NAD+ Precursors (NMN and NR): They Raise a Biomarker, Not Proven to Slow Aging

Summary

NAD+ precursors — nicotinamide mononucleotide (NMN) and nicotinamide riboside (NR) — are sold as anti-aging pills, and the biomarker they move is real: NAD+ genuinely declines with age, and both compounds reliably raise blood NAD+ in humans across double-blind trials, a legitimate and replicated pharmacodynamic effect, with a few scattered surrogate signals (muscle insulin signalling, a muscle NAD+ metabolome shift, an anti-inflammatory transcriptomic signature) — yet the outcome that is actually marketed, slowing aging or extending healthspan and lifespan, has never been shown in a single hum

Why Emerging

Emerging Evidence because the entry's verdict deliberately splits across two very different strengths, and the outcome that matters sits at the low end. The NAD+-raising effect is Moderate and replicated — real, double-blind, reproducible pharmacodynamics. But the healthspan/longevity outcome that the category is actually sold on has never been demonstrated in a human: the trials are small, short, and mostly null on function; the strong efficacy data are in mice; and the human "positives" are unreplicated surrogates, one of them formally contested. An entry inherits the confidence of its load-bearing claim, and the load-bearing sold claim here is unproven.

NOT Moderate because there is no adequately-powered human trial with a validated or hard endpoint supporting the outcome — only a raised biomarker and contested surrogates. The pharmacodynamic sub-claim reaches Moderate on its own and is marked as such, but it cannot lift the whole verdict, because the whole verdict is about aging, not about NAD+ levels.

NOT Experimental because the mechanism is not speculative — NAD+ decline with age is well documented, the precursors demonstrably raise NAD+ in humans, and real (if unreplicated) surrogate signals exist. This is not a mechanism searching for any evidence; it is a real effect searching for a proven use.

The per-claim split (read this, not just the headline):
• NAD+ declines with age; NMN/NR raise blood NAD+ in humans: Moderate (replicated, double-blind).
• Tolerability at studied doses: Moderate for the short term; long-term safety unknown.
• Human functional/metabolic benefit: Emerging (small, mostly null, one contested surrogate).
• Slows aging / extends healthspan or lifespan in humans: Experimental/unproven (no trial; mouse data only).

Practical takeaway

The framing to hold: NMN and NR are a real pharmacodynamic tool with no proven clinical use. The pill measurably raises your NAD+. It has not been shown to make you age more slowly, live longer, or function better. Buy the biomarker as fact; treat the longevity pitch as unearned.

Don't buy it for the reason it's sold.
• Anti-aging / longevity / "restore youthful function": there is no human evidence for any of this. If that is why you are considering it, the honest answer is that no one has ever been shown to get that benefit. Money spent here buys a raised biomarker, not a demonstrated outcome.

What is actually true, at its real confidence.
• It raises NAD+: yes, reliably, in humans (Moderate, replicated). That is the one thing you can count on.
• It is well tolerated at studied doses: yes, in trials lasting weeks to a few months. Long-term safety is unknown (see Risks).
• Surrogate signals exist: muscle insulin signalling (one small contested trial), a muscle metabolome shift and anti-inflammatory signature (small trials). These are real single studies, not established benefits, and not functional outcomes. Do not upgrade them in your head to "it improves my metabolism."

Don't be moved by the tells.
• The FDA reinstatement is not validation. "NMN is a legal supplement again" is a lawsuit outcome about classification, not evidence it works. Firewall the two.
• Mouse data is not human data. Impressive rodent healthspan results are the norm in this field and have a poor translation record. Discount any pitch that cites them as if they were human findings.
• Sinclair's name is not a receipt. The commercial through-line from Sirtris onward is a reason for more scrutiny, not less.

If you choose to take it anyway. Understand you are running an n-of-1 on an unproven longevity compound, buying a biomarker with no shown outcome, at your own cost and with unknown long-term safety. That can be a legitimate informed choice — it is not the choice the marketing is describing.

