back to Overview

HbA1c test results don't tell the full story

November 12, 2015 by Scott Johnson

It’s often viewed as the number to rule all numbers. But hemoglobin A1c (HbA1c) test results can be misleading and don’t tell the full story. As I learned in my teens, the HbA1c test shines a light on things I was trying to hide.

Overall, It’s not good at getting to the details of blood sugars, but when used with other pieces of information it can draw attention to (sometimes unseen) problem areas in our diabetes management, and that’s a good thing.

How do HbA1c test results work?

Let’s take a quick look at the basics of the HbA1c test. A certain amount of sugar in your blood sticks to your red blood cells and can’t be unstuck. It’s there for the life of the cell, which is, on average, about 8-12 weeks.

Those red blood cells in your body are constantly recycled, and by checking your HbA1c value every 8-12 weeks (or as often as recommended by your doctor – the ADA recommends at least twice a year), you get to see a fresh new grouping of them.

So – A higher blood sugar for a longer time means more sugar on more cells – which means a higher HbA1c. Get it?

Ideal HbA1c range

HbA1c goals are very individual, which makes sense. We’re all different, right? Of course, there are reference values as a guide, and that’s a good place to start.

The ADA suggests an HbA1c of 7%, but also say that “more or less stringent glycemic goals may be appropriate for each individual.”

Why have different goals? Because, as you know, there’s a lot to consider with diabetes. Avoiding lows (hypoglycemia) while pushing for lower A1c’s is really important because low blood sugars are immediately dangerous. It’s simply not safe to push for a very low HbA1c if you’re having a lot of lows, so your doctor might adjust your target up a little bit.

Additionally, your doctor should know if you have any conditions that may affect your HbA1c, such as sickle cell anemia, heavy bleeding, very low iron levels, kidney failure or liver disease.

Ideally, you’ll work together with your doctor to find a comfortable HbA1c goal that keeps you safe and healthy.

HbA1c’s dirty secret

The HbA1c is an average.

Dust off those old math lessons and think about how averages work. You can have two very different stories – one with pretty stable blood sugars, one with wild fluctuations – that equal the same HbA1c.

Dr. Stephen Ponder, a pediatric endocrinologist who’s been taming his diabetes monster for about 50 years, explains it very well in this graphic we’ve adapted (with permission) from his book, Sugar Surfing.

Graph adapted by mySugr from Sugar Surfing showing different groupings with same A1c

It’s a great reminder that HbA1c’s are not very good at details, and you still need to take a closer look.

Take the time to dive into your numbers and make sure your HbA1c is coming from details you want. Standard deviation is one way to do that. Standard deviation shows how much your BG’s are fluctuating.

According to another favorite, Think Like a Pancreas by Gary Scheiner, a good standard deviation is less than one-third of your average. Ideally, you want an in-target average with a low average standard deviation.

How to Estimate your HbA1c

You don’t even have to crunch any numbers. We do it all for you (along with average and average standard deviation) in mySugr.

Sketch showing mySugr Logbook's estimated HbA1c feature

From now on, the top right of your graph will display your estimated HbA1c – assuming you’ve logged enough blood glucose values. How many values do you need?

An approximate average of 3 BG’s per day over seven days

Sketch showing a calendar with the data requirements for mySugr Logbook's estimated HbA1c feature

Without enough data, the estimated HbA1c level is not displayed, as you can see here. The progress bars around the circle fill as you log more BG’s. Keep feeding mySugr data, and you’ll have your estimated HbA1c in no time.

But please, keep in mind that this is an estimate and may differ from the laboratory value. Enter more values and mySugr can do a better job estimating.

mySugr Logbook screenshot with estimated HbA1c

Keeping an eye on your HbA1c can make a big difference, as even small improvements have shown great reductions in the likelihood of complications. With the estimated HbA1c feature we hope we’ve made it easier to know where you are so you can get where you’d like to be.

And can we toast to no more nasty surprises when your doctor delivers your next HbA1c result? Cheers!

Pro Tip: Keeping your overnight blood sugar in the green is a quick and easy way to improve your HbA1c. Why? Think about it. It’s a large portion of the 24-hour period where you don’t have to do much, or deal with many variables. Sure, there will be some work up-front to troubleshoot problems, but once you get your overnight BG’s stable and on target, you fix one-third of your HbA1c. Easier said than done (like most things in diabetes), but totally worth the effort.


