The Science Behind Longevity & Cellular Health Supplements

Therapeutic dose: 4–12 mg/day (natural Haematococcus-derived, taken with a fat-containing meal)

• In a small 8-week RCT in young healthy women (Park 2010, n=42), oral astaxanthin lowered the DNA oxidative-damage marker 8-OHdG by week 4 (maximal at the lower 2 mg/day dose) and reduced C-reactive protein by week 8, while raising several immune-cell measures — but these are biomarker and ex-vivo endpoints, not clinical or aging outcomes.PubMed ↗
• The antioxidant mechanism is genuinely strong and well-characterized: astaxanthin's extended polyene chain and terminal hydroxyl rings make it an unusually efficient singlet-oxygen and lipid-radical quencher, and it does not flip to a pro-oxidant at high concentration the way some carotenoids can — a real structural advantage, but mechanism is not evidence of slowed cellular aging.
• No human trials test astaxanthin against senescence, NAD+, mitochondrial function, or longevity/lifespan endpoints; the cellular case rests entirely on reduced oxidative-stress and inflammation biomarkers in small, short, mostly manufacturer-funded trials (AstaReal/Fuji, BGG, Cyanotech). Improved biomarkers are not the same as a slowed-aging outcome — this biomarker-to-outcome gap is the core honesty caveat for astaxanthin.

Therapeutic dose: 250–500 mg/day (clinical-trial range 100–1,200 mg/day)

See ranked NMN (Nicotinamide Mononucleotide) products→

• Multiple short-term RCTs (Irie 2020, Yoshino 2021, Igarashi 2022, Pencina 2023, Yi 2023) consistently raise whole-blood NAD+ and NAD+ metabolites in a dose-dependent way — the single most reproducible human NMN finding. This is a target-engagement biomarker (proof the molecule reaches its pathway), not evidence of a downstream cellular-aging benefit.PubMed ↗
• Pencina 2023 RCT of pharmaceutical-grade microcrystalline NMN (MIB-626) in overweight/obese adults aged 45+ substantially raised circulating NAD+ and its metabolites and was well-tolerated. It is a physiologic/biomarker trial with no hard cellular, senescence, or longevity outcome, and it is sponsor-funded by Metro Biotech. Secondary metabolic signals (body weight, LDL, diastolic BP) appeared but were not the point — the reproducible result is NAD+ elevation, not clinical benefit.PubMed ↗
• No human RCT has shown NMN extends lifespan, healthspan, or any senescence/mortality endpoint. The 'reverses aging' framing rests on rodent data — most prominently Mills 2016, which improved several mouse healthspan markers but did NOT produce statistically significant lifespan extension in the C57BL/6N strain despite how it is popularly cited.
• The nearest human 'aging' signal — a blood-based 'biological age' estimate that shifted on NMN in the Yi 2023 dose-ranging RCT (n=80) — is exploratory and not a validated longevity surrogate. Nearly all positive human trials are small (n=10–108), 8–24 weeks, industry-funded, and often share a single NMN supplier, so the effective independent evidence base is thinner than the trial count suggests.

Therapeutic dose: 500–1,000 mg/day curcuminoids, but ONLY in a bioavailable form — Theracurmin 90–180 mg/day, Meriva (phytosome) 200–400 mg/day, or BCM-95/Curcugreen 500–1,000 mg/day. Standard 95% extract is ~1% absorbed and functionally undosed unless paired with piperine 5–20 mg or a fat carrier.

• The strongest human signal relevant to cellular aging is inflammatory, not longevity: a 2022 meta-analysis of 32 RCTs (Gorabi et al., Phytotherapy Research) found curcumin lowered C-reactive protein by a weighted mean of ~3.67 mg/L versus placebo, with the effect strongest at doses at or below 1,000 mg/day. This maps onto 'inflammaging' and oxidative stress, but CRP is a biomarker, not a cellular-aging clinical endpoint.PubMed ↗
• Mechanistically well-characterized for the pathways this category cares about — curcumin downregulates NF-κB, TNF-α and IL-6 and activates the Nrf2 antioxidant response, and preclinical work suggests senomorphic (senescence-dampening) activity. However, these senescence, mitochondrial and NAD+ effects are in-vitro and animal-model only; there are no human RCTs measuring senescent-cell burden, mitochondrial function or NAD+ in this literature.
• No curcumin RCT to date reports a longevity, cellular-senescence or mitochondrial-function endpoint in humans; the existing positive human trials target knee osteoarthritis pain, depression as an adjunct, and ulcerative colitis maintenance — adjacent inflammatory conditions, not cellular aging. Bioavailability is the gating problem: without a phytosome carrier (Meriva) or submicron/lipidated delivery (Theracurmin, Longvida) or piperine, plasma curcumin stays near-zero regardless of the label dose.

