NAD+ Precursors (NMN and NR)

Supplements · §532

The fuel molecule your cells run on — NAD+ — falls steadily from your thirties on. Two supplements, NMN and NR, reliably push it back up. Whether that does anything you'd actually notice is where the field splits. The honest summary: real biology, contested clinical translation, low harm, mid cost, and a regulatory back-and-forth that's mostly noise.

Do · Daily Evidence Mixed Chapter Supplements

The trials reliably show one thing: take the pill, and the level of this fuel molecule in your blood goes up. Trials don't yet reliably show much else — a flicker of endurance for runners already training, a quarter-step gain in walking speed for older men, a contested insulin-sensitivity result in one small trial. None of it transformative. The case for taking it is the longevity bet: the underlying biology fits a real story about aging, and at $30–60 a month it's a cheap call option on that story panning out. The case against is just as honest: the human evidence is thin enough that the same money buys you a year of creatine, an omega-3 supply, and equipment for the strength work that beats every supplement on the shelf.

Every cell in your body runs on NAD+ the way a phone runs on its battery — it carries energy from the food you ate to the engine that uses it. The engines (your mitochondria) burn through NAD+ continuously. So do the maintenance crews: a family of enzymes called sirtuins that quietly tune your metabolism in the background, and another crew (PARPs) that scramble to patch DNA whenever something nicks it. All of them grab NAD+ and spit out a depleted form. Your cells recycle the depleted form back into fresh NAD+ on a continuous loop.

The loop runs slower as you age. The recycling enzyme that does the heavy lifting (NAMPT) makes less of itself when inflammation creeps up — and inflammation creeps up with the decades. A separate enzyme called CD38, which destroys NAD+, ramps up in older tissues. By 70, skeletal-muscle NAD+ in humans is roughly half what it was at 30. The same drop shows up in skin, liver, and brain on the available tissue measurements (Verdin 2015).

NMN and NR are both shortcut substrates feeding into this loop. Swallow either one, and within a few weeks the NAD+ measurable in your blood goes up — usually doubling, sometimes more.

One detail matters for the marketing: it's not settled whether NMN gets into your cells intact, or whether your gut chops it down to plain nicotinamide first and your body rebuilds NAD+ from there. Imai's lab claimed in 2019 to have found a dedicated NMN gut transporter (Grozio et al. 2019); Charles Brenner's lab published a rebuttal months later arguing the same gene actually moves calcium, not NMN (Schmidt & Brenner 2019). The dispute hasn't been resolved in humans. The practical upshot for you: whether the pricier form (NMN) is doing anything the cheaper form (NR), or even plain niacinamide, can't, is genuinely open.

What the trials actually show

The blood NAD+ rise is the easy part. Every trial replicates it. What people actually want to know is whether the rise translates into anything you'd notice in your life. Here the story gets thinner.

Insulin sensitivity. The headline result came from a 2021 Science paper: 25 postmenopausal women with prediabetes took 250 mg of NMN daily for ten weeks, and the NMN group showed about 25% better skeletal-muscle insulin sensitivity than placebo. The paper made NMN famous overnight.

Endurance for people already training. A six-week trial of amateur runners in China — 48 people, training five-to-six sessions a week — gave the runners either placebo or NMN at 300, 600, or 1,200 mg/day. The runners on the higher doses got more efficient at sustained submaximal effort. Top-end power, VO₂max, and peak heart rate didn't move; the ceiling of what their bodies could do stayed put. What shifted was the sustainable middle of a long effort — the threshold below which you can keep going without falling apart (Liao et al. 2021). The honest reading: if you're training, NMN may push the floor of a hard session up a notch. It won't raise the ceiling.

Walking speed and grip in older adults. 42 healthy Japanese men over 65 took 250 mg/day of NMN for 12 weeks; gait speed improved from 1.50 to 1.60 m/s (placebo flatlined), and left-hand grip strength edged up (Igarashi et al. 2022). A larger dose-response trial in 80 middle-aged adults in India saw the six-minute walk distance rise meaningfully at every NMN dose versus placebo, with NAD+ blood levels and walk distance both plateauing around 600–900 mg/day (Yi et al. 2023).

