I was looking at this discussion because I’ve been trying to determine what the advantages of the insulin analog varients were, and how I might be able to use them if I needed different results than I have been getting -given the equipment available today.
From the very sparse data available, it appears that the makers of these products are reluctant to disclose objective statistical numbers regarding onset times. The published response curves are without vertical axes, based on pure glucose loads with a limited number of test subjects. Some “short” products seem to have faster initial rises than others.
What all short term insulin products share is for the same pure glucose load, over a 6 hour period, they all produce similar outcomes in test subjects. That is the first criteria the FDA uses for approval.
The second criteria is effects when used as directed. The FDA cares about new safety-related side effects, ones than those of the previously approved products. Hypoglycemia from an overdose of insulin is not a side-effect of a product intended to lower blood glucose. It is only a concern of the FDA if its effect is very non-linear with dose compared to approved products and requires special precautions.
If I were a test subject laying on a bed in a clinic all the time and being fed glucose, I don’t think that would care what kind of insulin I was being given. But that is nothing like my life.
The principles I use for my tech, insulin and bolus decisions are well known:
The “best fit” of an infused insulin bolus:
- delivers enough, early enough to slow BG rise
- prevents the BG from rising out of the linear response range into the hyoerglycemic nonlinear range of “glucose intolerance” or dropping into hypoglycemia duirg iut working period.
- matchs the peak of the insulin delivery to the peak of carb absorption.
- matches all the food converted to glucoise by digestion within it working period.
- accounts for the delay between infusion and onset of absorption.
- produces the lowest deviation from average.
My observations from 40 years of using insulin are:
- To get the highest degree of control over post-prandial glucose with any insulin, its delivery has to be matched to my rate of food conversion to blood glucose.
- An insulin with a fixed absorption profile and cannot closely match my wide variation in food composition and conversion to glucose from mea toi meal without adjusting rate of delivery.
- Fixed premeal boluses, at a fixed interval before eating, are not a close match.
- Pre- or post meal bolus whose interval depends on the approximate CH/L/P/F mix of the meal or not, may be a closer match if and when the estimate of the meal composition is accurate, and there are no other factors.
- There is a “depot effect” with infusion. Rather than a steady steam of insulin being conveyed to the bloodstream, it may be held within the interstitial layer, stored and/or delayed and delayed for a short but unpredictable period of time. This effect is used intentionally by long term insulin products, where a short delay variation isn’t critical.
As an engineer I know that any product with a known absorption profile and onset delay time (ODT) could be matched to this by controlling the infusion rate and initiation time (Ti), but without known, accurate BG measurements AND a close estimate of BG at (Ti + ODT), the result will be suboptimal.
As a PWD I know that no matter how hard I try, I won’t get perfect results every time. There are too many variables. Maybe in the future I’ll have a “robodoc” that tracks all my activity, my stress hormones, and my actual BG 24x7, not today.
What does this have to do with my choice of an insulin with my insulin pump?
ODT is more important than relative onset curve and will vary between sites and with a site over time. To know ODT I need to somehow monitor the effect after each meal to determine how well the last bolus is working.
If I know ODT, I can control that variable and bolus size, I can make an educated guess on an average delivery rate for the kind of meal I’m about to eat, and closely estimate the net carbs.
Unless I don’t learn or use that information, the onset curve doesn’t matter to me as a pump user. I can infuse FIASP a little faster or slower, earlier or later to match Novolog, or vice-versa, by using a 30 minute extended bolus and adjusting the percentage initial bolus.
If I don’t use the information available, I’d be totally at the mercy of a (first gen ) hybrid closed loop system.
I’ve seen how poorily it reacts if I don’t get food estimates right, if I don’t bolus at a reasonable to me before a meal. I’vee seen how my TDD varues when I do. If I were dependent on a HC worker, unable to oversee and provide the intelligence to manage the system, I’d rather go back to scheduled MDI than stay on a HCL pump. Staying alive and out of ERs is more important than keeping my GMI and SD low.
When it comes health, I don’t want to be an explorer or a pioneer, but a developer who benefits from from the knowledge of the pioneers. To me FIASP has several practical disadvantages over Humalog or Novolog.
- The pump I use has not been clinically tested with FIASP.
- Reported higher “infusion set” occulsion rates than Humalog. I chose Novolog over Humalog because of its reported slightly lower rate over 72 hours in clinical trials.
I’ve never experienced an “insulin set” occulsion. I have had bad infusion sites, and resolved them by relocating the steel cannula to another site. The ability to
do this was one reason I switched to steel from plastic cannulas.
With my pump being one part of an dynamic interactive system, high quality monitoring of what is happening is more important than the insulin being used.
So I decided that I will use whatever insulin product has been tested by the maker of whatever pump I’m using, and has the lowest cost to my insurer. I will validate every CGM sensor before relying on it. And I will use an realtime reporting/anaylsis app like Xdrip+ that lets me set multiple level alerts to let me know within my target range where I am.