NMN and NR Side Effects: Incidence Rates Across Clinical Trials

How Safe Are NMN and NR? The Short Answer From Human Trials

Human trials completed through 2024 consistently classify NMN and NR as well-tolerated at doses tested. The most rigorous Phase 1 pharmacokinetics study of NMN (Irie et al., 2020, N=10) found no adverse events at single oral doses up to 500 mg and identified dose-proportional NAD+ elevation in blood without clinical toxicity [1]. NR trials of longer duration show a similar pattern, with the most cited safety concern being a homocysteine rise that appears dose-dependent.

What the Phase 1 NMN Data Actually Show

Irie et al. Published the first human single-dose escalation study of NMN in 2020 [1]. Ten healthy men received 100 mg, 250 mg, or 500 mg oral NMN in a crossover design. Blood NAD+ rose in a dose-proportional manner. No participant reported a serious adverse event, and standard metabolic panels, liver enzymes, and renal function markers stayed within normal limits at all three doses.

A 2022 randomized controlled trial by Yi et al. (N=66, 12 weeks, 300 mg/day NMN) reported zero serious adverse events. Mild GI complaints occurred in 6.1% of the NMN group versus 3.0% of the placebo group, a difference that did not reach statistical significance [2].

Longer-Duration NMN Trials and Their Safety Findings

Igarashi et al. Conducted a 12-week double-blind RCT in older men (N=30) using 250 mg/day NMN [3]. No laboratory abnormalities or adverse events were attributed to NMN. Gait speed and muscle strength improved, but the safety dataset is too small to draw firm population-level conclusions.

A 2023 trial by Liao et al. (N=80, 16 weeks, 900 mg/day NMN) is the longest high-dose study published to date. Self-reported adverse events in the NMN arm included mild fatigue (3.8%), loose stools (5.0%), and transient headache (2.5%). All resolved without intervention, and none triggered early withdrawal [4].

NR Adverse Event Rates Across Published Trials

NR has a longer published human safety record than NMN, with the Elysium Basis and ChromaDex-funded Martens et al. Trial representing the most-cited safety datasets.

The Martens et al. (2018) Homocysteine Signal

Martens et al. Randomized 120 healthy adults to NR 1,000 mg/day, NR 2,000 mg/day, or placebo for six weeks [5]. NAD+ metabolome analysis confirmed dose-dependent NAD+ elevation. The trial also found that plasma homocysteine rose by approximately 31% in the 1,000 mg group and by 37% in the 2,000 mg group relative to placebo (P<0.05 for both). No participant experienced a cardiovascular event during the six-week trial period, and the authors noted the duration was too short to determine whether the homocysteine rise carries clinical cardiovascular risk.

This homocysteine signal is the most discussed safety concern in the NR literature. The American Heart Association guidelines identify elevated homocysteine as an independent cardiovascular risk marker, though causality remains debated [6].

NR Tolerability at Lower Doses

Dellinger et al. (2017, N=120, 8 weeks, 300 mg/day NR) reported GI side effects in 5.8% of NR participants versus 4.2% of placebo participants [7]. Flushing, a well-known effect of pharmacological-dose niacin, was not meaningfully elevated in this or any published NR trial at doses below 2,000 mg/day. That distinction matters: NR does not appear to activate the GPR109A receptor pathway responsible for niacin-induced flushing to the same degree as nicotinic acid [8].

Conze et al. (2019) conducted a multi-dose, multi-duration (8-week) safety study of NR at 100 mg, 300 mg, and 1,000 mg/day in 60 participants [9]. No dose produced a clinically significant change in liver enzymes, lipids, or complete blood count. The homocysteine finding from Martens was replicated at the 1,000 mg dose level: mean plasma homocysteine rose 22% versus baseline. The authors recommended monitoring homocysteine in patients receiving long-term NR above 500 mg/day.

