NMN Dosage: What Human Studies Have Actually Used

IntermediateStrong human evidence9 min read

Short answer Human NMN trials have used oral doses from 250 mg/day up to 2,000 mg/day, most commonly 250–900 mg/day, over periods of two weeks to six months. This page catalogs exactly what each of 13 identified trials studied — population, dose, duration, design, outcomes, safety observations, and funding/conflicts of interest — each independently verified against its primary source. It does not convert any of these figures into a recommended consumer dose; no trial was designed to establish one.

Two different questions: studied dose vs. recommended dose

This page answers one question only: what NMN doses have human clinical trials actually used? It does not answer, and is not designed to answer, "what dose should a person take?" Those are different questions with different evidence requirements. A dose that a research team selected for a short trial — often chosen for practical, funding, or pharmacokinetic-modeling reasons — is not the same thing as a dose demonstrated to be optimal, and no human trial to date has been designed to identify an optimal consumer dose. Nothing on this page or elsewhere on this hub should be read as a dosing recommendation.

A note on funding and conflicts of interest

Most of the trials in the table below were conducted or funded, at least in part, by companies that manufacture or sell NMN or NMN-related ingredients, or include authors employed by such companies. This is disclosed in the table's "Funding / COI" column and detailed in full underneath it. 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 trials below were found to have concealed or misrepresented their funding. But readers should be able to see this context alongside the findings, not have to look for it separately, particularly because the most heavily industry-connected trials also happen to include some of the higher doses in this table.

Human NMN trials by studied dose

The table below summarizes randomized, placebo-controlled human NMN trials identified through a review of the current published literature, current as of this page's last-reviewed date. It is not necessarily exhaustive of every NMN trial ever conducted, but reflects the trials with clear, verifiable dosing, population, and outcome data at time of writing, each independently checked against its primary source (PubMed and/or the publishing journal) rather than taken from secondary summaries.

