NAD+ protocols in the published literature

What amounts of NAD+ did the published studies use?

animal model

Two figures appear in the published summaries, and both are rodent figures written per kilogram of body weight. In the 2025 cuprizone experiment the mice in the cuprizone + NAD+ group received 250 mg/kg/day NAD+ intraperitoneally once a day, for as long as the poisoned chow was fed.1 In the 2025 anaesthesia experiment the figure was NADH (150 mg/kg, intraperitoneal)2, set against salt water as the comparison. The two other mouse papers name a molecule without giving a quantity in their summaries. One looked at acute supplementation of nicotinamide mononucleotide in mice made confused with a bacterial toxin, and its abstract gives no amount at all.3 The other gave nicotinamide riboside to Alzheimer's model mice, again with no figure in the summary. So the whole stated record is two numbers, both from mice and both injected, and neither of them describes a capsule or a clinic drip.

Who received those amounts?

human RCT

Mice made ill on purpose, and one group of patients with a named condition. That matters because an amount is chosen for a particular body with a particular problem, and it travels badly to anyone else. In the cuprizone work, six-week-old C57BL/6J mice were divided into three groups, two of them fed a chow that strips the coating off nerves.1 In the anaesthesia work the animals were adult male and female C57BL/6 mice (n = 8-10/group), put under with an inhaled anaesthetic.2 The only people were 80 CFS patients, CFS being chronic fatigue syndrome, a condition the trial report describes as severe disabling fatigue with no known cause.4 No healthy adult, no older adult as a group and no child appears among the people given any of these molecules in this material. A young mouse with damaged nerves and a tired adult buying capsules are different questions, and an amount picked for the first says nothing about the second.

How was NAD+ administered in the studies?

animal model

By injection into the belly of mice, which is a laboratory route rather than a clinical one, and by an unstated route in the one human trial. Both stated amounts went in the same way, with the cuprizone mice dosed by daily injection and the anaesthesia mice given a single shot. In that anaesthesia work the injection was given at baseline or during anesthesia, so the timing was tied to an operation rather than to a daily habit.2 The human trial summary describes supplementation taken twice a day without saying whether it came as a tablet, a capsule or anything else. That leaves the two forms people are actually sold, an oral product and an intravenous drip, without a stated route anywhere among these sources. A 2025 rodent paper names the gap from the consumer side, noting that with rising over-the-counter use of NAD, understanding their impact on anesthetic recovery becomes essential.2 An amount injected into a mouse cavity does not tell anyone what an amount swallowed or dripped into a vein would deliver, because the route decides how much arrives.

How often was NAD+ given, and for how long?

human RCT

The courses on record run from a single injection to eight weeks, and none runs longer. At the short end, the anaesthesia mice received one shot around the procedure, and the delirium mice were given their precursor as an acute course rather than a long one. In the cuprizone experiment the mice were fed the damaging chow for 4 weeks while the NAD+ injections were given once a day.1 The Alzheimer's model mice reached the longest animal course, where after supplementation with NR for 8 weeks their gut bacteria were measured again.5 In people the course was the same length, a proof-of-concept, 8-week, randomized, controlled, double-blind trial with doses twice daily.4 Eight weeks is therefore the ceiling in both species. A course length chosen to see whether a measurement moves is a design decision rather than a recommendation, and nothing in this material tells anyone what months or years of daily use would do.

Does an amount of NMN or NR count as an amount of NAD+?

animal model

Not directly, and this is where most of the numbers people quote go wrong. A precursor is raw material that the body converts into NAD+, so a milligram of precursor is not a milligram of the finished molecule. The delirium paper describes its compound as nicotinamide mononucleotide, a direct precursor of Nicotinamide adenine dinucleotide, and measured the animals rather than the conversion.3 The anaesthesia paper calls NADH a common NAD+ precursor, which is the molecule its 150 mg/kg figure belongs to.2 A 2025 heart failure review reports that supplementation with NAD+ precursors (e.g., β-nicotinamide mononucleotide (NMN), nicotinamide riboside, etc.) also significantly elevates myocardial NAD+ levels, myocardial meaning heart muscle.6 That sentence says the level rose, in work summarised by a review rather than run by its authors, and it does not say by how much per milligram. A label that lists an amount of NMN or NR is describing a different molecule from the one in its name, and there is no conversion among these sources to turn one into the other.

What are the registered human studies giving?

registered trial

Mostly precursors and branded products, and none of their amounts is reported in the material behind this page. Two registered entries name nicotinamide riboside rather than NAD+ itself. One, NR Supplementation and Exercise, is marked COMPLETED.7 Another, Effects of NR Supplementation on Metabolic Flexibility in Zone 2 Training, is marked RECRUITING.8 The only entry carrying a phase number for NAD+ itself is Nicotinamide Adenine Dinucleotide and Skeletal Muscle Metabolic Phenotype, filed as PHASE2 with a status of UNKNOWN.9 A completed entry with no published result is a study that stopped rather than one anybody can read. So the place where a human amount might eventually come from is a register of plans, and the plans mostly test a precursor rather than the molecule sold in the drip.

