A patch that makes vitamin D in skin worked in mice — but not in people

A skin patch designed to make vitamin D the way sunlight does — but without the ultraviolet rays that damage DNA — produced vitamin D in mice and in lab-grown human skin, then failed to raise vitamin D levels when tested in people [s1]. The honest headline from the British Journal of Dermatology trial is a negative one: the patch was safe but did not work, and its developers say so plainly [s1]. It is a clean illustration of why a clever idea that succeeds in the laboratory still has to clear a human trial before anyone should believe in it.

Vitamin D is made in the skin when UVB light converts a cholesterol precursor into the vitamin, which is why sun exposure raises levels and why deficiency is common in winter and at high latitudes. UV exposure, though, also causes sunburn, DNA damage and skin-cancer risk, so a way to trigger the same vitamin production without UV would be genuinely useful. The TransVitD patch delivers cholecalciferol phosphate into the skin, where the researchers showed it converts to cholecalciferol — vitamin D3 — locally [s1].

The promising groundwork

The preclinical results were the reason to run a trial at all. In animals and in ex vivo human skin, the compound penetrated the tissue and turned into vitamin D3, and in rodents it mimicked the biological response to UV exposure — raising the vitamin D-binding protein and improving whole-body vitamin D status — without the skin inflammation or DNA damage that UV light inflicts [s1]. On paper, that is exactly the profile you would want: the benefit of sun-made vitamin D without the harm.

What the trial found

The human study came in two parts, both at a single London centre and dosed over the winter months, when natural vitamin D synthesis is lowest [s1]. A dose-finding first part enrolled 16 participants across two escalating doses; it flagged no safety problem — blood calcium and vitamin D levels did not shift dangerously — but two people reported local skin irritation [s1].

The main part was a randomised, double-blind, placebo-controlled trial in 108 healthy volunteers aged 18 to 65 with suspected low vitamin D, randomly assigned across four arms [s1][s2]. Two of the active arms were stopped after four weeks, leaving a comparison of one active patch (a total dose of 140 mg) against placebo over eight weeks [s1]. The outcome that mattered was the change in serum 25-hydroxyvitamin D3, the standard blood measure of vitamin D status.

It did not move. In the per-protocol analysis, the active patch raised 25(OH)D3 by 19.9 percentage points more than placebo, but the result was not statistically significant (95% CI −1.3 to +41.0; p=0.065) [s1]. In the more rigorous modified intention-to-treat analysis, the difference was a negligible 0.52 ng/mL (95% CI −2.18 to +3.22; p=0.71) [s1]. The authors conclude that any benefit was no greater than roughly 3 ng/mL — the sort of rise you would get from 300 to 500 international units of oral vitamin D a day, which is a modest supplement dose available far more cheaply [s1]. Skin irritation was common but no more so on the active patch than on placebo: 42.9% versus 42.3% (p=1.00) [s1].

Why the gap matters

The split between the preclinical success and the human null is the whole story, and it is a familiar one. Mice are not people; lab-grown skin is not a living body with its own circulation, binding proteins and clearance; and an effect large enough to measure in a controlled tissue sample can vanish once a compound has to cross real human skin in useful amounts. The per-protocol hint of a rise is the kind of borderline, non-significant signal that selective reporting can inflate into a "promising" claim — which is why the intention-to-treat result, showing essentially nothing, is the one to trust.

What it means

For now there is no UV-free patch that reliably raises vitamin D in people, and anyone who is deficient has the same well-established options they did before: sensible sun exposure and inexpensive oral supplements, which work. The trial's value is not a product but a caution — against the steady stream of novelty vitamin D delivery gadgets and against reading animal or laboratory results as if they were clinical proof. A negative trial, reported honestly, is doing exactly the job trials exist to do.

Sources

• [s1] A UV-Free Strategy for Vitamin D Synthesis in Skin: Concept Verification and a Randomised Double-Blind Placebo-Controlled Clinical Trial of a Cholecalciferol Phosphate Patch (TransVitD). British Journal of Dermatology, 1 Oct 2026. https://doi.org/10.1093/bjd/ljag421
• [s2] TransVitD: A Study of a Cholecalciferol Phosphate Patch, ClinicalTrials.gov NCT06098846. https://clinicaltrials.gov/study/NCT06098846