Error bars indicate 95% confidence intervals. Participants were aged 2 through 19 years.
Ogden CL, Carroll MD, Flegal KM. High Body Mass Index for Age Among US Children and Adolescents, 2003-2006. JAMA. 2008;299(20):2401-2405. doi:10.1001/jama.299.20.2401
Author Affiliations: National Center for Health Statistics, Centers for Disease Control and Prevention, Hyattsville, Maryland.
Context The prevalence of overweight among US children and adolescents increased between 1980 and 2004.
Objectives To estimate the prevalence of 3 measures of high body mass index (BMI) for age (calculated as weight in kilograms divided by height in meters squared) and to examine recent trends for US children and adolescents using national data with measured heights and weights.
Design, Setting, and Participants Height and weight measurements were obtained from 8165 children and adolescents as part of the 2003-2004 and 2005-2006 National Health and Nutrition Examination Survey (NHANES), nationally representative surveys of the US civilian, noninstitutionalized population.
Main Outcome Measures Prevalence of BMI for age at or above the 97th percentile, at or above the 95th percentile, and at or above the 85th percentile of the 2000 sex-specific Centers for Disease Control and Prevention (CDC) BMI-for-age growth charts among US children by age, sex, and racial/ethnic group.
Results Because no statistically significant differences in the prevalence of high BMI for age were found between estimates for 2003-2004 and 2005-2006, data for the 4 years were combined to provide more stable estimates for the most recent time period. Overall, in 2003-2006, 11.3% (95% confidence interval [CI], 9.7%-12.9%) of children and adolescents aged 2 through 19 years were at or above the 97th percentile of the 2000 BMI-for-age growth charts, 16.3% (95% CI, 14.5%-18.1%) were at or above the 95th percentile, and 31.9% (95% CI, 29.4%-34.4%) were at or above the 85th percentile. Prevalence estimates varied by age and by racial/ethnic group. Analyses of the trends in high BMI for age showed no statistically significant trend over the 4 time periods (1999-2000, 2001-2002, 2003-2004, and 2005-2006) for either boys or girls (P values between .07 and .41).
Conclusion The prevalence of high BMI for age among children and adolescents showed no significant changes between 2003-2004 and 2005-2006 and no significant trends between 1999 and 2006.
In the United States, the prevalence of overweight among children increased between 1980 and 2004, and the heaviest children have been getting heavier.1- 4 We updated the most recent national estimates of the prevalence of high body mass index (BMI) among US children and adolescents using data from 2005-2006. Estimates of high BMI at 3 different levels are presented. In addition, trends were examined between 1999 and 2006 and racial/ethnic differences were evaluated.
Data were obtained from the National Health and Nutrition Examination Survey (NHANES), a complex, multistage probability sample of the US civilian, noninstitutionalized population conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention (CDC).5 The survey was reviewed and approved by the NCHS ethics review board, and participants provided written informed consent (parental consent for children). Height and weight were measured with standardized protocols and calibrated equipment during a physical examination in a mobile examination center. Body mass index was calculated as weight in kilograms divided by height in meters squared and rounded to 1 decimal place. Survey participants aged 16 years and older reported their race and ethnicity. A family member reported race and ethnicity for children younger than 16 years. Race and ethnicity were reported by participants based on a list that included an open-ended response. For the purposes of this analysis, racial/ethnic groups were categorized as non-Hispanic white, non-Hispanic black, and Mexican American. Sample sizes for other racial/ethnic groups were too small for meaningful analysis.
The overall unweighted examination response rate for children and adolescents in NHANES 2005-2006 was 84.3%, calculated as the number of examined children and adolescents divided by the total number selected into the sample. Response rates for 1999-2004 were similar and have been previously published.1- 4 Less than 1% of examined children had missing data for BMI and were excluded from the analysis. This report includes data for 4207 children and adolescents (aged 2-19 years) from 2005-2006 and 3958 children and adolescents from 2003-2004.
High BMI was defined based on the 2000 sex-specific BMI-for-age growth charts from the CDC.6 Prevalence estimates are shown based on 3 high BMI-for-age cut points: at or above the 97th percentile, at or above the 95th percentile, and at or above the 85th percentile.7- 9 Throughout the article, the percentiles refer to the reference population of the 2000 CDC growth charts.
