Hacking Sleep Cycles: My Perfect Bedtime Experiment
On his Insomnia blog, famous Sleep Doctor Michael Breus describes simple method that you can use to “hack” your sleep cycles and find the optimal time to go to bed. According to this method, you change your bedtime every 3-4 days in 15-minutes increments, until you can wake up without alarm most of the time. This January, I put together a little self-experiment, with the objective to test this method, and hopefully, find my perfect bedtime. Between January 7th and February 1st, I was going to bed at 11:00, 11:15, 11:30 and 11:45 (5 consecutive nights for each time slot), and waking up at 7 am every morning. At the same time, I was using Zeo, my rTracker log, and a couple of apps to track my sleep and some other data. While results of my experiment were not exactly perfect, I believe it still helped me to identify my optimal bedtime slot.
To identify the optimal bedtime, I tracked the following data points for each night and the following morning:
- whether I woke up without alarm that morning (5-15 minutes before 7 am)
- my subjective sleep quality (measured using 10-point scale, average score on “How long did I sleep” and “How well did I sleep” questions)
- my subjective “wake” score (measured using 10-point scale, “How easy was it to get out of bed” question)
- total time slept (measured using Zeo)
- total REM sleep (measured using Zeo)
- total deep sleep (measured using Zeo)
- my subjective physical and mental energy scores (measured using 10-point scale)
- reaction time, memory and executive cognitive functions (measured using Mind Metrics and Stroop Effect apps)
At the end of the month, I computed weekly medians for each of these metrics (in Excel) across all four weeks, and visualized them using Sparklines. As you can see, my subjective sleep scores were pretty much the same across all four weeks, however, getting out of bed was much harder during the week when I was going to sleep at 11:45. If I went to sleep at 11:15 or 11:45, I was more likely to wake up without alarm the next morning:
According to Zeo, going to bed between 11:00 and 11:30 resulted in longer sleep (20-22 minutes more), compared to 11:45 bedtime. The same was true for REM sleep:
In terms of subjective assessment of energy, going to bed at 11:45 would result in me feeling more tired, both physically and mentally than going to bed before or at 11:30:
Every morning, around 7:25-7:30 am (it is recommended to wait at least 20-30 minutes after waking up in order to avoid effects of sleep inertia) , I also measured my cognitive performance, using Mind Metrics and Stroop Effect apps. As you can see, my reaction time (PVT test) and memory (3-back test) were somewhat better in the morning following bedtime before 11:30. The Stroop Effect test did not show any notable differences in cognitive executive functions:
Based on the results of this experiment, I conclude that my optimal bedtime is between 11:00 and 11:30 pm. The 11:15 looks especially promising, as I am more likely to wake up without the help of alarm on the following morning. Perhaps, later this year, I will replicate this experiment to see if anything changes. But in the meantime, please no phone calls after 11:15 pm.