LongevityComparisonSep 2026

NAD+ Injections: Subcutaneous vs IV, What Trials Used

NAD+ injections get sold as one idea, but the route changes almost everything about what has actually been studied. NAD+, nicotinamide adenine dinucleotide, is a coenzyme found in every living cell, not a peptide and not an FDA-approved injectable drug.¹² This page compares the routes people ask about, subcutaneous, intravenous, intramuscular, and the nasal spray, on one narrow question: what did real trials and drug registries actually use? It lays out the doses on record, shows where the human evidence is thin, and marks where there is no comparison data at all. It does not tell anyone what to take.

This is general education, not medical advice. There is no FDA-approved NAD+ injection, and in Vietnam the infusion technique is not licensed.²²⁶ Every dose figure below is a number a study or a drug registry recorded, quoted so the evidence level is visible, never a dose to copy. Nothing here is an instruction, a schedule, or a source.

FDA-approved NAD+ injections

Registered trial with a subcutaneous arm

50-fold

Spread in the doses trials recorded

Subcutaneous vs IV: what actually separates them

The honest headline is asymmetry. Intravenous NAD+ carries almost the entire human record: two small randomized trials in heart failure, a pair of tolerability studies in healthy adults, and several Chinese trials that register the molecule as Coenzyme I for Injection. Subcutaneous NAD+ appears in exactly one registered study, and that study is still recruiting and measures injection pain rather than absorption or any health outcome.⁶ So a fair subcutaneous versus intravenous comparison is not a contest of results. It is a map of what each route has and has not been tested for.

The single most important gap: no published study compares how much NAD+ reaches the blood after a subcutaneous injection versus an intravenous one. The only trial that gives NAD+ by both routes head to head is built to record how much the injection hurts, not how much of it is absorbed.⁶ Anyone claiming one route is more effective than another is going past the evidence.

Route	What is on record	Evidence status
Subcutaneous	One recruiting trial arm, with an injection-pain endpoint (NCT06919328)	No results posted, and no absorption or outcome data for this route
Intravenous	Heart-failure trials, tolerability studies, and several Chinese dose-escalation trials	Several small trials; the recorded doses are in the list below
Intramuscular	The same recruiting trial as subcutaneous (NCT06919328)	No results posted
Nasal spray	One 90-day combination pilot (NCT07475546)	Bundled with other drugs, so no isolated result
Transdermal patch	One completed post-COVID pilot using skin patches (NCT04604704)	A patch, not an injection; no isolated result reported here
Oral, for context	The precursors NR and NMN, not NAD+ itself	Where most human data sits; a different question from injected NAD+

The row that surprises people is the last one. Most of the human data on raising NAD+ is about swallowed precursors, NR and NMN, not about injecting NAD+ at all. A 2026 systematic review of 113 intervention studies put the injection question bluntly: "No eligible outcomes trials evaluated intravenous or intramuscular NAD+ itself for anti-aging or wellness indications."⁵ That sentence frames everything below.

What NAD+ injections are

NAD+, nicotinamide adenine dinucleotide, is a coenzyme built from a nicotinamide nucleotide joined to an adenine nucleotide, molecular formula C21H27N7O14P2, listed by PubChem under the name Nadide at compound ID 5892.¹ It is a dinucleotide, not a peptide: it is not built from amino acids, and it is not the same class of molecule as the research peptides often sold beside it. That distinction is worth stating up front because injection guides for peptides do not carry over to it.

There is no FDA-approved NAD+ injection. The openFDA label database holds exactly one product carrying the substance nicotinamide adenine dinucleotide, and it is an oral multivitamin; the same substance restricted to the intravenous route returns "No matches found."² DailyMed returns no single-ingredient NAD+ injection either, only multi-ingredient homeopathic products.³ In 2019 the FDA evaluated NAD as a bulk substance for pharmacy compounding and proposed that it not be added to the list, stating the nominated substances "do not meet the criteria for inclusion." That was a proposed rule, and this page does not claim it was later finalized.⁴

The usual reason given for injecting rather than swallowing NAD+ is bioavailability, but the fetched evidence does not actually establish that swallowed NAD+ is destroyed in the gut, a claim repeated widely online. What the research shows is narrower: in one mouse study, researchers switched to intravenous nicotinamide riboside precisely because oral NR had failed to raise muscle NAD+ in earlier clinical trials (animal study, mice, intravenous, PMID 35198907).²² For the mechanism of NAD+ itself and the ways people try to raise it, the NAD+ profile and the NAD+ benefits guide cover that ground. This page stays on the injection routes.

