nmn

NMN (beta-Nicotinamide Mononucleotide)

A mononucleotide NAD+ precursor that enters the NAD+ salvage pathway one enzymatic step upstream of NAD+ itself. The premise for its use in aging: NAD+ levels decline progressively with age across tissues — driven by upregulation of the NADase cd38, increased parp activation from accumulating DNA damage, and reduced activity of the rate-limiting salvage enzyme nampt — such that supplementing NMN can restore NAD+, reactivate sirtuin deacylases, and improve metabolic and cellular homeostasis. Imai and Sinclair groups have led most of the mouse work; the Yoshino/Imai group published the first positive human RCT (muscle insulin sensitivity, Science 2021).

Identity

• PubChem CID: 14180
• InChIKey: DAYLJWODMCOQEW-TURQNECASA-N
• ChEMBL: CHEMBL610238
• Molecular formula: C₁₁H₁₅N₂O₈P
• Molecular weight: 334.22 g/mol
• Class: pyridine nucleotide (mononucleotide form of nicotinamide); immediate biosynthetic precursor to NAD+
• Related compound: Nicotinamide Riboside (NR) — one phosphate group fewer; NMN = NR + phosphate group

Mechanism of action

NAD+ biosynthesis route

NMN sits inside the NAD+ salvage cycle, one step from the end product:

Nicotinamide (NAM)
    |  NAMPT (rate-limiting)
    v
NMN  <-- also: NR + NMRK1/2 (NR kinase route)
    |  NMNAT1/2/3 (NMNAT enzymes)
    v
NAD+

The NMNAT-catalyzed reaction is: NMN + ATP → NAD+ + PPi. Three NMNAT isoforms have distinct subcellular localizations: NMNAT1 (nuclear), NMNAT2 (cytosol/Golgi), NMNAT3 (mitochondrial). All three can use NMN as substrate, so intracellular NMN can contribute to NAD+ in multiple compartments ¹.

Cellular uptake — the SLC12A8 controversy

Whether intact NMN enters cells is a subject of active dispute:

• Grozio et al. (2019) identified Slc12a8 (SLC12A8) as a specific NMN transporter. Slc12a8 is most highly expressed in the jejunum and ileum of the small intestine, and also in the pancreas, liver, and white adipose tissue at lower levels. Transport is sodium ion-dependent but chloride-independent; NR is not efficiently transported (IC₅₀ for NR ~77 µM vs ~23 µM for NMN). Whole-body Slc12a8KO mice (CRISPR) showed significant NAD+ deficits in jejunum and ileum during the dark period. Slc12a8 expression is upregulated in the aged ileum (compensatory mechanism). Authors and inventors hold patents on this transporter ². needs-replication
• Competing view: Several groups find that NMN must first be dephosphorylated to NR by CD73 (5’-ectonucleotidase) at the plasma membrane before cellular uptake, then rephosphorylated to NMN intracellularly. Under this model, NMN and NR are functionally equivalent orally. contradictory-evidence

Resolution matters clinically (it affects whether NMN has tissue-distribution advantages over NR), but the downstream NAD+ elevation from oral NMN is not in dispute.

Why NAD+ declines with age

Three mechanisms are additive and together produce a tissue-dependent decline in NAD+ (~30–90% across mammalian tissues; magnitude varies substantially by tissue per Yoshino 2018 review) in old vs young animals ¹:

1. CD38 upregulation — CD38 is an NADase expressed on immune cells; it is chronically upregulated during aging-associated inflammation (inflammaging).
2. PARP hyperactivation — accumulating DNA damage (see genomic-instability) chronically activates PARP1/2, which consumes NAD+ for poly-ADP-ribosylation.
3. NAMPT decline — the rate-limiting enzyme in the salvage pathway shows reduced expression/activity with age in several tissues including muscle.

