H(a)emoglobin Woes

Sunday, June 25, 2006 | 11 comment(s)

WARNING: The author of this post realizes that it is long, detailed, and embarassingly revealing of just how much of a freakshow
Ladies and Gentlemen!
Step Right Up!
Admission is FREE!
Don't Be Afraid!
Kids, don't try this at home!
he is. But if you feel as if you're ready to cash in your chips and click yourself away from here before getting to the end of the post, please check out some questions I have near the bottom that I would love to get some feedback on before you go. "We thank you for your support" (- Frank Bartles).

Ahhhh, the H(a)emoglobin
The Brits spell it "Haemoglobin" but in the States the 'a' is dropped. From here on out, I'll just use "Hemoglobin" but I wanted to give a nod to the folks over there on the other side of the pond. Oh and this is kinda interesting...(man, I love wikipedia!).
A1c Test (sometimes "HbA1c" or even shorter: just "A1c"). Whatever you call it, we are all familiar with it.
At Least…
We diabetics are certainly familiar with it, whether friends and family members (and thus some readers of this blog) are familiar with it, is another matter completely.

For those who don't know, the Hemoglobin A1c test is the 'gold standard' test of how well controlled one's diabetes has been over the previous 3 months. Below 5% is the normal, non-diabetic range, and while the American Diabetes Association suggests diabetics shoot for a reading of <7%, the American Association of Clinical Endocrinologist recommends a more stringent reading of <6.5% as the target for diabetic patients.

More Detail (just in case)
Hemoglobin are protein structures that are contained in your red blood cells. Lab tests measure how much glucose binds to these proteins. Because red blood cells are constantly dying and being created, the HbA1c test can give a measure of how much glucose one has had floating in their system over the lifespan of the red blood cell population (which is approximately 3 months). Since this population of cells is constantly in flux, HbA1c results are essentially a weighted average of blood sugar control. That is: about 50% of the reading represents the previous month's control and then 25% represents the previous two month's control, and the final 25% measures the previous 3-4 month's worth of control.

So I went to my endo appointment on Thursday morning and I got my test result back. As I wrote about previously, I was pretty anxious / excited about getting the test done this time 'round. In that post, I said I'd be "dancing a jig if I [hit my target A1c], and crying to fugues if I don't." Well, I have my result, and wouldn't say I'm dancing or crying, really. Instead I'd say I've got quite a salad of emotions about the dumb number. A partial list of ingredients would include: pissed off, confused, determined, depressed, and sprinkled with a little bit of happy, too (just for the crunch). Soooo... (without any further ado), my A1c reading was 7.1,
I know, I know...
This is a decent reading... a good reading, even. And I may be lambasted for complaining about such a score, and if so, I completely understand. I realize that HbA1c scores can be a sensitive topic (not unlike asking how one scored on a final exam or asking how much someone's salary is). And I've felt this way before too. When I've seen or heard people complain about blood sugars that bounce above 200 occasionally or A1c's that are in the low 6's or even the upper 5's I think they're crazy to be complaining and I get a little angry, too. But I think that largely stems from envy. But overall I'm happy for them, and I respect the amount of discipline they probably must exert to attain such great scores on a consistent basis. And so, I apologize if anyone reading this thinks that I have nothing to complain about, but I'm a wee-bit anal retentive and I've been particularly goal-oriented on this lately. Please excuse me.
and immediately after being told this, I vehemently protested: "But that just doesn't make any sense!"

Like I mentioned before, I've been busting my ass to maintain the best control I can (and I think that I have probably gotten to the best control that I ever have had). I've kept detailed records, tested my blood sugar many, many times per day, I've been eating the right things, I've been exercising, I've even lost some weight, yadda, yadda, yadda, etc., etc., etc., ad infinitum, ad nauseum. So beforehand, I was looking for a conversion chart between what my average bloodsugars have been and what I might expect for my A1c reading. I came across the following table in various formats and in various locations (though this one is lifted from the Children with Diabetes web page on the HbA1c test. And this table seems to originally come from a 2002 "Diabetes Care" article that analyzed data from the Diabetes Control and Complications Trial (DCCT) and I've seen the same table in numerous other places including ADA (which publishes "Diabetes Care") and NIH (i.e., reputable sources).

 A1c (%)    Mean Plasma Glucose    
From Diabetes Care - 26 (Supplement 1): Table 1

Based on this table, my 7.1 HbA1c reading indicates that my average blood sugar has been around 170 mg/dl. But based on the data I've collected with my glucometer, I again have to protest: "This simply cannot be!" Below is a table of data from my glucometer for the 7, 14, 30, 60, and 90 days prior to getting the blood drawn for the HbA1c test:

    Kevin's Average Readings    
   days       mg/dl       # tests       tests/day       S.D.   
7 145 145 20.7 65
14 151 270 19.3 67
30 148 563 18.8 62
60 150 1,141 19.0 62
90 150 1,658 18.4 65

So, a few things: my average blood sugars tend to be around 150 mg/dl (not great, but not awful, either). I test like a maniac -- on average about 20 times per day. And my standard deviation is in the mid-to-upper 60s (I believe I've read that one should be aiming for below 60 in terms of S.D. as there's a bunch of new research suggesting that the amount of glycemic variation one has may be more important than the actual average level in terms of complication risks).

