43 Hour Fast

     One of the virtuous things about using a Continuous Glucose Monitor, is the ability to study one's own body and how it naturally and uniquely works with food and insulin. To determine the accuracy of my basil rates in the past, I have used the Medtronic CGM, the test-every-hour method, and even the Gluco-Watch back in 03-04. But I prefer Dexcom's G4 CGM, as it has come a long way in accuracy and ease/comfort of use- even more than previous models. 

     A CGM can offer scientific evidence necessary to answer the following universal questions:

How much does my blood sugar drop per unit of insulin? 

How much do carbs raise my blood sugar?

     Between those questions lie the golden calculations: 

How much insulin do I need to drop my high blood sugar?

How much insulin do I need to eat this?

     At the time of writing this, I am currently in the middle of a 43 hour fast, trying to get a beat on my basil rate. I am also figuring out my insulin to carb ratios. I stopped eating at 1030 pm on Monday 2/22, and plan to fast through Wednesday 2/24 afternoon.

     Around 1 am on Tuesday morning, I happened to wake up and saw that my CGM reading was 140. I decided to take a unit of insulin to see what would happen. I woke with a 100 sugar, so I assume that a unit drops my sugar 40 points. This is not a completely pure test, since I had eaten 2.5 hours earlier.

     I woke up Wednesday morning at 715 am with a low notification on my pump at 69. The CGM required a recalibration, so a finger stick showed 56 - not a huge deviance. Obviously, my basil rate is too high at some point overnight, and I will look into that deeper. But with a medically necessary opportunity to consume a fixed amount of carbs, I decided to use this to check my carb to blood sugar ratio.

745a: drank 24g carbohydrates worth of pineapple juice

     After consuming a small can of pineapple juice (24g on the label), my sugar jumped up to 210, an increase of 154 points in about an hour. Simple math proves that my blood sugar has a:

6.4 point rise per 1 gram.

   At 1050a, my sugar was 190. This has given me an opportunity to test if a unit of insulin really brings me down 40 points. Novolog publishes the precaution that it begins working in 5-10 minutes and lasts 1-2 hours. So with "40 points per" unit in mind, I decided to take 2 units of insulin. And at 1250a, I reviewed my CGM to see the result.

2 hrs after a 2 unit bolus: 190 dropped to 125

     So at 2 hours, my blood sugar decreased by about 65 points. By 136 pm, it was down another 13 points to 112. That could mean that my basil rate is a little high, lowering my sugar more than the correction bolus alone, or it could be that insulin works longer than 2 hrs in my system. Either way, I have surmised that novolog creates a:

32-40 blood sugar point drop per 1 unit

     If about 5 grams of carbs raises my blood sugar the same amount that 1 unit of insulin decreases it, then in fruit juice form, I would need

 .2 units of insulin per 1 gram of carbohydrate 

     I don't fully believe this conclusion. I would never expect to give myself 3 units for 15 grams of carbohydrates. Typically, it is a one unit per 15 grams ratio. I feel like I'm invalidating my own controlled experiment with what I am accustomed to and what I expected.

     But not to worry. I am going to test this now. I have a Del Monte fruit cup totaling 12g of carbs, which, using my calculations, would require 2.4 units. It is 2 pm (finger stick = 103, CGM = 108) so I'll post the screen shot of what happens after this carb-insulin combo around 4 pm.

     So the moment of truth arrived. I am surprised I'm not hungry after going 43 hours with only eating a small can of fruit and having a small can of pineapple juice. But here we are, 4 pm, and the result look...accurate. The 2.4 units of insulin seem to have balanced out the intake of 12 g of fruit cup. Science and math have won over assumed perception.