Diabetes is the third-leading cause of death in the United States after heart disease and cancer, according to research published in PLOS ONE by Samuel Preston, a professor of sociology in the School of Arts & Sciences, and Andrew Stokes at Boston University. Diabetes accounts for 12 percent of deaths nationwide—a much higher percentage than previously thought.
“Another way of saying that is if diabetes were eliminated as a disease process, the number of deaths would decline by 12 percent,” says Preston, who is also a research associate at the Population Studies Center.
As part of ongoing work on obesity, the researchers turned their focus to diabetes. They pulled numbers from two nationally representative datasets, the National Health Interview Survey and the National Health and Nutrition Examination Survey.
Their goal was to follow the health records of those with and without diabetes from life until death. The survey pair provided a sample size of more than 300,000, a subset of which had diabetes. Some cases were self-reported, with people stating that they had diabetes; others came from high hemoglobin A1c levels, a disease indicator.
“People with diabetes have about 90 percent higher death rates than people without diabetes,” Preston says. “It’s huge.”
Despite that number, diabetes isn’t listed as the underlying cause of death as frequently as it should be. (The Centers for Disease Control and Prevention (CDC) puts it seventh overall.) Preston says that’s likely because many other health complications are associated with diabetes, and these complications can mask a single reason for a fatality.
“When someone who died had multiple conditions, the certifier must choose one as the underlying cause of death. That choice is, to some extent, arbitrary,” Preston says. “For example, whether it’s ‘diabetes’ depends, in part, on the density of cardiologists in an area. The most accurate way to identify diabetes’ contribution to mortality is to estimate how much it raises the risk of death for an individual, and combine that with estimates of diabetes prevalence in the population.”
Not doing so can skew health surveillance data, Stokes adds.
“The way we monitor population health in the United States and around the world is primarily through mortality statistics. In the case of diabetes, the burden is potentially obscured because of underreporting,” he says. “Our work aims to reveal that diabetes is a much more important cause than is appreciated.”
Only one other study has looked at such numbers, but using older data. It found that diabetes accounted for just 4 percent of total deaths. Disease incidence has skyrocketed since, jumping more than 300 percent from 1980 to 2014, according to the CDC.
Preston and Stokes stress the need for broad intervention to combat both obesity and diabetes.
“We need something on a population scale,” Stokes says. “It’s a major issue.”
-by Michele Berger
Deaths Attributable to Diabetes in the United States: Comparison of Data Sources and Estimation Approaches. Andrew Stokes, Samuel H. Preston. PLOS January 25, 2017. DOI:10.1371/journal.pone.0170219.
University of Pennsylvania