Cognition. This one is the cleanest miss. Older adults with early memory decline took 1,000 mg of NR daily for eight weeks. NAD+ went up; the cognitive battery didn't move (Orr et al. 2024). A larger long-COVID NR trial at twice that dose for twenty weeks had the same shape: NAD+ rose, fatigue and brain fog didn't reliably budge. There's one bright spot — Norwegian researchers gave early-stage Parkinson's patients high-dose NR for a month, and the half who showed brain-NAD+ rising on a scan also showed altered brain energy use and a small slip on the Parkinson's symptom scale (Brakedal et al. 2022). Promising, but Phase I, and the clinical signal blurs with their levodopa timing.

Heart and circulation. The peripheral-artery-disease trial published in 2024 is the largest cardiovascular signal so far: NR at 2 g/day for six months meaningfully improved how far patients could walk in six minutes — a clinical endpoint that genuinely matters in PAD (McDermott et al. 2024). Blood pressure and arterial-stiffness trends across smaller trials hint in the right direction but rarely cross the statistical line on their own (Martens et al. 2018) (Kim et al. 2023).

Longevity itself. No human trial measures lifespan, mortality, dementia incidence, or fracture risk. The longevity claim rests entirely on mouse data — modest lifespan extensions in some cohorts, none in others — and on the underlying NAD+/sirtuin biology being plausible (Imai & Guarente 2014). If you're taking NMN or NR for longevity, you're making a mechanism-based bet, not following a finding.

If you're going to take it

Pick one of NMN or NR; running both is pointless. NMN is pricier and the marketing premium is built on the contested bioavailability claim — pick NR if you want the cheaper, less-disputed option. The trials that found something used standard daily doses; bigger isn't better.

Try it for three months. If you're training, look at how recovery between sessions and your sustainable pace on a long effort change — that's where the trial signal lives. If you're not training, the felt-experience window is small enough that you may simply not be able to tell. That's fine; the longevity bet doesn't require you to feel anything.

When to skip

The standard safety story across the 8-to-12-week trials is clean: no serious adverse events at doses up to 1,250 mg/day NMN and 3,000 mg/day NR (Conze et al. 2019) (Brakedal et al. 2022). What's missing is the multi-year data. If you start in your forties and take it daily for thirty years, nobody has the trial that tells you what happens.

What gets oversold

"NMN is better than NR because of a special transporter." This is the line on every NMN sales page. The transporter was reported in mice in 2019 and rebutted four months later by a different lab using the same methods (Grozio et al. 2019) (Schmidt & Brenner 2019). Whether it exists at scale in human guts is unresolved. No head-to-head trial compares NMN against NR on any clinical endpoint. The premium price isn't backed by clinical data.

"Raising NAD+ reverses aging." Raising NAD+ in your blood is real and measurable. What that does to your actual aging trajectory is the open question. The trials that have looked at hard endpoints — cognition, insulin sensitivity in healthy adults, blood pressure — have mostly come back null or noisy. The clearest positive signals are in submaximal exercise markers and in disease populations, not in the wellness-shopper demographic the marketing targets.

"Plain niacin or niacinamide is much worse." The cheaper precursors raise NAD+ too. Whether they do it as efficiently as NMN/NR is partly the same Brenner-vs-Imai dispute and partly under-studied. Niacin causes flushing at lipid-lowering doses, which is why people moved off it; niacinamide doesn't flush and costs pennies a day. If the mechanism story collapses on the bioavailability question, you've been paying 10–50× the price of niacinamide for the same biology.

"The FDA banned NMN." The FDA temporarily excluded NMN from the supplement definition in late 2022 because a pharmaceutical developer had filed an investigational drug application — the rule designed to prevent supplement-makers from stealing drug-development work. The FDA reversed itself in September 2025 on the grounds that NMN was sold as a supplement before the drug filing (FDA 2025). NMN is now lawful again. The episode says more about regulatory mechanics than about NMN's safety or efficacy.

What else raises NAD+ — and may matter more

The biggest NAD+ lever is the one nobody sells. Hard exercise — both cardio and lifting — drives the recycling enzyme that rebuilds NAD+ in muscle. The published muscle-NAD+ increases from a few weeks of structured training rival what 12 weeks of supplement does to blood NAD+, and exercise comes with everything else exercise comes with.

