What did the July 2026 NAD+ trial find?

That 12 weeks of nicotinamide riboside at 1000 mg per day raised blood NAD+ 2.08-fold in older adults with amnestic mild cognitive impairment, while producing no significant improvement in cognitive function, total cerebral blood flow, blood pressure or arterial stiffness. It was published in Alzheimer's & Dementia and indexed on PubMed as PMID 42478598.

How large was the trial?

Small. 64 people were screened, 52 randomised, 49 started treatment and 42 completed, 22 on nicotinamide riboside and 20 on placebo. It is described by its own authors as a phase-II pilot.

Was the trial designed to detect a cognitive effect?

Cognition was the declared primary outcome, but the sample size was never calculated for it. The sample size came from effect sizes for blood pressure and arterial stiffness instead, no correction for multiple comparisons was applied, and only completers were analysed. That means a modest cognitive effect could have been missed, which is a limitation the authors state themselves.

Is nicotinamide riboside the same as NAD+?

No. Nicotinamide riboside is an orally dosed derivative of vitamin B3 that the body converts into NAD+. It is not a peptide, and in every trial discussed here it was taken orally as capsules, although nicotinamide riboside is also given intravenously in some commercial wellness settings. NAD+ itself, given by infusion, is a different pharmacological situation because it bypasses gut absorption and first-pass metabolism.

Do oral NAD+ precursors actually raise NAD+ in the blood?

Yes, and this part is not in dispute. A randomised placebo-controlled trial found that 100, 300 and 1000 mg per day of nicotinamide riboside raised whole-blood NAD+ by 22, 51 and 142 percent within two weeks, and a 2026 randomised study in 65 people found both nicotinamide riboside and nicotinamide mononucleotide raised circulating NAD+ over 14 days while plain nicotinamide did not.

Does raising blood NAD+ raise brain NAD+?

It is not established. The 2026 trial measured blood only and its authors write that whether the blood increase translated to brain NAD+ is not known. The available human brain data amount to one acute study in 10 healthy volunteers after a single dose and one 12-person pharmacokinetic study in which cerebral NAD+ rose only after about four weeks. Neither included a cognitive outcome.

Have other trials found a cognitive benefit?

Not in a randomised setting. Three trials in mild cognitive impairment or subjective cognitive decline, from 2024, 2025 and 2026, used three different cognitive instruments over 8, 10 and 12 weeks, and all three missed their cognitive endpoint. In the 2024 trial, cerebral blood flow in the default mode network decreased rather than increased.

What was the hippocampus finding?

An exploratory measurement. Blood flow in the left hippocampus rose in the treated group with p equals 0.033, the only subregion below 0.05 out of a dozen examined, with the hippocampus as a whole sitting exactly on 0.050. The authors state that none of these regional changes would have survived correction for multiple comparisons, and that the regional analyses were added after the trial began.

What about the widely quoted blood-pressure benefit?

It is weaker than it is usually presented. In the 2018 pilot from the same laboratory, the systolic and diastolic changes across all participants were not statistically significant after correction for multiple comparisons, and the often-cited 9 mmHg figure comes from a post-hoc subgroup of 13 participants with elevated baseline pressure, about which the paper states explicitly that no statistical inferences can be made.

Is there any randomised trial of intravenous NAD+?

Yes, one that we could locate: a 2026 trial in 180 patients with heart failure from ischemic cardiomyopathy, using 10 mg per day intravenously for seven days on top of standard therapy. It met its primary imaging endpoint, a change in ejection fraction at one month, at p equals 0.024, while all clinical secondary endpoints were non-significant trends. We could not locate any randomised trial of infused NAD+ with a cognitive endpoint, nor any at the higher doses used in wellness infusions.

How well tolerated is intravenous NAD+?

The only head-to-head record we could find is a retrospective review of a commercial infusion chain's own client records, which reported moderate to severe gastrointestinal symptoms, increased heart rate and chest pressure during NAD+ infusions, against minor tingling and mild cramping for intravenous nicotinamide riboside. Infusion time averaged 97 minutes versus 37 minutes. The people who conducted the review worked for the company whose records were reviewed.

