Questions abound about whether a blue-light hazard exists (MAGAZINE)

IAN ASHDOWN and STAN WALERCZYK respond to attendee questions posed during an LEDs Magazine webcast on the reality of a blue-light hazard and the path forward for human-centric lighting.

Back in January, Ian Ashdown and Stan Walerczyk, industry experts on color science and its relationship to human-centric lighting, presented a webcast entitled "LEDs and humans: Is blue light a hazard or can SSL improve our wellbeing." The webcast can still be viewed on demand. The event drew a tremendous number of questions given the contentious nature of the subject. The presenters took the time after the live event to answer many of the questions that were not able to be addressed in the one-hour webcast. For more background, you may also want to review a prior interview we published on the state of what the research community knows about non-visual impacts of lighting on humans that was based in part on one of the papers referenced in these answers.

Ian Ashdown

Q: What do you think is the reason there is such a furor over the blue-light hazard? It seems to be cropping up everywhere. Who benefits from designers not using blue-enriched light?

Ian Ashdown: Within the lighting industry, it is probably not so much a question of who benefits as that of people with strongly-held opinions. There is also an overlap with commercial companies selling blue-light blocking filters for eyewear and tablet screens. This, however, is a separate topic.

Stan Walerczyk: Many lighting professionals and architects were trained to think that incandescent is the holy grail of lighting and it can be difficult for them to change. And, as Ian stated, some lighting professionals just do not like high CCT at all and do not understand that others may.

Stan Walerczyk

Q: Is 460-nm light a problem for human vision health?

IA: At the light levels encountered with typical electric lighting, there is no evidence to indicate this.

Q: One of the points of one of the articles you cite, "Measuring and using light in the Melanopsin Age," was that the research isn't yet mature enough to advertise blue-light management for circadian control standards with spectral density management of illumination. Would you agree?

IA: This is a difficult but key question. The cited article was an opinion piece by 14 medical researchers, and they were quite correct in advising a cautionary approach. On the other hand, DIN SPEC 67600:2013, "Biologically Effective Illumination - Design Guidelines," is quite explicit in its recommendations. The specification, however, is also quite clear: "On the basis of current findings, it is recommended that humans spend about half an hour outdoors every day for the good of their health."

DIN SPEC 67600 takes a very pragmatic approach to the issue, basing its recommendations on current knowledge in the field and also experience gained from successful implementations. There will undoubtedly be claims made by small (and possibly large) companies regarding unsubstantiated benefits of their products, similar to some of the "full spectrum lamp" nonsense currently available on the Internet. From a lighting designer's perspective, however, perhaps the guiding principle should be primum non nocere - first, do no harm. We cannot advance the state of our art if we do not try.

SW: When there are multiple authors, like 14 in this case, they will not all agree. I asked Dr. Steven Lockley and others, and they have said that although we do not know everything, we know enough to do decent jobs now. Plus most existing lighting, especially if in the 3000-3500K-CCT range - is not that good. So even an imperfect human-centric lighting project is probably better than existing lighting.

Q: There is a huge difference between daylight, which has a fairly flat curve and a true color temperature, compared to light sources with specific blue peaks and discontinuous spectral power distribution that meets a specific correlated color temperature. Is it really appropriate to compare blue-peaky sources with daylight?

IA: Actually, the daylight spectrum is quite spiky. It is the CIE mathematical model of the daylight spectrum that is smooth. In answer to the question, however, it entirely depends on what the comparison criteria are. Our ability to discriminate colors is based on our ability to separate signal differences from three very wideband color sensors (i.e., our cones). As a result, we cannot tell the difference between a smooth daylight spectrum and white light generated by three laser lines at 450, 525, and 625 nm.

Q: Understanding that blue light supports our circadian cycle through stimulating the iPRGCs (intrinsically photosensitive retinal ganglion cells) leading to melatonin suppression, what is the dosing required for minimum stimulation? Likewise, at what exposure level are the iPRGCs bleached where no increased exposure can increase melatonin suppression?

IA: There are lighting industry recommendations on this issue (e.g., DIN SPEC 67600:2013, "Biologically Effective Illumination - Design Guides").

SW: People are different. For example, people in the Asia-Pacific rim are used to 6500K lighting even before going to bed and have adjusted to be able to sleep.

