OK there is a graph of insulin response for Ozempic on the Novomedlink site. It is copyright, so I cannot share it here. Basically it shows two stages of comparison between groups, one at baseline, and another after 12 weeks of treat,ment. The baseline comparison is the sample group vs the placebo group, and it can be seen that the two graphs are linear and track each other, so the behaviour of the pancreas is indeed glucose dependant as claimed. Looking at the 12 week graphs, the comparison changes to the intervention group vs "Healthy people", but not the placebo group. so not comparing the same test cohort members. Again the two graphs are linear and track each other.
First observation - although the GLP-1 agonist is present in the blood, the normal GLP-1r route from the alpha cells is still rolling and providing the correct stimulus. There is no evidence that the agonist med is actually contributing anything. To show that the comparison should have been against the placebo group.
Now normally insulin secretion rate used to be expressed as pmol/min. But drug companies prefer to use pmol/kg/min. The effect of this is that the rate is now totally dependant on body weight or BMI. Comparing within a group at the same time is valid, but where the primary intervention is weight loss, then over time, and as the weight decreases, so the insulin secretion rate will rise, which again is shown by these graphs. It is in the mathematics. Again, the 12 week
graphs do not show that the Ozempic has changed the pancreatic response only the magnitude of it which may just be from losing the kg.