A1c and CGM don't agree, I've finally discovered why, w/ update


I started using a continuous glucose monitor (CGM) almost 10 years ago. Reviewing my CGM data quickly became a daily habit for me. Time-in-range, time low, average glucose and standard deviation data numerically characterized the quality of my glucose metabolism.

I soon observed that my A1c (glycated hemoglobin) and its associated average glucose did not agree. My A1c number consistently sat about 0.5% above the average measured by my CGM and my finger-stick meter.

My first attempt to resolve this discrepancy was a hypothesis that my fingerstick meter, the device I used to calibrate my CGM, was inaccurate and the reason for the A1c/average glucose discordance.

I began to do three fingerstick meter measurements at the same time as a lab glucose blood draw. My meter, however, agreed closely with the lab draw. Last December, for example, three fingersticks came in at 92, 88, and 88 mg/dL (5.1, 4.9, 4.9 mmol/L). I employed three different pokes using three different fingers across both hands.

The average of these three equated to 89 (4.9). The lab plasma glucose weighed in at 89 mg/dL or 4.9 mmol/L. I’ve repeated this exercise four times per year for the last three years and found similar results. My meter was not the cause of the A1c/average glucose disagreement.

I’ve done lots of reading about this topic and the best, yet still unsatisfying reason, was Dr. Irl Hirsch’s explanation that the A1c just does not precisely equate to a discrete average glucose value. We’ve all seen the doctor exam room posters that display a 6.0% A1c that indicates an average BG of 126 mg/dL (7.0 mmol/L).

Yet, Dr. Hirsch has written that people with a 6.0% A1c experience average blood glucose levels anywhere from 100 mg/dL (5.6) to 152 mg/dL (8.4). People on the high side of this distribution curve, Dr, Hirsch called high glycators. At least the medical literature told me that my experience was not that unusual. But it didn’t reveal why my A1c seemed to always ride about 0.5% above my glucose experience.

I started seeing a naturopathic physician recently to seek help implementing lifestyle changes to address a coronary artery disease diagnosis last fall. Her recent lab order included a complete blood count or CBC panel. The results showed that my hematocrit, hemoglobin, and red blood cell count all came in at just below the lower lab threshold. I have iron deficiency anemia!

This was unexpected and news to me. I quickly searched my lab history and was surprised to see that the CBC panel had not been run since 2011 and 2009. And those labs were ordered by one physician who repaired a hernia and another doctor who was conducting a study. The 2009 numbers for hematocrit, hemoglobin, and red blood cell count were all within the lab range but at the low end. The 2011 showed all these measures trending downward with one of the three sinking below range. I’ve been trending anemic for almost nine years now.

None of my three endos during that time ordered a CBC panel. Yet I distinctly remembered commenting to them about the A1c/average glucose discordance. One of those endos diagnosed me with sinus tachycardia or rapid heart-rate. She quickly prescribed a beta-blocker drug and showed no curiosity about the root cause. Turns out anemia often drives a rapid heart rate. This makes sense. Anemia impairs the blood’s ability to distribute oxygen to all the body’s tissues. A threat to oxygen levels leads to the heart pumping more quickly to overcome that deficit.

Inappropriately cold hands and feet are another symptom I reported to my docs over the years and these comments elicited little follow-up or curiosity. Cold extremities are also a symptom of anemia.

I take pride in my consistent monitoring of the medical literature to inform health choices, but I totally missed this one. I guess, since I didn’t know I was anemic, I didn’t pay much attention to the literature that discussed anemia and its effects on A1c, but it’s clear now that I should have.

Here’s a 2014 National Institute of Health (NIH) study that describes my A1c/average glucose disagreement that puzzled me for the last 10 years.

These two figures tell the story.

Here’s the simple conclusion made in this study.

Iron deficiency anemia elevates HbA1c levels in diabetic individuals with controlled plasma glucose levels. The elevation is more in patients having plasma glucose levels between 100 to 126 mg/dl. Hence, before altering the treatment regimen for diabetes, iron deficiency anemia should be considered.

