The short answer: no side effects attributed to NMN, adverse events similar to placebo, and nothing tested past 24 weeks
The honest, unglamorous answer on NMN side effects is that the trials run so far have not found any to attribute to NMN.
A 2023 review in Advances in Nutrition, indexed as PMC10721522, identified 10 published human clinical trials of NMN, and it is the evidence base this page starts from. Its trials ran oral doses of 250 mg once daily for 6 or 12 weeks, 300, 600 and 1,200 mg once daily for 6 weeks, 300 mg once daily for 60 days, 300 mg daily for 30 to 90 days, and, at the top, 1,000 mg twice daily for 14 days. The review's own words on safety: "existing human clinical trials have not shown any side effects of NR and NMN," and "the administration of NMN orally is safe and has good tolerance."
The individual trials say the same thing in more detail. A 60-day trial of 80 healthy adults at 300, 600 or 900 mg a day (Yi et al. 2023, epub Dec 2022, PMID 36482258) logged nine adverse events in total: six in the placebo group, three in the 300 mg group, none at 600 or 900 mg, all mild or moderate, and none judged related to NMN. A 4-week safety trial at 1,250 mg once daily in 31 healthy adults aged 20 to 65 (Fukamizu et al. 2022, PMID 36002548) recorded five adverse events, each judged unrelated to the test food, and no severe adverse events. Two trials at 1,000 mg twice daily (Pencina et al., PMID 35182418 and PMID 36740954) reported adverse events similar across groups. The breakdown below keeps dose and duration side by side, since one blended "NMN is safe" headline hides that not every trial tested the same amount for the same length of time.
| Trial | Dose per day | Duration | Adverse events, as reported |
|---|---|---|---|
| The 10 trials counted by the 2023 review (PMC10721522) | 100 mg single dose to 1,000 mg twice daily | 5 hours to 90 days | "Existing human clinical trials have not shown any side effects of NR and NMN" |
| Three-arm dose trial, 80 healthy adults aged 40 to 65 (Yi et al. 2023, epub Dec 2022, PMID 36482258) | 300, 600 or 900 mg | 60 days | Nine adverse events: six on placebo, three at 300 mg, none at 600 or 900 mg; all mild or moderate, none related to NMN |
| High-dose safety trial, 31 healthy adults aged 20 to 65 (Fukamizu et al. 2022, PMID 36002548) | 1,250 mg once daily | Up to 4 weeks | Five adverse events (loose stool twice, one of them on placebo; a common cold; high blood pressure; acne), all judged unrelated; no severe adverse events |
| Highest-dose trials, overweight or obese adults (Pencina et al., PMID 35182418 and PMID 36740954) | 1,000 mg once or twice daily | 14 days and 28 days | Reported as well tolerated (14-day trial) and safe (28-day trial); adverse events similar across groups in both |
| Longest placebo-controlled trial we found, 14 men with diabetes aged 65 and over (Akasaka et al. 2023, PMID 36443648) | 250 mg | 24 weeks | Tolerable without any severe adverse events; no difference from placebo in grip strength or walking speed |
Dose list and safety wording per the human-trial review indexed as PMC10721522 (Song et al. 2023, Adv Nutr), read September 7, 2026, which names 1,000 mg twice daily for 14 days as the highest oral dose in the trials it reviewed. The Yi, Fukamizu, Pencina and Akasaka rows are cited to PubMed directly. This table covers NMN as an ingredient across trials, not any single branded formula.
Ingredient trials, not a finished-product test, and not proof of benefit
The review pools trials that dosed NMN itself, not any one capsule sold at retail, and a clean safety record is not the same as an efficacy record.
- A meta-analysis (Zhang et al. 2025, epub Aug 2024, Crit Rev Food Sci Nutr, PMID 39116016) pooled 12 studies and 513 participants to assess fasting glucose, triglycerides, total cholesterol, LDL-C and HDL-C. It found "an overall significant effect of NMN supplementation in elevating blood NAD levels," but "most of the clinically relevant outcomes were not significantly different between NMN supplementation and control group," and it rated five of the 12 studies at high risk of bias.
- The review itself says the trials were small: eight of them enrolled 8 to 66 participants, "primarily at the phase I clinical trial level," and most participants were aged 55 or older.
- "Well tolerated" describes trials lasting from a single day to 24 weeks at most, not a year of daily use, let alone five.
What this means for you: the safety picture so far is reassuring for a short trial, not a lifetime commitment. Both are true at once.
