NMN vs NR · The 2026 NAD+ Precursor Decision Tree (Evidence-Based)

60-second answer to the three things readers actually want to know

Which precursor has more human RCTs? NMN now has roughly thirteen published randomized controlled trials in humans, three independent confirmations of functional endpoints in older adults, a dose-finding multi-centre trial, the strongest single athletic-endpoint signal in the precursor class, and a twelve-trial meta-analysis. NR has roughly eight to ten randomized trials, the only positive cognitive-endpoint trial (mild cognitive impairment), the only large peripheral artery disease randomized trial in the class, and the longest published safety dose at 3000 mg/day for four weeks. Both are mature enough to defend evidence-based structure/function framing; neither is mature enough for a disease-treatment claim.

Does one raise blood NAD+ more than the other? Both reliably double blood NAD+ at typical chronic doses (NMN 250 to 1000 mg/day, NR 1000 mg/day). No head-to-head RCT has compared them on this endpoint or on any other.

Is NMN superior because of the SLC12A8 transporter? No. SLC12A8 is a contested hypothesis from Grozio 2019 (PMID 31131364) that Schmidt and Brenner 2020 (PMID 32694648) reported they could not reproduce. The dispute is unresolved; marketing claims of NMN superiority based on SLC12A8 outrun the data.

This article walks through the decision in five dimensions: (1) absorption pathway and the SLC12A8 controversy, (2) randomized trial portfolios, (3) blood NAD+ elevation, (4) regulatory and cost reality across four jurisdictions, and (5) side-effect data and long-term safety gaps.

Dimension 1 · Absorption pathway and the SLC12A8 controversy

How orally administered NMN gets from the gut into circulating blood, and then into cells, is more contested than the textbook salvage-pathway diagram suggests. The prevailing model in 2026: oral NMN is largely dephosphorylated in the small intestine by ectoenzymes (CD73 and 5'-nucleotidase) to yield nicotinamide riboside (NR); NR is then absorbed into enterocytes, rephosphorylated to NMN by nicotinamide riboside kinase 1 (NRK1), and only then converted by NMNAT enzymes to NAD+. Under this model, the in-vivo endpoints of oral NMN and oral NR converge considerably, because the molecules collapse onto the same intracellular intermediate.

In 2019 Grozio and colleagues from the Imai laboratory proposed a different model in Nature Metabolism (PMID 31131364): the transporter SLC12A8 as a mammalian-specific NMN transporter, allowing intact NMN to cross plasma membranes directly without dephosphorylation. This implied an absorptive route NR could not share, and it became the mechanistic basis for many marketing claims that NMN was inherently superior to NR.

In 2020 Schmidt and Brenner (Iowa) published a rebuttal in the same journal (PMID 32694648) reporting they could not reproduce the SLC12A8-mediated NMN transport in their own cell models. Grozio and colleagues subsequently published a reply reaffirming their original observations while acknowledging some details required further replication. The current state of the field: the SLC12A8 hypothesis has not been widely independently replicated, and the prevailing mechanistic model remains the dephosphorylation-to-NR route.

For consumer-facing writing the honest framings are: NMN is a direct precursor to NAD+ in the salvage pathway; oral NMN raises blood NAD+ approximately two-fold in healthy adults across multiple RCTs; multiple absorption mechanisms have been proposed, including a direct transport route via SLC12A8 first reported in 2019, and that mechanism remains under debate. The framing that does not survive critical appraisal is "NMN is more efficient than NR because of the SLC12A8 transporter."

What changes if SLC12A8 is real

Even if the SLC12A8 absorption route were fully independently replicated tomorrow, it would not by itself establish NMN clinical superiority over NR. Clinical superiority requires a head-to-head randomized trial in humans on a defined endpoint — no such trial exists in 2026. The mechanism debate is interesting science; it is not a substitute for outcome data.

