Cautious optimism on leveling of high BMIs.

The apparent leveling off in the prevalence of being overweight among U.S. children and adolescents is cause for celebration, but it might not necessarily be a trend, experts say.

A study published in JAMA by the National Center for Health Statistics (NCHS) found no significant trend in the incidence of high body mass index (BMI) among children and adolescents between 1999 and 2006, suggesting that rates of pediatric obesity might be stabilizing after tripling in the 1980s and 1990s.

"It's difficult to know whether or not we are seeing a true halt to the rise in the prevalence of childhood obesity and overweight based on just a few years of data. I think that we were pleasantly surprised by the data, but we can only speculate as to the reasons behind [them]," Dr. Gilbert P. August, chair of the Endocrine Society panel that developed guidelines for the prevention and treatment of pediatric obesity, said in an interview.

The study was conducted by Cynthia L. Ogden, Ph.D., and her associates at the NCHS. Working with data from the National Health and Nutrition Examination Survey, the researchers used logistic regression to model trends in high BMI for age over four time periods: 1999-2000, 2001-2002, 2003-2004, and 2005-2006. No significant trends were identified (JAMA 2008;299:2401-5).

The study also sought to identify recent changes in the prevalence of high BMI for age. To this end, the investigators analyzed height and weight measurements collected from 3,958 children and adolescents (aged 2-19 years) during 2003-2004 and 4,207 children and adolescents during 2005-2006.

The subjects' BMIs were calculated and compared with the CDC's BMI-for-age growth charts with attention to three measures of high BMI: at or above the 97th, 95th, and 85th percentiles. For each of these cutoffs, no statistically significant difference was found between the two 2-year periods.

Dr. Ogden and her associates pooled data for 2003-2004 and 2005-2006 to create population estimates for the prevalence of overweight among children and adolescents between 2003 and 2006. Within that period, 11.3% of children and adolescents had a BMI at or above the 97th percentile of the 2000 CDC growth charts, 16.6% had a BMI at or above the 95th percentile, and 31.9% had a BMI at or above the 85th percentile.

In an accompanying editorial, Cara B. Ehbeling, Ph.D., and Dr. David S. Ludwig of Children's Hospital Boston described the study by Dr. Ogden and her associates as a contrast to "years of unremitting bad news about increasing rates of pediatric obesity."

Dr. Ebbeling and Dr. Ludwig said more data will be needed to determine whether the study's findings indicate a true plateau in the obesity epidemic. But they speculated that the findings might indicate the beneficial effects of public health campaigns aimed at raising obesity awareness and improving the quality of school lunches (JAMA 2008;299:2442-3).

"The optimistic hypothesis is that increased awareness and some programs have made a difference, but we really don't know," Dr. Ogden said in an interview. She emphasized that the prevalence of pediatric obesity remains a critical issue. "We need to remember that it's still too high."

Dr. August stressed the importance of working to decrease the prevalence of high BMI among children and teens even if the study's findings reflect a true leveling off in the obesity epidemic.

"We should not be satisfied with just holding the prevalence rates at this high level," said Dr. August, who is a pediatric endocrinologist and professor emeritus of pediatrics at George Washington University, Washington. "We need to decrease the incidence of childhood overweight and obesity if we are to control the rise in the prevalence of type 2 diabetes mellitus."

Dr. August emphasized that change takes time. "It is questionable Whether these efforts have had time to affect the prevalence of childhood overweight," he said. "We need to remember that educational efforts to control smoking took a generation to show an effect."

The authors of the study reported no disclosures.

BY AUDREY KUBETIN

Editorial Intern