Most 'NAD' studies never gave anyone NAD. Here is how to check in a minute.

2026-09-20

Go looking for research on NAD and you will find plenty. Trials, reviews, registry entries, news coverage. It looks like a well-studied molecule.

Then you check what those studies actually gave people. Something odd shows up.

Most of them did not give anyone NAD.

This is not a conspiracy. It is a naming problem. Once you know its shape, you can spot it in about a minute.

Four molecules, one name

At least four substances get discussed under NAD-related names. They are not variations on a theme. Each has its own formula, its own registry number, and its own body of research.

Name	PubChem ID	Formula
NAD+ (the coenzyme itself)	5892	C21H27N7O14P2
Nicotinamide riboside (NR)	439924	C11H15N2O5+
Nicotinamide mononucleotide (NMN)	14180	C11H15N2O8P
Nicotinamide	936	C6H6N2O
Nicotinic acid	938	C6H5NO2

Look at the formulas. NAD+ has 21 carbons and two phosphorus atoms. NR has 11 carbons and no phosphorus. These are not near neighbours.

NR and NMN are precursors, sold as supplements. Nicotinamide and nicotinic acid are the vitamin B3 forms. They have their own nutrition reference values, set by the Institute of Medicine.

A result from one is not a result for another. Everything below follows from that.

Step 1: find what was given, not what it was called

Titles are written by people. People reach for the famous name.

One registered trial, NCT02942888, is called "The Effects of NAD on Brain Function and Cognition." Its registered intervention is oral nicotinamide riboside. The title says the coenzyme. The study used a precursor.

The registry shows how common this is. We queried the ClinicalTrials.gov API for "nicotinamide adenine dinucleotide" as an intervention in September 2026. It returned 264 studies. We then read the registered intervention for every one:

50 name nicotinamide riboside or Niagen
10 name nicotinamide mononucleotide or NMN
only 14 name NAD at all

And one of those 14 is not about this subject in any way. NCT02010151 studies cardiopulmonary resuscitation. There, NAD stands for Neighborhood Access Defibrillation. Same three letters, completely different field.

Several of the rest are oral consumer products branded "NAD+", not the coenzyme being given.

Step 2: match the formula, not the name

Names mislead in both directions.

So use the identifiers instead. NAD+ is CID 5892. The formula is C21H27N7O14P2. The CAS number is 53-84-9.

Open the methods section. If the substance there has a different formula or a different CAS number, it is a different molecule. The title does not get a vote.

Step 3: check the route

This step is the least obvious. It matters as much as the other two.

Liu and colleagues ran isotope-tracer work in cells and mice in 2018. They report that intravenous NR or NMN "delivered intact molecules to multiple tissues, but the same agents given orally were metabolized to nicotinamide in the liver."

Read that twice. Oral and intravenous are not two doses of one experiment. They are two experiments. The molecule reaching the tissue is not the same molecule in each case.

So a study might have done any of three things. Given a precursor by mouth. Given a precursor intravenously. Or given the coenzyme itself. Treating those as one pile is how a shelf of papers turns into a claim none of them made.

What survives the check

Run all three steps across the human literature and it thins out fast.

The famous human studies are precursor studies. Trammell 2016 opened that work. It gave oral NR, and its authors call it the first clinical trial of NR pharmacokinetics in people. Dollerup 2018 randomised 40 men to NR or placebo for 12 weeks. It reported that insulin sensitivity, glucose production, glucose disposal and body composition were not improved. Both studied NR.

Only one located human study gave NAD+ directly.

Grant 2019 ran a six-hour intravenous infusion and tracked where it went. The authors are blunt about the gap they were filling. Before their work, they write, "no data are currently available on the fate of directly infused NAD+ in a human cohort."

What they found: NAD+ was cleared from plasma rapidly and completely for at least the first two hours. They describe themselves as surprised that nothing changed in plasma until after the two-hour mark.

That study measured where a molecule went. It was not built to measure whether anything happened to the people in it, and it did not.

This is not a complaint about the research

The precursor papers are not bad papers.

Trammell, Dollerup, Liu and the others did what they said they did. They named their compounds correctly in their methods. Several published null results, which is more than much of this field manages.

The trouble starts afterwards. A study of one molecule gets filed under the name of another. By the time it reaches a product page, "NAD" covers all of it.

One more piece of context. Commercial interests run through this literature in the ordinary way. ChromaDex supplied the NR and sponsored the Trammell study. Two of the Grant 2019 authors are tied to an NAD+ research company and a wellness centre. None of that makes a finding wrong. It is disclosed in the papers, and it is worth knowing.

The short version

Before accepting any sentence that starts "studies show NAD":

Find the intervention, not the title.
Match the formula or the CID, not the name.
Check the route: oral or intravenous, precursor or coenzyme.

It takes about a minute. Most of the time, the answer narrows what that sentence was entitled to say.

Sources

All articles