The Unpredictability of Diabetes Predictions

800px-insulin_penAs if you don’t already feel guilty for double helpings of pie on Thanksgiving, Black Friday brought another reason to fret over an unhealthy diet: new diabetes projection numbers that suggest insulin injections could someday be a new shared family Thanksgiving ritual. In the journal Diabetes Care, Elbert Huang, assistant professor of internal medicine at the University of Chicago Medical Center, published the results of a new model that predicts more than 44 million Americans will suffer from Type 2 diabetes by the year 2034, a rate of disease that will cost the U.S. an estimated $336 billion each year.

Huang’s model, constructed with Anirban Basu, Michael O’Grady, and James Capretta, bases its projections on current prevalence of diabetes combined with projections about how demographics and diabetes care are expected to change in the coming 25 years. The near doubling of diabetes the paper predicts (currently, about 23.7 million Americans are diabetic) would result from the combination of more and more Baby Boomers entering age groups where diabetes is more common and people living longer with diabetes – an ironic byproduct of improved medical management of the disease. Surprisingly, Huang’s model does not predict much of a change in the rates of obesity, the largest risk factor for diabetes. That’s not entirely good news, though; Huang’s logic is that America has already reached a sort of flab ceiling where the proportion that is overweight or obese (currently holding steady at 65%) simply can’t get any higher.

Or as Huang told my colleague John Easton: “we anticipate that the population will reach an equilibrium in obesity levels, since we cannot all become obese.”

If Huang and his co-authors are wrong about that piece of their model, the diabetes rates could be even higher as we enter the 2030’s. And while that would mean Huang’s predictions were off, it wouldn’t be the first time diabetes researchers have underestimated America’s diabetes boom. As collected by Easton, scientists have regularly low-balled the numbers in their predictions about the diabetes epidemic, missing the mark by millions of cases.

In 1991, a study by the American Enterprise Institute predicted that diabetes numbers would double to 11.6 million by the year 2030. As mentioned above, here in 2009 we have 23.7 million cases of diabetes, more than double the AEI estimate – and they actually warned that they could be overestimating the threat “under the overly pessimistic assumption that there will be no scientific or medical discoveries to reduce the prevalence of diabetes.

A 1998 study that sought to foretell future diabetes rates for several countries around the world placed the United States at 21.7 million – a good guess except for the timing, since this was a projection for 2025. This model was equally conservative for the diabetes numbers in other countries, predicting 2025 figures of 57.2 million and 37.6 million for India and China, respectively. According to the International Diabetes Foundation, India has already passed the 50 million mark, while China is keeping pace at 43 million.

And as recently as 2001, a CDC estimate was only 29 million U.S. cases by 2050 –  a number we haven’t yet reached, but one that Huang’s model predicts the U.S. to zoom past early in the next decade. “Our projection of diabetes burden in the U.S. indicate that the situation may be more alarming than previously believed,” those authors concluded…well, who’s alarmed now?

The shortcomings of previous models likely resulted from underestimating just how rapidly the rates of obesity would grow in America, as well as misreading how U.S. population numbers would rise over the last couple decades, Huang said. But such long-term estimates, even if they prove wrong, are still valuable in formulating a plan for dealing with the future costs of a pervasive disease such as diabetes. Huang’s model, one of the first to assess the future costs of diabetes in addition to the size of the diabetic population, seeks to make health care policymakers aware of the more than $300 billion yearly diabetes bill that will have to be addressed in the future if the disease rates hold steady.

In a companion paper published this fall in the journal Health Affairs, Huang and his colleagues argued for using such long-term projections to inform health care policy. Currently, Huang said, the 10-year window used by the Congressional Budget Office to assess new health-care legislation marginalizes the impact of spending on preventive medicine, which can show small benefits in the short term but big results down the road.

“It takes many, many years for patients to develop obesity, diabetes, and then diabetes-related complications,” Huang said. “We also now know that efforts to prevent diabetes-related complications can oftentimes take decades of follow-up for there to be observable benefits.  That means preventing diabetes and diabetes complications takes a long-term investment.”

Thus, funding earlier medical intervention – such as the diet and exercise changes proven successful in the Diabetes Prevention Program trial – could save the U.S. money in the long run. Implementation of those measures could also be the torpedo that sinks Huang’s model, leading a blogger in 2034 to chuckle at what a huge overestimate 44 million diabetics turned out to be. But that’s a miscalculation Huang and his colleagues would be happy about.

“The diabetic population could certainly be much smaller if there are dramatic changes in the weight distribution of the country,” Huang said. “That will require significant changes in dietary and exercise habits. These habits could change due to the many public health interventions underway and the rising public consciousness regarding the obesity problem. The recent changes in smoking habits suggest that changes in public habits are possible.”

Update: MedPage Today interviewed Dr. Huang about his study, you can view the video below:

http://www.medpagetoday.com/Medpage-Player/17201/ window.onload = function () { var q = (document.URL); document.getElementById(“mptplayer”).src += q; }

About Rob Mitchum (526 Articles)

Rob Mitchum is communications manager at the Computation Institute, a joint initiative between The University of Chicago and Argonne National Laboratory.

%d bloggers like this: