That's an interesting finding. I overlooked it till now. This quote from the Luigi Fontana study
Effects of long-term calorie restriction and endurance exercise on glucose tolerance, insulin action, and adipokine production confirms what Chupoman said above. Namely, a higher FFA level is interpreted by the body as the "starvation mode" and leads to the skeletal muscles ignoring glucose in favor of the tissues that cannot utilize FFAs. This is an old evolutionary response to "starvation" noted back in the 1970s in pioneering research by Cahill and Owen.
The key here is high level of FFAs:
http://www.ncbi.nlm....rticle_9118.pdf
Fasting serum free fatty acids were significantly higher in the CR group than in the WD group [WD = western diet, EX = exercise group of endurance runners].
Free fatty acids (mEq/L) :
CR group 0.72±0.35
EX group 0.59±0.18
WD group 0.51±0.20
In order to improve glucose tolerance what's needed is more carbs and less fat in the diet. Simply cutting out fat did the trick for me, when I was getting super high BG upon completion of a ~2 weeks fast. Of course this does not concern people on a ketogenic diet (other than knowing this fact, as Chupoman reminds us). Since CR people are not on a ketogenic diet, they would need to prevent into going into the "starvation mode" associated with high FFAs, which will drive the insulin resistance regardless of all other parameters. I am not saying that I'd know how to accomplish this on CR, only that this is an old metabolic fact. The EX group normally develops higher glycogen capacity, which delays their going into burning fat (and corresponding high level of FFAs).
The CR-IGT group was leaner and went into the starvation mode sooner, resulting in higher level of FFAs. I'd guess that
the CR-IGT group is actually undernourished. IMO they need to up their calorie intake and add some exercise to improve their parameters.
Relevant quotes from the Fontana study:
Mean values for age, total body fat, and lean mass were not different between the two CR subgroups, whereas BMI was significantly lower in the CR-IGT subgroup than in the CR-NGT subgroup.
VO2max was markedly lower in the CR-IGT subgroup than in the CR-NGT subgroup.
.. our data suggest that severe chronic CR, in some individuals, may be associated with a relative peripheral insulin resistance mainly due to a low muscle mass with decreased capacity to take up glucose. [<-- that's related to glycogen capacity in endurance runners]
One possible explanation for the reduced glucose disposal following a glucose load in the CR practitioners may be a protective physiological adaptation to prevent hypoglycemia, which is considered part of the adaptive response to fasting.
...these individuals are in a state of “sensing” severe energy restriction.