Does this mean NAD+ research is pointless?

No. It means the evidence currently supports one claim and not the other. Precursors reliably raise blood NAD+; no randomised trial has yet shown a robust functional or cognitive benefit from doing so. Larger and longer trials could change that, and the authors of the 2026 trial argue for exactly that.

Who funded the 2026 trial?

It was supported by United States National Institutes of Health grants, specifically NIA K01AG054731 and NIGMS P20GM113125. The nicotinamide riboside and the matching placebo were supplied by Niagen Bioscience, formerly ChromaDex, under a material transfer agreement. The paper carries a statement that the authors have no conflicts of interest to disclose.

Is NAD+ approved as a medicine?

No. NAD+ and its precursors are not approved medicines in the EU. We supply NAD+ as a research compound for laboratory use only, not for human consumption, and we give no dosing or usage recommendations.

How can the identity and purity of a research compound be verified?

Through a batch-specific certificate of analysis from a third-party laboratory that reports identity, purity and the measured content per vial, and that can be checked against the issuing laboratory's own records. Our supplier lab reports are published per batch on our CoA page.

Blood NAD+ Doubled, Cognition Did Not Move: What the July 2026 Trial Actually Found

Q: Were there side effects?

No serious adverse events occurred in either arm. The most frequent complaints in the treated group were gastrointestinal symptoms such as nausea and diarrhea, plus mild headaches, all transient. Four participants discontinued because of adverse events, three of them on placebo.

Important notice: This article discusses published clinical research for scientific information only. NAD+ is a research compound in our catalogue, not intended for human consumption, and nothing here is a dosing recommendation, a protocol or medical advice. Where a trial dose is mentioned, it is a description of what researchers administered under an approved protocol, not a suggestion.

TL;DR: the cleanest test yet of a claim everyone assumes

The trial: A 12-week, double-blind, randomised, placebo-controlled phase-II pilot of nicotinamide riboside, 500 mg twice daily, in adults aged 65 and over with amnestic mild cognitive impairment. Published in Alzheimer's & Dementia on 21 July 2026 (PMID 42478598), 42 completers. The biochemistry worked: blood NAD+ rose from 23.36 to 48.52 micromolar, a 2.08-fold increase, p less than 0.001, with a very large effect size. Placebo did not move. The outcome did not: no significant improvement on any of the ten cognitive subscales, none on total cerebral blood flow, none on blood pressure or arterial stiffness. The one interesting signal: blood flow in the left hippocampus rose, p equals 0.033, and the authors themselves note it would not survive correction for multiple comparisons. Why it matters: this is now the third randomised trial in a mild-cognitive-impairment or subjective-cognitive-decline population to miss its cognitive endpoint, and the second in which blood NAD+ demonstrably rose while cognition did not. Raising a molecule and changing an outcome are two different claims, and only one of them has been demonstrated.

NAD+longevity

Essential cellular coenzyme that declines with age. Powers energy metabolism in every cell, activates sirtuins (longevity genes), and supports DNA repair. A cornerstone molecule in aging and longevity research.

What the trial did

The study ran at a single site, the University of Delaware, under an investigational new drug application, and was registered in 2018 as NCT03482167. Community-dwelling adults aged 65 or older with amnestic mild cognitive impairment took four capsules a day, two in the morning and two in the evening, each containing 250 mg of nicotinamide riboside, or a matching cellulose placebo, for 12 weeks. The compound and the placebo were supplied by Niagen Bioscience, formerly ChromaDex, under a material transfer agreement.

Of 64 people screened, 52 were randomised, 49 started treatment and 42 finished: 22 on nicotinamide riboside and 20 on placebo.

