Wednesday, January 9, 2013

Health Geekery: Do It Yourself Testing

During a recent talk I gave in New York City at the interesting software company Cyrus Innovation, I mentioned that self-tracking and -testing will be disruptive of conventional health care. It already is. Doctors will have to get used to patients coming in with their own generated data, and they'll have to adapt, be open to, and embrace this development. This is all a movement in the right direction. Self-testing has a demystifying effect on your own attitude. With more and varied data comes more valid interpretations.

Times they are achangin' in the health field; see PrimalDocs for one.

It simply makes sense from a philosophical and practical standpoint. It's your body, right? And what doctor has the time and inclination to properly test various biomarkers over time, and within the proper contexts?

What to test?

You could obviously test a lot of biomarkers or indicators of your fitness and well-being. Three of the easiest and cheapest are:
  • Resting heart rate (RHR): How many times your heart beats in 60 seconds, at rest.
  • Weight and body composition (body fat percentage)
  • Blood pressure (BP)

The initial investment for testing equipment might seem a little pricey, but pales in comparison with a typical doctor's visit, and all that prescription medication. A few hundred dollars at most (and less if you borrow, share, or buy the equipment used), will give you a:
  • Blood pressure cuff with a digital reading, including BP and RHR
  • A decent body composition scale
For the testing I'm doing today, I use the following devices: A Withings BC scale, which gives me weight and body-fat percentage; a pulse oximeter (PO) left over from my RestWise account, giving me my pulse and oxygen saturation; and a Homedics digital blood pressure cuff.


You should test several times a day under different contexts (e.g., after waking up; during fast; after eating; after exercise; before bed, etc.). These devices are not infallible. One reading does not cut it. You will get different readings throughout the day as your body fluctuates through natural stages (i.e., your BP will be different during a fast than within an hour after eating, or after exercise).

More data is better

A great analyzer of baseball statistics used to say, "You can tell more about a hitter when he's gone 200 for 1,000 at-bats than if he's gone 2 for 10." More data means more valid conclusions. I'll be testing throughout the day today, but it will be better when I'm done doing it throughout the week.

Imagine you go to a doctor's office, and you get a "white coat" BP reading of 132/85. He says "hmm, you have a little hypertension." But you reply, "Dude, your office makes me a little edgy. I tested myself 100 times last month (wanna see the Excel file?) and it averages 115/75." Now who's in charge?

For example, here's what my data looked like early this morning (BCOF = before coffee; ACOF = after coffee):

7 AM:

Pulse oximeter (BCOF): 49/99; 44/98 (ACOF). Explanation: the first number is my resting heart rate (RHR); the second is a measure of how well oxygen is being diffused throughout my body (it goes up to 100).

Blood pressure (ACOF): 109/71; RHR 54. Explanation: caffienated coffee affects pulse and BP, generally increasing BP and lowering pulse.

Body composition (BCOF): 143 pounds; 9.8% body fat


(A few hours later, ACOF + meal, breaking intermittent fast)

10 AM:

Blood pressure : 112/67, pulse rate 56

Pulse oximeter: 47/99

Body composition (BCOF): 143.4 pounds; 11.1% body fat


...Now after a walk and a cat nap:

3 PM:

Blood pressure : 107/67, pulse rate 53

Pulse oximeter: 51/99

Body composition (BCOF): 143.1 pounds; 10.4% body fat

You can see how the data jumps around, depending on the circumstance, the context, the activities. Body composition does not change day to day (hardly week to week either), but the data spit out by your scale will. But now we're getting into some data, we have something to work with.

View Data In Context

One hundred BP readings would be good data, including knowing how food and sleep patterns were affecting it. A lot of data points will allow you to spot patterns (e.g., how does time of day, circadian and hormonal rhythms, affect your RHR and BP? What about the foods your eating and physical movement?).

I'll continue this throughout the week and then reproduce the results. You really should become kind of a stat geek and figure out the mean, mode, and standard deviations of the readings.