Eating fewer calories, or compressing your eating into a shorter window, pushes the same lever. So does plain nicotinamide (niacinamide) at supplement doses, at a fraction of the price of NMN or NR.

If the goal is the recovery-and-training-edge claim NMN gets pitched on, creatine is the supplement with stronger evidence, lower cost, and a longer safety record. If the goal is the longevity bet, the same dollars across a year of NMN buy more durable improvements when spent on better sleep, a strength habit, and an omega-3 supply.

What you might actually notice

Be honest about the timescale. In the first few weeks, the only thing that's reliably changed is a number in your blood you can't feel. By two to three months, if you're a hard-training amateur runner or cyclist, the long efforts may feel a touch more sustainable — the version of you that used to fade in the back half of a workout fades a little less (Liao et al. 2021). People around you don't notice; you might.

If you're in your sixties and your walking pace has slowed, a 250 mg daily dose over three months may quietly add a step every couple of seconds — the kind of shift that doesn't feel like anything but shows up when you measure it (Igarashi et al. 2022). The rest of life looks the same.

Over years, the case the supplement is sold on is the longevity bet — that quietly keeping NAD+ higher through your decades trims the slope of decline. No human trial has measured this yet. If the bet pays off, it pays off in things you can't catalogue: the version of you at 75 who can still do the things the version of you at 50 could. That's a real possible payoff. It's also the kind of payoff no entry can promise.

Creatine — the better-evidenced training-and-cognition supplement, at one-tenth the cost.
Strength training — the largest single lever on age-related decline; the floor under any supplement stack.
Time-restricted eating and caloric restriction — the dietary patterns that raise NAD+ through the same pathway, for free.
Omega-3 supplementation — the longevity-stack staple with stronger trial evidence than NAD+ precursors.
Sleep duration and consistency — the upstream variable on every aging-related claim NMN gets sold on.

Related in the handbook

Alternatives

— The honest comparison: the same monthly spend buys creatine, which has far stronger evidence for the things people hope NAD precursors do.

— NMN and NR are pricey and lightly regulated. If you're betting on them, pay for a brand that's actually been tested.
— Both are supplements where dose and indication matter; omega-3 has a narrow but proven use, NAD precursors are still a longevity bet.
— The blunt counterpoint: the strength work this money could fund beats any anti-aging pill on the shelf for staying strong into old age.

Substance and claimed effects

NAD+ precursors are oral supplements — most commonly nicotinamide mononucleotide (NMN) and nicotinamide riboside (NR) — that act as substrates in the salvage pathway producing nicotinamide adenine dinucleotide (NAD+), the redox cofactor central to mitochondrial ATP production, sirtuin and PARP signalling, and a broad swath of metabolic regulation (Verdin 2015) (Imai & Guarente 2014). Tissue NAD+ falls with age in skin, muscle, liver and brain in animal models and in cross-sectional human data; the marketing claim is that restoring NAD+ via daily oral precursor reverses age-related metabolic decline, raises cellular energy, improves exercise capacity, supports cognition, and bends longevity trajectories. This entry covers the substance class (NMN + NR taken at typical supplement doses, 250–1,000 mg/day) and its meaningful consequences across cellular energy markers, age-related metabolic measures (insulin sensitivity, arterial stiffness, blood pressure), exercise capacity (aerobic endurance, gait speed, grip strength), cognition (mild cognitive impairment, Parkinson's brain NAD+), and the strength of supporting evidence. Mechanism, safety, dose-response, oral bioavailability controversy, and the FDA regulatory whiplash are all in scope.

Evidence by addressing question

mechanism

NAD+ is consumed continuously by sirtuins (SIRT1 nuclear, SIRT3 mitochondrial), PARPs (DNA damage response), and the cADP-ribose synthases CD38/CD157 (Verdin 2015). The salvage pathway regenerates NAD+ from nicotinamide via NAMPT (rate-limiting) → NMN → NMNAT → NAD+. NR enters the same pathway one step earlier, phosphorylated by NRK1/NRK2 to NMN. The aging story has two arms: NAMPT expression declines with inflammaging (TNF-α suppresses transcription), and CD38 expression rises in senescent-cell-recruited macrophages, accelerating NAD+ consumption. Verdin's 2015 model formalised the feed-forward loop: DNA damage activates PARP → NAD+ depletes → SIRT1 activity drops → NF-κB derepresses → more inflammation → more NAMPT suppression → lower NAD+ (Verdin 2015). Tissue NAD+ in skeletal muscle drops roughly 50% between age 30 and age 70 in cross-sectional human MRS data. Restoring NAD+ should, on this model, restore sirtuin activity, mitochondrial biogenesis (PGC-1α deacetylation), and the broader metabolic phenotype.