Evidence detail

Why This Entry Exists

NAD+ precursors sit on the exact fault line this library was built to police: a genuine, measurable effect in the body wrapped in an outcome claim the evidence does not support. The mechanism is not vaporware — NAD+ is a real cellular cofactor for redox reactions and DNA repair, it really does fall with age, and swallowing NMN or NR really does push blood NAD+ back up. That much is replicated and honest. The problem is the leap from "your NAD+ reading went up" to "you will age more slowly," which is the leap the entire commercial category is built on and which no human trial has ever earned. This entry exists to hold those two facts apart without collapsing into either the marketing or a lazy "supplements are useless" dismissal.

The failure modes run in both directions, and both are common. Over-claim, and a healthy person spends real money on a supplement whose sold benefit — a longer, younger life — has never been demonstrated in anyone, on the strength of a biomarker and a mouse. Over-correct, and you deny that the compounds do anything, which is also wrong: the NAD+-raising is one of the cleaner pharmacodynamic effects in the supplement aisle, and a handful of human surrogate signals are real single studies, not noise. The category also carries a specific cautionary history — the resveratrol/sirtuin story that seeded this whole field was a commercial and scientific failure — and a regulatory episode (the FDA's on-again, off-again treatment of NMN) that vendors are already spinning as vindication when it is nothing of the kind. Getting the ordering right — the biomarker is Moderate-grade and real, the longevity outcome is unproven marketing, and the two must never be welded together — is the whole job of the entry.

What bad advice this protects against, in all directions:
• "NMN/NR slows aging and extends your healthspan" → no human trial has ever shown this. The healthspan and lifespan data are almost entirely in mice; the human evidence stops at a raised biomarker and a few surrogate signals, and species translation for longevity compounds has a graveyard.
• "My NAD+ level went up, so it's working" → raising NAD+ is a pharmacodynamic effect, not a clinical outcome. A higher NAD+ reading is not living longer or better, and no one has been shown to convert one into the other. The surrogate-endpoint trap is the whole game here (defer the general argument to surrogate_endpoints_vs_outcomes).
• "NMN doesn't do anything" → the opposite over-correction. It reliably raises NAD+, is well tolerated at studied doses, and has produced real single-study surrogate signals. Dismissing the pharmacodynamics is as wrong as believing the marketing.
• "The FDA says NMN is a supplement again, so it's validated" → that is a legal outcome about product classification (an industry lawsuit won), not evidence that NMN works. The two must be firewalled.
• "David Sinclair's science proves the anti-aging mechanism" → the sirtuin/resveratrol precursor to this field collapsed under an assay artifact and outright fraud; the mechanistic story that seeded NAD+ marketing is a cautionary tale, not a foundation (see resveratrol_and_longevity_claims).
• "The insulin-sensitivity trial proved metabolic benefit in humans" → it moved a surrogate (muscle insulin signalling) in 25 women, left every clinical metric unchanged, and drew a published rebuttal in the same journal. That is a signal, not a settled outcome.

This entry owns the biomarker-real-outcome-unproven verdict for NMN and NR and the receipts behind it. It defers the general principle that a moved surrogate is not a clinical outcome to surrogate_endpoints_vs_outcomes — here it only NAMES the trap and shows the receipts. It defers the whole-of-category resveratrol/sirtuin collapse to resveratrol_and_longevity_claims, and the "does this clear the bar for near-universal use" question to universal_nearuniversal_supplementation.

Evidence

Organised so the split is unmissable: the biomarker moves, the outcome does not. Read the tier signal on each finding, not just the claim.

The biomarker claim is real and replicated (Moderate), but the human function endpoints are mostly null (Emerging-to-Experimental).

1. NMN reliably raises blood NAD+ in humans, but a systematic review of the human RCTs found physical-performance benefits were non-significant. Pooling ten NMN randomised controlled trials (437 patients, mean age 58, mean 9.6-week follow-up, doses 150–1200 mg/day), the review found NMN was well tolerated with no serious adverse events and "demonstrated non-significantly improved physical performance parameters." The biomarker moves; function largely does not. (Systematic review of 10 NMN RCTs, Cureus 2024, PMC11365583. Tier signal: NAD+-raising is Moderate and replicated; the function/healthspan outcome is Emerging-to-Experimental and mostly null. The NMN human-trial literature is heavily manufacturer-adjacent — several trials are industry-sponsored, and positive-spin write-ups cluster on vendor sites; cui bono runs through the positive framing.)

