• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

Hba1c Test: Biased Towards The Most Recent Weeks

Mr_Pot

Mr_Pot

Well-Known Member
Messages
4,573
Location
Berkshire
Type of diabetes
Type 2
Treatment type
Diet only
This bias is often vaguely mentioned but I have been unable to find and convincing references. I have tried to work it out for myself but I would appreciate a peer review in case of errors in my logic or my maths.
 

Attachments

Maths isn’t my strong point but this is very interesting. Thank you for taking the trouble to do the calculations.
 
Since reading your post I have tried to find references about the weighting towards the end of the period, and failed. It is mentioned from time to time on this forum, and I have just taken it for granted that it is correct. From your calculations, it does appear to be correct.
 
I wonder how MySugr works It out as you don’t need to put 12 week’s worth of readings in to get an estimate there? And I’ve always found it to be fairly accurate.
 
Rachox said:
I wonder how MySugr works It out as you don’t need to put 12 week’s worth of readings in to get an estimate there? And I’ve always found it to be fairly accurate.
Click to expand...

It doesn't work it out. If you go through it, all it seems to do is average all your entered data and converts it to mmol/mol. Or at least that is what I find on my data. I also enter my data on a spread sheet, and average all the numbers then convert to mmol/mol. It comes out exactly the same as MySugr. It isn't measuring glycated cells like the HbA1c does.
 
I downloaded the file but my tired old tablet refuses to open it, would someone be so kind as to give a synopsis, please? TIA
 
@Guzzler I've copied and pasted it, but the formatting has gone haywire. :arghh:

HbA1c Test: Biased towards the most recent weeks
Assume haemoglobin is renewed continuously over a period of 12 weeks, where the end of Week 12 is the date of the HbA1c test.
Assume a blood sample contains equal amounts of haemoglobin created in each of the 12 weeks.
Suppose 100 molecules of haemoglobin are glycated in each week (actually more like 100 billion) then after 12 weeks in the sample there will be….
1/12 x 100 = 8 molecules glycated in Week 1, as the remaining 11/12 glycated that week will have been renewed.
2/12 x 100 = 17 molecules glycated in Week 2, as the remaining 10/12 glycated that week will have been renewed.
And so on until Week 12 when all the 100 molecules glycated that week are still in existence.
Glycation During
Glycated Molecules in the Sample
Number
Percentage
Week 1
8
1.28%
Week 2
17
2.56%
Week 3
25
3.85%
Week 4
33
5.13%
Month 1
13%
Week 5
42
6.41%
Week 6
50
7.69%
Week 7
58
8.97%
Week 8
67
10.26%
Month 2
33%
Week 9
75
11.54%
Week 10
83
12.82%
Week 11
92
14.10%
Week 12
100
15.38%
Month 3
54%
In this table a total of 650 HbA1c molecules will be in the sample at the end of Month 3, of which 54% will be due to the glucose level in the most recent month.
 
Bluetit1802 said:
@Guzzler I've copied and pasted it, but the formatting has gone haywire. :arghh:

HbA1c Test: Biased towards the most recent weeks
Assume haemoglobin is renewed continuously over a period of 12 weeks, where the end of Week 12 is the date of the HbA1c test.
Assume a blood sample contains equal amounts of haemoglobin created in each of the 12 weeks.
Suppose 100 molecules of haemoglobin are glycated in each week (actually more like 100 billion) then after 12 weeks in the sample there will be….
1/12 x 100 = 8 molecules glycated in Week 1, as the remaining 11/12 glycated that week will have been renewed.
2/12 x 100 = 17 molecules glycated in Week 2, as the remaining 10/12 glycated that week will have been renewed.
And so on until Week 12 when all the 100 molecules glycated that week are still in existence.
Glycation During
Glycated Molecules in the Sample
Number
Percentage
Week 1
8
1.28%
Week 2
17
2.56%
Week 3
25
3.85%
Week 4
33
5.13%
Month 1
13%
Week 5
42
6.41%
Week 6
50
7.69%
Week 7
58
8.97%
Week 8
67
10.26%
Month 2
33%
Week 9
75
11.54%
Week 10
83
12.82%
Week 11
92
14.10%
Week 12
100
15.38%
Month 3
54%
In this table a total of 650 HbA1c molecules will be in the sample at the end of Month 3, of which 54% will be due to the glucose level in the most recent month.
Click to expand...
Thanks, Bluetit.
 
Bluetit1802 said:
It doesn't work it out. If you go through it, all it seems to do is average all your entered data and converts it to mmol/mol. Or at least that is what I find on my data. I also enter my data on a spread sheet, and average all the numbers then convert to mmol/mol. It comes out exactly the same as MySugr. It isn't measuring glycated cells like the HbA1c does.
Click to expand...
No I realise it’s only worked out on averages. It’s uncannily accurate though!
 
Back
Top