The oral bioavailability question is the live mechanistic dispute. Imai's group reported in 2019 that Slc12a8 is a dedicated intestinal NMN transporter, induced under low NAD+, allowing intact NMN absorption (Grozio et al. 2019). Schmidt & Brenner countered the same year that the gene encodes a cation/calcium transporter, not an NMN transporter, and that orally administered NMN is largely degraded to nicotinamide by gut bacteria and enterocyte CD73 before reaching peripheral tissues (Schmidt & Brenner 2019). The dispute is unresolved at the human level. Pharmacokinetic data are consistent with both readings: oral NMN raises whole-blood NAD+ reliably, but isotope-tracer work in mice shows the bulk of the label arriving at tissues as nicotinamide, not NMN, suggesting that whether you swallow NMN, NR, or plain nicotinamide, the systemic NAD+ rise is largely driven by salvage from nicotinamide. The practical implication: NMN's marketed premium over cheaper nicotinamide is largely arbitrage on a contested mechanism.

evidence

NAD+ elevation. The least-contested finding. Both NMN and NR raise whole-blood NAD+ in a dose-dependent fashion across every controlled trial. NR at 1,000 mg/day for 6 weeks elevated peripheral blood mononuclear cell NAD+ by ~60% versus placebo in healthy 55–79-year-olds (Martens et al. 2018). NMN at 250 mg/day for 12 weeks roughly doubled whole-blood NAD+ in older Japanese men (Igarashi et al. 2022). The Yi 2023 dose-response trial in 80 middle-aged Indian adults found whole-blood NAD+ rising from ~24 µM at placebo to ~42 µM at 500 mg and ~59 µM at 1,000 mg (Yi et al. 2023). High-dose NR (3,000 mg/day) in Parkinson's patients tripled-to-quintupled blood NAD+ and, critically, raised cerebral NAD+ measured by ³¹P-MRS — the first human demonstration that oral precursor reaches the brain (Brakedal et al. 2022).

Insulin sensitivity. Yoshino et al. (Science 2021) — the highest-profile NMN trial — gave 250 mg/day for 10 weeks to 25 postmenopausal women with prediabetes and reported a 25% increase in skeletal-muscle insulin sensitivity (hyperinsulinemic-euglycemic clamp) versus zero change in placebo (Yoshino et al. 2021). The trial has substantive methodological criticism: baseline hepatic lipid was 6.3% in the NMN arm versus 14.8% in placebo (P = 0.003) — a 2.4-fold imbalance on the variable most predictive of insulin sensitivity, in a 25-person trial. Brenner's published critique argues the randomisation failed and the apparent effect could be regression to mean. No replication in a larger or better-balanced trial exists. The Remie et al. AJCN trial of NR (1,000 mg/day, 6 weeks, healthy obese, crossover) found no change in insulin sensitivity, mitochondrial function, hepatic lipid, intramyocellular lipid, blood pressure, or inflammatory markers — only minor body composition shifts and altered skeletal-muscle acetylcarnitine (Remie et al. 2020). A 2024 systematic review of eight NMN RCTs concluded that short-term NMN does not improve glucose or lipid markers in generally healthy adults.

Arterial stiffness and blood pressure. Martens 2018 (NR 1,000 mg/day, 6 weeks, crossover, 24 adults) reported trends toward lower systolic BP (~−10 mmHg in the stage-1 hypertensive subgroup) and lower carotid-femoral pulse wave velocity, but the whole-cohort effects were not statistically significant (Martens et al. 2018). Kim et al. 2023 (NMN 250 mg twice daily, 12 weeks, 36 middle-aged adults) replicated the arterial-stiffness signal as a non-significant trend (Kim et al. 2023). The larger NCT03821623 follow-up to Martens (94 participants, 3 months, elevated baseline SBP) is the cleanest test of the BP claim; results have been mixed and the effect, if real, is small.