2. The flagship human NMN result moved a surrogate — muscle insulin signalling — not a clinical outcome, in a tiny sample, and was formally contested in the same journal. In 25 postmenopausal, prediabetic women who completed the trial on 250 mg/day for ten weeks, insulin-stimulated muscle glucose disposal and AKT/mTOR signalling rose, but body weight, plasma glucose, HbA1c, blood pressure and lipids did not change. A published Comment in Science challenged the interpretation. (Yoshino, Klein et al., Science 2021, doi:10.1126/science.abe9985; contesting Comment, Science, doi:10.1126/science.abj1696. Tier signal: Emerging — a single small trial on a surrogate endpoint, contested, and not replicated for any hard outcome. The NMN was industry-supplied; vendors reframed it as "metabolic benefits of NMN shown for the first time in humans," selling a surrogate as an outcome.)

3. NR (nicotinamide riboside) reproducibly raises NAD+ in blood and aged muscle, but the function endpoints came up null. In a six-week crossover trial NR was well tolerated and raised NAD+ by roughly 60%, with no significant effect on blood pressure or arterial stiffness overall — a signal appeared only in the subgroup with elevated baseline systolic pressure. In twelve aged men given 1 g/day for 21 days, NR raised the muscle NAD+ metabolome and produced an anti-inflammatory transcriptomic signature, with no proven clinical functional benefit. (Martens et al., Nature Communications 2018, PMID 29599478; Elhassan et al., Cell Reports 2019, PMC6702140. Tier signal: NAD+-raising is Moderate and replicated; clinical function is Emerging and null on the primary endpoints. NR is patent-protected and commercialised — ChromaDex funds a large share of NR human research, so cui bono runs directly through the trials.)

The mechanistic story that seeded the field failed (History/Experimental).

4. The sirtuin/resveratrol precursor to this whole field was a commercial and scientific failure — the cautionary tale for NAD+/longevity hype. GSK bought Sirtris (co-founded by David Sinclair) for $720M in 2008; its lead resveratrol drug SRT501 was halted in 2010 after kidney failures in a multiple-myeloma trial; a 2010 Pfizer analysis showed the claimed SIRT1 "activation" was a fluorophore assay artifact; GSK shut Sirtris in 2013. Separately, resveratrol researcher Dipak Das (UConn) was found to have fabricated data across roughly twenty retracted papers. (Tier signal: History/Experimental — the mechanistic narrative that marketing later transferred onto NAD+ collapsed under an assay artifact and fraud. The same principal, Sinclair, recurs across Sirtris → MetroBiotech (an NMN drug) → advisory roles at ChromaDex/Tally Health/InsideTracker — the commercial through-line of the anti-aging story. Detail deferred to resveratrol_and_longevity_claims.)

The regulatory episode is legal, not evidentiary — and cuts both ways (Context only).

5. The FDA's NMN saga exposes the drug-versus-supplement bait; the reinstatement is a legal outcome, not a finding that NMN works. In late 2022 the FDA ruled NMN excluded from the dietary-supplement definition under the DSHEA drug-preclusion clause, because MetroBiotech (Sinclair-linked) had authorized NMN for investigation as a new drug before it was marketed as a supplement. The FDA reversed this in September 2025 after a Natural Products Association lawsuit, and NMN can again be sold as a supplement. (NutraIngredients 2025-09-30; Venable LLP 2025. Tier signal: regulatory context, not efficacy evidence — reinstatement is a legal result of a trade-group lawsuit, not proof of benefit. The drug-preclusion basis is itself the tell: the molecule is being developed as a pharmaceutical precisely because supplement-grade evidence for a longevity outcome does not exist.)

The rest of the longevity-supplement class shares the pattern (Experimental/null).

6. Adjacent longevity-supplement "outcome" claims are equally unearned — the best-designed spermidine cognition RCT was null on its primary endpoint. In 100 adults aged 60–90 with subjective cognitive decline, twelve months of spermidine (0.9 mg/day wheat-germ extract) versus placebo produced no significant benefit on the pre-registered primary outcome (mnemonic discrimination). (SmartAge trial, Schwarz et al., JAMA Network Open 2022, PMID 35616942. Tier signal: Experimental/null — a well-powered, long, registered trial that failed its primary endpoint, the pattern across this supplement class. Owned by other entries and included here only as a class cross-check; senolytics such as dasatinib+quercetin remain early-phase open-label pilots — e.g. Justice et al. 2019, an IPF pilot with n=14 — with no hard-outcome RCT, the same surrogate-endpoint posture.)

