The result is the blue line in the chart below.
I knew there’d be some variation, but I wanted to figure out what is causing that. My attention was drawn to the sharp peak (73.0) at June 19th and so I looked back into my calendar to see what I was doing around that day.
Turns out, I woke up really early for some reason (might have been related to drinking a glass of alcohol the night before, but I haven’t kept track of that). So I looked at another peak on June 2nd. Turned out I had a doctor’s appointment early that morning.
I work from home and keep track of my working hours. I usually start work about 30 after I wake up, so these hours a pretty good indicator of wake up time.
I decided to plot wake up times that I could reliably reconstruct from my calendar. The result is the red line.
The earliest was 6:30 am, my usual and preferred time is 8:30 am and the latest recorded time was 10:30 am. I got up even later on some non-working days, but did not record them. I sleep about 9.5 hours a night on average and I almost always go to bed between 23:30 and 00:30.
Without doing any proper statistical analyses I think it’s safe to say that wake up time and weight are negatively correlated for me. My guess is that this is just water balance. I do drink a bit of water during the night, but not anywhere near as much as I drink during the day.
So here’s a trick. If you want to convince yourself that you are losing weight, simply get up 3 minutes earlier every day. Translated to a get-rich-quick-scheme: write a diet book that provides the reader with an excuse to get up earlier every day. Guaranteed result.
Is there anything useful that can be derived from this data? It tells me something about how my water balance changes between about 6:30 and 10:30. Any relevant parameters that can be derived from that knowledge?
Edit 3-7-2009 14:20: Here’s a scatter plot of wake up time versus weight. Google Spreadsheet finds a correlation of -0.6. It also finds a slope of -0.28 liters / hour.
Edit 3-7-2009 15:40: Armed with this knowledge I corrected my weight measurements for wake up time. The blue line shows the original measurements. The red line shows what they (probably) would have been at 8:30 in the morning. This leads to a reduced spread in daily weights, but it’s clear there are other factors involved here. I suspect bed time, temperature and daily water and food consumption play an important role too. I can’t be bothered to monitor all of those.
The worrying bit is that this plot has a positive slope of 0.65 kg / month (0.77 without time correction), which is about the opposite of what I have in mind… That said, if all goes according to plan, I’ll be upgrading from an average of 10 minutes of exercise a day in June to about 30 from August onwards. That should do the trick (I’ll be cycling at 25-30 km/h though, not walking 6 km/h). There’s plenty of room for diet improvements (-250 kcal / day) and experiments too and lower temperatures are also great for burning calories. Finally, weight is not a very good metric of health in the first place.