Average mmol/l to HbA1c Estimate Type 2

[Alineden]

Hi. Can we work out an estimated HbA1c based on a rolling average of mmol/l?

So if I get an average mmol/l of 10 mmol/l using a continuous blood glucose monitor, is there a conversion to predict approximately what my 3 month HbA1c will be?

 

[Rachox]

There’s a converter on this website, to be found here:

Average Blood Sugar Level Converter - HbA1c

Use this calculator to convert HbA1c to Average Blood Sugar Level. HbA1c indicates what your average blood glucose level has been in the past 30-90 days.

www.diabetes.co.uk

If you wear a CGM it will calculate an estimated HbA1c, well at least the Libre does.
Do bear in mind that any of these only estimates the number. I had my annual review recently and the Libre estimate was 4 below the lab test. The estimates are good to see trends though.

 

[KennyA]

Hi. Can we work out an estimated HbA1c based on a rolling average of mmol/l?

So if I get an average mmol/l of 10 mmol/l using a continuous blood glucose monitor, is there a conversion to predict approximately what my 3 month HbA1c will be?

I'm wary of "averages" - the HbA1c counts the actual number of glycated red blood cells, and that number gives you an idea/estimate of where blood glucose levels have generally been over the last 2-3 months. It isn't an average of anything.

I think most CGMs have a sort of "estimated HbA1c" function. However - CGMs are measuring glucose levels in interstitial fluid, and estimating a blood glucose vlaue from that. Then they're extrapolating from that estimate to a forecast HbA1c.

They seem to vary a lot in accuracy, however, which I don't think is surprising. Last time I used a CGM (Libre) it was predicting an Hba1c that was significantly lower than actual HbA1c at the next test - if I remember correctly it was a 32 predicted against a 38 actual. That's over 15% out.

Given that these are normal BG levels and it didn't do too well on those, I did wonder if it might give better accuracy at higher BGs, and be calibrated to work better in those levels.

I did find this small study which used CGM results from non-diabetic people. The blurb says that it was run mainly because CGM manufacturers didn't know what "normal" looked like for newer models.

Continuous Glucose Monitoring Profiles in Healthy Nondiabetic Participants: A Multicenter Prospective Study - PMC

Use of continuous glucose monitoring (CGM) is increasing for insulin-requiring patients with diabetes. Although data on glycemic profiles of healthy, nondiabetic individuals exist for older sensors, assessment of glycemic metrics with new-generation ...

pmc.ncbi.nlm.nih.gov

 

[Alineden]

Thanks for your replies. I appreciate that trends only offer estimates, but it would be good to know, for instance, that a certain average level of mmol/l would get me an HbA1c of approximately 50mmol/mol. It would be like breaking things down to a daily or weekly target rather than waiting 3 months to find out that you need to be trying harder.

When I was first diagnosed in 2018 I became almost obsessed with trying to eat a low carb diet. I got my HbA1c down to 41. I thought I had it cracked. But over time, my diet has slipped back into bad habits. In the past I learned a lot about what I should and shouldn't eat by using the Freestyle Libra. I decided I needed to bring my focus back so I've been using Dexcom for the last couple of weeks. I've been quite shocked at how high my average is.

 

[KennyA]

Thanks for your replies. I appreciate that trends only offer estimates, but it would be good to know, for instance, that a certain average level of mmol/l would get me an HbA1c of approximately 50mmol/mol. It would be like breaking things down to a daily or weekly target rather than waiting 3 months to find out that you need to be trying harder.

When I was first diagnosed in 2018 I became almost obsessed with trying to eat a low carb diet. I got my HbA1c down to 41. I thought I had it cracked. But over time, my diet has slipped back into bad habits. In the past I learned a lot about what I should and shouldn't eat by using the Freestyle Libra. I decided I needed to bring my focus back so I've been using Dexcom for the last couple of weeks. I've been quite shocked at how high my average is.

I agree, it would be very useful, but it doesn't seem to work that way. It's a bit like those stats they keep using for football these days. You'd think you should be able to use "shots at goal" and "possession" to predict actual goals, but the evidence is the other way.

I set myself a target of trying to keep all fingerprick testing results below 6.0mmol/l. I was (and still am) on around 20g carb/day. I don't remember why I chose that figure, it could have been something I read on the forum. At the time I was testing before all food and at +2hrs, as well as first thing in the morning. It worked really well for me, and although I was still seeing some sixes my Hba1c was normal within four months.