What the trial was actually powered to detect

This detail is easy to miss and it changes the interpretation. The sample size was not calculated for cognition. In the authors' words, "sample size estimates were based on effect sizes from our previous pilot study, targeting detection of improvements in blood pressure or arterial stiffness, which we hypothesized to be a key driver of cognitive benefits". No correction for multiple comparisons was applied anywhere, deliberately, to avoid an overly conservative design in a pilot. The analysis included completers only rather than everyone randomised, and the authors state that a future phase-III trial should use an intention-to-treat analysis.

The results, in the order that matters

Outcome	Nicotinamide riboside	Placebo	Result
Blood NAD+ (micromolar)	23.36 to 48.52	24.12 to 25.14	p less than 0.001, d 1.879
Total NAD metabolome	41.84 to 70.71	44.35 to 45.61	p less than 0.001, d 1.499
Cognition, 10 subscales	no significant change	no significant change	all p values 0.062 to 0.850
Total cerebral blood flow	48.5 to 55.7	47.2 to 47.2	p equals 0.135
Left hippocampal blood flow	51.7 to 58.0	55.6 to 51.7	p equals 0.033, exploratory

Blood NAD+ (micromolar)

Nicotinamide riboside: 23.36 to 48.52
Placebo: 24.12 to 25.14
Result: p less than 0.001, d 1.879

Total NAD metabolome

Nicotinamide riboside: 41.84 to 70.71
Placebo: 44.35 to 45.61
Result: p less than 0.001, d 1.499

Cognition, 10 subscales

Nicotinamide riboside: no significant change
Placebo: no significant change
Result: all p values 0.062 to 0.850

Total cerebral blood flow

Nicotinamide riboside: 48.5 to 55.7
Placebo: 47.2 to 47.2
Result: p equals 0.135

Left hippocampal blood flow

Nicotinamide riboside: 51.7 to 58.0
Placebo: 55.6 to 51.7
Result: p equals 0.033, exploratory

The cognitive battery was not a single crude score. It combined the California Verbal Learning Test third edition, two Wechsler Memory Scale subtests and the NIH Toolbox fluid cognition composite, reported across ten subscales. Not one showed a significant group-by-time interaction. The closest was delayed logical memory at p equals 0.062. Both groups improved slightly within themselves, which the authors attribute to practice effects from repeating the tests.

Adherence was 85 percent in the treated arm and 91 percent on placebo, a difference that was not significant. There were no serious adverse events. The most frequent complaints on nicotinamide riboside were gastrointestinal symptoms such as nausea and diarrhea, plus mild headaches, all transient.

The hippocampus finding, handled honestly

The one result that looks like something is the rise in left hippocampal blood flow. The authors do not oversell it, and neither will we. In their words: "Several subregions showed promising trends towards improvement with NR with moderately large effect sizes ... Most of these changes did not reach statistical significance except for the left hippocampus, and none would have survived correction for multiple comparisons." These regional analyses were also added after the trial began, so that the results could be compared with a study published in the meantime. That makes them hypothesis-generating, which is a real category of finding, and not evidence of benefit.

This is the third trial to land in the same place

The most useful thing about this result is that it is not an outlier. Three randomised trials have now given an NAD+ precursor to people with mild cognitive impairment or subjective cognitive decline, using three different cognitive instruments, and all three missed on cognition.

Trial	Population	Duration	Cognitive result
Orr et al. 2024, GeroScience	20 adults with MCI	10 weeks	Primary outcome was cognition (MoCA). Not improved. Cerebral blood flow in the default mode network went down, not up.
Wu et al. 2025, Alzheimer's & Dementia TRCI	46 randomised, subjective decline or MCI	8 weeks	Primary cognitive outcome (RBANS) not met. A plasma biomarker, pTau217, did fall relative to placebo.
Martens et al. 2026, Alzheimer's & Dementia	42 completers with amnestic MCI	12 weeks	Primary outcome was cognition. No improvement on any of ten cognitive subscales.

Orr et al. 2024, GeroScience

Population: 20 adults with MCI
Duration: 10 weeks
Cognitive result: Primary outcome was cognition (MoCA). Not improved. Cerebral blood flow in the default mode network went down, not up.