Randomized, placebo-controlled human NMN trials, by studied dose (ascending)
Study	Population	n	Dose	Frequency	Duration	Design	Key reported outcomes	Safety observations	Funding / COI	PMID	DOI
Igarashi et al., 2022, npj Aging[1]	Healthy men, ≥65y	20	250 mg/day	Once daily	6 or 12 weeks	RCT, double-blind, placebo, parallel	Blood NAD+ ↑; no effect on insulin sensitivity, skeletal muscle mass, or visceral fat area; nominal (non-robust) gait speed/grip changes	Well tolerated; no significant deleterious effect reported	Industry funded / author employment	35927255	10.1038/s41514-022-00084-z
Okabe et al., 2022, Frontiers in Nutrition[2]	Healthy adults, 20–65y	30 (29 completed)	250 mg/day	125 mg twice daily	12 weeks + 4-wk follow-up	RCT, double-blind, placebo, parallel	Blood NAD+ nearly doubled	No serious adverse events; minor GI symptoms in both arms	Industry funded / author employment	35479740	10.3389/fnut.2022.868640
Katayoshi et al., 2023, Scientific Reports[3]	Healthy middle-aged adults	36	250 mg/day	125 mg twice daily	12 weeks	RCT, double-blind, placebo, parallel	NAD+ metabolites ↑; pulse wave velocity (arterial stiffness) trended lower, not statistically significant	Well tolerated; no adverse events attributed to treatment	No competing interests reported	36797393	10.1038/s41598-023-29787-3
Akasaka et al., 2023, Geriatrics & Gerontology International[4]	Men ≥65y with diabetes and reduced grip/gait	14	250 mg/day	Once daily	24 weeks	RCT, double-blind, placebo	No significant improvement in grip strength or walking speed vs. placebo (small sample)	Tolerable; no severe adverse events	Not verified	36443648	10.1111/ggi.14513
Morifuji et al., 2024, GeroScience[5]	Healthy older adults, 65–75y	60	250 mg/day	Once daily	12 weeks	RCT, double-blind, placebo, parallel	Prespecified primary outcome (stepping test) not significantly different from placebo at 4 or 12 weeks; secondary outcomes: NAD+ ↑, shorter 4-m walking time than placebo, improved sleep quality (PSQI)	No adverse effects attributed to treatment	Industry funded / author employment	38789831	10.1007/s11357-024-01204-1
Yoshino et al., 2021, Science[6]	Postmenopausal women, prediabetic, overweight/obese	25	250 mg/day	Once daily	10 weeks	RCT, double-blind, placebo, parallel	↑ muscle insulin sensitivity and insulin signaling; ↑ muscle-remodeling gene expression	No severe adverse events reported	Industry ties disclosed	33888596	10.1126/science.abe9985
Huang, 2022, Frontiers in Aging (Uthever)[13]	Healthy adults, 40–65y	66	300 mg/day	150 mg twice daily	60 days	RCT, multicenter, double-blind, placebo, parallel	NAD+/NADH ↑38% vs. 14.3% placebo; SF-36 quality-of-life score ↑6.5% vs. 3.4%; HOMA-IR stable vs. worsened in placebo	Not detailed in the abstract; not independently confirmed beyond the efficacy findings	Industry funded / author employment	35821806	Not independently confirmed
Liao et al., 2021, J Int Soc Sports Nutr[7]	Recreational runners	48	300 / 600 / 1,200 mg/day	Once daily, 3 dose arms + placebo	6 weeks, combined with training	RCT, double-blind, 4-arm, placebo-controlled	Certain submaximal / ventilatory-threshold-related oxygen-uptake measures improved more at medium and high doses alongside training; VO2max itself did not significantly differ between groups	Well tolerated; no serious adverse events reported	No competing interests reported	34238308	10.1186/s12970-021-00442-4
Yi et al., 2023, GeroScience[8]	Healthy adults, 40–65y	80	300 / 600 / 900 mg/day	Once daily before breakfast, 3 dose arms + placebo	60 days	RCT, multicenter, double-blind, placebo, dose-dependent, parallel	Dose-dependent NAD+ ↑; ↑ 6-minute walk distance; improved self-reported wellbeing	Safe and well tolerated up to 900 mg/day	Industry funded / author employment	36482258	10.1007/s11357-022-00705-1
Pencina et al., 2023 (pilot)[9]	Overweight/obese adults, 55–80y	32	1,000 or 2,000 mg/day	Once or twice daily, block-randomized	14 days	RCT, double-blind, placebo, block-randomized	Dose-dependent NAD+ ↑; established basis for larger follow-up trial	No severe adverse events reported	Not verified	35182418	10.1093/gerona/glac049
Fukamizu et al., 2022, Scientific Reports[10]	Healthy adults, 20–65y	31	1,250 mg/day	Once daily	Up to 4 weeks	RCT, double-blind, placebo, parallel (dose-escalation safety design)	No changes exceeding normal physiological variation across hematology, biochemistry, urine, body composition	No severe adverse events reported	Industry funded / author employment	36002548	10.1038/s41598-022-18272-y
Christen et al., 2026, Nature Metabolism[11]	Healthy adults, mean age ~35y	65	1,000 mg/day	Once daily, 1 of 4 study arms (NMN / NR / nicotinamide / placebo)	14 days	RCT, open-label, placebo-controlled, 4-arm, parallel	NMN ~doubled circulating NAD+, comparable to NR; nicotinamide did not; gut-microbiome and short-chain-fatty-acid changes observed	Not detailed in available primary text	Not verified	41540253	10.1038/s42255-025-01421-8
Pencina et al., 2023, JCEM[12]	Overweight/obese adults, ≥45y	30	2,000 mg/day	1,000 mg twice daily	28 days	RCT, double-blind, placebo, 2:1 randomization	NAD+ significantly ↑, returning toward baseline ~28 days after last dose; ↓ LDL, non-HDL cholesterol, body weight, diastolic blood pressure; no significant change in strength measures	No serious adverse events; 24 on-treatment adverse events in 15 of 30 participants (2 moderate, remainder mild); frequency similar across groups	Industry funded / author employment	36740954	10.1210/clinem/dgad027

Funding and conflicts of interest, in full

The compact labels above summarize what each trial's own published conflict-of-interest statement says, verified directly against the primary source. "Not verified" means no funding/COI statement could be independently located in the primary material accessible for this page — it is not a claim that a conflict exists or does not exist.

Igarashi 2022: three co-authors (Y. Fukamizu, T. Sato, T. Sakurai) are employees of Mitsubishi Corporation Life Sciences Limited, a manufacturer of NMN ingredients.
Okabe 2022: four co-authors (K. Yaku, Y. Uchida, Y. Fukamizu, T. Sato/T. Sakurai per the paper's disclosure) are employees of Mitsubishi Corporation Life Sciences Limited.
Katayoshi 2023: the authors state "no competing interests."
Akasaka 2023: no funding or conflict-of-interest statement could be located in the primary abstract record accessible for this page.
Morifuji 2024: three co-authors (M. Morifuji, S. Higashi, M. Nagata) are employees of Meiji Holdings Co., Ltd., a food and nutrition company.
Yoshino 2021: a co-author discloses receiving "a part of patent-licensing fees from MetroBiotech (USA) and Teijin Limited (Japan)"; another co-author is listed as an inventor on an NMN-related patent application. This study was not funded by those companies, but the disclosed royalty interest is reported here for transparency.
Huang 2022 (Uthever): the sole author was, per the paper's own disclosure, "employed by Effepharm (Shanghai) Co., Ltd.," which markets NMN under the Uthever brand.
Liao 2021: the authors state "none of the authors had a conflict of interest."
Yi 2023: one co-author is an employee of Abinopharm, Inc.; two co-authors are employees of Aba Chemicals, Co. — both NMN ingredient suppliers.
Pencina 2023 (pilot, J Gerontol A Biol Sci Med Sci): no funding or conflict-of-interest statement could be independently located for this specific paper. Other trials from an overlapping author group (below) disclose funding from Metro International Biotech; this is noted for context but is not treated as confirmed for this specific study without its own disclosure.
Fukamizu 2022: the study was conducted at Pharma Foods Co., Ltd. and funded by Mitsubishi Corporation Life Sciences Limited, which also supplied test material; five co-authors are Mitsubishi Corporation Life Sciences Limited employees.
Christen 2026: no funding or conflict-of-interest statement could be independently located in the primary material accessible for this page.
Pencina 2023 (JCEM): "This study was funded by Metro International Biotech." A senior co-author is disclosed as a "consultant and equity owner in Metro International Biotech," the maker of the MIB-626 formulation tested in the trial; other co-authors report research funding from the same sponsor.

Industry funding does not by itself invalidate a study, and none of the disclosures above indicate concealed or misrepresented funding — every conflict noted here was disclosed by the study's own authors in the publication. It is reported so readers can weigh it alongside the findings themselves.

What the studied-dose range actually shows

Across these trials, studied daily doses span roughly an eightfold range — from 250 mg/day up to 2,000 mg/day — with 250 mg/day and the 300–900 mg/day band the most frequently used.[1][8] Higher doses have generally been tested over shorter durations (14–28 days), while the lowest common dose, 250 mg/day, has the longest track record among these trials, extending to 24 weeks.[4] No trial has directly compared a full range of doses against each other over a long duration in the same population, and no trial was designed to identify a minimum effective dose or a ceiling dose for any specific health outcome.

Why this page does not recommend a dose

A recommended dose requires evidence of a specific outcome, a demonstrated dose-response relationship for that outcome, and safety data at the recommended level over a realistic duration of use. For NMN, no combination of trials currently meets that bar for any specific consumer-relevant outcome. Some trials report dose-dependent increases in blood NAD+ within their own study population,[8] but a marker increasing in a dose-dependent way is not, by itself, evidence that a specific dose produces a specific health benefit. Converting any figure in the table above into a personal recommendation would go beyond what these trials were designed to show.

Item added to your cart