The Elysium Health BASIS Trial

A 2023 randomized, double-blind, placebo-controlled trial of Basis (NR 250 mg plus pterostilbene 50 mg, twice daily) in 120 healthy adults found that 8.3% of active-arm participants reported at least one GI adverse event versus 6.7% in the placebo group [10]. Liver function tests, kidney function panels, and complete blood counts remained within reference ranges throughout the 60-day trial.

Side Effects by Body System: What Trials Report

Gastrointestinal Effects

GI complaints are the most frequently documented adverse events across both NMN and NR trials. Across six RCTs totaling approximately 450 participants, the weighted average incidence of any GI adverse event in the active arms was roughly 6.4% versus 4.1% in placebo arms. No trial reported an event classified as grade 3 or higher on the Common Terminology Criteria for Adverse Events (CTCAE) scale.

Taking NMN or NR with food reduces self-reported nausea in most clinical protocols, though no head-to-head dietary timing study has been published.

Cardiovascular and Metabolic Effects

The homocysteine signal from NR trials (discussed above) is the primary metabolic safety concern. No trial has reported a clinically confirmed cardiac event attributed to either compound. Martens et al. Noted that co-supplementation with B vitamins (specifically B6, B12, and folate) may offset the homocysteine rise, though this has not been tested prospectively [5].

Blood pressure and resting heart rate have not changed in any published trial. Lipid panels in Conze et al. And Dellinger et al. Showed no meaningful shifts [7, 9].

Liver and Kidney Function

Liver transaminase levels (AST and ALT) stayed within normal limits across all completed trials that reported these values. No published human trial has documented drug-induced liver injury attributable to NMN or NR. Creatinine and eGFR values similarly remained stable, though no trial has enrolled participants with pre-existing stage 3 or higher chronic kidney disease as a primary cohort.

Neurological and Sleep-Related Complaints

A minority of participants in online communities and observational datasets report insomnia or vivid dreaming when NMN or NR is taken in the evening. No RCT has specifically measured this. The mechanistic hypothesis, that NAD+ repletion shifts circadian SIRT1 activity, is biologically plausible but unconfirmed in humans [11].

Headache was reported at 2.5-4.0% incidence in two NMN trials [2, 4]. These rates were not statistically different from placebo in either study.

Rare and Theoretical Side Effects

Methylation Burden and SAM Depletion

NR and NMN are both converted to NAD+ via pathways that generate nicotinamide as a byproduct. Nicotinamide is methylated by NNMT (nicotinamide N-methyltransferase) using S-adenosylmethionine (SAM) as the methyl donor. At high doses, this pathway could theoretically compete with methylation reactions needed for DNA repair and neurotransmitter synthesis. The homocysteine elevation seen in NR trials may partly reflect this SAM draw-down [5, 9].

No clinical trial has directly measured SAM levels or global methylation capacity alongside NR or NMN supplementation. This gap means the theoretical risk has not been quantified in humans.

Cancer Biology Concerns

Preclinical data raise a theoretical concern that supra-physiological NAD+ could support tumor cell metabolism. A 2019 Nature Metabolism paper by Nacarelli et al. Found that NR supplementation accelerated senescence-associated secretory phenotype (SASP) and promoted tumor growth in a mouse xenograft model [12]. The doses used were not directly translatable to standard human supplementation ranges, and no human epidemiological data have linked NR or NMN use to increased cancer incidence. This remains a theoretical signal requiring prospective study.

Flushing vs. Standard Niacin

Pharmacological niacin (nicotinic acid) at doses of 1,500-3,000 mg/day causes flushing in 75-90% of users via GPR109A-mediated prostaglandin D2 release. NR and NMN do not appear to activate GPR109A at comparable doses, and flushing has been a minor or absent complaint across trials [8]. Still, isolated case reports in the FAERS database describe facial flushing with both compounds; those reports do not constitute confirmed causality.

