How Red Light Therapy Works: Mitochondria & ATP

Red light therapy works. That much is clear from thousands of studies. But understanding why it works — at the molecular level — is what separates informed use from guesswork. The mechanism is grounded in mitochondrial biology, and once you understand it, the seemingly disparate benefits of photobiomodulation (PBM) begin to make complete sense.

This page explains the science in detail. Every claim is referenced to the primary literature.

Mitochondria: Your Cells’ Power Stations

Every cell in your body (with the exception of mature red blood cells) contains mitochondria — organelles that generate the adenosine triphosphate (ATP) your cells need to function. A typical human cell contains between 1,000 and 2,500 mitochondria, and they collectively produce approximately 40 kg of ATP per day.

Mitochondria generate ATP through a process called oxidative phosphorylation, which takes place along the electron transport chain (ETC) — a series of four protein complexes (I through IV) embedded in the inner mitochondrial membrane.

Here is the simplified sequence:

1. Complex I (NADH dehydrogenase) accepts electrons from NADH, a product of earlier metabolic steps (glycolysis and the citric acid cycle). It pumps protons (H⁺ ions) across the inner membrane into the intermembrane space.

2. Complex II (succinate dehydrogenase) provides a secondary electron entry point from FADH₂.

3. Complex III (cytochrome bc1) transfers electrons from ubiquinol to cytochrome c, pumping more protons across the membrane.

4. Complex IV (cytochrome c oxidase, or CCO) is the final step. It accepts electrons from cytochrome c, combines them with oxygen and protons to form water, and pumps additional protons across the membrane.

The resulting proton gradient across the inner membrane creates an electrochemical potential — essentially a stored energy charge. Protons flow back through ATP synthase (sometimes called Complex V), a molecular turbine that uses this flow to convert ADP into ATP.

This entire process depends on Complex IV — cytochrome c oxidase — functioning efficiently. And this is precisely where red light therapy acts.

Cytochrome c Oxidase: The Photoacceptor

The key to understanding PBM is a single enzyme: cytochrome c oxidase (CCO), also known as Complex IV of the mitochondrial electron transport chain.

CCO is a large transmembrane protein containing metal centres — specifically, two copper centres (CuA and CuB) and two haem groups (haem a and haem a3). These metal centres are chromophores: they absorb light at specific wavelengths.

Tiina Karu’s foundational research in the 1980s and 1990s established that CCO is the primary photoacceptor for red and near-infrared light in mammalian cells. Her action spectra experiments showed that the wavelengths producing the greatest biological response in cells matched the absorption spectrum of oxidised CCO — with peaks in the red (around 620-680nm) and near-infrared (around 760-900nm) ranges.

Key reference: Karu TI. “Multiple roles of cytochrome c oxidase in mammalian cells under action of red and IR-A radiation.” IUBMB Life. 2010;62(8):607-610. PMID: 20681024

This is not a vague association. The absorption spectrum of CCO and the action spectrum of PBM overlap precisely. When a photon of the correct wavelength hits a CCO molecule, it is absorbed by one of the metal centres, triggering a cascade of downstream effects.

Why Specific Wavelengths Work

The absorption spectrum of CCO explains why only certain wavelengths are therapeutically active:

Wavelength Range	Absorbing Component	Notes
620-640nm	Oxidised CCO (haem a3)	Moderate absorption, shallow penetration
650-680nm	Oxidised CCO (CuA, haem a)	Strong absorption peak; 660nm is the most studied
760-830nm	Reduced CCO (CuB, CuA)	Near-infrared; deeper tissue penetration
840-880nm	Reduced CCO (haem a3)	850nm is the most studied NIR wavelength

Light outside these windows — say, at 700-740nm — is poorly absorbed by CCO and produces little photobiomodulatory effect. This is why a generic red light bulb (which emits a broad spectrum peaking around 620-630nm, with significant output at non-therapeutic wavelengths) is not equivalent to a purpose-built PBM device emitting at 660nm and 850nm with narrow spectral bandwidth.

