5D Library › Crystal Piezoelectricity, Iron-Nickel Meteorites and Photobiomodulation · 中文
Chapter 5 Photobiomodulation: Cytochrome c Oxidase and the Biphasic Dose
From hair on mice to a clinical guideline
In 1967 the Hungarian surgeon Endre Mester tested whether a low-power ruby laser could cause cancer in mice. It did not; instead, the shaved patches of skin it shone on regrew hair faster than those of untreated mice (Mester et al. 1968). Over the following decades “low-level laser therapy” was tried for wound healing, pain and inflammation, at first with lasers and later with light-emitting diodes, and in recent years the field has adopted the name photobiomodulation (PBM): the use of red and near-infrared light, at irradiances too low to heat tissue appreciably, to change what cells do.
The proposed mechanism
Evidence
The leading explanation centres on cytochrome c oxidase (CCO), complex IV of the mitochondrial respiratory chain, which passes electrons to oxygen. Its copper and haem centres absorb in the red and near-infrared, and action spectra of cellular responses, measured by Tiina Karu, show peaks in the regions around 620, 680, 760 and 820 nm that resemble the absorption of CCO in its different oxidation states (Karu 1999). A proposed link in the chain is nitric oxide: NO binds to CCO in competition with oxygen and slows respiration; light can dissociate it, letting respiration and ATP production recover (Karu et al. 2005). Downstream, changes in mitochondrial membrane potential, a brief burst of reactive oxygen species and calcium signalling are thought to activate transcription factors that alter inflammation and cell survival (Hamblin 2017).
Note
The mechanism is not settled. Other proposed light absorbers include light- and heat-sensitive ion channels and the structured water at membrane surfaces, and different mechanisms may dominate at different wavelengths (Freitas and Hamblin 2016). The book uses the CCO account as the best-supported working model, not as an established fact.
Counting photons and joules
Photobiomodulation is described by its wavelength, its irradiance (power per area, mW/cm²) and its fluence or dose (energy per area, J/cm²), which is irradiance multiplied by time. Device labels often give only “power”; what matters to tissue is irradiance at the skin, which depends strongly on distance (Chapter 8).
Calculation
The energy of a photon at 660 nm and 850 nm, the number of photons in 10 mW/cm², and the time needed to deliver doses of 1, 4 and 10 J/cm² at three irradiances.
# Dose (fluence, J/cm²) = irradiance (W/cm²) × time (s).
h, c = 6.626e-34, 2.998e8
for nm in (660, 850):
E = h * c / (nm * 1e-9)
print(f"{nm} nm photon: {E / 1.602e-19:.2f} eV; 10 mW/cm² = {0.010 / E:.1e} photons per cm² per s")
print()
print("seconds needed for a dose of 1 J/cm² 4 J/cm² 10 J/cm²")
for mw in (5, 25, 100):
print(f" at {mw:3d} mW/cm² " + "".join(f"{H / (mw / 1000):10.0f}" for H in (1, 4, 10)))
# For scale: clear midday sunlight carries very roughly 30 mW/cm² between 600 and 900 nm.
print("100 mW/cm² of red and near-infrared ≈ 3 × that part of midday sunlight")
# 660 nm photon: 1.88 eV; 10 mW/cm² = 3.3e+16 photons per cm² per s
# 850 nm photon: 1.46 eV; 10 mW/cm² = 4.3e+16 photons per cm² per s
#
# seconds needed for a dose of 1 J/cm² 4 J/cm² 10 J/cm²
# at 5 mW/cm² 200 800 2000
# at 25 mW/cm² 40 160 400
# at 100 mW/cm² 10 40 100
# 100 mW/cm² of red and near-infrared ≈ 3 × that part of midday sunlightA strong panel at 100 mW/cm² delivers, in red and near-infrared, a few times what midday sunlight provides in the same band. It also warms the skin noticeably, which matters for every claim about its effects: warmth alone increases skin blood flow.
