October 8, 2007

A1c. Grrr.

I've been seeing the best blood sugars I've had in years these past few months. So I was really curious to find out what my latest A1c would be. My last one was in April and was 5.6%.

The best A1c I'd had since starting insulin was 5.5%, but that was taken two weeks after the Christmas holiday period when I'd been eating what you'd expect me to be eating that time of year, so I figured that was why it was higher than the 5.2% I'd been expecting.

But this time I'd been eating right for months and testing enough that I knew there were no hidden spikes going on. So I was looking forward to a pleasant surprise.

Wrong. The A1c came in at 5.8%.

The DCCT formula suggests that a 5.8% A1c results from running an average blood sugar of 129 mg/dl. No way!

My fasting blood sugar has been a rock solid 92 mg/dl for months. My three hour post meal blood sugars have been in the 90s or 80s. My peaks have been in the 120s, especially the during past two weeks when I'd made some changes to my daily regimen, but even before that I rarely went over 140 mg/dl ever and I was coming down into the 90s after meals.

The highest number I'd seen during this whole time was a single test result of 175 mg/dl which was measured half an hour after eating something starchy when I had injected my Novolog a bit late. Even so, my blood sugar had dropped back to 86 by 2 hours after the meal. And that was almost a month before the test.

The Nathan formula maps that 5.8% to a lower mean glucose number, 107 mg/dl. But that is still 10% higher than any average I see on my meter, total, fasting or post-meal.

Toss into the mix that the A1c is supposed to be heavily influenced by the last two weeks and it gets even sillier. I've been seeing 70s all afternoon the last few weeks.

Another test confirmed that I'm not living in a fantasy by thinking that my blood sugars have improved. My blood lipids were better than ever, in a way that confirmed that my exposure to blood sugars had dropped. With no cholesterol drugs, my triglycerides were 80. My HDL was 76, higher than ever before.

My conclusion is that either a) the A1c is useless for predicting mean glucose normal blood sugar levels or b) I'm a "high glycator" or c) the lab screwed up.

Whatever it is, I'll trust the evidence of my 5-6 blood tests a day over the A1c.

But it is annoying!


Robin said...

I am reading various things, sorry don't remember where, and new Internist is using a lower average BG/A1C chart for A1Cs in the 5s. I don't know what this means given 4.5 is more likely with normal blood sugars. I agree it's a mystery. I'm a 5.7 staying between 90 and 120, most under 110.

nonegiven said...

My last one was 5.2%, I was surprised because I hadn't been all that good. I've been really good the last few months so the next one should be lower, but who knows? My last 2 cholesterol tests were screwy also but in different directions, on one the HDL was much, much higher than previous tests with the other parts the same and on the other test the LDL did the same thing with the HDL back down to my normal range. One thing I'm going to try is to drop all my supplements a few days before the next test.

ORION said...

I used to be a Med Tech for years before I became other things ( an author!)
and I am stunned that so much has stayed the same - little progress with this disease.
BTW My SECOND favorite author is Jane Austin!

jpatti said...

I ran across this in Taubes book last night - fructose actually glycolyates proteins more than glucose does.

That's why the equations to convert blood glucose to A1c don't always work.

It seems to me that A1c is more of a direct measure of organ damage than bg is.

Jenny said...

Sounds like time to get a copy of that Taubes book. Everyone seems to be mentioning it!