I then took the HbA1c-to-Mean Plasma
This is important
So apparently when a lab takes a sample of blood to perform the HbA1c test, it filters out some of the blood first (though I don't know which parts or why) and only tests the blood plasma. When we prick our fingers, however, we are getting whole blood samples. The concentration of glucose in the whole blood sample is therefore slightly lower than in the plasma basically because the denominator is larger. Most meters these days do a quick internal calculation to adjust for this fact so that the readings displayed on your glucometers are in "plasma equivalents." (I think that almost all glucometers simply multiply your whole blood reading by 1.11 to get the "plasma equivalent").

Until about 6 months ago, I was still using the OneTouch Profile II glucometer. This might be the only meter on the market that still produced results in whole blood values (and in addition took 45 seconds to give up it's results (I really have no idea why it took me so long to upgrade to a more modern glucometer)). But I have actually had arguments with my endocrinologist about the importance of this matter when I originally started to notice a disconnect between what my average readings were telling me and what my HbA1c readings came back as. She basically mocked my concern and told me that there is no difference and that there are no meters that still use the whole blood measure -- two facts that are patently wrong and only a small portion of the list of why I need to find another endocrinologist (but that's a whole 'nother post).
Glucose table and did some interpolating between 6.0 and 7.0 as such:

 A1c (%)    Mean Plasma Glucose    
6.0 135.0
6.1 138.5
6.2 142.0
6.3 145.5
6.4 149.0
6.5 152.5
6.6 156.0
6.7 159.5
6.8 163.0
6.9 166.5
7.0 170.0

Based on this table and my glucometer readings, I was going into the test expecting a HbA1c result in the 6.5 range. 7.1? 6.5? What's the difference, right? (0.6). I know I shouldn't be upset about this, but I am. I even argued about it with my endocrinologist. Since then, I've been doing research on my own and reading medical articles with such riveting titles as: "Effects of Hemoglobin Variants and Chemically Modified Derivatives on Assays for Glycohemoglobin" and "Defining the Relationship Between Plasma Glucose and HbA1c." The later of which seems to be the original source where they analyzed average plasma glucose readings and statistically
Be Thankful
I originally intended to go over the linear regression coefficients and discuss their methods of calculating "average plasma glucose" but that would have required yammering on about arithmetic averages, weighted averages, the "area under the glycemic curve" and then perhaps some integral calculus and THAT would have been overboard.
compared them with HbA1c readings.

*************     KEY QUESTIONS     *************
So before I wrap this up, here are my questions for you:
Do your average glucose readings from your meter fit well with the expected HbA1c readings you get back from the lab (or from an HbA1c meter if you're lucky enough to have access to one at your endocrinologist's office)? Or do you regularly experience a disconnect between what you would expect your A1c reading to be based on what your average blood sugar readings have been?

To make a short story (even) long(er), I actually had a new blood sample drawn to be sent to a different lab on Thursday. In addition to an HbA1c test, I am also having a hemoglobin electrophoresis test done to see if I have any abnormalities in my hemoglobins that may bias my HbA1c test scores. I'll be sure to report back what I find out.

And so I'll wrap up this long-ass post with a selection from
Defining the Relationship between Plasma Glucose and HbA1c:

Several studies have suggested that, although intraindividual variation in HbA1c is minimal, there is evidence of wide fluctuations in HbA1c between individuals that are unrelated to glycemic status, suggesting that there are "low glycators" and "high glycators" (14–16). However, a recent study showed that when multiple observations per patient are used to minimize the effects of assay variation, the interindividual range of HbA1c results in nondiabetic individuals is actually quite narrow, <1% HbA1c (17). Therefore, for any individual patient, a consistent discrepancy between patient-monitored PG determinations and estimated HbA1c should be investigated; there may be other factors causing this discrepancy, such as improper meter use, laboratory error, a physical condition that alters red cell life span, or a variant hemoglobin interfering with the HbA1c assay method.

11 Comment(s):

Anonymous Anonymous said...

my average blood sugar according to my ultra smart meter for the 90 day period was 7.5mmol (135) my A1c came back at 4.6% so those meter's average are not that accurate.

Blogger Kassie said...

a little freaky, you are...

Remember that the meter is a snapshot and the A1c counts all the blood sugars. Though it is clear that you take lots of snapshots, and should reasonably expect a closer result. Maybe try switching up some testing times - swap a 1 hour post prandial for a 2 hour, etc.

Also remember that meters can be off by as much as 20%.