The biphasic dose response
Evidence
Across cell, animal and some clinical studies, the response to light follows a biphasic pattern, often compared to the Arndt–Schulz rule of pharmacology: too little light has no effect, a moderate dose has an effect, and a larger dose has a smaller one or inhibits. In a review, Huang and colleagues described effects in many cell studies at doses of the order of 1 J/cm² and below; in a mouse wound model, 2 J/cm² was more effective than 50 J/cm²; and doses of tens of J/cm² have often been inhibitory in vitro. The optimum depends on the wavelength, irradiance, cell type and the state of the tissue, and the same dose delivered at different irradiances may not have the same effect (Huang et al. 2009).
Calculation
An illustrative two-process model: a stimulating effect that saturates, minus an inhibiting effect that sets in at higher doses. None of the parameters is a measurement; the model only shows how a curve with a window can arise.
# A two-process picture of the biphasic dose response (all parameters illustrative):
# stimulation saturates: S(H) = H / (H + Hs)
# inhibition starts later: I(H) = (H/Hi)² / (1 + (H/Hi)²)
# net response: R(H) = S(H) − w · I(H)
Hs, Hi, w = 2.0, 20.0, 1.5 # J/cm², J/cm², weight
def net(H):
S = H / (H + Hs)
I = (H / Hi) ** 2 / (1 + (H / Hi) ** 2)
return S - w * I
for H in (0.3, 1, 2, 4, 6, 10, 20, 40, 60):
r = net(H)
bar = ("+" * round(20 * r)) if r >= 0 else ("−" * round(-20 * r))
print(f"{H:5.1f} J/cm²: net {r:+.2f} {bar}")
best = max((h / 10 for h in range(1, 600)), key=net)
zero = next(h / 10 for h in range(int(best * 10), 1000) if net(h / 10) < 0)
print(f"best dose ≈ {best:.1f} J/cm²; net effect turns negative above ≈ {zero:.0f} J/cm²")
# 0.3 J/cm²: net +0.13 +++
# 1.0 J/cm²: net +0.33 +++++++
# 2.0 J/cm²: net +0.49 ++++++++++
# 4.0 J/cm²: net +0.61 ++++++++++++
# 6.0 J/cm²: net +0.63 +++++++++++++
# 10.0 J/cm²: net +0.53 +++++++++++
# 20.0 J/cm²: net +0.16 +++
# 40.0 J/cm²: net -0.25 −−−−−
# 60.0 J/cm²: net -0.38 −−−−−−−−
# best dose ≈ 5.5 J/cm²; net effect turns negative above ≈ 26 J/cm²
The model’s practical lesson holds whatever the true parameters are: more is not better. A user who doubles a session “to be safe” may move from the stimulating part of the curve to the part where the effect disappears.
What has been shown in people
Evidence
The clearest clinical support is in cancer care. The Multinational Association of Supportive Care in Cancer and the International Society of Oral Oncology recommend intra-oral photobiomodulation, using low-level red laser light, around 630–660 nm, with specified parameters, to prevent oral mucositis, the painful mouth ulceration caused by chemotherapy and radiotherapy, in patients undergoing haematopoietic stem cell transplantation and in patients receiving radiotherapy for head and neck cancer (Zadik et al. 2019). That is a specific use, under protocols, delivered by trained staff.
For many other uses, such as muscle recovery, joint pain, hair loss, skin ageing and cognition, there are numerous trials, but they are mostly small, use widely different wavelengths, doses and devices, and are often difficult to blind, because light is visible and warm. Results are mixed. This book does not claim that red or near-infrared light treats or prevents any of these conditions.
Hypothesis · not established
Not established. In healthy skin, the short-term increase in blood flow after a red and near-infrared exposure of a few joules per square centimetre is explained by the warming of the skin, not by a photochemical effect. The hypothesis predicts that, against a sham that produces the same rise in skin temperature, active light produces no increase in perfusion large enough to matter. It would be refuted by the experiment in Chapter 8 if the active light exceeded the heat-matched sham by at least the pre-specified margin.
References
Fan, Y. (2026). Crystal Piezoelectricity, Iron-Nickel Meteorites and Photobiomodulation: Alpha-Quartz, Tourmaline Pyroelectricity, Widmanstätten Diffusion and Near-Infrared Cytochrome c Oxidase in Solid-State Physics and Biophotonics [晶体压电性、铁镍陨石与光生物调节] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23698498 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).