What trials actually used, route by route

Here is the full set of injected and infused NAD+ doses on record in registered or published human work, each quoted so the evidence level shows. Every figure is a dose a study or a registry recorded, not a recommendation.

A recruiting head-to-head trial in up to 70 healthy adults gives 100 mg NAD+ in 2 mL bacteriostatic water by subcutaneous, intramuscular, or intravenous-push injection, each against a matched bacteriostatic-water placebo (registered trial dose, not a recommendation, NCT06919328).
A placebo-controlled trial in 53 healthy adults gave 500 mg NAD+ in 500 mL of saline as a single intravenous infusion (registered trial dose, not a recommendation, NCT06382688).
A randomized, placebo-controlled trial in 180 adults with ischaemic cardiomyopathy heart failure used intravenous NAD+ at 10 mg per day for 7 days (trial dose, not a recommendation, PMID 40954388).
A trial in 60 older heart-failure patients used 50 mg NAD+ in 50 mL of normal saline for 7 days (trial dose, not a recommendation, PMID 39228487).
A Chinese phase 1 trial escalated intravenous Coenzyme I for Injection through 10, 20, and 50 mg per day for 21 consecutive days after cord blood transplant (registered trial dose, not a recommendation, NCT06558253).
A trial in unresectable liver cancer gave injectable coenzyme I at 100 mg per day by IV on days 1 to 3 of each 21-day cycle, alongside two cancer drugs (registered trial dose, not a recommendation, NCT06590844).
One registered longevity pilot uses an NAD+ nasal spray at 30 mg per day for 90 days, bundled with other agents (registered trial dose, not a recommendation, NCT07475546).
A completed post-COVID pilot used 400 mg NAD+ solutions on iontophoresis skin patches worn 4 to 6 hours once a week, which is a patch, not an injection (registered trial dose, not a recommendation, NCT04604704).

Two registered Chinese trials, one in vascular aging and one in sudden sensorineural hearing loss, list 7 consecutive days of infusion but state no dose in the registry record, and the hearing-loss record does not even name the route.¹³¹⁴ That is not an omission on this page; it is what those records actually contain, and where an abstract or registry gives no figure, this page says so rather than inventing one.²⁰ Across the injected human record, the per-administration amount runs from 10 mg to 500 mg, a fiftyfold spread with no shared rationale visible in any record.⁷⁸¹¹¹² The one completed study that delivered NAD+ through the skin used iontophoresis patches rather than a needle, which is why it sits outside that injection range entirely.¹⁶

A caution on the animal numbers: the rodent NAD+ studies you will see cited elsewhere dose by body weight, for example 75 mg/kg (animal study, rodent, intraperitoneal, PMID 41380826) or 250 mg/kg (animal study, mice, intraperitoneal, PMID 34174704) given into the abdominal cavity. Human trials use flat milligram amounts, and the intraperitoneal route that carries most of the rodent work has no human equivalent in these products. There is no valid way to convert a per-kilogram animal dose into a human dose, and this page does not.

Subcutaneous NAD+ in humans: one recruiting trial

Everything the human literature currently offers on subcutaneous NAD+ sits in a single registered study, NCT06919328, sponsored by the Nutraceuticals Research Institute and still recruiting up to 70 healthy adults.⁶ It gives 100 mg NAD+ in 2 mL bacteriostatic water by three routes at once, subcutaneous, intramuscular, and intravenous push, each against a matched bacteriostatic-water placebo by the same route (registered trial dose, not a recommendation, NCT06919328). That head-to-head design is exactly what a subcutaneous versus intravenous question needs, which is why it matters here.

The catch is what it measures. Both primary outcomes are discomfort: the McGill Pain Questionnaire and an open-ended subjective-discomfort measure, each taken one minute after the injection is complete.⁶ It is a tolerability study, not an efficacy study, and no results are posted. So the honest read on subcutaneous NAD+ is that its comfort against other routes is being formally studied for the first time, and its absorption, its blood levels, and any health effect are simply not on the human record yet. A separate tolerability trial measured total NAD by dried blood spot, but only for the intravenous route, so it does not close the subcutaneous gap.⁷

The NAD+ nasal spray

The nasal spray is a common search, so it is worth being precise about what exists. One registered human study uses an NAD+ nasal spray at 30 mg per day over 90 days (registered trial dose, not a recommendation, NCT07475546), run by AgelessRx in up to 30 subjects.¹⁵ The problem for anyone hoping to judge the spray is that it never appears alone: it is bundled with naltrexone, metformin, and B12, and in a third arm with rapamycin, a glutathione patch, and a proprietary supplement. Nothing in that design can attribute an effect to the spray by itself, and no results are posted.

