Insulin Treatment Linked to Elevated Mortality, but Study Deemed Flawed



PARIS – Patients with diabetes who received insulin treatment faced a markedly increased risk of death in a review of more than 3,400 French adults who were followed for 14 years.

Diabetes patients "treated with insulin at baseline were at increased risk of all-cause mortality," Dr. Emilie Bérard said at the annual congress of the European Society of Cardiology. This finding "provides further information to the debate on the risks and benefits of increasing hypoglycemic treatments" in patients with diabetes, said Dr. Bérard, an epidemiology researcher at Toulouse (France) University Hospital Center.

After adjustment for several potential confounding factors, patients with diabetes who received insulin treatment had a fivefold increased risk for death during a 14-year period, compared with adults without diabetes. In contrast, patients with diabetes who received treatment with other hypoglycemic drugs – metformin or a sulfonylurea – had relative risks of 1.5- to 2.2-fold higher than did those without diabetes. In fact, the relative risk for death faced by patients with diabetes who were treated nearly doubled the relative risk faced by patients with diabetes who received no treatment at all at the time they entered the study, whose risk for death ran about 2.8-fold above that of the people without diabetes.

This study "is on the right track," but the findings must also "be taken with a grain of salt," commented Dr. Lars Rydén, professor of medicine at Uppsala (Sweden) University. "The results indicate something, but we need further studies." The main potential flaw in Dr. Bérard’s analysis is that it applied several adjusters in an analysis of a relatively small group of 171 patients with diabetes. "I think [the analysis] is on the edge of controlling for too many things in a small number of people, which led to extremely wide confidence intervals," he said.

Another limitation of the study was that when the participants were initially assessed in 1995-1996, they did not undergo measurement of their hemoglobin A1c, which meant that the adjusted analysis could not take into account the quality of diabetes control that patients received at baseline, Dr. Bérard said. She speculated that insulin may have boosted mortality by leading 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 by stimulating the sympathetic nervous system and triggering vasoconstriction and promoting atherosclerosis.

Her analysis used data collected from 3,403 randomly selected French adults aged 34-64 in 1995-1996 as part of the third French survey performed as part of the MONICA (Multinational Monitoring of Trends and Developments in Cardiovascular Disease) project, a study sponsored by the World Health Organization. This group included 171 patients with diabetes, including 123 on some sort of hypoglycemic regimen and 48 who were not on any treatment at the time of their entry into MONICA. The treated patients comprised 37 on metformin only, 59 on a sulfonylurea alone or in combination with another hypoglycemic drug other than insulin, 16 on insulin only or in combination with a drug other than a sulfonylurea, and 11 on some other regimen.

Dr. Bérard and her associates determined the vital status of these people 14 years after their initial assessment for MONICA, and found that the mortality rates during that period ranged from 7% in the people without diabetes to 33% in those treated with insulin. Patients with diabetes who received no hypoglycemic treatment at baseline had a 23% mortality rate during follow-up.

The adjusted analysis controlled for demographic factors, duration of diabetes, history of diabetes complications, serious chronic comorbidities, alcohol use, smoking, blood pressure, and other potential confounders.

MONICA is sponsored by the World Health Organization. Dr. Bérard said that she had no disclosures.

Next Article: