A 65-person randomized trial compared them over 14 days. NR and NMN raised circulating NAD+ by a comparable amount. The FDA reinstated NMN in December 2025. The last real gap is the EU, where NR is capped at 300 mg/day and NMN is unauthorised.
A randomized trial of 65 healthy adults found that NR and NMN raised blood NAD+, a substance measured in the blood, by similar amounts after 14 days. Nicotinamide did not raise it. The study suggested that gut bacteria may convert both NR and NMN into nicotinic acid, but this does not prove longer-term health benefits. The trial was short, its doses were not publicly stated, and it measured no functional health outcome. Evidence for wider benefits remains weak: two major benefit claims rely on only 13 people each, while the largest metabolic trials for both products were null. In the United States, NMN was reinstated as a lawful supplement ingredient in December 2025; in the EU, NMN is unauthorised and NR is capped at 300 mg per day. Median listed prices were about $4.07 per gram for NMN and $3.67 per gram for NR.
What this means for you: Neither product clearly wins: both raised blood NAD+ similarly in one small, short trial, but health benefits remain unproven. For buyers, legality and price matter more than claims that one is biologically superior.
early evidenceUntil January 2026 every NMN against NR article rested on indirect comparison. Trials used different doses, different populations, different assays and different tissues, so nobody could say which molecule raised NAD+ more. That gap is now closed. A randomized, open-label, placebo-controlled trial gave 65 healthy adults nicotinamide, NR or NMN for 14 days and measured circulating NAD+ and the whole-blood NAD+ metabolome[1].
The headline is one sentence. Fourteen days of NR and NMN, but not nicotinamide, comparably increased circulatory NAD+. Two molecules with two decades of competing marketing behind them produced the same measured result in the same trial, in the same assay, on the same timeline.
The mechanistic half of the trial is more interesting than the headline. Only nicotinamide acutely and transiently changed the whole-blood NAD+ metabolome, which fits its known fast absorption. When human gut microbiota were incubated with each compound, NR and NMN both gave rise to nicotinic acid, and both specifically increased microbial growth and metabolism. Then the trial tested each molecule directly on whole blood outside the body. Nicotinic acid was a potent NAD+ booster. NMN, NR and nicotinamide were not.
Put those three results together and the proposed model is uncomfortable for both product categories. Neither NR nor NMN appears to raise blood NAD+ by acting on blood cells directly. Both appear to work partly because gut bacteria convert them into nicotinic acid, which then feeds the Preiss-Handler arm of NAD+ synthesis. Nicotinic acid is niacin. It costs cents per gram and has been sold as a vitamin since the 1940s.
Two caveats belong here. The trial ran 14 days, which is long enough to move blood NAD+ and far too short for any functional endpoint. And the doses used are not stated on the public trial record, so the result cannot be read as a milligram-for-milligram equivalence. What it does establish is that at the doses tested, the molecule choice did not decide the NAD+ outcome.
Swipe the chart sideways
The oldest sales line in this category is that NMN is one step from NAD+ and NR is two. The biochemistry says the opposite, and it has said so since 2016.
A kinase-tracing study using genetic gain and loss of function in mammalian cells plus stable isotope-labelled NR and NMN reported three things[2]. First, the kinase NRK1 is necessary and rate-limiting for the use of external NR and external NMN. Second, NRK1 is dispensable when the cell is fed nicotinamide or nicotinic acid instead. Third, and most awkward for the marketing, extracellular NMN is metabolised to NR, which is then taken up by the cell and converted to NAD+. Mammalian cells in that study required conversion of extracellular NMN to NR before uptake.
The enzyme doing the stripping is CD73, a cell-surface nucleotidase. In human tumour cell lines with the salvage enzyme NAMPT blocked, CD73 generated cell-permeable precursors from external phosphorylated substrates including NMN and NAD+, keeping intracellular NAD+ topped up[3]. A phosphate group is what separates NMN from NR. CD73 removes it on the outside of the cell, and NRK1 puts it back on the inside. The extra step exists in both directions.
There is a counter-argument, and it deserves a fair hearing. A 2019 mouse study reported that Slc12a8 is a sodium-dependent transporter that carries intact NMN, that knocking it down abolished NMN uptake in vitro and in vivo, that it transports NMN but not NR, and that it is upregulated in the aged mouse ileum[4]. If that holds, NMN has a private entrance that NR cannot use. The same journal then published a formal Matters Arising titled to say there is an absence of evidence that Slc12a8 encodes an NMN transporter[5], and the original group published a reply. The dispute has never been resolved in humans. The honest label for that route is disputed, not discovered.
