In my last post I mentioned that my sleep has improved tremendously over the past year, and this has been a significant boost in my happiness and productivity. In particular, I used to need about 10 hours of sleep to feel fully rested, and now I only need about 7.5 hours. Currently I sleep an average of 7.5 hours a night with about 45% REM/deep sleep and 55% light sleep.
Things that worked really well:
- Waking up at the same time every day: I wake up at about 7 AM nearly every day, including weekends and holidays, regardless of when I went to sleep or even which state I’m in. Conversely, I go to sleep whenever I’m tired, whether that’s at 2 AM, 6 PM or (more frequently) 11 PM. This means I rarely have to worry about an alarm clock, can wake up with the sun, and can adjust whenever I’m tired: it’s much easier for me to adjust when I go to sleep than when I wake up.
- Setting the temperature of my apartment to 74 degrees: when it’s too hot I get dehydrated and my sleep quality drops significantly (from about 50% light sleep to about 80%!) When it’s too cold I simply have trouble falling or staying asleep. The range seems to be very narrow, from my experiments 70 degrees is clearly in “too cold” territory.
- Using a neti pot: without intervention, my nose tends to get stuffed about every two weeks. Using a neti pot lets me breathe clearly and improves the quality of my sleep (as well as significantly improving the quality of my waking life.)
- Eating right before bed: most sleeping advise tells you to eat2-4 hours before going to sleep. This didn’t work for me; I tend to have a hard time eating enough and would always wake up too early. The gain from making sure I’m don’t wake up due to hunger seems to be significantly larger than any loss from eating so late.
- Not doing any work in bed: as somewhat of an accident, over the last year I’ve pretty much only used my bed for sleeping. This has given me a significant ability: I can fall asleep immediately once I lie down.
Things that worked poorly or didn’t seem to matter:
- Naps: I have tended to take naps throughout the day whenever I feel tired or mentally worn out. However, my Zeo tells me that I’m getting almost no REM or deep sleep from naps-in other words, I’m gaining nothing from actually being asleep. This was a major surprise as I’ve always felt better after naps, but it may just be a placebo effect. I successfully induced the same effect in ~5 minutes through self-hypnosis, so it does seem to be pure placebo.
- Caffeine: surprisingly irrelevant. Caffeine does not seem to have a significant impact on my sleep even if I consume coffee an hour before bed. Similarly, I’ve done weeks entirely without coffee and it hasn’t seemed to have a positive or negative impact on my sleep.
- Being on the computer right until I sleep: conventional advice indicates you should turn off any blue-light emitting devices an hour before you sleep, but this doesn’t seem to make a difference for me either way-it could be because I already use flux.
- What I eat: quantity of food seems to be important, but the type of food doesn’t seem to matter. The balance of carbs, fats, or proteins I eat before sleeping doesn’t seem to have an impact, even two slices of cake seems to satisfy the need for food.
Things I’m trying going forward:
- Melatonin: several people report significantly better sleep using small doses of melatonin, e.g. Gwern reports shaving roughly 50 minutes off his sleep schedule while remaining equally well-rested. Note that doses in pharmacies tend to be overly large, the recommended dose seems to be about .5 mg, and pharmacies typically sell 3-5 mg, or about 6-10x the recommended amount.
- Dual n-back tracking: the major worry of sleep experiments is degradation of mental performance, when you’re sleep-deprived you can’t even tell how stupidly you’re acting. So it’s important to have an objective measurement, which is why I’m tracking performance on the dual n-back game. I don’t have any useful data so far because I have yet to plateau, but once I hit a plateau experiments should tell me what is and isn’t affecting my brain.
- That’s it: unfortunately my Zeo headband seems to be dying, so I won’t be able to measure small changes in my sleep quality until someone comes out with a replacement. I’ve already tested most of the things that are supposed to have a large impact on your sleep quality (except melatonin), so trying to measure anything else would most likely cause me to be tricked by statistical artifacts-therefore I’ll stop.