NMN Safety and Side Effects: What Human Studies Tell Us

What this page claims — and what it doesn't

This page summarizes what published human trials report about NMN's short-term tolerability. It does not claim, and the evidence does not support the claim, that "NMN is proven safe." Proving something safe in the fullest sense requires long-term data across diverse populations at realistic doses — data that does not yet exist for NMN. What exists is a body of short-to-medium-duration randomized trials that have not detected a serious safety signal within their own limited scope. Those are different statements, and this page keeps them separate throughout.

What "no serious adverse events" does and doesn't mean

Trial reports use adverse-event language in ways that are easy to conflate. This page distinguishes four things that are not the same: (A) no adverse events of any kind occurred; (B) adverse events occurred but none were judged serious; (C) adverse events occurred at a similar rate to placebo, whether or not any were serious; and (D) a trial's published report simply doesn't detail adverse events at all. Most NMN trials fall into category (B) or (C), not (A) — non-serious adverse events (mild gastrointestinal symptoms, for example) do occur in these trials, in both the NMN and placebo groups, and are not hidden by the studies that report them. As one detailed example: Pencina et al.'s 2023 JCEM trial of 2,000 mg/day MIB-626 recorded 24 on-treatment adverse events across 15 of its 30 participants — 2 classified as moderate, the rest mild — with "no serious adverse events" and adverse-event frequency "similar across groups." [3] That is a materially more precise (and more informative) picture than a bare "no adverse events," and it is representative of what "no serious adverse events" actually means across the trials on this hub: not that nothing happened, but that nothing rose to the level of serious, at the doses and durations studied.

What human trials report on adverse events

Across the randomized, placebo-controlled human NMN trials catalogued on NMN Dosage: What Human Studies Have Actually Used, the consistent pattern is an absence of serious treatment-related adverse events at doses from 250 mg/day up to 2,000 mg/day. [1][2][3] Where minor symptoms were reported — such as mild gastrointestinal complaints in one 12-week trial — they occurred in both the NMN and placebo groups, making it unclear whether NMN itself was the cause. [1] No trial identified on this hub reports a serious adverse event attributed to NMN.

The current systematic-review picture

The Yang et al. 2026 meta-analysis pooled 15 trials from the published NMN literature — a subset of, not identical to, this hub's own 16-trial inventory described on NMN Human Clinical Trials — and found that NMN supplementation did not increase overall, serious, withdrawal-related, or system-specific adverse events compared with placebo, and did not significantly elevate liver enzymes (ALT/AST), across studies using doses from 250–2,000 mg/day and durations from 14 days to 24 weeks. [4] This is currently the most comprehensive available synthesis of NMN's short-term human safety data. It is a pooled analysis of other trials, not a new primary trial of its own, and its own authors frame the finding as short-term tolerability, not long-term safety — larger and longer trials were explicitly identified as needed before firmer conclusions can be drawn. [4]

A note on funding and conflicts of interest

Several of the trials this page and the Dosage page rely on were funded by, or include authors employed by, companies that manufacture or sell NMN or NMN-related ingredients. Most directly: the Pencina et al. 2023 JCEM trial [3] was funded by Metro International Biotech, maker of the MIB-626 formulation tested, and a senior co-author is disclosed as a consultant and equity owner in that company; the same sponsor funded the COVID-19/AKI trial referenced below. [6] The Okabe et al. 2022 trial [1] includes co-authors employed by Mitsubishi Corporation Life Sciences Limited, and the Yi et al. 2023 trial [5] includes co-authors employed by NMN ingredient suppliers Abinopharm, Inc. and Aba Chemicals, Co. Funding and conflicts of interest are reported here because they are relevant context when evaluating a body of evidence — industry funding does not by itself invalidate a study, and none of the disclosures found were concealed; each was published by the study's own authors. Full per-study detail is on the Dosage page.

Duration and sample-size limits

These reassuring short-term findings come with real limits. The longest individual NMN trial identified on this hub ran 24 weeks, in a sample of 14 people; [2] the largest individual trial enrolled 80 people, over 60 days. [5] No published human NMN trial identified for this hub has run for a year or longer. That means there is, at present, no direct human evidence — good or bad — about what happens with continuous NMN use over multiple years, which is how many consumers who buy long-term supplements actually intend to use them.

Who these trials studied — and who they didn't

Published NMN trials have enrolled generally healthy adults, along with some trials in older adults with diabetes or reduced physical performance. [2] They have excluded pregnant and breastfeeding people, and have not enrolled children or adolescents. Trials have not been designed to test NMN in people actively undergoing cancer treatment, and most exclude participants with significant uncontrolled medical conditions. One trial administered a high dose of an NMN formulation to hospitalized adults with COVID-19 and acute kidney injury under close medical supervision — a fundamentally different context from routine consumer use, and not evidence that NMN is appropriate for acutely ill people generally. [6] See Who Should Talk to a Doctor Before Taking NMN? for how this shapes who should be cautious.

Absence of evidence is not evidence of absence

No trial has reported a clear signal of serious harm from NMN at studied doses and durations. That is a genuinely reassuring, real finding — and it is not the same thing as evidence that no risk exists. Because trials have been short, moderate in size, and conducted in relatively healthy populations, rare adverse events, effects that take years to emerge, and effects in excluded populations (including drug interactions) could exist without yet being detected. This distinction — between "not observed so far, in the populations and timeframes studied" and "shown not to occur" — is the central limitation of the current evidence base, and it applies to NMN's safety profile specifically, not as a generic disclaimer.

What remains genuinely unestablished

Long-term (multi-year) human safety, safety during pregnancy and breastfeeding, safety in children, safety in people with active cancer or significant chronic illness, and interactions with prescription medications have not been directly studied in humans and remain unestablished. NMN supplements are not medications, are not FDA-approved to diagnose, treat, cure, or prevent any disease, and nothing on this page should be read as medical advice.