The age-related decline is tissue-specific in humans — and is not observed in whole blood. Trętowicz et al. 2026 (n=303 across 7 cohorts) found whole-blood NAD+ is stable with age (P=0.24–0.62, R²=0.012–0.051) using a UHPLC-HRMS assay rigorously controlled for pre-analytical artifacts. The negative blood result does NOT contradict tissue-level decline (Janssens 2022 muscle, Massudi 2012 skin remain valid), but invalidates blood NAD+ as an aging biomarker and undermines the simplest version of the NMN-restoration rationale (“supplement to correct age-related blood NAD+ deficit”). NMN supplementation can still raise blood NAD+ above baseline (confirmed across all NMN trials), but the baseline is not “low for age.” contradictory-evidence ³

Downstream effectors of NAD+ restoration

Restored NAD+ activates:

• Sirtuins (SIRT1–7): NAD+-dependent deacylases regulating gene expression, mitochondrial biogenesis (SIRT1/3), DNA repair (SIRT6), and metabolic adaptation.
• 2: paradoxically, adequate NAD+ substrate allows efficient DNA repair rather than the “substrate depletion” feedback loop.
• cd38: NADase activity is substrate-driven; sufficient NAD+ maintains signaling pools.

Effects on aging hallmarks

Hallmark	Proposed effect	Strength of evidence
deregulated-nutrient-sensing	NMN restores SIRT1-mediated nutrient sensing and insulin signaling in muscle	Mouse: strong; human: limited (one RCT)
mitochondrial-dysfunction	NMN increases SIRT3 activity → deacetylates ETC subunits; improves mitochondrial respiration in aged muscle and liver	Mouse: moderate; human: in-progress
epigenetic-alterations	SIRT1/6 deacetylate histones and regulate DNA methylation patterning; NMN-driven NAD+ supports these activities	Mouse: modest data; human: unknown
chronic-inflammation	Reduced SASP-like signaling downstream of SIRT1 activation; indirect via improved cellular homeostasis	Mouse: limited; human: unknown

Preclinical evidence (mouse)

Yoshino 2011 — foundational metabolic study

Yoshino et al. used two mouse models of type 2 diabetes. In the HFD model, wild-type B6 mice (3–6 months) fed a 42%-fat diet were given NMN (500 mg/kg/day i.p.) for 7–10 consecutive days; NAD+ levels were significantly restored in liver and WAT, glucose tolerance was completely normalized in diabetic females, and hepatic insulin sensitivity was improved (elevated liver p-AKT). In the age-induced model, naturally occurring diabetic male mice (15–26 months, screened for fasting glucose >120 mg/dL) received a single dose of NMN (500 mg/kg i.p.), which normalized impaired glucose tolerance (n=11). Lipid profiles were also corrected by NMN. Mechanistically, the authors showed SIRT1 is at least one mediator: SIRT1-specific inhibitor EX527 abrogated NMN’s restoration of Gsta2 expression in TNF-alpha-treated hepatocytes ¹. This was the first study showing that NAD+ restoration via NMN has tissue-metabolic consequences in vivo in mammals.

Dimension	Status
Pathway conserved in humans?	yes — NAD+ salvage is conserved; SIRT1 mechanism conserved
Phenotype conserved in humans?	partial — insulin resistance translates; magnitude unclear
Replicated in humans?	yes (Yoshino 2021 Science, limited scope)

Mills 2016 — long-term NMN administration in aging mice

NMN administered in drinking water at two doses (100 or 300 mg/kg/day) to male C57BL/6N mice beginning at 5 months for 12 months (aged to 17 months) produced broad improvements in age-related physiological decline ⁴. The two doses produced partially different effect profiles:

• Suppressed age-associated body weight gain (both doses; 300 mg/kg/day more pronounced, ~9% reduction)
• Improved energy metabolism: oxygen consumption significantly increased at both doses during the dark period; respiratory exchange ratio decreased (shift toward fat oxidation) in both groups
• Enhanced locomotor activity (100 mg/kg/day showed higher dark-period ambulations; 300 mg/kg/day showed slightly lower ambulations but other benefits)
• Improved eye function (electroretinography: both doses reduced fundus light-colored spots from the rd8 Crb1 mutation; 300 mg/kg/day improved scotopic a-wave)
• Improved bone density (dose-dependent increase in bone mineral density at 12 months)
• Improved insulin sensitivity at 12 months in body-weight-matched comparisons; plasma lipid changes were not significant at all time points
• Preserved myeloid-lymphoid immune composition
• No adverse effects; no difference in mortality rate by log-rank test

NAD+ levels tended to increase in liver at both doses (p=0.06 at 6 months; not significant in skeletal muscle or WAT). Mitochondrial oxidative capacity was significantly enhanced in 300 mg/kg/day skeletal muscle (high-resolution respirometry). Lifespan was not a primary endpoint and no survival difference was detected. needs-replication — One group, one genetic background, male-only.

