The direct answer

Current randomized trials do not show that NMN causes more overall adverse events than placebo. A 2026 meta-analysis pooled safety data from 10 trials within a 15-trial review and found no clear increase in overall events, serious events, withdrawals caused by events, system-specific events, or the liver enzymes ALT and AST.

The studied window was limited: 250 to 2,000 mg/day for 14 days to 24 weeks. The correct conclusion is “no group-level safety signal detected over the studied periods,” not “NMN has no side effects.” Trials are too short and too small to rule out uncommon reactions, long-term effects, or interactions in people unlike the participants.

The 2026 pooled safety result

This is the widest current view of randomized human NMN trials, with limits set by the underlying studies.

Safety fields in Yang et al. 2026
Field	Evidence set	Pooled finding	What remains unknown
Overall adverse events	10 trials contributed safety analyses.	No significant increase versus control.	Rare events and risks beyond 24 weeks.
Serious adverse events	Randomized trial reports.	No significant increase detected.	The dataset is not large enough to exclude uncommon harm.
Withdrawals due to events	Reported treatment discontinuations.	No significant increase detected.	Tolerance in broader clinical populations.
ALT and ASTLiver enzymes	Available laboratory outcomes.	No significant pooled increase.	Long-term liver safety and people with liver disease.
Dose and duration	250–2,000 mg/day; 14 days–24 weeks.	Defines the studied range.	Use outside that range or duration.

Absence of a detected difference is not absence of risk

A trial compares event rates between groups. If it finds no significant difference, that may mean the treatment truly adds little risk over the study period, or that the study was too small or too short to detect an uncommon difference. This matters when many NMN trials enroll dozens rather than thousands of participants.

Published adverse-event tables also depend on how events were collected and classified. A participant can report a symptom during a trial without the study proving NMN caused it. Conversely, a rare reaction may never appear in a small sample. We therefore report group comparisons rather than turning isolated symptom lists into a claimed NMN side-effect profile.

Three useful trial snapshots

Selected human NMN studies and their safety boundaries
Study	Population	Dose and duration	Safety reading
Okabe et al., 2022	30 healthy participants.	250 mg/day for 12 weeks.	The authors reported no obvious adverse effects. The sample cannot establish uncommon or long-term risk.
Yi et al., 2023	80 healthy middle-aged adults assigned to placebo or three NMN doses.	300, 600, or 900 mg/day for 60 days.	No treatment-related safety concern was identified in the report; company-employed authors were involved.
Christen et al., 2026	65 participants in the modified intention-to-treat analysis across NMN, NR, nicotinamide, and placebo arms.	NMN 1,000 mg/day for 14 days.	Short comparative exposure only; not a long-term safety study. Authors included Nestlé affiliations.

Five gaps that matter more than a generic symptom list

Long-term use: randomized evidence extends to 24 weeks, not years.
Uncommon events: pooled samples remain much smaller than the datasets used to identify rare drug risks.
Drug interactions: the trials do not provide a comprehensive interaction map for prescription medicines.
Understudied groups: trial results should not be generalized to populations excluded from the studies.
Combination formulas: a reaction to a finished supplement may involve another active ingredient, capsule material, excipient, or contaminant rather than NMN alone.

A COA can narrow the product-quality question by checking identity, potency, or listed contaminants for a tested lot. It cannot answer clinical interaction or long-term safety questions. See how to read an NMN certificate of analysis.

Before using an NMN supplement

Read the full formula: record every active ingredient and the total daily serving, not only the front-label NMN number.
Do not exceed the label to copy a study: a research dose belongs to its protocol, population, and monitoring conditions.
Check medications and conditions with a clinician: a lack of documented interaction is not proof that no interaction exists.
Stop using the product after a concerning reaction: seek medical care based on the severity of the reaction rather than waiting to identify the exact ingredient yourself.
Keep the bottle and lot number: those details connect a reaction or quality concern to the actual product.

If you are deciding among current products, the NMN comparison separates dose, finished-product testing, added actives, price, and conflicts. None of those fields substitutes for individualized medical advice.

NMN Side Effects: What Trials Report