Marks Daily Apple
Serving up health and fitness insights (daily, of course) with a side of irreverence.
21 Jun

How to Conduct a Personal Experiment: Yellow Light Exposure (plus an Announcement)

You guys ready for another personal experiment? I hope so. Even if don’t think you’re ready to take something on, I’m confident you’ll be able to handle this one, because it’s relatively simple, intuitive, and easy. It’s also something I’ve been discussing for a couple years now, so you’re most likely familiar and comfortable with the concept. But most importantly, today’s experiment is a gentle one that requires very little commitment. No jumping in freezing cold water, no drastic changes to your sleeping schedule. All I’m asking you to do is experiment with nighttime yellow light exposure.


Remember how I wrote about nighttime exposure to blue light affecting melatonin secretion and, subsequently, sleep quality and duration way back when? Yeah, that. In case you didn’t read it, I’ll give a quick explanation:

The color (or wavelength) of the light we perceive entrains, or “sets,” our biological clocks, also known as circadian rhythms. And in the natural environment, with its reliably consistent lighting schedule, it works pretty well. During the day, we see all the visible wavelengths provided by the sun, including violet, blue, green, yellow, orange, and red light, and this “tells” our bodies that it’s daytime, that’s it’s time to be active, to hunt, to gather, to build, to work, to exercise, to study, and so on. Secretion of melatonin, the “sleepy hormone,” is blunted. At night, when the only visible light is historically the longer wavelengths, the yellows, oranges, and reds which we create through campfires, or candles, or gas lamps, melatonin secretion is unaffected. We get sleepy like we should, when we should. All is well. And if we had evolved to be nocturnal, like rats, all that low-wavelength blue light exposure available during the day would let us know it’s time to sleep.

But we don’t use candles and oil lamps at night anymore, do we? We use white LED lights (blue light) and stare into laptop (blue light) and high-def TV (blue light) screens. We use our iPhones (blue light) or Androids (blue light) in bed, even waking up in the middle of the night just to check our email (blue light) because “why not, we’re up anyway!” To really get a sense of this, next time you take a nighttime stroll around your neighborhood, pay attention to the living rooms of the houses you walk past. If they’re got their plasma or LCD going, the lights off, and it’s dark out, the entire room will be bathed in an overpowering blue light. It’ll look like a scene from an alien abduction movie or something. Of course, whether the room lights are on or off, that blue light from the screen is still there, beaming directly into the eyes of those present and affecting the secretion of their melatonin.

And we wonder why we have so much trouble getting good sleep.

It’s not just sleep that’s affected (although that’s enough reason to take heed). Disturbing our circadian rhythms with improper light exposure may have a range of other health effects, including, but not limited to:

When I say “yellow light exposure,” what I’m really talking about is “blue light avoidance.” Today, I’m going to show you how to put together an experiment to test the effects of exposing yourself to yellow/orange/red light and avoiding blue light. Although that sounds like we’re testing two things, we’re really not, since yellow light has little to no effect on our melatonin production. For all intents and purposes, it and other, higher wavelength lights are neutral, while blue light is antagonistic to our circadian rhythms.

Okay, so how do I do it?

First, choose a goal that yellow light exposure and blue light avoidance might help make possible. Since we can’t really test our susceptibility to cancer in a short trial, nor does metabolic syndrome develop in mere weeks, let’s test the effect of yellow light exposure (and blue light avoidance) on some aspect of our sleep.

  • “I want to feel more refreshed in the morning.”
  • “I want to go to bed by 10 PM every night.”
  • “I want to have more melatonin in my morning urine” would be an effective way to test, but it also requires being able to test your urine for melatonin. Most people don’t have that on hand. Let’s go with the second one – getting sleepy and in bed by 10 every night – since that’s easy to quantify (did you go to bed at 10?). If you already go to bed at 10 PM, choose a time that’s earlier than your normal bedtime. The key is to find out if you get sleepy earlier without the blue light.

Come up with a hypothesis, such as:

  • “Since blue light exposure suppresses normal melatonin secretion, and melatonin makes us sleepy, exposing myself to yellow light and avoiding blue light after dark will help me get sleepy and go to bed by 10 PM.”

Next, identify some of the variables that could affect the results of your experiment:

  • Activity at night – Does what you do while avoiding blue light affect your sleepiness? Is reading more stimulating than hanging out with your spouse?
  • Electronics usage – Assuming you’ve taken steps to eliminate blue light exposure (wearing blue light-blocking goggles, installing f.lux on your computers), do electronics still stimulate you and keep you awake?
  • Yellow light source – Does a yellow light bulb act differently on you than candlelight?

Then, let’s take some measurements. What should you be measuring?

  • Bedtime – When did you get into bed?
  • Sleepytime – When did you start getting noticeably sleepy? Time of first yawn?
  • Sleeptime – When did you fall asleep? This is hard to measure without equipment, but you can probably approximate it. Personally, once I get too tired to read another page and find myself nodding off in the middle of a sentence, that’s my sleeptime, because I fall asleep as soon as I put the book down and turn off the (yellow) light.
  • Morning wakefulness – On a scale of 1-10, how awake and refreshed do you feel in the morning?

