Many surveys about body weight rely on participants to report their height and weight, because it is less expensive and easier than measuring. Under-reporting one's body mass index (BMI) is common across all gender and ethnic groups, according to the study, but the researchers don't believe the practice makes a huge impact on how we view the nation's obesity numbers.

"In terms of studies examining risk factors of obesity, I don't think the under-reporting is a huge problem," said lead author Ming Wen, Ph.D., an associate professor in the Department of Sociology at the University of Utah.

The study, which appears in the journal Ethnicity & Disease, relied on data from the 2007–2008 National Health and Nutrition Examination Survey (NHANES) and looked at reports from 2,672 men and 2,671 women. NHANES includes both self-reported and measured height, weight, and BMI grouped by gender and ethnicity. The researchers also looked at overweight and obesity status, education, age and poverty status.

Wen and

her colleague, Lori Kowaleski-Jones, Ph.D., found that in all ethnic groups, both men and women overestimate their height. Women also under-report their BMI more than men do, and White women are more likely to do so compared to Black and Hispanic women. The authors speculated this was because White women have a stronger social "desire for a lean body" and were more acutely aware of their weight problems. Those who were overweight, in the oldest age group and who had a college education were also more likely to under-report their BMI.

The researchers said, however, the under-reporting bias is "generally small" with the range of difference between measured and self-reported BMI falling within the 1 BMI unit range.

Wen said their results highlight the care that should be taken when making comparisons of BMI across different U.S. socio-demographic groups.

Dori Rosenberg, Ph.D., an assistant investigator at Group Health Research Institute in Seattle who specializes in obesity prevention and control, agreed.

"The paper underscores that there may be slight differences by race or ethnicity which should be emphasized when making conclusions at the population level based on reported height and weight," she said. "Therefore, when studies are able to use measured height and weight, they should do so."