Hba1c is confounded not only by red blood cell turnover, but also by the activity of fructosamine 3-kinase (FN3K). Surprisingly, lower Hba1c due to higher FN3K activity could actually mean MORE glycation if your downstream metabolism of 3-deoxyglucosone is not in order.

Mastering Nutrition Episode 012: What Is Measuring Our Hba1c REALLY Telling Us About Our Blood Glucose and Diabetes Risk?

Visit Us
Hba1c is confounded not only by red blood cell turnover, but also by the activity of fructosamine 3-kinase (FN3K). Surprisingly, lower Hba1c due to higher FN3K activity could actually mean MORE glycation if your downstream metabolism of 3-deoxyglucosone is not in order.

In response to popular demand, this week is glycation week. In this
episode, I discuss the strengths and limitations of using Hba1c to
measure our cumulative recent exposure to blood glucose and diabetes

Many people will be familiar with the fact that variation in red
blood cell turnover confounds this measurement. Less well known is that
variations in the deglycating enzyme fructosamine 3-kinase (FN3K) also
confound the measurement.

Counter-intuitively, if you have a higher rate
of this deglycating enzyme but a lower rate of downstream metabolism of
3-deoxyglucosone, your lower Hba1c could actually mean MORE glycation. I
conclude that Hba1c is a useful test, but only in the context of a
bigger picture put together with more information.

Listen on ITunes or Stitcher.
Click here to stream.
Right-click (control-click on the Mac) here and choose “save as” (“save link as” on Mac) to download.
Subscribe in your own reader using this RSS feed.

Read the show notes.
Leave a comment.

Want transcripts? Sign up for the CMJ Masterpass with this special link to get 10% off.

Read on for the show notes.

Show Notes for Episode 12

In this episode:

0:50 Sneak peaks of PUFA Report Part 2 on Snapchat (chrismasterjohn)
1:10 Facebook Live this Saturday, June 11, 2:00 PM Eastern time (remember to adjust for the time zone if necessary)
1:50 Cliff notes on Hba1c
4:14 Update on the audio quality
7:55 Diversify your working positions PSA: I recorded this in vajrasana
11:00 Intro to Hba1c as a marker of cumulative recent hyperglycemia
16:46 variations in red blood cell turnover
18:10 deglycation of Hba1c by fructosamine 3-kinase (FN3K)
20:38 genetic variation in FN3K activity (and maybe nutrition and metabolic rate?)
26:20 Hba1c does not seem to causally contribute to diabetes or its complications
31:57 small aldehydes, particularly methylglyoxal, and maybe 3-deoxyglucosone, do causally contribute to diabetes and its complications
38:15 lower Hba1c due to higher FN3K means MORE formation of 3-deoxyglucosone!
40:15 strong insulin signaling and polyphenol-rich fruits and vegetables are likely to protect against 3-deoxyglucosone accumulation
41:55 practical conclusions about using Hba1c

Research and Other Links Discussed in this Episode

This is the microphone I use. Please let me know in the comments if adding Garage Band's compressor for the first time in this episode seems to help the consistency of the audio quality.

Start here for all my writings on glycation.

This is my recent Examine.Com Research Digest editorial.

Here is a paper on red blood cell turnover (free), and here is one on genetic variations in fructosamine 3-kinase (also free).

Here is research showing FN3K knockout mice do not develop diabetes (free),
but that methylglyoxal causes acute hyperglycemia (free), and severe diabetes when administered chronically (free).

Here is a review on 3-deoxyglucosone and methylglyoxal (free) and a paper on the metabolism of 3-deoxyglucosone to 3-deoxyfructose (paywall).

Do you use Hba1c to measure your health status? How has it helped you? Does this episode change your interpretation at all?

Visit Us

You may also like


  1. Hi Chris, I’ve been reading your stuff for years but somehow missed that you had started a podcast! I’m really enjoying it.

