Hba1c test coming up.....

[Sulfate]

This probably sounds a silly question but do you think taking an average of the last 3 months of my BS readings (before meals) will give me an idea of what my result could be? I'm not sure why but the last month has been in the 6's the previous 2 months was in the 5's.

 

[Leadinglights]

This probably sounds a silly question but do you think taking an average of the last 3 months of my BS readings (before meals) will give me an idea of what my result could be? I'm not sure why but the last month has been in the 6's the previous 2 months was in the 5's.

Not really as they are measuring something different and is only a small number of readings which would not be taking into account the highest and lowest even during a day let alone a three month period of time.

 

[Dragonheart]

The more readings across a better representation of the whole day the more likely it is to correspond to hba1c.

Conversely a single reading a day will mean little, and a pattern of only the highest or lowest readings of the day will squew the estimate badly in one direction or the other.

 

[Docb]

I think you can use BG data to predict HbA1c but you have to be very nerdy to get anything near the lab result.

What you need to do is to record lots of results (several thousand) over several years (so you have several HbA1c results) so you can extract averages (overall 90 day average in the period before the HBA1c or a waking average, again in the 90 day period before the HbA1c) and then correlate them to the HbA1c result. You can use those correlations to predict the HbA1c you will get (together with a statistical error if you are really nerdy) to predict the HbA1c from a current set of results.

I have done this and have had a bit of a laugh with my DSN by predicting the test result when they have been taking blood. Last two, which I have mentioned on the forum, have been very close.

The keys are to correlate your averages with your HabA1c results and to have an awful lot of patience.

 

[Sulfate]

I think you can use BG data to predict HbA1c but you have to be very nerdy to get anything near the lab result.

What you need to do is to record lots of results (several thousand) over several years (so you have several HbA1c results) so you can extract averages (overall 90 day average in the period before the HBA1c or a waking average, again in the 90 day period before the HbA1c) and then correlate them to the HbA1c result. You can use those correlations to predict the HbA1c you will get (together with a statistical error if you are really nerdy) to predict the HbA1c from a current set of results.

I have done this and have had a bit of a laugh with my DSN by predicting the test result when they have been taking blood. Last two, which I have mentioned on the forum, have been very close.

The keys are to correlate your averages with your HabA1c results and to have an awful lot of patience.

Thank you🙂

 

[Sulfate]

The more readings across a better representation of the whole day the more likely it is to correspond to hba1c.

Conversely a single reading a day will mean little, and a pattern of only the highest or lowest readings of the day will squew the estimate badly in one director the other.

Thank you 🙂