Dimension	Status
Pathway conserved in humans?	yes
Phenotype conserved in humans?	partial — most endpoints untested in humans
Replicated in humans?	no (Yoshino 2021 tested only insulin sensitivity)

de Picciotto 2016 — vascular aging

NMN (300 mg/kg/day in drinking water) for 8 weeks in old male C57Bl/6 mice (26–28 months, from Charles River / NIA colony; n=13–22 per group) restored maximum endothelium-dependent dilation (EDD) and NO-mediated EDD to levels not significantly different from young controls, reduced aortic pulse wave velocity (aPWV) and elastic modulus, reduced superoxide production and nitrotyrosine abundance, restored aortic collagen I to young-control levels, and increased SIRT1 activity (reduced acetylated p65 NF-kB ratio) ⁵. Aortic NAD+ was not detectably increased in vivo during chronic NMN drinking (authors attribute this to rapid metabolic turnover), but in vitro acute NMN incubation (100 µM, 48h) produced a threefold increase in aortic NAD+. NMN had no significant effect on EDD or arterial stiffness in young mice. Effect was attributed to NMN-driven NAD+ restoration and consequent SIRT1 activation reducing oxidative stress. needs-human-replication

Human evidence

Irie 2020 — first-in-human safety and PK

Single oral doses of NMN (100, 250, 500 mg) were administered sequentially (each separated by ≥1 week) to 10 healthy Japanese men (age 47.9±6.0 years) in a single-arm, non-randomized sequential dose-escalation study. NMN was safe and well-tolerated at all doses; no significant changes in hemodynamics, ophthalmic parameters, or sleep quality scores. Lab values showed small but statistically significant changes (bilirubin increased 51.3%; creatinine, chloride, and glucose decreased slightly), all within normal ranges and likely attributable to 5h fasting. Plasma NMN itself was NOT successfully measured — the authors state that freezing plasma samples before extraction caused NMN degradation, consistent with NMN’s rapid clearance from blood; plasma NMN levels were therefore not reported. Downstream metabolites 2-Py (N-methyl-2-pyridone-5-carboxamide) and 4-Py (N-methyl-4-pyridone-5-carboxamide) increased significantly and dose-dependently in plasma, confirming NMN was absorbed and metabolized to NAD+ and catabolized ⁶. This was primarily a safety study; it was not powered for efficacy endpoints.

Yoshino 2021 (Science) — muscle insulin sensitivity RCT

The landmark human RCT ⁷:

• Design: randomized, double-blind, placebo-controlled, 10-week parallel-group trial (not crossover); NCT03151239
• Population: postmenopausal women with prediabetes who were overweight or obese (BMI 25.3–39.1 kg/m²); n=25 completed (12 placebo, 13 NMN)
• Dose: 250 mg/day oral NMN (supplied by Oriental Yeast Co., Ltd., Tokyo, Japan)
• Primary outcome: skeletal muscle insulin sensitivity assessed as insulin-stimulated glucose disposal rate per kg fat-free mass during hyperinsulinemic-euglycemic clamp
• Results:
  • Muscle insulin sensitivity was 25±7% greater after NMN than before (p<0.01); no significant change in the placebo group
  • Muscle NAD+ itself did not significantly change in either group; however, NMN increased skeletal muscle NAD+ turnover metabolites (N-methyl-nicotinamide, 2-PY, 4-PY) — indirect evidence of increased NAD+ flux rather than elevated steady-state NAD+
  • PBMC NAD+ (basal) increased significantly after NMN
  • Muscle AKT phosphorylation (Ser473 and Thr308) and mTOR phosphorylation (Ser2448) increased significantly during insulin infusion after NMN, not after placebo
  • Global transcriptome of quadriceps (RNA-seq) identified 308 differentially expressed genes during insulin infusion after NMN (vs 5 in placebo); most enriched pathway was PDGF receptor binding; specific genes significantly upregulated were PDGFRβ, CD90 (THY1), CD109, COL1A1, COL5A1, COL6A1 — markers of myogenic pericytes and muscle remodeling (not SIRT1, PIK3CA, or INSR as previously summarized)
  • Mitochondrial oxidative capacity did not change after 10 weeks of NMN or placebo (high-resolution respirometry of quadriceps biopsies)
  • No significant change in body composition, lipids, hepatic insulin sensitivity, adipose tissue insulin sensitivity, or systemic inflammation markers
• Notable limitations: n=25 women only (postmenopausal, prediabetic, overweight/obese); no men; parallel-group design; cannot generalize to broader populations. Authors note no men enrolled because NMN caused greater metabolic benefits in female than male mice in the preclinical work.

