[Jonathan_R]

I've tested this against ACCU-CHEK: A1C calculator and it's with in 100th of a percent close.

(BGL+77.3)/35.6

Reference: Calculating A1c from average BGL

The formula I posted is the closest, and the best one I have found. That is, with out getting overly complex. I saw some ones where you needed a med degree to work it.

[Lloyd]

Originally Posted by Jonathan_R

I've tested this against ACCU-CHEK: A1C calculator and it's with in 100th of a percent close.

(BGL+77.3)/35.6

Reference: Calculating A1c from average BGL

The formula I posted is the closest, and the best one I have found. That is, with out getting overly complex. I saw some ones where you needed a med degree to work it.

Don't I wish!

96 + 77.3 =173.3 173.3/35.6 = 4.867

Actual A1c 5.4

Here are some of my actual numbers and A1c's

138 A1c 6.9
99 A1c 5.7
96 A1c 5.4
90 A1c 4.9

All I can say is If I know my average, I can predict my A1c within .1

-Lloyd

[Jonathan_R]

Originally Posted by Lloyd

Don't I wish!

96 + 77.3 =173.3 173.3/35.6 = 4.867

Actual A1c 5.4

Here are some of my actual numbers and A1c's

138 A1c 6.9
99 A1c 5.7
96 A1c 5.4
90 A1c 4.9

All I can say is If I know my average, I can predict my A1c within .1

-Lloyd

I used this, ACCU-CHEK: A1C calculator to check my work. The problem is, it only goes as low as 100.

[DeusXM]

I used this, ACCU-CHEK: A1C calculator to check my work. The problem is, it only goes as low as 100.

That's because an AVERAGE reading of below 100 isn't healthy for anyone. From the calculator, you can extrapolate that a non-diabetic A1c is anything less than 5.9%, so I'd daringly suggest there's actually no benefits to pushing for anything lower than that.

[Subby]

Jonathan, not sure what your aim is here. Is this a way to perform those rough calculations instead of use an online calculator like the one you mention? Or are you looking for "real" or "improved" accuracy to predict or replace A1c tests?

[Grunch]

It was very accurate for me, less than .1 difference. That's weird cause I think the number of measurements and the moments when you measure can make a pretty big difference in avg BG.

[Jonathan_R]

Originally Posted by Subby

Jonathan, not sure what your aim is here. Is this a way to perform those rough calculations instead of use an online calculator like the one you mention? Or are you looking for "real" or "improved" accuracy to predict or replace A1c tests?

It's so if your doing a log or spreadsheet or away from the computer/internet, then you can calculate your A1C. If I wanted "real" I would have posted the medical formula. I didn't feel to many would have understood it, so I didn't. Can't say it's "improved" either. It's just what I said.

My A1C, according to the formula, for the month of July, is 7.6351, which compared to the accu-check A1C calculator is pretty accurate. I only reference the accu-check calculator because I have their meter and to check the formula. That's it.

[Subby]

Ah I see... thank you for clearing that up. Fair enough if you want indications away from the internet. But, as mentioned, there is one problem with any A1c calculation based on BG tests - and that is that there is no guarantee of a representative scatter, of actual blood sugar over the period, from the tests. That simply cannot be guaranteed. As long as that's clear... it's just part of the mathematical reality if the data - there are variables you simply cannot account for.

That doesn't mean it's necessarily "useless" to do so, I am not disparaging your thread or idea. Just that, just as warned at the Accucheck site, none of these extrapolations are to be taken as anything more than a potential indication of what an A1c test might actually be. It could be really accurate this time. It might be less so next time.

[Jonathan_R]

Originally Posted by Subby

Ah I see... thank you for clearing that up. Fair enough if you want indications away from the internet. But, as mentioned, there is one problem with any A1c calculation based on BG tests - and that is that there is no guarantee of a representative scatter, of actual blood sugar over the period, from the tests. That simply cannot be guaranteed. As long as that's clear... it's just part of the mathematical reality if the data - there are variables you simply cannot account for.

That doesn't mean it's necessarily "useless" to do so, I am not disparaging your thread or idea. Just that, just as warned at the Accucheck site, none of these extrapolations are to be taken as anything more than a potential indication of what an A1c test might actually be. It could be really accurate this time. It might be less so next time.

Exactly right. An A1C is supposed to be an average of many blood sugar levels anyway. The more, the better.

[ShottleBop]

Originally Posted by DeusXM

That's because an AVERAGE reading of below 100 isn't healthy for anyone. From the calculator, you can extrapolate that a non-diabetic A1c is anything less than 5.9%, so I'd daringly suggest there's actually no benefits to pushing for anything lower than that.

Even the ADA acknowledges that lowering one's A1c below 6% may reduce the risk of suffering microvascular complications; the ADA recommends 7% out of concern that diabetics striving for "normal" A1cs run too high a risk of hypoglycemia. Defending ADA's A1c Target:

It is important to be clear about exactly what each group maintains in its published guidelines. AACE states simply that its goal is ≤ 6.5%.1 ADA not only states that its goal is <7%, but also notes that it is critical to include the following “key concepts in setting glycemic goals”:

The goals should be individualized.

Certain populations such as the elderly, young children, and pregnant womenrequire special considerations.

Less intensive goals may be appropriate in those with a history of significant hypoglycemia or hypoglycemia unawareness.

More stringent goals (i.e., a normal A1C of <6%) may further reduce the risk of microvascular complications at the cost of increased risk of hypoglycemia.2

ADA suggests that one should strive for the lowest A1C appropriate for the patient based on these concepts.

[DeusXM]

Even the ADA acknowledges that lowering one's A1c below 6% may reduce the risk of suffering microvascular complications

Precisely. What I'm saying is that once you've got an A1c below 6%, there's no point trying to drive it any lower. If your A1c is 5.9, there's no point whatsoever in trying to get it down to 5.2% or whatever.

[Subby]

Originally Posted by Jonathan_R

Exactly right. An A1C is supposed to be an average of many blood sugar levels anyway. The more, the better.

Well, kind of... but not really. I feel like we are chasing around a bush or something a bit here... I'm just going to go back to basics if that's ok...

HbA1c is a test that measures the amount of glycated hemoglobin in your blood. Glycated hemoglobin is a substance in red blood cells that is formed when blood sugar (glucose) attaches to hemoglobin.

HbA1c: MedlinePlus Medical Encyclopedia

It's specifically measuring the state of elements your blood. The result is something that once translated, correlates to an average blood sugar. And a figure we can take meaning from in itself.

So, yes you are right, that a HbA1c result indicates an average blood sugar level for the last three months.

But, lots of blood sugar levels extrapolated into a potential HbA1c result, is not an "A1c". That process can only ever be approximate (whether close or not), given pretty much any reasonable use of BG testing. It just can't replace the real thing. And without the real thing, you can't really check how closely it predicts the mark.

[Lloyd]

Originally Posted by DeusXM

Precisely. What I'm saying is that once you've got an A1c below 6%, there's no point trying to drive it any lower. If your A1c is 5.9, there's no point whatsoever in trying to get it down to 5.2% or whatever.

Three advantages are gained when you have very tight control (low standard deviation and a low A1c).

1. When you make a mistake, your numbers are still ok.
If I forget to take my metformin at a meal, for instance, 2 hours later I will be 120 to 125, instead of about 90 - 95. Still ok.

2. Meal spikes don't take me over 140, usually.

3. High variability, in and of itself, is an increased risk for complications.

I'm a little bit low, in the 60's, 2% of the time. In the 50's only about twice a year.

Non-diabetics go in the 60-s, as tests with a continuous glucose monitor reveal.

My average of 96 is perfectly safe for me, my doctor agrees.

For a volatile T1 likely it would not be a good idea to go that low, as they would be likely to suffer from hypoglycemia.

-Lloyd

[DeusXM]

Three advantages are gained when you have very tight control (low standard deviation and a low A1c).

The advantages are purely from a low standard deviation rather than a reduction in A1c. If you have a low standard deviation and an A1c of 5.9, there are no benefits to reducing that A1c any further.

[Jonathan_R]

Last I met with my endo, they did not want me below 200. I have a tendency to crash when below 200. Or at least I used to. Because of this board, I am getting better. This month is off to a rough start, but I'll get it. Currently, my standard deviation is 350. That's right, 350. In the last couple days I have had 3-4 hypoglycemic attacks, and several spikes (usually following the hypos).

I am finding out that excersize make a huge difference in my insulin sensitivity. My average is still 205, but that's because of rediculas lows. I am also finding out that I am incredibly sensitive to any sort of change. It has huge impacts on me, and lasts for days.