Therapeutic dose: 300–1,000 mg/day (trials 100–2,000 mg/day; 1,000 mg/day raises blood NAD+ ~60%)

• The one reproducible cellular effect: NR raises blood NAD+ dose-dependently. Martens 2018 (RCT, 30 enrolled / 24 analyzed, healthy adults 55–79, 1,000 mg/day, 6 weeks) raised whole-blood NAD+ ~60%; the first-in-human PK trial (Trammell 2016) and a heart-failure safety RCT (Wang 2022, ~doubled NAD+) agree. This is target engagement — a biomarker moving — not a demonstrated health outcome.PubMed ↗
• But the mitochondrial payoff did not appear. Elhassan 2019 (RCT crossover, 12 aged men ~75, 1,000 mg/day, 21 days) raised the muscle NAD+ metabolome and produced an anti-inflammatory signature, yet did NOT improve mitochondrial bioenergetics — muscle energy-metabolism gene pathways were actually downregulated. Deeper metabolic-phenotyping RCTs (Remie 2020, clamps + biopsies; Dollerup 2018, 2,000 mg/day) confirmed no mitochondrial-function improvement. Molecular signals without functional confirmation.PubMed ↗
• No human RCT has shown NR extends lifespan or healthspan or improves any hard longevity outcome. The animal case is weaker than the marketing implies: NR failed to extend lifespan even in well-controlled mouse studies despite raising NAD+. Buying NR for 'anti-aging' is betting on a mechanism (NAD+ declines with age in animals), not on outcome evidence in people.
• Preclinical safety caution, not a benefit: a mouse study (Maric et al., 2023) tied NR-enriched diet to increased triple-negative breast-cancer growth and brain metastasis, consistent with NAD+ fueling proliferation and PARP-mediated DNA repair. Animal data only — not proof of human harm — but a biologically plausible reason anyone with active or recent cancer should avoid NR without oncology clearance.PubMed ↗

Therapeutic dose: 1.5–3 g/day (oxidative-stress trials used 1.5–3 g/day)

• The most direct human cellular evidence is a small RCT in adults with type 2 diabetes (n=50, 3 g/day for 8 weeks): taurine lowered malondialdehyde (a lipid-oxidation marker) and raised SOD and GPx antioxidant enzyme activity. These are surrogate oxidative-stress biomarkers in an oxidative-stress-elevated population, not longevity or senescence endpoints, and no hard clinical outcome was measured.
• The viral 'longevity nutrient' framing comes from Singh et al. 2023 (Science), which showed ~10-23% median lifespan extension in worms and ~10-12% in middle-aged female mice plus a healthspan signal in monkeys, driven mechanistically by mitochondrial tRNA stabilization. The human arm was only a cross-sectional EPIC-Norfolk (n≈11,966) plasma-taurine correlation. No human interventional trial has shown taurine affects lifespan, senescence, or NAD+.PubMed ↗
• A small postmenopausal pilot RCT (n=24, 1.5 g/day for 16 weeks) reported reduced oxidative stress and modest lipid improvements. At n=24 this is hypothesis-generating only, and it is frequently overstated as 'anti-aging' evidence.
• Honest limits: there is no human RCT on lifespan, cellular senescence, or NAD+ for taurine, and a 2025 follow-up (Fang et al.) reported plasma taurine does not consistently decline with age in larger human cohorts, weakening the central age-decline premise. Human cellular data is confined to surrogate antioxidant biomarkers in disease populations.