This paper is the strongest human evidence for NMN’s aging-relevant efficacy to date, but the effect is muscle-specific and restricted to insulin sensitivity; no systemic metabolic, body composition, or mitochondrial benefits were observed.

2024–2026 NMN trials and meta-analyses (recency refresh, R34)

A wave of moderate-quality NMN RCTs and meta-analyses 2024–2026. Findings below are from PubMed abstracts + Crossref metadata; full PDFs not locally verified for any of these (flagged for future verifier pass). needs-replication

• Igarashi 2024 (GeroScience) — older-adult walking and sleep ⁸: n=60 older adults, NMN 250 mg/day × 12 weeks. Primary endpoint (stepping test) was null at both 4 and 12 weeks. Secondary: 4-m walking time significantly shorter in NMN group at 12 weeks; blood NAD+ and metabolites significantly higher; PSQI sleep quality “Daytime dysfunction” and “Global PSQI” scores improved. UMIN000047871. First positive function/sleep signal in healthy older adults at the Yoshino-2021 dose.
• Tang 2026 (Nutrients) — blood pressure meta-analysis ⁹: 10 RCTs, 11 arms, n=349. NMN reduced resting DBP −2.15 mmHg vs placebo (95% CI −3.68 to −0.61, statistically significant). Resting SBP reduction was not significant overall, but significant −3.94 mmHg in age ≥60 subgroup (95% CI −7.06 to −0.82). Modest BP signal — clinically small but consistent.
• Caldo-Silva 2025 (J Cachexia Sarcopenia Muscle) muscle-function meta-analysis ¹⁰: NMN no significant effect on skeletal muscle index (MD −0.42), handgrip strength, gait speed, or 5CST in older adults across pooled RCTs. Class-level null on muscle-functional endpoints — challenges sarcopenia-prevention positioning.
• Song 2025 (Crit Rev Food Sci Nutr) glucose/lipid meta-analysis ¹¹: 12 RCTs, n=513. NMN raised blood NAD+ but most clinically relevant outcomes (fasting glucose, triglycerides, total cholesterol, LDL-C, HDL-C) were not significantly different between NMN and control. Risk of bias high in 5/12 studies, some concerns in 7/12. Authors conclude likely “exaggeration of benefits” in the field. Strong null signal on metabolic biomarkers.
• Schloesser 2026 (Nat Metab) — three-precursor head-to-head ¹²: n=65 healthy adults, 14-day open-label randomized. NMN and NR comparably raise circulatory NAD+; NAM does not. Ex-vivo human-microbiota fermentation: NR and NMN are converted to nicotinic acid (NA) by gut microbes, and NA (not NMN/NR/NAM directly) is the potent NAD+ booster in whole blood. Authors propose a gut-dependent mechanistic model. Reframes whether the Slc12a8 transporter route (Grozio 2019) is rate-limiting for systemic NAD+ elevation in humans. contradictory-evidence

Pencina 2023 — NAD augmentation in older overweight adults

MIB-626 (a microcrystalline formulation of β-NMN, 1000 mg twice daily = 2000 mg/day total) was given to overweight or obese adults (mean age 61.9±8.6 years, BMI 29.2±3.6 kg/m²) for 28 days in a randomized, double-blind, placebo-controlled, parallel-group study; n=21 MIB-626, n=9 placebo (total n=30 randomized) ¹³:

• Blood NAD+ and metabolites (2-PY, 1-methylnicotinamide, nicotinamide) rose substantially in the MIB-626 group, not in placebo
• Body weight decreased: between-group difference −1.9 kg (95% CI −3.3, −0.5; p=0.008)
• Diastolic blood pressure decreased: between-group difference −7.01 mmHg (95% CI −13.44, −0.59; p=0.034); systolic BP difference was not significant (p=0.051)
• Total cholesterol decreased: difference −26.89 mg/dL (95% CI −44.34, −9.44; p=0.004); LDL difference −18.73 mg/dL (95% CI −31.85, −5.60; p=0.007); non-HDL difference −24.56 mg/dL (p=0.002)
• No significant effect on muscle strength, aerobic capacity (VO₂peak), insulin sensitivity (HOMA-IR), hepatic fat, intra-abdominal fat, or intramuscular NAD+ by 7T MRS
• Safe and well-tolerated; no serious adverse events; no flushing or liver enzyme elevation

Diverges from Yoshino 2021 on insulin sensitivity — populations, doses, formulation (microcrystalline MIB-626 vs standard NMN), durations (28 days vs 10 weeks), and endpoints (HOMA-IR vs hyperinsulinemic clamp) differ substantially. The failure to detect muscle NAD+ increase by 7T MRS is consistent with Yoshino 2021’s finding that muscle steady-state NAD+ did not change. contradictory-evidence

Trial	NCT	Phase	Status	Population	Dose	Primary endpoint
Yoshino 2021 (Imai/WashU)	NCT03151239	—	Completed	Prediabetic postmenopausal women, n=25 (12 placebo, 13 NMN)	250 mg/day, 10 wk	Muscle insulin sensitivity (hyperinsulinemic clamp)
Pencina 2023 (Bhasin/MGH)	—	—	Completed	Overweight/obese adults n=30 (21 active, 9 placebo), mean age 62y	MIB-626 1000 mg bid × 28 days	Safety; NAD augmentation; body composition
NCT04903210	NCT04903210	Phase 4	Unknown	Hypertensive patients	not specified	Not specified
NCT07144527	NCT07144527	Recruiting	Active	Healthy older adults	not specified	Exercise tolerance

long-term-unknown — No human trial has exceeded 12 weeks; long-term safety and efficacy data in humans are absent.

Comparison with Nicotinamide Riboside (NR)

NMN and NR are both NAD+ precursors that enter the salvage pathway immediately upstream of NAD+; they share a biosynthetic relationship (NMN = NR + phosphate). Key practical differences:

Feature	NMN	NR
Entry point	NMN → NAD+ (1 step via NMNAT)	NR → NMN → NAD+ (2 steps)
Uptake route	SLC12A8 (if Grozio 2019 correct) or dephosphorylation to NR first	ENTs / equilibrative nucleoside transporters
MW	334 g/mol	255 g/mol
Oral bioavailability	Not precisely quantified in humans	Moderate; ~25% as NR, rest as metabolites
Human evidence	Yoshino 2021 (insulin sensitivity); Pencina 2023	Multiple RCTs (Trammell 2016 et al.)
Cost	Higher (more complex synthesis)	Lower

Neither compound has been shown superior in a head-to-head RCT with aging endpoints. contradictory-evidence

Pharmacokinetics

• Oral NMN in humans is rapidly absorbed; plasma NMN itself was not successfully quantified in Irie 2020 (freezing of plasma samples caused NMN degradation before extraction — a known pre-analytical issue). Downstream metabolites 2-PY and 4-PY peaked dose-dependently at ~300 min post-dose. Rodent studies (Mills 2016) show plasma NMN increases steeply within 2–5 min of oral gavage and is cleared within 15 min, with tissue NAD+ rising over 60 min. Whether this rapid rodent kinetic translates to humans is unconfirmed ⁶. The archival record marks publication year as 2019 (submission) vs 2020 (journal print) — treat as Irie et al. 2020.
• Tissue half-life of the downstream NAD+ elevation is longer and tissue-dependent; liver and muscle show sustained NAD+ elevation for ≥12 h in rodent studies.
• Rapid plasma clearance does not necessarily mean low efficacy — NAD+ is sequestered intracellularly once synthesized.
• No reported CYP interactions; metabolized to nicotinamide and downstream catabolites (MeNAM, 2-PY, 4-PY).

