Safety of light emitting diode-red light on human skin: two randomized controlled trials. Jagdeo J, et al. Journal of Biophotonics, 2020.
Why safety trials matter more than they sound
Most light therapy research asks whether light does something. These two trials asked the other essential question, at what dose does red light start doing harm, and answered it with the pharmaceutical playbook: phase I dose escalation, randomized, controlled, in healthy volunteers. For a field whose consumer devices ship with almost no dose guidance beyond a timer, this is foundational data.
Design
- Two trials, 115 subjects. STARS 1 (n=60) enrolled all Fitzpatrick skin types, deliberately including skin of color; STARS 2 (n=55) enrolled non-Hispanic Caucasian subjects.
- Exposure: high-fluence red LED light to the forearm, three times weekly for three weeks, escalating across groups from 160 up to 640 J/cm² per session, for scale, a typical consumer facial session delivers on the order of tens of J/cm².
- Control: mock irradiation, with blinded assessment of skin responses.
Findings
Red LED light was safe up to 320 J/cm² in skin of color and up to 480 J/cm² in the lighter-skinned cohort. Dose-limiting effects, blistering, prolonged erythema, appeared only at the top of the escalation: at 480 J/cm² in the all-skin-types trial and 640 J/cm² in the Caucasian cohort. The differential is attributed to melanin: darker skin absorbs more red light in the epidermis, so the same incident dose deposits more energy there.
How to read the numbers
- The safe thresholds sit an order of magnitude above consumer session doses, the practical message is wide margins, not danger.
- The skin-type differential is real but appears at extreme fluences; it argues for representation in trials and dose awareness in protocols, not for avoidance.
- Forearm skin was the test site; facial skin is thinner, one reason conservative dosing near the face remains sensible.
Limits worth stating
These are safety trials, they measured harm thresholds, not benefits, and say nothing about effectiveness at any dose. Sample sizes were phase-I sized. And a single anatomical site was tested. Within those limits, this is among the cleanest data in the consumer-relevant literature, and the rare study whose finding is directly usable: the ceiling is high, and it is not infinite.
Source: Journal of Biophotonics, 2020. DOI: 10.1002/jbio.201960014