Woman using a mirror to catch a thin sunset beam inside a thick-walled keep, while scholars measure outside.
Back to the library

Brain

Red light, mood and cognition

Shining near-infrared light at the forehead to influence mood or thinking is called transcranial photobiomodulation, and it is a separate field from anything a body panel does. Early pilot studies reported encouraging signals for reaction time, working memory and depressive symptoms, but the best-designed trial, a multicentre randomised sham-controlled study of 830 nm LED light, found no difference from sham and its authors described it as marking a threshold of inefficacy at low doses. Cadaver measurements explain why: very little LED light reaches the cortex through skin and skull. Breathe's panels are not transcranial devices and nothing here should be read as supporting mood or cognition.

Evidence so far
Early
Reading timeAbout 12 minutes

Start here

Why this is being studied

This path covers attempts to influence mood, attention and memory by shining near-infrared light at the head, usually at the forehead. The rationale is that brain tissue is metabolically demanding and that its mitochondria contain the same light-absorbing enzyme found everywhere else, so light reaching the cortex might in principle support energy production there. The whole question turns on whether enough light actually arrives, and the measurements on that point are not encouraging.

Where the evidence stands

Flag text

The mechanism

What is going on, and why this might help

What is going on in the brain

Brain tissue uses a large share of the body's energy and is densely packed with mitochondria, so the general photobiomodulation logic applies in principle: where mitochondrial function is reduced, as is proposed in ageing, depression and some neurological conditions, supporting it might matter. Animal work reviewed by Salehpour and colleagues reports changes in brain energy metabolism and inflammatory signalling after near-infrared exposure.

What the light would be absorbed by

The same enzyme as everywhere else. Cytochrome c oxidase in neuronal and glial mitochondria absorbs near-infrared light, and the model is the familiar one: photons displace inhibitory nitric oxide, oxygen consumption and ATP output rise briefly, released nitric oxide increases local blood flow, and a low-level burst of reactive oxygen species acts as a redox signal activating transcription factors linked to repair and to resolution of inflammation. None of that is in dispute as cell biology. The question is whether it ever happens inside a living skull.

The biphasic dose response

Dose is the whole argument here. The pilot that reported cognitive improvements used a 1,064 nm laser at 250 milliwatts per square centimetre and 60 joules per square centimetre per site. The multicentre trial that found nothing used an 830 nm LED at 55 milliwatts per square centimetre. If a threshold exists, the LED trial sat below it, and its authors framed their own result in exactly those terms. Above the window, the curve falls away again, so this is not an argument that more is always better either.

How far the light actually goes

This is the decisive point. Cadaver measurements show 830 nm transmitting better than 633 nm but both being heavily attenuated by scalp and skull, with wide regional variation, and modelling work by Henderson and Morries argues that meaningful fluence at brain depth requires far more surface power than typical LED devices supply. A whole-body panel at 15 to 30 centimetres, with goggles on and light spread across the body, delivers nothing comparable to a device pressed against the forehead. Breathe's panels are not transcranial devices and should not be described as supporting mood or cognition.

Explainer illustration

The studies

Every study is laid out the same way: what the researchers wanted to find out, who took part, what they actually did, what they found, why it might work and where the limits are. No jargon, and a link to the original if you want to read it yourself.

Each study is laid out the same way: what the researchers asked, who took part, what they did, what they found, why it might work, and where the limits are. The clinician layer at the end of each one carries the numbers.

Paper 1

Very low-level transcranial photobiomodulation for major depressive disorder: the ELATED-3 multicenter, randomized, sham-controlled trial

Iosifescu et al. · Journal of Clinical Psychiatry · 2022 · Multicentre randomised sham-controlled trial

Read the paper

What they asked

An earlier pilot study by the same group had reported a signal for depression using near-infrared light at the forehead, and the field needed a proper test. This multicentre trial asked whether twice-weekly sessions of 830 nm LED light over six weeks would reduce depressive symptoms more than an identical sham device in adults with major depressive disorder.

Who took part

Adults with a diagnosis of major depressive disorder were recruited across multiple centres and randomly allocated to active or sham devices. Using several sites rather than one reduces the chance that a result reflects the enthusiasm of a single team, which is one of the main criticisms of earlier positive studies in this area.

What they did

Participants received 830 nm LED light at about 55 milliwatts per square centimetre for a dose of roughly 66 joules per square centimetre per session, twice a week for six weeks, or an identical-looking sham. Neither participants nor the researchers assessing them knew which was which. Depression was measured with standard clinician-rated and self-rated scales before, during and after the course.

