Hypoglycemia, gluco-normals, aggressive targets


If you’ve spent any time interacting in this and other diabetes online communities, you’re well aware that hypoglycemia is the limiting factor to near-normal glucose control in insulin-treated diabetes. Anyone who has experienced severe hypoglycemia (defined as < 54 mg/dL or 3 mmol/L by the Hypoglycemia Study Group) knows the existential threat accompanied by sweatiness, confusion, and social embarrassment. It’s a situation, once encountered, we all try to avoid. We get why severe hypos are unhealthy and indeed dangerous.

On the other end of the glucose spectrum is extended time spent in high hyperglycemia (> 180 mg/dL or 10.0 mmol/L) is strongly associated with long-term increased risk of eye, nerve and kidney damage. So we live in the tension between those two extremes.

If we see doctors regularly to follow our diabetes, we likely hear warnings about avoiding hypoglycemia, even if that means we spend more time in hyperglycemia. In my 34 years living with diabetes, my clinicians have almost universally been hyper-phobic about hypoglycemia.

I fully understand we need to respect severe hypoglycemia, but most medical practitioners seem to over-react to “alert” levels of hypoglycemia. I place these levels at 60-70 mg/dL (3.3-3.9 mmol/L). This is a range that we do need to take note and action but don’t need to be reaching for the glucagon kit and calling 911.

I read a report several years ago about the Ambulatory Glucose Profile or AGP, a one page standardized report using continuous glucose monitor data. This report includes glucose text data as well as a good graphic that the clinician and patient can quickly and easily see where forward looking treatment gains may be made.

In reading through this 14-page report, I came across the AGP report for what the report authors considered as “representative of a normal reference population.” Here is the pertinent text-section of the report that caught my eyes.

Check out the “Glucose Ranges” box near the top center of the page, the one that gives details of the amount of time spent below 70 mg/dL or 3.9 mmol/L. This report shows that gluco-normals spend about 5.2% of their time below 70 (3.9). More simply stated, typical people without diabetes spend over 70 minutes on average each day in hypoglycemia. Not only that, non-Ds typically spend 30 minutes on average each day under 60 mg/dL or 3.3 mmol/L!

This surprised me. In my efforts to maintain my blood glucose as close to normal as is reasonably possible, it’s rational that in my control efforts, I am allowed to experience hypos similar to the non-D population and my clinician would not over-react to that circumstance.

Now I know that the risk and dangers of severe hypoglycemia are higher for people with insulin-treated diabetes. I respect that notion. But the game we all play driving a metabolic system in manual mode would be easier if our medicos would relax about our spending some time each day in the “alert” zone of hypoglycemia.


Good stuff, Terry.

Yeah. I know a local endo who is one of the best in the biz, and I don’t go to her because she has a fit regarding my 5% in the 60’s range. Heck, I target that as relatively normal. The 2 endos I go to (long story) simply look at the overall and say “Never seen a T1 with numbers like that!” I don’t care what the docs think. I know when I am ok and when I am not.


Thanks for this.

One thing that stuck out from this graphic which was odd were the granular numbers they had for less than 50, less than 60, less than 70, but then the really huge range of 70-180.

Kind of strange! Why not some granularity there? Like 120-140, 140-160, etc.

Also, I learned long time ago that non-diabetic marathoners are often less than 70 at later stages of races. I have always felt comfortable running at 65, as long as I am not dropping. I can go all day at a flat 65.

The big difference is the ability to maintain that 60ish number and not drop more. That’s the only advantage the non-D’s have.

Thanks for this post. I laugh at endos all the time when they criticize a sub 70 value. Nice to have some backup stats.


Thanks Terry. I’ve thought a lot about this lately since I switched from Dex G5 to Freestyle Libre and no longer have alarms. I think endos and docs are usually so rushed during an appointment that they revert to a one size fits all approach to low BG. That approach is based on people without a CGM, who can only randomly catch low BG’s on their meter.