After a few weeks you should have enough data to start making some observations about what does and doesn’t work for you, and from there you can decide on what to test next, if anything. That’s it for this experiment. Have fun!

Thus concludes this short but sweet series on self-experimentation. What’d you think? I for one am a firm believer in the power of the self-experiment. In fact, I think it’s the ultimate arbiter of an individual’s ideal path to health. Sure, you could read all the blog posts and studies and papers and research in the world, but if you personally experienced results that completely contradicted the advice of the experts, what would you do? Would you continue down the path that supposedly worked for this cohort or that quadrant of some population somewhere? Or would you stick with what worked for you?

Exactly; we are complex beings with physiological processes that even the experts who study them for their entire lives don’t fully understand. Everyone is different, and there are no real one-size-fits-all plans – not any honest ones, at least – and so self-experimentation (even if it’s just an informal thing) is absolutely crucial and highly effective.

In case you missed the newsletter announcement this week: next Wednesday, my new book, The Primal Blueprint 90-Day Journal, is being released! As usual, there will be a very special early-bird offer. I’ll be throwing in some limited-time freebies, and doing a big day-of announcement, so check back here on June 27 and be ready to take advantage of the offer.

Now let’s hear from you. If you have any questions or comments about this self-experimentation series, leave them in the comment section and if there are enough questions, I’ll do a Dear Mark on self-experimentation next week. Thanks for reading!

You want comments? We got comments:

Imagine you’re George Clooney. Take a moment to admire your grooming and wit. Okay, now imagine someone walks up to you and asks, “What’s your name?” You say, “I’m George Clooney.” Or maybe you say, “I’m the Clooninator!” You don’t say “I’m George of George Clooney Sells Movies Blog” and you certainly don’t say, “I’m Clooney Weight Loss Plan”. So while spam is technically meat, it ain’t anywhere near Primal. Please nickname yourself something your friends would call you.

  1. I’ve been using F.lux for years on my laptop. I wish something similar was available for my iPhone!

    gilliebean wrote on June 22nd, 2012
  2. I love self experimentation and agree with mark that it is great for finding your way to health. I even like when everyone shares the outcomes of their experiments. But many people do an experiment and immediately tell everyone x is the way to do things because of it. They tend to forget that who knows you might be an outlier or there might be a confounding variable that you didn’t recognize. I would like to see more people have an awareness of the limits of their n=1.

    CMHFFEMT wrote on June 23rd, 2012
  3. Don’t forget the lampshades, peoples – my lampshades are a light-to-medium brown, and at night they cast a lovely subdued golden light – very relaxing.

    Also, since I keep my home very dark at night after bed, on the rare occasion I need to get up in the middle of the night, I use a little electric tea light candle (also called flameless candles) that produces an orange light. It’s just enough light to see by since my eyes are already adjusted to the pitch black. In my experience, they are much less bright even than candles.

    Angel wrote on June 23rd, 2012
  4. I am a big believer in personal experimentation, and the first of June I started a 30 day experiment to not use any light but candles unless it was an unwise situation, like midnight poopy diapers on my 6 month old. I use a flashlight so I can properly clean bums. Anyway, I was concerned that with the little sleep I DO get (6 month old still up every 2 hours) I can’t waste time between getting in bed and falling asleep. My mind would be busy even though I was exhausted. I started changing how I did everything because I’m avoiding artificial light as much as possible. No more vacuuming at 9:30pm because there is not enough light to see. Now most of my chores happen early in the day. Anyway, I do fall asleep faster now.

    Kari wrote on June 23rd, 2012
  5. This is do-able. I can put all nighttime bulbs and a filter in our bedrooms and have the blue lights (always my preference for a lightbulb…most likely an inherent outside time craving) in the rest of the house.
    And is it blue light spectrum on the skin — or received in the eyes — or both? I always found going outside in the morning and letting the sun shine on my bare back for instance, even with eyes closed, was energizing.

    Olivia wrote on June 27th, 2012
  6. Would using regular sunglasses have an effect on blue light intake?

    Joe wrote on July 2nd, 2012
  7. Do kindles emit blue light?

    Rachel wrote on July 21st, 2012
  8. Did this years ago. Look at for spectra of all sorts.
    He’s tested fLUX (screen still emits significant blue even set for very warm). He’s tested “warm white” bulbs (they emit blue). He’s tested the GE “Post Light Bug Light” — that’s a good one, almost completely blocks the blue emission with a yellow dome. And you can look for “Turtle Safe Lights” +amber +LED and find monochromatic amber sources, no blue at all.

    Distinguish color temperature; a warm LED or CFL still has a big spike of emission in the area below 500nm.