Dimension 2 · Randomized trial portfolios — what each precursor has actually proven

NMN human RCT portfolio (13 trials)

The NMN human literature has grown rapidly since 2020 and now spans biochemical primary endpoints (blood NAD+ and safety), functional secondary endpoints in older adults and athletes, and a small cluster on cardiovascular surrogate endpoints that has not yet reached statistical significance on its primary outcome.

• Yoshino M et al. 2021 in Science (PMID 33888596). Double-blind, placebo-controlled RCT in 25 postmenopausal women with prediabetes, 250 mg/day NMN for 10 weeks. Statistically significant increase in skeletal muscle insulin signalling (AKT and mTOR phosphorylation) and muscle NAD+ metabolites. The first high-impact placebo-controlled human demonstration of tissue-level NMN effects.
• Yi L et al. 2023 in GeroScience (PMID 36482258). Multi-centre, double-blind, placebo-controlled dose-finding RCT in 80 healthy middle-aged adults at 300/600/900 mg/day NMN for 60 days. Clear dose-dependent rise in blood NAD+ across all NMN arms; all doses well tolerated. The closest the literature has come to a formal dose-response study on the primary biomarker.
• Okabe K et al. 2022 in Frontiers in Nutrition (PMID 35479740). 250 mg/day for 12 weeks in 30 healthy adults; blood NAD+ approximately doubled by week 4 versus placebo and maintained through week 12.
• Pencina KM et al. 2023 in J Clin Endocrinol Metab (PMID 36740954). MIB-626 (crystalline polymorph of beta-NMN) 1000 mg twice daily for 14 days in 32 middle-aged and older overweight adults; approximate doubling of blood NAD+ in the active arm; no serious adverse events.
• Liao B et al. 2021 in J Int Soc Sports Nutr (PMID 34238308). 48 amateur runners at 300/600/1200 mg/day NMN for six weeks. Dose-dependent increases in VO2 max, ventilatory threshold and skeletal-muscle oxygen utilisation. The strongest single-trial signal for an athletic-performance endpoint to date in the precursor class.
• Igarashi M et al. 2022 in npj Aging (PMID 35927255). 42 healthy older Japanese men at 250 mg/day for 12 weeks. Statistically significant improvements in walking speed (p = 0.033) and left-hand grip strength (p = 0.019). Standard geriatric functional measures.
• Kim M et al. 2022 in Nutrients (PMID 35215405). 108 older Japanese adults at 250 mg/day or placebo with further randomisation to morning versus afternoon dosing for 12 weeks. The afternoon-dosing arm showed statistically significant improvements in five-times sit-to-stand, sleep quality and self-reported fatigue; the morning-dosing arm did not. One of the cleaner chronobiology signals in the supplement literature.
• Katayoshi T et al. 2023 in Scientific Reports (PMID 36797393). 250 mg/day for 12 weeks in 36 healthy adults; trend toward reduction in brachial-ankle pulse wave velocity (baPWV, a surrogate of arterial stiffness) with rises in NAD-related metabolites. The baPWV trend did not reach statistical significance. Best summarised as a hypothesis-generating cardiovascular signal.
• Fukamizu Y et al. 2022 in Scientific Reports (PMID 36002548). 20 healthy adults at 1250 mg/day for four weeks — the highest dose-duration combination validated in a published RCT to date. No clinically significant abnormalities in liver, kidney, electrolyte or complete-blood-count panels.
• Zhang J, Poon ET, Wong SH 2025 in Crit Rev Food Sci Nutr (PMID 39116016). Meta-analysis pooling 12 NMN RCTs (n = 513). Statistically significant overall elevation of blood NAD+ in NMN versus placebo; functional and metabolic endpoints remain heterogeneous.

NR human RCT portfolio (8 to 10 trials)

The NR human RCT history is older and has been published predominantly in high-impact journals from Nature Communications to GeroScience.