Mechanism

This entry owns only enough mechanism to make the biomarker-versus-outcome split intelligible; the general surrogate-endpoint logic is deferred to surrogate_endpoints_vs_outcomes.

Why the biomarker is real. NAD+ (nicotinamide adenine dinucleotide) is an obligate cofactor for cellular redox reactions and a substrate consumed by DNA-repair enzymes (PARPs) and the sirtuins. Tissue and blood NAD+ genuinely decline with age, which is a documented observation, not a marketing invention. NMN and NR are precursors the body converts into NAD+ through the salvage pathway, so supplying them raises the available pool — and human trials confirm blood and even aged-muscle NAD+ rise on dosing. This is a clean, saturable pharmacodynamic effect: pill in, cofactor up. Nothing about that step is contested.

Why raising the cofactor is not the same as slowing aging. The marketing runs a syllogism: NAD+ falls with age; sirtuins and repair enzymes need NAD+; therefore restoring NAD+ restores youthful function. Each premise is individually defensible and the conclusion still does not follow, because a correlation between low NAD+ and aging does not establish that topping the pool up reverses the downstream biology in a living human. The body may already be operating on the flat part of the dose-response, the extra NAD+ may not reach the compartments that matter, and "the enzyme has more substrate" is not "the organism ages more slowly." That gap — between a moved cofactor and a changed outcome — is exactly where every unproven claim in this entry lives.

Why the surrogate signals are suggestive but not sufficient. The real human signals (muscle insulin signalling, a muscle NAD+ metabolome shift, an anti-inflammatory transcriptomic pattern) are all one step upstream of anything a person would feel or a clinician would measure as an outcome. They are consistent with the mechanism, which is why they are worth noting rather than dismissing — but a transcriptomic signature is a biomarker of a biomarker. Until a trial shows NAD+ elevation statistically mediating a functional gain, the mechanistic chain is plausible and unclosed.

Why the mouse-to-human leap is the load-bearing weakness. The strong healthspan and lifespan data are in mice, and this class has a specific record of failing to translate: resveratrol, the compound that launched the sirtuin narrative, looked spectacular in model organisms and collapsed in humans under an assay artifact and fraud. Mouse aging biology is not human aging biology, and longevity is the one endpoint you cannot shortcut with a surrogate. That is why the entry treats the mouse data as genuinely strong and still refuses to let it carry the human outcome.

Risks And Contraindications

• The main risk is opportunity cost and false reassurance, not acute toxicity. At the doses studied, NMN and NR are well tolerated with no serious adverse events reported in trials. The realistic harm is money spent on an unproven longevity claim, plus the false confidence of "doing something about aging" on the strength of a moved biomarker.
• Unproven long-term safety of a novel longevity compound. The human trials are short — weeks to a few months. Chronic, multi-year dosing of an NAD+ precursor has not been characterised for safety, and a cofactor that feeds cell-proliferation and repair pathways is not something whose long-term elevation should be assumed benign without data. Absence of harm in a six-week trial is not evidence of safety over years.
• Do not treat a raised NAD+ level as a health outcome. The specific cognitive risk of this category is mistaking the biomarker for the benefit — a person who sees their NAD+ rise (or simply believes it has) may deprioritise the interventions that actually move hard outcomes (sleep, training, diet). The biomarker is not the goal; do not let it displace the basics (see chronic_disease_risk_mitigation).
• The FDA reversal is not a safety or efficacy clearance. Reinstatement as a supplement resolved a legal classification dispute; it made no finding about benefit or long-term safety. Do not read regulatory availability as endorsement.
• Keep the surrogate signals at their real confidence. The insulin-signalling, metabolome, and anti-inflammatory findings are unreplicated single studies on surrogate endpoints. Treating them as established metabolic or anti-inflammatory benefits over-states the evidence in the same direction the marketing does.
• Avoid the opposite over-correction. "NMN/NR does nothing" is also wrong. The NAD+-raising effect is real and replicated, and NAD+ decline with age is well documented. The honest message is "a real pharmacodynamic effect with an unproven outcome," not a blanket dismissal.