Q:To understand dose, you need proper measurement and agreed-upon definitions. Are you aware of efforts to address these missing items?

IA: Yes. DIN SPEC 5013-100:2014, currently in the publication process, addresses the issue of melanopic illuminance.

SW: Just like different people need to take different amounts of medical pills at different times of the day, light dosing can vary depending on where people live (close or far away from the equator), their age, if they are morning doves or night owls, work during the day or night, etc. Some non-invasive physiological testing procedures and electronic devices are being developed that could automatically change the lighting or inform a person how he or she should adjust the lighting.

Q: How would you reply to concerns about installing cool LEDs in your home if you are supposed to avoid blue light before bedtime?

IA: This could be addressed by not using high-CCT lamps in bedrooms, or with color-tunable lamps.

SW: High-CCT lamps should be avoided in bathrooms before going to bed or in the case of having to go the bathroom during the night.

Q: It is my understanding, however, that blue light is damaging to the eyes of newborns. Do you agree with this proviso?

IA: There are population groups that are known to be especially sensitive to excessive blue-light exposure, including:

• Children (because of the transparency of their crystalline lens) and both aphakics (with no crystalline lens) and pseudophakics (with artificial crystalline lenses) who consequently either cannot or can only insufficiently filter short wavelengths (particularly blue light).

• Population groups that are already light-sensitive: patients suffering from certain eye and skin diseases, patients taking photosensitizing substances, etc., for whom blue light may aggravate their condition.

• Population groups highly exposed to LEDs (certain categories of workers: those installing lighting systems, theater and film industry professionals, etc.) that are subjected to high-intensity lighting, and are therefore likely to be exposed to large quantities of blue light.

These must be put into context, however, in that exposure to daylight delivers perhaps 100 times as much blue light as any interior electric lighting.

Q: Why would anyone encourage the use of high-CCT lighting for residential applications, assuming someone works during daytime hours? Following nature, it would seem that there is a dramatic reduction or elimination of blue light as the sun sets.

IA: We use electric lighting on cloudy days while we are working.

SW: If somebody works or studies at home in the evening, high-CCT lighting can be better than drinking caffeine.

Q: How do you see the market responding to color-tunable systems that vary CCT dynamically throughout the day?

IA: Color-changing lighting systems that mimic daylighting are logical.

SW: If we lived in a perfect world, going to bed and getting up based on the sun, having electric light mimic daylight would be good. But since we do not live in a perfect world, it may be better to have electric light to be more of a drug, providing the best intensity and spectrum for specific people for various tasks at different times of the day.

Q: Have you taken into account the physiology of the human eye in selecting indoor CCTs? Because of the change in size of the pupil and light scattering, I have found that 5500K outdoors is equivalent to around 4200K indoors.

SW: It is really difficult comparing CCT during day and night. At least during the day based on spectrally enhanced lighting, 5000K dimmed 10-15% is equivalent to 4000K.

Q: How do you feel about mixing CCT levels within an architectural lighting environment - for example, 4000K and 2700K? Where do you feel the ratios, difference, or contrasting CCT levels become undesirable? What about higher/lower illuminance levels?

IA: This is no different than mixing daylight and electric lighting. It therefore becomes the lighting designer's choice.

SW: As mentioned during the webcast, RGB+ systems can provide substantial 460-490-nm light while still being a relatively warm white.

Q: If you have the 460-490-nm blues alongside reds/ambers/whites, do you still get all the benefits of the blue, with the more appealing (to some) warmer white appearance?

IA: The melanopic sensitivity function is similar to the V-lambda photopic sensitivity function, although it peaks at 490 nm rather than 555 nm. We are insensitive to light, in terms of melanopic response, beyond about 550 nm (green).

Q: Having established that CCT does not seem be relevant within the discussed context, how does the underlying spectral power distribution (SPD) affect this? Would you require a minimum CRI, and what would that CRI be?

IA: Davis and Ginthner (1990) stated, "The effect of color rendering, which may be of more importance than correlated color temperature, could be studied by repeating many of the experimental procedures used here, using lamp types of equal color temperature but different color rendering properties." We are not aware of any such studies having been conducted in the intervening 25 years. Even if they were to be, however, there would remain two issues. Color rendering and color appearance metrics (two separate issues) are still contentious topics, although it is widely acknowledged that the CRI metric is a poor indicator of LED lighting performance. Even if a link between color rendering metrics and preferred illumination levels were established, there will likely remain a considerable gap between what is best and what is commercially available.

