I’ve made comments about this before and wondering if anyone else has experienced this and had to change the way they bolus since using Fiasp. I find the tail to be much shorter, maybe 3 hours at most. I was a big fan of using the “superbolus” technique, taking 1-2 hours of basal upfront with my boluses and suspending for the equivalent period of time to prevent post-meal spikes without later lows. With Fiasp, I found this doesn’t work well anymore, as because of the shorter tail I will prevent the post-meal spike but then have a spike about 2-3 hours later from suspending the bail (due to the shorter tail I’m assuming). My solution has been to lower my carb ratio, and the nice thing is because of the shorter tail so far I’ve been able to take more insulin for meals without risk of a low later on. One drawback though is I am finding extended boluses are necessary for many more meals using Fiasp, since I’m assuming the unpredictably long tail of NovoRapid often kept working to cover some slow-digesting foods that an upfront Fiasp bolus just can’t. I’m not talking about pizza or pasta, but meals like chicken, rice or potatoes and salad where I wouldn’t have had to extend before.
Yes - I find a similar issue and think it has to do with the protien. Picture below is what I have been seeing. I prebolus around 10-15 minutes and get a pretty flat line. Then about 2.5 to 3 hours later I start to see a rise. If I am paying attention I knock it down with a bolus. I obviously was not paying attention on this day and waited for a cgm alarm to go off.
This lunch would have been heavy on protien (and with 35g of carb was more carb than I would usually have at lunch.).
If I eat higher carb with no protien I do not see this.
Unlike you, I cut back on my insulin per carb with Fiasp, but I think with Humalog I gave more insulin per carb to cover the protien wih the Humalog tail.
I am just using a later.correction right now but am thinking about trying extended boluses and counting protien.
Yep, I agree protein is likely responsible for this, though at least for me, also high-fat meals that delay digestion. I’m wondering if pre-bolusing is actually detrimental with this insulin except for very fast-acting carbs. Fiasp works so fast that pre-bolusing might not be necessary for foods that will digest slowly, and if anything it will cause the peak to happen and finish before the food really kicks in, causing a spike at the 2-3 hour mark which is what both of us have seen. Either way, let me know if you have any success with extended boluses.
I enjoyed success for several years dosing every meal with an immediate (pre-bolus) for carbs followed by an extended bolus to cover 50% of protein grams and 10% of fat grams. The extended bolus time varied from 2-5 hours depending on the amount of protein and fat. It worked very well for me. The extended bolus was easy to live with and dosing errors were small and easy to compensate for. I limited my extended boluses to no more than 1.2 units per hours.
I’ve never used Fiasp and I’m following these topics with interest. I love seeing the graphs and narrative comments. Please keep it up!
I’m curious what solution you (or others) found to that later rise. I’ve been pumping with Fiasp for several days now. I find it’s incredibly aggressive up front, keeping things pretty level after breakfast and lunch, and getting me back into range faster than Humalog did, but about 4 hours after dinner I just start climbing and climbing and then hover around 13 or 14 all night despite multiple corrections. It’s a puzzle. I was thinking an extended bolus at dinner, over 4 or even 5 hours, might tackle the overnight problem.
I ended up going back to NovoRapid. I found the delayed rises to be much worse of a problem than post-meal spikes from Rapid, and I also was having spikes after exercise and waking up that I never experienced on NovoRapid and were difficult to bring down. It’s unfortunate, I think it might be useful to take as an injection for very high GI meals and fruits, but overall I think my life is easier pumping with the old stuff.
I’ve been having issues with my overnight readings where I was going quite high so adjusted my basal rate up, then started dropping very low overnight so dropped it back down, and now I’m back to having the problem with highs. I’m wondering if the inconsistency has to do with fat and protein from my last meal of the day, and maybe I don’t notice it during the day becuase I can just correct as soon as I begin to rise rather than staying high for hours. For those of you who have bolused for fat and protein, can you point me to some guidelines for how to calculate the amount to bolus and also how long to extend the bolus for?
While I don’t have access to Fiasp (I’m in the US), I have been following threads like this with interest. I’m wondering if successful use means not only overhauling mealtime insulin doses but also re-doing the basal rates when using a pump.