Digestion work points the same way as the 2026 trial. Under simulated saliva, gastric and small-intestinal conditions followed by fermentation with human microbiota, NMN was degraded stepwise to NR, then nicotinamide, then ribose, with niacinate as the final microbial product[6]. Fermentation also raised propionate by 88% and butyrate by 23% and shifted the microbial community. Very little intact NMN survives the trip in that model.
One human trial makes the same point without meaning to. At 250 mg/day for 12 weeks, whole-blood NAD+ rose significantly while blood NMN itself did not, and the metabolite that rose alongside NAD+ was NAMN[20]. A second trial at the same dose could not detect serum NMN above 2 ng/mL or serum NAD+ above 5 ng/mL at either timepoint[23]. If oral NMN worked by arriving intact and being phosphorylated one step to NAD+, blood NMN would rise. In the trials that looked, it did not.
A field review from 2018 is often quoted as saying the two molecules behave the same in mice. It does not. It says their fates appear to depend on tissue distribution and on expression of NAD+ biosynthetic enzymes, nucleotidases and presumptive transporters, and that the pharmacokinetics remain under investigation[31]. Eight years later, the human answer arrived and it was convergence, not divergence.
Swipe the chart sideways
NR has the longer clinical record and, more importantly, the only clean milligram-by-milligram percentage ladder in the entire field. An 8-week randomized, double-blind, placebo-controlled trial in overweight healthy adults tested 100, 300 and 1,000 mg/day and measured whole-blood NAD+. It rose 22%, 51% and 142% respectively, within two weeks and maintained thereafter[12]. Nothing on the NMN side comes close to that clarity.
The first-in-human pharmacokinetic work reported blood NAD+ rising up to 2.7-fold after single doses of 100, 300 and 1,000 mg, dose-ordered[7]. An open-label study in 8 volunteers up-titrated to 1,000 mg twice daily for 9 days and measured a 100% rise in whole-blood NAD+, p=0.001, with the change in NR tracking the change in NAD+ at R-squared 0.72[8]. A 120-person randomized trial in adults aged 60 to 80 using NR with pterostilbene found whole-blood NAD+ about 40% higher on the single dose and about 90% higher on the double dose at 4 weeks, sustained to 8 weeks, with no rise in the placebo arm[9].
At the top of the dose range, a phase I safety trial gave 3,000 mg/day to 20 people with Parkinson disease for 4 weeks. Whole-blood NAD+ rose about 3.7-fold, with individual responses ranging from 1.8-fold to 5.8-fold[15]. That spread matters more than the average. At a fixed dose, the person at the bottom of the range got a third of what the person at the top got.
Tissue penetration has been tested too. In 12 aged men taking 1,000 mg/day for 21 days in a crossover design, the skeletal-muscle NAD+ metabolome was elevated, with NAAD and nicotinamide clearance products up[13]. The molecule reaches muscle. What happened there is covered in the functional section, and it is not what the marketing implies.
| Design and population | N | Dose | Duration | NAD+ result | Functional or safety outcome | Ref |
|---|---|---|---|---|---|---|
| First-in-human pharmacokinetics | single dose | 100, 300, 1,000 mg | single dose | blood NAD+ up to 2.7-fold, dose-ordered | no functional endpoint | [7] |
| Open-label pharmacokinetics, healthy adults | 8 | 250 up to 1,000 mg twice daily | 9 days | whole blood +100%, p=0.001 | none measured | [8] |
| Randomized, adults 60 to 80, NR plus pterostilbene | 120 | one and two times the label dose | 8 weeks | whole blood about +40% and +90% | no serious adverse events | [9] |
| Randomized crossover, adults 55 to 79 with stiff arteries | 24 completers | 1,000 mg/day | 6 weeks | PBMC NAD+ about +60%, NAAD about 5-fold | systolic BP -3.9 mmHg overall, not significant | [10] |
| Randomized, obese insulin-resistant men | 40 | 2,000 mg/day | 12 weeks | not reported on the record | insulin sensitivity not improved | [11] |
| Randomized, three doses, overweight healthy adults | not stated on record | 100, 300, 1,000 mg/day | 8 weeks | +22%, +51%, +142% whole blood | no adverse-event difference, no flushing | [12] |
| Randomized crossover, aged men | 12 | 1,000 mg/day | 21 days | muscle NAD+ metabolome elevated | inflammatory cytokines down, mitochondria unchanged | [13] |
| Randomized crossover, overweight adults, full phenotyping | 13 completers | 1,000 mg/day | 6 weeks | NAD+ metabolome analysed | insulin sensitivity p=0.98 | [14] |
| Randomized phase I, Parkinson disease | 20 | 3,000 mg/day | 4 weeks | whole blood about 3.7-fold, range 1.8 to 5.8 | no moderate or severe drug-related events | [15] |
Populations differ by trial, so percentages are not interchangeable. One trial measured peripheral blood mononuclear cells, several measured whole blood, and one measured skeletal muscle.
Swipe the chart sideways
NMN has more recent trials, larger sample sizes in a couple of cases, and messier measurement. The pattern across ten trials is that blood NAD+ or its metabolites rise reliably at 250 mg/day and above, while the molecule itself is often undetectable and the endpoint definitions vary widely.
The best-powered NMN trial randomized 80 healthy middle-aged adults to 300, 600 or 900 mg/day for 60 days. Blood NAD rose significantly in every NMN arm at day 30 and day 60 against both placebo and baseline, all p-values at or below 0.001, with the highest values at 600 and 900 mg[25]. That is the strongest NAD+ result on the NMN side. It is a set of p-values, not a percentage ladder, so it cannot be laid alongside the NR curve.
A 12-week trial in 30 healthy adults at 250 mg/day found whole-blood NAD+ significantly elevated at weeks 4, 8 and 12, peaking at week 4 and returning to baseline by week 16, four weeks after the last dose[20]. That is the most practical dosing fact in the whole category and almost nobody publishes it: the effect is rented, not owned.
Pharmacokinetics at the high end come from a microcrystalline pharmaceutical-grade polymorph. In 32 adults aged 55 to 80 taking 1,000 mg once or twice daily for 14 days, day-14 blood NMN reached 1.7-fold and 3.7-fold baseline respectively by mean area under the curve, with substantial dose-related NAD increases and very little unmodified NMN in the urine[24]. Response was unrelated to sex, body mass index and age.
Then there is the trial that could not measure anything. In 36 healthy adults aged 40 to 59 taking 250 mg/day for 12 weeks, serum NMN stayed below 2 ng/mL and serum NAD+ below 5 ng/mL at both timepoints, under the limit of quantification. What did move was serum nicotinamide, up 57.6% against a 26.8% fall on placebo[23]. Assay choice, tissue choice and matrix choice decide whether an NMN trial reports a NAD+ effect at all.
| Design and population | N | Dose | Duration | NAD+ result | Functional or safety outcome | Ref |
|---|---|---|---|---|---|---|
| Single-arm dose escalation, healthy men | 10 | 100, 250, 500 mg single doses | 5 hours | NAD+ not measured, metabolites rose | vitals and eye exam unchanged | [16] |
| Randomized, overweight postmenopausal women with prediabetes | 25 completers | 250 mg/day | 10 weeks | PBMC NAD+ AUC +43%, muscle NAD+ unchanged | muscle insulin sensitivity +25%, p<0.01 | [17] |
| Randomized, four arms, amateur runners | 48 | 300, 600, 1,200 mg/day | 6 weeks | NAD+ not measured | VO2max unchanged, ventilatory threshold dose-ordered | [18] |
| Randomized, morning against evening dosing, adults over 65 | 108 | 250 mg/day | 12 weeks | not on the fetched record | sleep, fatigue and performance endpoints | [19] |
| Randomized, healthy adults 20 to 65 | 30 | 250 mg/day | 12 weeks plus washout | whole blood up at 4, 8, 12 weeks, peak at week 4 | returned to baseline 4 weeks after stopping | [20] |
| Randomized, healthy men over 65 | 42 randomized | 250 mg/day | 12 weeks | whole-blood NAD+ and metabolites up | gait speed p=0.002 at 12 weeks in about 10 per group | [21] |
| Randomized safety study, healthy adults | 31 | 1,250 mg once daily | 4 weeks | not the endpoint | no change beyond physiological variation | [22] |
| Randomized, healthy adults 40 to 59 | 36 randomized | 250 mg/day | 12 weeks | serum NMN and NAD+ below the assay limit | arterial stiffness -2.1%, p=0.097 | [23] |
| Randomized pharmacokinetics, microcrystalline NMN | 32 | 1,000 mg once or twice daily | 14 days | blood NMN 1.7-fold and 3.7-fold baseline | well tolerated, no sex, age or BMI effect | [24] |
| Randomized multicentre, three doses, middle-aged adults | 80 | 300, 600, 900 mg/day | 60 days | blood NAD up in all arms, p<=0.001 | 6-minute walk up, HOMA-IR not significant | [25] |
Swipe the chart sideways
Swipe the chart sideways
Blood NAD+ went up in almost every trial on both sides. Almost nothing else did. This is the section that most comparison articles skip, and it is the one that should decide how much you are willing to spend.
The strongest positive result on either side is on the NMN side. In 25 completers, 13 on NMN and 12 on placebo, all overweight or obese postmenopausal women with prediabetes, 250 mg/day for 10 weeks raised insulin-stimulated glucose disposal per kilogram of fat-free mass by 25 plus or minus 7%, p<0.01, measured with a hyperinsulinaemic-euglycaemic clamp[17]. Muscle AKT and mTOR phosphorylation rose and muscle-remodelling genes were upregulated.
Read the same paper's null list and the picture changes. Body weight, body mass index, fat mass, intra-abdominal fat volume, intrahepatic triglyceride, systolic and diastolic blood pressure, plasma glucose, insulin, free fatty acids, lipids, adiponectin, leptin, hepatic insulin sensitivity, adipose insulin sensitivity, skeletal muscle mitochondrial respiratory capacity, handgrip strength, knee torque and fatigue recovery all failed to change. So did skeletal-muscle NAD+[17]. A 25% insulin-sensitivity gain in 13 women, with no measurable rise in muscle NAD+, is a real finding and a thin one. It also drew a formal published Comment and Response in the same journal.
The most quoted NR benefit is a blood-pressure number. In 24 completers aged 55 to 79 taking 1,000 mg/day for 6 weeks in a crossover, systolic pressure fell 3.9 mmHg and diastolic 2.0 mmHg, neither significant after correction for multiple comparisons. In an exploratory subgroup of 13 participants with baseline pressure in the 120 to 139 over 80 to 89 range, systolic pressure was about 9 mmHg lower on NR, and the paper states plainly that no statistical inferences can be made from that subgroup[10]. Carotid-femoral pulse wave velocity moved in the right direction and also missed significance after correction. Brachial flow-mediated dilation did not change at all.
The gait-speed headline for NMN needs the same treatment. In healthy men over 65 at 250 mg/day, gait speed was higher on NMN at week 6 and week 12, p=0.023 and p=0.002. But a supplier error invalidated the 12-week data of 22 participants, leaving about 10 per group, and in that analysed subset baseline gait speed already differed significantly between groups, p=0.041[21]. Grip strength never reached a significant between-group difference at any timepoint.
Now the nulls, which are larger and better instrumented than the positives. The most thoroughly phenotyped NR metabolic trial used clamps, magnetic resonance spectroscopy, muscle biopsies, ex vivo mitochondrial respiration and whole-body energy metabolism in 13 completers on 1,000 mg/day for 6 weeks with 99.3% compliance. Peripheral insulin sensitivity came out at p=0.98. Hepatic insulin sensitivity was numerically worse on NR, p=0.30. Non-oxidative glucose disposal, carbohydrate oxidation and body weight were all flat[14]. A separate trial doubled the dose to 2,000 mg/day for 12 weeks in 40 obese, insulin-resistant, sedentary men and found no improvement in insulin sensitivity, endogenous glucose production, glucose disposal or glucose oxidation[11].
NMN has matching nulls. Arterial stiffness fell 2.1% against a 0.5% rise on placebo and missed significance, p=0.097[23]. In the 80-person dose-ranging trial, six-minute walk distance and self-reported health improved but HOMA-IR showed no difference from placebo at day 60[25]. In 48 amateur runners, VO2max did not improve in any arm, and neither did oxygen pulse, peak power, peak workload, maximum heart rate or heart-rate recovery. What did improve, dose-ordered, was the percentage of VO2max at the first ventilatory threshold: plus 2.1% in control, plus 3.5% at 300 mg, plus 6.5% at 600 mg and plus 10.3% at 1,200 mg[18]. That is a threshold shift, not a ceiling shift.
The 21-day muscle trial on NR is the most instructive of all. NR reached aged human muscle and raised the muscle NAD+ metabolome. Muscle transcriptomics then showed downregulation of energy-metabolism and mitochondrial pathways, mitochondrial bioenergetics were unchanged, and circulating inflammatory cytokines fell[13]. The molecule arrived, the biomarker moved, and the mitochondrial story ran backwards.
Swipe the chart sideways
This is where nearly every published comparison is now wrong, including the earlier version of this page. Two claims circulate widely and both are out of date: that NR received GRAS status in 2022, and that NMN is banned or excluded as a US dietary supplement.
