NR produced a promising but preliminary walking-distance result in people with peripheral artery disease. In the 90-person NICE randomized trial, the difference versus placebo at six months was 17.6 meters on a six-minute walk test. The study used a prespecified, relatively permissive statistical threshold appropriate to its phase II screening purpose. A larger confirmatory trial is needed before treating that result as settled treatment evidence. [1]
Peripheral artery disease, or PAD, limits blood flow to the legs and can impair walking. This was a disease-specific study, not a test of athletic performance in healthy adults. Our NAD and exercise article addresses that different question.
The main comparison
| Study and participants | Design and outcome | Finding | Main limit |
|---|---|---|---|
| NICE; n = 90 with PAD [1] | Double-blind randomized trial; NR, NR plus resveratrol, placebo; six months | NR vs placebo: +17.6m on six-minute walk; one-sided 90% CI lower bound +1.8m. | Phase II threshold p < 0.10; small study; no multiplicity adjustment. |
| Adherent subgroup [1] | Participants taking at least 75% of pills | NR vs placebo +31.0m; combination +26.9m. | Post hoc subgroup, not the primary randomized comparison. |
| Szarvas et al.; eight participants [2] | Four-week open-label pilot; vascular and cognitive outcomes | Some favorable before-and-after findings. | No control group; cannot establish causation or replicate the walking result. |
Participants were randomized to NR, NR plus resveratrol, or placebo. The primary question was change in six-minute walk distance at six months. This test measures how far someone walks in a fixed time; it is not the same as daily activity recorded by a wearable, a symptom questionnaire or maximal performance on a treadmill.
The NR group increased its walk distance by about seven meters, while the placebo group declined by about 10.6 meters. The 17.6-meter treatment contrast includes both patterns. Saying that everyone taking NR walked 17.6 meters farther would confuse the between-group estimate with an individual’s response. [1]
Why the confidence interval matters
The authors reported a one-sided 90% confidence interval with a lower bound of approximately 1.8 meters and an open upper end, and a one-sided p value of 0.08. They had prespecified p < 0.10 as the threshold. This is more permissive than the conventional two-sided 0.05 standard familiar from many confirmatory trials. It was a deliberate design choice to avoid dismissing a potentially useful intervention too early. [1]
The finding met that study’s rule. It should not be presented as if it met a different rule. The trial also did not adjust for multiple comparisons. Together, those details make “promising signal” a better description than “proven walking treatment.” Our evidence overview explains how study phase and endpoints affect confidence.
Resveratrol did not add a demonstrated benefit
The NR-plus-resveratrol group’s six-month difference versus placebo was about 3.65 meters and did not meet the trial’s significance threshold (p = 0.38). The study did not show that adding resveratrol improved the main outcome compared with NR alone. [1]
This is a useful counterexample to the idea that two plausible ingredients must work better together. It is also distinct from studies of NMN combined with resveratrol; the molecules and populations differ. The NMN and resveratrol comparison addresses the broader supplement claims without treating them as interchangeable with this PAD trial.
The larger adherence result was exploratory
Among people who took at least 75% of their study pills, the reported differences versus placebo were 31.0 meters for NR and 26.9 meters for NR plus resveratrol. That subgroup analysis was post hoc. People who adhere to treatment may also differ in health, behavior or ability to attend visits. Restricting the analysis after randomization can weaken the protection randomization provides against those differences. [1]
The 31-meter figure therefore should not replace 17.6 meters as the headline result. Nor does the comparison with walking gains reported in other research prove that NR substitutes for supervised exercise. NICE did not directly randomize participants between NR and an exercise program.
What the eight-person pilot adds
A later four-week, open-label study gave NR to eight people with PAD and assessed vascular and cognitive measures. It reported favorable changes in some outcomes and performed laboratory experiments using participants’ serum. Without a placebo group or blinding, it cannot separate a treatment effect from repeat-testing effects, ordinary variation or other changes over time. [2]
Its role is to suggest mechanisms and improve the design of future controlled studies. It does not independently confirm NICE’s walking endpoint, and it does not establish a cognitive benefit. For controlled cognitive findings, see the memory-trial comparison.
What would make the evidence more convincing?
A larger trial should reproduce the walking benefit with a clear confirmatory analysis, assess outcomes that matter to daily life and characterize safety over longer follow-up. NICE’s small sample, limited biopsy and treadmill data, and eligibility run-in all constrain generalization. The result supports further research in PAD; it does not justify changing established vascular care or buying an NAD⁺ injection instead of the oral precursor studied. [1]