NAD+ research: mechanism on the floor, randomized trials at the front.

The short version

This page collects the NAD+ research in layers. At the bottom is the mechanism: NAD+ (a fuel-handling helper molecule every cell uses) shuttles electrons to make energy and feeds the cell's repair-and-signaling enzymes. In the middle are the randomized trials of oral precursors (building blocks the body converts into NAD+ — NMN and NR), which reliably raise the NAD+ measured in blood. At the front is the most-replicated result: those blood-NAD+ increases. Furthest back, in honest haze, is the unproven part — whether raising blood NAD+ changes hard health outcomes in people. We label each layer for what it is.

How NAD+ works: redox carrier and signaling substrate

NAD+ does two jobs. As a redox carrier it cycles between an oxidized form (NAD+) and a reduced form (NADH), accepting electrons in glycolysis and the TCA cycle and donating them in the mitochondrial electron transport chain to drive ATP synthesis ^[5]. The ratio between the two forms — the NAD+/NADH redox couple — is itself a metabolic signal, reporting how oxidized or reduced the cell's energy state is at any moment.

As a signaling substrate NAD+ is consumed — not just recycled — by three enzyme families: sirtuins (NAD+-dependent enzymes that regulate metabolism and DNA repair), PARPs (DNA-damage repair enzymes, chiefly PARP1) and CD38 (an NAD+-consuming ectoenzyme) ^[5]. Each of these spends NAD+ rather than merely borrowing it, which couples the size of the NAD+ pool directly to the cell's capacity for DNA repair, deacylation signaling and immune regulation. When DNA damage activates PARP1, for instance, it can draw down NAD+ sharply.

Because those enzymes spend NAD+, the cell must constantly remake it. The dominant route in mammals is the salvage pathway, which recycles nicotinamide back into NAD+ through the rate-limiting enzyme NAMPT ^[5]. NAMPT expression is induced by exercise and follows a circadian rhythm, tying NAD+ supply to activity and time of day. In cultured and intact mouse skeletal muscle, knocking down NAMPT lowered NAD+ and impaired maximal respiratory capacity, while the precursor nicotinamide riboside restored NAD+ and respiration ^[15] — direct evidence that NAD+ salvage is required for mitochondrial function.

Why NAD+ falls with age

Tissue NAD+ declines across the lifespan, and a principal cause is rising consumption. CD38 activity increases with age and inflammation; in mice, CD38 deletion preserved NAD+ levels and SIRT3 activity and protected mitochondrial and metabolic health into older age ^[2]. The mechanism connects NAD+ loss to inflammaging: inflammatory factors secreted by senescent cells — the senescence-associated secretory phenotype — activate CD38-bearing macrophages, which then consume more NAD+, a feed-forward loop the CD38 study helped establish ^[2].

A foundational review frames the age-related fall as competition among the NAD+-consuming enzymes — sirtuins, PARPs and CD38 — for a shrinking pool, and positions restoring NAD+ as a candidate strategy against age-related decline ^[5]. One proposed downstream consequence is pseudohypoxia, a disruption of nuclear-mitochondrial communication that low NAD+ can produce even when oxygen is adequate, linking the coenzyme's decline to mitochondrial dysfunction ^[5].

Human observational data align with the mechanism. In muscle biopsies from 119 older men across three populations, sarcopenia tracked with a transcriptional signature of mitochondrial bioenergetic dysfunction, fewer mitochondria and low NAD+ through perturbed biosynthesis and salvage ^[13]. That the same NAD+-depletion signature appears across ethnicities strengthens the case that the decline is a general feature of human aging, not a quirk of one population.

NAD supplement research: what the trials actually measured

The NAD supplement literature is, in practice, a precursor literature — because oral NAD+ itself is poorly absorbed, the controlled human trials studied NMN, NR and nicotinamide. Their most consistent, best-replicated endpoint is whole-blood NAD+, the standard pharmacodynamic readout (direct tissue NAD+ sampling in humans is invasive and rare).