What they found

There was no difference between active and sham on either depression scale. Both groups improved to a similar degree, which is typical in depression trials. The authors concluded that the trial established a threshold of inefficacy for transcranial photobiomodulation at these low doses, meaning that whatever the approach might do at higher doses, it does not work at this one.

What it means

A well-conducted negative trial is worth more than several positive pilots, and this one should settle the question of whether LED light at these settings affects depression. It does not. For members the relevant point is even simpler: if a device aimed directly at the forehead at 66 joules per square centimetre does nothing measurable, a body panel several feet away is not going to.

Limits

A negative trial tests one dose, one wavelength and one schedule. Advocates of higher-power laser devices argue, with some physical justification, that this trial used too little light to reach cortex at all, which is a fair criticism of the dose rather than a defence of the field. The trial studied depression, not cognition or general wellbeing. It also cannot rule out effects in subgroups it was not designed to examine.

Paper 2

Beneficial neurocognitive effects of transcranial laser in older adults

Vargas et al. · Lasers in Medical Science · 2017 · Single-arm pilot study without a sham group, n=12

Read the paper

What they asked

Thinking speed and working memory tend to decline with age, and one theory links that decline to falling mitochondrial efficiency in the frontal cortex. This pilot asked whether weekly sessions of higher-powered near-infrared laser light to the forehead would improve measures of attention, reaction time and memory in older adults.

Who took part

Twelve older adults took part. There was no sham group, so everyone knew they were receiving the light and every participant's later scores were compared with their own earlier ones. Twelve people is a pilot in the truest sense: enough to see whether a study can be run, not enough to learn whether something works.

What they did

Participants received 1,064 nm laser light at about 250 milliwatts per square centimetre, delivering roughly 60 joules per square centimetre per site, once a week for five weeks. Cognitive testing covered reaction time, sustained attention and working memory, administered before and after the course.

What they found

The study reported faster reaction times and improved working memory scores after the five sessions compared with the participants' own baseline measurements.

What it means

This is the study most often cited when transcranial light is described as promising, and its value lies mostly in the dose. It used a 1,064 nm laser at far higher power than the LED devices that later failed, which is consistent with the physics argument that enough light has to arrive for anything to happen. As evidence of benefit it carries very little weight.

Limits

Twelve people, no sham group and no blinding. Cognitive tests improve on repetition simply because people get better at taking them, and without a comparison group there is no way to separate that practice effect from any effect of the light. The study was short, the participants were few, and its results have not been confirmed in a controlled trial. It is a reason to run a proper study, not a result.

Paper 3

Transcranial red and near infrared light transmission in a cadaveric model

Jagdeo et al. · PLoS One · 2012 · Laboratory transmission measurements in cadaveric human tissue

Read the paper

What they asked

Before asking whether light at the head changes anything, it is worth asking how much light gets in. This study asked a purely physical question: when red or near-infrared light is shone at the outside of a human head, how much of it actually reaches the other side of the scalp, skull and brain tissue, and does it vary by wavelength and by region of the head.

Who took part

There were no living participants. The measurements were made on cadaveric human tissue, which allows the layers of scalp, skull and brain to be measured directly in a way that is impossible in a living person. Tissue from more than one region of the head was tested so that regional differences could be seen.

What they did

The team shone light at 633 nm and 830 nm through samples of scalp, skull and brain tissue of known thickness, and measured how much emerged on the far side with a detector. By comparing wavelengths and regions they could quantify attenuation at each layer.

What they found

Light at 830 nm penetrated measurably better than light at 633 nm, which fits the known optics of tissue. Both wavelengths, however, were heavily attenuated, with the skull absorbing and scattering a large share, and transmission varied considerably from one region of the head to another.

What it means

This is the study that keeps the rest of the field honest. It explains why an LED device at the forehead can produce a null result while a high-powered laser produces a signal, and it supports the argument that surface power has to be far higher than most devices provide before meaningful light reaches cortex. It also settles a broader myth: if the skull stops most of it, so does the rest of the body.

Limits

Cadaveric tissue is not living tissue. Blood flow, oxygenation and the optical properties of living cells all differ, and measurements made on samples cannot be assumed to hold exactly in a living head. Thickness of scalp, skull and hair varies considerably between individuals. The study measures transmission, not biological effect, so it constrains what is plausible rather than proving what happens.