The performance characteristics of Fiasp move us closer to endogenous insulin. It seems like there should be some combination of tactics that would give us the benefit of the faster Fiasp peak without the detriment of its shorter tail. In fact I think a shorter tail would often be an advantage.
It took me a long time to learn how to use well the rapid acting analog insulins. Well, I’ll suppose I’ll get a chance to try my ideas once Fiasp gets the regulatory green light in the US.
Thank-you for your update on Fiasp.
Here’s a link to a Fiasp user report from Tim Street in the UK. He’s using Fiasp in a hybrid closed loop system but his blog entry includes comments from others using different delivery systems.
Yes, using Fiasp does mean overhauling ratios and basal rates, at least it did for me. I had to move all my basal rates back or forth (I forget which) by about an hour and lowered them all slightly. This issue of getting my nighttime readings flat is something I’ve always struggled with to some extent. When I fast through dinner to bedtime I often will stay flat, but sometimes I drop very low, and at aother times I go high and stay high all night (even with corrections). Part of it for me may be hormones as I am in that insulin resistant phase right now and the past few nights have been difficult to control (and yesterday I spent most of the day above 8 mmol/L despite upping my basal rates by 14 units and lowering ratios a similar amount in the past day or so). But I also wonder if it’s related to the Fiasp wearing off faster and the protein/fat from dinner continues to drive my blood sugar up. It seems to happen most reliably when I eat a lot of meat at dinner, less so when I eat something else like a salad.
I am still using Fiasp and am really happy with not having to prebolus as early as when I was taking Humalog. To me that is the big benefit.
I deal with the “late rise” by extended bolusing for 50% of the protien (and 10% fat if I am eating a lot of it.). I set the extened bolus time so I deliver about 1.25 to 1.5 units per hour. The Omnipod only allows for 1/2hr increments so I do not get hung up on the exact time too much. My carb ratio is around 1:7 so usually the time is about 1 to 2 hours depening on meal size. Longer if I eat a ton of protien. I find that this completely elimates the late rise.
There are many ways to think about dosing insulin that will get the same results. My thinking is with Humalog or Novolog/rapid, the longer tail naturally covers the protien and makes people think they do not need to bolus for protien. That was me pre-Fiasp. Post Fiasp I pretty faithfully bolus for protien unless I am going to excercise. When I exercise, I skip the extened bolus and let the excercise deal with the protien rise.
The one weird Fiasp thing is that for a high carb meal when I give a lot of insulin, I see a drop at about 6 hours. This is pretty consitant and I usually deal with it by having a snack (or a beer on the weekend.)
Here is a nighttime flatish line (although a little high) for last night… nudged down with a little correction…
Here was a “full on carb pig-out” at the beach yeaterday afternoon/night with Fiasp which set me up for the night-time flatline. Limited prebolsing and wild-a××-guessing at carb counts and over 100g of carbs in a few hours with fatty foods, gelatto etc. Not pretty but it worked and I mananged to keep it below 180 (10). Slight low at one point but not too bad… Note: I am normally a low carb person but I am on vacation.
This is almost exactly what I would try if I had access to Fiasp. Reading your experience increases my confidence that it would work. Before my adoption of Loop I used an extended bolus for every meal of 50% of the protein and 10% of the fat grams.
I’m reading some accounts of people giving up on Fiasp; I respect various people’s experience, but I would try to do what you have done regarding dosing.
This may remain an academic tactic for me as I used to get allergy “pump bumps” when I last used Novolog. I’m happy to read that people like yourself are figuring out how to dose Fiasp.
Accounting for the protein could be the ticket. I fall into the group @AE13 mentions, of people on Humalog who think they don’t need to bolus for protein, because Humalog lasts long enough to cover protein’s later rise. Tonight I had 128 g of roast chicken, but bolusing for 64 g carb for meat just seemed too weird, so I … chickened out. But the straightforward carb portion of the meal was 60 g, and my brain had no problem doing an extended bolus for that, so maybe it’ll balance out in the end. (Currently holding steady two hours later, and I’m sure I’ll be checking my Dexcom all night long.)
