Red Light Therapy

Deep-dive

Wavelengths and what they do. "Red light therapy" is shorthand for two bands. Red (roughly 630-700 nm) penetrates a few millimetres and is the main driver of skin and superficial effects. Near-infrared (NIR, roughly 780-900 nm, with 810 and 850 nm the most studied) penetrates deeper, 1-3+ cm, and is what reaches muscle, joints, brain, thyroid, and gonads. Most commercial panels combine 660 nm and 850 nm for this reason. Wavelengths outside this "optical window" (UV, blue, green, yellow, far-IR) are either filtered by water and haemoglobin or behave differently and aren't what people mean by red light therapy.

The primary mechanism: cytochrome c oxidase. The dominant model, established in Karu's foundational work and reviewed extensively by Hamblin, is that red and NIR photons are absorbed by cytochrome c oxidase (CCO), complex IV of the mitochondrial electron transport chain. CCO contains two heme groups and two copper centres which absorb in the 600-900 nm range. Under stress, hypoxia, or inflammation, nitric oxide (NO) binds CCO and chokes ATP production. Red/NIR light dislodges the NO, the electron transport chain speeds back up, and you get more ATP, a brief controlled burst of reactive oxygen species (which acts as a signalling molecule), and increased mitochondrial membrane potential. The freed NO becomes a useful signalling molecule on its own, driving local vasodilation and improved blood flow.

This sequence triggers downstream activation of transcription factors (NF-κB, Nrf2, AP-1), shifts in gene expression toward anti-inflammatory and pro-regenerative pathways, and the release of growth factors. The same machinery that responds to

GHK-Cu signalling effects and that

Creatine supports through phosphate buffering is being modulated here through light. The CCO model isn't the whole story, more recent work implicates light-gated TRP ion channels, opsins in skin and other "non-visual" tissues, and structured water layers near membranes, but CCO remains the most consistent point of action.

The biphasic dose-response is the most important practical concept. Huang, Carroll, and Hamblin's 2009 review and the 2011 update lay this out clearly: too little light does nothing, the right dose stimulates, and too much light suppresses or even harms. The curve isn't linear, more is not better. Most home users assume that longer sessions or closer-to-skin must be better; the data say the opposite past a certain threshold. The sweet spot for skin sits around 3-10 J/cm² per session. For deeper tissue (muscle, joint, thyroid, brain), 10-60 J/cm² at the target depth. Above roughly 100 J/cm² for most applications, you start losing benefit.

Skin: collagen, wrinkles, and tone. This is the most replicated finding. In a 2014 randomized controlled trial of 136 volunteers, 30 sessions of red light (611-650 nm) or polychromatic red plus NIR (570-850 nm) twice weekly significantly improved skin complexion, roughness, ultrasonographically measured intradermal collagen density, and clinical photograph ratings compared with controls. Both wavelength ranges worked similarly. A 2023 split-face trial in 137 women showed 660 nm light at 3.8 J/cm² reduced periocular wrinkle volume by around 30% over 10 sessions across 4 weeks. The mechanism is fibroblast activation, MMP regulation (which clears old damaged collagen), and increased synthesis of new collagen and elastin. The effects are real but gradual: typically 4-12 weeks of consistent use.

Hair: androgenetic alopecia. Multiple FDA-cleared devices exist for pattern hair loss. The Lanzafame 2013 multicenter RCT found a 35% increase in hair counts after 26 weeks of LLLT at 655 nm in men with AGA versus sham. A 2014 follow-up trial in women found a 51% increase in hair counts. A 2019 meta-analysis confirmed efficacy for AGA in both sexes, comparable in magnitude to

Minoxidil (oral & topical) . Mechanism is thought to involve direct stimulation of dermal papilla cells, prolonging the anagen (growth) phase, and increasing local microcirculation. It works best in early-to-moderate hair loss, not in totally bald patches where follicles have miniaturised beyond recovery.

Muscle recovery and performance. A 2018 systematic review and meta-analysis of 39 trials by Vanin and colleagues found that photobiomodulation applied before exercise consistently reduced muscle fatigue, improved performance metrics (repetitions, peak torque, time-to-exhaustion), and lowered post-exercise creatine kinase and lactate. A 2025 meta-analysis reached similar conclusions for delayed onset muscle soreness. Effect sizes are modest and dosing-sensitive, but the direction is consistent. The pre-exercise window matters: applied 30-60 min beforehand tends to work better than post-exercise application.