The dose ceiling that was actually tested, and where a standard capsule sits inside it
Two kinds of ceiling exist in the trials. The highest single daily serving tested for safety is 1,250 mg once a day, for up to 4 weeks, in 31 healthy adults (Fukamizu et al. 2022). The highest total daily amount is 1,000 mg twice a day, tested for 14 days in 32 overweight or obese adults aged 55 to 80 (Pencina et al., PMID 35182418) and for 28 days in 30 overweight or obese adults aged 45 and over (Pencina et al., PMID 36740954). All three trials reported the dose as well tolerated, with adverse events similar across groups or judged unrelated.
That is the upper edge of what has actually been tested in a human trial, as opposed to what a label happens to sell. A common single-capsule NMN product prints 500 mg on its Supplement Facts panel: 40% of the 1,250 mg once-daily ceiling, and inside the 300 to 900 mg range of the 60-day three-arm trial, where the 600 and 900 mg groups logged no adverse events at all.
The criterion worth applying to any label: milligrams per day tested versus milligrams per capsule on the bottle. One 500 mg capsule daily lands inside the studied range; two, at 1,000 mg, equals the once-daily arm of the 14-day Pencina trial, though no trial ran exactly two 500 mg capsules of a retail product.
Dose ceilings per Fukamizu et al. 2022 (PMID 36002548) and Pencina et al. (PMID 35182418, PMID 36740954); dose arms per Yi et al. 2023, epub Dec 2022 (PMID 36482258). Capsule math is arithmetic against a common 500 mg single-ingredient NMN label, not a claim about any specific trial arm.
What the adverse-event tables actually list, and the flushing confusion
When the trials itemize adverse events, the list is short and scattered. In the 1,250 mg safety trial: loose stool (twice, once on placebo), a common cold, one case of high blood pressure and one of acne, every one judged unrelated to the test food. In the 300 to 900 mg trial: hyperacidity, a skin problem and a mouth ulcer, all in the 300 mg group and none related to NMN, against six events on placebo. Every event in both tables was judged unrelated to NMN, and the 60-day trial logged more events on placebo (six) than on any NMN arm (three at 300 mg, none at 600 or 900 mg).
Worth clearing up separately is flushing, the warm, reddening skin reaction some people associate with a B-vitamin supplement. That reaction belongs to niacin (nicotinic acid), not NMN. The review lists flushing first among niacin's most common adverse effects, alongside rash, hyperglycemia, hyperuricemia and gastrointestinal disorders. The 1,250 mg trial states that NMN at that dose for up to 4 weeks was given "without causing hepatotoxicity and vasodilative flushing," and a 12-week trial of 250 mg in 108 older adults (Kim et al. 2022, PMID 35215405) reported no side effects from NMN ingestion.
The confusion is understandable: NMN and niacin both sit in the vitamin B3 family. Sharing a family name is not the same as sharing a side-effect profile. If flushing does appear after taking any supplement, check every ingredient on the label rather than assume NMN is the cause.
Adverse-event tables per Fukamizu et al. 2022 (Table 6, PMC9400576) and Yi et al. 2023, epub Dec 2022 (Table 4, PMC9735188). Niacin adverse effects and the NMN wording per PMC10721522; the flushing sentence per Fukamizu et al.; the 250 mg trial per Kim et al. 2022 (PMID 35215405).
What no study has measured yet
The trials the review counted ran from a single 5-hour session to 90 days (one trial ran 30 to 90 days). The longest placebo-controlled NMN trial we found, Akasaka et al. 2023 (PMID 36443648), ran 24 weeks, about six months, in 14 men with diabetes aged 65 and over, and that is still the outer edge of what has actually been tested.
The longest placebo-controlled NMN trial duration in the sources we read (PMID 36443648). Everything described as "well tolerated" here was tolerated for, at most, about six months.
The review's own words: "longer-term studies with large populations have not been performed," and "the long-term safety issue should be addressed about NMN supplementation." That is a gap in the evidence, not a finding of harm.
The review notes a mouse study in which NMN supplementation could "promote pancreatic ductal adenocarcinoma progression in a mouse model driven by oncogenic Kras," and adds that NMN's toxicological data are limited. No human trial has tested this; it is exactly why active cancer treatment is on the ask-a-doctor list below.
None of this is evidence that long-term use is unsafe. It is evidence the question has not been asked yet, and "no long-term data" is not the same claim as "proven long-term safety."
