Simultaneous posts. 
The point where glucose does or does not remain in control is a distinct change in the progression of IR. But yes, the time glucose is elevated will get longer in time as well as higher.
I’m looking for a link to the Crofts study that has all the data…
Thanks Google.
OK, this article explains it well. So 2hr BG could be nicely low, but the pancreas has had enough time to pump out loads of insulin and we wouldn’t know. Bugger. Well that kills this theory then.
Roll on the time when we have a retail continuous insulin monitor.
You can still say that if glucose is elevated there’s a problem. 
Question is, is it because even elevated insulin can’t keep it down, or insufficient insulin release because of transitory PIR on keto?
Good question?
Glucagon (pancreatic and stomach lining a-cells; islands of langerhans) to glucose to insulin ratios when less fat and carbs are involved explains Dave’s flatline when protein is the primary source of nutrition which is very slowly and steadily releasing stored endogenous glucose.
Almost as if glucose is acting the same as glucagon breaking down glycogen how strange?
Think of it as neligible liver glycogen, muscle glycogen is locked in muscle, glucagon-insulin regulates lipolysis (FFA). Introducing an external source of glucose doesn’t change glucagon much (baseline of individual?) but insulin cranks up to dispose of extra glucose into glycogen or adipose.
”…However, even under this acute situation, the excess of glucose remains, favoring its final conversion to fat. …”
…More
Thank you all for thisI I have been down the rabbit hole for hours since reading this thread and especially since @carolT posted the video. Of course when the @atomicspacebunny gets involved my cerebral cortex whimpers. I have yet to see the video carolT posted as they say see this first and then saw another video by Dr Bikman…lol I will get there. I have really gotten alot from this post and it’s very important to me. I really agree with Dr Bikman’s point that taking insulin is like giving an addict more of what they are addicted to. It’s insulin that’s the problem…for me. Can’t wait to get off it. Just finishing this one up.
https://www.youtube.com/watch?v=uIVwrlqcyUY
One thing Dr Bikman says is that producing Ketones even if less than .5 tells us you must have low insulin. Is this something we can agree on? I am new to Dr Bikman.
First, liver glycogen must be low to cause the need for ketogenesis. Insulin must be low enough to allow release of free fatty acids into the bloodstream from fat cells, and insulin will halt the production of ketones.
Just how low insulin needs to be for these various things is a bit of a mystery. I remember Richard saying it’s around 13 for the lipolysis part of it. So, if you have naturally high fasting insulin, it’s going to be a stuggle. If you have an elevated insulin release reaction to protein, you’re going to have diminished ketone levels. And let’s not forget glucagon - an altered insulin:glucagon ratio would affect liopoysis but maybe not alter ketone production otherwise(?). [I need to look into the liver side of the equation more]
My pet theory is you can directly relate insulin levels to GKI.
A more detailed explanation of the factors controlling ketone production. Insulin is a controlling factor in various mechanisms.
Control of ketogenesis
The rate of ketogenesis depends upon the activity of three enzymes: hormone‐sensitive lipase (or triglyceride lipase), which is found in peripheral adipocytes, and acetyl CoA carboxylase and mHS, which are found in the liver (Figure 4). The first two of these enzymes, hormone‐sensitive lipase and acetyl CoA carboxylase, are in turn exquisitely controlled by the level of circulating insulin26, which acts to inhibit ketogenesis, and epinephrine and glucagon, which act to stimulate ketogenesis27-30. Insulin inhibits lipolysis and stimulates lipogenesis through deactivation of hormone‐sensitive lipase and activation of acetyl CoA carboxylase, respectively. In other words, a low glucagon/insulin ratio inhibits ketogenesis while a high glucagon/insulin ratio, as occurs with fasting or diabetes, favors ketogenesis through promotion of lipolysis in the adipocyte and stimulation of β‐oxidation of free fatty acids in the liver. SOURCE
One of the original 2KetoDudes and part owner of this forum. I don’t @ Richard any more… he’s a busy man.
But if you search on his posts reagrding insulin you’ll have a lot more to read. 
I was tested - my idea, doc does not believe in taking insulin (!!). Think I came out with a HOMA/IR of 1.6.
Through DirectLabs.com you can get an insulin done for $34. They even test in Massachusetts, which most direct-to-consumer places won’t. They want $32 for glucose but you can probably handle that yourself.
I disagree, especially for those of us lww carb/keto for 5+ years. When I started out, I could get blood ketones over 3.0, no problem, even eating. Now, most days I start under 0.5 and it goes up all day, but rarely above 1.0, and even 36 hour fasts don’t get it to that level… I just did a 4.5 day fast, and the highest measured value I got was 1.9. And then I transitioned into 3 days of high fat, high calorie (Dave Feldman’s protocol), and my ketones went down. Down. And I’ve been guzzling cream, sour cream, and other high-fat foods.
At some point, you become very efficient with ketones. And my blood sugar is always “high” in the mornings if I’m eating. (Went from 72 4.5 days into the fast to 108 this morning, two days after stopping the fast, while eating a TON of fat.) Yet, I’m down about 60 pounds and have gained about 10 pounds of muscle (will know more at my next DEXA scan).
After testing almost 1,900 times with blood sugar (4 different meters), ketones (two different blood meters, ketonix for breath, and urine stirips), I’ve given up on testing. Furthermore, my insulin has been all over the map:
My fasting insulin is all over the map, while my HbA1c keeps going down and I keep losing weight. That doesn’t quite compute. (The ones highlighted in yellow were after 4.5 days of fasting.) I started low carb 1/1/14. The test on 3/6/19 was in the afternoon, fasting, but I had coffee and tea (both no cream) earlier in the morning. The other tests in the morning.
Which part are you disagreeing with? The GKI as proxy for insulin levels?
I agree it is difficult to test for such things given the amount of noise in the signal and infrequent insulin testing. I’m saying if you tested at the same time every day during a fast, you’d see insulin trending the same direction as GKI during that period where there are as few variables as possible to skew the data. Maybe long term, an A1C or CGM average would be more telling for glucose levels. And ketones do vary widely after adaptation. You’d have to add blood + breath + urine and hold exercise at a constant.
Oh and PIR would kick in at a point to raise fasting glucose, which is why a longer term average instead of a snapshot measure would be better.
Pretty much all of it. I have thousands of tests, and if GKI is a reasonable approximation of anything, I would know. I tried tracking GKI, but it was so highly variable as to be useless. And I don’t think taking data at the same time every day is going to elicit much benefit. Will it be better than comparing ketones in the morning with ketones at night? Absolutely. But based on my own testing, even tests taken at the same time each day are highly variable.
And then there’s this:
In order for GKI to be accurate, the tools have to be exact. And they aren’t. In fact, as you can see, they’re pretty bad. What’s my GKI?
I had a 2 hour OGTT with insulin, and my doctor at the time pronounced me insulin sensitive. This was my doctor, Cate Shanahan:
I did this test without going from keto to high carb.
@ctviggen I understand your frustration… My numbers are all over as well. MY BG seems to keep rising until noon before it stops dropping without eating or drinking anything other than black coffee. Very extended dawn for me. When I have fasted longer than 24 hours my numbers are nothing special and sometimes actually high. This takes patience alright and too much data is moot. While I have added some tools for more monitoring I have to understand…just keep doing the right things and don’t hold my breath to every check unless it’s a possible anomaly.
Ok so I have a lot more reading now and reviewing. Who is Richard?