Transcranial Photobiomodulation: The Delivery Problem
If you have spent any time reading about transcranial photobiomodulation, you have probably noticed something odd. The marketing is confident and the physics is missing. Devices promise sharper focus and steadier energy, yet almost none of them tell you how much light actually arrives at the brain after it has traveled through hair, scalp, and bone. That unanswered question is why this topic feels slippery, and it is where this guide begins.
The Question Every Vendor Skips: Does the Light Actually Reach the Brain?
It is the right question to ask first. Photobiomodulation is a light-based approach, and light that never reaches its target cannot do anything at that target.
The best available numbers come from a cadaveric study on cranial transmission, published as PMC10884051. Across the wavelengths tested, transmission through skull bone ranged from 0.01 percent at 405 nm to 0.95 percent at 655 nm.
Read that range again. The ceiling for the entire experiment was under one percent.
Between 99 and 99.99 percent of the light applied to the outside of a skull did not pass through it in the cadaveric model.
The study authors are upfront about the limits of their own work. Their samples "were exclusively derived from cadaveric skull bones," and they note that "the optical and biological properties of living tissues can differ considerably."
A peer-reviewed critical review in Frontiers in Neurology (2024, article 1398894) pushes harder on the same point. It reports that infrared light from a 0.5-watt LED "will not penetrate the scalp and skull," that it "would penetrate no further than 2 mm through skull," and that "longer exposure times do not yield deeper penetration."
The consumer-facing version of that number, from the neurosity guide, is that roughly 1 to 5 percent of applied light reaches the cortical surface. A separate 2024 review, PMC10840571, puts it plainly: "Only a small number of photons can effectively reach brain tissues."
So the honest answer to "does the light get there?" is this: some of it, in small amounts, and how much depends on variables most product pages never discuss.
This is the delivery problem, and it is the real subject of every honest conversation about transcranial near-infrared.
What Transcranial Photobiomodulation Is Trying to Do
The premise is straightforward. Red and near-infrared light can be absorbed by cytochrome c oxidase, an enzyme in the mitochondrial electron transport chain. That absorption is proposed to support mitochondrial energy production.
The mechanism is not specific to the brain. It is the same foundation behind every other application of the technology, and the general mechanism behind red light therapy is worth reading once rather than restating here.
The brain-specific version adds one more step. Near-infrared light therapy for the brain has to reach tissue that sits inside a skull, which changes the engineering problem completely.
Researchers have studied a wide range of parameters. A 2024 review (PMC10840571) cataloged wavelengths from 600 to 1300 nm, irradiance from 0.01 to 10 W/cm², and daily energy densities including 4, 8, and 32 J/cm².
That spread is not a sign of precision. It is a sign that the field has not settled on a standard.
Transcranial photobiomodulation is also proposed to influence cerebral blood flow. A single case report cited in that same review described regional blood flow increasing by 20 percent in the left anterior frontal lobe. That is one observation, not a pattern, and the review itself flags how much remains unresolved.
For readers who think about circulation as part of a broader cognitive support routine, Agape's heart and circulation page covers the nutrient side of that same system. CoQ10 belongs in that conversation too, since it supports normal cardiovascular function and works inside the same mitochondrial chain the light is proposed to act on, and Agape's CoQ10 100 mg is a straightforward way to cover it.
Two Delivery Routes, Two Different Targets
This is the single most under-covered idea in consumer coverage. There are two ways to deliver light toward the brain, and they do not reach the same place.
The transcranial route
Applicators sit against the scalp over the forehead, temples, or back of the head. Light must cross skin, subcutaneous tissue, the skull, and the meninges before reaching cortical tissue.
Everything measured in the cadaveric study applies here. It is the route with the most published research and the most difficult physics.
The intranasal route
A small applicator is placed inside the nostril. From there, light can reach the ventral prefrontal region without crossing the skull at all.
That bypass is the entire argument for intranasal photobiomodulation. It trades a large treated area for a shorter, less obstructed path to a smaller one.
The two routes are complements rather than competitors. A 2024 registered trial protocol (Frontiers in Neurology, 10.3389/fneur.2024.1371284) uses both at once: five applicators total, with one intranasal, one frontal, and three occipital.
That protocol is a useful reference point because it is unusually specific. It specifies near-infrared light from 810 to 1100 nm, 40 Hz intermittent photoemission, 60 J/cm² frontal, 45 J/cm² occipital, and 15 J/cm² intranasal, delivered in 20-minute sessions, three times a week for 12 weeks.