Trial duration ceiling per Akasaka et al. 2023, PMID 36443648. Long-term gap and the mouse cancer-model note per PMC10721522. The cell-growth question is an animal finding and an open caution, not a demonstrated human risk.
Who should not take NMN without asking a doctor first
Nothing here is medical advice. Each of the six situations below is a reason to talk to a doctor first, not a reason NMN is automatically off-limits.
Six situations the evidence base does not cover
- Pregnant or trying to conceive: the 4-week safety trial (PMID 36002548) excluded anyone pregnant, breastfeeding or planning to become pregnant, and the 60-day trial (PMID 36482258) listed confirmed pregnancy as a reason to withdraw a participant.
- Nursing or breastfeeding: the same safety trial excluded breastfeeding participants; effects on a nursing infant have not been studied in any trial we read.
- Active cancer treatment, or a history of hormone-sensitive cancer: the mouse-model finding above is why this needs an oncologist's input first.
- Taking blood thinners or any other prescription medication: dedicated interaction trials do not exist in the sources we read, so a doctor is the only reliable check.
- Managing diabetes or blood pressure with medication: in one 28-day trial, 1,000 mg twice daily lowered diastolic blood pressure, total cholesterol and LDL cholesterol significantly more than placebo (PMID 36740954), while a 24-week trial of 250 mg in men with diabetes reported no severe adverse events (PMID 36443648): active enough to flag alongside an existing medication, whichever direction it moves.
- Under 18: the trials we read enrolled adults aged 20 and over; none enrolled minors.
Checklist built from the enrolment criteria and gaps in the sources listed on this page. It is a list of open questions, not a diagnosis.
Purity as a safety variable: what a certificate of analysis should show
A dose that matches the label is only half of NMN safety. The other half is what else is in the capsule: heavy metals and microbial contamination are safety variables a milligram figure says nothing about.
A proper certificate of analysis should report three things: NMN content measured against the label claim, heavy-metal levels against a stated maximum in parts per million, and microbiology results such as salmonella and E. coli reported as absent.
| Heavy metal | Stated maximum | Measured, one supplier CoA | Reading |
|---|---|---|---|
| Lead | max 1.0 ppm | 0.39 ppm | Well within limit |
| Mercury | max 4.0 ppm | 0.02 ppm | Well within limit |
| Arsenic | max 4.0 ppm | 0.03 ppm | Well within limit |
| Cadmium | max 1.0 ppm | 0.002 ppm | Well within limit |
The worked example above comes from one NMN supplier's own certificate of analysis for a 500 mg capsule (measured content 506 mg per capsule against a "not less than 500" specification), dated October 24, 2025, read directly rather than taken on faith. Reading any product's paperwork the same way means matching four things: a stated maximum per metal, a measured value under it, a batch number matching the bottle, and a recent date. A label that says "third-party tested" with no certificate to open has told you nothing yet.
Heavy-metal limits and measured values from one NMN supplier's certificate of analysis, dated October 24, 2025, read directly. A supplier certificate is quality paperwork on its own batch, not the same as independent third-party verification, and it does not cover every bottle ever produced.
How to run a two-week safety self-check before committing to a full bottle
Trial data describes groups. It cannot tell you how one specific person will respond, which is what a short, deliberate self-check is for.
Day 0: set a baseline
Before the first dose, write down how you feel: digestion, sleep, energy. Start at the low end of a label's suggested range, one capsule, at the same time each day.
Days 1 to 14: log it daily
One line a day: GI comfort, sleep, energy, mood, anything new. Fourteen days is the length of the shortest placebo-controlled trial in our sources, and it was long enough for that trial to measure a rise in blood NMN and NAD (PMID 35182418).
When to stop immediately
A rash, chest pain, or a severe or worsening GI symptom means stop and call a doctor, not wait it out. A mild symptom unresolved after 48 hours means pause and reassess.
If two weeks pass with nothing on that log worth flagging, that is the point where a full bottle, rather than a trial supply, actually has something behind it.
Self-check structure derived from the trial durations and adverse-event tables cited above, not from any single product's instructions. The stop rules are general first-aid sense, not a trial finding.
Questions people also ask
Does NMN cause flushing like niacin?
No signal of that in the trials we reviewed. Flushing is the first adverse effect the 2023 review (PMC10721522) lists for niacin, alongside rash, hyperglycemia, hyperuricemia and gastrointestinal disorders. NMN is a different molecule: the 4-week trial at 1,250 mg once daily (Fukamizu et al. 2022, PMID 36002548) states it was given without causing vasodilative flushing, and the review says existing human clinical trials have not shown any side effects of NMN. Sharing the vitamin B3 family name is not the same as sharing a side-effect profile.
