A few authors I know of: D’Agostino; Noakes; Volek.
Part of the conclusion:
Conclusion
We demonstrated that a habituating to a LCHF for ≥ 4weeks in 30 to 50-years old competitive male athletes resulted in equivalent short duration, high-intensity performance without differences in calories, training load, and body composition across groups.
Not a surprise to me… I could have told them that! 
I thought a very interesting fact was that 30% of the study participants of middle aged trained athletes started as pre-diabetic, and they didn’t know. Very likely TOFIs. The high carb religion in athletics is strong.
I also thought that was interesting. And, in the US, they only use one marker, which is fasting glucose, to determine diabetes. You can sometimes pass that test, but be diabetic anyway.
I think if you’re not exercising “a lot” (and I don’t know the level of this), you can do quite well with keto. There may become a point where some carbs could be useful. Zach Bitter has stated that he can recover well with most of his routine, but if he starts exercising a lot, say twice a day, or doing tons of miles per day, he needs to add some carbs to recover. But he’s a guy running 100 milers, so he’s not the “normal” keto person.
I was under the impression that HbA1c was the primary diagnostic marker for pre/diabetes in the US, no?
Elevated fasting glucose (spot measure) might tip off a physician that ordering a longer term average metric is warranted for testing.
The HbA1c is still fraught with certain limitations - discussed elsewhere - but it’s probably more revealing of a chronic condition than a glucose check at any particular moment in time.
Still doesn’t really address the extent of glucose excursions around the average (nor perhaps the presumed average “life” of the red blood cells being gylcolated.)
I was just at a new (to me) general care practitioner. She said they don’t do HbA1c unless the fasting blood sugar is high.
I think you’re correct that HbA1c is the indicator for diabetes, but you only get it taken if you have higher blood sugar.
I can see the argument that fasting glucose = indicator of diabetes, but I wonder how true this is? I’ve seen results of a Kraft test where the person’s glucose response was great (glucose barely budged), but their insulin response was horrendous – way high. Could focusing on fasting glucose alone miss someone like that?
Maybe there’s a group where high fasting glucose = diabetes, but another group where “normal” fasting glucose = diabetes (or maybe “overactive” pancreas?)?
I would love to go back to 10 years ago, and see what my HbA1c and fasting insulin were. I was just over the “fasting” blood glucose (101) back then, but no one said anything. But I’d bet good money that my HbA1c was bad and my fasting insulin was terrible.
Alas, I didn’t even know what those terms were 10 years ago.
I think HbA1c, glucose, and fasting insulin should be the minimum you get. No one pays attention to insulin, when that might be the best marker, HbA1c next, glucose last. Which one do they use? Glucose. Only, unless it’s high. And they never take insulin. I had to ask for it.
Yes, I did read that study. Very limited in scope (4 weeks) The main take away for me was, improved glycemic control when eating the LCHF diet.
‘Glucose reduction during carbohydrate restriction predicted the elevation in fat oxidation rates during exercise suggesting that glucose response is linked to systemic fat oxidation.’ Nothing new here. “LCHF may represent a therapeutic strategy to improve glucose levels, particularly in those at risk for diabetes, without compromising high-intensity exercise performance in middle-aged athletes.”
While short-duration high-intensity exercise (6 × 800 m) would be sufficient to reduce muscle glycogen content they did not measure actually muscle glycogen content so they cannot say for certain what levels of muscle glycogen were achieved and if they were associated with elevated fat oxidation levels during the high intensity exercise.
At higher exercise intensities, ATP synthesis demand increases, and fat cannot entirely meet the rate of ATP synthesis, so glucose oxidation increases. At higher exercise intensities fat cannot synthesize ATP fast enough for the demands of skeletal muscle fibers. Yes, fat utilization does yield a higher amount of ATP, glucose utilization is much faster, therefore necessary for ATP synthesis.
Have a look at Dr. Asker Jeukendrup and his studies on triathletes. High carb intake was associated with faster finish times. Dr. San Milan also has great info. When it comes to sports and elite levels there is not one body of research that shows that carbohydrate are bad for performance.
For most of us normal folk, like me, it doesn’t really matter. I exercise at most 5 days a week, and within that time frame, I can easily replace whatever glycogen I need. And I’m only working out say 6 hours a week or maybe 7? And if I go slightly slower, it doesn’t really matter.
Athletes trying to perform at their best might be a different story.
If your exercise habits were illustrative of “normal folk” the health care industry would be half its current size.
No, I believe it’s still serum glucose above a certain level. Certainly doctors use HbA1C as well, but as Prof. Bikman points out, it relies on the assumption that our red blood corpuscles live for precisely 90 days. If we are healthy and they live longer, they are inevitably going to pick up more glycation, and the reading will be falsely high. If we are unhealthy and they don’t make it to 90 days, the measured level of glycation will be falsely low.
In other words, the test has to be used with caution. But that is true of every test in our medical armamentarium.
Bob
Spot on, mate. It frustrates me (a lot!) that our doctors are supposed to be the experts in all this, yet they don’t understand the basics. Insulin is the critical marker here… if it takes a shed load of insulin to stuff the muscles full of glycogen to get blood sugar down, this is a massive red flag and needs to be managed.
When they test only fasting glucose (as was the case for me), or even with HBA1C, they have no clue whether there is some metabolic issues going on (ie the level of insulin required to get the blood sugar down). It’s the high insulin that causes the chronic disease. It’s that that needs managing. Yet almost every dr simply thinks that because BG is under 6 we are all good… no, No, NO!
Grrrrrrr…… 
That was the late Dr. Joseph Kraft’s contention and the point of his life’s work as a diabetologist.
He could infer, from the pattern of insulin response, who was going to receive a diagnosis of Type II diabetes down the road. He believed that people actually developed diabetes as much as 20 years in advance of the official diagnosis.
Today we call it insulin resistance and hyperinsulinaemia. Dr. Kraft called it diabetes in situ or occult diabetes.
So I think we’re all describing the same thing…
Tests typically used to diagnose pre/diabetes are often ordered in an incomplete manner. The tests themselves are inherently imperfect. The real health culprit is insulin - but even this level in the blood is a fleeting metric, so in reality an insulin test at any moment is essentially as limited as is a serum glucose test.
It’s the extent of the range (excursions) of both insulin and glucose that tell the full story.
Solution: eat carbs very sparingly. Enjoy the ride.
Interestingly, Robert Lustig says that the American Diabetes Association actively discourages measuring insulin, for reasons he considers bunkum. And loss of glucose control remains the official diagnostic.
Can’t agree there, it typically (was) 3 A1C’s, and now many are using LP-IR since NMR’s are becoming pretty mainstream and they’re part of it unless you’re at an outdated cookie cutter doc.
Prior to Semaglutide I always had a high-ish fasting number, but since my LP-IR was amazing doc says totally ignoreable then.