 50, is more common than you'd think. The rule for the whole group: test, then treat.
Real, but only if you're low. The giant VITAL trial (Manson et al., NEJM 2019, ~25,871 adults) found no fracture, cancer, or heart benefit from routine supplementation in people who already had enough. The benefit shows up when you correct a genuine deficiency — common after 50, especially through northern winters. Test first, and don't megadose: very high doses can actually increase falls.
A clean fix-a-deficiency win — not a everyone-should-take-it. After 50, and especially on metformin or acid-reducers, or eating little meat, absorption drops and deficiency is common, causing fatigue, nerve problems, and memory fog. Correcting a real deficiency clearly helps; taking it while already replete does nothing measurable.
Genuinely common to run low on, and the symptoms (poor sleep, cramps, higher blood pressure) are real — but the supplement mostly helps if you're actually deficient. RCTs for sleep and blood pressure show modest, inconsistent effects. Leafy greens, nuts, and legumes fix most of it. A cheap, low-risk try if you suspect you're short.
The advice has shifted toward food. Calcium supplements cut fractures mainly when paired with vitamin D in deficient or institutionalized older adults; for everyone else the benefit is small, and there's an unsettled debate about whether calcium pills (not food) nudge cardiovascular risk. Get it from diet first; supplement only for a real gap.
The textbook “it depends.” High-dose purified EPA cut cardiac events in one major trial (REDUCE-IT), while standard fish-oil doses did almost nothing in others (VITAL, STRENGTH). The contradictions come down to dose, formula, and who you are — not a coin flip. Eat fatty fish; treat a supplement as a heart-risk-dependent personal call, and know high doses can raise atrial-fibrillation risk. Full omega-3 breakdown here.
It really does lower blood sugar — and “Nature's Ozempic” is still nonsense. Meta-analyses (e.g., a 2024 review of ~50 studies, ~4,150 people) show modest, real reductions in fasting glucose and cholesterol. But the effect is a fraction of a GLP-1 drug's, works by a completely different mechanism, and brings GI side effects and drug interactions. A legitimate metabolic nudge, badly oversold.
Strong in one specific place: diagnosed heart failure. The Q-SYMBIO trial (Mortensen et al., JACC: Heart Failure 2014, n=420) found CoQ10 roughly halved cardiovascular deaths in moderate-to-severe heart failure. But for healthy people, or for statin muscle pain, the evidence is weak — the cleanest statin-myalgia RCT came back negative. A drug-grade effect in a narrow group; hype outside it.
Turmeric's active compound has modest anti-inflammatory and osteoarthritis-pain evidence — with a serious absorption problem. On its own, curcumin is barely bioavailable; the trials that show benefit use special formulations or add black-pepper piperine. Real but small effects, and only in a form your body can actually absorb.
A strange split decision. For osteoarthritis pain, the big GAIT trial was largely null. But a huge UK Biobank cohort (Li et al., Annals of the Rheumatic Diseases 2020, ~495,000 adults) found regular glucosamine users had about 15% lower all-cause and 18% lower cardiovascular mortality. That's an association, not proof — almost certainly tangled with healthier-user habits — but it's why this cheap, safe supplement keeps getting a second look.
This is where the podcasts spend their time and the prices climb. It's also where the grades fall. Almost everything here is built on animals, test tubes, or a moved biomarker — not on people getting measurably better.
Raises NAD+; hasn't been shown to slow aging in humans. Small human trials report higher NAD+ and some metabolic and performance biomarker shifts — but none shows it extends healthspan or lifespan in people. Regulatory whiplash worth knowing: the FDA excluded NMN from supplements in 2022, then reversed course in late 2025 and reinstated it as a lawful dietary ingredient. Moving a biomarker is not the same as living longer. Read the full NMN evidence file.
The most-studied NAD booster in humans — which is exactly why its limits are clear. It reliably raises NAD+, and a few trials show condition-specific signals (e.g., better walking distance in peripheral artery disease in the 2024 NICE trial). But across healthy-aging endpoints the results are mixed-to-flat. It does the one thing it promises without proof that more NAD+ makes a healthy person younger.
