I don't believe the authors of the 'insulin increased mortality' study controlled for diet... that would certainly have been a useful add-on. I don't know if the data are available, but I would have thought also that a group of comparable Type 2s given no medication and controlling diabetes by diet and exercise alone would have made even more interesting reading than the actual comparator of metformin alone.
In terms of adding a caveat about 'misusing insulin etc.' I don't think that would be appropriate, as the evidence from this study stands, the authors are saying that insulin as such was the risk factor, not 'misused' insulin. As above,it would be of particular interest to control for diet and exercise.
On a purely personal. anecdotal note, I have found diet to make an enormous difference, as most of you also suggest. I have been vegetarian for 40 years and have never weighed above 8 stone ('old money', sorry!) On the doctor-suggested regime of 16 units levemir and a bolus novorapid of 1 unit per 10g carbs my blood glucose was a complete nightmare swinging up to 25+ and down to 4 and below on an almost constant basis, landing me in hospital on several occasions. This was despite consistently eating a low carb (less than 40g per day) diet (although this did definately help) and being sensible about monitoring for possible lows/highs. Recently, as posted elsewhere on the forum, I switched to a vegan diet (no other changes) and within less than a month, I was able to reduce the levemir to 12 units (now 10) and drop the novorapid altogether. In this crazily short time, my blood glucose is both more stable and lower than it has ever been since first diagnosed. What's more, despite no change in insulin, as I stay on this diet my morning readings are going steadily down, but not going hypo. Definately thumbs up for diet, but how many people would be willing to try being vegan!
