Time and again I’m asked by patients how they can protect themselves against the damage caused by blue light as emitted by LEDs, be it from visual display units or “living room” lighting LEDs.

Well, what should we do here? Nothing, there is no danger!

In the Internet and the asocial media there are loads of “reports” and “advice” along these lines: LEDs have a strong blue emittance, and blue is known to cause eye damage (eye cancer!), thus you should avoid LEDs or use filters. I will not link to any of this. To top it, the fear thus evoked is being monetised and filters are offered for glasses or your monitors.

Facts

  • Of course blue can harm, light in general, and blue more, because it has higher energy than red.
  • But it depends on the intensity!
  • Typical spectrum images, showing the “harmful” blue peak, are normalised to 100%, hiding the weakness of the blue.
  • Examples
    • With a cloud-covered winter sky I measured ≈5000 lx outside. A room-lighting LED spotlight, in contrast, illuminated with 120 lx in 2 m distance.
    • In summer the illuminance outside could be 100,000 lx (Wikipedia).
  • →The blue part emitted by LEDs is typically much weaker than the blue part of daylight. See animation.
  • Situations where light can cause sizable damage include:
    • Directly staring into a strong (blue) theatrical flashlight (so called ‘par’)
    • Looking without filter into “polymerisation lights” – this are used at the dentists for curing fillings
    • Kids competing “who can look longest into a laser” – yes, a number of such cases are documented. It’s even worse with a laser beyond the allowed power level, which unfortunately can easily be ordered on-line. Etc.

Scientific Information

I wrote the above a little simplified and with candour. But this isn’t just my “strong opinion”, there is good information to back it up, here some examples:

  • Lou L, Frishman LJ, Beach KM, Rajagopalan L, Hung LF, She Z, Smith III EL, Ostrin LA (2023) Long-term blue light rearing does not affect in vivo retinal function in young rhesus monkeys. From an excellent group I know well personally, with an objective measure (electroretinogram): “Long-term exposure to narrowband blue light did not affect photopic or scotopic ERG responses in young monkeys. Findings suggest that exposure to 12 h of daily blue light for approximately 10 months does not result in altered retinal function.”
  • Singh et al 2023, a Cochrane study, great! Despite the promising title “Blue-light filtering spectacle lenses for visual performance, sleep, and macular health in adults”, the authors systematically examine only the aspect of “visual fatigue” (also important, of course). And there they found no positive effects of blue-filtering lenses, even some weak negative effects (increase in depression, which is not entirely implausible, but not really relevant here due to the number of cases and subject selection).
  • Mainster, Findl, Dick, Desmettre, Ledesma-Gil, Curcio, Turner (2022) The Blue Light Hazard Versus Blue Light Hype. From the abstract: “The blue light hazard is misused as a marketing stratagem to alarm people into using spectacles and IOLs that restrict blue light. Blue light loss is permanent for pseudophakes with blue-blocking IOLs. Blue light hazard misrepresentation flourishes despite absence of proof that environmental light exposure or cataract surgery causes AMD or that IOL chromophores provide clinical protection. Blue-filtering chromophores suppress blue light critical for good mental and physical health and for optimal scotopic and mesopic vision.”
  • A BBC movie. Well explained, and also addresses the monetising aspects.
  • In December 2019 there was a OSA webinar, where O’Hagan explained this in detail.
  • O’Hagan et al. (2016) Low-energy light bulbs, computers, tablets and the blue light hazard. Clear statement here: “None of the sources assessed approached the exposure limits, even for extended viewing times.”
  • Aren’t there animal models that say that blue light damages the retina or promotes diseases (e.g. AMD)? The best known study on this is Krigel et al. 2016. I do not consider it appropriate to infer harmful effects for humans from this French study, because
    1. Rats are nocturnal animals; 12 h of light is stressful for them - continuous light exposure is actually used as a stressor in animal models of depression.
    2. Damage has only been observed in albino rats, and they have associated changes in the RPE65 gene that make them more sensitive to light damage.
    3. 2023 update: in primates (not rats!) no blue light damage was found in a similar experiment, see above: Lou L et al. (2023) Long-term blue light rearing does not affect in vivo retinal function in young rhesus monkeys.

Sleep and circadian rhythm

The alleged eye damage caused by LEDs is often confused with the influence of blue on the sleep-wake rhythm. However, this is something different and quite complex, here a brief synopsis:

  • Bright light before going to bed, regardless of the light source, is suboptimal for a good night’s sleep. Only at less than (melanopic²) 10 lx is there practically no influence on melatonin levels (Recommendations for healthy daytime, evening, and night-time indoor light exposure (preprint 2021), Fig. 2).
  • The widely cited work by Chang et al. (2015) reports that reading before falling asleep with a blue-dominated “e-reader” (iPad) triggers sleep disturbances compared to a (fluorescent-lit) book. BUT: The iPad was set to full brightness and therefore 47× brighter than the book! Thus this is a useless study, from which one can not conclude that blue light influences sleep.
  • For the circadian rhythm, it makes no difference whether, after cataract extraction, a blue-blocking (i.e. yellow) or clear intraocular lens is implanted (Brondsted et al. 2015).
  • Large-scale and methodologically sophisticated new work Schöllhorn et al. from the Basel Center for Chronobiology. 4 luminance levels were tested (27-284 cd/m²) under 2 conditions each: High vs. low effect on melanopsin-containing ganglion cells, although the screen always looked white – matching metameric lights) were generated with five light sources. The influence on sleep onset time, melatonin levels and alertness was measured. There was a clear dependence on luminance: on the one hand, at low luminance levels (27 cd/m²), there was no relevant effect on sleep time, but at high luminance levels there was a significant effect. Furthermore, the melanopsin-optimized lights had a significantly stronger effect, although they looked subjectively the same. I would summarize the results as follows:
    • Lights of low illuminance do not interfere with falling asleep; the 10-lux limit (see above) is obviously very safe.
    • If it has to be brighter, you can use suitable screens to mix lights that appear very bright, have no reduced blue component and therefore good color rendering, but have little effect on sleep behavior. Great.

→German version

² “melanopic” means that the wavelengths are weighted according to the spectral sensitivity of the melanopsin-containing ganglion cells in the retina; their maximum is in the blue-green range.