is a considerable amount of data regarding waist-hip ratios in adults
and children as it relates to the development of hyperlipidemia,
insulin resistance, hypertension, and diabetes mellitus. It has been
speculated that waist circumference alone may be a more useful and
more easily obtainable index in both adults and children. Fernández
and colleagues utilized the NHANES III data set to determine 10th,
25th, 50th, 75th, and 90th
percentiles for waist circumference by ethnic background. Data were
analyzed from African-American (AA), European-American (EA),
Mexican-American (MA) boys and for girls at each age from 2 years
through 18 years. The data were gathered by trained technicians using
a tape measure just above the uppermost border of the right ilium at
the end of normal expiration. The raw data were modeled utilizing an
indicator variable and logistic regression. The data set included
4769 boys and 4944 girls.
A number of important findings are observed. In
general, MA boys and girls have higher waist circumferences than AA
or EA children at each age, while AA boys have lower circumferences
than the other ethnic groups. AA boys have a slower rate of increase
in waist circumference as they age than do the other boys. MA girls
have the fastest rate of increase of all girls. At the 75th
percentile, 16 and 17 year old MA and AA girls exceed the waist
circumference cut-off point for obesity related co-morbidities in
adult women. See Table for the abdominal circumference data of the entire population studied.
authors carefully point out that these presentations are not
standards of optimal circumference, but rather descriptions of actual
findings among the United States childhood population. How the data
relate to obesity co-morbidities or whether they have any predictive
validity cannot be determined from these variables.
JR, Redden DT, Pietrobelli A, Allison DB. Waist Circumference
Percentiles in Nationally Representative Samples of African-American,
European-American, and Mexican-American Children and Adolescents J Pediatr 2004;145:439-44.
The mean values differ among specific populations as compared with the combined population. An example is at the 90th percentile.
Data are differences in cm as compared with the mean combined population in the above table. For complete date of each population across all percentiles the reader is referred to the original article and its additional tables by Fernández et al. J Pediatr 2004;145:439-44.
Comment: This manuscript should be an important reference
in every pediatrician’s office. Although, as the authors state,
the data merely describe observed waist circumference percentiles,
they do not establish safe or healthy norms. However, it would
certainly be accurate to consider those children whose waist
circumferences fall above the 90th percentile for age,
gender, and ethnic background at significant risk for obesity related
co-morbidities. Some may ask the rather obvious question, “Why
measure and record this variable, when one’s eyes can surely
tell that the child is obese?” Clearly, the morbidly obese
child is at a high risk regardless of waist circumference. What are
not known however, are the cut-off points of central adiposity for
serious co-morbidities at different ages for different ethnic
backgrounds and gender. Such information could be exceedingly
important and only by careful recording and plotting of variables
such as waist circumference will we be able to establish these
cut-offs. Although the information provided by Fernández and
colleagues may initially seem useful only to clinical researchers, it
will probably be important to those who care for children.
L. Clarke, MD