Key reference: Karu TI. “Mitochondrial signaling in mammalian cells activated by red and near-IR radiation.” Photochemistry and Photobiology. 2008;84(5):1091-1099. PMID: 18651871

The Three Primary Effects

When red or near-infrared photons are absorbed by CCO, three primary biochemical events occur. Together, these explain the broad therapeutic profile of PBM.

1. Increased ATP Production

The most direct effect is enhanced electron transport efficiency. When CCO absorbs a photon, it catalyses the reduction of oxygen to water more rapidly, accelerating proton pumping and ultimately increasing ATP synthesis.

Multiple studies have measured ATP increases following PBM treatment. Passarella S et al. demonstrated increased ATP production in isolated mitochondria exposed to HeNe laser light (632.8nm) as early as 1984. More recent work by Ferraresi C et al. (2015) showed that PBM increased ATP content in human muscle tissue.

Why does this matter clinically? Cells under stress — from injury, inflammation, infection, or ischaemia — often have impaired mitochondrial function and insufficient ATP for repair processes. By boosting ATP production, PBM provides the energy these cells need to heal, regenerate, and maintain normal function.

This is also why PBM tends to show the greatest benefit in damaged or stressed tissue rather than healthy tissue. A cell operating at full capacity has less room for improvement than one operating at, say, 60% of its mitochondrial potential.

2. Nitric Oxide Release

Under normal conditions, nitric oxide (NO) binds to the CuB and haem a3 centres of CCO, competing with oxygen and inhibiting enzyme activity. This is a normal regulatory mechanism — NO is a signalling molecule that helps control cellular respiration.

However, when CCO absorbs a red or NIR photon, the energy photodissociates NO from its binding sites, releasing it into the cell and the surrounding tissue. This has two simultaneous effects:

First, it relieves CCO inhibition, allowing the enzyme to bind oxygen more efficiently and further increasing ATP production. Under conditions of cellular stress — injury, inflammation, or ischaemia — NO binding to CCO is often increased, further suppressing already compromised mitochondrial function. PBM effectively rescues these cells from an energy-deficit state by clearing the NO blockade.

Second, the released NO acts as a potent vasodilator, relaxing smooth muscle in blood vessel walls and increasing local blood flow. Improved circulation brings more oxygen and nutrients to the treated area and carries away metabolic waste products. NO also functions as a neurotransmitter and immune signalling molecule, with downstream effects on pain perception, immune cell recruitment, and tissue remodelling.

This NO-mediated vasodilation is one reason PBM can produce immediate, visible effects — skin flushing and warmth in the treated area are common and reflect genuine circulatory changes. The effect is not merely cosmetic: increased perfusion delivers oxygen to hypoxic tissue, which is critical in wound healing, post-surgical recovery, and chronic inflammatory conditions where local blood supply is often compromised.

It is worth noting that this NO release mechanism also explains why PBM may interact with conditions involving NO dysregulation, such as cardiovascular disease and erectile dysfunction — though clinical evidence for these applications remains preliminary.

Key reference: Lane N. “Cell biology: power games.” Nature. 2006;443(7114):901-903. PMID: 17066004

3. Reactive Oxygen Species (ROS) Modulation

This is perhaps the most nuanced of the three primary effects. PBM transiently increases reactive oxygen species (ROS) production in the mitochondria — but at low, signalling-level concentrations, not the high levels associated with oxidative damage.

This mild, controlled increase in ROS acts as a cellular signal, activating transcription factors and signalling pathways that upregulate protective and restorative processes. Think of it as a hormetic stimulus — a small stress that triggers a disproportionately large adaptive response.

At the correct dose, this ROS signalling activates antioxidant defence pathways, ultimately leaving the cell better protected against oxidative stress than before treatment. At excessive doses, however, ROS production can overwhelm these defences — which is why overdosing PBM can be counterproductive (see biphasic dose response below).

Key reference: de Freitas LF, Hamblin MR. “Proposed mechanisms of photobiomodulation or low-level light therapy.” IEEE Journal of Selected Topics in Quantum Electronics. 2016;22(3):7000417. PMC4834554

Downstream Signalling Cascades

The three primary effects — ATP increase, NO release, and ROS modulation — trigger a cascade of downstream molecular events that produce the clinically observed benefits of PBM.

NF-kB Pathway

Nuclear factor kappa B (NF-kB) is a master transcription factor that controls the expression of hundreds of genes involved in inflammation, immunity, cell survival, and proliferation. PBM modulates NF-kB activity in a context-dependent manner:

• In acutely inflamed tissue, PBM tends to suppress NF-kB activation, reducing the production of pro-inflammatory cytokines (TNF-a, IL-1B, IL-6) and inflammatory mediators (prostaglandin E2, cyclooxygenase-2).
• In quiescent tissue, PBM can mildly activate NF-kB, promoting cell survival and proliferation.

This dual behaviour is one reason PBM can reduce inflammation in injured tissue while promoting growth and repair in healing tissue.

Anti-Inflammatory Cytokines

PBM shifts the cytokine balance from pro-inflammatory toward anti-inflammatory. Studies have shown increased production of IL-10 (a potent anti-inflammatory cytokine) and decreased levels of TNF-a, IL-1B, and IL-6 following PBM treatment.

Hamblin MR (2017) published a comprehensive review of the anti-inflammatory mechanisms of PBM, documenting consistent reductions in inflammatory markers across multiple tissue types and conditions. AIMS Biophysics. 2017;4(3):337-361. PMC5523874

Gene Expression Changes

PBM does not merely provide a temporary energy boost. It alters gene expression through the activation of transcription factors including NF-kB, AP-1, and CREB. Documented gene expression changes include:

• Upregulation of genes involved in cell proliferation, collagen synthesis, growth factor production (VEGF, FGF, PDGF), and antioxidant defence (SOD, catalase, glutathione peroxidase)
• Downregulation of genes involved in apoptosis (programmed cell death) and inflammatory signalling

These changes persist for hours to days after a single treatment session, explaining why PBM effects can outlast the brief treatment window.

Growth Factors and Tissue Repair

PBM has been shown to upregulate several growth factors that are central to tissue repair:

• Vascular endothelial growth factor (VEGF) — promotes angiogenesis (new blood vessel formation), critical for wound healing and tissue regeneration
• Fibroblast growth factor (FGF) — stimulates fibroblast proliferation and collagen synthesis
• Platelet-derived growth factor (PDGF) — recruits cells to wound sites and stimulates tissue remodelling
• Transforming growth factor beta (TGF-B) — involved in extracellular matrix production and wound contraction

These growth factor responses have been documented in cell cultures, animal wound models, and human clinical studies. They provide the molecular explanation for PBM’s well-documented effects on wound healing — the therapy does not merely reduce inflammation, it actively promotes the biological processes of tissue repair.

Stem Cell Activation

An increasingly studied downstream effect is PBM’s ability to stimulate stem cell proliferation and differentiation. Arany PR et al. (2014) demonstrated that PBM activated latent TGF-B1, which in turn stimulated dental stem cells to differentiate into odontoblasts (dentin-producing cells). This study, published in Science Translational Medicine, was notable for identifying a specific molecular pathway from photon absorption to clinical outcome — a level of mechanistic detail rare in PBM research. PMID: 24848257

Other studies have shown PBM-enhanced proliferation of mesenchymal stem cells, adipose-derived stem cells, and bone marrow stem cells, suggesting potential applications in regenerative medicine.

The Biphasic Dose Response (Arndt-Schulz Curve)

One of the most important — and most misunderstood — principles in PBM is the biphasic dose response, often illustrated by the Arndt-Schulz curve.

The principle is straightforward: too little light produces no effect, the right amount produces optimal benefit, and too much light inhibits the biological response or causes harm.

This is not unique to PBM. It is a well-established biological principle known as hormesis, observed across pharmacology, toxicology, and exercise physiology. A small dose of exercise improves fitness; an excessive dose causes overtraining and injury. A therapeutic dose of a drug heals; an overdose kills.