Bringing the full range of CGM plots to the appointment (including the bad days) really helps a lot to get a dialogue going about this with my endo. I want that dialogue, I don’t want to sit there and listen to advice based on clinical studies. So instead of just handing over my meter I print off daily plots for normal days and outlier days and we discuss those.

The “no alarms” feature on the Libre has caused me to rethink my evening BG levels before heading off to sleep, with the goal of avoiding a severe low when sleeping. I have settled in on two BG strategies: one when I am awake and another for sleeping. The sleeping strategy is necessarily a lot more careful than the daytime strategy: I want to be very confident that I will avoid dropping below 70 when asleep because I want a safety margin just in case I would sleep through it and head dangerously low. So my sleeping BG goal is zero percent below 70.

On the other hand when I am awake I am OK with watching for trends on the Libre and adjusting accordingly if I am headed low. Sometimes I’ll catch it at 65, sometimes at 60 but as long as I reverse the trend and start heading up again all is well. The only time my endo really gave me a hard time was a low in the 40’s (caused by overbolussing for a rising BG trend) that I had a hard time eating my way back out of. I want that dialogue so I brought the whole play by play on the CGM plot and we discussed. She does not get worked up at all about short duration lows in the 60-70 range provided they are not too frequent.


Those numbers are a bit surprising to me as well!
I know my wife had an episode where she felt DRAINED. I tested her blood and it was at 65. Told her to eat some carbs and juice.

1 Like

In the past, my attitude for this doctor position was fairly intolerant. I feel it’s mostly driven by a lack of perspective about what we, as people with diabetes, deal with every hour, day, and year. But then I’ve come to realize that clinicians see the full spectrum of diabetes control in their patients. They’re probably sat in a front-row seat to some real hypo tragedies.

Having said that, I wish that these doctors could refine their perspective of the hypoglycemia range and patients who have reduced their BG variability to a sane level should be counseled with that in mind.

Reduced BG variability is the prerequisite to lower average BGs. Yeah, when you finally take ownership of your diabetes, you don’t need approval from the diabetes “experts.” In fact, I feel confident in the deep knowledge that 300,000 hours of living with diabetes has given me. Our “skin in the game” authority ranks above the broader more general knowledge of any clinician when it comes to day-to-day diabetes management.


Contained in the 14-page report is some discussion about the 70-180 mg/dL target range. These doctors wanted a range that was within reach of the average diabetic. At least they wanted to keep the patient’s interest long enough to get them to use the report on a regular basis.

These AGP reports, like the ones in Dexcom Clarity, Diasend, and Libre allow the user to customize these glycemic thresholds. I set mine to 65-140 mg/dL. I don’t start hypo symptoms until I reach 65 mg/dL, so I want that range on my side of the ledger. The 140 mg/dL goal is driven by my wish to minimize damage done by hyperglycemia. And it’s a goal that requires some effort and I can usually reach


The notes explaining Dexcom Clarity’s grading of hypo risk into “minimal, low, moderate and high” examines hypo frequency, depth, and duration of low glucose episodes. I feel that a computer program like Clarity does a better job of assigning hypo risk than most clinicians.


I wish that I could pull up my diagnosing physician’s research paper that he showed me regarding hypoglycemia and the range of “normal” blood sugar 37 years ago. The paper reported the range of fasting blood glucose levels of a group of normoglycemic individuals. A not insignificant number of subjects had fasting glucose levels in the 40s (and some lower, I believe!) I suspect that all persons with type 1 diabetes who try to maintain “normal” glycemic profiles have, on occasion, found themselves well below the official normal range, but still functioning as if “normal”. The trouble arises when we “abnormals” still have IOB and don’t respond quickly enough. I wish that our endocrinologists could gain an appreciation for the difficulty of maintaining near-normoglycemia as a type 1 diabetic attempting tight control and be a bit more supportive of our efforts!


Some people, including doctors, are more empathic in their personal style. Even the most empathic individuals will have a hard time knowing what it feels like to live with manually operating a glucose metabolism every moment of every day with never a day off. In addition to all the other demands of family, work, and friends, we have another full-time job making making dozens of metabolic decisions every day that non-Ds never need to give a second thought.