Controversy

Nature: a supplement class with a genuine, replicated pharmacodynamic effect (it raises NAD+) sold on an outcome (slowing aging, extending healthspan) that no human trial has demonstrated, with error possible at both poles — believing the biomarker equals the benefit on one side, and dismissing the real pharmacodynamics on the other.

Position A — "NAD+ precursors have a real, replicated effect." The mechanism-is-real take.
• Best evidence: NAD+ genuinely declines with age; both NMN and NR reliably raise blood (and aged-muscle) NAD+ in double-blind human trials; the compounds are well tolerated at studied doses; and there are scattered real surrogate signals (muscle insulin signalling, a muscle NAD+ metabolome shift, an anti-inflammatory transcriptomic signature). The pill does something measurable.
• Where it goes wrong if overstated: it slides from "raises NAD+" and "moved a surrogate" to "slows aging" — welding a biomarker to an outcome no human has been shown to get.

Position B — "The marketed outcome has never been shown in humans." The debunk take.
• Best evidence: the healthspan/lifespan data are almost entirely in mice; the human RCTs are small (n=12–25), short, and mostly null on function; a systematic review of ten NMN trials found non-significant performance effects; the flagship insulin result moved a surrogate in 25 women and was formally contested; and a large commercial ecosystem sells the NAD+ number as the outcome. The anti-aging claim is unearned.
• Where it goes wrong if overstated: it can tip into "NAD+ precursors do nothing," which ignores the reliable, replicated NAD+-raising and the real (if unreplicated) surrogate signals.

The funding/bias dimension — cui bono, both ways. Toward over-claiming: nearly all the commercial pressure. The field runs on a small set of actors — ChromaDex (NR/Tru Niagen, patent-holder, funds a large share of NR human trials), Elysium Health (Basis), MetroBiotech (NMN-as-drug) — plus a recurring principal, David Sinclair, whose commercial trail runs Sirtris → MetroBiotech → ChromaDex/Tally Health/InsideTracker advisory and who is the loudest voice for the anti-aging framing. Many "positive" human studies are manufacturer-funded or manufacturer-supplied, and vendor sites reframe surrogate results as the sold outcome. Toward the corrective pole: cui bono the other way is weak — independent academics and the FDA drug-preclusion posture both point against supplement-grade longevity claims, and skeptics have nothing to sell. The resveratrol/Sirtris collapse (assay artifact + fabricated data) shows the same commercial machine previously over-promised and failed.

Realised Position: Both are true, and that is the point. NAD+ precursors are a real pharmacodynamic tool searching for a proven clinical use. Buy the NAD+-raising as fact (Moderate-grade, replicated). Treat every anti-aging, longevity, or healthspan claim as unproven marketing built on mouse data and a moved biomarker. Do not confuse a higher NAD+ reading with living longer or better — no human has been shown to. This is Emerging: a real signal attached to an unearned outcome.

Cross-Pillar Connections

NMN and NR are a diet/supplement topic whose whole controversy is an evidence-methodology one, so its connections run along the supplement-literacy and evidence-method lines.
• Foundations (surrogate_endpoints_vs_outcomes): owns the general argument that a moved biomarker is not a clinical outcome; this entry only NAMES the trap and supplies the NMN/NR receipts, and defers the principle there.
• Supplements (resveratrol_and_longevity_claims): owns the sirtuin/resveratrol collapse in full; this entry holds only that the failed mechanistic story seeded the NAD+ marketing, and hands the detail off.
• Supplements (universal_nearuniversal_supplementation): the reference point for which supplements clear the bar for near-universal use — NMN/NR conspicuously do not, which is the contrast this entry draws.
• Foundations (publication_bias_and_evidence_distortion): the mechanism behind the manufacturer-funded-positive and surrogate-sold-as-outcome pattern that inflates NAD+ precursors' apparent evidence.
• Foundations (cui_bono_industry_funding_bias): the framework for reading who benefits — here, structurally, the sellers — and why the conflict-clean counter-claims deserve more weight.
• Conditions (chronic_disease_risk_mitigation): the honest place to put longevity effort — the interventions with hard-outcome evidence — so a biomarker chase does not displace them.

