Three months pumping and really frustrated



I hear what you are saying, and I am ready to believe. But the risk if I am wrong is high. I think it is quite possible that K may react like @cardamom, and require more adjustments that will be a pain in the neck: right now, with our management difficulties, we can’t really deal with more nights up because of 24-hour-long sustained lows, if that ever were to happen. Again, 6 weeks ago his basal was stable at night at 0.55/hr, and last night it was nice and stable also but at 1.1U/hr, exactly double, yet he is not sick (he can do >1.5U/hr when sick). I can’t help doubting that Tresiba can deal with such extremes! I have no idea why his basal is running so high right now, I just know it is because anything lower sends him into stacked injection after injection. Yet he could very well be back around 0.75U/hr tomorrow or the day after.

It is possible that my son has unusually variable basal. Overall, I think it is a more conservative play for us to try Afrezza first because the downside of Afrezza not working is pretty much naught :slight_smile: We have enough trouble as it is that adding another risk factor is too much for us right now.


I could provide a good reason for his basal to be running high for “this” week as compared to the last “three” weeks.

If he was a teenage girl.



This is the classic pumpers disconnect I’ve mentioned several times. There is absolutely no risk whatsoever. If it’s not a godsend you can go back on the pump immediately as soon as your heart desires. You don’t have to fight for pump approval again, you don’t have to get your doc on board again, you don’t have to cough up another copay, nothing. You’d just start doing again exactly what you’re doing right now if you don’t see improvement.

It may not work phenomenally well for everyone, then again, obviously, neither does pumping. There is no potential downside to trying it, only potential gain.




First, congratulations for the outstanding results! You’d probably need to look far beyond your clinic to find another teenager who is doing as well as Kaelen.

I have to say that I am not familiar with how exactly 670g works and what data it produces, but I think it is entirely possible that they are simply misinterpreting 670g-generated data. If they looked at my (Loop) graphs they could conclude that my basal needs vary by as much as 2:1 from day to day. That’s absolutely not the case. There are days when I do not bolus at all and just let Loop work through meals and other corrections, which it does in the form of moving basals up or down. That does not mean at all that my actual basal needs suddenly increased by a factor of 2.

That might be the case for someone who has some residual endogenous bg regulation capability. I would definitely (and fairly easily) notice a difference of 0.1U/h.

I agree with this.

I still think that attempting to manually adjust basal rates from day to day or even within a day is a mistake. If the basal needs were so variable (which I doubt is the case), adjusting basal based on what happened yesterday or a few hours ago is already too late and may be counter-productive looking ahead. It is more likely that bg variability is due to shorter-term effects, including unaccounted-for meals (e.g. protein and fat) , other hormones, exercise, etc. As many have already commented, those should be dealt with using corrections, not by adjusting basals. If I were you, I’d consider a few options:

  1. Set basal to a reasonable fixed rate, and let it run for say a week, correct as needed (but resist the urge to mess with basal rates), take notes of meals, exercise, and suspected hormonal events, then look at the data and make (small!) adjustments based on week long data. Then let those adjusted rates run for a week etc. In a few weeks you should come to a steady-state basal. Well, not really steady-state because the needs will definitely drift over time, but not day to day. Or,

  2. Try Tresiba, which should be basically the same as setting the rate to a fixed value as suggested above in option (1), except that it would not allow you to overreact, which I think might be good in your case :wink:. Or,

  3. Setup Nightscout and repeat option (1), but enter all data through the NS portal. Then run DIY auto-tuning algorithm and consider algorithm-generated basal rate suggestions. A bit more work to setup, but machines tend to be better at crunching data than humans.


Last week, Wed thru Sat, my son was using 0.75U/hr of basal at night and his nights were reasonably stable. Between Sunday and yesterday, my son had been using about 1.1U/hr of basal, with his nights being reasonably stable.

The difference between the two is an increase of 45%. The stable nights with no IOB prove that he is getting a reasonable approximation of what he needs. You were writing earlier that your body can feel a difference of 0.1U/hr. There is no way that my son’s body can deal with a difference of 0.35U/hr without running constantly low or constantly high: it would be obvious to us, and to anyone, if these basal levels were way off.

In fact, the reason we changed them in the first place was because we got into a phase where my son needed a bolus every 90 minutes for 7-8 hours on end to remain stable. Conversely, he has been running sustained lows late last night and this morning (with no IOB), leading us to start adjusting back, first with a temp basal, and if it sticks, with a more permanent basal change. I don’t think anyone would expect a CWD to take sugar every 45 minutes for 48 hours waiting out to see if a basal trend may invert.