    I was wondering two things (since I am working through the podcasts in order you may have already covered them and I just haven’t gotten there yet, so my apologies if that’s the case):

    First, do you know about the work of Dr. Joseph Kraft? He suggests that looking at glucose measures (postprandial and fasting) are not actually the right thing to be measuring in diabetes and that instead we should be looking directly at insulin response. Not only that but that we should be looking at the postprandial pattern of insulin response. As I understand it from him, since diabetes is a disease of insulin, blood glucose levels are just the secondary measure and they only show up as elevated once the disease has progressed to an advanced stage. I find this approach very interesting and it caused me to think differently about our use of fasting and postprandial glucose (coupled with HbA1c) as the measure of diabetes.

    Second, someone else’s comment above made me wonder about our ability to measure how AGEs in food affect the AGEs in our body. I’ve seen the statements before that animal products contain more AGEs and that high heat cooking creates more AGEs, but does that then directly impact the actual number of them in our body? I wonder if this is another instance of assuming that just because a food has something that our body also has or needs, that eating that food will directly impact the quantity of that thing in our body – like the misunderstanding of cholesterol. The old thought process of – Food has cholesterol; our blood has elevated cholesterol; therefore the food we consume is what’s causing the elevated cholesterol. Then there was that whole “realization” that the body also makes cholesterol and that the cholesterol in our food really doesn’t impact our blood cholesterol (I actually wonder if this “realization” will also be the thing that exonerates saturated fat, since we make that too.) So my question is, do you know of any studies that show that the AGEs in our food directly impact the AGEs in our body?


    1. Hi Caitlin,

      I don’t agree that diabetes is a disease of insulin. T2D is a disease of energy overload and T1D is a disease of cellular damage. I think the glucose-centric models and insulin-centric models are both too simplistic.

      Regarding AGEs, first, most dietary data is worse than useless so most of what is said about food AGE content is false. Second, there is no convincing data that eating AGEs is the problem. I have written extensively on this, and I recommend following the links in my “Start Here for AGEs” post, linked above as “Start here for all my writings on glycation.”


      1. Thanks for your response! I’ll definitely take a look at your stuff on AGEs. That’s what I thought since our dietary data is pretty useless most of the time.

        I’m interested in what you said about T2D being a disease of energy overload. Could you point me to some of your writings that expand on this more?

        – Caitlin

  2. Your voice seemed quite clear. I bet you re-positioned around 6:40-6:50. Otherwise I didn’t miss a word.

  3. Chris,
    Do you think that substances with antiglycating activity, like carnosine, have the potential to reduce the adverse effects of diabetes?

    1. Probably, but I would focus more on limiting the accumulation of the precursors to AGEs.


      1. I found this reference.
        “Foods high in protein and fat, such as meat, cheese, and egg yolk, are rich in AGEs. Foods high in carbohydrates have the lowest amount of AGEs. In addition, increased cooking temperatures, like broiling and frying, and increased cooking times lead to increased amounts of AGEs.”

        Unfortunately, I am on a low-carbohydrate diet secondary to pre-diabetes. However, I do cook at low temperatures.

  4. Can you describe your experience with foam rollers? Are they good for those just starting out, and then later graduate to something else?

    Are some rollers better than others? Rumble Roller? Perform Better Elite?

  5. Great explanation of the relative significance of HbA1c! Can any aspect of AGE formation be measured in standard lab tests? I'm thinking that the only way to get a glimpse into potential AGE load is to measure oxidative stress markers which would be defending against them (such as cysteine/cystine ratio as an indirect marker of glutathione level). Of course patient history is of paramount importance in providing context, as well as HbA1c, insulin, fasting and post-prandial glucose levels.

  6. Hi Chris, Nice talk! In the paper for Dicarbonal metabolite formation, in just the first category – Glyoxal, all the following were listed, and this is just one category!

    1. Lipid peroxidation.
    2. Degradation of glycated proteins.
    3. Oxidative degradation of serine (via glycolaldehyde).
    4. Monosaccharide degradation.
    5. Degradation of nucleotides.
    6. Food and beverages.

