Four years ago, I showed my epidemiology class how misinterpreting p-values had lead to the headlines about how the obesity epidemic was leveling off in the US.
Two years ago, I wrote about essentially the same press release being issued, with new data and the same incorrect interpretation.
Then a week ago, the same headlines hit the airwaves - Obesity may have peaked was the optimistic line.
And I can predict with a high degree of certainty that you will see the same story in the second or third week of January, 2014.
There's nothing I like more than good public health news - strike that, there's one thing I like better: accurate good public health news. And I wish I could be so optimistic about the obesity epidemic in the US. Heck, I'd like to be more optimistic about my own obesity, but that's another story.
So what's going on here? Why do we get told, every two years, that the obesity epidemic is leveling off when the very data that these assertions are based on tell a very different story?
Every two years, the NHANES reports out the proportion of men and women in the US it finds to be obese. The NHANES is a very well-funded, very well conducted research study of the nation's health. The people in it are very well selected to be representative of the population. They are weighed very accurately; their height is measured precisely. And there are enough people in the study to say something definitive about the prevalence of obesity in the country.
But, there are not quite enough people in the study to say definitively whether the proportion of Americans who are obese has grown or stayed the same. The year-by-year growth in the proportion of people who are obese in the US has, according to these data, risen about 1% per year, which is a small change over two years, but a huge change over a decade or two.
So, when they compare the NHANES data from one year to the NHANES data from the year before, they find no significant differences. They then make the classic misinterpretation that "no significant difference" means "no difference" or "almost no difference", and out goes the press release stating that the obesity epidemic in the US may have peaked. And pundits opine on why the end of the growth in the obesity epidemic has arrived (or will shortly). But they all always overlook a much simpler interpretation. Every time these data come out, they are more consistent with that small year in, year out, growth rate than they are with slowing down the proportion of obese Americans.
"Obesity rates have increased since the 1976-1980 survey period. There was no significant change in obesity prevalence, however, between 2003-2004 and 2005-2006 for either men or women."
CL Ogden, MD Carroll, MA McDowell, KM Flegal. (Nov 2007). Obesity among adults in the United States--No statistically significant change since 2003-2004. NCHS Data Brief No. 1.
"No statistically significant linear trends in ... high BMI were found over the time periods 1999-2000, 2001-2002, 2003-2004, 2005-2006, and 2007-2008 among girls and boys."
CL Ogden, MD Carroll, LR Curtin, MM Lamb, KM Flegal. (2010). Prevalence of high body mass index in US children and adolescents, 2007-2008. Journal of the American Medical Medical Association 303(3):242-249.
"The increases in the prevalence of obesity previously observed do not appear to be continuing at the same rate over the past 10 years, particularly for women and possibly for men".
KM Flegal, MD Carroll, CL Ogden, LR Curtin. (2010). Prevalence and trends in obesity among US adults, 1999-2008. Journal of the American Medical Association 303(3):235-241.
"There was no change in the prevalence of obesity among adults or children from 2007-2008 to 2009-2010."
CL Ogden, MD Carroll, BK Kit, KM Flegal. (Jan 2012). Prevalence of obesity in the United States, 2009-2010. NCHS Data Brief No. 82.