Note the intranasal dose: one quarter of the frontal dose. That gap is deliberate, and it leads directly into the next section.
One context point matters here. That paper is a protocol, which means it describes how the therapy is delivered in controlled research. It is not a result, and its outcomes are not yet published.
The Wavelength Debate: Why 810 nm and 1070 nm Both Have a Case
Two leading device vendors publish content arguing that their own wavelength is the better one. Each cites real studies. Each reaches the conclusion that favors its own hardware.
Here is what the underlying evidence actually says.
The case for 810 nm
Cytochrome c oxidase absorption peaks near 810 nm, according to the manufacturer technical literature. If the mechanism runs through that enzyme, matching its absorption peak is a reasonable design goal.
There is a penetration argument too. A Harvard-led modeling study compared 670, 810, 850, 980, and 1064 nm and reported that 810 nm offered the highest light penetration. That finding deserves a flag: it is real, but it reaches most readers second-hand through a manufacturer's technical page, so treat it as reported rather than independently confirmed.
The case for longer wavelengths
Longer wavelengths scatter less and are often assumed to travel deeper. The counterargument is water.
Water absorption "increases significantly beyond roughly 950 nm," which limits how far those longer wavelengths can travel through tissue. That is why a 1070 nm device and an 810 nm device can both cite legitimate physics and still disagree about the conclusion.
It is also worth separating this from a related but different technology. Far infrared therapy works at much longer wavelengths and relies on radiant heat rather than the photon-absorption mechanism discussed here. The two get conflated constantly.
What the disagreement really tells you
When two well-resourced vendors read the same literature and land on opposite answers, the honest conclusion is that the question is not settled. Both positions appear throughout brain photobiomodulation research, and both can be defended from real physics.
A wavelength number on a box is a design choice, not a guarantee of performance. What matters more is how much usable energy a device delivers to the intended target, and whether that amount falls inside the range where the effect is positive rather than flat or negative.
The 1064 nm end of the spectrum has its own evidence. One sham-controlled study of 25 healthy adults used a 1064 nm laser at approximately 107 J/cm² delivered at 250 mW/cm² for 8 minutes over the right prefrontal cortex.
More Is Not Better: The Biphasic Dose Response
If you take one practical idea from this article, take this one.
Photobiomodulation follows a biphasic dose response. Low doses do little. A middle range produces the intended effect. High doses stop helping and can actively work against you.
The clearest illustration comes from a 2022 review (PMC9400541). In the model it describes:
- 18 J/cm² and 25 J/cm² were ineffective
- 32 J/cm² was optimal
- 39 J/cm² produced negative morphological changes, including brain swelling
- The 730 nm and 980 nm groups showed no significant improvements
A dose that is too low does nothing. A dose that is too high does harm. The useful window sits in the middle, and it is narrower than most people assume.
The same pattern shows up elsewhere. The Lin 2024 review notes that "excessive power density may also lead to the plateau effect," and that an excessive number of treatments in an animal model could temporarily inhibit recovery rather than accelerate it.
This is why "more power" and "more diodes" are poor buying criteria on their own. A device that delivers a larger dose is not automatically a better device. It may be pushing past the window where the effect occurs at all.
What the Human Evidence Actually Shows
This is the section the affiliate pages leave out, and it is the reason the rest of this article is worth trusting.
The evidence in humans is genuinely mixed. Researchers have studied several populations, and the results have not been uniformly positive.
- A 2024 review reported that NEST-3, a large stroke trial, found no measurable neuroprotective effect from transcranial photobiomodulation.
- The same review reported that patients in a multiple sclerosis cohort showed no statistical significance on the modified fatigue impact scale.
- A separate review (PMC9400541) noted that one Parkinson's trial "demonstrated no significant changes in performed measures."
- An Alzheimer's cohort improved over 12 weeks, but the review records that one month later, their scores declined.
The reviews are candid about why. Lin 2024 states plainly that there is "limited information about the optimal tPBM protocols," and that it remains "unclear whether the tPBM effects are nonspecific or whether they depend on specific characteristics."
The critical review in Frontiers in Neurology reaches a similar verdict from the opposite direction. It describes the supporting proof as "thin with marginal benefits which are largely transient."
Where the signal is real, and what it does not prove
There is positive data too, and it deserves the same careful reading.
A 2025 sham-controlled study in Frontiers in Human Neuroscience (10.3389/fnhum.2025.1704482) measured brainwave activity in 25 healthy adults after a single 1064 nm session. The active condition produced "sustained, cyclic activation patterns reflecting coherent network engagement," while the sham condition showed "diffuse, short-lived activation patterns" that the authors described as "suggestive of a transient placebo-driven effect."