Wu et al. 2025, Alzheimer's & Dementia TRCI

Population: 46 randomised, subjective decline or MCI
Duration: 8 weeks
Cognitive result: Primary cognitive outcome (RBANS) not met. A plasma biomarker, pTau217, did fall relative to placebo.

Martens et al. 2026, Alzheimer's & Dementia

Population: 42 completers with amnestic MCI
Duration: 12 weeks
Cognitive result: Primary outcome was cognition. No improvement on any of ten cognitive subscales.

Three trials, three instruments, three timeframes, one direction of result. Each is small, and none of them proves that NAD+ precursors cannot help cognition. What they do establish is that the effect, if it exists, is not large enough or fast enough to show up in the studies that have been run.

The pattern extends beyond the brain

The same shape appears in the wider precursor literature. A 12-week trial in 40 men with obesity at 2000 mg per day did not improve insulin sensitivity, its registered primary endpoint. A six-week trial combining the compound with supervised walking in 54 people with hypertension did not lower daytime systolic pressure more than exercise plus placebo: daytime systolic rose by 5.19 mmHg in the combination arm while it fell by 2.71 mmHg on exercise plus placebo. The most positive functional result, a 2024 trial in 90 people with peripheral artery disease, reported a 17.6 metre gain in six-minute walking distance with a one-sided p of 0.08 against a pre-specified one-sided threshold of 0.10. Under a conventional two-sided test at 0.05, that result would not be called significant. Even the widely quoted 9 mmHg blood-pressure reduction from the 2018 pilot was a post-hoc subgroup of 13 people, and the paper says in plain words that no statistical inferences can be made from it.

What IS solidly established

It would be just as dishonest to write this article as though nothing works. The bioavailability question is settled, and settled well.

In a 2019 randomised, double-blind, placebo-controlled trial in overweight but otherwise healthy adults, nicotinamide riboside at 100, 300 and 1000 mg per day raised whole-blood NAD+ by 22 percent, 51 percent and 142 percent respectively within two weeks, and the increases held for the rest of the study. A 2026 randomised study in 65 healthy participants found that both nicotinamide riboside and nicotinamide mononucleotide raised circulating NAD+ over 14 days, while plain nicotinamide did not.

So the first half of the marketing claim is true: oral precursors raise blood NAD+, dose-dependently and reliably. It is the second half, that this changes how a person functions, that has not been demonstrated for cognition. Outside cognition the record is contested rather than empty: a 2021 randomised trial of nicotinamide mononucleotide in postmenopausal women with prediabetes did increase insulin-stimulated muscle glucose disposal, but a technical comment in the same journal argues that the trial was not effectively randomised, because the 13 women given the compound started with a hepatic lipid content of 6.3 percent against 14.8 percent in the 12 on placebo.

Does blood NAD+ mean brain NAD+?

The 2026 trial measured whole blood only, and its authors say so directly: "Whether the observed increase in blood NAD+ translated to an increase in brain NAD+ in this study is not known." The human evidence that oral precursors raise brain NAD+ is thinner than it is usually presented. It amounts to one acute magnetic-resonance-spectroscopy study in 10 healthy volunteers after a single 900 mg dose, in which cerebral NAD+ rose from 0.392 to 0.458 millimolar and which its authors call preliminary, plus one 12-person phase-I pharmacokinetic study in which cerebral NAD+ increased measurably only after about four weeks. A third often-cited dataset is a secondary analysis of 22 participants from the 2018 pilot, not an independent cohort. Neither of the two independent datasets included a cognitive outcome.

The injectable question, which is a different question entirely

Everything above concerns capsules. NAD+ itself, infused or injected, is a separate pharmacological situation: it bypasses gut absorption and first-pass metabolism entirely, so oral-precursor results cannot simply be transferred to it in either direction.

The honest summary of the intravenous literature is short.

The pharmacokinetic pilot

A 2019 study infused 750 mg of NAD+ over six hours in eight men, with three saline controls. Plasma NAD+ did not rise measurably until after the two-hour mark, then peaked around 400 percent above baseline at six hours. It measured blood and urine chemistry only, with no clinical or cognitive outcome, and it was run at a clinic that sells the infusion, at a dose the paper says reflects a common clinic regimen, and funded by an NAD+ research company together with a private research institute.

