In the evidence link contained in the paper you quote, I note three things that jump out immediately
- GNG goes up on a keto diet by 14% and there is an acknowledgement that it may go up even more as people become more fat adapted. This does not match the conclusion unless 14% is considered negligible, but moreover, there is an acknowledgement that 11 days was both insufficient to make a claim, but suggests that it MIGHT suggest said claim
- None of this applies to people who are insulin resistant, whereby an increase in GNG is expected (I think many on this forum are diabetic, and truly diabetic, not just “was afraid I was becoming” or “should have been diagnosed as it was starting” but with high blood sugars and therefore greater insulin resistance).
- I do not believe that GNG for normal protein ranges (under 2 g /kg bodyweight) would greatly affect the results, but from my experiments, higher than 2 g/kg certainly does, and that is not a “normal” amount of protein for most people. I also note that it was written in 2012 with a hope to study this more, but I do not see further updates. Are you aware of any updates to that train of thought in the past decade since it was written?
Separate from the paper, Dave Feldman has noted that he has a rise in his BG when he eats keto, but hypercaloric (every time after many attempts). Again, it could be more food at the same ratio means more carbs that are the culprit, but with a healthy metabolism, that would only make sense if he was unable to keep up with storing the glucose and it seems more likely to be partially caused by protein (or imho less likely fat). If it happens with someone metabolically healthy like Dave, and in someone with diabetes like me (A1C of 10.5 6 months ago), then I would not be surprised it it happened in most people (the vast majority would never know since they do not eat 250+g of protein a day while wearing a CGM like Dave and I both have for testing purposes).