SW: CRI by itself is not adequate. At least R9 should be added, and often R10-R14 or R15 should also be included. SPDs are also very good. In many applications, I am not sure if 90 CRI is really needed compared to 80 CRI.

Q: When you speak about blue light, you mostly refer to 460-490-nm light. Should we be concerned about commercially-available LED lights that have substantial power in the visible violet range below 450 nm?

IA: These products are usually designed such that the violet radiation (typically 415 nm) is almost entirely absorbed by the phosphors and emitted at longer wavelengths. It would not make sense to do otherwise, as we are almost totally insensitive to such short-wavelength radiation.

Q: Is Kruithof not consistent with the approach in Smet and Ryckert's paper on "A Memory Colour Quality Metric for White Light Sources," Energy and Buildings (2012, 49:216-225)?

IA: Thank you, we were not aware of this paper: It looks interesting in that it appears to explain Kruithof's results.

Q: Warm CCTs are preferred for indoor lighting in countries of cold/temperate climates such as in Europe and North America. However, in warm countries, like in India, the preferred CCT for indoor lighting is cool daylight - 6500K. Most offices in India make do with much lower levels of lighting of 200 lx. In Western countries, which prefer warm 3500K-CCT indoor lighting, most require a minimum of 500 lx or higher. Is the higher requirement in Western countries a result of the lower-CCT lighting, which requires higher light levels to support proper vision?

IA: We did a review of this topic, and have concluded that for the most part it is an urban myth. According to our discussion with lighting professionals from South America and the Far East (including India), the preference for cool-white fluorescent lamps is historical, and based mostly on the fact that cool-white fluorescent lamp phosphors were more efficient than warm-white phosphors. Thus, it was mostly a matter of consumer economics rather than consumer color preferences. As for the different light-level requirements, it is instructive to note that the "right to light" laws in the United Kingdom specify light levels only sufficient to read a newspaper by - some ten times lower than what we would consider to be adequate today. These laws are still in effect today, a good indication that regulatory requirements do not always coincide with good lighting design.

SW: Based on spectrally enhanced lighting and TM-24-13, high-CCT lighting with at least 200 lx is almost equal to 3500K and 500 lx for visual acuity.

Q: I think the earliest research on the blue-light hazard relates to United States military experience for observers staring into the sky close to the sun, looking for aircraft. It would be great to unearth that information! My question is what do you think of general lighting products that are released without hazard labeling as a Risk Group 1 "Low Risk" products? There is a concern that these lighting products are not totally eyesafe.

IA: We would not call this a "blue-light hazard," as the most likely damage mechanism from full-spectrum daylight would be photothermal effects. A relevant reference would be R. Penner and J.N. McNair, "Eclipse Blindness," American Journal of Ophthalmology, 61:1452.

Q: Studies recently published and noted on a LinkedIn lighting discussion group explore retinal cell necrosis with exposure to blue versus green versus red light. Studies are done in vitro, and only looking at LED emission, not blue light from other sources. I have not vetted them using all of the metrics you are talking about now. Have you heard of these research studies?

• http://www.ncbi.nlm.nih.gov/pubmed/23898883

• http://www.ncbi.nlm.nih.gov/pubmed/22989198

• http://www.ncbi.nlm.nih.gov/pubmed/22364266

IA: We were aware of the first two papers. What all three papers have in common is that the studies were conducted in vitro with cultured cells. This is a completely different environment than in vivo studies, where there are many cellular repair mechanisms to repair damage. The second paper was noteworthy in that the topic was covered in the popular press, with the lurid headlines that LED Christmas lights pose a horrendous health hazard. Of course, there is no support whatsoever for this in the original paper.

Q: Is there an industry-wide accepted definition of "blue-enriched light?"

IA: No, it is merely a generic term, similar to warm or cool white light.

SW: I am trying to find or develop a definition for blue-enriched light.

IAN ASHDOWN is president of byHeart Consultants Ltd. andSTAN WALERCZYK is a principal at Lighting Wizards.