We demonstrate that the visual performance of those working under standard LED is significantly improved by exposure to incandescent lighting that has a spectrum similar to daylight with an extensive infrared component. These data are consistent with the hypothesis that LED lighting undermines human visual performance. This result is consistent with laboratory experiments where specific red/infrared wavelength ranges generated by LEDs have been used to improve visual function in animals and humans in a conserved manner^13,16,17. But there are three critical differences from these earlier studies. First, we have simply changed environmental lighting in a free moving work environment. Second, we have obtained significant balanced improvements in both the protan and tritan range. Previously, exposure to restricted experimental 670 nm resulted in improvements biased strongly in favour of only tritan function¹³. Hence, exposure to full spectrum lighting results in a balanced pattern of improvement in visual performance. Third, we have shown that improvements in visual function following incandescent light exposure are sustained for up to 6 weeks, and possibly beyond, whereas benefits from single LED restricted range red light were confined to around 5 days¹³. These three features change the way in which long wavelength light may be applied to improve human physiology by delivery in normal environments with lasting balance effects. These results are novel and may have public health implications.

Our study used 22 subjects but was statistically significant using both a before and after metric and also against an independent control group. They are also similar to group sizes in aspects of Shinhmar et al.¹³ (Figs. 2, 3, 4 and 5). However, future studies would clearly benefit from inclusion of a larger number of subjects.

The evolution of life on earth extends over 4 billion years, and that of humans over approximately 4–5 million years from the last common primate ancestor. This has all taken place under sunlight that has a spectral range of approximately 300–2500 nm+, within which there has been an invariant balance between short and longer wavelengths. Human adoption of fire 1–2 million years ago supplemented sunlight as they moved out of Africa as its spectrum is similar having a large infrared component. Likewise, development of the Edison filament luminaire, common until approximately the year 2000 had a spectrum similar to sunlight. However, around 2010 LED lighting with its highly restricted spectrum (350–650 nm) and energy saving characteristics became common, resulting in a loss of infrared light in the built environment¹.

The physiology of life forms are adapted to natural environmental light in a highly conserved pattern across species. Light impacts on mitochondrial function, which is a key regulator of metabolism and ageing in animals. When the balance of short and long wavelengths is shifted there are consequences for mitochondria. When shorter wavelength exposure is dominant, as in LED lighting, mitochondrial function declines. Mitochondrial complex proteins are reduced and there is reduced ATP production^2,3. With reduced mitochondrial demand for glucose there is increased body weight and disruptions to serum cytokines⁴. Consequently, consistent with the mitochondrial theory of ageing there is an increased probability of cell/organism ageing and death¹⁸. It is suggested that this is partly due to 420–450 nm light, dominant in LEDs, being absorbed by porphyrin and the subsequent production of oxygen singlets driving inflammation³.

Conversely, exposure to longer wavelengths is associated with increased mitochondrial membrane potential and increased concentration of mitochondrial complex proteins that have declined with ageing and disease. This in turn is associated with elevated ATP, reduced inflammation and extended average lifespan^7,9,10,19. The experimental use of longer wavelengths in such situations is commonly referred to as photobiomodulation.

The retina has the greatest metabolic rate in the body and a high mitochondrial concentration²⁰. Retinal metabolism declines with age, but this can be partly corrected with long wavelength light across species^16,21. In humans a single 3 min 670 nm exposure improves colour vision within 3 h, which is sustained for almost a week¹³. But what the authors of this study did not appreciate was that this was within a population who worked and lived mainly under LED lighting that may have undermined their baseline measurements. Here, we made no attempt to control light exposures or subject movements as would occur in laboratory-based experiments. Rather, our aim was to introduce wide spectrum long wavelengths into a work environment to improving human performance via mitochondrial manipulation in a translational step.

There is considerable evidence that introduction of longer wavelengths impact systemically. Durieux et al.²² stated in relation to experiments in C.elegans that “ We find that mitochondrial perturbation in one tissue is perceived and acted upon by the mitochondrial stress response pathway in distal tissue”. In mice there are significant distinct changes serum cytokine expression to exposures of both short and long wavelength light^4,23. Similarly, long wavelength exposures to the surface of the human body excluding the eyes significantly reduces blood glucose levels and increases oxygen consumption in humans. This is likely because mitochondrial upregulation will increase carbohydrate demand to support increased ATP production¹². Other systemic impacts can be found and are clear in experimentally induced Parkinson’s in primates. Light targeted by implants focusing on the substantia nigra are effective in reducing symptoms²⁴, but so also are those that are directed at distal locations²⁵.

Single 3 min 670 nm exposures remain effective for about 5 days¹³. But we show that with a wider spectrum they remain effective for 6 weeks, although we did not find the end of the effect. Here it is worth considering potential mechanisms of action which remain subject of debate. Historically, improvements with red light were thought to be due to light absorption by cytochrome C in the respiratory chain²⁶. However, positive effects are found in vitro in the absence of this. Consequently, it has been suggested that longer wavelengths reduce water viscosity around rotary ATP pumps allowing the rotor to increase speed²⁷. This cannot explain the sustained impacts of light exposure as this effect should be relatively transitory as viscosity would increase rapidly following light withdrawal. However, a key feature of long wavelength light absorption is increased respiratory chain protein synthesis. These proteins are in flux throughout the day²⁸ and complex IV is upregulated following red light exposure¹⁹. Hence, while red light may initially increase rotor pump speed there rapidly follows an increase in protein synthesis which may establish greater respiratory chain capacity. The life of these proteins could then determine the length of effect.

Only thirteen polypeptides are made in mitochondrial protein synthesis. This probably slows with age and likely contributes to aged mitochondrial decline¹⁸. But critically, we do not know the speed of mitochondrial protein synthesis, the life of such proteins or the pace of their decline. We suggest that these may be key events in the length of the effects from light exposure.

LED lighting clearly has the ability to undermine visual performance probably via reduced mitochondrial function. As light induced changes in mitochondrial ability have been shown to have systemic impacts^{4,15,22,23,25}, the effects of LEDs revealed here may be wider than initially anticipated. Given the prevalence of LEDs, this may represent an important issue in public health and clinical environments where changing lighting patterns in appreciation of this point can have significant positive outcomes²⁹.

Given our results, it is important to ask what solutions may be found to improve health in terms of lighting in the built environment. Incandescent lights that we reveal here to have significant positive impact over standard LEDs are being phased out universally for reasons of energy efficiency, where focus is only on the visible light produced.

A solution may be found in creating lighting units with multiple longer wavelength LEDs to cover a wider span of the near infrared. However, our attempts here have had limited success. Multiple closely associated spectral peaks do not produce a smooth spectral output as found in incandescent lights and sunlight, which is problematic in improving function and has yet to deliver. This possibly may be overcome using a greater number of spectral peaks with tighter spacing. But this raises a different series of problems regarding cost and increased energy consumption making this solution no better than retention of incandescent sources in terms of environmental sustainability.

Key to this issue is the question of how much infrared is needed to sustain improved function? Infrared has relatively few absorbers in the built environment and in current studies relatively little has to be added to the environment for effect. However, a viable option is to run an incandescent light at a lower temperature which results in both energy savings and increased life of the unit and also shifts the peak spectral output towards longer wavelengths.

If this is done with a halogen bulb, which is a type of incandescent tungsten bulb, the filament lasts for a longer period as evaporated tungsten is redeposited on the filament rather than blackening the bulb glass. Hence, using a halogen bulb at lower voltage is a realistic alternative in terms of health and energy consumption.