In PBM, the biphasic response has been demonstrated repeatedly at both the cellular and clinical level:

• Huang YY et al. (2009) showed that low fluences (0.5-4 J/cm²) of 810nm light stimulated human fibroblast proliferation, while high fluences (10-50 J/cm²) inhibited it. Dose Response. 2009;7(4):358-383. PMC2790317

The practical implications are significant:

• More treatment is not always better. Doubling the dose does not double the benefit — it may eliminate it entirely.
• The therapeutic window is relatively narrow. Typical effective fluences for most conditions fall between 1-10 J/cm², though this varies by tissue depth and condition.
• Consumer devices that are “too weak” may genuinely fail to reach the therapeutic threshold, while using a powerful panel for too long may exceed the optimal dose.

This is why dosing parameters matter enormously, and why studies that fail to report irradiance and fluence are scientifically unhelpful.

Water as a Chromophore at Longer Wavelengths

While CCO is the primary photoacceptor for wavelengths between 620nm and 900nm, longer near-infrared wavelengths (940nm and above) interact with a different chromophore: structured water.

Water molecules within cells and in the interfacial layers surrounding proteins and membranes are not randomly arranged — they form structured clusters with distinct optical properties. These water structures have absorption peaks at approximately 940nm, 1060nm, and higher wavelengths.

When NIR light at these wavelengths is absorbed by structured water, it can alter the vibrational state of water clusters, potentially affecting membrane dynamics, ion channel function, and protein conformation. This is a newer area of research and the mechanisms are less thoroughly characterised than the CCO pathway.

Pollack GH and colleagues have proposed that interfacial water (so-called “exclusion zone” or EZ water) plays a significant role in cellular bioenergetics and may be a secondary target for PBM at longer wavelengths. While this work is intriguing, it remains more speculative than the well-established CCO mechanism.

Practically, this provides a rationale for the therapeutic use of wavelengths like 940nm and 1060nm, which some newer devices incorporate alongside the standard 660nm and 850nm. However, the evidence base for these longer wavelengths is substantially smaller than for 660nm and 850nm, and consumers should be cautious about devices marketed primarily on the basis of novel wavelengths without corresponding clinical evidence.

Additional Proposed Photoacceptors

Beyond CCO and water, researchers have proposed several other molecules as potential photoacceptors for PBM:

• Opsins — light-sensitive proteins found not only in the retina but also in skin, adipose tissue, and other peripheral tissues. Melanopsin (OPN4) and neuropsin (OPN5) respond to blue and violet light, while other opsins may respond to longer wavelengths. Findings by Sikka G et al. (2014) suggest vascular opsins may mediate some PBM effects on blood vessels.
• Flavins and flavoproteins — these molecules absorb blue light (around 450nm) and are involved in cellular signalling. They are more relevant to blue light therapy than to red/NIR PBM.
• Porphyrins — light-absorbing molecules found in haemoglobin and myoglobin. Their role in PBM is debated but they may contribute to effects in blood-rich tissues.
• Transient receptor potential (TRP) channels — ion channels that can be activated by heat and potentially by light-induced thermal effects. TRP channels may mediate some of the pain-relieving effects of PBM.

These additional photoacceptors are not as well established as CCO and remain active areas of investigation. They may help explain PBM effects at wavelengths or in tissues where CCO-mediated mechanisms are insufficient to account for observed outcomes.

Why PBM Works Across Seemingly Unrelated Conditions

A common criticism of red light therapy is that it claims to treat too many things. How can one therapy help with pain, wound healing, hair growth, skin ageing, brain function, and oral mucositis?

The answer lies in the mechanism itself. PBM does not “treat” any of these conditions directly. It acts at the most fundamental level of cellular biology — mitochondrial energy production, blood flow, and inflammatory signalling — which are relevant to virtually every tissue and every pathological process.

Consider an analogy: if a factory loses power, every department shuts down — manufacturing, quality control, logistics, communications. Restoring the power supply does not specifically fix any one department; it enables all of them to resume normal operations. PBM is, in a simplified sense, restoring power to cells whose mitochondria are underperforming.

