Varying glucose sensitivity: why?

As you know, my 12-year-old T1D son recently (about 3 months ago) started puberty, and we have had a bit of a hard time adjusting our control practices.

Right now, based on his TDD, he is using about 0.35 insulin units/lb.day, so this indicates that, per this summary for instance, he probably still has some beta cells working (i.e. still in honeymoon but on the cusp) – since the expectation is about 0.5 unit/lb.day, going up to 0.7 while in puberty. I understand this varies a lot, which is why I am adding probably.

Since he has hit puberty, his glucose sensitivity has varied widely, which makes it challenging to deal with glucose corrections. It is more or less in one of two positions: either it is low at roughly 4-5 mg/dl.g (i.e. BG goes up by 4-5 per gram of carb), or it is high at roughly 12-15 mg/dl.g (i.e. BG goes up by 12-15 per gram of carb). We never quite now which it will be. Of course, we calculate this ratio only when he has no IOB outside of basal (his DIA is 5-5.5 hours).

My assumption is that his beta cells are sputtering widely: when they are on, his sensitivity is low; when they are off, his sensitivity is high. But, if that was the case, I would expect his TDD for the day to go widely up when his sensitivity goes up (since he stays in reasonably good control), which is not the case. So I am scratching my head.

What is your experience? If you are out of honeymoon, is your glucose sensitivity roughly stable? If you are not, do you see the same phenomenon? Is puberty a possible cause, and how?

And, most importantly, is there a way to predict better instant glucose sensitivity, since the day’s TDD trend does not seem to allow us to do so?

Also - what is your glucose sensitivity range? I would love to know what others see, and in what circumstances.

We suspect our son is out of honeymoon because he gets high ketones if he’s off his pump for any length of time. Our glucose sensitivity is roughly the same but we have random days when our son is wildly insensitive; for instance a few weeks ago he randomly jumped from a carbF of 18 to 11, then back down. Day to day, ISF varies by time of day and also by what he’s eaten. On a given day, he usually has an ISF of around 250 during the day and 315 at night. He can go down to about 230 on some days and up to 265 others; during the night it may vary between 300 and 335. This is based on his openAPS algorithm, which is automatically calculating ISF over time based on how much insulin it needs to deliver to keep him in range.

My understanding is that those wild fluctuations in sensitivity are largely to do with growth hormones/other hormones. Growth hormone and cortisol both make kids more insulin insensitive. Also I suspect that what he’s eaten also affects his sensitivity (studies in T2s show that fat inhibits insulin sensitivity, though in T1s it may be a different story.)

However, I would say that his carbF has more variability for small amounts of fast-acting carbs given with no IOB – i.e. let’s say he’s trending low and we give him a 4 gram glucose tab. That may raise his blood sugar by either 20 mg/dL or 80 mg/dL, depending on the time of day and other mystery factors. This level of variability seems to operate mostly when he has no insulin on board and is eating relatively few carbs. Once the insulin is flowing in his system, the amount of insulin he needs per meal varies much less – up to 25% from day-to-day during usual circumstances, and by up to 100% if he’s sick or some other unusual occurrence is going on. My guess is that you need some insulin to suppress the liver dumping of glycogen, but that once you’ve hit that bar the body’s level of sensitivity to insulin is a little more constant. But that’s just my theory.


You also seem like you have some extra confounders we mostly don’t have. You mentioned your son is in different athletics. From what I understand, aerobic exercise in the short-term can raise insulin sensitivity because, during that exercise, the muscles can ferry glucose into the cell without the use of insulin, and that ability increases. But exercise can also release cortisol which increases insulin resistance. And about 12 hours later the muscles will often try to replace depleted glycogen stores, leading to an increase in insulin sensitivity often at night. When you throw in the puberty hormones you’ve got quite a huge ball of string to untangle.
My hunch is that it won’t be so helpful for you to figure out whether these are honeymoon-related swings or diabetes-is-nuts-related swings. Either way you’ll have to rely on improvisational skills, short-term pattern matching, crisis prevention, and predictive, preemptive action.


@TiaG, I am so glad to be able to talk to other people who think of the same things I do. I cannot have this conversation with our endo, for instance.


The pattern recognition is in fact what we are doing. The hormones make him incredibly insulin resistant, we often have to treat 2-3x the amount of insulin to bring those down. For awhile they had the nice manners to only occur in the middle of the night, but now we are getting other hormone swings in the day, and for awhile we only had to deal with highs, and now we are dealing with a period of hormonal highs, followed by hard to cure lows. My son has had 10 double or triple treatment lows this week. Lowering basal and adding glucose has less effect than normal on these.

1 Like

We have had the same syndrome: at first the only occurred at night, but now they have spread throughout the whole day. [quote=“Chris, post:5, topic:482”]
for awhile we only had to deal with highs, and now we are dealing with a period of hormonal highs, followed by hard to cure lows. My son has had 10 double or triple treatment lows this week.
Same here as well. Our time in hypoglycemia has doubled.

This is interesting. We have some peaks that take a huge amount of insulin. But, for most hormone peaks, we find they take about 1/2 the amount of insulin as regular glucose peaks (I mean poorly compensated food peaks).

@Chris, how is glucose sensitivity outside of periods of IOB?

This is one wild ride…wish it would slow down just a bit…

1 Like