1) Lack of sleep coupled with disrupted day-night cycles predisposes to diabetes and prediabetes. Night-shift workers take note.
2) Compared to those using metformin alone, type 2 diabetics who also took insulin needed less insulin and had better blood sugar levels. Real-world benefits are not entirely clear.
Diabetic Mediterranean Diet 
As one using metformin plus insulin. On sufficient min dose ( 500 to 750 mg) that cuts liver back on its excess release of glucose, the met really cranks back on insulin add requirements.
The trick is to be aware of the liver cutback and factor into amount of insulin dose.
Met cuts back glucose at head end – liver; while insulin helps on back end on glucose already in blood stream to get it absorbed .
WHich is better? met is cheap; insulin is expensive.
For me needing to cut back dawn effect,met is real miracle.
Why my liver is pain is unclear to me.
Jim, I’m aware of a type 1 diabetic (on insulin of course) who recently started on metformin to reduce his need for insulin. It worked. Still needs insulin, but not as much.
Dr. Parker:
Most kind thanks for comments.
The other aspect I tend to think of is the exercise burn issues..
Using metformin tends to keep the excess stored glucose in liver when met works. If not and only insulin, one now has to either burn off excess glucose by more exercise or hope fat cells stores the stuff -add weight – maybe.
Lastly, in past when my liver was hammering mer hard , I would see Insulin resistance elevated and not see insulin work effectively till excess glucose burned off walking 2 miles and then I would BG drop and see insulin that had been previously circulating in my body take off.
Many times for me , back then, 25 units of 75/25 in early am was like sticking needle into wood post. It might as well have been water.
What all this means I am not sure but the metformin/insulin pair is working well. I am now on metformin and 1 to 2 units of humalog lispro and off lantus ( was 10 to 15) 2 units now craters me, off actos, off 26 units of 75/25 humalog.
Lastly thank you most kindly for your excellent websites and mediterranean /paleo diets and most helpful information for us all.
Even with metformin and insulin – proper diet/carbs control and hearty exercise most critical to keep mess contained.
Jim, in view of recent reports linking thiazolidinediones to bladder cancer, I’m even more inclined to recommend diet and exercise before drugs, when able.
(To be clear, metformin is not a thiazolidinedione, and it has a pretty good safety record.)
Dr. Parker:
many kind thanks for responding
do you know the connection between metformin and ldl? i recently watched a lecture by tara dall about clinical applications of lowering apo-b (ldl-p) using metformin in insulin-resistant patients, even before diagnosis of diabetes or syndrome x.
http://lecturepad.org/index.php?option=com_content&view=article&id=950:advanced-lipid-testing-comes-alive-part-1-of-4&catid=150:live-lectures&Itemid=384
any thoughts?
also, there seems to be an interplay between metformin and improving insulin signalling (at least in mice on a high fat diet… whatever that means… probably loaded with polyunsaturated fats). http://www.diabetesincontrol.com/index.php?option=com_content&view=article&id=6740
Hi, Ed. As I recall, improved serum lipid numbers on metformin therapy was a strong selling point early in metformin’s marketing history.
-Steve