dose-response-unclear — The relationship between oral dose, plasma exposure, tissue NAD+ elevation, and functional outcomes in humans is not established. The 250 mg/day dose in Yoshino 2021 showed muscle-specific insulin sensitivity improvement; the 2000 mg/day dose in Pencina 2023 raised blood NAD+ but did not improve insulin sensitivity (different formulation, population, and duration). Neither study detected a significant increase in steady-state muscle NAD+ by direct measurement. Dose-response curves in humans are absent.

Safety

• Human trials to date (up to 2000 mg/day, up to 10 weeks) report no significant adverse effects compared to placebo ⁶⁷¹³. The longest trial is Yoshino 2021 (10 weeks at 250 mg/day); Pencina 2023 used 2000 mg/day for 28 days.
• Theoretical concern: NAD+ catabolite nicotinamide (NAM) accumulates at high NMN doses; high NAM inhibits sirtuins (negative feedback). This has been documented in rodents at supraphysiologic doses. At supplement-range doses, NAM accumulation appears modest.
• Flushing: Unlike nicotinic acid (niacin), NMN does not cause prostaglandin-mediated flushing. Nicotinamide and NMN do not activate the GPR109A receptor responsible for niacin flush.
• Long-term safety data are absent. long-term-unknown

Limitations and gaps

• Limited human evidence: Only two completed RCTs; both small (n<30 randomized per trial), short (≤10 weeks), and narrow in population (postmenopausal prediabetic women; overweight/obese middle-aged and older adults). Effect sizes are moderate and not consistent across endpoints or doses.
• No lifespan data in humans — only physiological surrogates. The preclinical lifespan studies (Mills 2016) used one genetic background and did not report a survival curve as a primary outcome. needs-human-replication
• Uptake mechanism unresolved — the SLC12A8 transporter story (Grozio 2019) remains contested; several labs have not reproduced it. This matters for tissue targeting and comparison with NR. contradictory-evidence
• NMN vs NR head-to-head: No randomized human trial compares the two directly on aging outcomes.
• Sex- and age-specific effects: Yoshino 2021 enrolled women only; generalizability to men and non-prediabetic populations unknown.
• Optimal dosing unknown: Studies use 250 mg to 2000 mg/day. Saturation of NMNAT, tissue-specific uptake limits, and NAM feedback inhibition are not characterized in humans. dose-response-unclear
• Industry funding: Several trials and the Grozio 2019 paper involve investigators with commercial interests in NMN products. Conflict-of-interest should be noted during verification.

Cross-references

• Upstream biology: sirtuin, nampt, nad-biosynthesis (stub), cd38 (stub)
• Intervention class: nad-precursors (R23d class page) — NMN/NR/niacin/NAM class context
• Related compound: nr (parallel batch seed)
• Competing pathways: mtor, ampk, insulin-igf1
• Hallmarks targeted: deregulated-nutrient-sensing, mitochondrial-dysfunction, epigenetic-alterations
• Context pages: caloric-restriction, deregulated-nutrient-sensing
• Framework overlay: hallmarks-of-aging, sens-damage-categories

Footnotes

Footnotes

1. yoshino-2011-nmn-diabetes-mice · n=4–15/group (varies by experiment) · in-vivo (mouse) · p<0.01–0.001 · model: C57BL/6 mice, HFD-induced T2D (3–6 months) + naturally occurring aged diabetic males (15–26 months) · doi:10.1016/j.cmet.2011.08.014 ↩ ↩² ↩³

2. grozio-2019-slc12a8-nmn-transporter · in-vivo + in-vitro · model: mouse small intestine; HEK293 cells · doi:10.1038/s42255-018-0009-4 · contested — see SLC12A8 controversy section ↩

