Study: U.S. Obesity Underestimated
Statistics Often Use Self-Reported Weight & Height, Which May Be Flawed
WebMD News Archive
May 2, 2006 -- Obesity may be more common
that previously thought in the U.S.
In the Journal of the Royal Society of Medicine, researchers note
that national obesity statistics typically rely on self-reported weight and
height, which are often wrong.
Those inaccuracies often make people sound lighter or taller than they
actually are, write Majid Ezzati, PhD, and colleagues. Ezzati works at the
Harvard School of Public Health.
Self-reported weight and height don't always match reality, so U.S. obesity
statistics are too low, Ezzati's team argues.
The researchers recalculated America's obesity statistics, adjusting for
those errors. The result: The nation's obesity estimates went up.
Corrected Obesity Statistics
Obesity is defined as a body mass index (BMI) over
30, according to the CDC.
In 2002, 28.7% of men and 34.5% of women in the U.S. were obese, Ezzati and
The uncorrected estimate for that year indicated that 16% of men and 21.5%
of women were obese.
Ezzati's team based their corrections on data from two large, national
surveys of U.S. adults:
- Behavioral Risk Factor Surveillance System (BRFSS): Given by telephone
- National Health and Nutrition
Examination Survey (NHANES): given in person, with some participants measured
and weighed afterwards
Ezzati and colleagues compared BRFSS and NHANES data for similar years. They
found that people tended to report their height and weight more accurately in
person than over the phone, but that all self-reports generally missed the
Weight, Height, and Reality
Women tended to underreport their weight, the study shows. Men didn't do
that, but men aged 20-44 tended to overestimate their height more than women,
especially in telephone interviews.
After age 44, men and women overestimated height to a similar extent. Height
often dips with age. Middle-aged or older adults who haven't measured their
height lately may mistakenly think they're still as tall as in their youth, the
Such errors stack the deck in favor of a lighter BMI (body mass index). BMI
is calculated based on height and weight. A BMI of more than 25 but less than
30 is considered overweight, while a BMI of 30 or higher is considered
There are other ways to calculate size and shape, such as comparing waist
size to hip size. But researchers often use BMI to track obesity.
If height and weight aren't right, neither are BMI and obesity statistics.
It's like looking in a circus mirror that makes us look taller and leaner than
we really are.
Where Obesity Lives
Ezzati and colleagues identified the states (and Washington, D.C.) where
obesity was most common in 2000, based on the new calculations. Here are those
findings, along with the percentage of obese men or women in those areas.
Highest prevalence of obese men:
- Texas (31%)
- Mississippi (30%)
Highest prevalence of obese women:
- Alabama (37%)
- Washington, D.C. (37%)
- Louisiana (37%)
- Mississippi (37%)
- Texas (37%)
- South Carolina (36%)
Lowest prevalence of obese men:
- Colorado (18%)
- Washington, D.C. (21%)
- Montana (21%)
Lowest prevalence of obese women:
- Montana (16%)
- Colorado (24%)
- Massachusetts (27%)