Diabetes App

Do something good for your diabetes with mySugr

mySugr makes diabetes suck less – during crazy adventures or even just a normal workday.

  • Andy Collins

    Beautifully written – I’ve never stressed about one number! Your App has made living with Type 1 easier, for that, I’m really thankful GREAT APP!

  • Thanks a million, Andy! We really appreciate you stopping by. And I think you’ve got a great approach – not stressing about one number.

    Keep in touch, and let us know how we can help more.

  • Brant Puttkammer

    I agree this is a great feature! I just wanted to get clarification and ask if the readings input are given weight based on the time between tests. So if I log a reading at 6:00 am and then another at noon, does the reading at noon have a 6 hour weight or 1/4 of a day when used in the calculation?

  • Brandon Haller

    Very nice article. Thank you. I am a recently diagnosed Type 2. I have found the mysugr app an extremely helpful tool to monitor and manage my blood sugar levels. My doctor has instructed me to test my blood sugar levels 3 days per week, 3 times on each of those days. But, to use the new HbA1C feature one needs to test blood sugar much more frequently, as noted above. In estimating HbA1C why not use the the most recent 7 days worth of entries, or a minimum number of entries, even if those observations aren’t on consecutive days?

  • Thanks, Brant!
    The formula is a bit more complicated than that, even. But I haven’t come up with an easy to digest way of explaining it yet. Give me a little more time and I’ll get back to you with more details.

  • Hi, Brandon! Thanks for taking the time to read it! I’m so happy to hear you’re enjoying mySugr Logbook so far.
    At first glance, that makes a lot of sense. But estimating an A1c based on BG readings is complicated. A1c’s are looking at blood sugar levels in your body 24-hours of the day, every day of the week, whereas BG readings are just one tiny little snapshot moment in time.
    I’m working to pull together more details around the logic behind why we need what we ask for, and will post that here soon.
    Thanks again!

  • Chris MK

    I get a more accurate Hba1c from a 4 week or 90 day average. My last result was 6.2%, my excel workbook calcilated 6.27% and mysugr was 6.3%. If taken weekly this value tends to shift if you have variability in your glucose levels and less representative of the short-medium run.

    I’m taking 6-7 blood tests a day at the moment and mysugr has been great as an electronic diary. Using the accu chek aviva has been accurate in terms of my results, tested other meters like Carsens, TruYou and One touch vario but daily test results have been sognificantly off compared to my long term use of Accu chek. Meter accuracy is important in estimating a reliable result, as well as having a good number/cross section of blood tests to achieve this.

    If you’re talking specifically about control the the devil is in the detail, intensive monitoring.

  • Hi, Brant!
    I spent a little more time digging into the details, chatting with the developers, our risk management group, and medical advisors.

    The approach we’re using doesn’t apply a weight or relevance metric to any specific individual BG reading.

    We’re coping with the significance of individual measurements by requiring a relatively large amount of data that has been spread out over a (again, relatively) large period of time.

    In other words, we try to distribute the weight of enough readings so that no single reading upsets the balance and creates an inaccurate A1c estimation.

    Does that make sense?

  • Hi, Chris!

    Thanks for sharing. That really lines up with what we’ve found, too. That the A1c is a long-term tool that needs to look at a big pool of data over longer periods of time.

    I spent some time over the past few days chatting with our developers, risk management group, and medical advisors to learn more about how we arrived at the decisions we made.

    We tried to shorten the time requirements as best we could, and also minimize the number of readings per day to the lowest number possible, and the cutoff where we started to see too much variation was where we had to draw the line.

    We have a lot of anonymized data to work with, and looked at the formula results versus actual A1c results that people have logged, and it all correlates very nicely until we get below the minimum required data.

    So, Brandon, with only 3 days per week, and 3 checks per day on those days, we don’t think we can provide an accurate estimation, and decided that no estimation was better than a broken one. I think that’s the best approach, and I hope you’ll also agree.

  • Hi, Brandon! I embedded a reply to your question inside my comment to Chris – wanted to make sure I let you know about it. Thanks!

