Insulin Treatment: A Possible Risk Factor for Mortality   (September 11, 2011)

A recent news story discussed a presentation given at the European Society of Cardiology last month, 14-year risk of all-cause mortality according to hypoglycemic drug exposure in a general population.

The authors of this abstract studied 183 people with diabetes (48 untreated and 135 who were treated with insulin or oral diabetes medications) who had been part of a 14-year study of mortality of 3403 middle-aged people (35-64 years old) in France. Thirteen percent of these PWD were on insulin alone or on insulin in combination with other hypoglycemic drugs -- except sulfonylureas; 30% were on metformin alone; 48% on sulfonylurea alone or in combination with other hypoglycemic drugs except insulin, and 9% were on other, undefined, treatment programs. (I must add that I can't quite figure out why these categories were set up as such - there logically should have been some people on both insulin and sulfonylurea therapy, and if so, I'm not sure where they would have been categorized. But this is what they chose to do.)

The data indicated that death rates during the 14 years varied from 7% in people without diabetes up to 33% in those treated with insulin. The rates were 23% for untreated diabetes and 21% in diabetes treated with hypoglycemic drugs. The authors concluded that "diabetics treated with insulin at baseline were at increased risk of all-cause mortality."

How might insulin cause increased risk of death? The news story quotes the lead researcher speculating that insulin might have led to a "relatively high rate of hypoglycemia, it may have caused weight gain, or it may be a marker for more serious diabetes. In addition, insulin treatment may have caused cardiovascular morbidity or mortality..."

My biggest concern is that the conclusion of this study could lead to the belief that it was the insulin therapy that somehow caused the people with diabetes to die.  I don't think that that is an appropriate conclusion; I suspect that the correct conclusion might well be that the insulin therapy was merely a marker for more serious disease.

Let's look at some of what's missing from the abstract.

We don't have any information on a baseline characteristic of great importance: what was the A1C at time of entry into the study. We also don't know why some of these patients were on insulin, some on no therapy, and some on other hypoglycemic drugs. One might logically assume that the people on insulin were the "tricky" cases where physicians had already tried other treatment programs which had failed, and the patients were already at high risk: their diabetes control may have been terrible at the time of entry into the study. Also, we don't know if some of these patients were under the care of experienced diabetes teams, and whether the outcome was different for people treated by specialists. And there was only a very small number of people with diabetes in this evaluation.

Additionally, there's no breakdown of type 1 vs. type 2 diabetes. The authors describe adjusting their data for a huge number of factors including "duration of diabetes, history of diabetes complications (diabetes with renal, ophthalmic, neurological, peripheral arterial disease or atherosclerotic cardiovascular disease) or severe chronic disease (chronic renal, heart or respiratory failure, cirrhosis or cancer), living area, age, gender, educational level, alcohol consumption, smoking, blood pressure, and LDL- and HDL-cholesterol." Why not also clarify if people with T1D were included or excluded? One might assume that only a few of these patients (if any) had T1D, but it would be interesting to know, and to know if their outcomes were different from people with T2D.

It is clear that any treatment (or nontreatment) of diabetes in this study resulted in increased likelihood of death over the 14 years. Or to put it differently, a group of middle-aged people with diabetes tend to die more over 14 years than do middle-aged people without diabetes. It's also clear that people on insulin therapy were even more likely to die.

But saying that being on insulin therapy was the reason they died is inappropriate. We simply don't know.