Since the beginning of my career documenting my self experimentation on this blog I’ve been dealing with a borderling elevated HbA1c.
Here’s the first post on my HbA1c from March 2013 while doing my ketosis experiment:
The Effect of Nutritional Ketosis on HbA1c
Background:
Before I go further, for newer readers to the site and those unfamiliar with what an HbA1c actually is, I’ll copy and past what I wrote from that initial post:
- Everyday your body makes new red blood cells (RBCs) and breaks down old RBCS at an equal rate so you have a constant number of RBCs (assuming you are healthy).
- The lifespan of a RBC is somewhere between 100 – 120 days.
- The main function of the RBC is to bring oxygen to your body by using an oxygen carrying protein called hemoglobin
- When your RBCs are exposed to glucose, some of this glucose sticks onto the hemoglobin protein forming a sugar-hemoglobin
- The more glucose there is, the more hemoglobin proteins it sticks to, forming more sugar-hemoglobins
- Also, the longer your RBC is alive, the more it is exposed to glucose, hence it will have more sugar-hemoglobins
- The HbA1c test measures the average percentage of sugar-hemoglobins of ALL the RBCs in your body, giving you an idea of how much sugar they’ve been exposed to in the past 90ish days.
- So if you are diabetic, and your blood sugar is high all the time, you will have more sugar-hemoglobins, and thus a higher HbA1c.
- This is an important tool that doctors use to track diabetics because even though they can have a normal blood sugar measurement (which is a snapshot in time), if their HbA1c is high, it tells the doctor that their blood sugar has been poorly controlled on average in the past 3 months.
- HbA1c is a function of your blood glucose level and how long your red blood cells live. If this were put into an equation it would look something like this:
- HbA1c ~ (RBC lifespan) x (avg blood glucose concentration)
[otw_is sidebar=otw-sidebar-1]
Summary of My Previous HbA1cs
So in that initial post from March 2013 after puzzling over why my HbA1c didn’t improve despite being in documented ketosis for 72 days, I concluded that:
My low carb diet and subsequent low blood sugars actually help my RBCs to live longer. RBCs that live longer are then exposed to blood glucose for a longer period of time causing me to have a higher HbA1c. My ABNORMAL lab result actually reflects a HEALTHIER change in my body.
The curious thing is, there is virtually no data out there in the scientific literature examining the effects of a low carb or ketogenic diet on HbA1cs.
Inuitively it should look something like this:
- Low Carb -> Lower Blood Glucose -> Lower HbA1c
But from all the data I’ve just presented, it looks like it’s actually more like:
- Low Carb -> Lower Blood Glucose -> Longer RBC lifespan -> Higher HbA1c
I calculated that a HbA1c of 5.7% equates to an average blood glucose of 117 mg/dL, and so I wanted to prove to my self that my blood sugars weren’t actually that high. Since I didn’t have a continuous glucose monitor, I tested my blood sugar at 17 points over the course of a day, and found that my blood sugar ranged between 71-91 mg/dL with an average of 83.5 mg/dL.
That’s a 33.5 mg/dL difference from the calculated 117 mg/dL from my HbA1c which further supported my initial conclusion.
I tried a Carb Nite, a cyclic ketogenic diet, which also left my HbA1c unchanged at 5.7% in October 2013.
Later on I discovered a test that measured something called fructosamine that can also be used to measure an average blood glucose. At the conclusion of 10 weeks of Carb Nite, I measured both my fructosamine and HbA1c just for kicks. While my HbA1c was steady at 5.7%, here is what my fructosamine level was which muddied the waters a bit:
Fructosamine
This test is similar to the HbA1c in that it calculates your average blood sugar, except that it reflects values in the prior 2-3 weeks rather than the 3 months that HbA1c calculates. This is because this test measures how much glucose binds to proteins in your blood (primarily albumin) which has a turn over of around 2-3 weeks.
The reason I wanted to get this test is to obtain another marker for my average blood glucose that is independent of RBC lifespan, which would confirm my explanation of why my HbA1c is higher than expected.
Fructosamine = 229 umol/L (Normal 0 – 285)
According to the equations found here and here:
HbA1c = 0.017 x Fructosamine + 1.61
and
Estimated Avg Blood Glucose = { [ (0.017 x fructosamine) + 1.61 ] x 28.7} – 46.7
Therefore, plugging in my fructosamine of 229 umol/L, my HbA1c should be 5.5% and my estimated average blood glucose should be 111 mg/dL.
While these numbers are slightly better than what my HbA1c indicates, I’m not quite sure it confirms my hypothesis of an extended RBC lifespan causing a falsely elevated HbA1c.
In October of 2014 I tested my HbA1c again and it remained stable at 5.7%.
[otw_is sidebar=otw-sidebar-2]
When my lipidologist tested my HbA1c near the end of 2015 we saw that it had creeped up to 5.9%, and this is what she thought about it:
- Despite a normal fasting glucose and insulin, my HbA1c is up.
- HbA1c of 5.9 – Technically classifies me as pre-diabetic. This can explain my high particle count
- (This is also the worst it’s been. I think this could be related to my recent increased intake of carbs or possibly just due to differences in lab technique… but in any case, this is something I’ll need to keep a close watch on).
- Because I’m not eating any sort of high carb diet and am pretty fit, this tells her that there is clearly a genetic predisposition to insulin resistance.
- I can try to address this via life style, or can consider taking it a step further by taking a medication.
- She’s had experience using Metformin or Actos, to help address people with pre-diabetes and elevated particles and ApoB. In some cases the particles can drop significantly.
- There are studies on Actos lowering LDL-Particles and less data on Metformin, but a lot more data on how it addresses pre-diabetes.
- Given my current data, she is confident that I have borderline insulin resistance.
- My insulin isn’t elevated, so this tells her that I don’t have severe insulin resistance
After fiddling around for a few months with various supplements for my cholesterol I tested things again a few months later and saw that my HbA1c came back down:
HbA1c: When the lipidologist checked my HbA1c in July it had bounced up to 5.9 from being rock solid at 5.7 for the past two years, which made me worry a little. It’s back down to 5.7 now, despite all the changes I’ve made, including increasing my carb intake. This makes me think that the 5.9 value was likely due to lab variation. I normally get my blood tests through Labcorp, but the blood tests done with my lipidolologist went through her lab, Health Diagnostic Laboratory.
My Current HbA1c:
A slight bump up to 5.8%… with an estimated average glucose of 120 mg/dL, despite my finger sticks never going higher than 95 mg/dL while I tested during the first 15 days of the Whole 30.
[otw_is sidebar=otw-sidebar-3]
… Along came a reader
One of my favorite things about this blog is interacting with readers who comment on the blog or send me e-mails. I’m always learning something new and in this case, reader Saf directed me to a great resource:
hi caveman – you could try estimating your RBC lifetime, to validate your hypothesis, using the method here: http://www.precisionnutrition.com/blood-sugar-testing
What does Precision Nutrition say?
Things started out sounding very familiar…
First, folks with healthy blood sugar levels may have longer-lived red blood cells than those with poor glucose regulation.
The better you regulate and manage glucose, the longer your red blood cells can survive. The longer your red blood cells live, the higher your circulating hemoglobin. And if circulating hemoglobin is high, that will probably show up in the A1c blood panel.
But then things quickly get interesting:
How long do your red blood cells live?If longer-lived red blood cells can lead to higher hemoglobin A1c levels, despite healthy glucose levels, maybe doctors should be evaluating the lifespan and turnover of red blood cells in their patients.
Here’s a calculation for doing that. It’s just an estimation, as blood chemistry calculations aren’t perfect. Still, it may give a bit of insight into your personal red blood cell lifespan and it offers food for thought.
To do this calculation, you’ll need to know your reticulocyte count and your hematocrit.
Reticulocytes are early red blood cells. Produced in the bone marrow, they’re released into circulation as reticulocytes, and in a few days transform into fully mature red blood cells.
Reticulocytes can be used as a marker of red blood cell production.
For example, in someone who is losing blood (for example, from a bleeding ulcer or heavy menstruation) or in someone with short-lived red blood cells, the reticulocyte count may be higher. This is because the body will attempt to increase blood cell production to make up for the loss.
On the other hand, a low reticulocyte count can indicate that the body is generally happy with the amount of red blood cells or their lifespan, and doesn’t need to pump out as many reticulocytes.
The equation for determining how long your red blood cells are surviving is:
Red blood cell survival (days) =100 / [reticulocytes (percent) / reticulocyte life span (days)]
Here is an example:
Let’s say your reticulocyte count is 0.8% and your hematocrit is 45. Pulling from corrected reticulocyte count tables, the number for the reticulocyte life span (RLS) number would be 1.0.
Thus your equation would look like this:
100/[0.8/1] = 125 days
If your hemoglobin A1c number is a little higher than you’d expect given your current diet and lifestyle, and your red cell survival is longer than 120 days, your longer-lived red blood cells may be the reason.
Ah… now I know what I needed to do!
[otw_is sidebar=otw-sidebar-4]
How long do MY red blood cells live?
In order to calculate the life span of my RBCs I needed to know two things:
- Reticulocyte count
- Reticulocyte life span which is based on the hematocrit.
Check and check:
So, since my hematocrit is 44%, using this table we can see that my reticulocyte survival aka life span = 1.0 days
Plugging the variables into the equation for RBC survival from the Precision Nutrition post:
Red blood cell survival (days) = 100 / [reticulocytes (percent) / reticulocyte life span (days)]
Reticulocyte % = 0.4 %
Reticulocyte life span = 1.0 days
RBC Survival = 100/[0.4/1] = 250 days!
If you want to nerd out more about where this equation comes from, here’s a good resource.
