In plain words

In five placebo-controlled trials, people who took NMN by mouth for two weeks to three months had more NAD or its breakdown products in their blood than people on placebo. In the hours after a single dose, two studies saw no rise in NMN itself in the blood.

The short answer

All five randomized trials found higher blood NAD or NAD breakdown products with NMN than with placebo: 15 people on 1 g a day for 14 days, 32 on 1,000 mg once or twice daily for 14 days, 34 and 60 on 250 mg a day for 12 weeks, and 80 on 300 to 900 mg a day for 60 days. In the hours after one dose, blood NMN itself did not change in the 14-day trial of 15 people or in a before-after study of 13 people. Certainty is moderate: the trials agree, but they are small and none ran past 12 weeks.

Where the evidence sits

Higher NAD or breakdown products

■ Randomized trial, Christen et al., 2026 · 15 people · 14 days · blood NMN flat over 4 hours
■ Randomized trial, Pencina et al., 2023 · 32 people · 14 days
■ Randomized trial, Katayoshi et al., 2023 · 34 people · 12 weeks
■ Randomized trial, Morifuji et al., 2024 · 60 people · 12 weeks
■ Randomized trial, Yi et al., 2022 · 80 people · 60 days
■ Before-after study, Wakabayashi et al., 2026 · 13 people · 60 min · blood NMN itself unchanged
■ Before-after study, Irie et al., 2019 · 10 people · 5 h after each intervention
◪ Pooled analysis, Zhang et al., 2024 · may overlap studies counted here

■ measured by the source's own study · ◪ pooled analysis, may overlap studies counted here · □ reported, reviewed or argued, no new data

Evidence weight7474 / 100?

Who says what

SourceFindingDesignPeopleDose and lengthFunding and independence
Christen, 2026NAD+ higherRandomized trial151 g, 14 daysNot in ledger
Pencina, 2023NMN higherRandomized trial321,000 mg, 14 daysNot in ledger
Katayoshi, 2023Nicotinamide higherRandomized trial34250 mg, 12 weeksNot in ledger
Morifuji, 2024NAD+ higherRandomized trial60250 mg, 12 weeksNot in ledger
Yi, 2022NAD higherRandomized trial80300 to 900 mg, 60 daysNot in ledger
Wakabayashi, 2026NMN unchangedBefore-after study13250 mg onceNot in ledger
Irie, 2019Breakdown products roseBefore-after study10Up to 500 mg onceNot in ledger
Zhang, 2024NAD higherMeta-analysis; may include the trials above513Not in ledgerNot in ledger
Christen, 2026

NAD+ higher

Design
Randomized trial
People
15
Dose and length
1 g, 14 days
Funding
Not in ledger
Pencina, 2023

NMN higher

Design
Randomized trial
People
32
Dose and length
1,000 mg, 14 days
Funding
Not in ledger
Katayoshi, 2023

Nicotinamide higher

Design
Randomized trial
People
34
Dose and length
250 mg, 12 weeks
Funding
Not in ledger
Morifuji, 2024

NAD+ higher

Design
Randomized trial
People
60
Dose and length
250 mg, 12 weeks
Funding
Not in ledger
Yi, 2022

NAD higher

Design
Randomized trial
People
80
Dose and length
300 to 900 mg, 60 days
Funding
Not in ledger
Wakabayashi, 2026

NMN unchanged

Design
Before-after study
People
13
Dose and length
250 mg once
Funding
Not in ledger
Irie, 2019

Breakdown products rose

Design
Before-after study
People
10
Dose and length
Up to 500 mg once
Funding
Not in ledger
Zhang, 2024

NAD higher

Design
Meta-analysis; may include the trials above
People
513
Dose and length
Not in ledger
Funding
Not in ledger

Did blood NAD rise over weeks of use?

Yi and colleagues (2022) gave 80 healthy middle-aged adults 300, 600 or 900 mg a day, or placebo, for 60 days:

Blood NAD concentrations were statistically significantly increased among all NMN-treated groups at day 30 and day 60 when compared to both placebo and baseline (all p ≤ 0.001).

Morifuji and colleagues (2024) found higher blood NAD+ and metabolites than placebo in 60 adults on 250 mg a day for 12 weeks.

Does NMN itself reach the blood?

Christen and colleagues (2026) sampled blood for 4 hours after a 1 g dose in 15 people:

the low NR and NMN concentrations in whole blood ( < 2 µM) were not affected

After 14 days at 1 g a day, whole-blood NAD+ in those 15 people was 43.1 µM above placebo. Wakabayashi and colleagues (2026), in a before-after study of 13 people given 250 mg once, saw no change in blood NMN over 60 minutes; NAD+ was higher at 30 minutes when NMN was taken under the tongue.

Were the trials measuring the same molecule?

Katayoshi and colleagues (2023) gave 34 adults 250 mg a day or placebo for 12 weeks:

NAD+ and NMN in the serum were detectable in both groups, but the values were below the lower limit of quantification.

Serum nicotinamide, an NAD+ metabolite, was higher with NMN in that trial. Yi measured serum NAD+ plus NADH, and Christen measured whole blood. Irie and colleagues (2019) measured two breakdown products, which rose with dose in a before-after study of 10 people given 100, 250 or 500 mg once and followed for 5 hours.

Why the sources disagree

No independent measurements contradict each other; the studies differ in timing and measure. The two single-dose studies followed people for 1 to 4 hours and saw no rise in blood NMN itself, while Pencina found blood NMN higher in 32 adults after 14 days of 1,000 mg once or twice daily. The trials measured whole blood, serum or breakdown products. The ledger has no funding details for any source.

What would settle it

An independent, placebo-controlled trial of oral NMN at 250 and 1,000 mg a day for at least six months, measuring whole-blood NMN, NAD+ and breakdown products with one assay at baseline, 2, 6 and 24 hours after the first dose, then monthly, with a pre-registered analysis and disclosed funding.

How this page was made

The 8 sources are 7 human studies and one meta-analysis, a pooled estimate that may include the trials above, found through a PubMed search for human NMN studies; studies of other precursors, disease groups, animals and unpublished protocols were left out. Every claim was copied word for word from its source into the Longevity Circle ledger and checked against the original text, and a person approved each one. "Leaning" means the independent studies point mostly one way; certainty follows the GRADE levels used in medical reviews. Drafting was assisted by AI.

Changelog: 28 Sep 2026, first version, 8 sources.