If you think it's the meter (and not just the difference between home monitoring and lab monitoring), call the meter company and ask for a new meter.

Look also for hidden highs... can you get your hands on a trial CGMS? You might be running high during your sleep and not catching it.

Just some random thoughts. I suspect you may already have considered these points but figured, what the heck.

If all else fails, resort to my nunberless A1c scale (http://noncompliant.blogspot.com/2006/01/whats-your-a1c.html) :)

Blogger Nic said...

I feel your frustration. My last A1C was 6.4, which reflected the many 40s and 30s I had had but not the more recent 300s. Now, I'm happy that it didn't reflect the 300s, but it did show me how flawed the whole system is and made me wonder what glucos testing and AiC's don't tell us in general. And what price our bodies pay as a result.

But...kudos to you on your regimen. Please don't give it up...it will pay off.

Anonymous Anonymous said...

Sorry that it wasn't what you wanted ... I know you have been pushing hard for this. As one of the other comments said "it will pay off."

On another note ... "salad of emotions, with a little bit of happy just for the crunch" may be one of the funniest lines I have read in awhile. Clever. You've been reading your Nick Hornsby!


Blogger Scott K. Johnson said...

Hi Kevin,

First of all, I'm going to say "Congratulations" for a great A1C. Maybe not what you were hoping for, given all the work you've done, but great nevertheless.

I've never paid that much attention to the difference between my BG averages and my A1C's, mostly because I know they are both higher than I want, and I'm not close enough to target to warrant that level of scrutiny.

I think that as Nic said, it will pay off. Do you feel you are able to keep up with all the work? Has it become a lifestyle? If that's the case, your A1C's will fall into line, maybe with no logical explanation for it (it is possible to over-analyze you know).

If the work is too much, then to find a good balance between control and being happy with how things are is what you are after.

I give you a lot of credit for really whipping things into line, and you are testing an insanely high number of times per day (how the hell do you get so many strips?), and actually doing something with those numbers.

Kassie has some other good thoughts too - maybe you're just not catching the highs. Switch it up a little bit.

This scratches the surface of a very deep topic - the lack of reward for so much hard work. There's a blog in itself on that.

Great post - I always enjoy reading what you have to say.

Blogger floreksa said...

First off - Congrats on the 7.1!!! Even if you were expecting lower, its still great!

While I was pregnant, I was a charting lunatic. I found that if I averaged each day's readings, then divided by 20 I got a pseudo A1c #. Then when I'd average all those together for the time period, I was actually coming pretty darn close to what my actual A1c was.

Anonymous Anonymous said...

Hi Kev,
I am so in your shoes! My recent A1c (taken at the ADA Conference, see http://www.diabetesmine.com/2006/06/ada_conference_.html) was also 7.1 -- although I also got a result of 6.4 there the previous day, which I knew must be bogus.

My early A1c's after diagnosis were deliciously low, but in those days I was also having one to two severe lows per day! Not fun.

Soooo, I'm struggling to get "tighter" while still be able to (occassionally) eat what I want.

Aaargh. Hopefully the Dexcom will help.

Blogger Minnesota Nice said...

I take readings before & after breakfast and lunch; before supper and before bedtime. If the average of these readings is below 140-150, I'm usually at about 6.8-6.9 AlC - very consistently.
I have tried the Metris at home AlC test and it seems to be waaaaaaay off............a waste of money, in my opinion.

Anonymous Anonymous said...

So my husband was working as hard as usual with controlling his blood sugars (43 years diabetic/56 years old/tests 6 times minimum/day)A1C's usually 7.2 and when it is 6.8 or 6.9 it usually means several lows are entering into the equation. Last A1C we expected GREAT things......average on meter around 140....2week/one month & 90 days!! So results come back 7.6 - - worse one he has ever had and we thought we got punched in the stomach!!! Didn't think it was right and the doc and diabetic nutriionist said there may be "hidden" highs. Only thing we could think of was that 3 months prior there was a string of 3 days with all 300/400's. Thought they would be out of the mix. AND the day the A1C was drawn it was a 300 reading!! So we were convinced both contributed to these numbers. Waited another 6 weeks - - good daily/weekly numbers now averaging 130 and the A1C yesterday was 6.4!!!! VERY HAPPY - and we are convinced that bad string was the result of the extremely high A1C as this test "always remembers".

Anonymous Anonymous said...

DR. Khalid (Not a medical doctor)
My A1c is measured in three labs within 10 days, results are 5.3, 6.3 and 6.9
My average home BG is 115!!

There is NO accurate conversion table between A1c and avg. BG. U can use only the detla (or cahnge) in A1c to see ur control. I salute u for this work and wish u can publish it.

Anonymous Anonymous said...

Who knows where to download XRumer 5.0 Palladium?
Help, please. All recommend this program to effectively advertise on the Internet, this is the best program!

Post a Comment || Go Home