That is a real electrophysiology finding. It is also exactly what it claims to be and nothing more: a measurement of brainwave dynamics in healthy adults, not a measurement of memory, focus, or symptoms.
A device can produce a measurable change in brain activity without producing a benefit you would notice.
That distinction is the whole game in this category. It is also why the honest framing is that this technology is being studied for its support of normal brain energy metabolism and healthy cerebral blood flow, not presented as a finished answer.
For readers working on the fundamentals, Agape's brain and memory page is organized around the nutrient inputs that research links to normal cognitive function.
How to Evaluate a Photobiomodulation Device
If you decide to shop, evaluate the device the way an engineer would rather than the way a marketer wants you to. A photobiomodulation device is an optical instrument, and it should arrive with optical specifications.
Ask these six questions:
- What wavelength does it emit, and how much of its output sits in that band? A device described as "1064 nm" may allocate only a fraction of its emitters to that wavelength.
- What is the irradiance at the distance you will actually use it? Power measured at the surface of the emitter tells you nothing about power reaching your scalp.
- What total dose does a session deliver? Fluence, measured in J/cm², is the number that relates to the biphasic window.
- Where do the applicators sit? Frontal, occipital, and intranasal placements reach different regions. A single-point device and a five-applicator device are different instruments.
- How long is a session, and how often? The research protocol above used 20 minutes, three times a week, for 12 weeks. Extreme protocols are a warning sign, not a feature.
- Does the manufacturer publish its dosimetry? If a company will not state irradiance and dose in plain numbers, it is not giving you enough information to make a decision.
More diodes is not the metric. Dose at the target, inside the effective window, is the metric.
Practical Use and Safety
A few practical points matter regardless of which device you consider.
- Protect your eyes. Never look into an emitter, and follow the manufacturer's eye-protection instructions exactly. Intranasal applicators in particular should be used only as directed.
- Check your medications. Some medications increase photosensitivity. If you take one, talk to your prescriber before starting light therapy of any kind.
- Talk to your clinician first if you have an active medical condition, take prescription medication, are pregnant, or are managing anything under medical supervision. This is general guidance, not a substitute for individualized advice.
- Start conservatively. Because the dose response is biphasic, beginning with longer or more frequent sessions than a protocol specifies is not a shortcut.
- Keep expectations calibrated. The evidence supports a modest, still-being-characterized effect. Nothing in the literature supports treating a device as a replacement for medical care.
Where Nutrition Fits: Supporting Normal Brain Energy Metabolism
Light therapy and nutrition address the same underlying system from different directions. The brain runs on energy, and the nutrients that support normal brain energy metabolism are well characterized.
Two are worth understanding:
- Magnesium participates in hundreds of enzymatic reactions, including those involved in energy production. Agape's guide to magnesium and brain health covers where it fits, and a well-absorbed magnesium glycinate is one practical way to cover that input.
- Plasmalogens are phospholipids concentrated in brain tissue, and their role in brain health and cellular membrane function is an active area of research.
Neither of these replaces a device, and a device does not replace them. They are inputs to the same system, and the nutritional side has far more human data behind it.
That is the honest ordering of priorities for most readers.
What We Recommend
If the guide above has narrowed your options, these are the products in Agape's catalog that map most directly onto what the research describes. They are ordered by revenue rank within each group.
Brain Photobiomodulation Devices
Vielight's Gamma 4 is a transcranial and intranasal photobiomodulation device built around a single frequency, 40 Hz, the gamma setting used in the registered trial protocol described above. Listed at $1,799, it pairs a head module with an intranasal applicator, so both delivery routes are covered by one unit. It is the closest catalog match for a reader who wants to work with the 40 Hz parameters that appear most often in the brain photobiomodulation literature.
The Alpha 4 uses the same transcranial plus intranasal hardware as the Gamma model, set to pulse at 10 Hz instead of 40 Hz. The 10 Hz rate corresponds to the alpha band, the slower of the two rhythms the research protocols specify. It is the unit to weigh directly against the Gamma 4 if you are comparing the two frequencies, and the Duo below combines both.
The Duo 4 combines both settings in one device, running the 40 Hz and 10 Hz modes, plus the intranasal applicator. Listed at $2,399, it removes the need to guess which frequency suits you before you buy. It is the most complete single unit in the catalog for the two-route, two-frequency approach the research protocols use.
The MIP 470-633-655-810 is an intranasal device that ships with a set of interchangeable applicators covering 470, 633, 655, and 810 nm. Listed at $499, it is the most direct way to add the intranasal route described above without buying a transcranial head unit. Vielight describes the 810 nm channel as its brain stimulation wavelength, and the controller runs at both 10 Hz and 40 Hz.
Nutrition Support for Normal Brain Energy Metabolism
Prodrome Sciences, Glia Plasmalogens Protector 60 softgels
This is the plasmalogen product matching the nutrient discussed above. It delivers plasmalogen molecules in a daily softgel, which matters because plasmalogens are broken down in the gut and are difficult to raise through diet alone. Listed at $99, it suits a reader who wants the specific nutrient rather than a general brain formula.
Integrative Therapeutics, Active B-Complex 60 Capsules
A balanced B complex supplying B1, B2, B3, B5, B6, B7, and B12 alongside L-methyl folate and choline bitartrate. B vitamins support normal energy metabolism, which is the same system the light is proposed to act on. Listed at $18.25, it is the lowest cost addition to this list and the simplest way to cover the nutrient side of this article.
References
- Lin et al. (2024). Transcranial photobiomodulation for brain diseases. PMC10840571. https://pmc.ncbi.nlm.nih.gov/articles/PMC10840571/
- Dosimetry in cranial photobiomodulation therapy: effect of cranial thickness and bone density. PMC10884051. https://pmc.ncbi.nlm.nih.gov/articles/PMC10884051/
- Nizamutdinov et al. (2022). Transcranial near-infrared light in treatment of neurodegenerative diseases. Frontiers in Pharmacology, 13:965788. PMC9400541. https://pmc.ncbi.nlm.nih.gov/articles/PMC9400541/
- Frontiers in Human Neuroscience (2025). 1064-nm laser tPBM modulates frequency-specific cortical source dynamics. Article 1704482. https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2025.1704482/full
- Yokoi et al. (2024). A randomized sham-controlled trial of transcranial and intranasal photobiomodulation. Frontiers in Neurology, 10.3389/fneur.2024.1371284. https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2024.1371284/full
- Frontiers in Neurology (2024). Critical review of infrared light for brain applications. Article 1398894. https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2024.1398894/full
Frequently Asked Questions
Can light actually get through the skull?
Some of it, in small amounts. A cadaveric study found transmission through skull bone ranging from 0.01 percent at 405 nm to 0.95 percent at 655 nm, with the authors noting that living tissue may behave differently. A 2024 critical review concluded that infrared light from a 0.5-watt LED will not penetrate the scalp and skull, and that light would travel no further than 2 mm through skull. Consumer-facing sources cite roughly 1 to 5 percent reaching the cortical surface.
What is the difference between 810 nm and 1070 nm?
They represent two different bets. Cytochrome c oxidase absorption peaks near 810 nm, which supports the shorter wavelength. Longer wavelengths scatter less, but water absorption rises significantly beyond roughly 950 nm, which limits how far they travel. Two vendors publishing opposite conclusions from real literature is the clearest sign that this question is unresolved.
Is transcranial or intranasal photobiomodulation better?
They reach different targets rather than competing directly. Transcranial delivery covers a wider cortical area but must cross scalp, skull, and meninges. Intranasal delivery bypasses the skull to reach the ventral prefrontal region through a shorter path. Research protocols often combine them, using a lower intranasal dose than the frontal dose.
How long should a session be, and how often?
There is no standard. One registered trial protocol specifies 20 minutes per session, three times a week, for 12 weeks, with 60 J/cm² frontal, 45 J/cm² occipital, and 15 J/cm² intranasal. A 2024 review states plainly that there is limited information about optimal protocols, so following a device's own instructions is the practical baseline.
Is transcranial photobiomodulation safe, and who should avoid it?
Eye protection is standard practice and should never be skipped. Anyone taking a photosensitizing medication, managing a medical condition, pregnant, or under medical supervision should speak with a clinician before starting light therapy. No device in this category should be described as proven safe and effective for any medical purpose.
Does transcranial photobiomodulation work, or is it placebo?
The evidence is mixed, and both halves of that answer matter. NEST-3 found no measurable neuroprotective effect in a large stroke trial, and cohorts in multiple sclerosis and Parkinson's research showed no significant changes on their measures.
There is positive data as well. A 25-adult study found that active treatment produced coherent network engagement that sham did not. A 2024 critical review still called the proof "thin with marginal benefits which are largely transient."
The honest position is that the mechanism is plausible, the delivery is difficult, and the human evidence does not yet support strong claims.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.