  7. Hey Chris,
    First just let me say that I really am enjoying your podcast. Especially since I am a biochem geek. I am involved full time now with robotic automation as it pertains to laboratories that perform some type of nucleic acid extraction and purification (which at this point is just about everyone) and other similar functions such as NGS, but I got my start in clinical molecular diagnostics/esoteric testing for the monster clinical reference shop that is LabCorp. More specifically I spent 12 years at their C.M.B.P. facility in R.T.P. NC before moving on to robotics. So I MAY be a bit biased, but I really am glad that people like you, and others I follow, enjoy helping with my continuing education, and I really can't say enough good things about what I am learning through you. So, Thanks A Lot. It is very much appreciated.
    I have a question about insulin and I wasn't exactly sure where to ask it but I thought I would just try here since it sort of ties in to this podcast and the last few. I know you are super busy so if it's just easier maybe to answer this in a podcast episode discussion or some other way than here, then I completely understand. No worries. So here goes…

    If we consider a person with healthy metabolic function and we have them on 2 separate diets at different times for extended intervals, both being based in healthy ancestral guidelines containing nutritionally dense foods with complex carbohydrate sources like vegetables, good fats and moderate proteins, and an appropriate amount of calories, but we adjust the macros so that in one instance the person is under more normal carbohydrate metabolic dominance and then the other time we adjust the macros so that they become nutritionally ketotic for an extended period of time, then COULD the level of insulin production during NK be enough to provide adequate and robust insulin signaling equal to what it would be on the more carbohydrate dominated diet? Here I mean minimally carbohydrate dominant and certainly NOT the standard "Western Diet" I have been thinking as I have been writing this and I suppose the simple answer would be, no, the level of insulin release shouldn't be as robust as in even a barely minimal CO dominant diet because insulin levels would never easily fall off enough to alleviate their pressure on hormone sensitive lipase. Obviously with protein intake here being the primary fulcrum around which these 2 competing diets could/would swing. I suppose my real question here would be, do you believe it could be possible, through genetics or training and diet or a combination of both, to be so sensitive to insulin that its signaling strength would be just as robust on either diet? My first thought would be if it is possible then there is probably a genetic bias occurring.
    Forgive me if this sounds like a truly odd question. My mind tends to get lost aimlessly down odd dead end intellectual alleys late at night much of the time.
    Anyway, Thanks Again!

    Dan W.

    1. Hi Dan,

      Thanks for your kind words.

      I think the answer is a clear no. Unless the mechanism of ketogenesis is MCT oil or exogenous ketones, then it by definition requires low insulin signaling to be achieved.

      I would also say that I would not regard as ketogenic diet as ancestral. A tool, for sure, but it is only a possibility that indigenous tribes of the arctic were chronically in a ketogenic state and even that may be dubious. Even if correct, those groups are outliers and human evolution occurred in the presence of abundant plant food include considerable starch (as evidence, for example, by salivary amylase duplications).


  8. Coincidentally today I was reading what C. Kresser wrote about HbA1c before finding your blog post https://chriskresser.com/why-hemoglobin-a1c-is-not-a-reliable-marker/
    A few years back when I was suffering with hyperinsulinemia symptoms (today I recognize them) my HbA1c was 4.8 and fasting blood sugar 76. My endo sure was proud of me despite my awful skin and ruptured blood vessel followed by thrombosis. Evidently Warfarin didn't make me feel any better.

    1. I'm sorry to hear about your health problems. Warfarin inhibits vitamin K metabolism, which could contribute to awful skin and ruptured blood vessels. I have written about vitamin K extensively, so if you haven't already read those writings check them out.

    2. Hi Chris, thank you so much for the reply. I probably had very high homocysteine and insulin back when my HbA1c was 4.8 and then had endothelial damage. I was on warfarin for 1.5 years after that (which made everything worse of course), then 1 year on Gamma E. I found your work a couple of years ago when I wasn't thinning my blood anymore, tried to take K2-MK4 but couldn't get past the 1st dose of 250mcg – too much pain all over the body (de-calcification of soft tissues?). But the hormonal effect was amazing. I wish I could have that. Well, my point is that doctors trust HbA1c too much as you said so well on this elucidative podcast.

Leave a Reply

Your email address will not be published. Required fields are marked *