Hypoglycemia: When Is It Problematic? (How Low Is Too Low?)

I disagree if the phrase “too many moderate lows” is only based on the number.

Instead of a number, how about:

  • How long were they?
  • Did you need help to recover? Or could you manage it yourself?
  • Was it a quick drop below and then back up?
  • Did you see it coming, or was it a total surprise?
  • Was there a bad rebound?

The problem with using a number instead of those types of questions - is 56 okay but 55 is not?

I’d like to hear if you have some of these questions that could be used for lows. What am I missing?

You don’t need to duck. These are good points for discussion.

Well, i don’t have an opinion on whether @Irish is having too many lows or now, but I would mention that if you’re looking at studies evaluating the risk of hypoglycemia, many use 55 on finger stick as the threshold. Do we know that 56 is “fine” while 55 is dangerous? No, not *really. And I suspect it’s not much different.

But on the other hand, it’s really the only low threshold that’s been evaluated in any thorough way. From what I’ve read, hypoglycemia in the <55 range doesn’t seem to have a huge effect on cognitive function in adults long-term (aside from hypo unawareness). However, there are some (vague, circumstantial) studies that somewhat equivocally link cardiac autonomic neuropathy with hypoglycemia. None of these studies are very large or very definitive, and the data goes back and forth. Some studies suggest the causal arrow points the other way (in other words, a predisposition to glucose variability and hypoglycemia may be actually tied to some other factor that also causes the neuropathy). Regardless, I would argue that it’s not silly to use an admittedly arbitrary threshold like 55, which has actually been studied in lots of research, as a relevant metric for glucose control. It’s not perfect for one data point, but once you’re talking hundreds or thousands of these points, frequent lows <55 each week will suggest a pattern that you can then match up against all these studies. How you interpret the studies of course, is a matter of opinion.


So are all 55’s the same? If I drop below for a few minutes, is that considered the same as being below for an hour? I think there is a lot missing with just a number like that.

@Eric – there no doubt is a lot of information missing in this data and briefly touching down to 55 is most likely different from sitting there for an hour – just like A1C is just one number that is missing a lot of context. When I am treating lows I certainly operate as if a one-reading 55 is different from 10 readings below 55.

But unfortunately, just like A1C, if you’re a clinician (or even if you’re just a person with diabetes trying to manage it using evidence in the clinical research as opposed to hunches or penumbras or emanations from the research), you have to go on (and make recommendations) on what we have. And all we really have is data vaguely correlating finger-stick BGs less than 55 with some potentially negative outcomes. We don’t really have data that maps out low frequency versus time spent low because unfortunately most studies to date haven’t analyzed CGM data. In the next few years I suspect more studies will rely on that CGM data. I personally am avidly looking forward to the insights it will reveal.

1 Like

Eric, I will pretend that you don’t know the answer to your question. Of course when making a definition for a study you have to pick a number, so they went with 55 in many of them. Does that mean that 55 is low, well for purposes of the many of the studies yes. These studies have shown that there is an effect of having lows. Does that mean that 55 for one second equals 55 for 10 minutes, the short answer is we don’t know.

I would hope that in the next few years the auto-basal trending pumps will be good enough to keep everyone between 100-130 for most of the day. [fingers crossed]

I am in the same boat as Britt, if it was me, I would strive to raise my average numbers a bit to decrease the number of lows per week, if that could be accomplished without too many highs.


My sister, who is not a diabetic, was training for a half-marathon and had a sub 40 test after a training run. Of course her body corrected it. But the idea that non-diabetics never drop below 80 is totally wrong. Non-diabetic marathoners are often in their 60’s when running.

Not all 50’s are the same. It’s just not. They use that number because of a lack of better measurements.

There are 50’s where I just drink some Prime and keep running. And there are 50’s where I can’t get off the couch. It is interesting to me that lows when running don’t feel as bad as lows when I am at rest or sleeping or working. It is totally opposite of what you might expect. The big problem with these studies is that they treat all these lows as the same. It is my contention that they are not even close!

So here is where I think you are a little misguided.

You say Not all 50’s are the same, and of course I have observed that as well. But, just because you don’t feel bad doesn’t mean damage isn’t being done. A young man who I am acquainted with lives in the 250’s - 300’s, he feels perfectly fine there, and in fact feels terrible when his blood sugar is normalized. Does that mean he isn’t doing damage to his body? Of course not, we all know the A1c over 11 and the sustained 300’s are going to damage his body.

I think we all need to strive to live in the “normal” range as much as possible and try to prevent excursions either way. To say lows are better than highs would be a mistake in my mind. Sure, lows are probably better than highs, but the data isn’t there to make a definitive statement.

