Longer wavelengths of sunlight do not stop at the skin. In a 2025 Scientific Reports paper from University College London and collaborators, near-infrared energy in the 830 to 860 nm band was measured after it had crossed the human thorax. The same laboratory then delivered a controlled 15-minute 850 nm exposure to the back and retested color contrast 24 hours later. Vision improved, including in participants whose heads were wrapped in foil so the eyes received no direct light. Clothing reduced intensity. It did not act as a wall.
That finding sits on a mitochondrial claim that is older than this paper. Cytochrome c oxidase and related chromophores absorb in the red to near-infrared window. Longer wavelengths can raise membrane potential and ATP output when those systems have declined. The outer retina is mitochondrially dense, which is why color contrast, especially the metabolically expensive tritan axis, has been used as a distal readout. The 2025 study is important because it treats the body as an optical volume, not only as a surface to be aimed at.
Methods and scale must be stated without inflation. Forty Caucasian adults aged 25 to 63 were studied under ethics approval ETH2324-0250. Sunlight transmission through the chest was recorded with a spectrometer and radiometer housed in a light-tight aluminum box on the sternum while subjects stood back-to-sun, unclothed from the waist up. An 850 nm ceiling panel (9.18 mW/cm2 at 50 cm; half-power roughly 830 to 865 nm) was then used in the dark. Thirteen subjects received 15 minutes of back exposure with no upper clothing. Five more received the same exposure with the head wrapped in foil. Seven sham controls stood in the same position with the panel off. Color contrast thresholds were measured with Chroma Test on protan and tritan axes, using the same isoluminant letter protocol as Shinhmar et al. Secondary measures showed 850 nm light passing a hand and remaining detectable through layered wool and polyester. Independent summaries of the paper report on the order of 16 percent tritan and 9 percent protan improvement when eyes were free, and about 7 percent tritan improvement when the eyes were blocked. Controls did not move. Those percentages should be read as the authors’ reported shifts in a modest sample, not as a population guarantee.
The optical physics is consistent with known tissue windows. Water and hemoglobin dominate absorption across the visible. Near 850 nm those absorbers leave more path length, which is why infrared cameras can image transmission through a hand and why a thorax measurement is physically plausible even if only a tiny fraction of incident energy exits the far side. The paper’s own framing is that a small transmitted fraction can still be biologically relevant if mitochondria integrate that signal.
Methods and scale must be stated without inflation.
This 2025 experiment is sham-controlled and includes an ocular-block arm. It is not a large double-blind randomized trial. Treat that distinction as part of the evidence, not as a footnote to bury.
Comparable peer-reviewed human work from the same research line:
Shinhmar H, Hogg C, Neveu M, Jeffery G. Weeklong improved colour contrast sensitivity after single 670 nm exposures associated with enhanced mitochondrial function. Scientific Reports. 2021;11:22872. doi:10.1038/s41598-021-02311-1. Single 3-minute morning 670 nm exposures improved aged cone contrast, with reported mean tritan improvement around 17 percent lasting up to 1 week when delivered in the morning.
Shinhmar H et al. Optically improved mitochondrial function redeems aged human visual decline. Journals of Gerontology Series A. 2020;75(9):e49–e54. doi:10.1093/gerona/glaa094. Two weeks of daily 670 nm improved color contrast in adults older than about 40, with the largest reported effect on the tritan axis.
Kaynezhad P and colleagues, ARVO/IOVS 2023 abstract: randomized 670 nm versus 535 nm green-light control in adults 30 to 73. Tritan thresholds improved about 15 percent after 670 nm; between-group comparison versus green was weaker. This is one of the cleaner wavelength-control designs in the vision-PBM literature.
Double-blind clinical photobiomodulation trials in retina, different indication and device but same wavelength family:
LIGHTSITE II, prospective double-blind randomized trial in intermediate dry AMD using multiwavelength PBM including 590, 660, and 850 nm, reported visual acuity benefit versus sham.
LIGHTSITE III, prospective double-blind randomized multicenter trial in early-to-intermediate dry AMD (Valeda system; 100 subjects across 10 US centers), likewise sham-controlled.
