LED lighting badly suppresses human melatonin production at night

(Post author here, huh, it hit HN...)

This is not consistent with the studies I've read. It's a "total lumineous flux in the 460-490nm range" concern. The deep reds, up past 600nm, don't seem to have any impact on melatonin, nor do violets below ~420nm - though they can have other impacts on the retina and such.

Bright light will, all other things being equal, be worse than dim light, but "dim blue light" is still really bad for melatonin levels and sleep.

Some of the older studies did find an impact up in the greens, but they were for fairly short duration exposures - 30 minutes or so. The impact of the longer wavelengths, over a longer term, seems to be muted by [mechanisms somewhere in the system], leaving the blues with the most impact.

You're welcome to trust Huberman, but I'd encourage you to read the papers yourself (I've read dozens and simply picked the most relevant here). There's quite a bit of work on this matter from the past 20-30 years, and they all seem to conclude that "blues matter." Greens and above don't have nearly the same impact, and they don't have the longer term impact. I'll also suggest being careful about conclusions drawn purely from the absorption spectrum of the molecules involved, as the body doesn't seem to react directly to them, but to some skewed version. Unit tests are nice, but end to end tests cover all the details in the process.

>find "r9" >no results >find "cri" >no results in post

:/. This is definitely an important and I think unappreciated issue right now, and I'm normally all for long form investigations too. But I don't think this piece structured it very well. It started losing me right at the beginning with what felt like an outdated description of technology applied too broadly, and then not actually bothering to cover the very real improvements changing things until the very end, and even then putting off concrete coverage of who deserves kudos until later. It goes

>They’re internally blue LEDs with a phosphor coating that shifts some of the energy from blue over to the rest of the spectrum - the reds, yellows, and greens.

>They look white to the eye, but they’re actually three separate peaks of light, and they don’t do a very good job rendering colors.

But this seems unnecessarily hand-wavey. We have actual specific measures for this in the form of CRI (color rendering index, 0-100 with natural outdoor light being 100) and R9 (a measure of strong red rendering, which most people tend to like, incandescents tend to be in the 90s). It's definitely true that lots of common cheap LED bulbs have poor CRI and R9, and even these aren't necessarily linked either. But it's also perfectly possible to get decent bulbs that do both ok or focus well on things like helping sleep. The plain old Ikea Tradfri E27 White (like $12) has a CRI of 90 and an R9 of 91. There is a specific sleep focused "bedtimebulb" that has a CRI all the way up at 95 and a respectable R9 in the mid-80s. The Ketra line of lighting gets specific about this as well. This isn't some impossible problem.

Which the article does kind of acknowledge, but only at the very end. If more people knew to look for and ask for these measures, then the market could better respond accordingly. It could be worked into various lighting standards as well for work and government. But "LED lights and screens are killing you" from the beginning doesn't strike me as a good approach.

Basically, the lede should have been the "future reviews" bit and the author should have waited on doing them first because otherwise it's just defeatist. Give people something that is actually actionable, not a "well you're screwed and that's that" because nobody is giving up their screens. Structure should be "these things matter and are in fact being done with more to come, but most don't know to look for these measures and they are far too hard to find, this is something that action should be taken on." Then get into the nitty gritty of why. The goal should be something people can take with them, like at a schoolboard meeting parents being able to ask about the CRI/R9 and general quality of their children's lighting.

There are also probably very personal and biological variables involved too that the post doesn't mention. I stare into monitors and phones all day long. Every light in the house is full-blown LED. I can't help but fall asleep at 20:00 in an led lit room watching led TVs, while my led phone drops onto the floor after I've fallen asleep. I've been in an extreme LED environment for over a decade and never have problems waking up at 6:30 and falling asleep in bed at 22:00. I totally agree with concern raised by the article, but I also think it lacks the whole picture of someone's lifestyle. People have struggled with sleep issues well before LEDs.

I am a big fan of LEDs but the strobe effect is maddening and it only gets worse as they are dimmed. AC power crosses zero volts with every cycle so you need a driver circuit that can compensate. Most don't do a good job and I have replaced bulbs and dimmers multiple times to come up with something acceptable.

At the family camp, I retired my kerosene and propane mantle lamps in favor of solar battery powered LEDs many years ago. It was a matter of safety and convenience at the time, especially with children everywhere. The LEDs are powered directly from DC and I have no problem at all with them.

https://imgur.com/gallery/bqS6uSV

https://imgur.com/gallery/YEa3V63

I would indeed appreciate you reading the whole thing... I did write it to be read, and while it's long, I promise I don't blather on pointlessly. At least by my standards.

But I agree with you. I think we absolutely can make multi-emitter LED bulbs that reproduce the spectrum of incandescent bulbs, being low in the blue, etc. I know people who have looked into exactly that sort of thing, have researched relevant LEDs, good dimmers, etc. They run into the problem that to do something like that, you're looking at a BOM north of $100 (implying a far higher retail price), which is rather at odds with "anyone actually buying the thing."

So I have to ask, "Why are you sure?" I've explained how our current white LEDs work (blue pumped is the most efficient, violet pumped is somewhat better but less efficient), and the work involved in blue and white LEDs literally led to a Nobel Prize (https://www.nobelprize.org/prizes/physics/2014/press-release...).

Physics doesn't have to grant us what we want, and if your goal is "cheap white LEDs that reproduce incandescent spectrums," nobody, so far, has come up with such a thing.