Is a lower HbA1c always better?

Not to muddle the thread with another thing tangential to variability, but for those curious about the idea that “higher blood sugar is always bad, the higher the worse it is”.

I’ve never found that for certain.

Here’s a large study done in Japan about cardiovascular risks, among diabetics and non-diabetics that are relatively lean (BMI 22-24), a1c 4-7%+

The ideal number seems to be about 5.7% for most factors, but 5.3% for coronary heart disease.

Anything below 5.3% are as bad as numbers above 6.3%.

We can spin off a different thread to talk about these kinds of studies, but I hope it gives some relief to those already in the (according to study) optimal range of 5.3-6.3. lower isn’t always better!



But the link you referenced is not entirely relevant for us here, because we all are dealing with known diabetes. The counter-regulatory hormones are completely different for diabetics and non-diabetics.

2 Likes

If you had bothered to read past the headline, 1,800 dignosed diabetics (at start of study. Their average a1c was 7%.) are also included. As well as over 1,000 (not yet diagnosed) people above an a1c of 6.5% (some far far above), studied over 10 years. If they aren’t diabetic, I’m not sure I understand the term anymore! Also, 3,800 people who began the study in at least pre-diabetic range 6-6.5%
All included.

If I went to a doctor with an a1c of 6.3%, and stayed there for years, I don’t think they’d say “You aren’t diabetic! You’re completely normal!”

As for hormones, too much insulin seems to have a negative effect. I don’t know about glucagon, but some diabetics have too much, others too little, so hard to say. :man_shrugging:
The lack of epinephrine response seems to be minor overall.
Outside hormones, tissue resistance to glucose uptake may be a factor.
As well as recurrent hypoglycemia (which would vary a lot among diabetics, being very rare for untreated ones).
Hypoglycemia is partially why I think the “normal” people with a1c <5% have such an increased risk of illness according to the study.

Effects of hyperlipidemia, obesity, cholesterol, blood pressure, smoking, etc. should all controlled for in this study.

There are other studies that back up this one, but a 10+ year study of 30,000 people with a1c 4-10%+ is quite comprehensive!

[SECTION EDITED into new thread: How Glucose Variability impacts HbA1c]

@Sam the study that found the close, linear relationship between a1c & average glucose was quite rigorous. But there are indeed about 3 big, legitimate criticisms of it.
One relevant one is that the original study was only of people with good control, so we don’t know if or how GV factors into it (other than the one study above)!

Later I’ll try to make a thread about A1c validity and meaning for you and Jen, and another one about BG/a1c ranges and their effects for Eric.

It’ll just make a mess of this Glucose Variability thread. :+1: :call_me_hand:

The study says this:

“The observed CVD risk among people with low HbA1c levels and no known diabetes is particularly important, because this increased risk could not be related to hypoglycemia11 “

They’ve reached the exact opposite conclusion.

" Furthermore, the patterns of the association between low HbA1c levels and each CVD subset differed from that for diabetes or high HbA1c levels. Low HbA1c levels were associated with an increased risk of stroke, especially hemorrhagic stroke, while diabetes and high HbA1c levels were associated with increased risks of coronary heart disease and stroke, especially ischemic stroke. The observed CVD risk in individuals with diabetes in this study was also consistent with accumulating evidence of diabetes as a risk factor for CVD.26,27 These findings emphasize a possible increased CVD risk among people without known diabetes and with low HbA1c levels. "

No conclusion is made regarding people with diabetes with low HbA1c levels. This is likely because the study inclusion criteria required that none of the individuals were diagnosed with diabetes. I assume that after their first A1c came back above a certain range, they were advised of their diagnosis, but I’d need to read the study a bit more to determine that.

[Edit: The study has one subgroup of individuals originally diagnosed with diabetes, but they do not list their hazard ratios or information by HbA1c. In addition, it looks like they only managed to obtain HbA1c information from 65% of all the individuals only at baseline. As A1c can change quite a bit for D people, that seems very insufficient and not particularly helpful].

The study’s limitations state:

" our results may not be applicable to other populations, especially Western populations, because East Asians tend to have a higher incidence of stroke and lower incidence of coronary heart disease compared with those in Western populations.35 Therefore, the nonlinear relation between HbA1c and stroke might be especially relevant to Asians."

While I hope this study is very valuable to Asian non-diabetic individuals with low HbA1c levels, I don’t feel that any of their conclusions are particularly relevant to me or my life. I do not think it was at all appropriate to extrapolate pieces of their conclusions to mean anything about people with D on this site.

While this may be a bit rude, I’m afraid that I would now have to individually look at every other study you’ve posted on this thread to be convinced that the results you’ve stated are actually the results of the study. I do not wish to spend my time doing that though.

4 Likes

I do not think it is rude to firmly state your position.

4 Likes

Since you are being snarky about it…

The terms “without known diabetes”, “without diabetes”, and “nondiabetic” are used a total of 27 times throughout the article, from beginning to end.

The term is also used in the very last paragraph…

In conclusion, both low and high levels of HbA1c were associated with a higher risk of CVD in a general Japanese population without known diabetes. These data support the notion that very low and high HbA1c levels may be markers for identifying people with an increased health risk.

Additionally, the conclusion states - “…our results may not be applicable to other populations, especially Western populations, because East Asians tend to have a higher incidence of stroke and lower incidence of coronary heart disease compared with those in Western populations.”

And as an additional limitation of the study, most of the participants in the study only performed a single A1C test, and did not test throughout the length of the study. So for most of them, their A1C during the months that lead up to their death was unknown. That is an extremely limiting factor for such a conclusion.

Please don’t assume what I am and what I am not reading. You have not been here long enough to know much about any of our members.

