# Study of A1C vs GMI **URL:** \ **Category:** D-News\ **Created:** [March 16, 2021, 2:15pm UTC](https://forum.fudiabetes.org/t/study-of-a1c-vs-gmi/11082 "2021-03-16T14:15:43Z")\ **Posts on this page:** 3\ **Page:** 2
**Author:** ![MarkP](https://avatars.discourse-cdn.com/v4/letter/m/b3f665/32.png) [@MarkP](https://forum.fudiabetes.org/u/MarkP)\ **Post date:** [April 10, 2021, 4:44am UTC](https://forum.fudiabetes.org/t/study-of-a1c-vs-gmi/11082/21 "2021-04-10T04:44:47Z")
Glad to hear someone else say the same thing. I start feeling pretty bad above 130 and have had the sense that dexcom has been biasing their results downwards. If anything, a “TIR” measure should be symmetric around a non-diabetic mean and be allowed to be corrected if a number is fixed via juice but hasn’t shown up in the cgm readings yet. It makes more sense to consider a person’s average and sample standard deviation than their ability to stay between 70 and 180, I could go on about this for days and wish I had the qualifications and courage to show more “research” than the accumulated literature. Years ago, I was once chewed out by one of the renowned endos for having too many lows on my machine’s memory. What she didn’t consider, is that testing blood was expensive and I wasn’t going to waste strips if I knew the result was 80. ---
**Author:** ![Beacher](https://avatars.discourse-cdn.com/v4/letter/b/a6a055/32.png) [@Beacher](https://forum.fudiabetes.org/u/Beacher)\ **Post date:** [April 10, 2021, 1:37pm UTC](https://forum.fudiabetes.org/t/study-of-a1c-vs-gmi/11082/22 "2021-04-10T13:37:22Z")
> [@MarkP](#): > > a “TIR” measure should […] be allowed to be corrected if a number is fixed via juice but hasn’t shown up in the cgm readings yet Many of us get frustrated when we’ve treated, and we know we’re fine now, but the CGM still shows that we’re low, because we think the TIR is going to show we’re low more than we really are. But consider that the CGM lags at both ends. You may well have dropped below your lower limit 10 or 15 minutes before your CGM alerts you. By being late to alert you that you’ve dropped out of your range, and then late to show that you’re back in range, the TIR measure already sort of self-corrects. ---
**Author:** ![jbowler](https://avatars.discourse-cdn.com/v4/letter/j/e274bd/32.png) [@jbowler](https://forum.fudiabetes.org/u/jbowler)\ **Post date:** [July 4, 2021, 11:50pm UTC](https://forum.fudiabetes.org/t/study-of-a1c-vs-gmi/11082/24 "2021-07-04T23:50:52Z")
> [@MarkP](#): > > I start feeling pretty bad above 130 and have had the sense that dexcom has been biasing their results downwards. So far as I can determine from my period comparisons against a Contour Next-1 fingerstick the G6 biases _ **outwards** _ from something around 120 - i.e. readings below that are lower and above that are higher. This does make a curious sense; Dexcom would not want to be blamed for over-estimating a low BG or under estimating a high. > [@MarkP](#): > > If anything, a “TIR” measure should be symmetric around a non-diabetic mean and be allowed to be corrected if a number is fixed via juice but hasn’t shown up in the cgm readings yet. It makes more sense to consider a person’s average and sample standard deviation than their ability to stay between 70 and 180, I could go on about this for days and wish I had the qualifications and courage to show more “research” than the accumulated literature. 🙂 I don’t lack courage, but I am with you every step of that argument. We have the absolute, baseline, qualification; we really are fighting this war and we really do know what is happening. [Previous page](https://forum.fudiabetes.org/t/study-of-a1c-vs-gmi/11082.md?page=1)