I've come to a sketchy, off-the-wall, hypothesis that thermoregulation -- more to the point, recurrent drops in body temperature to near-hypothermic levels -- may be either a primary issue, manifesting in secondary hyperglycemia (aka diabetes), or a trigger for the expression of T1 or T2 diabetes.
Since in most people, body temperature tends to drop to its lowest about 2-3 AM, the same time as the infamous "middle of the night lows", I further insinuated (but did not postulate explicitly) that the hyperglycemic episodes known as "Dawn Phenomenon" could be the body's attempt to protect itself from what it perceives as hypothermia.
Where this hypothesis seems to fail is in the case of people like my Other Half, who wake up shivering, feeling extremely cold, and with low blood glucose levels. It also fails where he is concerned in that his average body temperature is in the "normal" (98-point-something to 99-point-something) range. While he is not completely insensitive to cold, the weather has to drop way below freezing before he'll consider the possibility of putting on a pair of gloves for an extended stay out of doors. Put us both in a room that is regulated to "normal" thermal comfort levels, and he'll be sweating bullets in a T-shirt while I'm shivering in a sweatshirt.
In defense of the thermoregulation theory, as far as we can tell, he did not go persistently hyperglycemic until this winter at work, when his normal work environment got chilly even for him, and he spent half the late autumn driving a vehicle that did not have a working heater. He's on Metformin, which inhibits glycogen release, so his body is not able to compensate for a low.
On the other hand, The Other Half had a "waking up both freezing AND hypo" episode some years ago, when he was still considered "borderline" enough to check blood glucose regularly and try to follow an ADA diet plan, but not in need of any controlling medications. We had been walking all around a Renaissance Faire the day before, and camped in an chilly, unheated cabin that night. So environmental issues may have played a role. Or "Dawn Phenomenon" may be more of an issue of excess calories consumed at night than of thermoregulation. Or something else.
I also mentioned in my previous post that the thought experiment flashed me back to grade school. One girl in my fourth grade class, regardless of how warm or cold it was in the (reasonably well-heated) school, would wear her winter jacket in class. She said she was cold, even in the jacket. This was a marked change from third grade, when she was -- like us -- in shirt-sleeves in the classroom. Mid-year, we were subjected to mandatory four-tine tuberculin tests. Shortly afterward, this girl disappeared from my classes. The rumor I'd heard was that she had tested positive on the tuberculin test and had been sent away for treatment. The flip side of this flashback is recalling that Dr. Faustman is looking at tuberculosis vaccines in her T1 research. I don't know if persistent chill is a symptom of tuberculosis. I don't know if diabetes causes a false-positive on a tuberculin test, and I don't know how the tuberculosis vaccine under study is expected to inhibit the expression of Type 1 diabetes. But the coincidence -- chills, tuberculosis (vaccine or exposure), and diabetes -- reminded me of that girl, and gave me cause to wonder.