The only study that isolates intranasal NAD is not a human one. In mice with chemically induced loss of smell, plus cultured human olfactory stem cells, intranasal NAD was reported to improve olfactory function and repair the olfactory lining, but the abstract states no dose (animal study, mice, intranasal, PMID 42380286).¹⁷ That is the whole of the isolated evidence for the nasal route: one mouse and cell study with no dose, and one human pilot where the spray is one ingredient among several.

What the evidence does and does not show

Start with the strongest single statement available. The 2026 systematic review that pooled 33 human and 80 rodent intervention studies concluded that "No eligible outcomes trials evaluated intravenous or intramuscular NAD+ itself for anti-aging or wellness indications." It found one nonrandomized intravenous NMN study and a single intravenous NAD+ pharmacokinetic pilot that it counted only as contextual evidence.⁵ The review also noted that human tolerability language, "generally well tolerated over weeks to months," applied to oral NR and NMN, while functional outcomes were "heterogeneous and often null or endpoint-specific."⁵ None of that is about injected NAD+.

The injected NAD+ trials that do exist are small and mostly outside the wellness framing. In the largest, a randomized, placebo-controlled trial in 180 heart-failure patients with an ejection fraction of 45 percent or below, the intravenous NAD+ group reached a left ventricular ejection fraction of 45.44 percent versus 42.44 percent on placebo at one month (p = 0.024), with a six-month composite MACCE rate, major adverse cardiac and cerebrovascular events, of 14.6 percent versus 24.7 percent that did not reach significance (p = 0.089). That is a reported trial result in a specific patient group, not a general claim about NAD+, and not a reason for anyone to seek an infusion (trial result, not a recommendation, PMID 40954388).⁸ A second trial in 60 older heart-failure patients reported that its NAD+ group did better than saline on cardiac markers, but the difference was not statistically significant (PMID 39228487).⁹ A 50-case open-label pilot of intravenous NAD+ combined with an enkephalinase inhibitor in substance use disorder reported improved craving, anxiety, and depression scores, but it had no placebo group, and its dose, schedule, and duration are not stated in the abstract, so no figure can be quoted (PMID 36118157).¹⁰ A 2022 narrative review argued for the intravenous route in post-acute COVID-19 without presenting a trial (PMID 35769168).²⁴

The animal picture is larger but points sideways from what these clinics sell. Rodent NAD+ studies dose by body weight and use the intraperitoneal route: 75 mg/kg given once at reperfusion in a spinal-cord injury model reduced injury markers (animal study, rodent, intraperitoneal, PMID 41380826), and 250 mg/kg daily in an autoimmune model lessened disease severity (animal study, mice, intraperitoneal, PMID 34174704).¹⁸¹⁹ A negative result deserves the same weight: the reduced form NADH, not NAD+, at 150 mg/kg by the same route did not speed anaesthesia recovery in mice (animal study, mice, intraperitoneal, PMID 39789056).²¹ One study did not even state the route, describing only daily NAD+ injections in mice (PMID 33290962).²⁰ Much of the injectable animal work is on precursors, not NAD+: intravenous NMN at 300 mg/kg in mice spiked serum NMN that fell within 15 minutes while serum NAD stayed unchanged (animal study, mice, intravenous, PMID 35593333), and the intravenous nicotinamide riboside study above raised muscle NAD+ without improving respiratory capacity or insulin sensitivity (animal study, mice, intravenous, PMID 35198907).²³²²

Put together, the evidence is a strong biological rationale, a handful of small trials in defined medical conditions, and a rodent literature that uses a route and a dosing basis with no human counterpart. It is not a demonstrated wellness or anti-aging benefit for any injection route, which is exactly what the systematic review said.⁵

Side effects and the safety unknowns

There is no approved US label for an NAD+ injection, so there is no official side-effect list to quote, for any route.
The one head-to-head route trial treats injection discomfort as its main outcome, pain and subjective comfort one minute after the shot (NCT06919328), which is the honest safety framing for a subcutaneous versus IV comparison right now.
The dose-escalation transplant trial defines its stopping rule by harm, a treatment-related grade 3 or higher reaction, but posts no results, so no tolerated dose can be reported (NCT06558253).
The only adverse signal in the fetched animal set is from the reduced form NADH, not NAD+: 150 mg/kg intraperitoneally decreased activity in mice after anesthesia (animal study, mice, intraperitoneal, PMID 39789056).
In a mouse pharmacokinetic study, intravenous NMN at 2 g/kg produced no acute toxicity over 14 days (animal study, mice, intravenous, PMID 35593333), which is animal data on a precursor, not human safety data on NAD+.