Exercise capacity. Liao et al. 2021 randomised 48 amateur runners 27–50 years old to placebo or NMN at 300/600/1,200 mg/day for 6 weeks during a structured training programme. Ventilatory threshold (the lactate-threshold proxy, an endurance-relevant submaximal marker) rose dose-dependently in the 600 mg and 1,200 mg arms versus placebo. VO₂max, peak power and O₂-pulse did not differ (Liao et al. 2021). Authors framed the effect as muscle-mediated, not cardiac. Igarashi 2022 in older men found gait speed rose ~7% (1.50 → 1.60 m/s) in the NMN 250 mg arm while placebo flatlined; left-hand grip strength also improved (Igarashi et al. 2022). Yi 2023 found the 6-minute walk distance increased significantly in all three NMN doses over placebo at days 30 and 60 (Yi et al. 2023). The NICE trial (NR 1,000 mg twice daily, 6 months) in peripheral artery disease patients reported a meaningful improvement in 6-minute walk distance versus placebo — the largest exercise-capacity signal to date (McDermott et al. 2024).

Cognition. The cognitive-benefit story is weak. Orr et al. 2024 ran a randomised, crossover, placebo-controlled trial of NR 1,000 mg/day for 8 weeks in older adults with subjective cognitive decline and mild cognitive impairment, using the RBANS battery as primary endpoint. NR safely raised NAD+ but did not move cognition (Orr et al. 2024). Brakedal et al. 2022 in early Parkinson's found that the subset of NR responders with brain-NAD+ elevation showed altered cerebral metabolism (FDG-PET) and mild MDS-UPDRS improvement, but the trial was Phase I and the clinical signal was confounded by levodopa dose-timing (Brakedal et al. 2022). A 2023 long-COVID NR trial (2,000 mg/day, 20 weeks) raised NAD+ but missed primary endpoints on cognition, fatigue, sleep and mood. The pattern is consistent: NR reaches the brain, but raising brain NAD+ has not yet translated to detectable cognitive benefit in any rigorous human trial.

Longevity, hard endpoints. No human trial reports mortality or major disease-incidence endpoints — the trials are weeks-to-months long, in populations of dozens-to-hundreds. The longevity claim rides entirely on rodent data (lifespan extensions of 5–15% in some NMN-fed mouse cohorts, mixed in others) and on the mechanistic hope that restoring NAD+ recapitulates the sirtuin-activation arm of caloric restriction.

protocol

Published trials cluster at 250–1,000 mg/day for NMN and 300–2,000 mg/day for NR. The Yi 2023 dose-response found NAD+ elevation and 6-minute walk benefit plateauing around 600–900 mg NMN (Yi et al. 2023). NR safety has been demonstrated at sustained 1,000–3,000 mg/day in healthy adults and Parkinson's patients (Conze et al. 2019) (Brakedal et al. 2022). Both compounds are administered once or twice daily, oral capsule, with or without food (no formal food-effect data showing a difference). Sublingual and liposomal formulations are marketed; no controlled head-to-head pharmacokinetic data support them as superior to plain capsule.

contraindications

Pencina-style trial safety reports across NMN and NR up to 12 weeks at standard doses show no significant adverse-event signal versus placebo (Conze et al. 2019) (Igarashi et al. 2022) (Yi et al. 2023). Two theoretical concerns:

Cancer. NAD+ supports DNA damage response (PARPs) and metabolic flux that proliferating tumours exploit; the theoretical concern is that chronic high-dose NAD+ precursor in someone with an undiagnosed or pre-clinical malignancy could accelerate growth. No human-trial evidence of this exists; the concern is mechanism-led and frequently raised by Brenner and others. Anyone with active cancer or recent cancer history should consult oncology before supplementing.
Methylation load. NMN and NR both deliver nicotinamide downstream, which is methylated to methyl-nicotinamide for excretion. Sustained high doses theoretically draw down SAM/methyl-donor pools. No clinical signal of methylation deficiency in human trials at supplement doses, but very-high-dose niacinamide users sometimes co-supplement methyl donors (TMG, choline) on this rationale.

