Nicotinamide riboside raises NAD⁺ in humans, reproducibly and dose-dependently. Almost every other benefit it is sold for has now been measured in a randomized trial, and most of those trials came back empty.
That combination is what makes NR worth writing about: it is the one compound in this field with enough properly designed trials that you can go claim by claim and find an answer for each. This page does exactly that. What NR is, and how its pharmacology was worked out, is a separate page; this one is only about outcomes.
Metabolic health: measured with the reference method, and null
The claim that NR improves how your body handles sugar and fat came from rodent work, and it was tested properly. Forty healthy, sedentary men with a BMI above 30, aged 40 to 70, were randomized to 1,000 mg of NR twice daily or placebo for twelve weeks, with insulin sensitivity measured by hyperinsulinemic euglycemic clamp — the reference technique, not a fasting proxy. Four measures came back unimproved — insulin sensitivity, endogenous glucose production, glucose disposal and glucose oxidation — and so did resting energy expenditure, lipolysis, lipid oxidation and body composition [1].
A crossover trial reached almost the same place by a different route. Thirteen healthy overweight or obese men and women took 1,000 mg of NR daily for six weeks against placebo, with clamps, magnetic resonance spectroscopy, muscle biopsies and ex vivo mitochondrial measurement. NR did something real and specific: fat-free mass rose (62.65% ± 2.49% versus 61.32% ± 2.58%, a change of 1.34% ± 0.50%, p = 0.02), sleeping metabolic rate rose, and skeletal muscle acetylcarnitine rose (4558 ± 749 versus 3025 ± 316 pmol/mg dry weight, p = 0.04). And then: no effect on insulin sensitivity, mitochondrial function, liver or intramyocellular fat, cardiac energy status, ejection fraction, ambulatory blood pressure, inflammatory markers or energy metabolism [2].
Read those two together and you have the shape of the whole NR literature. The molecule is doing something measurable inside muscle. The clinical outcomes people buy it for are not moving.
Cognition: the trial designed to find it, in the people most likely to show it
A phase II randomized, double-blind, placebo-controlled trial gave NR or placebo for twelve weeks to older adults with amnestic mild cognitive impairment — the population with the most room to improve. Forty-two participants completed (NR = 22, placebo = 20). Blood NAD⁺ doubled in the NR group. Adherence was similar and there were no serious adverse effects. There were no improvements in cognitive function, which was the primary outcome, nor in total cerebral blood flow or blood pressure; exploratory analysis raised the possibility of regional cerebral blood flow changes, particularly in the hippocampus [3].
The fatigue and brain-fog claim has its own trial now too. A 24-week, double-blind, placebo-controlled study randomized 58 people with long COVID to NR 2,000 mg a day or a placebo lead-in. NAD⁺ rose 2.6- to 3.1-fold within five to ten weeks and stayed elevated. There were no significant between-group differences in cognition (p-values 0.47 to 0.74), fatigue severity (p = 0.59), sleep quality (p = 0.69), anxiety (p = 0.84) or depression (p = 0.20) [4].
Post-hoc analyses of that trial, pooling everyone during their NR phase and comparing against their own baseline, did find changes in executive function, fatigue, sleep and depression. The authors present those as exploratory and unadjusted for multiplicity, which is the correct framing, and it is also the framing a marketing page will drop. A within-group before-and-after is exactly what a placebo arm exists to discount.
Muscle and physical function: pooled, and still nothing
A 2025 systematic review and meta-analysis pooled the randomized trials of NR and NMN against placebo in adults with a mean age above 60. For NR specifically, the narrative synthesis found a longer six-minute walking distance in people with peripheral artery disease — a genuine signal in a specific disease — but lower short physical performance battery scores and a slower five-time chair stand in people with mild cognitive impairment. The authors concluded that current evidence does not support NR or NMN for preserving muscle mass and function in that age group [5].
Note the direction of that second finding. NR did not merely fail to help in one population; the pooled numbers point the other way. One result in one review is not a warning, and it is the kind of detail that belongs in front of anyone buying NR for strength. The wider picture for training and performance is its own page, and it does not improve.
Skin and nerve: a phase 2 that measured tissue, not blood
A placebo-controlled, double-blinded phase 2 study used a validated capsaicin model of cutaneous nerve degeneration and regeneration to ask whether oral NR protects small sensory nerve fibers. The result is unusual and worth knowing: NR supplementation did not raise plasma NAD⁺ at all, though it produced a small increase in NAD⁺ measured in skin samples. It did not prevent capsaicin-induced degeneration of epidermal nerve fibers, and there was no difference in reinnervation at day 90. The authors state that at the doses used, oral NR cannot be recommended to prevent neuropathy or improve nerve regeneration [6].
That plasma finding is the reason to read the paper rather than the headline. Target engagement is usually the one thing NR can be relied on to do, and here it did not happen in blood. Biomarker responses vary between studies and populations more than the marketing admits.
Where NR did produce an outcome
It would be dishonest to end the list without this one. A double-blind, randomized, crossover, placebo-controlled trial in patients with Werner syndrome — a rare hereditary condition in which age-related disease arrives decades early — gave 1,000 mg of NR or placebo daily for 26 weeks before crossing over for a further 26. There were no serious adverse events. Arterial stiffness measured by the cardio-ankle vascular index improved, skin ulcer area decreased, heel pad thinning showed a declining trend, and blood creatinine fell significantly [7].
That is a real clinical result on real endpoints, in the longest NR exposure published. It is also a rare disease defined by a DNA repair defect, in patients with documented NAD⁺ depletion — a population where replacing a depleted cofactor has an obvious rationale that does not transfer to a healthy 45-year-old. The pattern recurs across this field: where NAD⁺ biology is demonstrably broken, repletion sometimes does something. Where it is merely age-typical, it mostly does not.
What the census says
A PRISMA-guided systematic review published in 2026 swept human and rodent intervention studies of NAD-related compounds from January 2010 to October 2025 and identified 113 eligible studies, of which 33 were human intervention studies and 28 randomized. Its summary of the human evidence is that oral NR and NMN consistently demonstrated biochemical target engagement and were generally well tolerated, while effects on functional, metabolic, vascular and other healthspan-relevant outcomes were heterogeneous and often null or endpoint-specific [8].
That is the same conclusion this page reached claim by claim, written by people with no stake in the answer. It is not a reason to think NR does nothing — target engagement is real and rare in this market. It is a reason to buy it for what has been shown rather than for what is implied, and to pay accordingly: an oral precursor costs a fraction of an infusion that has no outcome trial of any kind behind it. If you are choosing between the two precursors, NMN versus NR compares them directly, and the product whose ingredient most of this research used is where the trials above got their capsules.