That is awesome that you have such good control! I so wish all our lows were above 60. We work very hard towards that but not always successful
I just had this very discussion at the clinic: the nurse actually told me that she didn’t like it when K skips a meal (which he occasionally does) because his liver will deliver more sugar, which makes his BG less predictable – she totally agrees with you!
It appears you have dawn phenomenon! My son is starting to show signs of it too. Although he has something else too: feet-on-the-ground insulin, i.e. the moment he wakes up his BG goes up 20-30.
Isn’t it strange how endos have a hard time understanding so many things we see every day?
If you think about how long CGM’s have been readily available, it isn’t that long. Also looking at the daily CGM information is certainly not in every office. i.e. they seem to look at the accumulated CGM data and look for patterns, but our physician certainly isn’t looking at the daily tracing and trying to divine something. It will be even longer until the CGM data on highly compliant patients is looked at enough to help physicians. Someone will eventually write a nice paper on using the trends to tightly manage diabetes, but that isn’t out there now.
We see a bit of a dawn phenomenon too: To account for it with openAPS we reduce his ISF an hour before he typically wakes up (at 5am) and then raise basal about an hour before he eats (around 6am). It doesn’t totally curb the rise but it’s the best tradeoff between causing lows before he eats and having him start breakfast at 190 that we can find. Usually he’s around 120 or 130 when he sits down to eat. For us that’s fine because it allows a little cushion on mornings when he drops a lot before he starts rising.
If it’s DP then it’s a very aggressiv one - it starts at 6am and goes until 4 pm - if I don’t spend the day continually injecting humalog it will go to 270 by 4pm without eating any food - I typically have to inject about 20 units of humalog during that period to keep it below 130 then at 4 it stops and my nightly dose of tresiba keeps it flat until 6am easily - extremely flat - the cgm doesn’t even show a blip unless I eat some carbs - if I time the humalog in the pm incorrectly when the liver calms down at 4pm I used to go into hypo at 5-6 all the time - without the cgm it was almost impossible to figure out what was happening
To go back to an earlier post I made on using trulicity it’s why I’m using it to see if it helps my aggressive liver from 6am to 4pm - as I said earlier it is helping by halving my humalog needs during that period coupled with eliminating all carbs as well - but the liver is still very t City and unpredictable
I wonder if you might consider a shorter long-acting insulin like Levemir as an adjunct? For many users, it only lasts 1-12 hours. @Eric might be able to provide some good info, since he uses it regularly just for that purpose (shorter long-acting insulin).
Agree with Chris also - my endo refuses to look at daily cgm records when I bring these up to discuss issues - only the average graphs - he says daily can’t be relied on as it’s a one off occurrence - not if it happens in multiple days !!!
Yes Eric suggested levemir earlier - I’ve rung the doctor and asked him - we are discussing it when I visit him next end of this month - he said ‘you’ve been on those diabetic chat rooms again haven’t you’ !!
I think that’s one of the trickiest things for analysis. I definitely see patterns but maybe they only occur 60 percent of the time. If you look at the 90-day average, our CGM trace looks pretty flat, although the top 10% looks higher at certain times (like breakfast) – how on Earth do you extract valuable information from it?
I agree with @TiaG, this is a tool that I can’t extract a lot of info from. Early on I was really excited about it, but I hardly ever look at it these days.
Besides, if you have a problem, you’d have to not correct for it for a lot of days before it starts showing up there
The duration of Levemir depends on how much you are taking. But it’s not difficult to figure out. The Link below give some details. Just make sure you convert your weight to kg for the math.
Plank et al24 also found a dose-dependent duration of action for insulin detemir in type 1 diabetes from 5.2 hours at the lowest dose of 0.1 U/kg to 23.2 hours at the highest dose of 1.6 U/kg and a flat and protracted pharmacodynamic profile (Figure 2). Limitations of this study include it being a first-dose study and not a steady-state study.
For the curve above: GIRmax = max Glucose Infusion Rate (needed to neutralize insulin) i.e. how long it takes for it to get to maximum activity.
AUCGir/100 = Area Under Curve GIR, i.e. the total glucose infused to neutralize insulin action, i.e. the accumulated activity for the insulin in question.
I don’t get anything useful from a 90 day view. I do get good information from the 14-day. My son and I review this every week to look for places to optimize. We basically look at the top of the highs and the bottom of the lows to try to see if we can get away with an increased basal. Something like the picture below, although the decisions I highlighted are a little simplified.
Yeah, there is a lot of stuff in it, but here is a key part for the duration:
For doses in the interval of 0.2 to 0.4 U/kg, LEVEMIR exerts more than 50% of its maximum effect from 3 to 4 hours up to approximately 14 hours after dose administration.
As an example, if you weigh 170 pounds and take 23 units a day (basal):
Convert pounds to kg:
170 pounds = 77.1 kg
Then take your dose and divide:
23 units / 77.1 kg = 0.298 units / kg
So you would be within range of being under 14 hours of duration.
I’m guessing the most useful timescale has to do with who the person is and what physiological factors likely dominate any variation in their BGs.
Not to nerd out too much but I wonder if something like Wavelet analysis may help people figure out which views are most useful. Based on a 20-year-old class I vaguely remember those transforms are useful for pulling out periodic information from a repeating signal with a lot of noise.
