NAD+: The Molecule in the Bag Never Gets Inside a Cell

What it is

NAD+ is nicotinamide adenine dinucleotide, PubChem CID 5892, C21H27N7O14P2, 663.4 g/mol, a redox cofactor rather than a peptide. Nicotinamide riboside and nicotinamide mononucleotide, the two things most people have actually taken, are not NAD+; the published literature describes them as precursors, boosters of NAD+ rather than the molecule under study here. (PMID 41540253) Hold those two apart for the rest of this piece, because every argument for a drip depends on letting them collapse together.

The bag does not put NAD+ inside a cell

Intact NAD+ does not enter cells. Extracellular NAD+ is broken down by ecto-enzymes, CD38 and CD73 among them, into nicotinamide mononucleotide, nicotinamide riboside and nicotinamide, and of those the nucleotides do not cross the plasma membrane at all. NAD+ does not cross. NMN does not cross. What crosses is the nucleoside, nicotinamide riboside, and the base, nicotinamide, with NMN getting in only once CD73 has stripped its phosphate off and turned it into NR, and those fragments rebuild the pool inside. (PMID 34142751)

It is the wardrobe that will not go up the stairs. The delivery men take it apart on the pavement, carry the panels up one at a time, and what arrives in the bedroom is panels, whatever the invoice said. I found no study in an infused human measuring what share of a dose arrives intracellularly as NAD+ against as a fragment, the number that decides whether a drip beats a capsule.

What the human studies did

Grant and colleagues (Front Aging Neurosci 2019;11:257) gave 750 mg of NAD+ intravenously over six hours, at approximately 2 mg per minute, to eight men aged 30 to 55, with three further men as controls. It is still the only peer reviewed human pharmacokinetic study of intravenous NAD+. Plasma NAD+ did not change at all for the first two hours, which the authors read as rapid and complete removal from the plasma. Two hours of infusion with nothing in the blood to show for it. It then climbed to about 400 percent above baseline by the end of the six hours, and no adverse events were observed in either the NAD+ or the saline group. (PMID 31572171)

Reyna and colleagues (Front Aging 2026;7:1652582) read the records of a commercial infusion clinic, where 500 mg of lyophilised NAD+ or NR was diluted in 500 mL of normal saline and infused on four consecutive days, six clients on NAD+ and eight on NR, no randomisation, no placebo arm. All six who received NAD+ reported moderate to severe symptoms during infusion: abdominal cramping, diarrhoea, nausea, vomiting, raised heart rate, throat pain, congestion and chest pressure. The NR group had minor tingling and cramping, and symptoms stopped when the infusion stopped. On tolerability grounds the NAD+ infusions had to be run far more slowly, 97 minutes on average against 37 for NR. (PMID 41704678)

Only one injected NAD+ study is sized to say anything about an outcome. Yu and colleagues (Am J Cardiovasc Drugs 2026, ChiCTR2200059169) randomised 180 adults with ischaemic cardiomyopathy, ejection fraction 45 percent or below, NYHA class II to III, to intravenous NAD+ at 10 mg per day for seven days or to placebo, on top of guideline-directed medical therapy. Ejection fraction at one month was 45.44 plus or minus 8.55 percent against 42.44 plus or minus 9.09 percent, p = 0.024, while the six-month composite of major adverse cardiac and cerebrovascular events was 14.6 percent against 24.7 percent, p = 0.089, not significant. (PMID 40954388) Note the dose. Ten milligrams a day.

The rate question, which is the useful one

Grant put 750 mg into eight men at about 2 mg per minute and recorded nothing at all. The clinic in Reyna 2026 put 500 mg into six clients at roughly 5 mg per minute and every one of them was ill. Bigger dose, slower, no symptoms; smaller dose, faster, moderate to severe symptoms in everybody. That contrast is my own arithmetic on their published figures and it is a hypothesis rather than a finding, and I found no study that has tested infusion rate as a variable for intravenous NAD+. The endotoxin problem described further down muddies even that, although sterility was not in question at Reyna's clinic.

