How accurate are obesity predictions?
Forecasts state that obesity is still on the rise, long term, in America. How are these predictions made?
David Paul Morris/Bloomberg via Getty Images.
A new forecast by researchers at Duke University predicts that 42 percent of American adults will be obese by 2030, even though obesity rates have stayed level in women and risen only slightly in men over the past 10 years. How accurate are obesity forecasts?
They’re usually too large, but not by much. In 2003, the Centers for Disease Control and Prevention predicted that 40 percent of Americans would be obese by 2010. The actual 2010 adult obesity rate was 35.7 percent [PDF]. In 1999, the British Department of Health projected that 25 percent of Britons would be obese by 2010, a prediction that it updated in 2006 to 33 percent. The actual rate was closer to the more modest 1999 forecast: According to the most recent Health Survey for England, 26.1 percent of adults in England were classified as obese in 2010.
In general, it’s too soon to know how accurate obesity forecasts are. Obesity forecasting started to take off only about 10 years ago, so we have yet to reach the time periods for which most researchers have made predictions. However, different obesity projections sometimes conflict with one another, which means that at least some of them are inaccurate. The Duke study [PDF] out this week used variables such as unemployment rate, fast food prices, alcohol prices, and fuel prices to come up with its prediction of 42 percent. That contradicts a relatively simple 2008 forecast that used the change in American obesity rates between 1970 and 2004 to come up with a predicted 51 percent rate of obesity in 2030.
Researchers sometimes have to hedge their bets because they don’t trust people to be honest about how fat they are. A 2009 report [PDF] estimated that American obesity rates in 2018 could be anywhere between 38 and 47.5 percent; the study’s authors attribute that nearly 10-point margin to “the tendency of individuals to understate their weight in telephone surveys.” Some American obesity forecasters base their projections on data from the more reliable National Health and Nutrition Examination Study, a CDC program that clinically measures participants’ height and weight—but which uses a much smaller sample size than studies conducted via phone survey.
Researchers use the body-mass index, the ratio between a person’s body mass and the square of his or her height, to determine obesity rates. Beginning in 1985, the National Institute of Health defined obesity as having a body mass index above the 85th percentile, based on gender. In 1998, the NIH changed its definition of “obese” to having an absolute BMI of 30 or higher and added a new category, “overweight,” defined as having a BMI above 25.
Explainer thanks Karen Hunter of the Centers for Disease Control and Christopher J. Ruhm of the University of Virginia.
L.V. Anderson is a Slate assistant editor. She edits Slate's food and drink sections and writes Brow Beat's recipe column, You're Doing It Wrong. Follow her on Twitter.