The one real randomised trial

A 2026 trial in 180 patients with heart failure from ischemic cardiomyopathy gave intravenous NAD+ at 10 mg per day for seven days on top of standard therapy. Its primary endpoint, change in ejection fraction at one month, was met at p equals 0.024 (45.44 versus 42.44 percent). Every clinical secondary endpoint, including hospitalisation and major adverse cardiac events, was a non-significant trend. Note the dose: 10 mg per day, far below the amounts used in wellness infusions.

The tolerability comparison

A 2026 retrospective review of records from a commercial infusion chain compared intravenous NAD+ with intravenous nicotinamide riboside over four consecutive days. The NAD+ recipients reported moderate to severe gastrointestinal symptoms, increased heart rate and chest pressure during infusion, and their infusions took an average of 97 minutes against 37 minutes. The review was conducted by people reviewing their own company's records.

What nobody has published

We looked specifically, and so did our fact-checkers. We could not locate any randomised controlled trial of infused or injected NAD+ with a cognitive endpoint, nor any randomised trial at the 500 to 750 mg doses used in wellness settings. That is an absence of evidence rather than evidence of absence, but it is worth knowing before anyone treats an infusion as a tested intervention.

The general lesson, which outlives this trial

The most transferable idea here has nothing to do with NAD+ specifically. It is the difference between a biomarker and an outcome.

A biomarker is a thing you can measure. An outcome is a thing that matters. Raising a biomarker is a necessary step for many interventions, and it is often the easiest step. It is also where a great deal of marketing stops, because the biomarker moves reliably and photographs well in a chart, while the outcome is slower, noisier and frequently refuses to move at all.

This trial is a clean demonstration: the biomarker did exactly what the theory required, doubling in twelve weeks with a very large effect size, and every outcome the researchers had pre-specified stayed flat. Whenever a compound is sold on the strength of a mechanism, that is the question worth asking: has anyone shown the outcome moving, or only the mechanism?

Our own framing

We sell NAD+ as a research compound, and this article is what the current human evidence supports: the precursors reliably raise the molecule in blood, nobody has yet shown a robust cognitive benefit in a randomised trial, and the injectable literature is thinner still. Our supplier batch reports for every product, from a third-party laboratory, are published on our CoA page, and the half-life reference table covers what is and is not known about the pharmacokinetics of the compounds we list.

Related research compounds

Longevity & Anti-Aginglongevity

Mitochondrial function, NAD+ metabolism, telomere maintenance

The compound in this article

NAD+longevity

Other longevity research compounds

MOTS-clongevity

Mitochondrial-derived signaling peptide (16 amino acids) that mimics the effects of exercise at the cellular level. Activates AMPK, improves glucose uptake, and enhances fat metabolism - a key tool in metabolic and longevity research.

Epitalonlongevity

Tetrapeptide (Ala-Glu-Asp-Gly) that activates telomerase, the enzyme responsible for maintaining telomere length. One of the most studied peptides in longevity research, developed by Prof. Khavinson at the St. Petersburg Institute of Bioregulation.

SS-31longevity

Mitochondria-targeted tetrapeptide (Elamipretide) that stabilizes cardiolipin and prevents ROS formation at the source.

FAQ

Sources

Martens CR, Decker KP, DeConne TM, et al. "A phase-II randomized controlled pilot study of nicotinamide riboside supplementation in older adults with amnestic mild cognitive impairment." Alzheimer's & Dementia 2026;22(7):e71605. PMID 42478598. https://pubmed.ncbi.nlm.nih.gov/42478598/
Orr ME, et al. GeroScience 2024;46(1):665-682. Randomised pilot of nicotinamide riboside in mild cognitive impairment. PMID 37994989. https://pubmed.ncbi.nlm.nih.gov/37994989/
Wu CY, Kupferschmid AC, Chen L, et al. Alzheimer's & Dementia: Translational Research & Clinical Interventions 2025;11(1):e70023. Eight-week nicotinamide riboside trial in subjective cognitive decline and mild cognitive impairment. PMID 39817194. https://pubmed.ncbi.nlm.nih.gov/39817194/
Martens CR, et al. "Chronic nicotinamide riboside supplementation is well-tolerated and elevates NAD+ in healthy middle-aged and older adults." Nature Communications 2018;9:1286. PMID 29599478. https://pubmed.ncbi.nlm.nih.gov/29599478/
Conze D, Brenner C, Kruger CL. Scientific Reports 2019;9:9772. Dose-dependent whole-blood NAD+ increases. PMID 31278280. https://pubmed.ncbi.nlm.nih.gov/31278280/
Dollerup OL, et al. American Journal of Clinical Nutrition 2018;108(2):343-353. Twelve-week nicotinamide riboside trial in obese, insulin-resistant men. https://pubmed.ncbi.nlm.nih.gov/29992272/
McDermott MM, et al. Nature Communications 2024;15:5046. Nicotinamide riboside in peripheral artery disease (NICE trial).
Yu X, Xu J, Cao J, et al. American Journal of Cardiovascular Drugs 2026;26(1):97-106. Randomised placebo-controlled trial of intravenous NAD+ in ischemic cardiomyopathy. PMID 40954388. https://pubmed.ncbi.nlm.nih.gov/40954388/
Reyna K, Heinzen G, Patel N, et al. Frontiers in Aging 2026;7:1652582. Retrospective tolerability comparison of intravenous NAD+ versus intravenous nicotinamide riboside. PMID 41704678. https://pubmed.ncbi.nlm.nih.gov/41704678/
Grant R, et al. Frontiers in Aging Neuroscience 2019;11:257. Pharmacokinetics of a single intravenous NAD+ infusion. PMID 31572171. https://pubmed.ncbi.nlm.nih.gov/31572171/
Lin Y, Zeidan RS, Lapierre-Nguyen S, et al. GeroScience 2025;47(6):6895-6908. Nicotinamide riboside combined with exercise in hypertensive middle-aged and older adults. PMID 40770531. https://pubmed.ncbi.nlm.nih.gov/40770531/
Christen S, Redeuil K, Goulet L, et al. Nature Metabolism 2026;8(1):62-73. Comparison of three NAD+ boosters on circulatory NAD and microbial metabolism in humans. PMID 41540253. https://pubmed.ncbi.nlm.nih.gov/41540253/
Nanga RPR, Wiers CE, Elliott MA, et al. Magnetic Resonance in Medicine 2024;92(6):2284-2293. Acute nicotinamide riboside supplementation increases human cerebral NAD+ levels in vivo. PMID 39044608. https://pubmed.ncbi.nlm.nih.gov/39044608/
Berven H, Svensen M, Eikeland H, et al. iScience 2026;29(3):114764. Phase-I pharmacokinetics of NAD augmentation in blood and brain with oral precursors. PMID 41858901. https://pubmed.ncbi.nlm.nih.gov/41858901/
Vreones M, Mustapic M, Moaddel R, et al. Aging Cell 2023;22(1):e13754. Secondary analysis of neuronal-origin plasma extracellular vesicles from the 2018 crossover trial. PMID 36515353. https://pubmed.ncbi.nlm.nih.gov/36515353/
Yoshino M, Yoshino J, Kayser BD, et al. Science 2021;372(6547):1224-1229. Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women. PMID 33888596. https://pubmed.ncbi.nlm.nih.gov/33888596/
Brenner C. Science 2021;373(6554):eabj1696. Technical comment on the randomisation of the 2021 nicotinamide mononucleotide trial. PMID 34326206. https://pubmed.ncbi.nlm.nih.gov/34326206/

Research disclaimer: All content serves scientific information only. NAD+ and its precursors are not approved medicines and the products we supply are not intended for human consumption. Trial doses are reported as study facts, never as recommendations.