Does NMN Cause Cancer? What the Research Shows

Written by David Roberts, MPH, Co-Founder and Managing Partner, Mara Labs

Short viral video can move faster than a correction - especially when it instills fear. A physician says a supplement causes cancer, cites a recent study, and the claim reaches a few million people before anyone reads the paper.

The papers behind this particular claim are real. What they found is not what the claim says they found, and the gap between the two is a specific, teachable error about study design.

Does NMN cause cancer in humans?

No published human study has shown that NMN causes cancer. That is the direct answer, and it holds before any qualification.

Two statements are true at the same time, and most of the argument online comes from holding one and discarding the other. There is no human evidence that NMN causes cancer. There is also not enough long-term human data to call it proven safe over decades. [4]

Everything between those two statements is a question about what the available studies were built to measure.

What did the study behind the viral claim find?

Two recent papers are the likely source, and neither tested whether NMN causes cancer.

The first, published in Cancer Letters in 2026 by a Case Western Reserve team, found that NMN, nicotinamide riboside, and nicotinamide protected pancreatic cancer cells against three chemotherapy drugs. [1]

The cancer cells were already there. The finding concerns chemotherapy resistance in existing tumors, and the authors explicitly state that their work does not suggest that these supplements are dangerous for healthy people.

The second, published in Biosensors and Bioelectronics in 2023, was primarily an imaging paper. Researchers built a bioluminescent probe to watch nicotinamide riboside uptake in living mice. [2]

They reported that breast cancer cells took it up at high rates, with increased tumor burden and spread to the brain in supplemented animals.

Three design details matter. The mice were immunocompromised, meaning they lacked the immune surveillance that clears abnormal cells in a healthy body.

The tumor cells were injected, so the cancer was placed there by the researchers. And the university that published the press release issued a correction weeks later, clarifying that this was an imaging study in animals.

What is the difference between causing cancer and feeding one that already exists?

These are two different biological questions, and the studies behind the viral claim answer only the second one.

Causing cancer means turning a normal cell into a malignant one. That generally happens through genotoxicity: the compound damages DNA, produces mutations, breaks chromosomes, or interferes with repair, and a damaged cell eventually escapes normal growth control.

Feeding a tumor assumes the malignant cell already exists. NAD+ is a coenzyme every living cell uses for energy production, redox balance, and DNA repair through the PARP enzymes. A rapidly dividing tumor cell needs more of it than a resting normal cell does.

Supplying more NAD+ precursor does not create a cancer cell. It can, in principle, leave an existing one better fueled and better able to repair the DNA damage that chemotherapy is deliberately inflicting. [1][12]

That is the mechanism the pancreatic paper reports. It explains how the same molecule can be unremarkable in a healthy person and unhelpful in someone on active treatment.

Has NMN been tested for its ability to damage DNA?

Yes, and it came back negative across the standard battery. The European Food Safety Authority reviewed four assays in its 2026 opinion on beta-NMN. [3]

Those were a bacterial reverse mutation test, an in vitro micronucleus assay, an in vitro chromosomal aberration test, and an in vivo bone marrow chromosomal aberration test in mice.

The panel's conclusion was that there are no concerns regarding genotoxicity. A 90-day rat study set a no-observed-adverse-effect level of 400 mg per kg of body weight per day, and EFSA judged beta-NMN safe up to 300 mg per day in adults. [3]

Here is the limit of that evidence, stated plainly. No two-year rodent carcinogenicity bioassay has been published for NMN. "Not genotoxic" is supported. "Not carcinogenic" is not a claim the data can carry, and anyone making it is going past the research.

How long have people taken NMN in studies?

Not long, and not in large numbers.

The longest human NMN trial ever published ran 24 weeks, in 14 older men with diabetes and reduced physical performance, seven per arm. [5] The largest ran 12 weeks in 108 older Japanese adults. [6]

A 2023 dose-ranging trial gave 300, 600, or 900 mg daily to 80 healthy middle-aged adults for 60 days and reported no safety signals. [7]

Pooled across the literature, a 2026 systematic review covering 15 randomized trials found no increase in overall, serious, or withdrawal-related adverse events, with a safety pool of 383 participants. [4]

The reviewers were careful to add that this should not be read as excluding rare, delayed, or long-term risks.

No human NMN trial has ever measured cancer as an endpoint. Detecting a change in cancer incidence requires thousands of people followed for years. A 14-person study running 24 weeks cannot find such a signal, and it cannot rule one out either.

Do any studies show NMN slowing tumors?

Some report exactly that, and most are weaker than the internet claims in the other direction. 

A 2023 paper in Cancers reported that high-dose NMN suppressed lung adenocarcinoma growth by promoting ferroptosis. [8] The dose was 10 to 100 mM injected directly into the tumor, a local tissue concentration delivered by needle rather than an oral capsule.

The same paper reports that at lower doses, NMN increased cancer cell proliferation.

A 2023 study in Nutrients found that nicotinamide riboside suppressed liver cancer progression in mice, at 400 mg per kg per day by oral gavage. [9] That is roughly five to nine times a typical human supplement dose.

The effect on primary tumor growth appeared only in mice with intact immune systems, which suggests the benefit was immune-mediated rather than direct.

A widely shared claim about triple-negative breast cancer comes from a 2023 Oncogene paper that tested NAD+ supplementation, not NMN, and measured metastasis rather than tumor growth. [10]

A peer-reviewed 2025 study of oral NMN in a UV-induced skin cancer model found no effect on tumor development in either direction. [11]

The pattern across both sides is the same. The doses are far from what people take, the routes are often not oral, and the models rarely resemble a healthy human body.

Who has a real reason to be careful with NAD+ precursors?

Anyone with an active cancer diagnosis, particularly during chemotherapy. This is the part of the viral claim that is significant, even though it is aimed at the wrong audience.

Several chemotherapy drugs work by inflicting oxidative stress and DNA damage on rapidly dividing cells. NAD+ is a cofactor for the repair machinery that undoes exactly that kind of damage.

Raising NAD+ availability during treatment should be first discussed with your doctor, and the Cancer Letters authors framed their own findings that way. [1]

The interest in NAD+ metabolism as a cancer target is not new or fringe. NAMPT, the rate-limiting enzyme in the NAD+ salvage pathway, is overexpressed across many malignancies and has been pursued as a drug target for years. [12]

No NAMPT inhibitor has been approved, and several trials were halted for toxicity or lack of efficacy. The dependence is well established, and drugging it has proven hard.

If you are in active treatment or under an oncology workup, this belongs in a conversation with your oncologist, who knows your cancer type, your protocol, and your labs. That is a different conversation from the one a healthy 55-year-old is having about her supplement stack.

Do the amino acids in Perfect Amino cause cancer?

No study links this product, or essential amino acid powders generally, to cancer. Searching the published literature turns up nothing connecting the two.

The dose comparison closes the question. One serving contains 5 grams of a proprietary blend, and that 5 grams includes non-amino-acid components, so the essential amino acid load is under 5 grams.

A serving delivers less than half the essential amino acids in a chicken breast, and less than three eggs. If the amino acids in the powder caused cancer, dinner would be the larger exposure.

One note: The proprietary blend means individual amino acid amounts are undisclosed, so no one outside the company can say how much leucine or methionine a serving contains. That is a transparency problem, and it is a separate issue from the cancer question.

Why do people think amino acids feed cancer?

Because there is real literature on amino acid metabolism in tumors, and it is routinely read backwards.

The most-cited example is a 2014 Nature Medicine study finding that elevated blood levels of branched-chain amino acids preceded a pancreatic cancer diagnosis by two to five years, with more than double the risk. [13]

The authors' own conclusion was that the tumor causes the elevation. Early pancreatic cancer drives breakdown of the patient's own muscle tissue, releasing those amino acids into the blood.

Dietary intake was never studied. The finding points toward an early-detection biomarker, not a risk factor.

A 2019 Nature study found that restricting dietary methionine slowed tumors and improved response to chemotherapy and radiation in mouse models. [16] Showing that removing an amino acid slows an existing tumor is not evidence that normal intake starts one.

The human portion of that study was six healthy volunteers measured for blood metabolites, with no cancer outcomes at all.

Leucine's preclinical record runs in both directions. A 2024 review documents pro-tumor effects through mTORC1 activation in some models and anti-tumor effects in others, including enhanced antitumor immunity, and concludes that clearer conclusions require more work. [15]

The largest relevant human dataset points the other way. A 2020 BMJ meta-analysis of 32 prospective cohorts covering 715,128 participants found no association between protein intake and cancer mortality, with a pooled effect size of 0.98. [14]

Higher total protein was associated with lower all-cause mortality.

How can you tell a real supplement warning from a viral one?

Four questions separate a finding worth acting on from a headline. They work on any scare, not just this one.

1. Humans or animals? A result in mice is a hypothesis about humans, not a finding in them.
2. Was the cancer already there? If tumor cells were injected or already established, the study cannot speak to whether anything caused the cancer.
3. What dose, by what route? Molar concentrations injected into a tumor, or ten times a human dose by gavage, do not translate to a capsule.
4. Does the study's own author agree with the headline? Researchers frequently say publicly that coverage went beyond their findings. That statement is usually one search away.

Frequently Asked Questions

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. Consult a healthcare professional before beginning any new supplement, especially if pregnant, nursing, on medication, or managing a condition. Individual results may vary.