• Trammell SAJ et al. 2016 in Nature Communications (PMID 27721479). The landmark pharmacokinetic demonstration: NR at 100/300/1000 mg single doses produces dose-dependent rises in blood NAD+ metabolome in healthy adults. This established NR as orally bioavailable in humans and set the dose ranges that subsequent trials would use.
• Martens CR et al. 2018 in Nature Communications (PMID 29599478). 24 healthy middle-aged and older adults at 1000 mg/day NR for six to eight weeks (crossover design); approximately 60 percent elevation of blood NAD+. A non-significant trend toward systolic blood pressure reduction in the sub-group with elevated baseline blood pressure. The most frequently cited NR efficacy reference in healthy aging.
• Vreones M et al. 2023 in Aging Cell (PMID 36515353). 22 older adults at 500 mg twice daily for six weeks; blood NAD+ rose approximately 2.4-fold; biomarkers of neurodegenerative pathology showed downward trends.
• Orr ME et al. 2024 in GeroScience (PMID 37994989). 20 adults with mild cognitive impairment at 250 to 500 mg/day for 10 weeks; significant elevation of whole-blood NAD+ and improvement on selected cognitive sub-domains. The first positive randomized trial on a cognitive endpoint in the NAD+ precursor class.
• McDermott MM et al. 2024 (NICE) in Nature Communications (PMID 38871717). 90 adults with peripheral artery disease at 1000 mg/day for six months; improvement in the six-minute walk test (secondary endpoint). The largest NR cardiovascular trial published to date.
• NR-SAFE 2023 in Nature Communications (PMID 38016950). 20 adults with Parkinson disease at 3000 mg/day for four weeks; no serious adverse events; substantial blood NAD+ elevation. Establishes the highest published short-term NR safety dose.
• Conze D et al. 2019 in Scientific Reports (PMID 31278280). 140 overweight adults at 100 to 1000 mg/day NR for eight weeks; safety profile clean across the dose range; clean tolerability at the upper bound.

How to read the two portfolios side by side

NMN trials cluster around the 250 mg/day chronic dose with positive functional endpoints in older adults (three independent RCTs converging) and an athletic-endpoint signal (Liao 2021). NR trials cluster around the 1000 mg/day chronic dose with the broadest indication portfolio (mild cognitive impairment, peripheral artery disease, Parkinson safety, healthy aging). Neither portfolio supports a head-to-head superiority verdict — they cover overlapping but non-identical use cases.

Dimension 3 · Blood NAD+ elevation — the most replicated biochemical finding

Both precursors reliably elevate blood NAD+ in healthy adults at their respective standard chronic doses. The honest summary across trials:

• NMN at 250 to 1000 mg/day raises blood NAD+ to approximately two-fold baseline in multiple RCTs (Yoshino 2021, Okabe 2022, Yi 2023, Pencina 2023). The Zhang 2025 meta-analysis (PMID 39116016) confirmed the elevation across 12 trials.
• NR at 1000 mg/day raises blood NAD+ approximately 60 percent in healthy middle-aged adults (Martens 2018) and approximately 2.4-fold in older adults (Vreones 2023). The Trammell 2016 PK study established the dose-response curve from 100 to 1000 mg.

What blood NAD+ elevation does not establish: that any specific clinical endpoint will respond, or that one precursor outperforms the other on a clinical outcome. Blood NAD+ is the biochemical primary endpoint most reliably modified by either compound, but it is a surrogate. Translation to clinical endpoints depends on the indication.

Dimension 4 · Regulatory and cost reality across four jurisdictions

NMN and NR sit in sharply different regulatory positions as of 2026. This dimension matters more for purchase decisions than most consumers realize.

United States — FDA

NMN: Permitted as a dietary supplement ingredient following the 29 September 2025 withdrawal of the 2022 IND exclusion determination. Subject to NDI notification under 21 CFR 190.6, the standard FDA disclaimer ("These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease."), 21 CFR Part 111 cGMP, and structure-function claim limits. The phrase "FDA approved" is inaccurate for any dietary supplement and should not be used.

NR: Permitted as a dietary supplement ingredient via GRAS notifications (Niagen and other NR-containing branded raw materials) cleared by FDA. Same DSHEA / 21 CFR 190.6 / 21 CFR Part 111 framework. No equivalent IND-exclusion history.

