This 2019 pilot reports random allocation of eleven healthy men: eight received direct NAD+ and three received saline control. Several plasma metabolites and urinary excretion rates changed during the eight-hour assessment. The findings describe short-term handling of selected chemicals; they do not show improved energy, cognition, longevity, addiction outcomes or tissue NAD+ replenishment.
The important distinction is between what was measured and what the authors inferred. No statistically significant early plasma increase is not proof of zero change or complete uptake by cells. Likewise, a later increase in a blood measurement is not evidence of clinical benefit. The primary Frontiers paper supplies the methods, results, figures and disclosures summarized below.
Study at a glance
| Feature | What the publication reports |
|---|---|
| Participants | Eleven healthy men aged 30–55 years, with body-mass index below 30 |
| Allocation | Randomized to eight active participants and three saline controls |
| Study material | Direct NAD+ supplied for a supervised research protocol; not an oral precursor |
| Observation | Six-hour exposure, with sampling through eight hours from the start |
| Measurements | Plasma NAD+, nicotinamide, ADPR, methyl-nicotinamide and NMN; urinary NAD+, nicotinamide and methyl-nicotinamide |
| Safety assessment | Clinical observation during exposure and selected laboratory comparisons at eight hours |
| Identifiers | PMID 31572171; PMCID PMC6751327; DOI 10.3389/fnagi.2019.00257 |
| Ethics | William Carey University IRB protocol 2017-12; an ethics identifier, not a trial-registration number |
The historical exposure description identifies the experiment. It is not a preparation, administration or treatment recommendation.
Randomized does not mean fully documented
The Methods explicitly say participants were randomized. Calling this study nonrandomized would contradict that account. However, the paper does not describe sequence generation, allocation concealment or masking. These are reporting gaps, not proof that a particular method failed. The distinction is explained in randomization, masking and control groups.
Participants were recruited through advertisements and screened for selected health conditions and medication use. The narrow population and three-person control group limit generalization. Baseline differences are visible in some plasma measurements; a very small control group cannot reliably balance all relevant characteristics.
Figure captions use n=8 and n=3, but the paper does not supply a participant flow diagram or detailed endpoint-by-endpoint missing-data account. Those labels are not proof that every planned sample was obtained or that a verified intention-to-treat analysis was performed. Repeated measurements from one participant also do not become additional independent participants.
Diet and fasting were standardized. A Methods statement about dextrose in the infusion solution does not clearly establish whether the control was composition-matched. This unresolved detail matters to interpretation; the paper's preparation and calorie wording is not reproduced as practical guidance.
What the sampling can resolve
Blood was sampled before exposure and at 30, 60, 120, 360 and 480 minutes. There is a four-hour gap between the two-hour and six-hour measurements. Consequently, the data cannot locate the exact onset of a later rise or establish a precise moment at which a metabolic pathway became saturated.
After a baseline urine sample, urine was collected over intervals associated with 30 minutes, two hours, six hours and eight hours. The urinary percentage comparisons in the Results use the 30-minute collection as their reference, not an untreated pre-exposure baseline.
The laboratory used liquid chromatography–tandem mass spectrometry. Isotopically labeled internal standards were used for analytical measurement; they were not administered human tracers. This is therefore not a tracer study that follows the administered molecule into particular tissues or measures all its metabolic destinations.
Plasma results: keep the comparator attached
The Results report no significant early plasma increase, followed by several significant findings at six hours. Selected six-hour results are reproduced with their stated comparisons:
| Plasma analyte | Reported change from active-group baseline at six hours | Six-hour comparison with saline control |
|---|---|---|
| NAD+ | +398%; P<0.0001 | P<0.001 |
| Nicotinamide, or NAM | +409%; P<0.0001 | P<0.001 |
| ADPR | +393%; P<0.0001 | P<0.0001 |
These percentages describe changes from baseline, not the size of an active-minus-control treatment effect. An increase of 398% means approximately 4.98 times the starting value, not 3.98 times it. The percentages do not measure how much administered material reached cells.
At eight hours, NAD+ remained elevated relative to baseline and control. NAM and ADPR were no longer significantly different from control at that time, although ADPR remained significantly above its own baseline. Those are different comparisons: a positive within-group result does not establish a positive between-group result.
Methyl-nicotinamide also rose, with significant baseline and control comparisons at six and eight hours. The text's six-hour percentage wording and the plotted relationship are not straightforwardly consistent, so this summary does not choose a disputed exact percentage. NMN was reported significantly elevated only at eight hours; this does not establish a beneficial intracellular response.
Urinary findings and numerical caution
Urinary NAD+ and methyl-nicotinamide excretion rates rose at six hours relative to the 30-minute reference and fell by eight hours. The Results also report a significant six-hour NAD+ comparison with saline control. Urinary nicotinamide showed no significant change over time or difference from control. That does not mean nicotinamide was absent from urine.
Figure 2 reports excretion rates in nanomoles per hour. These are not urine concentrations, renal clearance estimates or tissue uptake measurements. A change in excretion cannot by itself establish where material was metabolized, how much entered cells or whether renal handling was beneficial.
Some narrative urinary percentages do not map cleanly to the plotted relationships. Exact disputed increases and declines are omitted here rather than corrected by reading coordinates from pixels. The broad time pattern is retained; the apparent numerical inconsistency remains a limitation of the report.