Can a blood or cell level stand in for a dose?

registered trial

It is the closest thing to a dose check the registered studies propose, and it measures something other than benefit. One entry, Effectiveness of Qualia NAD+ Supplementation on Intracellular NAD Levels, sets a named brand against a reading taken inside cells.10 Another, Nicotinamide Adenine Dinucleotide Responses to a Nutritional Supplement, is marked COMPLETED.11 Both are built to ask whether a product raises a level, which is the first step before anyone could ask how much raises it by how much. The heart failure review frames the same idea from the other side, noting that the inhibition of NAD+-degrading enzymes increases the tissue intracellular NAD+ content, so a level can rise because less is broken down rather than because more was taken.6 A higher reading is a measurement of the molecule rather than of a person feeling or working better, and the two can move separately.

Is there an NAD+ dosage chart?

animal model

Not one drawn from a study, and the reason sits in what the published work did not set out to do. A chart lists amounts for people of different sizes or goals, which requires trials that gave several amounts and watched what followed. Every stated amount in this material is a single level chosen by a team for one experiment in mice. The cuprizone team, for instance, tested one amount against saline and reported that supplementation with NAD+ increased the value on a water maze memory test (28.78% vs. 16.32%, p = 0.023) in those mice.1 One amount against a blank tells you whether that amount did something in that mouse model, and it says nothing about whether half or double would have done more, less or harm. A chart printed on a vial or a clinic menu is therefore a commercial document rather than a summary of measured amounts, because there are no graded amounts among these sources for it to summarise.

Why is a study amount not a protocol?

animal model

Because a protocol is a decision about a person, and every figure here was a decision about an experiment. The anaesthesia team picked its amount to answer one question, and the answer was negative for the outcome they hoped to see. Their conclusion was that this study demonstrates that NADH does not appear to hasten recovery from anesthesia, so the one stated NADH amount is attached to a result nobody would want to repeat.2 Per-kilogram figures from mice cannot be scaled into a figure for a person, and we do not convert them. Mice clear and use small molecules at a different rate from people, and no study among the research behind this page has measured how body size changes what an NAD+ injection delivers. A number lifted out of a mouse experiment keeps its decimal point and loses everything that gave it meaning: the species, the illness induced, the route, and the question being asked.

What was the starting amount in the studies?

animal model

There was no starting amount, because no experiment among these sources stepped up from a low figure to a higher one. A starting amount comes out of dose-finding work, where people or animals receive increasing quantities while a team watches for effect and harm. The nearest thing to an early-stage question is the delirium team's own framing, that whether Nicotinamide adenine dinucleotide supplementation may be beneficial for delirium has not been explored yet.3 That is a team asking whether to bother rather than a team asking how much. The human trial began everyone at the same twice-daily schedule from day one, which is a fixed design rather than a gradual increase. So a beginner amount offered for NAD+ has no experiment we could find behind it, and the absence is the finding rather than a gap waiting to be filled from somewhere else.

Is there a cycle or a break in the published work?

human RCT

No study among the research behind this page stopped NAD+ or a precursor and then kept measuring, so there is no off period to report. Cycling, meaning weeks on followed by weeks off, assumes somebody watched what returns to baseline after stopping. In the human trial the questionnaires on fatigue, pain and sleep were evaluated at baseline, and then reassessed at 4- and 8-weeks through self-reported questionnaires, which ends the record on the last day of supplementation.4 In the Alzheimer's model work the decreased diversity and perturbated microbial compositions were normalized in AD mice by the end of the course, and the paper reports that end point rather than what happened once supplementation stopped.5 A schedule of on and off weeks sold with a product is a pattern rather than a result, and these experiments are silent on it in both directions: they neither support a break nor show that one is unnecessary.

How do you reconstitute an NAD+ vial, and how many draws does it hold?

That part is arithmetic rather than evidence, and it is the one question on this page with a clean answer. Divide the milligrams of powder by the millilitres of diluent and the result is the concentration in milligrams per millilitre; multiply that by 1000 to express the same concentration in micrograms per millilitre. A 500 mg vial mixed with 5 mL therefore holds 100 mg/mL, and the same vial mixed with 10 mL holds 50 mg/mL. The number of draws a vial holds is the total milligrams divided by whatever amount is drawn each time, and that second figure is exactly the one the published work does not supply. Those calculations describe a solution rather than a target, and the reconstitution calculator on this site does the same division without suggesting an amount. One caution belongs to NAD+ in particular, because the label may name NAD+, NADH or a precursor, and the arithmetic is only as good as the identity of the powder.