European Union — EFSA Novel Food

NMN: Not currently permitted as a food or food supplement. Regulated under Regulation (EU) 2015/2283 on Novel Foods. Six applications are in evaluation as of early 2026, the most advanced being EffePharm's Uthever (public consultation completed February 2025). The earliest plausible authorisation is late 2026 to early 2027 for the leading application.

NR: Authorised as a Novel Food at specified maximum daily intake levels via EFSA opinions on nicotinamide riboside chloride (ChromaDex Niagen authorization 2017 onwards). NR is therefore legally available in EU member states as a food supplement within the authorised intake bounds.

Brazil — ANVISA

NMN: Prohibited across the full supply chain under Resolução-RE No. 1,139 of 7 April 2022. Covers commercialisation, distribution, manufacture, importation, advertising and use, and mandated recall of any NMN-containing products already on the market. Covers cross-border e-commerce as well as domestic supply. NMN is not available in Brazil through any legal channel.

NR: Subject to case-by-case ANVISA review as a functional food ingredient. No full-chain prohibition equivalent to NMN. Commercial pathway exists but specific products require individual approval.

Australia — TGA

NMN: Added to the Therapeutic Goods (Permissible Ingredients) Determination No. 4 of 2025 on 10 December 2025. Maximum recommended daily intake 500 mg/day (stricter than the 1250 mg/day RCT-validated safety ceiling); adults only; ARTG registration required. SyncoZymes holds two-year market exclusivity through 10 December 2027. The first formal regulatory approval of NMN as a therapeutic-product ingredient anywhere in the world.

NR: Available as a listed therapeutic-goods ingredient under earlier TGA determinations; no equivalent recent first-permission event.

Cost reality (educational only, not a procurement recommendation)

As of early 2026 typical retail consumer-direct pricing per gram of active sits in the same order of magnitude for NMN and NR, with NMN tending toward the slightly higher end for pharmaceutical-grade material and NR tending toward the slightly lower end for established branded raw materials. The cost-per-gram conversation is dwarfed in importance by the regulatory and evidence considerations above.

Dimension 5 · Side-effect data and the long-term safety gap

Across all published RCTs reviewed in this brief, neither NMN nor NR has produced serious adverse events. The most commonly reported adverse events are mild gastrointestinal disturbances and occasional headaches at rates statistically indistinguishable from placebo.

• NMN short-term safety ceiling: 1250 mg/day for four weeks validated in a dedicated safety RCT (Fukamizu 2022 PMID 36002548). The 2000 mg/day total in the Pencina 2023 MIB-626 trial for 14 days also ran without serious adverse events but used a specific crystalline polymorph.
• NR short-term safety ceiling: 3000 mg/day for four weeks validated in the NR-SAFE Parkinson trial (PMID 38016950). The Conze 2019 study (PMID 31278280) confirmed clean safety up to 1000 mg/day for eight weeks in 140 overweight adults.

The long-term safety gap that affects both compounds equally

No RCT of either compound has followed participants beyond six months. The phrase "safe for long-term use" is not supported by trial data for either precursor. This is the single largest evidence gap in the precursor class and applies symmetrically to both NMN and NR.

Contraindications and precautionary populations (apply to both)

• Pregnancy and lactation. No human safety data for either; both should be avoided.
• Children and adolescents (under 18). No human safety data.
• Concurrent glucose-lowering medication. Because both precursors may modulate insulin signalling (most directly demonstrated for NMN in Yoshino 2021), patients on hypoglycaemic agents should consult their physician.
• PARP inhibitors and oncology medication. Because NAD+ is a substrate for PARP and NAD+ availability has been studied in tumour metabolism, oncology patients should not self-initiate either precursor outside clinical guidance.