Therapeutic dose: 500–1,000 mg twice daily (1–2 g/day) as used in clinical trials

• The headline anti-aging evidence is in-vitro, not human: McFarland & Holliday (1994) reported that carnosine (studied at ~20–50 mM in culture) retarded replicative senescence and reversibly rejuvenated the morphology of cultured human diploid fibroblasts — but it did not prevent the Hayflick limit, the youthful appearance reverted when carnosine was withdrawn, and these millimolar culture concentrations are far above anything oral dosing achieves in tissue.PubMed ↗
• A follow-up cell-culture study (Shao et al., 2004) found normal fibroblasts grown continuously in ~20 mM carnosine showed a slower telomere-shortening rate and less telomeric DNA damage across population doublings. This supports a plausible mechanism for cellular aging, but it remains an in-vitro finding at supraphysiological concentrations with no human replication.
• Human RCT evidence is limited to oxidative-stress surrogate markers in metabolic disease, not aging or senescence: a double-blind placebo-controlled trial in type 2 diabetics (500 mg twice daily, 12 weeks) reported reductions in malondialdehyde and protein carbonyls versus placebo. These are lipid-peroxidation/protein-oxidation markers in a diseased population, not measures of longevity or cellular senescence.
• A 2024 systematic review and meta-analysis (9 trials, ~350 participants) of carnosine and histidine-containing dipeptides on inflammation and oxidative-stress biomarkers found significant reductions in CRP, TNF-α and MDA but no effect on IL-6, adiponectin, GSH, SOD or total antioxidant capacity — modest and inconsistent effects across markers. No human study has tested carnosine against cellular-aging endpoints such as senescence burden, NAD+ status, or lifespan. For the longevity outcome specifically, the honest answer is: no human trials exist.

Therapeutic dose: 100–300 mg/day with a fat-containing meal (absorption roughly triples vs fasted). Doses derived from clinical trials: Q-SYMBIO used 300 mg/day ubiquinone in heart failure; migraine prophylaxis RCTs used 300 mg/day. Ubiquinol form preferred for adults over 50 and statin users (raises plasma CoQ10 ~2–8× more than matched-dose ubiquinone in pharmacokinetic work). Not a %DV-based figure.

• Mechanistically load-bearing but category-limited: CoQ10 is the obligate electron carrier between Complexes I/II and III in the mitochondrial electron transport chain, and endogenous synthesis declines with age (myocardial CoQ10 in 80-year-olds is roughly half that of 20-year-olds). The mitochondrial-function rationale is unimpeachable, but it remains a mechanism, not a demonstrated longevity or senescence outcome.
• Human target engagement is confirmed, not clinical benefit: pharmacokinetic work in healthy adults (Hosoe 2007, ~36 subjects, 4 weeks) showed the reduced ubiquinol form raises plasma total CoQ10 and the antioxidant-active ubiquinol fraction several-fold more than matched-dose ubiquinone, with the gap widening with age. Solid PK but industry-funded (Kaneka, the ubiquinol patent holder), and elevated plasma CoQ10 has not been translated into oxidative-stress or cellular-aging clinical endpoints in non-deficient adults.PubMed ↗
• Honest null on the strongest mitochondrial hypothesis in the nearest hard-endpoint domain: the large NIH-funded QE3 trial (n=600, doses up to 2,400 mg/day CoQ10) in early Parkinson's was terminated for futility, showing no slowing of neurodegenerative progression despite a clean mitochondrial-dysfunction rationale. A direct caution that CoQ10's cellular mechanism has not reliably converted into hard disease-modifying or aging outcomes.PubMed ↗

Therapeutic dose: Clinical trials used roughly 600 mg/day (COPD mucolytic) up to 2,000–3,000 mg/day (psychiatric augmentation) — for unrelated endpoints. No validated dose exists for any cellular-aging or longevity outcome.

