Do men and women lie about their weight? Of course they do—but a new study from researchers in Ireland shows we're getting much worse at our ability to estimate how much we actually weigh, and it may be profoundly affecting what we know about obesity epidemics.
Researchers measure obesity using BMI (Body Mass Index), which is a ratio of height and weight that's used clinically to assess whether an individual's weight falls within a healthy range.
Some BMI measurements actually take place in clinics, but much of the population's BMI data is self-reported. So when researchers use that data to calculate obesity prevalence, it matters whether that self-reporting is somewhat accurate.
Previous studies have shown people tend to overestimate their own height and underestimate their weight, and researchers have factored this a bit into their statistical analyses about rates of obesity.
But new research from the University College of Cork in Ireland recently published in the peer-reviewed, open-access research journal PLOS ONE shows that as many as half of the self-reported BMI scores for men and women in the obese and overweight categories may be wrong—down from 80 percent when the Irish survey began 10 years earlier.
There are lots of reasons that men and women lie about their weight. They may not want to admit to themselves they're obese, or they may want to tell their friends and family that they weigh less than they actually do. They may be in self-denial of their unhealthy weight, or they may not want to be labeled as obese.
But because self-reports are so routinely used for large-scale epidemiological studies of obesity, the fact that as many as half of self-reported BMI scores in the obese or overweight category may simply be wrong could skew everything we know about obesity epidemics.
In other words, it could be much worse than we thought.
Based largely on such self-reports, studies routinely indicate that more than a third of Americans are considered obese and another third are considered overweight—which has all sorts of healthcare and disease prevention implications.
But if the Irish research is any indication, the situation could be considerably worse because people routinely underestimate their own weight.
The new data comes from three national surveys (the Survey of Lifestyle Attitudes and Nutrition) done in Ireland in 1998, 2002 and 2007. Each survey questioned 23,000 adults. The surveys identified a "sensitivity score"—the percentage of people in a specific category who are identified correctly.
In 1998, the Irish researchers reported about 80 percent of the scores in the overweight and obese category were correctly identified. In other words, only one in five underestimated (or lied about) their weight. But by 2007, that percentage had fallen to 53 percent—which means that a lot fewer heavy people were accurately categorized.
And because both men and women tend to overestimate their height a bit, which further skews self-reported BMI, the statistics get even murkier when researchers try to use such self-reported data to assess obesity epidemics.
So what can researchers do about this increasing bias in self-reported weight? Unfortunately, not a whole lot. Public education could help, because people really do need to know whether they're obese and at risk for serious complications. Lying to yourself about your weight does both you and the researcher no good.
But the Irish researchers found that, over time, the bias towards reporting a lower weight was most notable in those who are obese—the group that also most needs to know its accurate BMI scores and the health implications.
Closing this growing gap between self-reported BMI scores and accurate, clinically reported BMI scores is important because it "brings us one step closer to accurately estimating true obesity levels in the population," the researchers wrote in their journal submission.
Which raises a troubling question. If half of us are lying to ourselves about how much we actually weigh, then how likely are we to take the needed steps to deal with an obesity epidemic that may, in fact, be a great deal worse than we thought?