Two megatrials on Type II diabetes and a covering editorial . The detail is complex but the take home message is that intensive management of blood glucose increased mortality. Received wisdom is that blood glucose control prevents renal failure and blindness (which is due to small vessel disease) but doesn't help coronary artery disease and stroke (which are due to large vessel disease). So pay attention to smoking and weight, statins, blood pressure and aspirin.
I'm tempted into some generalisations about not reducing complex systems (the human body, health services etc) into single targets.
The other message is of course that you can't take anything for granted - you actually need to do the trial. Most people would have thought it was obvious that better blood glucose control was a good thing, given that the primary problem in diabetes is high blood glucose. (But then again the high glucose is the effect of something else - insulin or insulin receptor problems, and insulin affects more than just blood glucose, as can be seen from people who have very high insulin levels from insulin producing tumours.)