The molecule that launched the longevity-supplement craze — and mostly fizzled in humans. The red-wine/SIRT1 story is real in yeast and mice, but resveratrol is poorly absorbed in people, and human trials have been disappointing and inconsistent, with no replicated longevity benefit (2024 reviews are blunt about it). A great science story; a weak supplement.
Hyped for memory and “autophagy”; the dedicated trial didn't deliver. The 12-month SmartAge RCT (Schwarz et al., JAMA Network Open 2022, n=100) found no improvement in memory or biomarkers versus placebo, with only faint subgroup hints. It's abundant in foods (wheat germ, aged cheese, natto) — eating those is fine; the pill hasn't earned its claims.
A genuinely exciting senolytic — in mice. The animal data on clearing “senescent” cells is striking, but the human results aren't in: the main human trial (Mayo Clinic's AFFIRM, in older women) is still running, with results not expected until roughly 2026–2027. Promising enough to study hard; far too early to take on the strength of mouse studies.
Mostly rides on other things' coattails. The headline “senolytic” results come from quercetin combined with the prescription chemotherapy drug dasatinib, in very small trials — not quercetin alone. On its own, as an anti-aging supplement, the human evidence is thin. Cheap and safe, but unproven for the claims stacked on it.
A cautionary tale about hype cycles. A 2023 Science paper showed taurine extended life in mice and tracked with age in monkeys — and sales exploded. Then 2025 human data pushed back hard: a study in Aging Cell (Marcangeli et al., ~137 people) found no consistent evidence taurine declines with age or drives it in humans, and Nature ran the headline “anti-ageing effects of taurine challenged.” The human case got weaker, not stronger.
Modest evidence for stress, anxiety, and sleep in small short trials — but a safety signal that isn't on the label. Ashwagandha is now listed in the NIH's LiverTox database as a probable cause of clinically apparent liver injury, with a growing set of case reports between 2023 and 2025, including at least one that ended in a liver transplant. If you try it, keep it short-term and stop at any sign of liver trouble. This is not the harmless “calm” herb it's marketed as.
The uncomfortable pattern. Run your eye down the scorecard. The single A is a supplement from the 1990s. The B's are either fixing a deficiency or amplifying exercise. And the priciest, most-hyped molecules — the ones with the slick longevity branding — cluster at the bottom on animal data and moved biomarkers.
That's not an accident. It's what happens when you ask “did it work in actual people, on something real?” instead of “is there a mechanism that sounds plausible?”
Our anti-aging guide collects every intervention that survives the evidence, with the human studies behind each one and zero supplement hype. The short list that actually moves the needle after 50.
See the Full GuideCreatine monohydrate — the only A on this list. A 2025 meta-analysis of about 20 trials in roughly 1,093 older adults found creatine plus resistance training added about 2 kg to one-rep-max strength versus training alone, with decades of safety data. The catch is that it works alongside strength training, not on its own.
Both are interesting mechanistically and unproven for human aging. NMN reliably raises NAD+ but no human study shows it extends healthspan or lifespan; the FDA reinstated it as a lawful dietary ingredient in late 2025 after earlier excluding it. Resveratrol is poorly absorbed and its human trials have been disappointing, with no replicated longevity benefit.
Mostly, with one real caveat: it's now listed in the NIH LiverTox database as a probable cause of clinically apparent liver injury, with a growing number of case reports between 2023 and 2025 including at least one liver transplant. If you use it, keep it short-term and stop at any sign of liver trouble.
Most do less than the marketing implies. The clearest wins are correcting a genuine deficiency (vitamin D, B12) and creatine paired with strength training. Many popular supplements only help if you're already deficient, and several trendy longevity molecules move biomarkers without proving any real-world benefit. Food and exercise still beat the aisle.
— Scott Covert, 60, skeptic, not a physician. I built this for my own medicine cabinet first, after getting tired of paying for hope. Think I graded one wrong, or have a study I missed? Tell me and I'll dig in.