The sharpest safety point comes from Vietnamese clinicians quoted in the reporting below: there is no established target for the normal physiological NAD+ level when it is infused into a person, so there is no research on what too much or too little would do, and because the technique is unlicensed, much of the solution used is hand-carried and hard to verify for safety.²⁷ That is a different and more basic gap than a side-effect list. It is the absence of the reference point a side-effect list would be measured against.

NAD+ injections in Vietnam

NAD+ has no Vietnamese marketing authorization, and the infusion technique itself is not licensed. In July 2026 the Ministry of Health stated it has not licensed the technique of infusing NMN or NAD+ in treatment or cosmetic care, and asked the Hanoi health department to inspect, with law enforcement, a cosmetic facility reported to be advising the infusions without a licence.²⁶ The Ho Chi Minh City Department of Health has said the product is not licensed for circulation in Vietnam and appears in no Ministry of Health treatment guideline.²⁷

What inspections found is the most useful fact for a Vietnamese reader. At a dermatology clinic in District 3 of Ho Chi Minh City, inspected in 2024, the clinic told inspectors that the solutions billed to two patients as NAD+ were in fact ordinary B-vitamin infusions, vitaplex, pantogen, and polymia kabi, costing between 60,000 and 125,000 dong, and it could not produce the NAD+ product or any invoice for it. At a District 10 polyclinic, inspectors found three patients receiving an unlabelled yellow solution, alongside vials marked "NDA+ 1.000 mg" and glutathione that lacked origin and quality information.²⁷ In other words, part of the risk is not the pharmacology of NAD+ at all; it is not knowing what is in the bag.

On price, the honest sourcing matters. A newspaper reported that a social-media account quoted 13 to 16 million dong per infusion session, and 39 to 48 million dong for a three-session course, and that one international polyclinic was said to offer around ten NAD+ infusion packages at 6.8 to 9 million dong per session, with staff declining to commit to any result.²⁷ Those are a reporter account of what advisers said, not a clinic published price list; no clinic own price page was verified for this guide, so nothing here should be read as a current quoted price.

In retail pharmacy, the chain Long Chau lists no NAD+ injection of any route: a search for NAD+ returns only products matching the letters AND, and a search for nicotinamide returns niacinamide skincare and oral B-vitamin drugs. Oral NMN supplements do appear, filed under functional food rather than medicine, for example NMN Premium 21600 Jpanwell and NMN PQQ Kenko, each listed around 8,900,000 dong per box, but those are swallowed capsules, not NAD+ injections.²⁵ Two other chains, Pharmacity and An Khang, could not be read from this box during research, a tooling limit, not evidence that they carry or do not carry anything. The Vietnamese terms to recognize are truyen NAD+ for the infusion and thuc pham chuc nang for the functional-food category the oral products sit in.

The short version

NAD+ is a coenzyme, a dinucleotide, not a peptide, and there is no FDA-approved NAD+ injection of any route.
Intravenous NAD+ carries almost the entire human record; subcutaneous NAD+ appears in one recruiting trial that measures injection pain, not absorption or effect.
No published study compares how much NAD+ reaches the blood by the subcutaneous versus the intravenous route.
Recorded human doses run from 10 mg to 500 mg per administration, a fiftyfold spread with no shared rationale; rodent per-kilogram doses do not convert to people.
A 2026 systematic review found no outcomes trials of injected NAD+ for anti-aging or wellness; the trials that exist are small and in specific conditions.
In Vietnam the infusion is not licensed, and inspectors found solutions sold as NAD+ that were actually ordinary B-vitamin drips.

Frequently asked questions

Is subcutaneous or intravenous NAD+ better?+

There is no evidence-based answer, and this is not a recommendation. No published study compares how much NAD+ reaches the blood by the subcutaneous versus the intravenous route, and the only registered trial that gives NAD+ by both routes head to head measures injection discomfort, the McGill Pain Questionnaire and subjective comfort one minute after the injection, rather than absorption or any health outcome (NCT06919328). Intravenous NAD+ carries almost the entire human record; the subcutaneous route appears in that one recruiting trial and has no posted results.

