Guider 

Red versus NIR: compare the target, not just the label

Red is part of the visible spectrum. Near-infrared is the neighboring infrared region beyond it, which human eyes do not normally see. “Infrared” is a broader term than NIR and does not identify a particular wavelength or treatment. [1]

For fine lines, the useful question is whether a defined exposure improved a wrinkle measure against a suitable control. A collagen-related measurement, an elasticity measurement, and a participant's impression are different outcomes. Do not silently exchange one for another when deciding which light is “better.”

A fair comparison asks what NIR adds to a particular red-light exposure. It also asks whether that difference matters to the reader's concern. A device using both bands may have supporting evidence for the complete intervention without evidence that either band is indispensable.

Why “deeper” does not mean “better”

Light entering skin is absorbed and scattered rather than traveling straight to a sharply defined stopping point. Tissue optics depend on wavelength and on components such as melanin, blood, and water. Scattering also changes with wavelength and tissue structure. These relationships help explain why parts of the red-to-NIR region are useful for delivering light into tissue. They do not create a universal millimeter-depth chart. [2]

Longer wavelength is not a rule that depth increases indefinitely: absorption spectra have structure, including water absorption features. A meaningful depth estimate must identify the tissue, optical assumptions, and what “penetration” means. A detectable amount of transmitted light and enough light at a biological target to produce a measured response are different questions. [2]

The missing step in a depth-based claim

“This light reaches deeper, therefore it reduces wrinkles more” skips the clinical comparison. To support that conclusion, a study needs to measure the relevant skin outcome under defined conditions. A penetration model can help explain delivery; it cannot report a result that participants were never shown to achieve.

This is also why a cell experiment is not a facial result. Photobiomodulation (PBM) concerns light-induced changes in biological activity, but a laboratory signal does not establish visible improvement from a particular mask. Human evidence and the actual exposure remain necessary. [3] [4]

Read the study design before declaring a winner

A broader spectrum did not automatically outperform a red-dominant source

Wunsch and Matuschka's 2014 controlled study included 136 volunteers. It used large-area gas-discharge lamps: one source was predominantly red at 611–650 nm; the broader source covered 570–850 nm. After 30 treatments, the authors reported improvements in skin roughness, ultrasound-based collagen intensity, and photographic wrinkle assessments, with no significant difference between the two light groups. [5]

The qualification matters. This was not a narrow-band red-LED mask versus an NIR-LED mask. The untreated control group was recruited without randomization, treatment could not be blinded, and the study was industry funded. The red-dominant source was not exclusively red. Failure to detect a difference does not prove equivalence, and the findings do not establish that adding NIR to every facial device is useless. [5]

LED evidence includes several settings and mixed outcomes

A 2018 systematic review of randomized LED trials included red, NIR, and combined settings in skin-rejuvenation research. It identified limitations including small samples, missing blinding or sham controls, and varied parameters. Its skin-rejuvenation evidence received a grade C recommendation within that review's grading system. This is evidence to evaluate, rather than a general endorsement of any device sharing a wavelength. [3]

A newer combination study may answer a different question

A 2025 exploratory trial compared two yellow-light wavelengths, each combined with the same 620 nm red and 850 nm infrared exposure. The publisher's abstract describes 30 enrolled patients with facial photoaging and photographic/imaging assessments. Because both groups received red and infrared, this design does not isolate the benefit of NIR over red alone. Full methods were not accessible for this article, so it is used only to explain that comparison boundary. [6]

When reading any study, distinguish improvement from baseline, improvement versus a control, and superiority over another active exposure. Those are separate findings. A before-and-after change in a mixed-light group cannot, by itself, answer which wavelength caused it.

A practical checklist for red-versus-NIR claims

Use these questions to identify what is established and what remains untested:

1. Name the outcome. Does “better skin” mean a wrinkle score, roughness, elasticity, or a subjective rating?
2. Identify the comparison. Was NIR tested against red, added to red, or bundled with other wavelengths in both groups?
3. Match the light delivery. Check the spectrum, irradiance at the treatment site, radiant exposure, pulsing, and illuminated area—not wavelength alone. [4]
4. Match the participants and setting. Evidence for one skin concern or a supervised research setting does not automatically establish another use.
5. Separate the result from permission to use. A study protocol does not replace the current instructions for another device.

The FDA's January 2023 PBM submission document illustrates why multiple exposure characteristics matter. It remains draft guidance, not for implementation, with nonbinding recommendations for specified US medical-device submissions; some low-risk general-wellness products may be outside its scope. It is not a worldwide operating standard or clearance for a device mentioned elsewhere. [4]

Three misleading shortcuts

• “More wavelengths must be better.” Ask for the comparison that demonstrates the extra contribution.
• “The deeper band is the anti-aging band.” A depth claim needs a separate link to a measured appearance outcome.
• “The same wavelength means the same result.” A shared label leaves delivery conditions and clinical applicability unresolved.

Further questions

If NIR is invisible, how can I tell whether it is working?

Your eyes cannot verify the NIR spectrum or delivered output. Use the manufacturer's documented specifications and approved function checks. A visible indicator establishes only what the manufacturer says it indicates; it is not a measurement of the light reaching skin. Do not look into emitters to test them. [1] [4]

Does “infrared LED light therapy” mean the same thing as an infrared heat lamp?

No. The infrared label alone does not identify an intervention's intended effect. PBM research and devices designed to produce therapeutic heating must be evaluated separately. Likewise, laser, intense pulsed light (IPL), and photosensitizer-based photodynamic therapy (PDT) results cannot be transferred to ordinary cosmetic LED exposure because their names include “light.” [4]

Can I compare two studies by their percentage improvement?

Only with care. Different measures, controls, participants, and follow-up can make percentages describe different things. Ask whether the studies measured the same endpoint in a comparable way. Otherwise, a larger number does not establish a superior wavelength.

What a useful comparison should leave you knowing

Red versus NIR is a question about a defined skin target under defined exposure conditions. If the only evidence offered is a color label or a deeper arrow in a diagram, a better facial result remains unestablished. Look for the matching human comparison and the exact device instructions before turning that claim into a decision.