Statistical analyses were done using SAS (version 9; SAS Institute Inc, Cary, North Carolina) and SUDAAN (version 9; Research Triangle Institute, Research Triangle Park, North Carolina). All analyses excluded pregnant adolescents. Sample weights were used to account for differential nonresponse and noncoverage and to adjust for planned oversampling of some groups. Standard errors were estimated with SUDAAN using Taylor series linearization.
Trends in high BMI for age over 4 time periods (1999-2000, 2001-2002, 2003-2004, and 2005-2006) were modeled using logistic regression. Time period was included as an ordinal variable in the models while controlling for age group and racial/ethnic group. Because significant interactions were found between sex and racial/ethnic group, models were run separately for boys and girls. Trends are displayed graphically, and P values based on the Satterthwaite adjusted F statistic are presented in the text.
Differences between estimates from 2003-2004 and 2005-2006 and by sex in 2003-2006 were tested using t tests at the .05 significance level controlling for multiple comparisons using the Bonferroni method. Differences by age and race/ethnicity were tested in logistic regression models using data from 2003-2006.
In 2005-2006, 10.9% (95% confidence interval [CI], 8.6%-13.2%) of children and adolescents aged 2 through 19 years were at or above the 97th percentile of the 2000 BMI-for-age growth charts; 15.5% (95% CI, 12.7%-18.3%) were at or above the 95th percentile; and 30.1% (95% CI, 26.7%-33.5%) were at or above the 85th percentile of BMI for age.
Analyses of the trends in high BMI for age showed no statistically significant trend over the 4 time periods (1999-2000, 2001-2002, 2003-2004, and 2005-2006) for either boys or girls (P values between .07 and .41 for the 6 sex- and BMI-level–specific models). Figures 1, 2, and 3 show the estimates of high BMI for age (using the 3 cut points) over the 4 time periods by race/ethnicity. Results were similar for non-Hispanic white, non-Hispanic black, and Mexican American boys and girls. Trends were not statistically significant for any racial/ethnic group (P = .05).
No statistically significant change in high BMI for age was found between 2003-2004 and 2005-2006. Although the estimate of BMI for age at or above the 95th percentile of the CDC growth charts decreased from 17.1% to 15.5% for children and adolescents aged 2 through 19 years, the change was not statistically significant. None of the estimates for non-Hispanic white children of BMI for age at or above the 95th percentile were higher in 2005-2006 than in 2003-2004. One estimate was the same while all the others were lower in 2005-2006 compared with 2003-2004, but the differences were not statistically significant. Among non-Hispanic black or Mexican American children, we did not see this pattern.
Because no significant differences were found between 2003-2004 and 2005-2006, these 2 two-year survey periods were combined to make detailed population estimates for the prevalence of high BMI for age by sex, age, and race/ethnicity. For 2003-2006, 11.3% (95% CI, 9.7%-12.9%) of children and adolescents were at or above the 97th percentile of BMI for age from the 2000 CDC growth charts. For the same period, 16.3% (95% CI, 14.5%-18.1%) of children and adolescents had a BMI for age at or above the 95th percentile of BMI for age, and 31.9% (95% CI, 29.4%-34.4%) were at or above the 85th percentile. Using 4 years of data, these estimates were based on a larger sample size and were thus more stable than those from only 2 years of data (Table 1).
For each of the 3 cutoffs, high BMI for age differed significantly by age and racial/ethnic group but not by sex. Results from the logistic regression models indicated that 2- through 5-year-old boys and girls were significantly less likely to have a high BMI for age than 12- through 19-year-old adolescents (Table 2). Non-Hispanic black and Mexican American girls were more likely to have a high BMI for age than non-Hispanic white girls. As seen in Table 1, almost 28% (95% CI, 23.8%-31.6%) of non-Hispanic black teen girls aged 12 through 19 years and almost 20% (95% CI, 17.0%-22.8%) of Mexican American teen girls were at or above the 95th percentile of the 2000 BMI-for-age growth charts compared with 14.5% (95% CI, 10.4%-18.6%) of non-Hispanic white girls. A similar pattern is seen among the adolescents at or above the 97th percentile; 19.6% (95% CI, 16.5%-22.7%) of non-Hispanic black girls and 14.1% (95% CI, 11.4%-16.8%) of Mexican American girls were at or above the 97th percentile while 9.1% (95% CI, 5.8%-12.4%) of non-Hispanic white girls were at or above the 97th percentile. Among boys, Mexican Americans were significantly more likely to have high BMI for age than non-Hispanic white boys. Non-Hispanic black boys, however, were only more likely than non-Hispanic white boys to have high BMI for age at the highest BMI-for-age level (BMI for age ≥ 97th percentile).
