Food, vitamins, nutrients, measurements

Wow, two months since my last post, huh? What a nice break. Enjoyed the good weather, saw a concert or two, explored my home country, worked in the garden, tried to live a healthier life the geek way :)

When last I mentioned I had quit smoking and tried to get some more exercise, I was using a spreadsheet to record what I stuffed into my mouth and made a stab at guessing how much calories, proteins, carbs, and fats were in there. This went reasonably well, but after I found the USDA food database I became interested in tracking more data, like amount of vitamins/minerals gained from foods vs. the dietary reference intake (DRI). Trying to do this in a spreadsheet by hand would be bothersome.

So I was very happy to find cron-o-matic, a free Java program which uses the USDA food database and which will give you clear feedback on how you are doing nutrient-wise and which allows you to plan ahead as well.

The program is very friendly but takes a bit of fiddling to set up - one of the java methods it uses in setting up the basic dietary requirements is deprecated and replaced, so to get optimal use you'll have to run an older JVM.

Once it is set up it is really easy to use, there are shortcut keys for nearly everything, many measurements (teaspoon, cup, stalk etc) already preprogrammed where appropriate, it is dead easy to combine ingredients and so make custom 'recipes' for items you frequently consume. It will even create nifty little nutrient reports for any given period of time.

Of course, you have to recognise and understand the limits of software like this. For instance, it is known that vitamine D is heavily underreported in the database. In fact, there are a couple of uncertainty factors:
a) database errors (under/overreporting, typos)
b) natural differences in foodstuffs (dependent on growing conditions, available nutrition, genetics, storage time, etc)
d) measuring errors (yours truly)

However, you know you are on the right path when your weight gain/loss corresponds with your self-reported energy intake:
Blue line/left axis is weight in kg, red line/right axis is energy intake in kcal. As you can see a spike in energy leads to a spike in weight one or two days later. The relationship is not 1:1 as other factors (like physical activity, water intake, etc.) weigh in as well. Those huge energy peaks? Social events, BBQs, parties. Even moderation should be done in moderation. If you can't have a little fun once in a while, what's the point?

But waaaaait a minute, I hear you say, weren't you trying to lose weight? I don't really see that in the graph! True, this 2-month graph shows very little weight loss. I want this weight loss to be sustainable. If I starve my body and then start eating normally again I expect a jo-jo effect. And I want to get all the proper vitamins and minerals, without using supplements, so my daily energy intake is very near to 'normal intake'. As I approach my 'ideal weight' I expect it to go even slower. At the same time I am also working out so I am replacing fatty tissue with muscle. But perhaps this chart showing progress since January works better:
Yeah, that is more like it! (the absence of measurements between ~12/2 and 12/3 are due to lots of business travel and very little opportunity to weigh). Also, I can't believe I had let my weight slip to 98 kg or so.

All this data you are gathering in the CRON-o-meter software is saved as XML. You can grab this data and plug it into Excel or other spreadsheet software to see how you are doing in the long-term:
I've used conditional formatting to color it by how close to or over the RDI/target I was - light green is under, dark green is in range, yellow is too much. As you can see I'm hitting most targets most of the time in the vitamins and mineral sector, without any supplements. Since I had this spreadsheet and was tracking daily data anyway, I also started tracking other things, ranging from hours of sleep to subjective mood during a day to far more, um, intimate things. Because as it turns out, tracking stuff is fun! (for geeks like me, I suppose).

For instance, check this chart highlighting the relationship between hours of sleep and subjective mood:
Fun, what? Of course this chart is a little funky in that (a) I tend to sleep in more during weekends and (b) I generally feel happier in weekends as I don't have to contend with all that work malarky. But still there is a clear correlation. More sleep means less murder.