We treat peaks with corrections, and lows with sugar. When we detect a pattern that is sustained, we set a temp basal. When the pattern is confirmed after a day or two, we switch the basal baseline.

This does not even take into account the constant change in basal needs induced by the highly varying sports schedules and practices, btw.


Has Kaelan read this thread? I’m curious to know his thoughts?


He is tired and short-tempered, as are we all.


We have been logging all this data in MyNetDiary since a few weeks after he was diagnosed. What we have not done is write down the BG issues in a qualitative manner (vs the raw BG track), which would allow us to review things much more quickly. So I will take that on and see if we can exploit this info better.

Given the quick basal changes required by sports practices, I think Tresiba carries too high a risk for us. We are pushed hard enough right now that we can’t afford this additional risk. But we are seriously looking at going back to Basaglar for a few weeks.

Let’s not forget that we have not had any trouble with basal tuning in the past. We are not total morons in matters of BG management. It is easy, from the outside, to assume that we are too quick on the trigger, but I can assure you that we bend over backwards to make sure we don’t. We pay the price with too many highs and too many lows when we don’t change fast enough because we want to make sure the trend is real.

It is definitely something we want to do: it has been on our todo list for many months now. Life has a way to make things difficult though. Right now we are bandwidth limited. We will try to get there as soon as we can. I would like to use Nightscout as the general data management and storage system piece in the future.


Please consider giving tresiba a trial. I would argue all day that it’s the safer option. I’ve exercised with Lantus plenty and it always caused exercise induced basal lows for me. Tresiba does not. There’s no additional risk. If you had good results with Lantus I expect it will be even better. It is approved for children younger than any other basal. Children are active. Your sons condition and situation are not highly unique. If you’re not willing to do that, go back to basaglar to figure out what his 1 or 2x daily basal dose should be then give it a try. It’s just a trial not a commitment. I wouldn’t be suggesting it if I didn’t truly believe it would make your life easier.

What’s the worst that could happen? A couple sleepless nights (you’re already doing that) then you give up on it… I don’t think that’s very likely.

What’s the best that can happen? Stable blood sugars outside of meals and throughout the night. 8 hours of sleep every night, not feeling like crap (both you and your son), better control, less hormone peaks, less basal lows.

Or maybe it’ll be somewhere in between…

Basaglar is a knockoff of 20 year old technology. Is that’ newer or older than our cars that we’ve joked about being old beaters? Tresiba is the cutting edge best that modern science has to offer. What’s going on in your sons body has a lot to do with how his body is responding to insulin and not as much as you think to do with other factors imo. granted being an athletic teenager is certainly a factor also, but it’s far from an unmanageable one.

Nobody thinks you don’t know what you’re doing, we understand that you do… but there’s a lot of collective wisdom here that you can tap into and use it to add to your arsenal.


@Michel, if you are going back to MDI, I would suggest trying Tresiba or Levemir vs Baslagar/Lantus—even with the issues re: Tresiba that I mentioned above, Lantus was a %#^show in comparison. I did not realize how inconsistent the absorption was until I switched, and that was with 2x a day dosing (which was better than 1x a day, but still not great). I think that Levemir is probably the better choice if you want a long acting with shorter duration and easier adjustability—my endo wanted to switch me to that prior to Tresiba coming out, but I am a super weirdo who gets big red hive-like things at injection sites for some insulins, including Levemir and old-school Lente, so that was never something I could stick with.

@Eric or someone else can probably speak to the Levemir part better than I can.


Different basals are slowed down in different ways. Tresiba forms soluble multi-hexamers to slow it down (using hexadecanedioic acid). Lantus uses higher-order aggregates of insulin hexamers. Levemir is bound to albumin in the blood which buffers its absorption. NPH uses a protamine and zinc to slow it down.

If you want, you can argue that Tresiba’s absorption is not as affected by exercise as the other basals. But it is completely wrong to say that exercise does not affect a body’s use of insulin.

We know that insulin increases the GLUT4 translocation to the cell membrane and increases glucose uptake. But muscle contraction and exercise also increase glucose uptake into the skeletal muscle.

The mechanisms that regulate glucose uptake during exercise have redundant pathways, such as calcium ions, AMP-activated protein kinase (AMPK) (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2779044), reactive oxygen species (ROS) (https://www.ncbi.nlm.nih.gov/pubmed/26791627), and nitric oxide signaling.

