Rajiv Mehta asked some questions about using one-legged standing to improve sleep. I do three sets of two (left leg, right leg) each day.

Q. How do you spread out your three sets (have you found some minimum time between sets, say 3 hours)?

A. I make sure there’s at least 4 hours between sets. The effect was weaker with only 2 hours between sets. The time of day doesn’t matter but for convenience I usually do one set in the morning, another set in the afternoon, and a third set in the evening.

Q. They say exercise before bed is not a good idea. Do you make sure your last set is at least X hours before bed?

A. No. If anything this particular exercise will make you more sleepy, not less.

Using the Tonic app for this.

Christine Peterson’s poster of her Zeo research was one of the highlights of the QS conference for me, as I said. Here’s why.

The correlation between Sleep Stealer score and time awake. When her Sleep Stealer score was 5 or less, she was awake about an hour during the night. When her score was more than 5, she was awake about two hours — a big difference. There should be a big difference, but you could fail to see it for a thousand reasons. The large difference is a validation of the whole thing — above all, an indication that her Zeo is working correctly.

Even when her Sleep Stealer score is low, she is awake a long time. This means there are major determinants of sleep depth not captured by the Sleep Stealer score. With the right Sleep Stealer score — assuming the correlation reflects cause and effect — she can improve from two hours to one hour (one hour difference) but that leaves one hour. This implies that the determinants of time awake not in the Sleep Stealer score are just as important as those that it contains.

Even when she is at the best level of important factors, she is awake a long time. When she had no drinks, she was awake 56 minutes/night. When other people didn’t disrupt her sleep at all, she was awake 54 minutes/night.

The average wake time for women 50-59 is half an hour. That’s a lot of lost time, day after day, night after night. Note however that the data is from Zeo users, who may have worse sleep than average.

It only took three months to collect the data. This isn’t on the poster. Yet this is a solid contribution, in the sense that I learned from it. With perhaps nine months of data and better data analysis, it might be publishable. The main point such a paper would presumably make is that even when you do everything right (Sleep Stealer score = 0) you’re still awake a lot. This point is nowhere in the sleep literature.

Christine, if you would like to sleep better I suggest:

1. Don’t eat breakfast until at least three hours after you wake up.
2. Get at least one hour of sunlight early in the morning — e.g., 6 to 7 am. You can do this by working outside. (I work outside several hours every morning.)
3. Stand on one leg to exhaustion four or more times per day. (I do it six times/day.) You can do this while reading — it should not reduce your free time.

I am at the First Quantified Self Conference in Mountain View. The attendees are much more relaxed and cheerful than at the academic conferences I’ve attended, presumably because they chose to come. Some are from Europe. My overall take is that the conference’s theme is the romance of tracking, in the sense that the typical presentation is something like: isn’t it wonderful that I’m measuring this? Or hypothetical. (Of course, the research presented at typical academic conferences is almost never shown to have practical value.) I think this is entirely reasonable. In my experience, it is very hard to learn something clearly useful and takes a long time. For example, I measured my sleep for about 10 years before figuring out how to improve it.

Sean Ahren‘s presentation was one of the best I heard, and illustrated the difficulty. He has Crohn’s Disease. He wondered if hookworms would help. Day by day, he measured how much pain he felt, and for some of the time took hookworms. There was no clear difference between the two periods (with and without hookworms). He learned plenty of useful stuff — how easy/difficult it was to do the measurements, what the data look like, the apparent ineffectiveness of one brand of hookworms – but when contrasted with the goal of learning how to reduce pain from Crohn’s, it doesn’t seem like much. Perhaps the average Crohn’s sufferer would say it’s great you’re doing this but think how does this help me? I think his observations lasted about 8 months. Perhaps if he continues for 6 years, by then the amount of learning will be larger and more tangible. Overall it’s a good example of the way scientific progress and job don’t mix well. When you have a job, you make tangible progress quickly: you fill someone’s order, for example. They wanted something, you gave it to them. Tangible. Whereas trying to clearly improve one’s Crohn’s Disease might take ten years. Too long if your motivation is connected to making a living. Too long for professional scientists.

At a breakout session on sleep experiments, I learned that someone had great success wearing blue-blocking glasses (which look orange) after 9 pm. Something I want to try. I’ve heard about these glasses before but these results were especially impressive. The glasses quickly reduced how long it took him to fall asleep. Someone else was told he had sleep apnea. But when his acid reflux got better, so did his sleep.

You can read about many talks, including mine, in great detail at Ethan Zuckerman’s blog.