I have to say this report is the best report out there right now. My goals are set to this report now. And when my doctors and I look at it, of course the first number they all look at is the very low number ( we try for 0) and in target (70% or higher), and standard deviation (lower than 45). These 3 numbers will keep me safe and not bouncing all over the place. Again, everyone has different goals, rightly so!
But I think one of my biggest takeaways was the numbers of people who don’t have diabetes. People with a normal functioning pancreas can swing up and down out of range. So when those of us with a lazy, useless pancreas, swing up and down, it’s ok. We are like everyone else some of the time.
So thanks for pointing out that we also can be within normal ranges when our blood sugar falls to 50!

1 Like

For non-diabetics, who expect to shoot up at least some after a meal AND THEN THEIR PANCREAS TAKES CARE OF IT, I’m not sure it really makes sense to break it down that finely.

For you or me, where we need to know what kind of insulin dose is required to correct it, then I think that sort of fine distinction can be very important. (Although shooting based on an after-meal number before it settles down can be tough in itself!)

1 Like

I am just curious about it. I think it would be helpful to know what is normal. Gives us something to shoot for.

1 Like

The full report on the AGP contained one paragraph that I thought contained most of the value in the entire report and it concentrates on hypoglycemia risk. It focuses on using the AGP graphic to identify the actual time of day to take possible action going forward.

Smoothed curves representing the median (50th), 25th, and 75th (IQR) [inter-quartile range] and 10th and 90th frequency percentiles define the 24 h AGP. At a glance, one can observe the time(s) of day when glucose is most consistently low or high and when the most variability is occurring [the width of 25–75 (50% of reading) or 10–90 (80% of readings)] that needs to be addressed. This is an exercise clinicians can do together with patients in a matter of minutes. For instance, without dependence on numbers, formulas, or derived indices, clinicians and patients can quickly become skilled at identifying the risk of hypoglycemia. For example, if the 10th percentile curve crosses 70 mg/dl or lower, there is moderate risk of hypoglycemia at that time since consistently 10% of the values fall in this range. However, if the 25th percentile curve crosses into hypoglycemia, this implies a marked risk since more than 25% of the glucose values fall in the hypoglycemic range, and consequently, this should be addressed before additional therapy is instituted to treat accompanying hyperglycemia as is often seen with significant GV [glucose variability].

Here’s a sample graph taken from the cited report to use as a reference when reading the above quote.

I’ve read and reread the above quoted paragraph many times. It absolutely helped me to put hypoglycemia incidence into a rational space. I can’t ignore it but I needn’t over-react to it like many clinicians. I’ve been fortunate to often keep the 10th and 90th percentile lines contained within my upper and lower BG boundaries. When they do start to cross my boundaries, I take action by changing pump settings and prebolus times. I’m working to educate my doctors on this relatively simple system – it’s a work in progress!

1 Like

Normal is 80 :wink: I know that’s not very helpful of me. I’m T1D (no insulin), my wife has in the past been identified as “pre-diabetic”, her close family are mostly Type II now (we’re in our 50’s).

She tests herself periodically and normally comes out so close to 80 that it is 80; there is no variation, but once she has registered 70 and once 240. I don’t think “normal” helps, because we aren’t.

I’ve come to the same conclusions as the OP. It comes down to this; what doesn’t kill us makes us stronger and what always kills us is blood sugars over 250. It’s just not that quick, so the medical professionals concentrate on the quick deaths, the ones that happen before they, themselves, draw their pension.

I currently target 80 (I’m an Omnipod user and do negative and positive corrections). In the past I experienced severe overnight low blood sugar however when I started using the Omnipod I calibrated my basal (it’s now set to 12IU/day) and since then haven’t had any problems. In other words I can go without eating for any arbitrary period of time and not have either low or high blood sugar, but I do become very “brittle” (as it used to be called) on no food.

I still have serious problems with high blood sugar (high HbA1c), but the problem seems to be errors and inaccuracies in my carb intake calculations.

