Does NR Do Anything Beyond Raising NAD+? What the Trials Show

Creator: Erin Rose
Published: 2026-07-18

By Erin Rose · Updated July 18, 2026 · Methodology

Educational summary — not medical advice. No NR trial has proven a disease treatment, lifespan, or anti-aging benefit. This statement has not been evaluated by the FDA. This product is not intended to diagnose, treat, cure, or prevent any disease.

The honest answer

Reliably, yes, at the one thing it's built to do — raise NAD+. Beyond that, not yet demonstrated. Five independent, non-industry-funded RCTs each tested a different organ system: cardiovascular (Martens), whole-body metabolism (Dollerup 2018), skeletal-muscle mitochondria (Dollerup 2020, Elhassan), and heart-failure function (Wang). All five confirmed NAD+ rose. All five found the functional or clinical endpoint null or explicitly underpowered. NADPARK, in Parkinson's patients, is a partial exception with a correlational signal — but it's a 30-day Phase I safety trial, not proof of a treatment effect.

Five independent trials, five organ systems, five null functional results

This is the pattern that matters most in NR's evidence base, and it plays the role NMN's 2024 meta-analysis (Chen et al.) plays on that cluster's hub — except here the honest counterweight isn't one pooled analysis, it's five separate, independently-run, non-industry-funded RCTs that each tested a different organ system and each landed in the same place: NAD+ up, functional endpoint null or underpowered.

The five null (or underpowered-null) functional trials, verified PMIDs
Study	Organ system tested	Dose & duration	NAD+ result	Functional result
Martens 2018 PMID 29599478	Cardiovascular (blood pressure, arterial stiffness)	1000 mg/day, 6 weeks × 2 crossover periods, n=24	PBMC NAD+ metabolome rose significantly	Systolic BP and pulse-wave velocity: NOT significant after multiple-comparison correction, whole group. Exploratory subgroup trend only, "no statistical inferences" per authors.
Dollerup 2018 PMID 29992272	Whole-body metabolism (insulin sensitivity)	2000 mg/day, 12 weeks, n=40 obese men	(NAD+ not the primary metabolic readout here)	NO improvement in insulin sensitivity, glucose disposal, resting energy expenditure, lipolysis, or body composition.
Dollerup 2020 PMID 31710095	Skeletal-muscle mitochondria	Same cohort, muscle-biopsy sub-study	NO change in muscle NAD+ despite a 14% NAMPT protein drop	NO change in mitochondrial respiratory capacity, abundance, or morphology.
Elhassan 2019 PMID 31412242	Skeletal-muscle mitochondria (independent cohort)	1000 mg/day, 21 days, n=12 aged men	Muscle NAD+ metabolome rose (metabolomics)	Mitochondrial bioenergetics (actual respiration) unaltered, despite transcriptomic gene-expression changes.
Wang 2022 PMID 36644285	Heart failure (HFrEF) function	2000 mg/day, 12 weeks, n=30	Whole-blood NAD+ ~doubled	NO significant difference in 6-min walk distance, LVEF, or quality of life. Authors state the trial was underpowered (~172 participants needed for LVEF, ~6x this enrollment).

The partial exception: NADPARK, and why it isn't proof NR treats Parkinson's

The NADPARK study (Brakedal et al. 2022, Cell Metabolism) is the one trial in this evidence base with a positive-leaning clinical signal, and it deserves the most careful handling of any citation here. Thirty newly-diagnosed, treatment-naive Parkinson's patients took 1,000 mg/day of NR for 30 days in a randomized, double-blinded Phase I design. Cerebral NAD+ rose significantly, measured directly by 31P-MRS, and correlated with altered cerebral metabolism and what the authors describe as "mild clinical improvement." Inflammatory cytokines decreased in both serum and cerebrospinal fluid.

Read exactly what that is: a Phase I safety and mechanistic trial, n=30, 30 days, with a correlational, exploratory clinical observation — not a powered efficacy trial designed or sized to prove a treatment effect. The study authors themselves frame NR as a candidate that warrants larger trials, not a demonstrated therapy; an open-label extension study (NOPARK) is following up specifically because this result needs replication in a design built to test it. This page states that plainly: NADPARK is real, published, peer-reviewed evidence that NR reaches the brain and moves NAD+ there, paired with a genuinely interesting but unproven clinical correlation. It is never accurate to summarize NADPARK as "NR treats Parkinson's disease," and no page on this site should ever be read that way.

The verdict NR reliably does the one thing it's designed to do — raise NAD+ — and has the cleanest regulatory paper trail of any NAD+ precursor on the market. Whether raising NAD+ this way changes how a person actually feels or functions is still an open question in every organ system tested so far. Five independent trials looked for a functional payoff and didn't find one (or found the trial too small to tell); one trial found a hint worth following up on, in a disease population, at Phase I. That's the honest state of the evidence in 2026.

Why this reads differently from NMN's meta-analysis, and why that's not a ranking claim

NMN's null counterweight is a single 2024 systematic review pooling 8 RCTs (Chen et al.) that found no significant glucose or lipid benefit. NR's null counterweight is structurally different: five separate, independently-run trials, each with its own population and organ system, each individually null or underpowered on its functional endpoint. Neither pattern is "worse" or "better" evidence than the other in a simple sense — a pooled meta-analysis and five converging independent null results are different kinds of honest evidence, and both point the same direction: a reliably-moved biomarker without a demonstrated downstream benefit. Consistent with the live NMN vs. NR comparison, this page does not reframe NR as having stronger or superior efficacy evidence to NMN — NR's trial volume here is comparable in count (8 PMIDs) but individually smaller (n=8-40 per trial, 21 days to 12 weeks) than NMN's. The honest contrast between the two compounds is regulatory cleanliness, not efficacy superiority in either direction.

Frequently asked questions