See ranked NAC (N-Acetylcysteine) products→

• Mechanism only: NAC supplies cysteine, the rate-limiting substrate for glutathione synthesis — the cell's primary endogenous antioxidant and the plausible link to oxidative-stress theories of aging. No RCT has tested NAC against cellular-aging, cellular senescence, mitochondrial function, or NAD+ endpoints in humans.
• The 'glutathione booster' claim is weaker than marketed: in healthy people, measured glutathione increases from oral NAC are modest in the few studies that quantified them. NAC demonstrably restores glutathione mainly in depleted/pathological states — e.g. the IV acetaminophen-overdose antidote protocol (Smilkstein 1991, n=179) — which is emergency medicine, not a model for longevity supplementation.PubMed ↗
• No human longevity, lifespan, or anti-senescence trial exists for NAC. Its actual RCT evidence base sits in unrelated categories (COPD mucolytic, and small adjunctive psychiatric trials in bipolar depression, schizophrenia and trichotillomania — with a null result in OCD augmentation), none of which measure cellular-aging outcomes — so a cellular-health rating rests entirely on mechanistic plausibility, not replicated clinical effect.

Therapeutic dose: 600 mg/day racemic (~300 mg/day R-ALA equivalent) — trial-anchored dose from diabetic-neuropathy studies, not validated for any longevity or cellular-aging endpoint

See ranked Alpha-Lipoic Acid (R-ALA) products→

• No human RCTs test alpha-lipoic acid against any cellular-aging, senescence, NAD+, or longevity endpoint. The nutrient's own verdict is explicit: anti-aging and general antioxidant claims 'have no good RCT support.' All eight of its logged trials target diabetic neuropathy, glycemic control, or body weight — not cellular healthspan.
• The category rationale is mechanistic only: ALA acts as a mitochondrial cofactor for the pyruvate dehydrogenase complex, scavenges reactive oxygen species, regenerates other antioxidants (glutathione, vitamins C and E), and chelates redox-active metals. This is a plausible oxidative-stress mechanism, but mechanistic and in-vitro signals do not support clinical longevity claims in humans.
• The closest human data sit one category away: a 2012 meta-analysis pooled 15 RCTs of ALA for diabetic peripheral neuropathy (an oxidative-stress-driven condition) and reported improved nerve conduction velocity and symptom scores — but this is a disease-specific nerve outcome in diabetics, driven largely by short IV-dosed trials, not a cellular-aging or mitochondrial endpoint in healthy older adults. Do not extrapolate.PubMed ↗

Therapeutic dose: 500 mg three times daily (1.5 g/day total), taken with meals; standard berberine HCl is ~5% bioavailable, so the split dose is load-bearing. Note: this is the metabolic (glucose/lipid) trial dose — no cellular-aging trial has established a longevity dose.

• No human RCT in our database measures a cellular-aging endpoint for berberine — mitochondrial function, oxidative-stress markers, NAD+, or senescence were not assessed in any of the 8 trials on file. All human evidence is metabolic (glucose and lipids), so any longevity claim is currently an extrapolation from mechanism, not a measured outcome.
• The longevity rationale rests on one shared mechanism: berberine activates AMPK, the same cellular energy sensor metformin acts on, which is the basis of geroscience interest. The Lan 2015 meta-analysis (27 RCTs, n=2,569) and the Yin 2008 head-to-head against metformin confirm the AMPK-driven metabolic effect in humans — but they measured it via HbA1c and glucose, not aging biomarkers.PubMed ↗
• The most reproducible human signal is lipid-lowering: the Dong 2013 meta-analysis of randomized trials found consistent reductions in total cholesterol, LDL-C, and triglycerides, attributed to ERK-dependent LDL-receptor upregulation. This is cardiometabolic-risk reduction, adjacent to healthy aging, but it is not a direct cellular-aging or lifespan endpoint.PubMed ↗
• Honest limits: the viral 'natural Ozempic' framing has no head-to-head trial behind it, and no berberine study has tested senescence, mitochondrial biogenesis, or lifespan in humans. Treat berberine as a metabolic agent with a plausible AMPK longevity mechanism — not as a validated anti-aging compound.

Therapeutic dose: 500–1000 mg/day (clinical trial range; plain quercetin is poorly absorbed, which is why effective trial doses are high). No established RDA or upper limit.