Where 500 to 1500 mg came from

The intravenous convention traces to something specific: the Springfield Wellness Center BR+NAD addiction protocol, whose supporting 60-patient study gave NAD+ intravenously at 500 to 1500 mg daily for about ten consecutive days, five to ten hours daily, alongside vitamins, oral amino acids, NAC and as-needed medications, and was presented as a Society for Neuroscience poster in 2014 with no peer-reviewed publication I could find. NAD+ was one of several things going in.

Vendors and clinics converge on 250 to 1000 mg per intravenous session over two to four hours, and roughly 100 to 500 mg per subcutaneous injection. That is convention, reported here as convention. The guide I read most closely cites no study at all for its intravenous range, and justifies its subcutaneous range with a trial of oral NMN at 300 to 600 mg per day, a different molecule by a different route.

Now go and look at what hospitals have registered. China approved an injectable as Coenzyme I for Injection, NMPA registration H41024721, for coronary artery disease, leukopenia and myocarditis, in a 5 mg vial. NCT06776510 gives 100 mg intravenously daily for a week in immune thrombocytopenia, the highest registered intravenous human dose these searches found. The one adequately sized randomised trial used 10 mg a day, and against that anchor the convention is fifty to a hundred and fifty times larger, and I found no published human dose-finding or dose-response study of injected NAD+ by any route to account for the gap. For the subcutaneous route, which is the most widely sold home protocol, I found no published human study at all.

The precursor trials, which are about something else

A 2026 Nature Metabolism trial ran the three oral options against each other in 65 healthy participants over 14 days, at 0.5 g nicotinamide, 1 g NR and 1 g NMN once daily, and found NR and NMN comparably raise circulating NAD+ while nicotinamide does not. (PMID 41540253) All oral, all precursors, none of it evidence about injected NAD+. Nor is it flattering on its own terms: a 2023 Science Advances review of the entire human oral NR literature concluded that supplementation has displayed few clinically relevant effects. (PMID 37478182)

What the regulators have done

NAD+ is not an FDA-approved drug for any indication in the United States. In a proposed rule published 5 September 2019 under docket FDA-2018-N-4845, FDA proposed that it not be included on the section 503A bulk drug substances list, one of 26 substances proposed for exclusion, because it degrades on exposure to light, moisture, alkaline pH or room temperature, because the non-clinical and clinical data were inadequate, and because no published studies supported the nominated use. Whether that proposal was finalised I could not confirm; every fda.gov page I tried returned a 404, so no compounding category appears here.

Then the manufacturing. FDA has published a compounding notice saying compounders are using food-grade NAD+ to make intravenous products, that food-grade material is unsuitable for sterile compounding without appropriate processing, and that it has received adverse event reports of severe chills, shaking, vomiting and fatigue after NAD+ injectable drugs, some requiring medical treatment, consistent with excessive endotoxin. It classified a GenoGenix LLC recall of NAD+ for Injection, 100 mg/mL and 200 mg/mL, as Class I for elevated endotoxin, recall D-0094-2026.

My read

What decides this compound for me is the membrane. You are paying for a line, a chair and four hours to receive a molecule that cannot get into a cell without being taken to pieces first, and the pieces are sold in capsules. The precursor literature does the persuading, and the precursor literature is oral.

Yu 2026 is the one result I take seriously, and carefully: one trial, one country, 10 mg a day for a week in people with damaged hearts already on full medical therapy, a three percentage point ejection fraction difference at a month, an event rate that missed significance, no replication I could find. A cardiology finding at a hospital dose. I cannot see what it says about a gram in a wellness clinic.

The experiment I would want funded is the boring one: the same dose at three infusion rates, symptoms recorded. It would cost very little and it has been sitting in plain sight since Reyna published.

If you have had a drip, dig out the paperwork: milligrams, bag volume, and minutes on the pump. How fast did yours run, and did you feel it?

---

This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.