Nicotinamide riboside (NR) is the nucleoside form of vitamin B3, a nicotinamide molecule joined to a ribose sugar. Within this catalogue it stands apart. Almost everything else here is a synthetic peptide whose evidence base is preclinical or limited to a handful of small studies. NR is the opposite: an established nutrient that has been through dozens of randomised human trials and holds food-ingredient regulatory status in both the US and the EU. This page is built around that difference, and around what follows from it: when a compound is studied long and carefully, you get to see what it does and, just as clearly, what it does not do.

2004: a vitamin identified in retrospect

Nicotinamide riboside had been known as a chemical since the 1940s, identified as a growth factor for Haemophilus bacteria. But it only became a vitamin in 2004, when Bieganowski and Brenner described the nicotinamide riboside kinase genes (NRK1 and NRK2) in yeast and humans. The cell turned out to have a dedicated enzyme that phosphorylates NR to NMN, a distinct route to NAD⁺ independent of the Preiss-Handler pathway.[1]

This is an unusual sequence in biochemistry: rather than fitting a new substance into a known pathway, a long-known substance was found to have a pathway of its own. That discovery is what turned NR from a chemical curiosity into an industry, and it is why NR is discussed separately from nicotinamide and nicotinic acid even though all three are formally vitamin B3.

NRK1 as the bottleneck

Twelve years later, Ratajczak and colleagues showed that NRK1 is not merely involved: it is necessary and rate-limiting for the use of exogenous NR. In cells lacking NRK1, external NR stops raising NAD⁺ altogether.[2]

The same work produced a second result that later became central to the NR-versus-NMN argument: NMN does not enter the cell as NMN. It is dephosphorylated outside the cell to NR, taken up as NR, and only rephosphorylated to NMN inside. At the cellular level NMN is essentially an NR prodrug, and both compounds converge on the same enzyme.

2025: the route turned out to be considerably messier

The most interesting development in NR's story is recent, and it cuts against the tidy picture above. Yaku and colleagues (Science Advances, 2025) traced labelled NR and NMN in vivo and found that after oral dosing only a small fraction is absorbed directly from the small intestine. Most is broken down to nicotinamide, deamidated by the gut microbiota to nicotinic acid, absorbed, and used for hepatic NAD⁺ synthesis via the classical Preiss-Handler route. A similar pattern appeared even after intravenous dosing: the compounds were rapidly degraded to nicotinamide and secreted into bile, entering an enterohepatic cycle.[12]

If this holds up, it reorganises the field. NR's supposed advantage as a molecule that enters cells intact largely dissolves: most oral NR appears to work through the microbiome and liver rather than through NRK1. This does not make NR ineffective, it raises blood NAD⁺ reliably, but it relocates the explanation of why. Worth keeping in mind when reading any marketing copy about "direct cellular uptake".

NR versus NMN: what the difference actually rests on

The practical answer comes from the pharmacokinetic work of Trammell and colleagues (2016), the first study to demonstrate oral NR bioavailability in humans and a dose-dependent rise in whole-blood NAD⁺.[3] The head-to-head evidence base for NR versus NMN is far thinner than vendors of either compound suggest, and much of the comparison rests on indirect arguments, molecular weight, presence or absence of a transporter, rather than direct randomised comparisons within a single study.

The review by Yoshino, Baur and Imai (Cell Metabolism, 2018) remains the most measured account of where the data end and extrapolation begins.[15] The metabolism of the coenzyme itself, biosynthesis, redox chemistry, consuming enzymes, is covered separately in the NAD⁺ monograph; here we deliberately stay at the level of a single precursor molecule.

The human data: NAD⁺ goes up, and mostly that is where it ends

This is the most important section on the page. NR is one of the few compounds in this category with enough randomised human data to draw conclusions from. The conclusion is inconvenient.

What replicates reliably: NR raises NAD⁺. Martens and colleagues (2018), in a double-blind crossover trial (500 mg twice daily, 6 weeks), reported roughly a 60% increase in blood-cell NAD⁺.[4] Elhassan and colleagues (2019) confirmed NR reaches aged human skeletal muscle, raising the muscle NAD⁺ metabolome.[6]

What did not replicate:

  • Dollerup and colleagues (2018): 2000 mg/day for 12 weeks in obese, insulin-resistant men, no improvement in insulin sensitivity.[5]
  • The same cohort, separate analysis (J Physiol, 2020): no change in mitochondrial respiration, content or morphology in skeletal muscle. The authors state plainly that the data do not support the hypothesis.[7]
  • Remie and colleagues (2020): 1000 mg/day for 6 weeks, crossover, muscle NAD⁺ synthesis markers rose, acetylcarnitine metabolism shifted, and there was a small gain in fat-free mass, but no effect on insulin sensitivity, mitochondrial function, hepatic or intramyocellular lipid, blood pressure, or inflammatory markers.[8]
  • Elhassan (2019): despite the raised muscle NAD⁺ metabolome, mitochondrial bioenergetics were unchanged.