3. tretowicz-2026-blood-nad-stable-aging · doi:10.1038/s42255-026-01537-5 · observational + intervention pooled · n=303 across 7 cohorts · whole-blood NAD+ stable with age (all 6 age comparisons null: P 0.24–0.62, R² 0.012–0.051) and stable across exercise/protein interventions in older adults; positive control: 5-month NR supplementation in twin-pair cohort raises whole-blood NAD+ as expected · UHPLC-HRMS validated for pre-analytical artifacts · Nature Metabolism 2026-05-14 · Trętowicz MM et al. (Houtkooper laboratory) · archive: not yet in archive · verified: false (Results/Discussion paywalled; Abstract + Reporting Summary + 8 source-data XLSX directly verified) ↩

4. mills-2016-nmn-aging-mice · n=9–15/group · in-vivo (mouse) · p<0.05 · model: C57BL/6N male mice, 12-month NMN supplementation (100 or 300 mg/kg/day in water) from 5 months of age · doi:10.1016/j.cmet.2016.09.013 ↩

5. depicciotto-2016-nmn-vascular-aging · n=13–22/group · in-vivo (mouse) · p<0.05 · model: old male C57Bl/6 mice (26–28 months), 300 mg/kg/day NMN in drinking water × 8 weeks · doi:10.1111/acel.12461 ↩

6. irie-2020-nmn-first-human-pk · n=10 · single-arm non-randomized sequential dose-escalation · model: healthy Japanese men (age 40–60), single oral doses 100/250/500 mg · doi:10.1507/endocrj.ej19-0313 ↩ ↩² ↩³

7. yoshino-2021-nmn-insulin-sensitivity · n=25 (12 placebo, 13 NMN) · rct (randomized double-blind placebo-controlled parallel-group) · p<0.01 (insulin sensitivity) · model: postmenopausal prediabetic overweight/obese women, 250 mg/day × 10 weeks · NCT03151239 · doi:10.1126/science.abe9985 ↩ ↩²

8. igarashi-2024-nmn-older-adults-walking-sleep · n=60 · rct (randomized double-blind placebo-controlled parallel-group) · primary endpoint (stepping test) null; 4-m walking time shorter in NMN group at 12 weeks · model: older adults · NMN 250 mg/day × 12 weeks · UMIN000047871 · GeroScience 2024 · doi:10.1007/s11357-024-01204-1 · archive: pending download · integrated from abstract pending PDF verification ↩

9. tang-2026-nmn-blood-pressure-meta-analysis · meta-analysis · 10 RCTs / 11 arms / n=349 · NMN reduced DBP −2.15 mmHg (95% CI −3.68 to −0.61); SBP −3.94 mmHg in age ≥60 subgroup · Nutrients 2026 · doi:10.3390/nu18060890 · archive: not in archive · integrated from abstract pending PDF verification ↩

10. caldosilva-2025-nmn-nr-muscle-meta-analysis · meta-analysis · NMN: no significant effect on SMI, HGS, gait speed, 5CST in older adults · Journal of Cachexia, Sarcopenia and Muscle 2025 · doi:10.1002/jcsm.13799 · archive: pending · integrated from abstract pending PDF verification ↩

11. song-2025-nmn-glucose-lipid-meta-analysis · meta-analysis · 12 RCTs, n=513 · NMN raised blood NAD+ but no significant change in fasting glucose, TG, TC, LDL-C, HDL-C vs control · Critical Reviews in Food Science and Nutrition 2025 · doi:10.1080/10408398.2024.2387324 · archive: pending · integrated from abstract pending PDF verification — high risk of bias in 5/12 included studies ↩

12. schloesser-2026-nad-precursors-comparison · doi:10.1038/s42255-025-01421-8 · n=65 · rct (open-label, placebo-controlled, three-arm comparison NR vs NMN vs NAM) · NR and NMN comparably raise circulatory NAD+; NAM does not · ex-vivo: gut microbes convert NR/NMN to nicotinic acid; NA is the potent NAD+ booster · 14 days · Nature Metabolism 2026 · archive: pending download · integrated from abstract pending PDF verification — substantively reframes mechanism debate ↩

13. pencina-2023-nmn-nad-augmentation · n=30 (21 MIB-626, 9 placebo) · rct (randomized double-blind placebo-controlled parallel-group, 2:1 allocation) · model: overweight/obese adults mean age 62, MIB-626 1000 mg bid × 28 days · doi:10.1210/clinem/dgad027 ↩ ↩²