This is why:

• Wound healing improves — fibroblasts, keratinocytes, and immune cells need ATP for proliferation, migration, and matrix synthesis
• Pain reduces — inflammation-mediated pain decreases when pro-inflammatory cytokines are downregulated and local circulation improves
• Hair follicles reactivate — follicular cells in the resting (telogen) phase may re-enter the growth (anagen) phase when given sufficient mitochondrial energy
• Skin collagen increases — fibroblasts increase collagen and elastin production when cellular energy supplies are adequate

This does not mean PBM will work for every condition in every person. Dose matters. Wavelength matters. Tissue depth matters. The severity and nature of the underlying pathology matter. But the breadth of applications is consistent with, not contradicted by, the mechanism.

Summary of the Mechanism

The molecular mechanism of photobiomodulation can be summarised in five steps:

1. Photon absorption. Red (620-680nm) or near-infrared (760-900nm) photons are absorbed by cytochrome c oxidase in the mitochondrial electron transport chain.

2. Primary biochemical events. ATP production increases, nitric oxide is released from CCO, and a transient, low-level increase in reactive oxygen species occurs.

3. Signalling cascade activation. These primary events activate transcription factors (NF-kB, AP-1, CREB) and modulate cytokine production, shifting the cellular environment from pro-inflammatory toward anti-inflammatory.

4. Gene expression changes. Upregulation of genes for cell proliferation, collagen synthesis, growth factor production, and antioxidant defence. Downregulation of pro-apoptotic and pro-inflammatory genes.

5. Clinical outcomes. Faster wound healing, reduced pain and inflammation, improved tissue repair, enhanced cellular function across the treated area.

The science is not speculative. The mechanism is documented at every level — from isolated mitochondria to cell cultures to animal models to human clinical trials. Where the research still needs to catch up is in determining optimal protocols for each specific clinical application.

Tissue-Specific Responses

While the core mechanism is universal (any cell with mitochondria can respond), the clinical outcome varies by tissue type because different tissues have different baseline mitochondrial densities, different depths below the skin surface, and different rates of cellular turnover:

• Skin — highly accessible to red light (minimal penetration depth needed), fast cellular turnover, abundant fibroblasts. This is why skin-related applications (wound healing, collagen production, acne) often show consistent results: the target tissue receives the full dose.
• Muscle — deeper, requiring NIR wavelengths (810-850nm) for adequate penetration. PBM effects on muscle recovery are real but depend critically on delivering sufficient dose to the target depth.
• Joints — covered by skin, subcutaneous fat, and capsular tissue. Reaching articular cartilage requires high-irradiance NIR. This explains why WALT-recommended doses for joint conditions are higher than for superficial applications.
• Brain — the skull attenuates approximately 96-98% of incident NIR light. Transcranial PBM is feasible (enough photons reach the cortex to produce biological effects) but requires high surface irradiance and longer treatment times.
• Internal organs — largely inaccessible to external PBM. While some photons at 850-940nm can penetrate to abdominal organs, the attenuation is severe. Claims about PBM treating liver disease or gut conditions from external application should be viewed with significant caution.

Understanding these tissue-specific factors is essential for realistic expectations. PBM is not equally effective everywhere — its impact is proportional to the dose that actually reaches the target cells.

Core references for this page:

• Hamblin MR. “Mechanisms and applications of the anti-inflammatory effects of photobiomodulation.” AIMS Biophysics. 2017;4(3):337-361. PMC5523874
• Karu TI. “Multiple roles of cytochrome c oxidase in mammalian cells under action of red and IR-A radiation.” IUBMB Life. 2010;62(8):607-610. PMID: 20681024
• de Freitas LF, Hamblin MR. “Proposed mechanisms of photobiomodulation or low-level light therapy.” IEEE J Sel Top Quantum Electron. 2016;22(3):7000417. PMC4834554
• Huang YY et al. “Biphasic dose response in low level light therapy.” Dose Response. 2009;7(4):358-383. PMC2790317