Is NMN safe to take long term?
Not yet answered in either direction. The trials counted by the 2023 review ran from a single 5-hour session to 90 days; the longest placebo-controlled NMN trial we found ran 24 weeks (Akasaka et al. 2023, PMID 36443648). The review itself says longer-term studies with large populations have not been performed and that the long-term safety issue should be addressed. Good short-term tolerability is not the same claim as proven long-term safety, and right now only the first one has trial data behind it.
Can NMN cause insomnia if taken at night?
No trial we read reports insomnia as an adverse event, and the one trial that tested clock time did not find evening dosing hurting sleep. Kim et al. 2022 (PMID 35215405) gave 108 older adults 250 mg of NMN or placebo once a day for 12 weeks, either between waking and noon or between 6 PM and bedtime, and reported no side effects from NMN ingestion. On the Pittsburgh Sleep Quality Index there was no statistically significant group-by-time interaction, but the evening NMN group showed the largest effect sizes of the four groups on the sleep quality score and the total PSQI score. Most other trials do not specify a time of day, so "take it in the morning" is a label default, not a trial finding. See our full breakdown in best time to take NMN.
Does NMN interact with blood pressure or diabetes medication?
No dedicated drug-interaction trial exists in the sources we read, which is itself the caution. NMN is metabolically active enough to matter: in a 28-day trial, 1,000 mg twice daily lowered diastolic blood pressure, total cholesterol and LDL cholesterol significantly more than placebo (Pencina et al. 2023, PMID 36740954), while a meta-analysis of 12 studies (Zhang et al. 2025, epub Aug 2024, PMID 39116016) found most clinically relevant outcomes not significantly different from control. A 24-week trial of 250 mg in men with diabetes reported no severe adverse events (PMID 36443648), but it enrolled 14 people. Anyone on blood pressure or diabetes medication should talk to a doctor before adding NMN, precisely because that combination has not been studied on its own.
How we built this page
This is a literature-and-label page, not a lab test, and not a review of any specific brand. Every number on it comes from a document we read, on a date we print.
- Trial count, dose list and safety wording come from the human-trial review indexed as PMC10721522 (Song et al. 2023, Advances in Nutrition), read September 7, 2026, which identified 10 published human clinical trials of NMN.
- Individual trial details come from the PubMed abstracts, and where we quote adverse-event tables, the PMC full text, of Fukamizu et al. 2022 (PMID 36002548), Pencina et al. (PMID 35182418 and 36740954), Akasaka et al. 2023 (PMID 36443648), Yi et al. 2023, epub Dec 2022 (PMID 36482258) and Kim et al. 2022 (PMID 35215405).
- The metabolic-outcomes finding comes from a meta-analysis (Zhang et al. 2025, epub Aug 2024, Crit Rev Food Sci Nutr, PMID 39116016), cited only for what its abstract states.
- The heavy-metal worked example comes from one NMN supplier's own certificate of analysis for a 500 mg capsule, dated October 24, 2025, read directly.
- We did not run a trial or a lab test of our own. Every number on this page is one we read in a source; where a source did not print a number, neither do we.
- Nothing here is medical advice. A new supplement while pregnant, nursing, on medication, or during cancer treatment is a conversation with your doctor first.
The capsule where we checked the label against the paperwork
If the safety picture makes sense to you, the next question is whether the capsule you buy actually contains what the label says, which is exactly what we checked line by line in our PartiQlar Pure NMN review.
The certificate-of-analysis example above comes from that same product: label claim against an independent lab's measured result, and the supplier's own purity paperwork read in full.
Before trusting any "third-party tested" claim, ask to see the actual certificate. A batch number and a recent date on the document are what make it checkable against the bottle in your hand.
Start with a single bottle, not a bundle. It caps the cost of a failed two-week self-check, and it is the only sensible order size when a brand's guarantee, like the 60-day window on the capsule we review, only accepts bottles that are still unopened and sealed.
Affiliate disclosure: the review linked below contains affiliate links, and a purchase made through them may earn us a commission at no extra cost to you. This page carries no affiliate link of its own.
Read our PartiQlar Pure NMN reviewBoth certificates of analysis, read line by line, plus the checkout totals we read, each with the date we read it.