Exercise greatly increases glucose uptake. Making a blanket statement that people will not need to adjust basal during exercise because Tresiba does not require it is wrong.

It is not Tresiba that is the issue, it is the exercise. I cut basal for exercise. There is no magic to it, that’s just how the body works.


Totally! Reducing my basal an hour before yoga class means I no longer go low in class! :+1:t3:


I asked for levemir because I needed adjustability and my doc gave me tresiba instead. I was mad as hell. I was certain it was “high risk” because of my fluctuating basal needs. Boy was I wrong…

The point is that he should be trying all of these options to see what best works instead of continuing what he’s realized isn’t working and isn’t allowing either one of them to be happy or healthy or balanced. The “pumpers dilemma” (I made up that term) is a very real and potentially harmful psychological phenomenon when they convince themselves contrary to all reason that that can’t try anything but the pump


I suspect @Michel, that you are at the point where you need some encouragement, some sympathy and some sleep and quite possibly fewer suggestions from us peanut gallery members? We can all try to troubleshoot but you guys are very sharp, you’re an engineer (I think?) and have a handle on all the additional tools you could employ at this point. I suspect that all these suggestions will percolate and at some point you’ll try them when you feel ready.

One thing I’m realizing with insulin is that there are multiple strategies but that ultimately, what you want is the right amount of insulin in the bloodstream, binding to cell receptors, for the 4 minutes that insulin is actually active. You can do that with temp basals, with injections, with boluses, with exercise, with food – but ultimately what you care about is that insulin activity for very specific, short, windows of time. And there are lots of ways to skin that cat and get the approximate right amount of insulin at any given time. You guys have to pick the way that you feel most comfortable with, which is not to say that all these suggestions might not work just as well either.

My only suggestion is sleep, and I have no idea how to make that happen for you as I’m not getting any either.


If they looked at my (Loop) graphs they could conclude that my basal needs vary by as much as 2:1 from day to day. That’s absolutely not the case.

@dm61, have you looked at some of the posts by another person who is Looping with an Animas ping? I believe he’s named Marius? He has created his own looping algorithm in which the only variable the loop changes is ISF, and that goes up and down based on past data every few minutes. From what I understand he’s gotten very good results with his toddler – likely even better than what most Loopers are accomplishing using more traditional, 3-variable models.

Anyways, all that to say that I am not sure there really is a true “basal” rate anyways that is physiological. There’s an average need for insulin in the body at any given time of day, and that is moderated by a bunch of stuff, from food to stress to exercise. Your “basal” rate is just some average that works reasonably well as the baseline from which to make these minute-to-minute deviations. At least, that’s my opinion. Given that, I think any strategy that provides the best jumping off point for those deviations will work reasonably well.

That said, I tend to agree that making daily changes to basal programs is problematic – mainly because it requires a lot more work in troubleshooting. But that’s based on my tolerance for data analysis and my ability to tease out different variables. Someone who is more meticulous in data tracking and perhaps more intuitive might be fine with it.


I agree with this and the rest of your two posts, and would also add that a potential issue with constant tweaking of basal programs and creating a lot of them would be something that in statistical modeling we call “overfitting to the data”—if you try to get your model to be highly complex in order to reflect the nuance in all available data points, odds are very good that you are actually modeling considerable amounts of error (vs the meaningful signal) that will make your model less able to accurately predict future outcomes than a simpler model that doesn’t overfit to the data. In this case, that means there’s so much random, unknown, and not necessarily repeating stuff contributing to blood sugars, we can get off-track if we try to respond and especially make future decisions based on all of it—often it may not reflect replicable components of what’s going on.


This is well-stated. But to go a bit further though, I don’t think any of it is random. There is a reason for everything that happens. Sometimes it takes a lot of figuring out to be able to know the reason.

The first step, given enough time and working with it, is to be able to look back at any spike or drop that happened and know why it happened. Maybe you weren’t able to predict it, but afterwards it makes sense. To understand there is a reason for anything that happened.

So I think a key point is for people to replace the idea of it being “random” with the concept that it is “unknown reasons”, and then work on identifying the reasons after-the-fact.

Once you get to that point of understanding the reasons, managing it becomes much easier. That’s when you can start to work on making better predictions.


“In 1936, Hagedorn and B. Norman Jensen discovered that the effects of injected insulin could be prolonged by the addition of protamine obtained from the “milt” or semen of river trout.”

I mean - who thought of that? Well obviously Hagedorn and Jensen, but what were they thinking?

(Just trying to inject some humor into @Michel’s life)…


I wonder if they had to buy the trout some drinks first.