Six months down the line, you’d think that everything was settled and stable. At least, that’s what I’d expect. And then that not so shiny and new insulin chucks a few curveballs in there. Or maybe it’s life. Or maybe it’s just something else. Whichever of these it is, I’m rather glad there’s help from OpenAPS to smooth it out.
So what am I talking about? I started trying more aggressive Super Micro Bolusing (SMB), and at first this was working very nicely. Then over the period of testing, I noticed that the ability to restrict the glucose rises was getting worse, going from a rise of 3mmol/l to a rise of 6 or 7mmol/l for similar amounts of carbs.
And then the house of cards collapsed spectacularly. Using SMB normally, with a 75% bolus on eating and allowing SMB to catch the rest? Fail. Glucose rise of 4mmol/l. I’d not seen that previously with Novorapid nor Fiasp.
And then with full bolus prior to food, still a 4mmol/l rise with SMB trying to help.
The worst bit being that glucose levels are going up, and even with top up insulin, they are then staying up, in spite of additional SMB adding in further insulin (like a dual wave on a pump). Overall, I’ve needed to use nearly twice as much insulin to deal with food. And of course, then you start chasing highs, which can only result in unwanted lows.
This is reflected in my AGP plot, which shows notable overall increases in the periods when I eat, and then slower recovery overnight.
To add to the strangeness of the situation, when I look at the overnight periods, without food, my glucose levels have stayed reasonably flat, and the basal profile is not showing additional insulin outside my normal amounts. What appears to be happening is that when I eat, specifically in relation to food, my IC ratio has dropped from 1u:9g to 1u:5g – practically in half. And yet when I’m running basal alone, my ISF seems to remain the same.
The result is going through the process of adjusting bolus timing, bolus amounts, ratios, etc to try and get it back on track. But it’s not entirely obvious why it’s doing what it’s doing.
A further example is below. A prebolus of 2u, some 35 mins before eating that 14g of carbs should be enough to stop a rise, and given the absorption time that Fiasp is supposed to show, should have started to show a drop (indeed, if you look at the examples from fasted when I first started to use it, would almost certainly have shown a notable reduction starting by the time I ate). I’ve seen enough of a change to go back and have a look to see if I can determine whether I was mistaken early on.
When I look back though, I see that there were a number of occasions where I saw unexpected behaviour. Not loads, but enough. And they all look similar to what I’m seeing now. The main difference is that these are repeatedly happening and making it more difficult to use.
What’s also interesting is that over the past week or so I’ve seen a dramatic increase in “feet to the floor” syndrome and the “Coffee” effect.
You can see that from roughly 5am, there’s a fairly sharp rise, and similarly from about 7.15am. These coincide with getting up and then a cup of coffee. You can also see OpenAPS trying to control those rises. The thing that both of those incidents have is that they are both “glycogenolysis” events, where the liver produces glucose from glycogen stores. Looking back over the past few days, there’s a pattern of “feet to the floor” rises occurring that haven’t been as clear in previous reviews of my data.
My first inclination was that the insulin from the vial was bad. So I chose a different vial. That made no difference. Then I swapped to insulin I’d kept from a different batch at work. Still no difference. And finally back to a further, different vial. And still, no, it’s made no difference. So I need to look for a different solution.
Following a discussion with Scott Leibrand, I wonder if the issue is Glucagon/Glycogen related. I’d had no issues really, up until this point, but since this started, the trajectory of the rise after each meal, large or small has been almost the same. If it was solely down to the food, this shouldn’t be the case, as different glycaemic loads should demonstrate different slopes in terms of carb absorption. Instead, I wonder if what I’m seeing is akin to Bernstein’s “Salad” effect, where eating a whole head of lettuce results in the liver dropping a load of glucose into the blood.
That would certainly make sense if I am looking at a model where I was seeing if the SMB could handle larger carb loads, so that while there was little insulin on board, the liver dropped glucose into the body, but bolusing ahead of eating, which some of the more recent tests have done, should see the faster absorbing insulin at least partially hitting the liver (even if in small amounts) to stop that release taking place.
So what’s going on here? I have to admit I’m not really sure. But a theory is that the continuous subcutaneous infusion of B3 as an excipient of Fiasp is in some way making the liver resistant to insulin and therefore stopping it from listening to the tiny amount of exogenous insulin that should be telling it to stop producing glucose alongside food or in response to other hormones.
This translates to a need to bolus more insulin to allow the rest of the body to absorb both the carbs in food and the additional liver released glucose. It seems plausible. I don’t know if it’s correct. What it is is a conundrum that’s making life quite tricky.
I’m lucky. I’ve got an algorithm doing it’s best to stop me going high and low, but even that’s been challenged with what’s been going on.
So far, the best way to manage it seems to be to bolus with more insulin, effectively halving my IC ratio. It also seems to have slowed down its action time. So I’ve been considering what the options are to try and deal with this. They seem to be:
- Stop using Fiasp for a week and see what my body’s response is on restarting.
- Increase exercise intensity – while I’ve been walking way more than 10,000 steps each day, due to work commitments, I’ve not had the same level of resistance training intensity, and I know this reduces GLUT-4 concentration.
- Try and reset all my ratios, starting from scratch again.
While I should probably do the latter, I’m going to start with option 2, and see what happens. If there’s no difference, then it’s option 1, mainly as an interesting experiment.
What’s certain is that this is proving a tough nut to crack and is showing that Fiasp is perhaps not the panacea I’d hoped. It’s one of those times where we just have to keep on keeping on, and hope that we come to a conclusion. And if anyone has any ideas, please feel free to add them in the comments.
Footnote: If you’re a Fiasp user, and especially if you’ve seen similar, I’m trying to collect together user experiences so that they can be shared with those who are interested. You can find the submission form here. The responses so far can be found by following this link.