I have also thoroughly looked through the medical literature (even ones behind paywalls) after I kept reading about diabetics being ridiculously anal about their variability.
I have not found any evidence at all to prove variability is a risk factor on its own.
New diabetic guidelines say higher A1Cs do not significantly increase mortality and morbidity risk compared to hypoglycemia risk of lower A1Cs.
Guidelines for GCK MODY, who are permanently in the 120-140 range, say no treatment is necessary as no known extra risk.
There is evidence high variability is correlated to more hypoglycemic episodes (spiking lows and overcorrections) which are the true risk doctors worry about.
You also have numerous diabetics now living well after 80yo. Those people had very poor treatments until the 1980s, and spent half their life with awful control.
I have found no description of pathways where temporary glucose spikes during the day harm tissues in any significant way. If you avg 300 5% of the time, then maybe you are just 300 5% of the time. Maybe that’s 95% less risk than someone who is 300 all the time. Or maybe more or maybe less. No evidence at all about it.
After finding absolutely no proof that momentary spikes cause any health issues, I agree with others that being anal-retentive over variability is stupid.
Stress and anxiety are well known killers. I’d take the certain health benefits of less stress over the completely unknown benefit of tight control, and the greater risk of hypos with tight control. That’s a very easy choice. Very.
I stumbled upon this research article, and I think there is a portion of the article that is interesting and relevant to this thread.
" Glycemic variability has been considered a HbA1c-independent risk factor for diabetic complications.63–65 Although, in another analysis of the same population, they found that GV showed significant impact on the MPG–HbA1c relation- ship only in type 1 diabetes mellitus (T1DM) patients, leading patients with high GV to higher HbA1c levels for the same MPG;66 they concluded that mean glycemia and HbA1c rather than postprandial or GV showed stronger, consistent associations with cardiovascular disease risk factors.62 As many trials demonstrated, microvascular and macrovascular complications are mainly dependent on dysglycemia, which has two components: chronic sustained hyperglycemia and acute glycemic fluctuations from peaks to nadirs.65,67–69 Both components lead to diabetic complications through two main mechanisms: excessive protein glycation and activation of oxidative stress. However, a study involving patients with T1DM failed to find a relation between high GV and levels of oxidative stress in these patients.70 The explanation offered by the authors was that, unlike type 2 diabetes mellitus (T2DM) patients, T1DM patients are not sensitive to GV as a stimulator of oxidative stress because of different underlying pathophysiological mechanisms and the more accurate laboratory method (mass spectrometry) used to measure urinary excretion of prostaglandin, which gave lower results than in previous studies."
See, there is only about 1 or 2 citations in that article (which is about a completely different topic about validating eAG to HbA1c) that say variability has any significant effect.
And they are a bit old. If you search through papers that cite them, some just assume there are risks, but no studies. There are.more studies showing glucose variability has no effect, than studies that show there is an effect.
This does not support the hypothesis that targeting GV would be beneficial in reducing subsequent secondary cardiovascular events.
But there are many more like this study. All null results.
It’s a lot of effort and stress for tight control (and extra risk for dangerous hypoglycemic events) for having only negative results in multiple studies.
@oni, you mentioned before your belief that targeting GV has extra risk for “dangerous hypoglycemic events” because of tight control.
But you are making two assumptions that I don’t think are valid. First, you assume that targeting GV requires increasing hypos. That is not necessary. It is possible to try to obtain tight control around a higher average BG control point.
Second, you assume that increasing hypos is automatically bad. But, again, that is not automatically true. There have been, on this forum, numerous discussions and quotes of multiple studies, whose conclusions in general are that hypos do not have long term impact on brain chemistry. On the other hand, for every minute of high BG we know we pay a price.
I don’t think one can blindly assume that increasing hypos is automatically bad, if it has other positive effects. We are totally willing to have a decent amount of mild hypos if we can lower average BG and GV. The danger, imho, is (a) the number of severe hypos, that might increase the risk of a deadly one, and (b) the amount of time in hypo, that might increase the risk of hypoinsentivity.