What sits in between: the longest trial, Lapatto and colleagues (Science Advances, 2023), gave NR to BMI-discordant monozygotic twin pairs at escalating doses up to 1000 mg/day for five months. Here NR did increase muscle mitochondrial number, improve myoblast differentiation, and alter gut microbiota composition and DNA methylation.[9] But adiposity and metabolic health did not improve. Longer exposure produced real biological change, just not the kind that converts into a clinical outcome.

One correction worth making explicitly: the reductions in blood pressure and aortic stiffness in the Martens trial were exploratory secondary outcomes and did not reach statistical significance in the overall group. The authors themselves stressed that no inferences can be drawn from the subgroup analysis. This is frequently miscited.

Neurodegeneration: the one area with a distinct signal

NR research in Parkinson's disease has developed differently from the metabolic line. In the phase I NADPARK trial (Brakedal et al., Cell Metabolism, 2022), 30 newly diagnosed, treatment-naive patients took 1000 mg NR for 30 days. NR was well tolerated and significantly, though variably, raised cerebral NAD measured by ³¹P magnetic resonance spectroscopy. In those participants whose brain NAD did rise, cerebral metabolism shifted on FDG-PET, and this was associated with mild clinical improvement.[10] This was phase I; it was not powered for efficacy.

It was followed by the high-dose safety trial NR-SAFE (Berven et al., 2023)[11] and then NOPARK, a randomised, double-blind, placebo-controlled phase III trial in 400 patients with early Parkinson's disease (500 mg twice daily, 52 weeks, primary endpoint MDS-UPDRS). Recruitment and dosing completed in 2025. As of writing, the NOPARK results have not been published. This is the largest test of the NAD⁺ hypothesis in neurology, and until it reports, any claim of neuroprotection by NR in humans is premature.

Safety and regulatory status

This is where NR diverges most sharply from the rest of the catalogue. Long-term rat toxicology (Conze et al., 2019) and the clinical record have not surfaced signals limiting use at studied doses.[13] In 2019 the EFSA (European Food Safety Authority) panel assessed nicotinamide riboside chloride as a novel food and considered it safe; the European Commission authorised it for market, and in the US it holds GRAS (generally recognized as safe) status and has been through NDI notification.[14]

This needs reading correctly. Regulatory clearance as a food ingredient means confirmed safety at defined intakes, it does not mean demonstrated efficacy against ageing, metabolic disease, or any condition. The two questions are judged by different standards, and NR passes the first comfortably while the second remains open.

What remains unknown

  • Why raising NAD⁺ in blood and muscle so consistently fails to translate into functional outcomes. Candidate explanations: NAD⁺ is not limiting in healthy tissue; the wrong pool rises (cytosolic rather than mitochondrial); or the compartment that matters is never reached.
  • How far the 2025 enterohepatic-circulation finding transfers to humans, that work was done in animal models.
  • Whether populations exist in which NAD⁺ really is depleted and the effect would differ. Neurodegeneration is the most plausible candidate, which is precisely why NOPARK matters.
  • Whether duration matters. Five months in the Lapatto trial produced biological changes absent from 6-12 week studies.

Reagent status

Supplied as nicotinamide riboside chloride (NR-Cl), a crystalline powder; the compound is hygroscopic and less stable than nicotinamide, so it needs protection from moisture and light. Product page: Nicotinamide Riboside (NR); adjacent NAD⁺-pathway compounds sit in the NAD+ / metabolic longevity category, including the NNMT inhibitor 5-Amino-1MQ, which approaches the same pathway from the other side, not adding precursor but slowing nicotinamide clearance.

Labelled RUO: research use only. Not for human or animal consumption; not a medicine, food or cosmetic. The doses quoted above are those used in published research protocols, cited for accuracy in describing the evidence base; they are not and cannot be a recommendation for use.