Pregnancy and breastfeeding are excluded from all human trials; no safety data exist for those populations.

misconceptions

Three common errors in the marketing layer:

"NMN is superior to NR because of Slc12a8." The transporter's existence at scale in humans is contested (Schmidt & Brenner 2019). Head-to-head pharmacokinetic data show both raise blood NAD+; no clinical-endpoint head-to-head has been published.
"Raising NAD+ reverses aging." Trials show NAD+ rising and isolated biomarkers shifting (insulin sensitivity in one trial, gait speed in another, brain metabolism in PD). No trial shows broad reversal of age-related decline, no trial measures mortality, no trial replicates the rodent lifespan signal.
"Cheaper precursors are inferior." Plain nicotinamide raises NAD+ at a small fraction of the cost. Whether NMN/NR's mechanism differs meaningfully from nicotinamide salvage in vivo is the unresolved Brenner-vs-Imai question; commercially, NMN/NR market at 10–100× the per-gram price of nicotinamide.

practicalities

At typical 500 mg/day NMN, monthly cost in the US runs $30–60, $360–720/year. NR (Niagen-branded) sits at a similar monthly cost. Higher-dose regimens (1 g/day) double the spend. Third-party purity testing matters: a 2022 ConsumerLab assay found multiple NMN products under-delivering on label claim, some by >30%. FDA regulatory status is unstable: NMN was excluded from the dietary-supplement definition in November 2022 when MetroBiotech's MIB-626 IND triggered the drug-preclusion clause; the September 2025 FDA reversal restored NMN's supplement status on race-to-market grounds (FDA 2025). Each manufacturer still owes a New Dietary Ingredient Notification. NR has no equivalent regulatory dispute.

stakes / payoff

For the typical 40-something reader optimising healthspan, the felt-experience case is thin. The signals are submaximal exercise markers (ventilatory threshold, 6-minute walk), one contested insulin-sensitivity result, and biomarker shifts that don't map to anything most readers can perceive. The honest payoff is "you may have raised an aging biomarker that animal data suggest matters, and there's a non-trivial chance the daily 500 mg captures a 5–15% endurance/recovery edge if you're training" — not "you'll feel transformed." The stakes case for skipping is symmetric: there's no documented harm at typical doses, and the dollars freed up buy other interventions (creatine, omega-3, strength training equipment) with stronger evidence.

alternatives

Exercise raises NAD+ more than any supplement dose; aerobic and resistance training both increase NAMPT expression in skeletal muscle. Caloric restriction and time-restricted eating raise NAD+ via NAMPT induction. Niacin (nicotinic acid) is the cheapest NAD+ precursor and a regulated drug at lipid-lowering doses; flushing limits use. Nicotinamide (niacinamide) is the cheapest non-flushing precursor — a meta-question across the field is whether anything NMN/NR delivers can't be reached by nicotinamide at one-tenth the price. Creatine is a competing supplement for the same training-recovery and cognitive-resilience claims, with substantially stronger evidence.

The credibility range

Optimist case

NAD+ declines with age in tissue; the mechanism (NAMPT down, CD38 up, PARP overdraw) is well-established (Verdin 2015). Oral NMN/NR reliably restore peripheral NAD+, demonstrably reach the brain at high dose (Brakedal et al. 2022), and have shown clean signals on submaximal exercise capacity (ventilatory threshold in trained runners (Liao et al. 2021), gait speed in older men (Igarashi et al. 2022), 6-minute walk in PAD patients (McDermott et al. 2024) and middle-aged adults (Yi et al. 2023)). Yoshino 2021 — even with its methodological asterisk — showed real muscle insulin-sensitivity improvement in a prediabetic population that would benefit from any margin (Yoshino et al. 2021). Safety profile across 12-month-and-shorter trials is clean. The longevity bet is precisely that: a bet whose downside is bounded (the cost) and whose upside is the rodent lifespan extension translating partially in humans. For someone training hard, over 50, or with metabolic syndrome features, the expected value is positive.

Skeptic case

No human trial has shown a hard clinical endpoint — no mortality reduction, no diabetes prevention, no dementia prevention, no fracture-or-fall reduction. The flagship Yoshino 2021 result rests on a 25-person trial with a 2.4-fold baseline imbalance on the most relevant covariate; Brenner's published critique is hard to dismiss (Yoshino et al. 2021). The Remie 2020 NR trial in healthy obese adults found nothing on insulin sensitivity, mitochondrial function, or metabolic markers despite robust NAD+ elevation (Remie et al. 2020). NR in mild cognitive impairment moved no cognitive endpoint (Orr et al. 2024). The exercise signals are all on submaximal endurance markers (ventilatory threshold, 6MWT) — never VO₂max, never strength, never time-to-exhaustion. The oral bioavailability of intact NMN to peripheral tissues is contested at the mechanism level (Schmidt & Brenner 2019); if Brenner is right, NMN's value proposition collapses to "expensive nicotinamide." The cancer-acceleration mechanistic concern has no human evidence on either side. The supplement market is a multibillion-dollar industry on a base of trials that wouldn't pass a Phase II bar for any clinical claim.

Author's call

Land mildly skeptic-leaning, in the "real biology, contested clinical translation, reasonable bet for some readers" zone. NMN/NR demonstrably raise NAD+; that's the easy part. The clinical-endpoint evidence is a scattered set of small-positive signals (exercise markers, gait, one insulin-sensitivity trial) that haven't converged on a robust, replicated, large-trial finding. The honest framing for the reader: this is a "decent mechanism, weak-to-moderate evidence, low harm, mid cost" supplement — comparable in evidence tier to creatine before 2010 but priced 10×. evidence scores 2 (sparse and contested clinical translation despite robust biomarker effect). controversy scores 3 (active dispute among credible researchers — Brenner vs Imai on mechanism; Yoshino reviewers vs authors on insulin trial; FDA litigation on regulatory status). The article should recommend it as a consider-if-you're-training-hard-and-can-afford-it option, not a default daily.

Stakeholder and incentive map

ChromaDex (NR / Niagen). Holds the patented NR (Niagen) supply, funds the External Research Program, has Charles Brenner as Chief Scientific Adviser. Most peer-reviewed NR trials are CERP-supplied; the conflict is disclosed but real. Commercial interest in NR-superiority framing.
NMN manufacturers (SyncoZymes, Effepharm, dozens of smaller). Highly fragmented post-FDA-reversal market. Commercial interest in the Imai/Slc12a8 mechanism narrative and Yoshino-style insulin trial.
MetroBiotech / MIB-626. Pharmaceutical developer pursuing NMN as a drug. Their IND triggered the 2022 FDA exclusion. Incentive to maintain the drug-vs-supplement distinction.
Academic NAD+ field (Imai, Sinclair, Guarente, Brenner, Verdin). Reputational and IP stakes on opposing sides of the mechanism question. Sinclair's commercial associations (InsideTracker, Tally Health, prior involvement in InVivo) inform his public advocacy for NMN.
FDA. Position has reversed twice (NDI approval May 2022 → exclusion Nov 2022 → reversal Sept 2025). Industry trade groups (Natural Products Association, Council for Responsible Nutrition) drove the litigation.
Wellness and longevity influencers. NMN is a flagship of the Sinclair-aligned longevity media ecosystem; podcast hosts and supplement-stack influencers carry significant promotional weight. Skeptical voices (Brenner, Peter Attia in shifting form, Examine.com) push back with mixed reach.

Population variability

Age. Baseline NAD+ falls with age; the absolute room to restore is larger in older adults. Trials in adults <40 are sparse and tend to show smaller biomarker shifts. The plausible-responder profile skews 50+.
Metabolic status. Yoshino 2021's positive insulin-sensitivity signal was in prediabetic postmenopausal women; the healthy-obese Remie 2020 trial found nothing. The plausible niche for the metabolic claim is overt prediabetes/metabolic syndrome, not generally healthy adults.
Training status. Liao 2021's ventilatory-threshold benefit was in actively-training amateur runners (Liao et al. 2021). The endurance-capacity signal may require a training stimulus to translate.
Disease state. NR's brain-NAD+ delivery and modest clinical signal in Parkinson's (Brakedal et al. 2022) and the 6MWT signal in PAD (McDermott et al. 2024) suggest the largest absolute benefits may be in disease populations rather than wellness use.
Sex. The flagship insulin-sensitivity trial was women-only; the flagship exercise-capacity trials are male-skewed. Sex-stratified effects are under-studied.

Knowledge gaps

No large, long-duration trial (years, >500 participants) on any hard clinical endpoint. The Tzoulis NOPARK trial (52-week MDS-UPDRS in Parkinson's) is the largest disease-specific Phase II in flight.
No head-to-head between NMN, NR, and plain nicotinamide on clinical endpoints. Without this, the cost premium of NMN/NR over niacinamide is unjustified by data.
Replication of Yoshino 2021's insulin-sensitivity result in a well-randomised, larger sample is overdue.
The Slc12a8 transporter question — does intact NMN reach peripheral tissues in humans, or does everything pass through nicotinamide? — is unresolved and load-bearing for the NMN-vs-NR commercial story.
Long-term (>2 year) safety in cancer-survivor populations, given the PARP/sirtuin mechanism's relevance to tumour metabolism, has not been studied.
VO₂max, strength, and time-to-exhaustion endpoints — where positive results would matter most for a "boosts performance" claim — have been negative or unmeasured across trials.

Substance framing. Treated NMN and NR as one entry (NAD+ precursors), not two. Same upstream biology, same downstream claims, same shopping decision. The article and meta scores reflect the substance class. A separate entry for plain nicotinamide is a future candidate — it's the cheaper precursor whose under-study is itself the relevant question.

Brief scope. The input description named cellular energy markers, age-related metabolic measures, exercise capacity, cognition, and evidence strength. All five are covered in the body. Cognition gets a candid null reading (Orr 2024, long-COVID NR) rather than padded coverage — the honest answer is the evidence isn't there yet.

Rating difficulties.

health_short_term at 1 vs 2 was the closest call. Yoshino 2021 plus Igarashi 2022 plus Yi 2023 do show real submaximal metabolic and motor signals, but the felt-experience threshold for a healthy reader is high enough that 1 is the honest score. Bumped above 0 because the trial signal is real, not because most readers will perceive it.
longevity at 1 is a mechanism-bet score with no human hard endpoint. Could argue for 0; landed at 1 because the underlying NAD+ decline / sirtuin biology is well-established enough that a mechanism-led tier-1 is defensible.
controversy at 3 captures three live disputes: Brenner vs Imai on Slc12a8 mechanism, the Yoshino 2021 baseline-imbalance critique, and the FDA 2022→2025 regulatory reversal. Could have justified 4 if the field were more polarised; 3 reads as active-debate-not-trench-warfare.
evidence at 2 reflects: NAD+ elevation replicates everywhere; clinical endpoints are scattered, small, often unreplicated. A 3 would overstate the field; a 1 would understate the mechanism work.

Excluded from the article.

NAD+ IV drips and intranasal sprays — different administration profile, different population (longevity-clinic / wellness-clinic users), warrant their own entry. Flagged as separate-entry candidate.
NMNH and reduced NAD+ precursor variants — early-stage, no human RCTs. Out of scope.
MIB-626 (the MetroBiotech NMN drug) — referenced only in the FDA-regulatory misconception note. Drug-development pathway is a separate story.
Skin / topical NAD+ precursor formulations — different mechanism story, no good RCTs. Could become a lookmaxxing-category candidate.
Resveratrol / NR-resveratrol combinations — NICE trial (McDermott 2024) found no resveratrol additive benefit; resveratrol warrants its own entry rather than being smuggled in here.

Future-link candidates. Creatine (referenced as a stronger-evidence comparator), strength training, time-restricted eating, omega-3 supplementation, sleep duration. All called out in out-of-scope; should be wired to their entries once they exist.

Voice/structure calls. Headingless first section (mechanism) flows directly out of the dek per the spec's editorial pattern. No stakes section: the felt-experience story for skipping is genuinely small (no documented harm, opportunity cost only), and writing a stakes passage would have been padding. Payoff carries the time-scaffolding voice instead, honest about the years-to-decade onset of any longevity bet landing.

Hard decision. Action set to do rather than decide. The decision-tradeoff framing was tempting given the contested evidence, but decide in the spec implies clinician input (TMS, HRT). This is a supplement decision the reader makes themselves; the contested evidence rides on the evidence: 2 + controversy: 3 scores, not on the action verb.