Modern humans have radically changed their relationship with light. We spend roughly 90 per cent of our lives indoors, increasingly behind energy-efficient glazing and under LED illumination. That has undoubtedly reduced energy consumption, but it has also unintentionally changed the spectrum of radiation reaching our bodies.

A provocative hypothesis explored by Graham Lawton in the 15 August 2026 issue of New Scientist suggests that this may have biological consequences. The idea has been termed “red-light starvation”: modern life may deprive us of some of the red and near-infrared wavelengths to which humans were routinely exposed throughout our evolutionary history.

The underlying science is credible and increasingly interesting. The larger claims about its importance to human health, however, remain unproven.

What have we lost?

Natural sunlight isn't simply the visible rainbow. It contains substantial amounts of long-wavelength red and near-infrared radiation.

Historically, humans also received these wavelengths from fire and, much more recently, incandescent lighting. Modern LEDs are very different. They produce plenty of visible light but comparatively little radiation beyond about 750 nanometres. Modern energy-efficient windows may remove still more infrared radiation.

Consequently, someone spending most of the day inside a modern building may receive a substantially narrower spectrum of light than somebody outdoors. That would be merely an interesting consequence of technological progress were it not for growing evidence that our bodies can actually use these wavelengths.

The mitochondrial connection

The most important part of the story concerns mitochondria - the structures within our cells responsible for producing much of their usable energy in the form of ATP.

Red and near-infrared wavelengths penetrate biological tissue much more effectively than shorter wavelengths. A substantial experimental literature shows that exposure to particular red and near-infrared wavelengths can influence mitochondrial function, oxidative stress and cellular signalling.

One proposed mechanism involves the mitochondrial electron transport chain. Red or near-infrared photons appear capable of affecting the movement of electrons through this system, potentially allowing mitochondria to produce ATP more efficiently. Glen Jeffery of University College London describes the phenomenon as “photometabolism”.

The precise mechanism isn't settled. Cytochrome-c oxidase has frequently been proposed as an important target, while other researchers are investigating whether water molecules associated with mitochondria absorb the photons and change their physical behaviour in ways that facilitate electron transport. Whatever the precise mechanism proves to be, the broader proposition that red and near-infrared light can produce biological effects is supported by considerable experimental evidence.

A surprising effect on blood sugar

One of the most intriguing pieces of human evidence comes from research by Michael Powner and Glen Jeffery published in 2024. Healthy participants received 15 minutes of 670-nanometre red light before undergoing an oral glucose tolerance test.

The results were striking. Total blood-glucose elevation over the following two hours was reduced by 27.7 per cent, while the maximum glucose spike was reduced by 7.5 per cent.

This doesn't demonstrate that red light prevents or treats type 2 diabetes. It was a relatively small experiment in healthy people rather than a long-term clinical trial. It does, however, provide evidence for something potentially important: red-light exposure can measurably alter human glucose metabolism.

Jeffery has also reported experiments comparing office environments illuminated by LEDs with environments containing some incandescent illumination. Blood glucose was reportedly lower under the latter conditions, apparently because incandescent bulbs restore some of the long wavelengths largely absent from LED lighting. That finding requires replication, but it raises an intriguing possibility: perhaps some benefits attributed to red-light therapy represent correction of an environmental deficiency.

“21st-century scurvy”?

Jeffery makes the deliberately provocative comparison with scurvy. Scurvy wasn't understood until people realised that something routinely present in a normal diet was necessary for health. Remove vitamin C for sufficiently long and disease gradually develops.

Could removing a component of our ancestral light environment have analogous consequences? Jeffery suggests that inadequate red and near-infrared exposure might contribute to the mitochondrial dysfunction associated with ageing and potentially conditions including obesity, type 2 diabetes and neurodegenerative disease.

It's an intriguing hypothesis. It isn't yet an established medical fact.

Dementia and Parkinson's disease

Photobiomodulation - the deliberate therapeutic application of red or near-infrared light - has attracted interest in neurological disease. Small human studies in Parkinson's disease have produced encouraging signals, while animal studies provide stronger evidence that photobiomodulation can influence neuronal function and mitochondrial biology.

Research in Alzheimer's disease and mild cognitive impairment has similarly produced some positive findings. A 2025 pilot study using 810-nanometre near-infrared treatment reported improvements on cognitive tests, but only nine patients completed treatment and there was no sham-treated control group.

More recent research in mild cognitive impairment has provided additional encouraging evidence, including improvements following several weeks of home-administered transcranial photobiomodulation. These results justify larger clinical trials. They do not yet demonstrate that red-light therapy prevents dementia - still less that inadequate environmental red light causes it.

What about cardiovascular disease?

The New Scientist article discusses a University of Edinburgh analysis involving more than 400,000 people. Greater habitual exposure to sunlight was associated with lower risks of cardiovascular and all-cause mortality.

