PRB

The prevalence of childhood obesity in the United States is often tracked using outdated, self-reported data. Self-reported data—when a child (or the child’s parents) is asked for the child’s height and weight—tends to be unreliable.

Research shows that people frequently overestimate their height and underestimate their weight. Most people describe themselves as taller and thinner than they actually are.¹ Shaving off five pounds may seem harmless, but problems may arise when researchers use misreported measurements to monitor characteristics of the obese and overweight population, particularly for children.

Small Errors Make a Big Difference

Typically, height and weight measurements are used to calculate a person’s body mass index (BMI), which is a reliable measure of a person’s body fat.² BMI is highly related to other measures of body fat—such as caliper tests, water displacement tests, or x-rays—but these measures are more difficult to obtain because they require medical equipment. When height and/or weight are misreported, the BMI is skewed, making this otherwise-reliable measure inaccurate.

For people under age 20, BMI categories are more variable than for adults over age 20—the range of each BMI category is much smaller, and it changes with each month of age.³ A narrower range of healthy BMIs means that slight variations in height and weight could lead to more children and teens being misclassified as overweight and obese, compared with adults. For example, a 10-year-old girl who is 4 feet 8 inches tall and weighs 80 pounds has a BMI of 17.9—in the healthy range for her age. If, however, she is 4 feet 7 inches tall and weighs 87 pounds, she has a BMI of 20.2 and is considered overweight. Similar changes to an adult’s height and weight would not have as significant an effect on the resulting BMI category.

Adolescents who misreport their measurements do not grossly underestimate their weight, perhaps because the stigma attached to excess weight is not as strong in childhood as it is in adulthood.⁴ Therefore children might not be concerned with their exact weight, leading to only a small margin of misreporting. However, a difference of even 1 inch and 7 pounds would reclassify a 10-year-old girl from being healthy to being overweight.

Ideally, BMI would be calculated from measured height and weight and not from self-reported or parent-reported height and weight. But the additional cost of measuring height and weight and the permissions required to obtain such measurements makes this method hard for researchers to implement, so many surveys rely on self-reported height and weight (usually for older adolescents and teenagers) or parent-reported height and weight (for younger children).

Reliable State-Level Data Unavailable

The National Health and Nutrition Examination Survey (NHANES), conducted by the National Center for Health Statistics, provides regular estimates of childhood obesity based on height and weight measurements taken by trained professionals. However, NHANES estimates are only available at the national level. Surveys that provide estimates of childhood obesity at the state level, such as the National Survey of Children’s Health (NSCH) and Youth Risk Behavior Surveillance System (YRBSS), are based on self-reported or parent-reported height and weight. In addition, YRBSS data are not available for all states, and NSCH data are only collected every four years, so estimates of obesity and overweight may be outdated before the next release is available.

Currently available data lead to different findings regarding the prevalence of childhood obesity and overweight. Although the age ranges are not identical, these surveys show the differences between estimates based on self-reported data and measured data.

• The YRBSS, based on self-reported height and weight, indicates that 13 percent of high school students were considered obese in 2009.⁵
• Data from NHANES, based on measured height and weight, showed that more than 18 percent of children and youth ages 12 to 19 were obese in 2009-2010. For youth ages 2 to 19, almost 17 percent were found to be obese.⁶
• In the 2007 NSCH, based on parent-reported measurements, more than 13 percent of children ages 14 to 17 were found to be obese; for ages 10 to 17, more than 16 percent were obese.⁷

The different methods of collecting information on height and weight clearly affect the reported prevalence of obesity for youth in the United States. The prevalence of childhood obesity ranges from 13 percent to 18 percent, depending on the measure used. A 5 percentage-point difference represents about 1.5 million children ages 12 to 18 who may be misclassified as obese.⁸ With so much attention being paid to obesity in the United States, especially among youth, a timely and reliable measure of obesity is needed to fully assess the extent of the problem. Furthermore, surveys that use self-reported height and weight often include a variety of behavioral variables (such as hours of television viewing or sports participation), which tempt researchers into using these data beyond their intended purpose. But these types of surveys underestimate obesity among youth, potentially compromising the associations that researchers find in these data.

Current projections show that over the next two decades, obesity among adults will increase by 33 percent and severe obesity (where the BMI is greater than or equal to 40) will increase by 130 percent.⁹ Two cities and two states (New York City and Philadelphia, and Mississippi and California) recently reported modest declines in child obesity based on measured height and weight.¹⁰ However, these four geographic areas do not present a complete picture of childhood obesity in the United States. In order to fully assess obesity today and in the future, researchers need accurate, detailed, and up-to-date information on childhood obesity for all states.

Rachel Cortes is a research associate at the Population Reference Bureau.