• The 'quercetin as anti-aging senolytic' story traces to a single 9-patient open-label first-in-human pilot in diabetic kidney disease (Hickson 2019, EBioMedicine), where a 3-day course reduced senescent-cell markers in fat and skin tissue. The senolytic effect was driven by pairing quercetin with dasatinib, a prescription tyrosine-kinase inhibitor — it says almost nothing about quercetin taken alone.PubMed ↗
• For oxidative/inflammatory endpoints, a meta-analysis of 7 RCTs found a small but significant CRP reduction (WMD −0.33 mg/L), strongest at ≥500 mg/day and in people with baseline CRP under 3 mg/L. This is a modest anti-inflammatory biomarker signal, not a demonstrated effect on cellular aging or hard longevity outcomes.PubMed ↗
• The proposed mitochondrial-biogenesis mechanism (PGC-1α), seen in rodents, does not appear to translate at human oral doses: in Nieman's ultramarathon RCT (2007, ~39 completers at 1000 mg/day), quercetin raised plasma levels but failed to blunt exercise-induced muscle damage, inflammation, or cytokine and hormonal changes.PubMed ↗
• There are no human RCTs testing quercetin monotherapy on NAD+, senescent-cell clearance, or longevity endpoints. Much of the cellular-aging enthusiasm comes from cell-culture concentrations that poorly-absorbed oral quercetin cannot reach in human plasma.

Therapeutic dose: 1–2 g/day combined EPA+DHA (status-repletion range used to reach an Omega-3 Index >8%); no cellular-longevity-specific dose is established

• No omega-3 trial in the underlying EPA or DHA evidence base measured a cellular-aging endpoint. The RCTs cover cardiovascular events (REDUCE-IT, STRENGTH, VITAL, OMEMI), major depression, pregnancy and infant outcomes, cognition and Alzheimer's, macular degeneration and dry eye — none tested mitochondrial function, NAD+, oxidative-stress or cellular-senescence outcomes. For this category the honest read is: mechanism only, no category-specific human trials.
• Mechanistic rationale (EPA): EPA displaces arachidonic acid in membrane phospholipids and shifts eicosanoid production away from inflammatory PGE2/LTB4 toward less-inflammatory PGE3 and pro-resolving E-series resolvins. Because chronic low-grade inflammation ('inflammaging') is one proposed driver of cellular senescence, this is a biologically plausible longevity pathway — but it is inferred from anti-inflammatory data, not demonstrated on any senescence or lifespan endpoint in humans.
• Mechanistic rationale (DHA): DHA is the dominant structural fatty acid in neuronal and photoreceptor membranes and shapes membrane fluidity, lipid-raft composition and GPCR signaling — the cleanest biophysical mechanism in the omega-3 family. However, the DHA dossier itself notes that translating membrane biophysics into measurable clinical change in already-adequate adults has repeatedly been null: repletion helps depleted tissue, while supraphysiologic dosing in adequate adults rarely moves outcomes.
• Even where cellular-relevant human signals might exist, they would be biomarker-level at best, not longevity or senescence outcomes — and none appear in the EPA or DHA trial dossiers underlying this category, which measured cardiovascular, mood, pregnancy, cognition and eye endpoints only. Treat any 'omega-3 for cellular aging' positioning as mechanism-stage, not trial-supported.

Therapeutic dose: 150–1,000 mg/day trans-resveratrol (Japanese knotweed-sourced, standardized for trans- isomer); note under ~1% reaches plasma intact, and gram-level doses cause GI distress

• The most-cited human 'calorie-restriction mimetic' study (Timmers 2011, Cell Metabolism) gave 11 obese men 150 mg/day trans-resveratrol for 30 days and reported muscle AMPK/SIRT1 activation, lower sleeping metabolic rate, and improved mitochondrial markers — but with only 11 participants, industry-supplied material, and a 30-day crossover, it is a proof-of-concept, not evidence of durable cellular benefit.PubMed ↗
• At the population level the longevity claim does not hold: in ~783 older Italian adults followed 9 years (Semba 2014, InCHIANTI, JAMA Internal Medicine), urinary resveratrol metabolites were not associated with all-cause mortality, cardiovascular disease, cancer, or inflammation — directly undercutting the anti-aging narrative, though this is observational dietary intake, not supplementation.PubMed ↗
• The famous SIRT1/'sirtuin' lifespan-extension findings come entirely from yeast, worms, flies, and obese mice; no human trial has shown resveratrol extends lifespan or healthspan, and its poor oral bioavailability (under ~1% reaches the bloodstream intact) makes it uncertain whether the in-vitro concentrations behind these mechanistic claims are ever reached in human tissue at tolerable doses.
• A caution for active, aging users of a longevity supplement: in physically inactive men (~65 years), 250 mg/day trans-resveratrol blunted several cardiovascular adaptations to 8 weeks of exercise training versus placebo (Gliemann 2013, J Physiol) — the compound interfered with one of the best-evidenced healthspan interventions rather than adding to it.PubMed ↗

