Intensive Blood Glucose Control, Good or Bad

Someone posted a scientific study from a peer-reviewed journal on this message board recently which said that the same elevated blood sugar recommendations are now being made for type 1 patients as have now been made for a long time for type 2 patients: that is, strict control is not good, but rather, a blood sugar around 7 is better. If this is the case, then how does this affect the logic of the bionic pancreas?

That’s just typical inanity that comes from the medical community with regard to the disease. Since some people are not able to have tight control without severe hypos, and since hypos kill some diabetics, the medical community makes a blanket statement that tight control is not good for anyone.

Here is the study; the finding is that blood sugars which are too high (A1c 11.8%) or too low (A1c 5.6%) are dangerous:

J Clin Endocrinol Metab. 2014 Mar;99(3):800-7. doi: 10.1210/jc.2013-2824. Epub 2014 Jan 1.
Glycemic control and all-cause mortality risk in type 1 diabetes patients: the EURODIAB prospective complications study.
Schoenaker DA1, Simon D, Chaturvedi N, Fuller JH, Soedamah-Muthu SS; EURODIAB Prospective Complications Study Group.

Glycemic targets and the benefit of intensive glucose control are currently under debate because intensive glycemic control has been suggested to have negative effects on mortality risk in type 2 diabetes patients.
We examined the association between glycated hemoglobin (HbA1c) and all-cause mortality in patients with type 1 diabetes mellitus.
A clinic-based prospective cohort study was performed in 2764 European patients with type 1 diabetes aged 15-60 years enrolled in the EURODIAB Prospective Complications Study.
Possible nonlinearity of the association between HbA1c and all-cause mortality was examined using multivariable restricted cubic spline regression using three (at HbA1c 5.6%, 8.1%, and 11.8%) and five knots (additionally at HbA1c 7.1% and 9.5%). Mortality data were collected approximately 7 years after baseline examination.
HbA1c was related to all-cause mortality in a nonlinear manner after adjustment for age and sex. All-cause mortality risk was increased at both low (5.6%) and high (11.8%) HbA1c compared with the reference (median HbA1c: 8.1%) following a U-shaped association [P overall effect = .008 and .04, P nonlinearity = .03 and .11 (three and five knots, respectively)].
Results from our study in type 1 diabetes patients suggest that target HbA1c below a certain threshold may not be appropriate in this population. We recognize that these low HbA1c levels may be related to anemia, renal insufficiency, infection, or other factors not available in our database. If our data are confirmed, the potential mechanisms underlying this increased mortality risk among those with low HbA1c will need further study.

The natural inference from the high death rate in type 1 patients with an average A1c level below 5.6% might be that it is accounted for by the excess death rate for hypoglycemia, but the study authors ruled that out. They say:

“In our study, severe hypoglycemia events more often occurred in the lower compared with the higher HbA1c quintiles (P for trend < 0001); however, additional adjustment for severe hypoglycemia events did not attenuate the association between low or high HbA1c with all-cause mortality.”

Well, this is one explanation. Plus, many people struggle to get their A1Cs, anywhere close to 7, so devices like this would certainly be helpful even if extremely low A1Cs aren’t the goal. You say you keep your A1C in the 4s, I’m still confused why you do that if you don’t think strict control is worth it and might even be counterproductive.

Well, this study is completely new, dramatic, information for me, especially its recommendation that the best A1c value for type 1 patients to maintain is between 7% and 8%, rather than anything lower. Since most people can achieve an A1c in that range without a pump of any sort, this latest study has revolutionary implications for therapy, if confirmed.

It is interesting that this study made so little impact on patients when it was first posted a while ago, since they just went on with the usual approach of assuming that the closer they could get to normal the better. I suspect that the message of the study was simply too revolutionary for it to sink in for most people.

These studies are not saying, “The best A1C for Seydlitz is between 7% and 8%”. They are making a very weak statement that for diabetic patients in general, the best A1C is between 7% and 8%.

And they don’t even assign a cause. They find a very weak correlation, but no causation.

I could easily find a correlation between people who buy cigarettes and a higher incidence of heart disease. But I would be a fool to say that buying cigarettes causes heart disease. Obviously it is the smoking of them, and the reasons for that have been shown.

But this A1C study does not even give a reasonable explanation. The idea that counter-regulatory hormones for Type 1’s for hypoglycemia is foolish. Type 1’s do not have the same counter-regulatory hormones that T2’s have.

I hope this study does not change what you are doing for your control.

1 Like

It seems to me it is sensible for patients to try to fit themselves into the average A1c category which yields the lowest death rate. The study says, and I quote:

In line with our findings, the lowest mortality risk in the standard treatment group was associated with average HbA1c between 7% and 8% (19).

Please don’t let statistics mislead you!

