phoenix said:
An Hba1c of 9.5% is the equivalent of an average blood glucose over the last 3 months of about 12.6 mmol (see
http://professional.diabetes.org/glucosecalculator.aspx)
While I would follow the sentiment that HbA1c of 9.5% is high, I am leery of statements like "the equivalent of an average blood glucose
over the last 3 months"
The first thing is that not everyone agrees on the conversion factors. For example:
http://www.diabetesdaily.com/calculator/ for a A1c of 6.0% it gives a whole blood "average" of 6.7 mmol/L whereas your link gives 7.0 mmol/L a 4% difference. I first found this a few years back when comparing my blood sugars with an American friend (I am in the UK) and I found that for seemingly the same HbA1c he was being given different "average" values.
These "average blood glucose" values are
not measured. It is the percentage glycosylated haemoglobin (HbA1c) that is measured and so you should only pay attention to the percentage, not the "average blood glucose" value. I agree that it helps to visualise what a percentage HbA1c means in terms of the blood glucose you can measure, but don't put any more meaning in it.
The other issue I have is with the statement "over the last three months". This is widely quoted, but I think it is misleading. I first became worried about this when I had two HbA1c tests within two weeks performed by the same lab. There was a difference of 2% between the two (if I remember rightly 9.5 -> 7.5), which is a large difference. Between the two tests I had made changes in my control, but when you think about
averages, a two week improvement in three months (12 weeks) should not have had a significant effect because if it is a true average (arithmetic mean) then two weeks could not have had such a significant effect. Also, since I saw a fall, then the actual blood glucose reduction would have had to be huge.
Think about what averages mean, as an example, take six values (say, weekly rainfall in mm): 10, 10, 10, 10, 10, 10 this is a total of 60 and an average of 10 (say, the weekly rainfall averaged over 6 weeks is 10mm). Now take a range of 10, 10, 10, 10, 10, 4, that is, the range is the same, except for the last value) the total falls to 54 and the average falls to 9. So a huge change in the last value (the spot value) has very little effect on the average. If I want my average to fall to 5 then it would take many more values of 4 to achieve that (in fact, 10, 4, 4, 4, 4, 4 gives an average of 5).
In fact, if you think about what the HbA1c measures, you can see that it is a
weighted average. See the explanation here:
http://www.leedsth.nhs.uk/sites/diabetes/tips/HbA1c.php. This says that red blood cells have an average lifetime of 120 days. Other sources suggest a lifetime of between 60 and 120 days. Three months is 90 days, which is the mid point between 60 and 120, and I guess where the "over 3 months" comes from. I don't know what is the correct value for the age of red blood cells, but the important point is that blood cells do not live
exactly 90 days (or 120 days, or 60 days, or whatever), instead there will be a range of ages. This means that a sample of blood will have blood cells of a range of ages, and (like in the general population of humans) there will be more younger cells than older cells. So more of the "younger" cells will contribute to the HbA1c than the "older" cells.
This means that the HbA1c is weighed towards the more recent past and that the "over three months" statement really means that the value is affected far more by the last few weeks and although values from three months ago will have some effect, the effect will be minuscule.
This would explain the big change I noted above, that the HbA1c tests I had were showing that the blood sugar over the last two weeks were contributing to the HbA1c far more than the other 10 weeks of the three months.
Another empirical example. Over the last three months I have been battling with high morning blood sugars. And I mean high - 24 mmol/L was typical, and I even achieved a personal record of 31. My night time insulin was at fault, and I have recently changed to a different insulin and now I have morning blood sugars of 4 to 7. Of course, with a morning blood sugar like that I had to do something, and so a combination of missing breakfast and taking extra short acting insulin brought my blood sugar down to normal by lunch and remained normal until bedtime. I went to my diabetic clinic and my HbA1c "over three months" (ie the period that I had those high morning values) was 8.5. Not wonderful, but not concerning (and less than the original poster's value). However, I clearly had a serious problem, but my efforts to get a normal blood sugar was masking the effect of the high blood sugars in the HbA1c value.
If you think I am cynical about using HbA1c as an accurate measure of diabetic control, you are right. However, there is a way to determine the quality of your diabetic control, and this has already been suggested: taking blood glucose tests yourself. If you want an "average" blood glucose level then measure your blood 5 times a day for three months, add up all the values, divide by 450 and you have your average over three months. I am not seriously suggesting that you do that, but in my opinion the only way that you can control your blood sugar is to take frequent blood glucose tests. Don't rely on HbA1c as being any more than a suggestion of how good control you have, do not celebrate when you achieve a "good" value, because it may not mean that your control is as good as you think.
Richard