Goal and lifestyle integration · which precursor fits which user

For the Longevity Stack reader

The longevity-stack reader wants the most-evidence-developed NAD+ precursor available. The two precursors are close to a tie at this level: NMN has the denser functional-endpoint base in older adults (Igarashi 2022, Kim 2022, Morita 2024) and the strongest athletic signal (Liao 2021); NR has the only positive cognitive trial (Orr 2024) and the only large peripheral vascular trial (NICE McDermott 2024). Either is defensible in a longevity-stack discussion; both are inappropriate for lifespan-extension claims, because no trial of either compound has measured lifespan.

For the Cognitive Support reader

NR has the only published positive randomized cognitive-endpoint trial in the precursor class — Vreones 2024 (PMID 37994989) in mild cognitive impairment showed significant blood NAD+ elevation and improvement on selected cognitive sub-domains. NMN cognitive evidence remains mechanistic only — biologically plausible through NAD+ in brain energy metabolism and sirtuin-mediated neuroprotection, but no human cognitive-endpoint RCT on NMN has been published. For consumers prioritizing cognitive support specifically, NR has the evidence advantage at present.

For the Athletic Performance reader

NMN has the single strongest published athletic-performance signal in the precursor class via Liao 2021 (PMID 34238308): 48 amateur runners, six weeks, 300/600/1200 mg/day, dose-dependent improvements in VO2 max, ventilatory threshold and skeletal-muscle oxygen utilisation. NR athletic-endpoint evidence is sparser; the closest equivalent is the NICE 2024 NR trial (PMID 38871717) reporting six-minute walk test improvement in peripheral artery disease, which is mobility rather than performance. For competitive or recreational athletes, NMN has the more directly relevant trial.

For the Heart Health reader

NR has the larger cardiovascular evidence base via the NICE peripheral artery disease trial (McDermott 2024 PMID 38871717) and the Martens 2018 systolic blood pressure trend in healthy aging. NMN cardiovascular evidence is limited to Katayoshi 2023 baPWV arterial stiffness trend that did not reach statistical significance. Neither precursor has measured hard cardiovascular endpoints (MACE, mortality, stroke). For heart-health framing both remain hypothesis-generating; NR has the slight evidence edge.

For the Senior 60+ reader

This is the lifestyle context where the precursor evidence base is densest, and NMN has the more developed portfolio: three independent RCTs (Igarashi 2022, Kim 2022, Morita 2024) targeting older adult cohorts with functional improvements (walking speed, grip strength, five-times sit-to-stand, sleep quality, fatigue). The chronobiology finding — afternoon dosing outperformed morning dosing in older adults in both Kim 2022 and Morita 2024 — is one of the more reproducible timing signals in the supplement literature. NR has the NICE peripheral artery disease trial in this age cohort and the Vreones 2023 healthy-aging trial. For senior functional support specifically, NMN at 250 mg/day in the afternoon has the densest evidence anchor.

For the Athletic Performance Lifestyle reader

See the athletic-performance Goal section above. For lifestyle athletes training four to six times per week, NMN at 600 to 900 mg/day on training days has the most directly relevant trial; NR remains a mechanistically reasonable but evidentially thinner alternative on this endpoint.

For the GLP-1 Companion reader

Neither precursor has been studied in a randomized trial in a GLP-1 receptor agonist co-medication cohort. The mechanistic rationale — muscle preservation during rapid GLP-1-mediated weight loss via NAD+-supported mitochondrial energetics — applies symmetrically to both. This is currently an extrapolation, not an evidence base, and should be framed as such.

Transparent disclosure · why PMID precision matters in NAD+ writing

One of the recurring quality issues in NAD+ precursor consumer content is PMID drift — identifiers being attached to the wrong trial because three or four studies with overlapping topics get collapsed onto a single shared PMID in informal indexing. This article uses identifiers verified directly against PubMed esearch and esummary on 26 May 2026; readers can audit every citation by clicking through to PubMed.