NR in the United States. The GRAS notice for nicotinamide riboside chloride is GRN 000635, filed in March 2016, and the FDA response letter is dated 3 August 2016[33]. The letter says the agency has no questions at this time regarding the notifier's conclusion, and states explicitly that the FDA has not made its own determination of GRAS status. The intended uses were vitamin waters, protein shakes, bars, gum, chews and powdered beverages, at a maximum use level of 0.0057% by weight as consumed. That is a food-use clearance from 2016, not an efficacy endorsement and not a 2022 event.
NMN in the United States. The history runs like this. NMN received the first NMN new dietary ingredient acknowledgement in May 2022. In October and November 2022 the FDA issued two letters concluding that NMN was excluded from the dietary supplement definition, reasoning from the drug-preclusion provision, and the ingredient's notification status was withdrawn. Trade bodies filed citizen petitions in March and May 2023. The agency missed its 180-day deadline, litigation followed in 2024, a district judge stayed proceedings, and the FDA instituted a period of enforcement discretion. Then on 29 to 30 September 2025 the FDA answered the petition and reversed itself. The letter revised the agency's reading of the race-to-market clause at section 201(ff)(3)(B) of the Federal Food, Drug, and Cosmetic Act, concluding that its earlier requirement that prior marketing must have been lawful was not the best reading of the statute[37][39]. On 2 December 2025 the FDA sent new letters to two ingredient suppliers, covering new dietary ingredient notifications 1247 and 1259, confirming reconsideration and reinstating notification status. NMN may be lawfully marketed as a US dietary supplement[38].
The practical consequence is blunt. If you were choosing NR over NMN because of US legality, that reason expired in December 2025.
The European Union is the real asymmetry now, and it runs the other way. Nicotinamide riboside chloride was authorised as a novel food by Commission Implementing Regulation (EU) 2020/16 of 10 January 2020 and added to the Union list[34]. The conditions of use are specific: a maximum of 300 mg/day for adults excluding pregnant and lactating women, and 230 mg/day for pregnant and lactating women. Regulation (EU) 2022/1160 extended the conditions in July 2022 to cover foods for special medical purposes, total diet replacement at 500 mg/day, and meal replacements at 150 mg per meal to a maximum of 300 mg/day[35]. Five-year data protection ran from 20 February 2020 to 20 February 2025, which is why generic EU NR only became possible in 2025.
NMN is absent from the Union list of novel foods[36]. A new novel-food application for NMN was received on 21 August 2025[39]. Until that dossier completes assessment and authorisation, NMN is not a legal food supplement ingredient in the EU.
The EU cap has a consequence that almost nobody writes about. The trial doses that produced the largest NAD+ increases sit above the legal ceiling. The 8-week dose-response trial measured +142% whole-blood NAD+ at 1,000 mg/day and +51% at 300 mg/day[12]. An EU buyer can legally access roughly one third of the elevation the best-controlled NR trial demonstrated, and cannot legally buy the alternative molecule at all.
| Market and molecule | Status | Instrument and date | Dose limits | Ref |
|---|---|---|---|---|
| United States, NR | Lawful | GRAS notice GRN 000635, no-questions response letter dated 3 August 2016 | No supplement dose cap set by that letter | [33] |
| United States, NMN | Lawful again | Excluded from the supplement definition in November 2022, exclusion reversed 29 to 30 September 2025, new dietary ingredient status reinstated 2 December 2025 for notifications 1247 and 1259 | No dose cap | [38] |
| European Union, NR | Authorised with a hard cap | Novel food under Regulation (EU) 2020/16, conditions extended by Regulation (EU) 2022/1160 | 300 mg/day for adults, 230 mg/day if pregnant or lactating, 500 mg/day in total diet replacement | [34] |
| European Union, NMN | Not authorised | Absent from the Union list of novel foods. A new dossier was received on 21 August 2025 | No legal supplement dose | [36] |
Swipe the chart sideways
There is a commercial tell worth noting. The company that fought the NR patent case now also sells an NMN product in our catalogue. When both sides of a molecule war sell both molecules, the war is over.
Swipe the chart sideways
The cost calculation that actually matters is cost to match a trial. Matching the +142% NAD+ arm means 1,000 mg/day of NR, which at the cheapest tested NR in our catalogue is about $2.67 a day, roughly $80 a month. Matching the 250 mg/day dose used in the clamp-verified insulin-sensitivity trial means about $0.16 a day of the cheapest tested NMN powder, roughly $5 a month[44]. The cheapest evidence-matching dose in the entire dataset is on the NMN side, for the simple reason that the positive NMN trial used a four-times smaller dose.
NR chloride is the only NR form with human trials. Every trial cited on this page used nicotinamide riboside chloride, and the EU authorisation names that exact salt with a mandatory label designation[34]. Newer malate and tartrate salts exist in patent filings[42] and have no published human trial.