5 Likes

I mean, to be fair, those of us with Asian ancestry have been taking medicines, adhering to medical guidelines, and been given medical advice that was drawn almost exclusively from European cohorts for decades :slight_smile: Obviously there are limitations but for the most part, Asians are making it work.

So while the conclusions probably need a caveat, I’m not sure you would completely discount them just because they come from an Asian population.

3 Likes

I wrote this post up a few times and then deleted it. Let me be explicitly clear: I am not discounting the study because it happens to have occurred in an Asian country on an Asian population. However, the study itself states the limitation in applying the result to Western countries (I assumed due to dietary and lifestyle differences).

Perhaps there are individuals from Asian countries that look at this forum though. In any case, I feel like I listed multiple reasons why the study was problematic. I think the most significant is that no conclusions were drawn about low A1c levels and people with diabetes.

2 Likes

@Katers87, I’m sure that there are also many genetic factors that make Asian populations, on average, a little bit different than predominantly European ones as well too! (Sadly, while lifestyle factors are probably a bigger factor than genetics, lifestyle differences are certainly shrinking with every passing decade as people increasingly adopt a Westernized diet in Asia. :frowning: ) Which is probably not great for the health of people in those countries.

Anyways, my point wasn’t to say that there couldn’t be other limitations to the study – I actually haven’t looked through the study itself so I can’t comment on the study design just yet.

My point was just that people of Asian descent sort of “make do” with medical advice that is built for a different population and for the most part it works okay. Not perfect of course, but okay. So that particular caveat just seems a little less convincing to me. Especially when we’re talking about something like A1C, which is sort of a broad marker that probably is affected by many different factors, including lifestyle and genetics, so you’d maybe expect some of those differences to be counteracted or swamped in some ways.

The real issue is whether this V-shaped curve in CVD risk vs. A1C is replicated in other places. A1C has been so well studied I find it hard to believe this is the only study that looks at it. I’d want to know what other studies say before I take this as evidence that low A1C is harmful. That’s my main takeaway.

3 Likes

All the technical mumbo jumbo aside… this is pretty basic.

At zero blood pressure, nobody lives. At absurdly high blood pressure, nobody lives for very long. Somewhere in between there is an optimal range. This is common sense. The same applies to A1C, cholesterol, tax revenue, and everything else in the world.

5 Likes

Hey, thanks for splitting this for us too!

There are many many studies on different aspects of high & low BGs, risk for diseases, cognitive ability, etc. It’ll take a long time to go through. So let me work on it for a week or a few. Will come back when I get the time and continue the discussion with you all!

1 Like

A1C and CVD in Indians who are nondiabetic --a roughly linear relationship

In American Indians, A1C is only predictive of CVD above 6.5

In Chinese T2Ds, there’s a Jshaped curve in A1C vs CVD, with those below 6.5 and above 8 having significantly higher risk of CVD

In a meta-analysis of several studies, A1C below 6 for people with diabetes and below 5 for those without is associated with higher mortality:
https://bmjopen.bmj.com/content/7/7/e015949

In a European cohort, there’s a straight, linear relationship between A1C and CVD; the lower the number the lower the risk. Only 2 percent of the population had diabetes:
https://ebm.bmj.com/content/ebmed/10/2/57.full.pdf

Anyways, those are just the few I’ve found with a tiny bit of Googling. I’m sure I could find others. But the short answer seems to be that it’s complicated when it comes to CVD and A1C.

3 Likes

Here is a graph I just came across based on the DCCT that shows minimal difference between 6 & 7% A1c for T1D. _____8022081_orig

Well but that graph is scaled in such a way that you can’t see the differences. So for instance, the first dot seems to be around a risk of 1 and the second dot seems closer to 2. That’s a substantial increase in risk; it’s just that it’s dwarfed by the magnitude of the risk as you go up the curve… but if you’re in the 7 A1C group and go down the 6 A1C group, that reduction may not be minimal.

3 Likes

You are right, but I think it is useful to show how exponential the risk is after 7%.

For many, a result under 7% but above 6% should come as a relief, not shame stress or self-criticism. Like I see sometimes…

1 Like

And this one from DCCT, certainly more directly relevant to those here:
Excellent control (<7.2) over 30 years is associated with a dramatic reduction in risk on all major metrics:


Interestingly, the CVD and stroke factors were quite strongly reduced in this analysis, which I haven’t seen other places.

3 Likes

I think I had a lot more hypoglycemia a few years ago (before the cgm) at a higher A1c value, and I don’t think that’s necessarily uncommon. I’m not sure that it’s appropriate to attribute CVD at lower A1c values (if this conclusion is valid) to hypoglycemia.

If it’s not hypoglycemia, I would definitely be interested in knowing what is causing the lower A1c value. We’re making the assumption that the A1cs of these people align with their mean blood glucose level, but we don’t really know that (or at least, not that I can find). While these studies are supposed to control for anemia, are we sure that they have?

The study that @oni posted yesterday regarding high CVD at lower A1c values in non-D people (not known at least) was suggesting that there may be some other factor (aside from hypoglycemia) that’s influencing the CVD risk of people with low A1c values. I would be interested to know what that might be.

1 Like

no I agree with that. In general I think it’s great to hit that 7 and people should be really proud of that. It’s not easy for many to do that.

I also agree.

I initially considered them to be fasting a lot, but actually, caloric restriction is one of the few things known to predict extremely long, healthy lives.

However, just because low <5% A1c people are not diabetic, does not mean they don’t have one of many illnesses that encourage hypoglycemia. There seems to be many out there with a fasting BG in the 70s, getting down to 50s overnight.
I could imagine many daily issues with energy, cognition, exercise, etc. at chronically low levels.

But would need much more time to research how the brain, heart, blood pressure, etc. may react to being consistently right on the hypo/euglycemic line.

1 Like