These AMD trials test diseased retina and a clinical device protocol. They do not prove that 15 minutes of clothed sunlight will change acuity. They do show that sham-controlled, masked PBM designs exist in ophthalmology and that 850 nm is already inside those device spectra.
Powner MB, Jeffery G. Light stimulation of mitochondria reduces blood glucose levels. Journal of Biophotonics. 2024;17(5):e202300521. doi:10.1002/jbio.202300521. PMID: 38378043. Fifteen minutes of 670 nm light, in a human glucose-tolerance protocol, reduced the integrated rise in blood glucose over two hours by 27.7 percent and cut the peak spike by 7.5 percent. That is a different endpoint from color contrast, but it is the same laboratory’s published human evidence that a regional long-wavelength input can move a whole-body metabolic readout.
Glen Jeffery, Professor of Neuroscience at University College London’s Institute of Ophthalmology and corresponding author of the 2025 paper, has stated the transmission result in his own words. In an August 2025 interview with Kirk Hamilton that Jeffery approved for publication: “We found that infrared light from the sun does penetrate through the torso, and we measured it coming out the back.” On the vision arm: infrared exposure “improved vision, even when the eyes themselves weren’t directly exposed.” On mechanism, in a 2021 UCL statement on the 670 nm work: “Mitochondria have specific sensitivities to long wavelength light influencing their performance: longer wavelengths spanning 650 to 900nm improve mitochondrial performance to increase energy production.” Those are the investigator’s statements, not a press rewrite of the methods.
What a careful reader should keep: The transmission claim is instrumented, not anecdotal. The vision claim is a 24-hour color-contrast change after 15 minutes of 850 nm to the back, including an eyes-blocked arm. Clothing is optically leaky at 850 nm in the fabrics they tested. Mechanism remains photobiomodulation of mitochondria plus a still-unsettled systemic carrier, possibly cytokines or a circulating metabolic signal. Limits are real: small n, Caucasian cohort, one time point, no large double-blind sunlight trial, and media language that runs ahead of the methods.
What is proven, as measurement rather than hope: near-infrared in the 830 to 860 nm band can be recorded after it has crossed a human thorax. Clothing in the fabrics they tested reduced intensity and did not stop it. A 15-minute 850 nm exposure to the back was followed, 24 hours later, by improved color-contrast thresholds, including in an eyes-blocked arm, against sham controls who did not move. Earlier papers from the same line show 670 nm improving tritan contrast in adults over about 40, with a morning-timing effect. The 2024 glucose-tolerance study shows a smaller post-meal glucose rise after 670 nm. LIGHTSITE II and III show that sham-controlled PBM trials at related wavelengths exist in ophthalmology. That is the published record.
What is theorized and not yet proven in these subjects: that cytochrome c oxidase, or any named chromophore, was the causal absorber in the 2025 thorax experiment. Mitochondrial ATP was not measured in those volunteers. The circulating carrier — cytokine, metabolic signal, or something else — is unnamed. Outdoor, fully clothed sunlight has not been shown, in a large trial, to match the 850 nm panel. The reported percentages are shifts in a modest Caucasian sample, not a population guarantee. Language that sunlight “heals mitochondria in minutes through clothes” runs ahead of the methods.
What upcoming work would have to show, and how it might help: a larger, more diverse, double-blind trial of sunlight or 850 nm to the body, with vision and metabolic endpoints, including skin phototypes the 2025 cohort did not cover. AMD device trials (Valeda / LIGHTSITE) will keep testing whether 590, 660, and 850 nm PBM can move acuity in diseased retina — a different claim, and the closest clinical path. Metabolic follow-ups of the 2024 glucose result would tell whether long-wavelength exposure is a useful adjunct around meals. If those hold, the help is modest and specific. Older adults may get a measurable lift in color contrast from brief morning long-wavelength light. Some people may see a smaller glucose spike after a 670 nm session. Going outside remains the everyday, uncontrolled source of the same band. None of that is a license to stare at a high sun. It is a reason to treat the body as an optical volume and to keep the methods in view while the next trials run.
Jeffery et al., Sci Rep 15, 24435 (2025)