I think there are two separate parts to this:

  1. The basal. Despite what has been said on this thread I think there is a tendency for medical professionals to overdose the basal. Doing so creates a buffer which limits highs after eating at the cost of requiring eating on a regular basis.

  2. The bolus. This is, I believe, the major threat to our lives. It’s pretty much impossible to get insulin into our bodies fast enough to match the carbohydrate intake. Try going to a restaurant, waiting for the meal, then waiting 15 minutes after it is on your plate before you start eating :wink:

The excess basal helps with (2), but then if the meal is delayed you end up eating glucose tablets for dinner. Since I stopped using Lantus (basal disaster area) and calibrated my basal I’m left dealing with (2) but without the daily concern of hypoglycaemia.

John Bowler


Have you tried Afrezza @anon939029? Do you live in the U.S.? It kicks in really fast. I have to take it after I eat or I’ll drop low.



Alfrezza: no, I haven’t. I don’t use a CGM (utter failure of a DexCOM some years ago and they wouldn’t refund the money) and my problem seems to be that I am unaware of my BG heading up through 150mg/dl. I can normally tell somewhere past 200, though it’s not that simple; my body just pinged me but my BG measured at 171mg/dl. I always thought I was detecting rate of change, not absolute levels, but I’m starting to think that my skin BG lags my core BG in the same way that skin temperature doesn’t match core temperature.

Is Alfrezza substantially faster than Humalog? Humalog doesn’t start working any faster than regular (Humalin R) insulin, but it wears off a little faster; I’m considering going back to Humalin R because of the enormous cost difference in the US. It’s the pump that makes the difference, not the insulin.

Something that does cut in in 12 minutes might work well for me. I absorb glucose after, at most, 5 minutes but I avoid eating any dense carb; no sugars (unavoidable with restaurant food unfortunately) and none of the staples (bread, rice, potatoes etc.) I think it works but I don’t typically test immediately after a meal; another disadvantage of not using a CGM. Alfrezza might work well outside the home.

John Bowler


Yes, Afrezza begins working within 10 minutes for me. While that time frame is similar to Humalog, Afrezza’s peak action time is 35-45 minutes whereas Humalog’s is 1.5- 2 hours. Afrezza is out of my system in around 1.5-2 hours, but it depends on the dose (just as it does for Humalog). Here’s some info about it: https://www.afrezza.com/hcp/afrezza-action-profile/

The dosing is very different with Afrezza, and I do use it with a cgm. I love my Dexcom. There’s definitely a learning curve, so a cgm can be a huge asset. Afrezza is especially helpful with bringing down a stubborn high though. You get back in range incredibly quickly. However, for higher fat/protein meals, you may have to dose again 2 hours later because it lasts a shorter period of time. I tend to use Afrezza when I go out to eat, when I eat higher carb meals, and when I’m correcting for a high. I’ll use Humalog for high protein/fat meals.

Afrezza is still fairly new to the market, so not all insurance plans cover it yet. It’s covered under my plan as a nonformulary brand name drug.

I found an example of when I used Afrezza to correct a high. I rolled onto my sensor in the middle of the night and stopped getting readings. It didn’t start working again until I woke up at 7:20 and then loaded the last few hours (I was running high because of a migraine/stress). I dosed around 7:25, and I’m completely back in range (without a severe low) by 8:45. Sweet deal! Look at how beautiful that slope is.


1 Like

One way to deal with this is to prebolus for about half of what you estimate the carbs will be. That gives the insulin a head start, but there shouldn’t be enough to crash you. (If you’re worried about going low, the bread basket or a carb-containing beverage can help tide you over.) Once your food is in front of you, you can calculate the remaining dose to give. It’s not perfect, but I have fairly good success with this method. And if I over-estimate my bolus, well, that’s what dessert is for!


For me the timing of the pre-bolus depends on the restaurant and the type of food I am eating and what is happening with my BG.

Could be when going into the restaurant, when ordering, or when the food arrives.

When I am going to have a pitcher of margaritas with dinner, I pre-bolus when driving to the Mexican restaurant. Then I can drink a whole pitcher without spiking.

It just depends on the situation.