Preclinical vs. Human NMN Evidence

Why this page exists

Almost every dramatic claim about NMN traces back to a study that was never conducted in humans. That doesn't make those studies worthless — it makes them a different category of evidence that answers a different question. This page explains the distinction this entire hub relies on, and shows, with specific NMN examples, why a finding in one category does not automatically transfer to the other.

Two categories, defined

Preclinical evidence includes cell culture (isolated human or animal cells grown in a dish), isolated tissue experiments, animal models (most commonly laboratory mice), and other mechanistic experiments conducted before, or instead of, testing in humans. Human evidence includes observational studies (watching what happens to people without intervening), pharmacokinetic studies (tracking how a substance moves through the human body), randomized controlled trials, other controlled intervention studies, and meta-analyses that pool multiple human studies together. Preclinical research is how almost all biomedical hypotheses are generated and initially tested; human evidence is what's required to know whether a hypothesis holds in the organism it's ultimately meant to help.

Why findings can fail to translate

A result observed in a preclinical model can fail to hold in humans for several distinct, well-documented reasons, and more than one often applies at once:[1]

• Species differences in relevant biology — enzyme kinetics, transporter expression, metabolic pathway usage, and lifespan-related biology all differ between mice and humans.
• Dose differences — doses used in animal studies, especially on a per-kilogram basis, are frequently far higher than what is used or tolerable in humans.
• Route differences — injected, intraperitoneal, or diet-mixed administration in animals is not the same as human oral supplementation, and can bypass steps (like first-pass gut and liver metabolism) that matter in humans.
• Metabolic differences — a compound's breakdown pathway, rate, and byproducts can differ meaningfully between species.
• Lifespan and aging-biology differences — a mouse study demonstrating a lifespan effect is conducted over roughly 2–3 years in an organism that ages far faster than humans, under lab conditions that don't reproduce human genetic diversity, comorbidities, or environment.
• Disease-model differences — genetically engineered or chemically induced disease models in animals are simplified representations of human disease, not the disease itself.
• Outcome-measurement differences — a behavioral or biochemical readout in an animal (a maze-navigation time, a tissue biomarker) is not always a valid proxy for the corresponding human outcome (a cognitive test score, a clinical diagnosis).

Systematic reviews of translational failure across many disease areas have found that methodological weaknesses in the original animal studies — small sample sizes, lack of randomization or blinding, publication bias toward positive results — further widen the gap between what preclinical research reports and what later replicates in human trials.[1]

Four concrete NMN examples

Lifespan. Specific genetic mouse models (such as SIRT6-overexpressing mice) have shown extended lifespan, and long-term NMN administration in mice has been associated with mitigation of age-associated physiological decline (see NMN and Aging). No human NMN trial has tested, or could feasibly test on any relevant timescale, an effect on human lifespan. This hub does not present mouse lifespan findings as evidence of a human lifespan effect.

NMN cellular transport. A 2019 paper proposed that the transporter Slc12a8 allows NMN to enter mouse intestinal cells directly.[2] A rebuttal published in the same journal months later found no evidence that Slc12a8 functions as an NMN transporter, disputing the original claim.[3] This dispute is specifically about a proposed mechanism in mice; the human route by which NMN or its breakdown products enter cells remains separately unresolved, as covered in full on NMN Transport and Cellular Uptake. A contested mouse mechanism should not be cited as an established human one.

CD38 and aging. Mouse studies have shown that genetically removing CD38, an NAD+-consuming enzyme, prevents age-related NAD+ decline and mitochondrial dysfunction in those animals.[4] This is a real, peer-reviewed finding about mouse genetics, covered in full on NAD+ and CD38. No human study has causally demonstrated that CD38 activity, rather than correlating with, drives age-related human NAD+ decline — the human evidence for CD38's role remains observational and mechanistic-by-extension, not causally established the way the mouse knockout evidence is for mice.

Cognition. A peer-reviewed mouse study found that NMN supplementation rescued cerebromicrovascular endothelial function and improved measures of spatial working memory in aged mice (see NMN and Cognitive Research). The one human trial that directly tested cognitive function after NMN found no significant effect on two validated instruments. This hub does not use the mouse cognitive finding to imply, soften, or offset that human null result — they are reported as separate findings from separate categories of evidence.

What this page is not saying

None of this demeans animal or cell research. Preclinical work is how CD38's role in NAD+ metabolism was discovered, how the NAD+ salvage pathway was mapped, and how NMN became a plausible candidate for human study in the first place — without it, there would be no hypothesis to test in humans at all. The point is narrower and more specific: a preclinical finding establishes that something is biologically possible and worth testing in humans, not that it has been shown to happen in humans. This hub labels every claim PRECLINICAL or HUMAN accordingly, and readers should treat the label, not the impressiveness of the result, as the guide to how much weight the finding should carry for a human decision.

Related reading

For the complete human trial record, see NMN Human Clinical Trials. For the disputed transport mechanism, see NMN Transport and Cellular Uptake. For CD38, see NAD+ and CD38. For the honest state of human aging evidence, see NMN and Aging.