The largely inadequate data available in the era before CGMs simply could not establish the veracity of a GV hypothesis. It remains to be seen if a modern study can. Because of its very nature, it will always be a difficult hypothesis to prove, requiring high quality data and a subtle experimentation process. For obvious data reasons (easier to obtain because requiring a grosser level of detail) is easier to make the case for the impact of long term glycemic variability.
So, in the absence of data, one is left with assumptions, and risk management. We can decide that, in our opinion, it is best to go one way or another. But, imho, we should not either assume or affirm that one approach is better than another.
For instance, in my opinion, being more often in mild hypo is better than more often hitting highs above 160. While I certainly think that your opinion of the nefariousness of hypos in general vs high BG is an opinion (vs fact) that is reasonable to hold, I don’t agree with your premises and I don’t think one should assume they are a fact.
for every minute of high BG we know we pay a price.
I’ve yet to see proof that the human body is so delicate and non-resiliant to not be able to recover and repair from one minute of high BG.
But I also reject popular concepts like the Linear No-Threshold model of radiation, as I believe the human body is fundamentally resistant to (even strengthened by) minor damage.
The danger, imho, is (a) the number of severe hypos, that might increase the risk of a deadly one, and (b) the amount of time in hypo, that might increase the risk of hypoinsentivity.
Yes. That is the main danger and my point. I don’t see where we disagree here.
you assume that increasing hypos is automatically bad
Nope. I believe exactly what you just said you believed. Hypos can kill you now. Hypers not so much. I lived mostly fine and symptom-free with BGs between 300-700 for at least 3 months pre-Dx. Couldn’t say the same if I was even 50 for 3 days.
you assume that targeting GV requires increasing hypos.
It does for me.
CGM readings I often see on diabetic sites are people trying to keep their BG always around 75-85. That allows only 20mg/dL of room before serious, immediate, issues begin. My range is much larger, because I have dropped 100pts in 30 minutes. With food. And no basal. Just a couple units of bolus to correct high BGs before a substantial meal. I stopped Humalog corrections completely since the last time.
That kind of tight control would put me in a coma before I could find a bag of sugar. No thanks.
we should not either assume or affirm that one approach is better than another
I’ve only shown that research shows only null results and why I personally think stress and anxiety are more significant factors in destroying quality of life than allowing high variability. Other people can do whatever they feel is best for themselves, knowing there is no evidence (yet) to support it.
They’re essentially saying that your A1c and mean blood glucose levels are more important than variability.
In addition to those two measurements, the article states that sustained highs and severe peaks/valleys contribute to both microvascular and macrovascular complications. So these should be avoided. I was not aware of the oxidative stress component. I found that very interesting.
Addressing those 4 items seems like a very reasonable and logical approach. Of course, if you’re minimizing the severity of the peaks/valleys, that’s one way of addressing glycemic variability.
@Katers87
The article’s references do not exactly back up those claims as you think.
The referred studies on complications with sustained highs were again, A1c studies. Not intra-day or daily studies. If you find one, please let me know!
Actually, it would be good to see how significant effects are between 5-7 A1c. Project for another day…
The referenced article does not show proof that either glycation or oxidation from GV is significant enough to cause complications. Just that those effects are likely to exist (they did no study).
Here is another interesting, very recent study on uncontrolled T2 diabetics. A1c (and its variability) are a predictor of complications, but weak evidence GV is an independent predictor:
Nevertheless, no direct link between acute fluctuation of glycemia and oxidative stress has been consistently reproduced in human studies
the role of glucose variability in the development of vascular complications in diabetes remains unresolved. Translating into hard clinical end points the laboratory evidence that implicates glycemic fluctuations in complication risk has thus far proven difficult. We are also in no position to know whether reducing this variability will lead to a reduction in excess risk—if it exists in the first place. Knowing whether there is benefit in reducing variability beyond that of simply reducing the risk of hypoglycemia is of utmost importance, as it raises the possibility of allowing patients to help avoid hyperglycemia-related vascular disease without running the same risk of hypoglycemia that a strategy focusing purely on lowering HbA1c (by whatever means) might cause. I
And yet that is probably the one thing we know for sure about D: the percentage of time you spend high highly correlates with complications.