What would change our mind

Industry bias note

Sources (8)

• Systematic review of 10 NMN randomised controlled trials (2024). Cureus, PMC11365583. (Includes industry-sponsored/-supplied trials; positive-spin write-ups cluster on vendor sites.) — 437 patients, mean age 58, mean 9.6-wk follow-up, 150–1200 mg/day; NMN well tolerated, no serious AEs, "non-significantly improved physical performance parameters."↗
• Yoshino J, Klein S, et al. (2021). "Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women." Science, doi:10.1126/science.abe9985↗; with the contesting Comment, Science, doi:10.1126/science.abj1696.↗ (NMN industry-supplied; reframed by vendors as first-in-human metabolic benefit.) — 25 postmenopausal prediabetic women completed, 250 mg/day × 10 wk; insulin-stimulated muscle glucose disposal and AKT/mTOR signalling rose; body weight, plasma glucose, HbA1c, BP, lipids unchanged; interpretation formally challenged.
• Martens CR, et al. (2018). "Chronic nicotinamide riboside supplementation is well-tolerated and elevates NAD+ in healthy middle-aged and older adults." Nature Communications, pubmed.ncbi.nlm.nih.gov/29599478↗/" target="_blank" rel="noopener">PMID 29599478↗. (NR is commercialised; ChromaDex funds a large share of NR research.) — 6-wk crossover; NR well tolerated, raised NAD+ ~60%; no significant BP/arterial-stiffness effect overall, signal only in the elevated-systolic subgroup.
• Elhassan YS, et al. (2019). "Nicotinamide Riboside Augments the Aged Human Skeletal Muscle NAD+ Metabolome and Induces Transcriptomic and Anti-inflammatory Signatures." Cell Reports, PMC6702140. (Same commercial context as above.) — 12 aged men, 1 g/day × 21 d; raised muscle NAD+ metabolome and produced an anti-inflammatory transcriptomic signature; no proven clinical functional benefit.↗
• Sirtris/resveratrol history: GSK acquisition of Sirtris (2008, $720M); SRT501 halted (2010); Pfizer analysis of the SIRT1 fluorophore assay artifact (2010); GSK closure of Sirtris (2013); Dipak Das (UConn) data-fabrication finding (~20 retracted papers). (Cautionary precedent for the whole class; detail owned by resveratrol_and_longevity_claims.)↗
• FDA NMN classification: exclusion from the dietary-supplement definition under the DSHEA drug-preclusion clause (late 2022, on the basis of prior authorized new-drug investigation by MetroBiotech); reversal after the Natural Products Association lawsuit (Sept 2025). (NutraIngredients 2025-09-30; Venable LLP 2025. Regulatory/legal context, not efficacy evidence.)↗
• SmartAge trial — Schwarz C, et al. (2022). "Effects of Spermidine Supplementation on Cognition and Biomarkers in Older Adults With Subjective Cognitive Decline: A Randomized Clinical Trial." JAMA Network Open, pubmed.ncbi.nlm.nih.gov/35616942↗/" target="_blank" rel="noopener">PMID 35616942↗. (Class cross-check; owned by other entries.) — 100 adults 60–90, 12 mo spermidine (0.9 mg/day) vs placebo; no significant benefit on the pre-registered primary memory endpoint. Senolytic (dasatinib+quercetin) pilots — e.g. Justice JN, et al. (2019), IPF pilot, n=14 — remain early-phase/open-label with no hard-outcome RCT.
• Funding notation: the load-bearing counter-claims (the null function trials, the mouse-only longevity data, the surrogate-and-contested human "positives") come from the trials' own primary endpoints and from conflict-clean academic scrutiny, and skeptics have nothing to sell — so those claims are trustworthy. The most commercially-motivated claims (the anti-aging framing, the surrogate results reframed as outcomes, the FDA reinstatement spun as validation) are exactly the ones the entry marks and hedges. The bias vector runs almost entirely toward over-claiming benefit.*↗