Therapeutic dose: No established dose for cellular-aging endpoints. Green tea catechin trials (non-aging outcomes) used roughly 400–600 mg/day of catechins, delivering on the order of 400 mg EGCG in the decaffeinated prostate-chemoprevention trials. Keep total EGCG under 800 mg/day and take with food (EFSA hepatotoxicity threshold).

• No human RCTs test green tea extract on mitochondrial function, NAD+, cellular senescence, or oxidative-stress-of-aging endpoints. Its antioxidant and anti-angiogenic activity is documented only in cell and animal models; that mechanistic plausibility has not translated to measured human aging outcomes.
• The nearest category-relevant human trial was null: the Kumar 2015 Polyphenon E RCT (400 mg EGCG/day, decaffeinated, n=97 men with precancerous prostate lesions, 12 months) found no difference in progression to prostate cancer (5/49 catechin vs 9/48 placebo, p=0.25) — negative on its primary endpoint. It was the better-powered follow-up to a tiny 2006 pilot (n=60, Bettuzzi) that suggested a benefit but was never replicated.PubMed ↗
• For a supplement positioned around healthy aging, the hard human signal points the wrong way: the US Drug-Induced Liver Injury Network attributed 40 of 1,414 liver-injury cases (3%) to concentrated green tea extract — 35% judged severe, 3 requiring transplant — with the HLA-B*35:01 risk allele in 72% of those cases. EFSA flagged a hepatotoxicity signal at supplemental EGCG of 800 mg/day or more, especially fasted.PubMed ↗

Therapeutic dose: 200–500 mg/day (plasma saturates around 200 mg/day per inpatient pharmacokinetic data; higher oral doses spill into urine)

• Physicians' Health Study II (RCT, 14,641 male physicians aged 50+, 500 mg/day vs placebo, mean 8 years) found no reduction in total cancer incidence and, in its cardiovascular arm, no reduction in major cardiovascular events or all-cause mortality — the largest, longest test of vitamin C as a chronic-disease/longevity insurance policy came back null.
• SU.VI.MAX (RCT, 12,735 adults, 120 mg vitamin C within an antioxidant cocktail with vitamin E, beta-carotene, selenium and zinc, 7.5 years) showed no effect on cardiovascular events; a modest cancer-incidence reduction appeared in men only and cannot be attributed to vitamin C given the cocktail design.
• Mechanism only for cellular aging: vitamin C regenerates reduced vitamin E and supports glutathione recycling, but this antioxidant framing has not been tested against mitochondrial-function, NAD+, or cellular-senescence endpoints in humans — there are no category-specific RCTs, and the general antioxidant hypothesis has repeatedly failed to translate into disease or longevity outcomes in controlled trials.

Therapeutic dose: 20–40 mg/day, from the human trials: 0.2 mg/kg single dose and 0.3 mg/kg/day repeated in the 2013 crossover (~14–21 mg for a 70 kg adult), and 20 mg twice daily in the 2024 combination trial. Not derived from %DV.