Let me put this out as an analogy:

Which group of diabetics is more likely to be made of firefighters and race car drivers - a group with A1C’s near 5.5, or the ones with A1C’s between 7-8?

It is possible that people who have lower A1C’s are just in general more risk-taking, and the ones with the 7-8 A1C are more conservative. So the mortality may have nothing to do with the A1C.

You can can twist statistics into saying all kinds of things if you want.

But there is more support for this hypothesis than you presume. It has long been established that intensive control in type 2 diabetics produces counter-intuitive results, in that it increases death rates. This discovery was greeted with a chorus of scepticism when it was first made, and it was subjected to countless challenges and tests, and much to everyone’s surprise, confirmed. But they would always add, every time they added this surprising result, that since it had not been established as true for type 1 diabetics, type 1 patients should continue to try to normalize blood glucose levels as rigorously as possible, just as they used to advise type 2 patients.

But I always wondered why type 1 patients should be so different in this regard from type 2 patients? After all, in both cases, we are looking at the response of patients with defective self-regulation of glucose taking something to increase their insulin levels, so why should normalization of glucose by this method be dangerous for one group but not for the other? Well, this study just completes the logical inference, which is that regardless of how your insulin levels became inadequate, whether by genetic predisposition and lifestyle factors or an autoimmune attack on the pancreatic beta cells, trying to normalize glucose levels increases mortality risk.

This most recent study found that the distribution of mortality was distributed like a smile, peaking above an A1c of 11.8% and beyond at the high end and then again at an A1c of 5.6% and beyond at the low end, so the human body seems to be responding negatively to extreme values. This is not unusual in physiology, as a very high or very low blood pressure, for example, can both be harmful, but what is really surprising here is that even approaching normal glucose levels for diabetics increases mortality.

Because it is a totally different disease.

Type 1’s do not have the same counter-regulatory hormone responses to hypoglycemia that Type 2’s have. For Type 1’s, hormone responses like cortisol and glucagon are not triggered by lows like they are with Type 2’s.

And also, if I remember from that study, I believe the curve they drew for low A1C’s was inferred, not observed.

You express the concern that we cannot infer a causal mechanism from a mere correlation, but that also means that we cannot infer that the defective counter-regulatory response of type 1 diabetics to hypoglycemia must necessarily establish a mechanism making the damaging effects of lower blood sugar levels found in type 2 patients inapplicable to type 1 patients, since there could be other mechanisms explaining the harm in both groups. We are encouraged in that assumption because we have now found a correlation showing the same harm from approaching normoglycemia in both groups. In fact, it has been found that clinically low blood sugar should cause damage in both type 1 and type 2 patients, since a mechanism for that happening has been found in mammals: cf. N. Kajihara, et al., “Low Glucose Induces Mitochondrial Reactive Oxygen Species via Fatty Acid Oxidation,” Journal of Diabetes Investigation, 8 (6) 750-761 (2017). Perhaps, insofar as lower but still normal A1c values indirectly measure the frequency and duration of blood sugar dips below normal, Kajihara and colleagues may have identified the causal mechanism explaining why there is a higher death rate in patients with a ‘better’ average A1c.

Of course, science always prefers to explain via causal mechanisms and correlations rather than just by correlations. David Hume, the 18th century Scottish philosopher, drew attention to this problem by pointing out that constant conjunction is not causality. So, for example, if the London train always arrives in the Edinburgh station when the hand of the clock strikes four, the motion of the clock’s hand does not pull the train in, since even though the two events are always linked, they are linked as associations rather than as cause and effect.

But medicine often operates quite confidently with mere correlations even though it cannot explain the underlying causality supporting them. For example, aspirin was used for over a century with no one being able to explain why it reduced pain and fever, and it was not until 1979 that the biochemical explanation was found, but the mere correlation between taking aspirin and reducing these symptoms was enough for aspirin to be approved. Even the link between high blood sugar and diabetic complications is known as a correlation rather than being confidently established on the basis of a demonstrated causal mechanism. For example, B. Hoffmann, et al., “Hyperglycemia-Induced Glycosylation: A Driving Force for Vascular Dysfunction in Diabetes?” The FASB Journal, 1282.13 (1 April 2016) states that “Hyperglycemia … is among the causes leading to these vascular complications [in diabetes], although the underlying mechanisms of dysfunction are not well understood.” Or again, D. Deb, et al., “Critical Role of the cAMP PKA Pathway in Hyperglycemia-Induced Epigenetic Activation of Fibrogenic Program in the Kidney,” the FASB Journal, 31, 2065-2075 (2017), says that “Hyperglycemia is a major pathogenic factor that promotes diabetic nephropathy, but the underlying mechanism remains incompletely understood.” And yet, despite all this uncertainty even to this day about exactly how hyperglycemia causes diabetic complications (it doesn’t in hummingbirds, chickens, and about a third of those who survive diabetes for more than half a century), doctors in clinical practice still instruct their patients to avoid hyperglycemia to avoid complications. In fact, they have been doing so ever since the DCCT established the mere correlation between hyperglycemia and damage, with absolutely no indication of the mechanism by which it did so, many, many decades ago. So if we have been so ready to accept a mere correlation as sufficient grounds to establish the harmfulness of hyperglycemia, why not accept the mere correlation in the present study as showing the harmfulness of approaching normoglycemia in this special subset of humans?