The reduced forms beat both molecules in cells and have no human data. Dihydronicotinamide riboside, written NRH, raised cellular NAD+ 2.5-fold to 10-fold over control and produced greater increases than NR and NMN in every instance at equivalent concentrations, acting within as little as one hour[26]. It works through a different route: the kinase is adenosine kinase, NRH goes to NMNH and then to NAD+, and blocking adenosine kinase pharmacologically or genetically stops the effect[28]. A separate group showed NRH is orally bioavailable in mice and uses an NRK1-independent path[27]. There is no published human randomized trial of NRH. Anyone claiming a bioavailability ranking that includes NRH in humans is extrapolating from mice.
Sublingual and liposomal NMN rest on no human comparison. No published human trial compares sublingual, liposomal and oral NMN pharmacokinetics. That absence is the whole story, because the price gap is enormous. One brand in our catalogue sells sublingual NMN powder at $0.65 per gram and a liposomal NMN at $10.53 per gram, a 16-fold premium from the same company for a delivery claim with no published human comparative data[44].
The mechanism cuts both ways here, which is why the question is genuinely open. If oral NMN is degraded to NR, nicotinamide and finally nicotinic acid in the gut[6], bypassing the gut sounds attractive. But if the 2026 model is right and the NAD+ rise depends on microbially-produced nicotinic acid[1], then the gut step is not an obstacle to route around. It may be the mechanism. Sublingual delivery could reduce the effect rather than increase it. Nobody has tested it.
Pharmaceutical-grade NMN exists and behaves as expected. A microcrystalline polymorph of beta-NMN doubled day-14 blood NMN when the dose doubled, from 1.7-fold to 3.7-fold baseline, with very little unmodified NMN in urine[24]. It is a drug-development asset, and its existence under investigational new drug status is exactly what triggered the 2022 US exclusion in the first place.
Storage. NMN in powder form is hygroscopic and degrades faster warm and damp than cold and dry, which is why suppliers ship it with desiccant and recommend refrigeration for opened powder. No human trial has tested a deliberately degraded product, so the practical rule is procedural rather than evidence-based: buy from a seller who publishes a certificate of analysis with a recent test date, keep powder sealed, cold and dry, and do not buy a year's supply of an open-scoop product.
Across roughly 500 randomized participants on either molecule, the trial record shows no serious adverse events attributable to the supplement.
NR. Preclinical toxicology found no mortality at 5,000 mg/kg oral, no genotoxicity in the Ames, chromosome-aberration and micronucleus assays, and set a no-observed-adverse-effect level of 300 mg/kg/day with a lowest-observed-adverse-effect level of 1,000 mg/kg/day in 90-day rat feeding[29]. In humans, the 8-week three-dose trial found no significant difference in adverse events between NR and placebo or between doses, no reports of flushing, no LDL cholesterol elevation and no disruption of one-carbon metabolism[12]. The 6-week crossover recorded 14 treatment-emergent events in 7 of 30 participants, all self-reported and mild, no serious events, and two side-effect dropouts that both occurred during the placebo phase[10]. Twelve weeks at 2,000 mg/day in 40 obese men produced no serious adverse events and normal safety bloods[11]. Four weeks at 3,000 mg/day in 20 people with Parkinson disease was designed around adverse-event frequency as the primary outcome, and all 20 completed[15].
NMN. Single doses of 100, 250 and 500 mg produced no significant change in heart rate, blood pressure, oxygen saturation or body temperature, with small dose-independent laboratory shifts all inside normal ranges and no change in eye examination or sleep scores[16]. Four weeks at 1,250 mg once daily in 31 adults produced no changes exceeding physiological variation in anthropometry, haematology, biochemistry, urinalysis or body composition, and no severe adverse events[22]. Twelve weeks at 250 mg/day in men over 65 was well tolerated[21]. All 80 participants in the 60-day dose-ranging trial completed without protocol violation[25]. Mild gastrointestinal complaints are widely asserted for NMN but were not quantified in any of these trial records, so treat that claim as unverified.
The cancer question deserves a straight answer. The mechanistic concern is real and it is not about the supplements. NAMPT, the rate-limiting enzyme of the NAD+ salvage pathway, governs the proinflammatory senescence-associated secretome independently of growth arrest, acting through suppression of AMPK, then of p53-mediated inhibition of p38 MAPK, and finally through enhanced NF-kappaB activity[30]. That is a cell-biology finding with no human participants, no doses and no NR or NMN supplementation in the paper. It says high NAD+ can amplify the inflammatory output of senescent cells in a dish.
