Insulin Use Associated With Increased Mortality in Older Women

October 20, 2006

ATLANTA -- Excess mortality, common among Medicare-age diabetics, is greatest for women who use insulin to control their diabetes, according to an analysis of data from the Cardiovascular Health Study.

ATLANTA, Oct. 20 -- Excess mortality, common among Medicare-age diabetics, is greatest for women who use insulin to control their diabetes, according to an analysis of data from the Cardiovascular Health Study.

Women who used insulin had an almost fourfold higher mortality risk (hazard ratio 3.61 (95% confidence intervals 2.83-4.60) P

There were 2,391 deaths during follow-up.

Among the findings:

  • Deaths due to dementia and renal disease were more common among patients using oral hypoglycemic agents.
  • Median survival for patients using insulin was about 7.5 years, with 75% dead by the 12-year follow-up.
  • Median survival for patients using oral hypoglycemic agents was slightly less than 10 years.

Earlier studies have confirmed that diabetes significantly increased cardiovascular and coronary artery disease mortality among young and middle-age individuals, but this study provided evidence that the increased risk extends into old age, the authors wrote.

In this study participants with pharmacologically treated diabetes were roughly twice as likely to die of cardiovascular disease as non-diabetic controls (oral hypoglycemic agents, HR:1.99, 95% CI, 1.54 to 2.57 and insulin treated: HR2.16 95% CI, 1.54 to 3.03); and 2.5 times more likely to die of coronary artery disease (oral agents: HR 2.47, 95% CI, 1.89 to 3.24, and insulin: HR 2.75, 95% CI, 1.95 to 3.87).

At baseline all participants were asked about prescription medications taken in the two weeks prior to study entry. Baseline assessment also included psychosocial evaluation. At a subsequent clinic visit, fasting blood was collected and frozen plasma and serum samples were sent to the study's central laboratory. Participants also underwent resting ECG, ankle and brachial blood pressure measurements, and carotid artery ultrasound imaging.

Only patients who reported using oral hypoglycemic agents or insulin were considered diabetic in the study, a definition that was so strict that it may have excluded people who were considered diabetic in prior studies, a methodological difference that might explain some of the findings.

Another potential limitation was the lack of data about the duration of diabetes, an important shortcoming, the authors acknowledged, because duration is an important contributing factor in diabetes mortality.

The study was also limited by the decision to restrict enrollment to community-dwelling individuals, which excluded institutionalized patients who were likely to have greater diabetes-related mortality. As result, the authors said their findings are likely "conservative estimates."