The most common drift cases observed in third-party NAD+ writing and worth flagging for any reader:

• Pencina KM 2023 MIB-626 trial. Correct identifier is PMID 36740954 in J Clin Endocrinol Metab. Some third-party NMN literature has cited the Yi 2023 PMID 36482258 in this slot, which actually points to the dose-finding GeroScience trial.
• Kim M 2022 chronobiology trial. Correct identifier is PMID 35215405 in Nutrients. Again sometimes conflated with the Yi 2023 identifier.
• Katayoshi T 2023 baPWV trial. Correct identifier is PMID 36797393 in Scientific Reports. Some informal indexing has stored a PMC identifier in the PMID column.
• Zhang J, Poon ET, Wong SH 2025 NMN meta-analysis. Correct attribution is PMID 39116016 in Crit Rev Food Sci Nutr. Some pre-publication indexing attributed this paper to "Zhong O 2024."
• Song Q et al. 2023 NMN antiaging review. Correct attribution is PMID 37619764 in Advances in Nutrition. Some informal indexing attributed this paper to "Nadeeshani H."

None of these are scientific errors in the underlying trials; they are citation-hygiene issues that propagate through consumer content when authors copy-paste from prior summaries rather than verifying against PubMed. The reason we surface them here is educational: when an article tells you a trial used a certain dose for a certain duration with a certain endpoint, the PMID is the auditable anchor. If the PMID does not resolve to the claimed trial, the cited evidence is not actually in support of the claim.

Statements that go beyond the evidence in any of these jurisdictions

Bottom line · how to actually choose between NMN and NR in 2026

1. If your goal is geriatric functional performance (walking speed, grip, sit-to-stand, sleep) — NMN at 250 mg/day taken in the afternoon has the densest replicated evidence base (Igarashi 2022, Kim 2022, Morita 2024).
2. If your goal is cognitive support in early decline — NR at 250 to 500 mg/day has the only positive randomized cognitive-endpoint trial in the precursor class (Orr 2024 PMID 37994989).
3. If your goal is amateur or recreational athletic performance — NMN at 600 to 900 mg/day on training days has the strongest single-trial signal (Liao 2021 PMID 34238308).
4. If your goal is peripheral vascular or heart-health mobility — NR at 1000 mg/day in statin-treated patients has the largest randomized trial (NICE McDermott 2024 PMID 38871717).
5. If you live in Brazil — NMN is not legally available; NR remains case-by-case.
6. If you live in the EU — NR is the only precursor legally available as a food supplement; NMN Novel Food authorisation is pending.
7. If you live in Australia — NMN now has explicit permitted-ingredient status at 500 mg/day maximum under TGA Determination No. 4 of 2025; NR remains available under earlier listings.
8. If you live in the United States — both are legal dietary supplement ingredients post 29 September 2025; the regulatory tiebreaker disappears and the choice reverts to the evidence-by-goal matching above.

No head-to-head human RCT has been published, and consumers should treat any "NMN beats NR" or "NR beats NMN" framing as marketing rather than evidence. The honest 2026 verdict is precursor-by-use-case rather than precursor-by-precursor.

Frequently asked questions

Goals this evidence informs

• Longevity Stack — NMN and NR are the two most-studied human NAD+ precursors; the longevity-stack discussion lives or dies on these two compounds.
• Cognitive Support — NR has the only positive randomized cognitive-endpoint trial to date (Orr 2024 mild cognitive impairment); NMN cognitive evidence remains mechanistic.
• Athletic Performance — Liao 2021 (PMID 34238308) NMN dose-finding RCT in 48 amateur runners is the single strongest precursor-class athletic-endpoint signal.
• Heart Health — NR NICE trial in peripheral artery disease (McDermott 2024 PMID 38871717); NMN baPWV arterial stiffness trend (Katayoshi 2023 PMID 36797393); evidence remains preliminary on both sides.

Lifestyles this evidence informs

• Senior 60+ — Three independent NMN RCTs in older adults (Igarashi 2022 · Kim 2022 · Morita 2024) plus the NR-MCI Orr 2024 trial converge on the 60+ cohort as the precursor evidence base.
• Athletic Performance — NMN dose-finding athletic RCT (Liao 2021) plus mechanistic NR support; lifestyle athletes are the second-densest evidence cohort after older adults.
• GLP-1 Companion — Mechanistic adjacency via mitochondrial energetics and muscle preservation during GLP-1-mediated weight loss; no dedicated RCT cohort yet.

Sibling ingredient pages

• NAD+ Cluster Hub
• NMN (β-Nicotinamide Mononucleotide)
• NADH (Reduced NAD)

Related Evidence Articles (peer)

• 25 Years of NAD+ Clinical Evidence — the 25-year chronology this decision tree builds on; cross-read for historical RCT context (niacin → NMN/NR).
• Reduced NADH vs NMNH — NAD redox-form adjacent decision; closes the redox-axis precursor framing.

These peer evidence articles cover the broader NAD-axis evidence base behind the precursor choice.

References (21 PubMed-verified citations)

1. Yoshino M et al. — NMN Increases Muscle Insulin Sensitivity in Prediabetic Women (PMID 33888596). Science 2021.
2. Yi L et al. — NMN Dose-Finding RCT 300/600/900 mg/day n=80 (PMID 36482258). GeroScience 2023.
3. Okabe K et al. — 250 mg NMN 12 weeks blood NAD+ doubling (PMID 35479740). Frontiers in Nutrition 2022.
4. Pencina KM et al. — MIB-626 NMN 1g BID 14 days NAD+ doubling (PMID 36740954). J Clin Endocrinol Metab 2023.
5. Liao B et al. — NMN Dose-Finding in 48 Amateur Runners (PMID 34238308). J Int Soc Sports Nutr 2021.
6. Igarashi M et al. — NMN Walking Speed + Grip Strength in Older Japanese Men (PMID 35927255). npj Aging 2022.
7. Kim M et al. — NMN Afternoon-Dosing Chronobiology RCT (PMID 35215405). Nutrients 2022.
8. Katayoshi T et al. — NMN baPWV Arterial Stiffness Trend (PMID 36797393). Scientific Reports 2023.
9. Fukamizu Y et al. — NMN 1250 mg/day Safety Ceiling RCT (PMID 36002548). Scientific Reports 2022.
10. Zhang J, Poon ET, Wong SH — NMN Meta-Analysis 12 RCTs n=513 (PMID 39116016). Crit Rev Food Sci Nutr 2025.
11. Grozio A et al. — SLC12A8 Proposed as NMN Transporter (PMID 31131364). Nature Metabolism 2019.
12. Schmidt MS, Brenner C — SLC12A8 Non-Reproducibility Rebuttal (PMID 32694648). Nature Metabolism 2020.
13. Song Q et al. — NMN Anti-Aging Human Trials Update (PMID 37619764). Advances in Nutrition 2023.
14. Irie J et al. — NMN Single-Dose Safety in Japanese Men (PMID 31685720). Endocrine Journal 2020.
15. Trammell SAJ et al. — Nicotinamide Riboside Is Orally Bioavailable in Mice and Humans (PMID 27721479). Nature Communications 2016.
16. Martens CR et al. — Healthy Aging NR 1000 mg/day RCT (PMID 29599478). Nature Communications 2018.
17. Vreones M et al. — NR Lowers Neurodegenerative Biomarkers in Older Adults (PMID 36515353). Aging Cell 2023.
18. Orr ME et al. — NR in Mild Cognitive Impairment RCT (PMID 37994989). GeroScience 2024.
19. McDermott MM et al. — NICE Trial NR in Peripheral Artery Disease (PMID 38871717). Nature Communications 2024.
20. NR-SAFE: High-Dose NR Safety in Parkinson Disease (PMID 38016950). Nature Communications 2023.
21. Conze D et al. — NR Safety and Tolerability 8-Week RCT (PMID 31278280). Scientific Reports 2019.

All identifiers verified via PubMed E-utilities (esearch + esummary) on 26 May 2026. This article does not constitute medical advice and is not a substitute for clinical judgement. Regulatory status varies by jurisdiction; see Dimension 4 for the four primary markets in scope (FDA, EFSA, ANVISA, TGA).