• The only direct human study of PQQ on mitochondrial-related endpoints is a small crossover in 10 subjects (Harris/Rucker, J Nutr Biochem 2013): a single 0.2 mg/kg dose plus a short repeated-dose arm (0.3 mg/kg/day) lowered plasma CRP and IL-6, shifted urinary mitochondria-related metabolites, and altered TBARS antioxidant markers. These are surrogate biomarkers in 10 people, not a clinical longevity or direct mitochondrial-function outcome.PubMed ↗
• A 2024 randomized, double-blind, placebo-controlled trial in 34 older adults with mild cognitive impairment tested a combination product delivering PQQ alongside dihydrogen-producing minerals (twice daily, 6 weeks), so any effect cannot be attributed to PQQ alone. Its primary endpoint — serum BDNF, a neurotrophic proxy — rose significantly versus placebo, with some brain-oxygenation and cognitive signals, but the design confounds PQQ with the co-administered actives and it measured no direct mitochondrial-biogenesis or cellular-aging endpoint.PubMed ↗
• The proposed mechanism — upregulation of PGC-1alpha/SIRT1-driven mitochondrial biogenesis and Nrf2 antioxidant signaling — is characterized almost entirely in in-vitro and rodent work. Increased mitochondrial biogenesis has not been directly demonstrated in humans; human data are limited to inflammatory and urinary-metabolite proxies.
• No human trials evaluate PQQ against the defining cellular-aging endpoints for this category: there are no controlled studies on cellular senescence, lifespan, or NAD+ levels in people. Claims linking PQQ to longevity or NAD+ extend beyond what current human evidence supports.

Therapeutic dose: RDA 55 µg/day covers selenoprotein function; no cellular-aging benefit shown at the 200 µg/day dose used in the cancer-prevention trials (SELECT, NPC). Upper limit 400 µg/day — do not exceed.

• SELECT (Lippman 2009, JAMA), the largest cancer-prevention RCT ever run for selenium — 35,533 healthy men on 200 µg/day L-selenomethionine — found no reduction in prostate cancer and was halted early for futility. This is the most rigorous test of the oxidative-DNA-damage rationale that underpins any cellular/longevity claim, and it was null in an already selenium-replete population.PubMed ↗
• The Stranges 2007 secondary analysis of the Nutritional Prevention of Cancer trial (NPC, 200 µg/day selenium-enriched yeast) found increased incident type 2 diabetes in the selenium arm, concentrated in participants with the highest baseline selenium. For a cellular-metabolic-health framing the supplementation signal points toward harm, not benefit, and matches later observational U-shaped curves.
• Mechanistically, selenium is the cofactor for glutathione peroxidases and thioredoxin reductases that neutralize lipid hydroperoxides and regenerate antioxidant systems — settled biochemistry at the RDA level. But this antioxidant role has produced no human trials on mitochondrial function, NAD+ metabolism, cellular senescence, or longevity endpoints; moving from replete to supraphysiological status has not been shown to slow cellular aging.

Therapeutic dose: Trials used 0.9–3.3 mg/day (wheat-germ extract) up to 40 mg/day (synthetic 3HCl); no dose has changed a cellular or longevity outcome in a controlled trial

• The most-cited longevity signal is observational only: in the Bruneck community cohort (~829 adults followed roughly two decades), the top tertile of DIETARY spermidine intake had a hazard ratio near 0.6 for all-cause mortality vs the bottom tertile. High-spermidine eaters also ate whole-grain, vegetable-heavy diets, so this cannot show that a 5 mg capsule reproduces the effect of the eating pattern.
• The one human trial that tested whether the supplement even engages the pathway was essentially null: 40 mg/day of high-purity spermidine trihydrochloride (well above any retail dose) for about 4 weeks in healthy older men produced minimal rise in circulating polyamines and no meaningful change in clinical, lipid, chemistry or hematological parameters — undercutting the premise that oral spermidine meaningfully moves the target pathway in humans.PubMed ↗
• The autophagy mechanism (spermidine inhibits acetyltransferase EP300, deacetylating autophagy proteins) is well-characterized and extends lifespan in yeast, worms, flies and mice — but this is model-organism and in-vitro data, not a human cellular-aging or oxidative-stress outcome. No human trial has measured autophagy, senescence, mitochondrial, or NAD+ endpoints with supplemental spermidine.
• As a read on overall quality of evidence: the largest, best-designed spermidine RCT to date (SmartAge, 100 older adults, 12 months, non-industry-funded) was null on its primary endpoint — an adjacent cognitive outcome, not a cellular one, but it shows the strongest controlled trial available failed to confirm benefit, which is the honest backdrop for any cellular/longevity claim.