    Hank Roberts wrote on September 24th, 2012
  9. Use f.lux on your computers 😉 Save your eyes and sleep well. 😉

    sober wrote on September 25th, 2012
  10. You can buy removable & reusable iPhone, iPad, and tv filters on

    Kelly wrote on November 15th, 2012
  11. “Will Blue Light Affect Your Sleep? | Mark’s Daily Apple” was in fact a relatively good article, . I hope you keep authoring and I’ll try to continue to keep following!
    Thank you -Marianne wrote on January 2nd, 2013
  12. Aren’t there receptors in the skin as well? I remember there was some experiment involving something like that. I’ve heard a lecture about lab animals mentioning similar and heard Rob Wolf mention skin receptors as well.

    Moon light of the brightest fullest closest moon on a clear starry night may seem bright but measures around 1 lux(about a candle) it just that it is very diffused making it seem brighter. Also according to photo sites it measures 4000k (color temp) making it visually warmer than sunlight. so intensity and wavelength make a difference.

    If I remember correctly (maybe not) he Lab animal lecturerer mentioned light of about either 5watt or 5 lux would disturb lab animals circadian rhythms and lead to raised stress in animals. drats to my sieve like memory. It must be all the low quality sleep.

    dave wrote on January 11th, 2013
  13. I’d like to try a pair of Uvex Skypers but I already wear glasses and they look too snug. Would a pair of these work for me?

    The Skypers look a lot nicer, the other pair is not going to win points from the Mrs in the romance department!

    Arp wrote on February 28th, 2013
  14. HELP! I live in the uk and I need to find a descent priced pair of orange sunglasses. I am not paying 80+ dollars and having it shipped over here for something costing 2 pounds to make in china.

    GEORGE wrote on April 9th, 2013

      I bought the ~50% yellow and the 16% amber; the yellow does block blue wavelengths but allows much more total light through, so are quite unobtrusive. The amber ones are much darker, though you also get used to them pretty quickly. The non-fitover style fits the curves of your face pretty well to block out light from the sides, while the fitover style is less fitted but still pretty good.

      Ben wrote on April 9th, 2013
  15. If you’ve got a Mac, you can use Candlelight. It removes the blue component of your Mac’s display. For more information, go to:

    Oliver wrote on June 13th, 2013

    is a page of spectra from various sources. Craig at LEDMuseum has checked most of what’s been suggested.

    The “post light bug light” has only a very slight emission in the range of 400-500+ nanometers that interferes with sleep.

    The amber LEDs (“turtle safe” or hallowe’en LED lights) have no blue at all.

    But the “warm” lights emit a lot of blue; so does your computer screen even using “Flux” — because those are driven by a blue emitter with a phosphor over it.

    We started this same experiment at home about five years ago. It works very well –as long as you don’t fool yourself. Don’t use logic. Get a spectrograph or look up the spectrum being emitted.

    It helps to understand how fluorescents of all types including “white LEDs” emit blue to drive the light and a layer of phosphor to shift some of that blue to longer wavelengths that seem ‘warmer’ — and that the blue still comes through which is what you don’t want for this experiment)

    You can buy filter material to block blue from any theatrical supplier. Again look at the spectra being blocked.

    You can look this stuff up.

    Some places to start

    Hank Roberts wrote on January 22nd, 2014
  17. Aside — the first publication I saw that really explained this is Janet Raloff’s articles, found at
    part 1
    part 2

    Hank Roberts wrote on January 22nd, 2014
  18. Thank goodness the news is starting to report on this. I use the SleepShield iPad filters ( ) and they work great. I also put the new Good Night LED Sleep Lights in my bedside table ( ).

    Jeff wrote on January 28th, 2014
  19. I’m very aware of the effects of light during night and the sleep deprivation it causes, so the hour before going to bed I avoid tv, Pc, and turn on a very soft and warm yellow light.

    I think it’s advisable to use these features in a pc:
    – the high contrast yellow skin
    -the gooogle chrome add-on “change colors” I have already selected yellow for main text, orange por links and red for already visited links
    -and F. Lux, a free software that warms up your computer display at night, to match your indoor lighting by adjusting a computer display’s color temperature.

    andrenio wrote on April 8th, 2014
  20. Wonderful post however I was wanting to know if you could write a
    litte more on this topic? I’d be very grateful if you could elaborate a little bit further.
    Appreciate it!

    Derick wrote on May 9th, 2014
  21. So does a 1900K LED qualify as a good source of low-blue, yellow light?

    Sammy wrote on June 25th, 2014
  22. Hi,

    In addintion to F.lux software there is any other protection needed, for example any glasses or an screen for the monitor…


    Melvin wrote on August 4th, 2014
  23. I sleep early at 8 or lately 10 pm in winter. I want to sleep late. So if i use halogen bulbs does it affect my sleep ? Do i sleep late night with halogen bulbs while im studying or reading?

    celine d wrote on September 6th, 2014

Leave a Reply

If you'd like to add an avatar to all of your comments click here!

© 2016 Mark's Daily Apple

Subscribe to the Newsletter and Get a Free Copy
of Mark Sisson's Fitness eBook and more!