I’m happy to have discovered this connection after so many years. I’m disappointed in my doctors’ failure to pick up on this basic health measure. If they had invested a tenth as much attention to symptoms that I repeatedly reported to them as they were hyper-vigilantly distracted by potential hypoglycemia, they would have detected my hiding-in-plain-sight anemia. I’m also disappointed that I didn’t put 2 +2 together sooner. Live and learn!

I’ve started a daily iron supplement and I suspect in the months ahead that my A1c will finally come into agreement with my average glucose as measured by my CGM and finger-stick meter.


That is surprising that CBC, CMP, Lipids, UA were not run as standard lab work as part of your normal annual physical exam.

I find the trending of the numbers from the various test components over a period of time to be more relevant for the individual as compared to a one time test.

CBC - Complete Blood Count (with differential)
CMP - Comprehensive Metabolic Panel
UA - Urinalysis


My endos consistently ordered the lipid panel. Since I refuse statin treatment, it’s as if they wanted an abundant record of their due diligence to protect against any legal exposure.

I can’t lay the responsibility for this failure at the feet of any one doctor since three endos over a nine year period failed to discover my anemia. This suggests a systemic practice failure.


When I was diagnosed with anemia, my doctor postponed the next a1c because she said it would not be accurate. Because my red blood cells were not functioning normally.


Great post @Terry4. In depth research and analysis - and you just explained one thing I had no idea of. I’ve been anemic for 31 years now. Originally thought to be caused by cyclophosphamide given to me in an attempt to slow down a motor neuron disease I have, I was put on beta blockers 15 years ago.

Fast forward to today - I weaned myself off beta blockers and my resting heart rate is about 85 BPM, increasing to 100+ when I stand or walk. Aside from T1D, I have 2 other inflammatory auto-immune disorders. My doctors now describe my anemia as “anemia from chronic disease”.

Sadly, they don’t seem inclined to give me anything for it (no iron infusions or supplements). I wish you success with your treatment!


Glad you received appropriate care. I don’t think it’s wise to ever give up checking and double-checking the basis for all medical decisions. I think blind faith in doctors is hazardous to your health. The patient plays an important role and, of course, feels the consequences, whether good or bad.


When I last changed doctors, I chose a DO. Doctor of Osteopathic Medicine
The are educated and license the same as MD. But the basic of treatment is holistic, not just prescribing meds (they do prescribe meds when needed)
She seems to be a good fit for someone who does diet and exercise only.


Something I have not mentioned here, but I was always confused as to why T1 would be missed diagnosed as T2, I know it happens. My diagnosing doctor clearly laid out that since my age was 63 it would generally mean I was T2, but since my a!c was 12.0 that I could be a type 1 (1.5, Lada). She explained my options was to be treated as a T2 with meds (she did not like the meds at that time) or revise my diet and exercise and use Metformin, I could go straight to insulin, but she rather I work on My diet and exercise to see if I responded. If my numbers did not normalize in 6 months we would start insulin and do the antibody testing. (Insurance would pay for it then). This was back in 2009,
Plus for some reason, more of the people my age were diagnosed T1 that T2,

I did a intensive diabetes education course at a local hospital based on low carb woe. The course was great got my A1c down to 6. Unfortunately the hospital was bought out by Johns Hopkins and they eliminated the program. I have no idea what really went down in all of that, but there was a law suit by the program against JH, and JH lost and paid a financial settlement.

So I took a course that had 27 hours of classes, 3 individual meetings with a nutritionist, diabetes education, psychologist and exercise trainer. But that was closed down and the standard education classes became. Required.
It was sad, the hospital charged $750 for the course, my insurance paid $600, and a local family trust paid the rest (or all if you did not have insurance.)
The first day they put a pile of food on the table and made us learn how to count carbs. There were T2’s and two adult newly diagnosed T1’s in the class.
We had to eat to our meeters

Everyone should have access to all that good information.