Thyroid and Hashimoto's. This is one of the more interesting under-recognised applications. The Höfling 2013 RCT treated 43 patients with hypothyroidism from chronic autoimmune thyroiditis with 10 sessions of 830 nm LLLT or placebo. At 9 months follow-up, the laser group needed 64% less levothyroxine to maintain euthyroidism, had lower TPO antibodies, and showed improved thyroid echogenicity on ultrasound. A 6-year follow-up found the improvements persisted without adverse effects. A 2020 Hashimoto's cohort and a 2023 trial of PBM combined with supplements reproduced the pattern. The mechanism is thought to be reduced autoimmune attack on the gland plus improved follicular cell function. Worth knowing if you have Hashimoto's, the data are limited but consistent and the downside risk is minimal.

Brain: cognition and mood. Transcranial photobiomodulation (light applied to the forehead/scalp) penetrates a few mm into the cortex. The Cassano group at Harvard/Mass General have run multiple sham-controlled trials in major depression. The ELATED-2 pilot showed antidepressant effects at the right dose. ELATED-3, a larger multicenter trial, was negative at a much lower dose, illustrating the biphasic issue head-on. A 2024 RCT in adults over 50 found transcranial PBM improved cognitive scores and raised serum BDNF compared with placebo. The cognitive and mood applications are real but more dose-sensitive than skin or hair, and home devices marketed for this purpose vary wildly in actually-delivered fluence at the brain.

Testosterone, sperm, and fertility (men). The popular claim that shining red light on the testicles raises testosterone is heavily over-marketed. The serious data are mostly animal studies showing improvements in testicular function after chemical or radiation injury, plus in vitro work on human sperm. A 2025 in vitro study on asthenospermic human sperm found 810 nm light significantly improved sperm motility and energy production, while 635 nm reduced it. Translation: wavelength matters, NIR good for sperm, red alone may not be. There are no good human RCTs showing red light on the testes raises serum testosterone in healthy men. The mechanism (Leydig cells are mitochondria-rich, PBM should plausibly help) is sound but the human data are absent. Treat the testosterone claim as plausible, unproven, and not worth chasing if your levels are normal.

Women. Most of the skin, hair, muscle, thyroid, and brain research has included women throughout, and the effects look similar between sexes. The Höfling thyroid trial www.notion.so was conducted predominantly in women (Hashimoto's is around 8-10x more common in women). The Lanzafame female AGA trial showed a 51% increase in hair counts. For fertility, a 2024 case series on transdermal PBM for unexplained female infertility reported encouraging live birth outcomes, but this is a case series, not a controlled trial. Animal and in vitro work suggest PBM improves oocyte mitochondrial function and ovarian blood flow, which is biologically plausible. The strongest practical use cases for women are skin, hair, Hashimoto's (which disproportionately affects women), and possibly perimenopausal/postmenopausal skin and bone applications. No sex-specific dose adjustments are needed. Avoid the lower abdomen and uterine area during pregnancy, not because of documented harm but because there's no safety data and the prudent default is to skip.

Limitations of the evidence. Despite hundreds of trials, the field is plagued by inconsistent dosing, wavelength reporting, and outcome measures, which is exactly why some studies are negative. A negative trial at the wrong dose doesn't mean the modality doesn't work, but it also means you can't just trust a marketing claim that "studies show." The strongest signals are in skin, hair, muscle recovery, and Hashimoto's. The weakest are in serum testosterone, weight loss, and most "systemic" claims about whole-body panels. Whole-body panels are convenient but a 2025 systematic review found the evidence base for whole-body PBM (vs targeted) is much thinner than the marketing implies.