That readout moves reliably. NR at 100/300/1000 mg/day for 8 weeks raised whole-blood NAD+ by 22%/51%/142% ^[4]; a single 1000 mg NR dose raised NAD+ roughly 2.7-fold over 24 hours ^[6]; 1000 mg/day NR for 6 weeks raised it about 60% ^[7]. NMN at 300-900 mg/day raised blood NAD+ significantly versus placebo across a 60-day multicenter trial ^[3]. The result generalizes across precursor, dose and study population: raising blood NAD+ with an oral precursor is a settled pharmacodynamic fact. What it produces downstream is the open question of the next section.

Reported outcomes in the research literature

Beyond raising NAD+, several trials measured functional NAD+ benefits, with mixed and generally preliminary results. Ten weeks of NMN at 250 mg/day improved muscle insulin sensitivity (by hyperinsulinemic-euglycemic clamp) in prediabetic, postmenopausal women, without changing body composition or HbA1c ^[1]. Twelve weeks of NMN at 250 mg/day raised ventilatory thresholds during incremental treadmill testing in amateur runners, interpreted as improved skeletal-muscle oxygen utilization ^[10]. The 60-day NMN trial reported improved walking distance and quality-of-life scores alongside the blood-NAD+ rise ^[3]. NR's randomized crossover showed a trend toward reduced aortic stiffness and lower systolic blood pressure, not a definitive effect ^[7].

Preclinical signals are broader. In a mouse mitochondrial-myopathy model, NR restored muscle and liver NAD+ and was associated with greater exercise capacity ^[12]; in reproductively aged mice, an AI radiomic analysis classified 60% of NMN-treated aged oocytes as having a "young" morphology ^[11]. These are animal findings; no human trial establishes such outcomes.

Injectable and IV NAD+: the published pharmacokinetic and tolerability data

Injectable NAD+ — the marketed NAD injection — bypasses oral absorption but carries the weakest controlled evidence of any route. A human pharmacokinetic pilot using a continuous intravenous infusion found NAD+ is extensively metabolized extracellularly and rapidly cleared from plasma, with near-complete plasma removal within roughly the first two hours of infusion ^[9].

Reported wellness-clinic infusion protocols run roughly 250-1000 mg per session over several hours. Tolerability data note that infusions can cause chest tightness, abdominal discomfort, flushing and nausea if run too fast, with symptoms resolving on completion. A distinct safety concern is product quality: a compounded injectable NAD+ has been subject to an FDA Class I recall for elevated bacterial endotoxin. IV NAD+ is therefore presented here as an unapproved compounded therapy with documented quality risks and thin efficacy data — never as an approved treatment.

What controlled studies say about IV NAD+ therapy

Controlled evidence for IV NAD+ therapy is limited and mostly pilot or retrospective. The clearest controlled datum is pharmacokinetic, not clinical: infused NAD+ is rapidly cleared and heavily metabolized before cells take it up ^[9]. Historical and small pilot reports describe cognitive or addiction-related changes, but rigorous randomized trials are lacking, so the route's clinical claims rest on far weaker ground than the oral precursor RCTs.

Where the evidence stands

Read in layers, the NAD+ literature has a firm base and an unsettled top. The base is mechanism and pharmacodynamics: NAD+ is a central redox and signaling coenzyme ^[5], it declines with age partly through CD38 ^[2], and oral precursors raise whole-blood NAD+ dose-dependently and reproducibly across trials ^[4][3][6][7]. That much is well established.

The unsettled part is clinical translation. Some trials report functional gains — muscle insulin sensitivity ^[1], aerobic capacity ^[10], physical performance ^[8] — but the effects are modest, not always replicated, and concentrated in specific populations. The most current authoritative synthesis, a 2025 Nature Metabolism review, concluded that human efficacy data remain limited, that age-related NAD+ decline has been documented consistently in only a limited number of human studies, and that tissue-specific NAD+ data are sparse ^[16]. The honest reading: raising blood NAD+ is proven; what raising it does for human health is still being measured, and rodent results should not be read as human conclusions.