The direct question was addressed in a safety-focused review. Nicotinamide and NR were not found to be genotoxic. Nicotinamide, NR and NMN were not found to be directly carcinogenic. No evidence was found that prolonged NR or NMN treatment stimulated tumour development in animals. The review still calls for more work, because raising NAD+ could influence tumour biology indirectly and NAD+ may have stage-dependent effects, with adequate NAD+ described as protective in early carcinogenesis[32].
The defensible position is narrow: a theoretical, stage-dependent concern, no human signal in any trial on this page, and a clear reason for anyone with active cancer or a cancer history to raise it with their oncologist before starting.
The modifier that predicts response across these trials is baseline dysfunction, not age and not sex. Two independent trials, one on each molecule, found effects only in their metabolically worst-off subgroups.
In the NR blood-pressure crossover, every participant was aged 55 to 79 with impaired flow-mediated dilation, so the trial cannot speak to age at all. Its subgroups were split by baseline blood pressure. The 13 participants starting at 120 to 139 over 80 to 89 mmHg were about 9 mmHg lower systolic on NR. The 11 starting below 120 over 80 showed no change[10]. Anyone who reports that trial as showing older adults respond less strongly has misread which variable defined the subgroups.
In the NMN arterial-stiffness trial, no sex difference was detected in any endpoint after 12 weeks. The exploratory subgroups that did move were participants with above-average body mass index and above-average blood glucose[23]. Same pattern, different molecule.
The largest NMN pharmacokinetic dataset is explicit that changes in NMN or NAD levels were unrelated to sex, body mass index and age across 32 adults aged 55 to 80[24]. If you are hoping for a dose adjustment by demographic, the data do not support one.
One population does stand out, and it is narrow. The only clamp-verified functional gain on either molecule was in overweight or obese postmenopausal women with prediabetes at 250 mg/day NMN[17]. That result has not been replicated, and the same trial found no change in muscle NAD+.
The design bias underneath all of this is worth stating. One 12-week NMN trial enrolled only men over 65[21]. The positive NMN insulin trial enrolled only postmenopausal women[17]. The 2,000 mg NR trial enrolled only men[11]. There is no adequately powered mixed-sex functional trial for either molecule.
Our database holds 316 products. 39 contain NMN and 27 contain NR, of which 34 and 17 respectively have that molecule as their highest-milligram ingredient. Price per gram is the product price divided by the total grams of the named active, computed the same way for both molecules[44].
| Molecule | Brand | Product | Score | Dose/serving | Servings | Price | $ per gram | Public lab certificate |
|---|---|---|---|---|---|---|---|---|
| NMN | Renue By Science | NMN Pure Powder, sublingual | 8.3 | 500 mg | 200 | $64.95 | $0.65 | Yes |
| NMN | DoNotAge | Pure NMN | 8.1 | 1,000 mg | 30 | $65.00 | $2.17 | Yes |
| NMN | Nootropics Depot | NMN Capsules | 8.1 | 250 mg | 90 | $49.99 | $2.22 | Yes |
| NMN | ProHealth | NMN Pro 1000 | 8.0 | 1,000 mg | 30 | $79.00 | $2.63 | Yes |
| NMN | Double Wood | NMN Supplement | 7.9 | 250 mg | 60 | $48.95 | $3.26 | Yes |
| NR | DoNotAge | Pure NR | 8.1 | 500 mg | 30 | $55.00 | $3.67 | Yes |
| NR | Nootropics Depot | Nicotinamide Riboside Capsules | 8.0 | 300 mg | 30 | $34.99 | $3.89 | Yes |
| NR | Life Extension | NAD+ Cell Regenerator, 300 mg | 7.8 | 300 mg | 30 | $24.00 | $2.67 | No |
| NR | Thorne | NiaCel 400 | 7.7 | 400 mg | 60 | $65.00 | $2.71 | Yes |
| NR | Tru Niagen | Tru Niagen PRO 1000 mg | 7.7 | 1,000 mg | 30 | $104.40 | $3.48 | No |
Ranked by our six-pillar overall score inside each molecule. All ten products are tested by independent lab. No brand pays for placement[44].
Three observations from the table. First, the price inversion is visible at the top of both lists: the best-scoring NR powder costs $3.67 per gram, more than four of the five top NMN products. Second, only one NR product in the catalogue reaches the 1,000 mg/day dose used in the trials that produced the largest NAD+ increases, and it costs $104.40 for 30 servings. Third, one brand sells both molecules as single-ingredient, third-party-tested powders with public certificates, at $2.17 per gram for NMN and $3.67 for NR, which is the cleanest like-for-like price comparison available anywhere in the dataset[44].