  • Brandon Haller

    Yes, thanks Scott. I work with statistics and agree that no estimation is better than one that may be faulty. You have developed a reasonable approach, and I appreciate your transparency. Please keep up the good work!

  • Chris MK

    I also suppose you can get analytics from your user database, you can produce a Hba1c number, however, the more data your consumers produce the more potentially accurate their result will be.

    From what i’ve established with Accu chek, I have got a very close estimation of the lab test based on 6-7 tests a day. When I was using the Carsens meter my predicted results tended to be more off track. So different meters also certainly affect this. Quality of data and measurement is highly important in a reliable data set.

  • Gary Purvis

    When should you take samples. First thing in morning, before meals, two hours after meals.

  • Pingback: Highs & Lows | HbA1c Estimator()

  • Hi, Angela!

    Great question, and thanks for sharing.

    I can think of one or two possibilities, and it’s nothing you did wrong – it’s just the crazy nature of diabetes and how we have to become professional detectives and problem solvers to uncover mysteries like this.

    Blood sugar checks only capture a snapshot of that single moment in time. It’s wonderful that you’re checking so often (many people don’t), but you’re still only capturing 6 to 8 moments of the day, compared to the A1C which is looking at 24h of the day. And it’s completely possible that those checks are missing problem areas, especially if you’re always doing them at the same times through the day.

    For example, if you’re testing before a meal and then some hours after the meal, it’s possible that your blood sugar is spiking quite high from the meal then dropping back to a lower number by the time you test again.

    There are a couple of ways to uncover some of these things. One is to juggle your testing schedule around and get a picture of what your blood sugar is doing at times that you’ve not normally checked before. Another is to ask your doctor about wearing a continuous glucose monitor for a little while. Some offices have loaner systems just for situations like this. Additionally, there’s just been an announcement for the approval of Abbott’s FreeStyle Libre Pro device in the US and their consumer system is available in much of Europe (though I’m not sure where you live).

    Hopefully, this is at least a little bit helpful and makes sense? Please do keep us posted about what you uncover!

    Thanks again for reading and commenting, I appreciate it!

  • Hi, Melissa!

    Oh goodness, what a situation you and your family are in. I’m sorry to hear about that.

    It’s frustratingly common in the world of diabetes to have to become a tireless advocate for yourself (or your loved one), and that often comes in the form of pursuing a number of healthcare providers until you find a team that suits your needs.

    In this case, a few questions come to mind. Has the endo team seen the up and down records? Or are they just looking at the HbA1c results? I can’t imagine an endo group turning away an adolescent dealing with low BG’s like that. If so, that brings me to my next question – are there other endo teams or diabetes specialists you can turn to? I know we’re often limited by our insurance coverage and options, but perhaps they (your insurance group) can even suggest some other providers who can help. Additionally, if there’s no endo’s available, there are some good internal medicine providers who might be able to help, too.

    I hope this helps a little bit, and again, I’m so sorry to hear you guys are having to fight for care. That stinks.

  • MarkAttack

    Hi Scott. Just trying out the app for the first time. Looks promising so far, but is there a way to have the HbA1c estimate be based on my Dexcom CGM numbers which feed in through the Health app? I only test 2-3 times a day with my glucometer, so my CGM numbers would be a much more reliable A1c estimate.

  • Hi, MarkAttack!

    Thanks for giving us a try! Unfortunately, at this time we’re only able to provide estimated HbA1c’s based on glucometer entries.

  • MarkAttack

    Thanks for the response Scott. I think this would be a very useful feature to add, so please consider doing so

  • Isabel

    It would be nice in the options for types of there was an option for hypoglycemia. I have non-diabetic reactive hypoglycemia. Which is not an option, but I like the app to have a visual of my sugars.

  • Very good point, Isabel – thanks for the feedback!

    Can you tell me a little more about what your diabetes management routine looks like on a daily/weekly basis? I’d love to better understand how it relates to what we already know.