What doses of NAD+ have injection trials actually used?+

Across registered and published human work, the per-administration amount runs from 10 mg to 500 mg, a fiftyfold spread with no shared rationale visible in any record. Reported figures include intravenous NAD+ at 10 mg per day for 7 days in a heart-failure trial (trial dose, not a recommendation, PMID 40954388), 50 mg in 50 mL of saline for 7 days in older heart-failure patients (trial dose, not a recommendation, PMID 39228487), 100 mg in 2 mL bacteriostatic water by subcutaneous, intramuscular, or IV push in a recruiting tolerability trial (registered trial dose, not a recommendation, NCT06919328), and 500 mg in 500 mL of saline as a single infusion (registered trial dose, not a recommendation, NCT06382688). These are figures studies recorded, never recommendations.

Is there an FDA-approved NAD+ injection?+

No. The openFDA label database holds exactly one product carrying the substance nicotinamide adenine dinucleotide, an oral multivitamin, and returns no matches when the same substance is restricted to the intravenous route. DailyMed lists no single-ingredient NAD+ injection either. In 2019 the FDA proposed that NAD not be added to the list of bulk substances allowed for pharmacy compounding, stating it does not meet the criteria for inclusion; that was a proposed rule.

Does the NAD+ nasal spray have evidence behind it?+

Only one registered human study uses an NAD+ nasal spray, at 30 mg per day over 90 days (registered trial dose, not a recommendation, NCT07475546), and it is bundled with naltrexone, metformin, B12, and in one arm rapamycin and other agents, so no result can be attributed to the spray on its own. The only study that isolates intranasal NAD is in mice with chemically induced loss of smell, where the abstract reports improvement but states no dose (PMID 42380286). There is no isolated human nasal-spray result to cite.

Are NAD+ injections legal in Vietnam?+

No. Vietnam Ministry of Health has stated it has not licensed the technique of infusing NMN or NAD+ for treatment or cosmetic use, and the Ho Chi Minh City Department of Health says the product is not licensed for circulation in Vietnam and appears in no treatment guideline. Health inspectors in 2024 found that solutions billed to patients as NAD+ were in fact ordinary B-vitamin infusions costing 60,000 to 125,000 dong, and that clinics could not always produce the product or its invoices.

Sources and references

The load-bearing sources are listed here; the full reference list, which every citation marker links to, follows at the foot of the page. Numbers match the markers in the text.

2.openFDA drug label API. A search for the substance nicotinamide adenine dinucleotide returns one oral multivitamin and no matches for the intravenous route (fetched 2026-09-06). open.fda.gov/apis/drug/label
4.US Federal Register, FDA proposed rule 2019-18951. FDA proposed that NAD not be added to the 503A bulk substances list for pharmacy compounding, stating the nominated substances do not meet the criteria for inclusion. federalregister.gov / 2019-18951
5.Systematic review of NAD supplementation, Ageing Research Reviews, 2026. No eligible outcomes trials evaluated intravenous or intramuscular NAD+ itself for anti-aging or wellness indications. PMID 41655607. pubmed.ncbi.nlm.nih.gov/41655607
6.ClinicalTrials.gov NCT06919328. The only registered study with a subcutaneous NAD+ arm: 100 mg in 2 mL bacteriostatic water by IM, SQ, and IV push, with injection pain as the primary endpoint. Recruiting, no results posted. clinicaltrials.gov / NCT06919328
8.IV NAD+ in ischaemic cardiomyopathy heart failure, randomized, placebo-controlled, n = 180. Intravenous NAD+ 10 mg per day for 7 days. American Journal of Cardiovascular Drugs, 2026. PMID 40954388. pubmed.ncbi.nlm.nih.gov/40954388
15.ClinicalTrials.gov NCT07475546. The only registered human NAD+ nasal-spray study, 30 mg per day for 90 days, bundled with other agents in a longevity pilot. No isolated result. clinicaltrials.gov / NCT07475546
26.Bao Dan tri. Vietnam Ministry of Health has not licensed the technique of infusing NMN or NAD+ for treatment or cosmetic use, and ordered an inspection of a cosmetic facility advertising it. Reported 2026-07-31. dantri.com.vn
27.Bao Thanh Nien. HCMC Department of Health inspection findings: solutions billed to patients as NAD+ were ordinary B-vitamin infusions, and clinics could not always produce the product or its invoices. Reported 2024-08-03. thanhnien.vn