Full rankings: Best NMN Supplements 2026 and Best NR Supplements 2026.
Six situations, six answers. Each card shows the dose the trial used and links to the trial that produced the number.
Dose: 250 mg/day, the dose used in four randomized trials
Cost: about $0.16/day at the cheapest tested powder in our catalogue
Whole-blood NAD+ rose at 4, 8 and 12 weeks in 30 adults and peaked at week 4.
Evidence tier A minusRead the 12-week trialLegal cap: 300 mg/day for adults, 230 mg/day if pregnant or lactating
What that buys: about +51% whole-blood NAD+, not the +142% seen at 1,000 mg
NMN has no EU authorisation. A dossier was only received on 21 August 2025.
Regulatory certaintyRead Regulation (EU) 2020/16Trial: N=65, 14 days, three arms, randomized
Result: NR and NMN raised circulatory NAD+ comparably
Buy on price per gram and third-party testing, not on the molecule.
First direct comparisonRead the 2026 trialTrial: N=25 completers, 10 weeks, clamp-verified
Result: insulin-stimulated glucose disposal +25%, p<0.01
13 women against 12 on placebo. Muscle NAD+ did not change and neither did weight or blood pressure.
Evidence tier B plusRead the trialTrial: N=48 amateur runners, 6 weeks, four arms
Result: %VO2max at the first ventilatory threshold +6.5% and +10.3%
VO2max itself did not change in any arm. The gain is in threshold intensity, not ceiling.
Evidence tier C plusRead the trialHuman signal: none across roughly 500 randomized participants
Animal evidence: no NR or NMN tumour promotion reported in the safety review
The concern is mechanistic. High NAD+ can amplify the inflammatory secretome of senescent cells in culture.
Read the safety reviewDoses are stated as the trials stated them. Nothing on this page is medical advice, and no trial on this page ran longer than 12 weeks at a functional endpoint.
The first head-to-head trial found NR and NMN raised circulatory NAD+ comparably over 14 days[1]. Both are lawful US dietary ingredients as of December 2025[38]. Only NR is legal in the EU, capped at 300 mg/day[34]. So the decision is geography, then price per gram, then whether the brand publishes a certificate of analysis. The molecule is the least important variable in the purchase.
The first direct human comparison, 65 adults over 14 days, found that NR and NMN raised circulatory NAD+ comparably and that nicotinamide did not. Neither molecule won. Choose on price per gram, third-party testing and what is legal where you live.
No. NMN was excluded from the dietary supplement definition in November 2022. The FDA reversed that reading of the race-to-market clause on 29 to 30 September 2025 and reinstated new dietary ingredient status on 2 December 2025 through letters to two ingredient suppliers. NMN may be lawfully marketed as a US dietary supplement.
Not legally as a food supplement. NMN does not appear in the Union list of novel foods and a new application was only received on 21 August 2025. NR is authorised under Regulation (EU) 2020/16 at a maximum of 300 mg/day for adults and 230 mg/day for pregnant or lactating women.
In the cleanest single dose-response trial, whole-blood NAD+ rose 22% at 100 mg/day NR, 51% at 300 mg and 142% at 1,000 mg over 8 weeks. A 4-week trial at 3,000 mg/day in Parkinson disease reached about 3.7-fold. NMN has no published mg-by-mg percentage ladder.
There is no trial of the combination. Both funnel through the same cell-surface and microbial steps, so the mechanistic case for stacking them is weak. Two products at half dose each costs more than one product at full dose.
Not at the median. Across our catalogue the median price per gram is $4.07 for NMN and $3.67 for NR. NMN has a wider spread, from $0.44 to $35.60 per gram, so the cheapest and the most expensive options are both NMN.
No published human trial compares sublingual, liposomal and oral NMN pharmacokinetics. One brand in our catalogue sells sublingual NMN powder at $0.65 per gram and liposomal NMN at $10.53 per gram, a 16-fold premium resting on no head-to-head human data.
In a 12-week trial at 250 mg/day NMN, whole-blood NAD+ peaked at week 4, stayed elevated at weeks 8 and 12, and was back at baseline 4 weeks after the last dose. NAD+ elevation is a maintenance effect, not a stored one.
Neither has a flushing signal in the trial record. An 8-week randomized trial at up to 1,000 mg/day NR reported no flushing at all. In a 6-week crossover, three participants reported flushing and two of those reports came during the placebo phase.
Corrections and source queries: info@no1gevity.com.