    Much appreciated,

  • Isabel

    Well I’m still pretty new to this diagnosis (almost a year) and I still have not seen a endocrinologist or any kind of specialist yet, so I’m sure I can’t help at all. Recently I’ve been having to check more and more because when I eat, my sugars drop and for some reason when I don’t they jump. I’m still trying to figure all of this out especially this weekend where I will jump/ drop 30+ points in less than an hour. It is making me really sick because my body can’t handle the rapid extremes. The app is however giving me a visual for myself and so I can show my doc if I can see them next week. Sorry I couldn’t help much.

  • That’s very helpful, and I’m sorry to hear you’re having such a rough go of things. Keep us posted, and hopefully, you can get in to see an endo or specialist before too long. I’m sure they’d be a lot of help.

  • Hi, Linda!

    Great question! And you’re right, it’s certainly easier said than done.

    Diabetes is a complicated thing to figure out, and I’d definitely recommend recruiting the help of a specialist or endocrinologist. Additionally, if you have access to test strips, start experimenting and gathering some data. For example, wake up just to do a quick BG check at a few different times of the night (doesn’t have to be all on the same night) to get a feel for if your BG is rising gradually all night, or maybe it’s stable most of the night then spikes up at a certain time.

    If you’re limited in test strips, you might be able to ask for an increase in your prescription or perhaps buy some affordable over the counter ones just for the purpose of this experiment.

    Some doctor offices even have loaner continuous glucose monitoring devices you can use – that would be a tremendous way to see what’s going on.

    Stick with it, Linda! You’ll get to the bottom of what’s happening and then you and your doc can come up with a plan.

    Best of luck!

  • andy957

    Terrific article, and still informative even after all these years of having to deal with ‘the monster’. Thank you, Scott! 🙂

  • You’re welcome! Thanks for taking the time to stop by and give us a read. 🙂

  • Kelli Mceldowney-Cackler

    Scott thank you for this article. I learned a lot of good information. My doctor has me taking my blood sugar only once per day. in order to do the A1C test would I need to take it 3 times A-day for maybe a week?

  • Hi Kelli! Thanks for taking the time to read! Yes, that’s right. In order for mySugr’s estimated HbA1c calculations to work, the system needs a minimum of 3 checks per day for about one week. But keep in mind, the more information you can provide, the better.

  • Khalid Ibrahim

    Quesiton to Scott K. Johnson please. Today, I randomly checked my Glucose level through strip after 1 hour of having breakfast( just after having some herbal medicine for my spinal stenosis which was happened to be very sweet), my blood sugar level came up as 238mg/dL. Immediately after i went for HBA1c test and it came up as normal (4.9%). After dinner blood sugar level was 98mg/dL. Should i be worried at this stage ?? thank you.

    History. My mom recently been diagnosed with Type 2 and i have four elder Brothers

  • Hi Khalid, thanks for reaching out. While I’m aware that people without diabetes do experience blood sugar fluctuations, I’m not a trained medical professional and am not qualified to give you an answer. I’m sorry. I have to recommend that you ask your healthcare team about this.

  • Iva Vrzic

    Great article, thank you! I’ve been living with type 1 for 13 years now and have always kept my A1C around 7 (anywhere between 6.7 and 7.3). But have been wondering about the A1C’s “dirty secret” myself, so thank you for explaining that further and offering tips on better control over that!

  • Pamela

    Interesting article. the a1c averages made me realize, after all these years, just how out of control Diabetes can get.

  • Terry Aman- Sumner

    Thank you for this article & shedding some light on how this test works. I love the app & find it helpful in tracking my numbers etc as I work to control these numbers as I am a visual person.

  • Hi Terry! Thanks for stopping by. You’re very welcome and I’m so glad this was helpful for you.

  • Hi Pamela! Yes, it’s a reminder that keeping ourselves educated is an important part of staying healthy with diabetes. Thanks for stopping by!

  • Robert

    I was just recently diagnosed with type 2 diabetes. I’m still learning about it. This app is helping me more than you know. Thank you Scott.

  • Alan

    Why is it that you have to test 3xday to get an HABC1 and not at least once a day for 7days? Why can’t I export report using the free app vice the Pro?

  • Thanks for the questions, Alan.

    One test per day doesn’t give enough data to provide a reliable estimated HBA1c.

    You can export your data as a CSV file with the free version. To get the nicely formatted PDF report or an Excel file, you have to be using mySugr Pro.

Newsletters, blog posts and tips & tricks from the monster workshop.