Precision and statistical limitations
The publication describes two-way analysis of variance with Bonferroni post-hoc comparisons and a P<0.05 threshold. It does not explain how the model handled correlations among serial observations from the same person. Naming an adjustment also does not establish that every outcome, comparison and analysis belonged to one controlled testing family.
Table 1 explicitly labels its summaries as means and standard deviations. The error-bar type in the figures is not defined in the captions or Methods read for this review, so those bars should not be relabeled as standard deviations, standard errors or confidence intervals.
The NAM–ADPR correlation was calculated between group averages across time points. It is not a correlation among eleven independent participant responses and does not establish a causal metabolic pathway. With eight active participants, three controls and multiple measurements, the results warrant replication and better-defined uncertainty. See confidence intervals, P values and multiple testing.
Proposed pathways are not measured tissue uptake
The authors interpret the early pattern as rapid removal of NAD+ from plasma and propose contributions from extracellular metabolic enzymes and cellular uptake. The pathway diagrams describe possible fates; the experiment did not measure their individual fluxes.
The distinction matters even if a mechanism is biologically plausible. The study did not directly measure tissue NAD+, tissue ATP production, enzyme-specific activity or the proportion of administered material reaching an intracellular compartment. Plasma appearance, distribution, metabolism and excretion are different processes, and the observed pattern does not uniquely identify their contributions.
In particular, statistical nonsignificance during the first two hours does not demonstrate complete removal, zero accumulation or a precise saturation time. The paper's stronger mechanistic language should be read as interpretation, not as a direct measurement of those quantities.
Safety observations are narrow
No adverse events were observed during the six-hour exposure in either cohort. This is a report about eight active participants and three controls over a short observation period, not proof that adverse events cannot occur.
At eight hours, the active group had statistically significant within-group changes in bilirubin, GGT, LD and AST. The authors judged the changes not clinically significant. The small control group showed no significant laboratory changes, but “significant in one group” and “not significant in the other” do not establish a significant difference between groups.
The discussion suggests improved hepatic integrity. Selected short-term laboratory changes do not establish liver protection, and this study did not test that clinical claim. Uncommon, delayed, repeated-exposure or formulation-specific harms remain unresolved. The evidence does not support general safety assurances.
Funding, registration and material identity
NAD+ Research Inc. and the Australasian Research Institute jointly funded the work. It was conducted at Springfield Wellness Center, and Archway Apothecary supplied the study material without charge. Richard Mestayer was a director of NAD+ Research Inc. and the center's medical director; Susan Broom received consulting fees from NAD+ Research Inc. The remaining authors declared no relevant commercial or financial conflicts. These disclosures provide context, not an automatic verdict on the measurements.
The paper says datasets are available on request; no participant dataset was obtained for this review. It reports ethics approval but no public trial-registration identifier. No original protocol, statistical analysis plan or registry history was reviewed here, so prospective endpoint commitments have not been verified. That is not a claim that no registration or protocol exists anywhere. See trial registration and outcome switching.
The experiment studied direct NAD+ under one research protocol. It did not compare NAD+ with NR, NMN, niacin, nicotinamide, NADH or an oral supplement. A shared name also cannot establish another material's identity, purity, formulation, stability or sterility. The paper does not validate a catalog lot, compounded preparation or commercial service.
Frequently asked questions
Was this a randomized study?
The Methods report randomized allocation to eight active and three control participants. Sequence generation, concealment and masking are not described, so the quality of those procedures cannot be verified from this paper alone.
Did NAD+ remain unchanged for the first two hours?
No statistically significant early increase was reported. This does not prove zero change, complete tissue uptake or a fully measured balance between input and removal.
What did the six-hour increases establish?
They establish reported changes in selected plasma measurements and specified comparisons with baseline or control. They do not establish clinical benefit or the amount of NAD+ delivered into tissues.
Does unchanged urinary nicotinamide prove complete reabsorption?
No. A nonsignificant excretion-rate result does not by itself identify a renal mechanism. Nicotinamide was measured in urine, and the study did not directly quantify complete tubular reabsorption.
Did the study show improved energy or longevity?
No patient-centered efficacy endpoint was measured. Plasma or urinary metabolites are not validated substitutes for those outcomes in this experiment.
Did the absence of observed adverse events establish safety?
No. The exposure and laboratory windows were short, and only eight participants received NAD+. General, delayed or repeated-exposure safety was not established.
Can these results be transferred to a different NAD+ product or precursor?
No. This study did not establish equivalence across materials, formulations, routes or precursor compounds, and it cannot authenticate another preparation.
Related methods and source review
For context, use the Study Library, how to read a research study and surrogate endpoints versus clinical outcomes.
Sources: the PMC record, PubMed record and DOI identifier identify the same paper as the Frontiers full text above; they are not independent replications.
Editorial evidence review: September 3, 2026. This independently worded summary is based on the full publisher article, its HTML Table 1 and actual Figures 1–3. Figure 4 was reviewed as a caption only. The publisher text was accessible; separate PMC/PubMed access limits were retained, and no exhaustive correction/retraction search, participant-data reanalysis or independent specialist certification is claimed. The source is Grant, Berg, Mestayer, Braidy, Bennett, Broom and Watson (2019), Frontiers in Aging Neuroscience 11:257, published under CC BY. Report a factual correction through Contact.