A good friend and myself decided to play with my dexcom last night. So after a 4h hike in the woods we arrived at his home ready to eat. I've had my Onetouch meter calibrated against lab results in the last year with a reasonable correlation.
I started off with a calibration BG. Since I knew that my Dexcom 7 plus updated every 5 minutes I determined the minute it received the update and ran onetouch tests on that minute at 5 minute intervals. I used the same finger (oh my poor pinky) until I ran out of test strips after 19 pinky bleeds.
Finally, with the help of my friend Levente the data was tabulated and compared for correlation. Ultimately we determined that the result from the dexcom correlated closest with a 10 minute delay. With additional research - something he's far better at than me we found an article: Real-time continuous glucose monitoring
Bruce Buckingham, Kimberly Caswell and Darrell M. Wilson
Curr Opin Endocrinol Diabetes Obes 14:288–295. ß 2007 Lippincott Williams & Wilkins.
It says:
The current sensors measure interstitial glucose, which
is a distinct physiologic space. When the blood glucose
levels are changing rapidly, interstitial glucose levels
will lag behind blood glucose levels by 8–18 min [5–7].
Bingo. So our result of 10 minutes of lag turned out to be in the proper range. I guess it's really not new info but it does seem to contradict some advice given on line of 20-30 minutes.
Sorry for those who are more comfortable with the US measure, please multiply the numbers by 18. The reason for added precision on the "Monitor" column is that my Dexcom is a US version and we had to divide everything by 18...
A few smaller points on shortcomings... With better planning I should have recorded a few readings on the Dexcom before starting the finger pricks as this would have provided a stronger sample for the correlation tests.
Marked on the graph you can see when I took 10 units of Novo Rapid for an incoming calorie packed meal with desert. I overdid it a little on the insulin but in the hours that followed it did work out. In hindsight I probably would have been better with 8 units. If there had been a large spike then the delay would have been much easier to see. I was also mostly stationary during this experiment and I assume more activity and more circulation may have changed the delay.
Other background info: This relates to me, my body type and such, it may not apply to everyone. I'm a 34 year-old at 6'1" 1.86m and 184lbs 84kg. I assume larger people may have longer delays.
-Michael
1941-GlucoseTest.pdf (21.9 KB)