OMG Terry, I get at least two CBC’s done in one year! Man, your doctors really dropped the ball. Shame on them. I am very happy that you now have some answers.


My PCP does full testing at annual visit, and endo has access to results. So I review them with both. But for a couple years, I did not have a PCP, and my endo ordered them.

I agree they should be done yearly.


@Terry4 Thanks for the article and charts!

I have thalassemia, a blood disorder that causes red blood cells to be abnormally small and odd shaped. It shows up as anemia in CBC tests, and the MVC will be below normal range.

I also have consistently higher a1c than my cgm calculates, but only by a small about, 0.3 higher.


@Marilyn6 @Terry4 - I see my GP at least 4 times a year, in addition to several other docs. Every one of them runs a full CBC with each appointment, likely 8 or 10 a year. The numbers change as fast as the weather forecast. Creatinine values change even more quickly; when they think I might have an issue it can dramatically change in as little as 36 hours.

All (most) of these test results are highly dynamic, and over a 30 year period they can look back and trend your “normal” range for results.


I didn’t know that anemia can elevate A1c. I’m also currently iron deficient and have dipped in and out of low lab values for some of the blood levels (MCV, MCH, etc.) for the past five or so years. However, my red blood cell count trends towards high, and my hemoglobin and other values have always remained normal. I’ve had these tests done regularly for years, but doctors have never seemed concerned about these results just outside normal range until I began seeing a cardiologist for unexplained significant tachycardia (that has persisted for about eight years). She is trying to rule out all possible causes, so prescribed iron supplements to get my iron levels up.

My last A1c was 6.8%, which honestly surprised me and was higher than I had expected. Maybe the fact that I’ve had low iron levels and low lab values for a while now explains why, even when eating low carb, I’ve never been able to hit an A1c below 6.0%. I’ll be very curious what effects taking iron supplements may have when I get my A1c done in three and six months. My cardiologist plans to take me off the beta blocker I’ve been taking in about six months, so I’ll also be curious if I’m able to stop without the tachycardia coming back.


After reading the cited study and others, it appears that a iron deficiency anemia does elevate the A1c beyond what the underlying glucose exposure justifies. I’ll be curious to see if this holds true for me. I’d be interested to read about your experience, too.


Hey maybe you can break that 6,0 barrier next time. Good luck.


I so agree you can’t have blind faith in Doctors. They try their best and just like everyone else have faults. Some a lot more than others.

The problem I see is you don’t know what to ask to question what they say? So you don’t know what to look into. Just like the anemia, how do you even know to think that?

So you’re not going to ask for a blood test. That seems like a real lack as my endo did a complete work up every year, but if you don’t have it done, you just don’t know.


I Learned about anemia from my trainer, I had work with her for four years.
She noticed my energy was low and I was looking a bit pale. A bonus for working with her so long.

I have had good experience with my doctor now. But I have fired two who I thought were incompetent. I was also married to a Nurse (we are still friends) so she is my back up medical advise. So between a trainer and an ex wife, they know me pretty well.


It won’t be next time as I’m not eating low carb right now and I don’t think I can hit an A1c like that without eating low carb. But yes, perhaps if the low iron and low test results is the cause of a slight discrepancy in A1c (this time around I was 0.4% above what Dexcom said I should be), then perhaps I can break through at some point. I have been thinking about low-carb a lot lately, and doing some research, but it remains to be seen whether I can put together a sustainable menu without dairy, eggs, avocado, tomato, (possibly salmon), or processed meats that is adequate for frequent travel as well (as I can’t eat out so have to bring most/all of my own food).


That’s some great research you’ve done. Kudos.


I’ve had a couple of my better doctors, run CBC’s on me and specifically comment that they were looking to see if I were a “low glycator” or “high glycator”.

I do not think this is yet standard but just an extra cross-check done by a couple of the better docs.

While all my docs have done the BMP panel all the time, only a few have done one-off CBC’s.