1 Like

So then why do you aim for the 4s? I still don’t understand, especially when even most of us superstars on here arguing for tight control aim for the 5s.

As I always say, I have had an average A1c in the four range for the last ten years, so that is a report about my past practice, not my future intentions. Since the initial report that the approach to normoglycemia prompted a chorus of critical responses from the conservative forces in the diabetes patient punishment community, I expect that considerable research will reply to this latest result for type 1 patients, and I want to study those articles before coming to any definite conclusions for my own treatment.

If you are interested in looking into the possible mechanisms by which approaching normal blood glucose levels may be damaging to diabetics, theories about these were extensively developed during the long debate over the implications of the studies showing that it was harmful to type 2 diabetics, and I suspect that many of them may be reasonable candidates for explaining why this same phenomenon has been found in type 1 patients. Who knows, perhaps along with the cluster of genes disposing diabetics to develop hyperglycemia under the appropriate triggers there are some which cause them to need a higher blood sugar level to function healthily.

A limitation of the Saith, et al. study is that it is measuring blood sugar by random glucose testing rather than by A1c values, as the opposing study does, and the latter measure gives a more accurate picture of the glucose status of the patient.

But this is what happens in a science of correlations like medicine: the debate goes back and forth with studies opposing studies for a while, until it finally settles down to a new consensus. But since it has long been established that approaching normoglycemia is bad for type 2 patients, it seems reasonable to suppose that it is bad for type 1 patients, and it has indeed been shown that there are causal mechanisms by which lower blood sugar may damage the body.

In a science based on more formal frameworks, however, like physics, where you are studying the characteristics of universal structures like time, space, and matter, you can perform a single decisive experiment and settle something once and for all.

I don’t agree with the tenets of this :arrow_up: statement.

  • Looking at A1C numbers are not a valid way of determining if a person has spend time with normal blood sugar. The A1C is simply a measure of glycated hemoglobin. Stating that a low A1C equates to low blood sugar is supposition.

  • It is not normoglycemia that is harmful to T2’s. Dips into hypoglycemia that cause the hormonal counter-regulatory responses which lead to hyperglycemia are harmful. And time spend in hyperglycemia are harmful.

  • T2 and T1’s do not have the same counter-regulatory responses, so it is not reasonable to infer what is bad for one is bad for the other. It is a different disease.

I apologize for being part of the thread-hijacking here, but this conversation is very interesting. @Stemwinder_Gary, can you split this thread into a new one?

Eddie2: I agree this part of the message thread should be split. I would add, though, that the findings that tightening up control for type 2 diabetics below 7% was harmful did not depend on the blood sugar falling into the hypoglycemic range, but only into the lower range, i.e., between 5% and 7%, and the debate goes on as to what mechanism underlies this harmful effect.

If it is a truly random and sufficient sampling of BG values, I could argue that it is superior to an A1C test result.

A1C tests do not give any true indication of hypo’s. It only gives an indication of glycated hemoglobin, or lack of glycation. A random sampling of BG values - provided it is properly done at a variation of times and of sufficient size - would show actual highs and lows.

Additionally, A1C values are influenced by the lifespan of an individual’s RBC’s. They never factor this into the A1C value.

Calculating a person’s RBC lifespan could be done very easily by using a person’s hematocrit value and getting the reticulocyte life span from a corrected reticulocyte count table. And then doing a test for reticulocyte count, and calculating RBC lifespan like this:

RBC lifespan = 100 / (reticulocyte count / reticulocyte life span)

But it is clearly a possibility where a person who has shorter RBC lifespan may have other issues that would affect their health. A shorter RBC lifespan equates to a lower A1C, but may also indicate other health issues.

As a simplified way of expressing that thought, what they are saying is somewhat like a researcher saying this:

People with an iron deficiency, chronic kidney disease, and anemia who have an A1C below 6.0 have a shorter life expectancy than those without an iron deficiency, chronic kidney disease, anemia and an A1C above 6.0.

That :arrow_up: is a preposterous statement, but a person’s health conditions can reduce their RBC lifespan and reduce their A1C.

Thanks for the split, @Stemwinder_Gary !