Lol
So maybe I missed somehing but if chicken is about 30% protien and you had 128g of delicious roast chicken - that would be about 38g of protien. If you bolus for 50% that would be the equivalent of about 19g of carb which isn’t too scary.
For me with an IC ratio of 1:7 that would be 2.7u which I would extend over 2 hours.
If I was completely wrong about the bolus I would only have to eat like two cookies to cover the insulin so getting it wrong is easy to fix.
I am trying to understand the “50% of the protein and 10% of the fat”.
So for an example of an egg which might have 6 g protein, 5 g fat and minimal carb. 50% of the protein is 3 g, 10% of the fat is 0.5 g. Added together is 3.5. So would you consider this egg to have the dosing equivalent of 3.5 g of carbs which you would then does on your regular I:C ration for that time of the day?
Or is that “almost” so rather you would does for an equivalent of 3.5 g of carbs on the I:C ration but use an extended (square, combo, whatever the pump calls it) bolus and if so, what percent of the extended bolus is up front and how long would you extend for?
@Tim35 - That is exactly it. An egg would have the “equivalent” of 3.5g of carbs and you dose based on your I:C ratio for the day. If I ate a 60g carb breakfast and had one egg, the 3.5g of carb would just be noise and probably within my carb counting error range. If I had 4 eggs for breakfast (14g) and no carbs - then around 2.5 hr later, my CGM would start to slowly go up by about 60 (3.5 mmol/l).
As you noticed, there is only 0.5g carb equivalent in fat in one egg, so I often just ignore the fat and focus on the protein.
For protein and fat, I always use an extended bolus. 0% up front and 100% extended.
I aim to deliver about 1.25 U/hr to cover the protein - but round to the nearest 1/2 hour. So for 1 egg (based on my I:C ratio of 7), I would deliver over 0.5 h and for 4 eggs either 1.5 hours or 2 hours depending on my blood sugar at the time. If BG is lower, I would use 2 hr / higher I would use 1 hr.
So where does the 1.25 U/hr come from? - I started with what @Terry4 used and adjusted it a bit 
- Since it works for me, I have never bothered to look into it further. I suspect if you wanted to go there, the number would be based on the rate at which your body converts protein (ok - amino acids) into glucose. It is pretty easy to measure your carb absorption rate (eat some glucose and see how long it impacts your glucose for). I would assume one could come up with a “protein absorption rate” (a very unscientific name but you get the idea) based on a bunch of variables.
Protein is alleged by some to have effect on blood glucose in the 3-6 hour range (depending on where you look on the internet the numbers are different). For me, I start to see an effect at the 2-2.5 hour mark if I eat protein only (ok protein and fat - like cheese) when I am flat lining. The 1.25U/hr thing usually gets enough insulin in slowly enough to match my “protein absorption rate.”
There is a old TAG group here that may have more info:
No, it’s me who missed something, being shall we say not very good with math. I went with 50% of weight, rather than 50% of the 30% portion of the chicken that is protein. Which is why I plunged in the night. Found a very useful TAG calculator spreadsheet in this forum (@Terry4’s?), but unfortunately I’m not usually near a computer at mealtime. Your explanation about eggs makes sense. Will keep trying.
Did you ever try Humalog Terry, or did you switch straight to Apidra when you were getting pump bumps? I’ve noticed I am getting them too and they are getting worse and worse. Last night my site was so painful I removed my Silhouette early and have a large red welt that is still painful today. I am going to try Humalog and see if it solves the problem.
I used Humalog when it was first introduced in 1996. I used it for several years but one of my doctors wanted to switch me to Novolog and I used it for quite some time. I don’t have good records for my Novolog use but it took some time for the allergic reaction to occur. It was then, about ten years ago, that I switched to Apridra.
I think switching to Humalog is worth the trial. My red and swollen pump bumps took a week or more to completely resolve in me. I think the largest risk is damaging the tissue and compromising its ability to absorb insulin well in the future. You are smart to switch now. Good luck!
Thanks Terry! Humalog seems to be working almost too well right now, but I’m sure that won’t last. No redness around the cannula so far!