So did my son, and he ended up in DKA, about 12 hours from death. This is the danger with anecdotes: they are illustrative but not probative.
I have no problem with your opinions: God knows how much variability there is in D, and there is plenty of room for many opinions:-)
But I do not agree with presenting opinions as facts and validated theories, even if that is only to make a point! Many more people read these threads than participate in it, and it is important to be very clear on what we think, and what we know is true.
Specifically:
Focusing on reducing glycemic variability does not automatically increase hypos (although that may have happened to you)
A control strategy that leads to more hypos does not automatically lead to negative consequences. It could well be, on the whole, more positive to the patient: we can’t know without looking at the whole proposition.
You may be more comfortable spending 3 months at 300-700 than 3 days at 50, but many others might feel otherwise. I, for one, would pick the 3 days at 50 for my son.
One of the things we actually know for sure about diabetes is that time spent high statistically leads to complications.
At the same time, I can totally see that your PERSONAL opinion may be different, and I respect that opinion But, again, I think it is important to separate opinion from facts! Let’s be careful, when we are rhetorical, not to let our enthusiasm get away too far from data
How high is high? Is 90 high? Is 110 high? How long at 120 causes complications? How much extra risk for 3 hours at 130? Is it the same as 260 for 1 minute? The official recommendations for GCK MODY (avg 5.5-7.5% A1c) is no treatment necessary and no known extra mortality risk.
For me, there is quite a bit I think isn’t known about that subject. Not for certain.
I hope the dozen articles I have shared from my collection are sufficient facts to satisfy readers over our opinions
Since I already covered that the GV oxidative stress theory has been disproven (so far) for T1Ds and T2Ds, let’s do the last theory, glycation: High GV has no significant effect on T2Ds. For T1Ds, its effect on Glycated Hemoglobin (HbA1c) is extremely minor the lower your mean glucose is. If your average glucose is 140mg/dL, then the difference between low and high variability on A1c is only 0.27%.
And here is one showing hypoglycemia is a much greater concern than glucose variability:
There is one good study from this year that shows fasting GV predicts cardiovascular, mortality, and hypoglycemic risk.
They blame the cardiovascular & mortality risk on increased hypoglycemia, however. For those who think more hypos “do not automatically lead to negative consequences”, such a claim should be surprising.
There are some other minor studies, like one showing how children with variability in the normoglycemic range (but not persistent hyperglycemia) lose cognitive function. Because of more hypos? …
But I hope I have been thorough enough already with the major research
Happy Reading!
But again… we don’t know that. That is an assumption.
We can say we reasonably know that the percentage of time we spend with an elevated A1C highly correlates with complications… But the whole takeaway I see in this thread, all the studies cited in it, and everything I’ve read in general is that it hasn’t been well demonstrated by science that general glucose variability is highly correlated with complication risk…
That’s not to say that minimal GV isn’t a worthwhile goal. In general it is much easier to control diabetes reasonably well when and by striving for minimal variability, for me. That doesn’t necessarily mean that those who are not able to minimize their variability as well as I can are necessarily at a higher complication risk than I am, all other factors being equal…
It’s also distinctly possible that glucose variability is, in and of itself, inherently bad for you… but evidence to that effect is scanty imo… whereas it is abundant that elevated A1C is correlated with complications.
If we say 100 is normal and 200 is high, just for the sake of nice round numbers… is a a person better off spending 50% of their time at 100 and 50% at 200? Or better off spending all of their time at 150? Would they be better off spending 90% of their time at 100 and 10% of their time at 400? I don’t know.
But, again, I think it is important to separate opinion from facts! Let’s be careful, when we are rhetorical, not to let our enthusiasm get away too far from data 
…