That is interesting evidence for the potential benefits of sunlight, but it cannot establish the case for red light specifically. The researchers were studying ultraviolet exposure as a proxy for sunlight. People who spend more time outdoors also differ in physical activity, vitamin-D status, circadian exposure and potentially many other respects. Sunlight itself can influence nitric-oxide biology independently of infrared radiation.

It would therefore be a mistake to conclude from this study that infrared light produced the cardiovascular benefit.

Therapy and ordinary daylight aren't the same thing

A photobiomodulation device delivering precisely controlled 670 or 810-nanometre radiation isn't equivalent to walking outside. Nor is sitting under a red LED necessarily equivalent to exposure to natural sunlight.

Wavelength, intensity, duration and the area of the body exposed all potentially matter. Photobiomodulation also appears to exhibit a dose-response relationship in which more isn't necessarily better. Excessive exposure may cease to provide benefit or potentially become counterproductive. This is one reason to be cautious about the rapidly expanding consumer market for powerful red-light panels.

How convincing is the evidence?

The science can usefully be divided into several levels of confidence:

The distinction matters because an intriguing biological mechanism can easily become an exaggerated health claim. There is a substantial difference between demonstrating that red light influences mitochondria and demonstrating that the replacement of incandescent lighting with LEDs has contributed to modern epidemics of diabetes, dementia or cardiovascular disease. We currently have good evidence for the former and insufficient evidence for the latter.

The experiment we really need

The red-light-starvation hypothesis is testable. An ideal experiment would compare people living or working for an extended period under conventional modern LED illumination with a similar group exposed to lighting supplemented with physiologically realistic amounts of red and near-infrared radiation.

Researchers could then measure glucose control, insulin sensitivity, mitochondrial biomarkers, inflammatory markers, sleep and circadian parameters and, eventually, clinical outcomes. A sufficiently large and well-controlled study could determine whether restoring these missing wavelengths produces meaningful improvements in health.

Until such trials exist, “red-light starvation” should be regarded as a serious research hypothesis rather than a newly discovered deficiency disease.

So should we all buy red-light lamps?

Probably not - at least not on the strength of the evidence currently available. The more interesting conclusion from the research is considerably simpler: go outside.

Natural daylight provides the broad spectrum of radiation under which humans evolved. Red and near-infrared wavelengths penetrate clothing and skin much more effectively than ultraviolet radiation, so obtaining them doesn't require sunbathing or deliberately increasing potentially harmful UV exposure.

Some hospitals are already exploring this principle architecturally. King's College Hospital in London has created outdoor space allowing patients access to daylight rather than keeping them continuously behind glass.

For people confined to offices, restoring modest amounts of longer-wavelength light may eventually prove worthwhile. Intriguingly, an old-fashioned incandescent bulb produces substantial red and infrared radiation, although at the cost of much poorer energy efficiency than LEDs. What isn't yet justified is assuming that an expensive, high-powered red-light therapy panel must therefore be beneficial.

The genuinely important question

The most compelling aspect of this research isn't red-light therapy itself. It's the possibility that technology has changed an environmental variable we didn't realise was biologically important.

Over little more than a century, humans have moved from spending substantial portions of their lives outdoors - supplemented by fire and incandescent illumination - to spending most of their time behind infrared-filtering glass illuminated by LEDs whose spectrum bears relatively little resemblance to sunlight.

We made that transition without knowing that wavelengths outside conventional visible illumination might influence cellular metabolism. It is entirely plausible that the change is biologically irrelevant. It is also plausible that restoring some of those wavelengths will ultimately prove beneficial. We don't yet know.

But given the accumulating evidence that red and near-infrared light interact with mitochondrial metabolism, it is a question worth answering properly.

In the meantime, the sensible response isn't to regard red light as the next vitamin D or to spend thousands of pounds on photobiomodulation equipment. It is simply to recognise that our biology evolved outdoors - and natural daylight may be doing rather more to us than merely allowing us to see.

Selected scientific sources

• Powner MB et al. (2024). Light stimulation of mitochondria reduces blood glucose levels. Journal of Biophotonics. PMID: 38378043.

• Peer-reviewed reviews of photobiomodulation and mitochondrial mechanisms, including work examining cytochrome-c oxidase, ATP production, oxidative stress and cellular signalling.

• Human pilot and feasibility studies of photobiomodulation in Parkinson's disease (including PMID: 34215216 and PMID: 38094162).

• Pilot studies of near-infrared photobiomodulation in Alzheimer's disease and mild cognitive impairment, including PMID: 39910867.

• University of Edinburgh/UK Biobank observational research examining sunlight/UV exposure and mortality (PMID: 39094281).

• Lawton G. “Red-light starvation.” New Scientist, 15 August 2026, pp. 30-33.