[otw_is sidebar=otw-sidebar-5]
Final Thoughts
According to this my RBCs live more than twice the average life span of 100-120 days… which at first sounds amazing, but on second thought makes me doubt the veracity of this equation. While they probably don’t live 250 days, they definitely live somewhere between 120 and 250, which is enough to cause a false elevation in HbA1c.
In any case, I feel satisfied in finally getting to the bottom of why my HbA1c is what it is, and that is because my RBCs live longer than normal, and I now have data to prove this.
I probably won’t be testing my HbA1cs anymore unless things change drastically with my blood glucose.
This also made me think… the typical reference ranges provided by these labs are calculated from actual patients that have been measured. I wonder how many of those patients are young healthy active individuals. I mean if you have normal blood sugars and are reasonably fit, there is NO reason a doctor would even think about testing your HbA1c. It just wouldn’t be worth the cost… and therefore none of these subjects would be factored into the referene ranges of the labs. I imagine if you tested the HbA1cs of healthy athletes, they’d like higher than expected also.
I hope you learned something from this post… I know I did! So glad to finally put this issue to rest.
*Image found here
This is a little different question, but have you experimented with intermittent fasting to see if your numbers / possible insulin resistance improves?
Yes I have experimented with IF… Generally 14-16 ish hours of fasting. Doesn’t make a difference.
Also, after going through all of this with the HbA1c, I have to say that I don’t think I am insulin resistant. Blood sugars are fine. HbA1c is spuriously high and not accurate in my situation. Fasting insulin is low. Even post prandial sugars are ok. All of these point to me being insulin sensitive.
A bit confused.
Where did the 0.4 originate from? Your hemocrit was .44. Was this just the rounded down number or am I missing something?
Thanks for a terrific post.
Thanks for asking this question. I get where the confusion is coming from and added more into the section to make it more clear.
0.4% is my Reticulocyte count.
I used my Hematocrit to figure out my reticulocyte survival, aka reticulocyte life span, which is 1.0.
That’s where all the numbers came from.
Ketosis increases HIF1-alpha (HIF1a) levels. HIF1-alpha has been shown to induce transcription of EPO genes, which in turn increases red blood cell longevity.
https://books.google.com/books?id=lXkeDQAAQBAJ&pg=PA220&lpg=PA220&dq=ketosis+epo&source=bl&ots=hNZnFvcMuC&sig=lz8OYiEomN_bANK5x76vhqd6slc&hl=en&sa=X&ved=0ahUKEwiAusLjo97QAhWKj1QKHR9LCkEQ6AEILzAD#v=onepage&q=ketosis%20epo&f=false
https://www.ncbi.nlm.nih.gov/pubmed/20204773
https://en.wikipedia.org/wiki/HIF1A
Well, you made me say Wow
🙂
I am opposite my A1c went from 6.1 to 5.3 in two months of keto but can’t get my fasting blood sugars down from 115. so confusing
That’s strange! Are you in documented keto? With pee sticks or blood sticks?
What would happen to your HbAIC if every 12 weeks you made a blood donation?
Regularly donating blood usually has a beneficial effect on insulin sensitivity
http://www.clinchem.org/cgi/pmidlookup?view=long&pmid=15976100
That’s interesting. Didn’t realize the connection between insulin sensitivity and iron.
I imagine that donating blood would lower HbA1c simply by increasing RBC turnover.
Donating blood lowers A1C numbers. I have hereditary hemochromatosis and confirm that. Iron overload affects blood glucose regulation because it attacks the pancreas amoung other organs. 60% of people with iron overload from hereditary hemochromatosis end up with diabetes.
I am following lchf diets 11 months Fbs are in between 70 to 85 but hba1c is higher value.
Before lchf hba1c 6.3
After lchf hba1c 6.3
Ppbs also within 110
Random bs within. 90
It’d be interesting to see what your CBC and Reticulocyte count looks like to estimate your RBC life. That could potentially explain things.
6.3 does seem high though.
try donating blood before the next HbA1C
With the amount of blood I give for all of my blood tests… it’s at least 1/2 a pint!
Are you concerned by your low reticulocyte count? Did you get it checked again?
Great work here! I love diligence in citizen science. I have discovered on keto now 16 weeks that my blood glucose 10 tests per day, vary greatly with hydration, exercise, total fat intake and number of meals. Fasting glucose 5.0-5.4 ave. Then a walk and water only and goes up to 6.5. But my post prandial response to low-carb eating keeps me in the 5 to 5.8 range always. These numbers are not what constitutes diabetes. Steady state post prandial is of no health concern. I know people on metaformen who are very happy if their blood glucose stays under eight postprandial. Anything under 7.7 mmols is considered good. For those of us testing our blood sugar so frequently and seeing consistently mid five readings, the HbA1c is of little value. Too much glucose in the blood and too much insulin in the blood is the major concern of diabetes. Changing out fuel sources changes everything in terms of statistical data geared to a glucose fed population.