NAD patches promise the NAD+ of an infusion without the needle or the hours in a chair, and the pitch rests on one assumption: that NAD+ can pass through intact skin into the blood. That assumption can be checked against what is known about skin, and the answer is not encouraging.
What people given NAD+ by vein actually showed is covered in NAD+ injections benefits. This page is about the skin.
Why skin keeps most molecules out
The outer layer of skin, the stratum corneum, is a thin wall of flattened dead cells sealed with fats, and its job is to stop things getting in. Small, fat-soluble molecules can slip through it; large or water-loving ones mostly cannot.
In 2000, two Dutch researchers argued that a compound must weigh under 500 daltons to be absorbed through skin [1]. Their evidence was that nearly all common contact allergens, all common topical drugs and all known transdermal-patch drugs at the time fell below that size.
The 500 dalton figure is a rule of thumb built from that pattern, not a hard physical cutoff, and delivery technology has since been designed specifically to push larger molecules past it. It remains a useful first test for any patch claim.
663.4
molecular weight of NAD+, in daltons (g/mol)
PubChem
500
the size ceiling for skin absorption proposed in 2000
Bos 2000
Untested
whether a patch raises NAD+ in human blood: no published study has measured it
Where NAD+ falls
NAD+ has a molecular weight of 663.4 [6]. It also carries two phosphate groups, which make it highly water-soluble, so on both size and chemistry it is the kind of molecule the skin barrier is good at stopping.
That does not prove no NAD+ crosses. It means a patch claim needs direct evidence, and for NAD+ patches on intact human skin, none has been published.
What the delivery studies actually tested
The studies that exist show how hard researchers have to work to get NAD+ or its precursors into skin at all, and none of them is a patch worn on an arm.
A 2026 study packaged NAD+ in ion-coupled transfersomes, flexible lipid carriers, and found better NAD+ penetration into pieces of pig and human skin in the lab than free NAD+ or its precursors achieved [2]. The rest of its results come from cells and worms, and most of its authors work for a consumer-products company.
A 2024 study coated microneedles, tiny needles that pierce the outer skin layer, with NMN, an NAD+ precursor [3]. In human skin samples in the lab, about 189 micrograms passed through over 24 hours; the rest of the work was in a mouse ear. Microneedles work precisely because they break the barrier a patch has to cross.
The one study in people treated melasma, a skin pigment disorder, in 36 women [4]. Each session used microneedling first, then a sterile NAD+ product applied to the skin. Melasma scores fell from 16.8 to 6.9 over 21 weeks, but there was no control group, and the needling alone could explain part of it.
What happens even when NAD+ reaches the blood
Suppose some NAD+ did cross the skin. A pilot study that infused NAD+ directly into a vein for six hours found no change in plasma NAD+ for the first two hours [5]. The authors concluded it was rapidly and completely removed from the plasma over that period.
If an infusion delivered straight into the bloodstream barely moved blood levels at first, a patch delivering a small fraction of that through skin has a much steeper climb. That comparison is an inference, not a measurement, because no patch study exists to test it.
What this means
An NAD patch rests on a delivery claim that no published human study has tested, for a molecule larger and more water-loving than the skin normally admits. That does not show patches fail; it shows nobody has shown they work.
The other needle-free route, the nose, has its own evidence problems, set out in NAD+ nasal spray. The Breeze Meds review covers one seller that offers prescribed NAD+ as an injection or a nasal spray.
The route that skips the skin and the gut altogether is covered in NAD IV therapy. For approaches to healthy aging with human data, start at the longevity goal page.