When I posted my first 48 hour review, less than 24 hours ago, I mentioned this graph, which suggested that it starts to take effect very quickly.
This seems to be really quite quick, but it looks like there are compounding factors in the picture, such as basal initiating a drop, so I decided to test the same thing again this morning. This time a little earlier, thanks to it being a weekday and having to get up and go to work, but still, worth a check, to try and make sure there were no outstanding factors.
Although the Drop on the y-axis is slightly larger in the second example (which was expected as I am more insulin sensitive at the time of day the test was done), the graph profile and timings are very similar.
In both cases, we see that there’s a fairly short period of time before the glucose levels start to drop, and then cease dropping after an hour or so (in the second case, I’m going to suggest it’s around 15-20 mins due to the loss of connection and what looks like an almost immediate sudden drop that indicates a few Dexcom issues at the time).
So what does this mean? Well, it’s an n=2 sample, so statistically, not much, but the similarity between the two plots appears to show that it’s not a one off, and this is just how the insulin works. It also suggests for me that on smallish doses of this insulin, my Insulin duration needs to be around two hours, but that doesn’t take into account any tail, which if I’m fair, I’ve not been able to yet see in the results of these experiments. At this level of dose, it looks like it is shorter, or more likely, less pronounced, with more of the available insulin being used earlier in the cycle, than NovoRapid. As I’ve previously said, I’m suspicious that on larger doses there is more of a tail, but that would tie in with the same amount of time overall, but proportionally less insulin in the tail, meaning the effects are more pronounced with larger doses. Others that I’ve spoken to have talked about using three hours for DIA and it working well.
It also means that we need to do some work on the IOB calculations for OpenAPS and Loop given the difference between this and the existing algorithms. Time to get buried in excel I think.
It begs the question about how people use this with MDI or CSII. For meals with a slow absorption profile, there is a real risk of hypoing here if the upfront bolus is too high. It looks like it will be beneficial in the use of a Square wave or dual-wave/combi bolus though. I think these factors will need to be considered in changing to this insulin. People may need to relearn their bolusing strategy dependent on the foods that they eat. Something for both PWDs and HCPs to be aware of when looking at changing.
But I think it also answers the question “Is it really faster?”. That’s two out of two where the same behaviour has been observed and it’s definitely shown itself to be faster acting, even if n only equals 2. To be fair, it’s not like moving from NovoRapid to Afrezza, but from what I’ve seen, there are definite benefits to its use. So it’s doing what it says on the tin, and for me, that’s got to be a good thing….