Dosage:

• Wavelengths: Red (630-670 nm, with 660 nm the most studied) for skin, hair, superficial tissue. Near-infrared (810-850 nm) for muscle, joint, thyroid, brain, deeper tissue. A combined 660+850 nm panel covers most use cases

• Dose at target tissue: 3-10 J/cm² for skin, 10-60 J/cm² for deeper tissue. More is not better, the response is biphasic. A typical panel at 6 inches delivering 100 mW/cm² hits 6 J/cm² in 60 seconds, so most skin sessions are 1-3 minutes per area, not 20

• Distance: 6-12 inches from the body for most panels. Closer is not better, the intensity rises sharply but you risk overdosing and heating tissue. Check your specific device's irradiance curve

• Session length: 5-20 minutes total covering the areas you care about, 3-5 times per week. Sessions over 20 minutes are rarely doing more than the first 10. Daily is fine if the per-session dose is moderate; if you're running high power close to the skin, every other day is enough

• Skin and anti-aging: Face or specific area at 6-12 inches, 1-3 min per area, 3-5x/week. Expect changes at 6-12 weeks.

• Hair (AGA): Scalp coverage, 10-15 min, 3x/week. Consistency matters more than session length. Most clinical trials ran 24-26 weeks before evaluating

• Muscle recovery/performance: Apply 30-60 min before training or immediately after, 3-10 min per major muscle group at 6-8 inches. NIR (810-850 nm) penetrates deeper, use it for muscle, not red alone

• Thyroid (Hashimoto's): Front of the neck over the thyroid, 5-10 min, 2-3x/week. The Höfling protocol used 830 nm, so NIR is the wavelength of interest here. Don't expect to drop levothyroxine without monitoring TSH/T4 alongside, this is a "do it and watch your labs" application

• Transcranial (cognition/mood): This is where dosing gets touchy and home devices are inconsistent. 810 nm at 100-200 mW/cm² for 8-20 min to the forehead in published trials. If you're doing this for depression rather than for fun, do it with a clinician who's tracking it

• Eye protection: Wear blackout goggles for face/scalp work, especially at higher power. Red light isn't directly damaging the retina the way UV would, but bright NIR at close range over time isn't well studied for ocular safety. Cheap goggles included with most panels are sufficient. Just close your eyes and look away isn't enough at close range

• Timing of day: Morning or daytime is generally preferred. Some people get mild alertness afterward which can disrupt sleep if done late

Here's what you can expect:

Side effects & risks:

• Skin: redness, mild heat, dryness during or briefly after sessions, especially at high irradiance or close range. Usually settles within an hour. If it doesn't, you're running too hot or too long

• Eye irritation, blurry vision, headache if you're not wearing eye protection during facial sessions, especially with NIR. Cumulative ocular exposure over years isn't well studied, so use goggles. People with macular degeneration or retinal disease should consult an ophthalmologist before transcranial use

• Worsening of melasma and pigmentation conditions in some cases. Red light is generally considered safer than visible blue or UV for pigmentation, but if you have active melasma, test on a small area first and watch for darkening

• Headache, jitteriness, sleep disruption with transcranial use, especially at high dose or late in the day. Drop the dose, move the session earlier

• Photosensitising medications. If you're on tetracycline antibiotics, isotretinoin, St John's Wort, certain diuretics, or amiodarone, your skin response to red light may be amplified. Patch test first

• Active skin cancer or undiagnosed pigmented lesions in the treatment field, avoid until evaluated. Red light isn't thought to be carcinogenic (no UV/DNA damage component), but stimulating cell proliferation on top of an undiagnosed lesion is not what you want

• Pregnancy. Skin and limb use is generally considered fine. Skip lower abdomen, uterus, and breast/chest applications during pregnancy, not because of documented harm but because there's no controlled safety data

• Thyroid hyperfunction. If you have Graves' disease or hyperthyroidism, the same mechanism that helps Hashimoto's by improving follicular function could plausibly make hyperthyroidism worse. Avoid direct neck/thyroid application until your thyroid is stabilised

• Pacemakers, implants, tattoos. Red and NIR light don't interact with pacemakers or metal implants. Dark tattoos absorb light strongly and can heat up; expect warmth, dial back distance and time if uncomfortable

• Whole-body panels and tanning-bed-style cabins deliver less dose per cm² than people assume because of distance, but at extended sessions (20-30+ min daily) you're in territory where cumulative biphasic suppression becomes a real possibility. Less is more

• Overuse syndrome. Sessions too long, too close, too frequent can produce the opposite of the desired effect: worsened recovery, paradoxical fatigue, skin that looks more inflamed rather than less. If results have plateaued or reversed, cut the dose in half before adding any more