That is often referred to as basal bolus ratio. There have been several discussions on this if you search this site.

Often 50-50 is starting point for new diagnosis. Some doctors think it should be followed more strictly than others. With today’s pumps and CGMS providing more data to base insulin dosing, I think that ratio is not used as often.


I agree with other commenters. What you are describing is the split between basal and bolus insulin (and I agree, on a low-carb diet it’s virtually never 50/50).

Although you can calculate standard deviation manually (and you could do it for blood glucose values or units of insulin taken or number of tests per day or anything with a data set), it’s freakin’ tedious and not something I’d ever do outside of an exercise in a statistics course (which is the only place I’ve done it manually).

I personally have never really had an endocrinologist discuss standard deviation at all. But it is a useful statistic, and something that’s mentioned in many forums, blogs, and books about diabetes management.


I think if you upload the Libre reader results to Diasend instead of the Libre software you can see SD.


Yes, Diasend definitely shows standard deviation (that’s where I puload my CGM, pump, and meter).


SD can be applied to any set of numbers.

For example, what if 2 students took 10 tests, each with an average grade of 90 (out of 100).

Student A got 5 100s and 5 80s
Student B got 5 91s and 5 89s.

Student A has a higher standard deviation. (More variability).

Generally, a brittle diabetic would have high BG variability even when food, insulin and activity were kept constant.


I was told on another message board that if one uploads one’s reader to the cloud-based LibreView, that this includes standard deviation calculations. I haven’t tried it yet, so can’t verify.


You’re right. I used to upload my Dexcom/Animas/meter/fitbit to Diasend. I’m a fan of the AGP report and Diasend has an attractive graph.


Diasend doesn’t work with the Libre for US customers yet :roll_eyes:. I tried to use it last night and when it didn’t work I contacted Diasend and that’s what they said.


Some doctors are just too lazy to help the patient so they label them with this label.


I changed my post it didn’t make sense.


Instead of “brittle,” we could say, “data complexity.” As in, “My numbers are behaving in a complex way.” Meaning, despite my best efforts to interpret the numbers, the patterns are a bit nonsensical, or are behaving in a complex way that is unaccounted for.


I am amazed to learn that brittle is defined as being difficult to control but also includes frequent hospitalizations… I wouldn’t describe myself as phobic, but I don’t go to the hospital over blood sugars. I’ve been very low many times but I know what to do for that without hospital visits. In 37 years of uncontrolled diabetes, the only times I’ve been hospitalized were becaus I got sick and went into DKA. I’ve worn an insulin pump for over 20 years, all three generations of CGM, but even with doctors monitoring insulin pump settings, my diabetes has never been considered controlled. Hospitalization, or not, should not define brittle…


I had never heard of worrying about the BBR until about 8 years ago and I’ve had diabetes for 27 years. I had a CDE trying to get me to 50:50 while pregnant and she kept getting irritated that I kept changing my rates around because my blood sugar readings were all over the place with the BBR at 50:50. That was interesting.


It’s always puzzled me that any clinician would latch on to some people with diabetes coincident ratio (basal to bolus) weighing in at 50:50 as having any special significance. And then turn it into a therapy goal! I suspect someone must have published a paper about this and some practitioners imbued it with more meaning than it has, mere coincidence.