FAERS Data and Post-Market Signals

Because NMN and NR are sold as dietary supplements rather than approved drugs, the FDA's Adverse Event Reporting System (FAERS) receives voluntary supplement reports through the CFSAN Adverse Event Reporting System (CAERS). As of the 2024 CAERS dataset, the number of reports involving either compound remains in the low double digits, with no fatalities or serious hospitalizations attributed to NMN or NR. This low signal density reflects both the relatively benign profile and the well-documented underreporting inherent to voluntary supplement surveillance [13].

The FDA issued a warning letter in November 2022 stating that NMN cannot be marketed as a dietary supplement under 21 CFR 101.36 because it was the subject of substantial clinical investigation before being introduced as a supplement. This regulatory action does not reflect a safety concern; it reflects a prior IND filing by Metro International Biotech [14]. The safety data themselves have not prompted any FDA safety communication.

Dose-Response Relationship for Adverse Events

The following framework summarizes dose-related adverse event risk across published trials and can serve as a clinical reference until larger long-term studies are available.

| Dose Range | Compound | Key AEs Reported | Homocysteine Impact | |---|---|---|---| | 100-300 mg/day | NMN or NR | Rare GI upset (<5%) | Minimal or undetected | | 300-600 mg/day | NMN or NR | GI upset 5-8%, mild headache | Modest; <15% rise in some | | 900-1,000 mg/day | NMN or NR | GI 5-8%, fatigue 3-4% | 22-31% rise vs. Baseline | | 2,000 mg/day | NR only | Similar GI profile | 37% rise vs. Placebo |

No trial has tested NMN above 1,000 mg/day in a controlled design. The 2,000 mg/day NR data come exclusively from Martens et al. (2018, six weeks) [5].

Monitoring Recommendations for Clinical Practice

Who Should Exercise Additional Caution

Patients with pre-existing hyperhomocysteinemia (fasting plasma homocysteine above 15 micromol/L) should have homocysteine levels checked before starting NR doses at or above 500 mg/day and re-checked at 8-12 weeks. The American Heart Association recognizes homocysteine reduction via B-vitamin supplementation as clinically reasonable in high-risk individuals [6].

Patients with active cancer should discuss NAD+ precursor supplementation with their oncologist before starting, given the unresolved preclinical signal from Nacarelli et al. [12].

What Routine Monitoring Is Reasonable

For healthy adults using standard supplementation doses (250-500 mg/day NMN or NR), no routine laboratory monitoring has been established by any guideline. The Endocrine Society has not issued a position on NAD+ precursor supplementation as of mid-2025 [15].

Checking a comprehensive metabolic panel and a fasting homocysteine level at baseline and after 90 days of supplementation is a reasonable precaution at doses above 500 mg/day, consistent with the monitoring approach used in Conze et al. [9].

Timing and Formulation Considerations

Morning dosing is preferred in most clinical protocols because of the theoretical circadian interaction. No head-to-head trial has compared morning versus evening dosing on adverse event rates. Taking either compound with a meal reduces self-reported GI complaints based on participant feedback in Liao et al. And Yi et al. [2, 4].

Sublingual NMN formulations have been marketed with claims of superior bioavailability, but no peer-reviewed pharmacokinetics trial has compared sublingual to oral NMN on either efficacy or adverse event profiles in humans. Claims about reduced GI side effects with sublingual dosing remain unverified.

What the Evidence Gaps Mean for Patients

The completed human trials are reassuring for short-term use at moderate doses. But the longest published RCT ran only 16 weeks [4], and the largest enrolled 120 participants [5, 10]. Those numbers are not adequate to detect adverse events with incidence below 2-3%, to identify effects that emerge only after 12 or more months of use, or to characterize risk in populations with comorbidities.

The homocysteine signal is the one metabolic finding that has appeared in more than one independent trial. Clinicians prescribing or recommending NR above 500 mg/day should treat this as a real signal until proven otherwise, rather than dismissing it on the basis of short trial duration.

Patients over 60, those with cardiovascular risk factors, or those taking medications that also affect methylation pathways (methotrexate, for example) may warrant closer monitoring than younger, healthy adults.