# NAD+ Research: Mechanism, Human Trials, and the Evidence Order Book > NAD+ research from primary sources: the redox and signaling mechanism, the CD38 decline, NMN and NR human trials, and where controlled evidence runs thin. Cited study by study. Mechanism, the age-related decline, the human precursor trials, and the honest gaps — each logged to the study that measured it. ## The short read This page lays out what the NAD+ literature has actually shown. NAD+ (a fuel-handling helper molecule every cell uses) does two jobs: it carries electrons to make energy, and it is a consumed substrate for repair and signaling enzymes [5]. Levels fall with age [2][5]. The reliable human finding is that the precursors NMN and NR raise blood NAD+ [3][4]; the unreliable part is whether that changes health outcomes in people [21]. Below, each finding is tagged with its species, dose, and route so an oral-precursor trial is never mistaken for "taking NAD+" or for the IV route. ## Mechanism: a redox carrier and a consumed signaling substrate NAD+ runs two distinct jobs. As a redox couple (NAD+/NADH — redox is the chemistry that shuttles electrons to release energy), it cycles between oxidized and reduced forms to carry electrons through glycolysis, the TCA cycle, and the mitochondrial electron transport chain, driving ATP synthesis [5]. As a signaling substrate, it is consumed — not merely recycled — by three enzyme families: sirtuins (SIRT1-7, NAD+-dependent maintenance enzymes that regulate metabolism and DNA repair), PARP1 (a DNA-damage repair enzyme), and CD38/CD157 (NAD-consuming ectoenzymes) [5][12]. The body makes NAD+ through several routes. The dominant one in mammals is the salvage pathway, which recycles nicotinamide back into NAD+ via the rate-limiting enzyme NAMPT [11]. That salvage flux is not optional for tissue function: in mouse myoblasts and intact skeletal muscle, knocking down NAMPT lowered NAD+ and impaired maximal respiratory and oxidative-phosphorylation capacity, while the precursor NR restored NAD+ and raised respiratory capacity [10]. NR enters through a separate, NAMPT-independent route using the NRK1/NRK2 kinases — a structural basis worked out from crystal structures of human NRK1 [12]. De novo synthesis from tryptophan and the Preiss-Handler pathway from niacin supply the rest [5]. ## Why tissue NAD+ falls with age The age-related decline in NAD+ is not only reduced production; it is increased consumption. CD38 is the principal NAD+-consuming enzyme whose activity rises with age, and CD38-knockout mice are protected against the age-related fall in tissue NAD+, preserving SIRT3 activity and mitochondrial function [2]. A foundational review frames the competing NAD-consuming enzymes — sirtuins, PARPs, and CD38/CD157 — as drawing down a shared NAD+ pool, and positions restoring NAD+ as a candidate strategy against age-related disease across yeast, worm, mouse, and human models [5]. The decline is visible in human tissue. In muscle biopsies from 119 older men across three populations, sarcopenia tracked with a transcriptional signature of mitochondrial dysfunction, fewer mitochondria, and low NAD+ through perturbed NAD+ biosynthesis and salvage [9]. ## Human precursor trials: blood NAD+ rises, dose-dependently The most reproducible human result is that oral precursors raise blood NAD+ in proportion to dose. NR at 100, 300, and 1000 mg/day for 8 weeks raised whole-blood NAD+ by 22%, 51%, and 142% respectively in healthy overweight adults, with no flushing and no significant adverse-event difference from placebo at any dose; NR did not elevate LDL cholesterol or disrupt one-carbon metabolism [4]. NMN at 300, 600, or 900 mg/day for 60 days raised blood NAD+ at days 30 and 60 across all groups versus placebo (p ≤ 0.001) in a multicenter, double-blind, randomized trial, identified 600 mg/day as the optimal dose, and reported improved walking distance with no safety issues [3]. Functional metabolic data exist but are narrower. Ten weeks of NMN at 250 mg/day raised muscle insulin sensitivity in prediabetic, postmenopausal women on a hyperinsulinemic-euglycemic clamp, while leaving body composition and HbA1c unchanged [1]. ## NAD+ injections and IV infusions in the literature Intravenous NAD+ carries the weakest controlled evidence of any route. The most-cited human data is a pilot: five consecutive daily infusions of 750 mg/day improved cognitive performance on 6 of 8 neuropsychological tests, versus 2 of 8 for saline controls [7]. That is a pilot design, not a definitive trial. Infused NAD+ is also cleared rapidly — pharmacokinetic data from the same IV-NAD+ human work found near-complete plasma removal within roughly the first two hours of infusion [7]. IV NAD+ also carries documented quality risk: a compounded injectable NAD+ product was subject to an FDA Class I recall for elevated bacterial endotoxin. Compounded injectables are unapproved and not FDA-reviewed. The cognition and neurodegeneration evidence sits on [NAD+ and the brain](/nad-and-the-brain). ## Recent reviews: the honest ceiling on human efficacy The 2024-2026 literature converges on a cautious read. A 2025 Nature Metabolism review of NAD+ precursor supplementation in human ageing concluded that trials have shown limited efficacy for hard endpoints, that age-related NAD+ decline has been consistently observed only in a limited number of human studies, and that tissue-specific NAD+ data remain sparse — underscoring the need for more clinical study rather than rodent extrapolation [21]. The sharpest positive signals come from rare DNA-repair diseases. A review of NR in Werner syndrome, ataxia-telangiectasia, and Cockayne syndrome reported a 52-week Werner-syndrome RCT (n=9) with a roughly 140% plasma NAD+ rise and improved arterial stiffness, HDL counts, and kidney function, plus improved ataxia scores in ataxia-telangiectasia trials, with no moderate or severe adverse events [13]. In a 24-week long-COVID RCT (n=58), NR significantly raised NAD+ but did not show a significant group-level cognitive difference versus placebo, though within-group improvements in fatigue, sleep, depression, and executive function were observed in NR-treated participants [14]. --- The NAD+ literature read like an order book — the coenzyme and its precursors NMN and NR each listed apart, the blood-NAD+ the trials actually moved quoted to source, and the rapidly-cleared IV route flagged as the weak-data listing; no clinic behind this desk and nothing here ordered, infused, dispensed, or sold.