1 Like

I know it is different with youngsters. Absolutely. I agree with that. But since the OP was Irish, I was aiming my comments at adults.

The link I posted above from a study led by Joslin and posted in the New England Journal of Medicine does definitively say that for adults low is better.

Lows can kill you quickly and right now. You can die in your sleep or have a car crash. But highs will kill you later, and more slowly.

Look at people who are in their 80’s and been diabetics their whole life. How did they do it? By focusing on avoiding highs more than lows.

I am not advocating people have lows. But if you want to run your BG in the normal range, lows are an unfortunate consequence.

From the link:
Further study of hypoglycemia and cognitive function over time in people younger than 13 is needed.

Yes, but the average A1c in the treatment group was 7.1, so while what you say is true, there is evidence that intensive treatment is better than conventional treatment that resulted in an average A1c of 9.0, I think the evidence that maintaining an A1c as low as Irish (4.7) is not there. If she could maintain an A1c of 5.5 and reduce her lows by half, that would be a reasonable goal in my opinion.

Additionally, the intensive treatment group had double the rate of serious hypoglycemic events defined as coma or seizure as the conventional group. Now, I would never advocate for a treatment plan that results in an A1c of 9.0, I believe that caution should be paid to ensure you give yourself a fighting chance to feel the low, so you don’t have the serious events resulting in coma or seizure.


All these lows might well be different, but there is really no experimental evidence to suggest that’s the case as yet (anecdata is intriguing but needs to be backed up).

And while, yes, most non-diabetics do spend a fair amount of time below 80, they spend relatively little below 60 from the little they have been studied. Beyond that, we don’t know if people who are nondiabetic but frequently experience BGs less than 60 or 55 have elevated risk of adverse outcomes, because we haven’t studied it. It may well be that frequent moderate hypoglycemia is bad in everyone, but those non-D people get folded into the “healthy” general population in risk studies and no one ever realizes that they had greater incidence of CV or death because of an undiagnosed problem with hypoglycemia. All we know is that nondiabetics are unlikely to die or go into a coma from hypoglycemia, not that it has no long-term effects on them. So in my opinion, we don’t know enough about normal BGs to say definitively that lows below 55 are perfectly safe if you don’t “feel low.”

Finally, I agree the evidence for long-term negative cognitive outcomes is absent. I’d argue that’s only one of the potential side effects of repeat hypoglycemia. Others, as I’ve mentioned, could be autonomic neuropathy, while a few studies suggest that hypoglycemia in healthy individuals can fuel a rise in pro-inflammatory, pro-thrombotic chemicals – potentially meaning a higher propensity for stroke or heart attack:

A study showing that hypoglyemia in nondiabetic patients was associated with a higher risk of cardiac ischemia:

ADA/JDRF Type 1 Diabetes Sourcebook, Excerpt #13: Setting Treatment Targets, Part 2 of 2

Blood glucose goals in pregnancy are undoubtedly the most stringent. @Eric, please note that there is no target <60.

Sure in a perfect world you can do that. But in a perfect world, none of us would need to be here with the disease. You can’t just pick your A1C number. You target best BG you can, and then see what A1C number turns up.

From the link you posted:

Our study found no evidence of substantial long-term declines in cognitive function in a large group of patients with type 1 diabetes who were carefully followed for an average of 18 years, despite relatively high rates of recurrent severe hypoglycemia.

These results do not mean that severe hypoglycemia is entirely benign. It is well established that an extended episode of profound hypoglycemia, such as one with a blood glucose level below 1.0 mmol per liter (18 mg per deciliter), can induce massive cerebral energy failure with a corresponding development of neuronal necrosis.26 Less severe episodes of severe hypoglycemia (e.g., those with a blood glucose level in the range of 2.8 to 3.6 mmol per liter [50 to 65 mg per deciliter]) are also known to disrupt brain activity transiently, can lead to short-term cognitive impairment, and can increase the risk of motor vehicle accidents. Nevertheless, with the exception of several small, cross-sectional studies, most researchers either have failed to find effects or have found only relatively weak effects of recurrent hypoglycemia on brain structure and function in children and adults with diabetes.

Of course @britt_j , nobody targets 60. Please tell me you don’t think am suggesting that. I would never target 60 for myself, certainly not anyone else either.

All I am saying is that sometimes <60 happens when you are trying to stay within range. And for adults, it is not a worrisome thing as long as you correct it and don’t crash.

Of course, we’re diabetic, highs and lows happen!

The idea is to minimize them!

@Irish asked

What would you say to a patient who achieved an A1 C of 4.7 looking like this?

Please re-read my replies. I simply suggested that she focus on variability and easing up just a bit to avoid the frequent moderate lows.