HbA1c Home Meter

Status
Not open for further replies.
D

DaPa

Member
Relationship to Diabetes
At risk of diabetes
are they worth purchasing, or should I stick with the GP?
 
Depends what you want to see and how often you want to check...

I get 6 monthly HbA1c from the docs but also test my blood sugar levels daily so I can tell what the results "should" be.
 
Personally I couldn't have managed to reduce my BG without a monitor, although I rarely use it these days unless I'm eating something new. Since you are only at risk rather than being a full blown diabetic it may not be worth the expense. If you do purchase one, the Gluco Navii and the Spirit Tee 2 are both reasonable and have cheapish testing strips. If you do decide to get one it's useful to keep a log of what you've eaten, the amount of carbs consumed and your BG levels before eating and 2 hours after.
 
I have the BG finger prick one.

I was curious as to getting the device that measures the HbA1c, 3-month readings
 
Ooh, I didn't know you could buy them. I will investigate, thank you
 
DaPa said:
are they worth purchasing, or should I stick with the GP?
Click to expand...
I doubt they're worth purchasing. But perhaps that's just me. Hypothetically, if you had such a thing how would knowing the results (presumably more often than you can get them from your GP) help you?

It seems clearer how regular BG tests can help: for example you can see what your readings are before lunch (which might suggest a morning walk would be sensible, or a different breakfast). I'm less clear about HbA1c.
 
I have a dexcom and it provides estimated a1c over 14, 30, 90 days. V useful
 
I’ve used an app for years logging my BG results that calculates HbA1cs.
My sensor app can do the same.

They always show a little higher than what is fed back from the surgery.
 
I have been testing for a while now and have put all the readings into a data base. I have found good correlations of HbA1c with 90 day averages prior to the HbA1c test, of both of all readings and waking readings.

For me,

HbA1c = 10x (90 day average waking reading ) - 16

If I had an HbA1c taken today I would quite confidently expect a result of around 50 mmol/mol.

I repeat my plea for any of you data nerds out there to look at your data to see if they see the same. If there is a widespread correlation it would provide a fairly simple method estimating HbA1c from finger pricks.

PS... I'm not panicking over the HbA1c of 50. There are no short term issues and at my age I'm not over concerned by long term problems. So provided it stays there I'm happy and I will know soon enough if it begins to creep up.
 
Docb said:
I repeat my plea for any of you data nerds out there to look at your data to see if they see the same. If there is a widespread correlation it would provide a fairly simple method estimating HbA1c from finger pricks.
Click to expand...
Doesn’t work here. Last a1c 70, things have improved a bit since and libre estimating 63. Seems about right, it’s probably 65ish.

Average bg 10.0 (from libre) x 10 -16 = 84 no way

If I use the 6-9am libre average 3 month reading of 8.1 I do get 65 though
 
I wouldn't bother. Get yourself a BG monitor that provides you with at least an estimated HbAC1 reading over a monthly period. That should be enough to help you manage things.
 
Wow, @Lucyr. Your three month, 6-9 am libre average, which is a similar statistic my 90 day waking average predicts an HbA1c from my equation that is in the same ball park as the libre! Amazing.

I like it.

Anybody else prepared to join in?

Deepest apologies to @DaPa for hijacking your thread!!!
 
Tdm said:
I have a dexcom and it provides estimated a1c over 14, 30, 90 days. V useful
Click to expand...
Libre gives the same sort of thing, but (after the novelty) is it really useful? Doesn't feel useful to me day to day. It's useful for setting my expectations of what an upcoming HbA1c blood test is likely to be (roughly), but apart from that I never really look at it.
 
I don't find the HBA1C estimates are always under for me whether it be mySugr or Libre Link.
 
Bruce Stephens said:
Libre gives the same sort of thing, but (after the novelty) is it really useful? Doesn't feel useful to me day to day. It's useful for setting my expectations of what an upcoming HbA1c blood test is likely to be (roughly), but apart from that I never really look at it.
Click to expand...
I find it useful to give an idea of the longer term trends. Knowing that my last a1c was 70 and that I’ve been trying to make improvements, and seeing libre say 63 means I know my changes are starting to make the bigger picture improve.
 
Bear in mind any HBA1c based on finger pricks will be an estimation because it does not take into consideration what happens between pricks and may miss peaks and troughs.

Libre also does an estimate but, again, this is an estimate. My Libre consistently estimates 10 less than a blood test.

Both are useful things to track but I considered them different. I will track how my Libre estimate is progressing but will not think of it as the same thing as the real HBA1C which I track separately.

Whilst I track them, I make little change in my management based upon the HBA1C results. I use the individual BG readings/Libre graph to adjust my management.
 
Last edited:
This is the foundational study from the working group which empirically established the standard best fit between blood glucose and HbA1c: https://diabetesjournals.org/care/a...nslating-the-A1C-Assay-Into-Estimated-Average

This study produced the best-fit formula which is used by BG monitors and CGM's to estimate HbA1c from BG.

The best-fit correlation is pretty good but not exact:

m_zdc0080870710001.jpeg


Even with precise, accurate, fine-grained measurements of BG, your HbA1c is unlikley to be what yr BG monitor/CGM predicts, and it's common for the variance to be as much as 20%.

It's not because of any inaccuracy in the test procedure, rather it's mainly due to the fact that the average age of red blood cells varies from person to person, generally in the range of 90 - 110 days. If two people with exactly the same BG levels have a 20% difference in their red blood cell age, then their HbA1c values will differ by 20% also.

As CGM's become more prevalent and more accurate the usefulness of HbA1c becomes less & less. It was always just a clever trick for roughly estimating avg BG levels. As it becomes more possible to measure them directly the need for the rough estimate reduces.
 
Thanks for the reference @Eddy Edson.

They give an equation for the correlation as:

AG = 28.7 × A1C − 46.7,

AG is the average blood glucose in mg/dl and A!C is HbA1c as a %, so to compare that equation with mine you have got to sort out the pesky units!

It was easier to write a query in my database to convert my units from UK standard to the foreign stuff and got a comparative equation for my correlation as:

AG = 23.5 x A1C +3.9

At this stage I would suggest that 28.7 and 23.5 as the two slopes isn't too bad an agreement. Beyond my brain these days to do any stats but I doubt, bearing in mind the scatter, that they will not be statistically significantly different by any reasonable measure. The difference in intercepts is a different matter and no doubt has something to do with the testing method and in particular the testing frequency and times of testing relative to eating.

One of the reasons I looked at, and got, a correlation with waking BG average was to take testing time out of the equation. It was the one test least likely to be affected by all those extraneous variables which influence blood glucose levels. Obviously I cannot sort the waking data from the dataset in the paper so cannot make a comparison with my waking data.

Numbers are much more fun than opinions don't you think.
 
Status
Not open for further replies.
Back
Top