What it is
NAD+ is the oxidized form of nicotinamide adenine dinucleotide (C21H27N7O14P2, MW 663.43 Da, CAS 53-84-9, PubChem CID 5893), a dinucleotide formed by pyrophosphate linkage of adenosine monophosphate to nicotinamide ribonucleotide. Every mammalian cell synthesizes NAD+ de novo from tryptophan via the kynurenine pathway and — supplying ~85% of the steady-state pool (Frontiers in Cell & Dev Biol 2024) — recycles it through the salvage pathway, in which nicotinamide (NAM) liberated by NAD+-consuming enzymes is converted to nicotinamide mononucleotide (NMN) by NAMPT and then to NAD+ by NMNAT1–3. Intracellular concentrations are 200–500 μM in most tissues, with 40–70% of the pool in mitochondria. The redox cycle (NAD+ + 2e⁻ + H⁺ ⇌ NADH) powers glycolysis, the TCA cycle, β-oxidation, and oxidative phosphorylation. The non-redox roles — sirtuin deacylation, PARP poly-ADP-ribosylation, CD38/CD157 hydrolysis to cyclic ADP-ribose — consume NAD+ stoichiometrically, which is why NAD+ is a flux metabolite, not an inert cofactor. Age-related decline in tissue NAD+ (Massudi 2012 in human skin; Zhu 2015 by 31P-MRS in brain; Camacho-Pereira 2016 implicating CD38) is well-documented and motivates the entire precursor-supplementation program. For grading purposes, 'NAD+' covers three distinct therapeutic classes: intravenous/subcutaneous NAD+ itself (longevity-clinic protocol, no FDA approval, minimal controlled human efficacy), oral NR (dietary supplement, best-characterized human PK), and oral NMN (drug-excluded by FDA in October 2022, Yoshino 2021 Science data). The evidence varies drastically by which one is being asked about.
How it works
- 01
Sirtuin cofactor function (SIRT1–7)
NAD+ is the obligate co-substrate of the seven mammalian sirtuin NAD+-dependent deacylases (Imai 2000 first characterized SIRT1 as the NAD+-dependent enzyme). Each deacetylation consumes one NAD+ and releases nicotinamide and 2'-O-acetyl-ADP-ribose. SIRT1 activity scales with NAD+/NADH ratio; its IC50 for inhibition by released nicotinamide is ~175 μM (Guarente 2016, npj Aging Mech Dis). SIRT1 deacetylates PGC-1α, p53, FOXO3a, and NF-κB p65, which is the molecular basis of the longevity / caloric-restriction-mimetic narrative. SIRT3 (mitochondrial) deacetylates MnSOD, OPA1, and multiple TCA-cycle enzymes; SIRT6 (nuclear) regulates telomere maintenance and base-excision repair. This mechanism is well-characterized in biochemistry — but sirtuin activation is necessary, not sufficient, for any clinical outcome.
- 02
PARP and CD38 NAD+-consumption economy
PARP1 (Alano 2010, J Neurosci) and CD38 (Camacho-Pereira 2016, Cell Metab; Chini 2020) are the two dominant NAD+-consuming enzymes in aging tissue. CD38 expression rises with age in multiple tissues and is the proposed molecular explanation for age-related NAD+ decline. CD38 inhibitors (78c, Chini 2018) raise tissue NAD+ in mice without exogenous precursors — an alternative therapeutic axis to precursor supplementation that has not yet been tested in humans. PARP overactivation under chronic DNA damage can deplete cellular NAD+ and trigger parthanatos; this is the molecular rationale for NAD+ precursor therapy in PARP-inhibitor-induced toxicity (EBioMedicine 2025) and in chemotherapy adjuncts.
- 03
Salvage pathway — NAMPT is rate-limiting
The salvage pathway converts nicotinamide (released by sirtuin, PARP, and CD38 reactions) back into NAD+ in two steps: NAMPT (nicotinamide phosphoribosyltransferase) adds PRPP to yield NMN; NMNAT1/2/3 add AMP to yield NAD+. Revollo 2004 (J Biol Chem) and the Imai lab established NAMPT as rate-limiting. NAMPT expression oscillates circadianly under CLOCK/BMAL1 control (Ramsey 2009, Science; Nakahata 2009, Science) — so plasma and hepatic NAD+ oscillate with ~24-h periodicity, which is the mechanistic rationale for timed NAD+-precursor dosing. NAMPT declines with age in most tissues (Frontiers Mol Biosci 2024 review), which is part of why older adults have the biggest NAD+ rise from precursor supplementation.
- 04
Oral NR pharmacokinetics — why the biomarker grade is earned
Trammell 2016 (Nat Commun) administered single oral NR doses of 100, 300, and 1,000 mg to 12 healthy adults, showed dose-proportional whole-blood NAD+ elevation, and identified a distinct NR-specific metabolic signature (the NAD+ metabolome) that distinguished NR from equivalent doses of niacin or niacinamide. Airhart 2017 (PLoS ONE) showed NR 1 g/day × 8 days doubled whole-blood NAD+ in healthy middle-aged volunteers. Conze 2019 (Sci Rep) confirmed chronic safety and continued NAD+ elevation over 8 weeks at 1 g/day. Multiple labs, multiple cohorts, consistent direction and magnitude — this is the strongest part of the NAD+ literature.
- 05
Intravenous NAD+ pharmacokinetics — why the IV grade is low
Grant 2019 (Front Aging Neurosci) infused 750 mg NAD+ over 6 hours in three healthy adults and sampled plasma at 0, 2, 4, 6, and 8 hours. Plasma NAD+ rose, but parallel plasma nicotinamide rose far more sharply — most of the infused NAD+ was being hydrolyzed in the circulation (plausibly by cell-surface CD38 and plasma NAD+-glycohydrolases) and entering tissues as nicotinamide rather than as intact NAD+. By Grant's estimates only ~3% of the infused dose remained as NAD+ in the systemic pool at 2 hours. This undermines the 'direct NAD+ repletion' story for commercial clinic protocols, which typically run 500–1,000 mg over 2–4 hours (faster than Grant's 6-hour infusion, which was selected because faster infusions cause severe 'chest pressure' side effects — a known tolerability issue of clinic IV NAD+).
- 06
What the mechanism does not explain
The mechanism is strong enough to motivate the program but does not explain the biomarker-to-outcome gap. Oral NR reliably raises whole-blood and skeletal-muscle NAD+ by 1.5–2.7× (Martens 2018, Elhassan 2019), but does not improve insulin sensitivity (Dollerup 2018), VO2max (Stocks 2021), or grip strength / SPPB (Dolopikou 2020) in placebo-controlled trials. The hypothesis that NAD+ rise in whole blood corresponds to NAD+ rise in the tissues that actually perform the claimed function (e.g., hepatocyte nuclear NAD+ for insulin signaling, mitochondrial NAD+ for oxidative phosphorylation) is not directly tested in most trials. Compartmentalization — the mitochondrial SLC25A51 NAD+ transporter, identified only in 2020 (Luongo 2020, Nature; Kory 2020, Nature) — is likely why increasing systemic NAD+ does not automatically raise the functional pool in the tissue of interest.