No statistically significant trend in high BMI for age was found over the time periods 1999-2000, 2001-2002, 2003-2004, and 2005-2006. Because there were no significant differences in prevalence of high BMI for age between 2003-2004 and 2005-2006, data for those 4 years were combined to provide more stable estimates than when using only 2 years of data. For 2003-2006, 11.3% of children and adolescents were at or above the 97th percentile of BMI for age from the 2000 CDC growth charts. For the same period, 16.3% of children and adolescents had a BMI for age at or above the 95th percentile, and 31.9% were at or above the 85th percentile.
As a sample of the US population, estimates from NHANES have some variability. The trend analysis presented here includes the most recent data. Analysis of trends using 4 data points provides more precise estimates of trends and sampling error than analysis with 3 data points. Previous analyses of trends with only 3 data points through 2003-2004 showed a significant increase in BMI for age at or above the 95th percentile over the period 1999-2004 (P = .03 for boys and P = .04 for girls).2 However, the same test for trends with additional data from 2005-2006 did not show a significant trend over the time period 1999-2006 (P = .19 for boys and P=.41 for girls) in BMI for age at or above the 95th percentile.
Similar to previous analyses of NHANES data, these results show some significant differences in high BMI for age by race/ethnicity, particularly for girls.1- 4 Adiposity levels at different BMI-for-age levels, however, may vary by race/ethnicity.10
In comparison with a general reference population, the 85th and 95th percentiles have been used as cut points for high BMI.7,9 The 97th percentile provides an even higher cut point to identify the heaviest children. Strictly speaking, obesity refers to excess body fatness or adiposity and overweight to total body weight in excess of a weight standard.11 In practice, measurement of body fat is difficult both in clinical applications and in population studies. In addition, there are no well-accepted standards for body fatness for children. Thus, in general, weight adjusted for height is used rather than a more direct measure of body fat. A variety of methods have been used to adjust weight for height, but currently the most common, both for children and for adults, is BMI.12 For children, BMI varies considerably with age, so generally the BMI of a child is compared with the BMI of a reference population of children of the same sex and age. A variety of reference data sets exist for BMI in childhood. There is no standard definition of adiposity or excess body fat in childhood against which given BMI levels can be compared.
The 2000 CDC BMI-for-age growth charts are often used in the United States to screen for overweight. These charts were constructed from reference population samples of US children in the 1960s, 1970s, and 1980s (and early 1990s for children younger than 6 years).6 In the construction of the 2000 CDC growth charts, selected percentiles were smoothed with a variety of methods, but data were too sparse to estimate percentiles below the 3rd or above the 97th percentile. Using a different approach from that developed by Cole,13 estimates of 3 parameters, L (a skewness parameter), M (the median), and S (a measure of variation), were calculated by solving equations that used the values for the smoothed percentiles ranging from the 3rd to the 97th. Because percentiles less than the 3rd or greater than the 97th are beyond the range of the data from which the LMS values were estimated, “extrapolation beyond this range should be done with caution.”6 Consequently, we present estimates of high BMI for age based on the 97th percentile.
The increase in the prevalence of high BMI for age among US children that was previously seen between NHANES III (1988-1994) and NHANES 2003-2004 was not observed between 2003-2004 and 2005-2006. Even at the highest BMI for age level, no change in prevalence was found between 2003-2004 and 2005-2006, either overall or by racial/ethnic group. Data from 2007-2008 are needed to further examine the trends.
Corresponding Author: Cynthia L. Ogden, PhD, National Center for Health Statistics, Centers for Disease Control and Prevention, 3311 Toledo Rd, Room 4414, Hyattsville, MD 20782 (firstname.lastname@example.org).
Author Contributions: Dr Ogden had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Ogden, Flegal.
Analysis and interpretation of data: Ogden, Carroll, Flegal.
Drafting of the manuscript: Ogden, Carroll, Flegal.
Critical revision of the manuscript for important intellectual content: Ogden, Carroll, Flegal.
Statistical analysis: Ogden, Carroll, Flegal.
Financial Disclosures: None reported.
Role of the Sponsor: All data used in this study were collected by the National Center for Health Statistics, Centers for Disease Control and Prevention. The Centers for Disease Control and Prevention reviewed and approved this report before submission.
Disclaimer: The findings and conclusions in this report are those of the authors and not necessarily of the agency.
Additional Contributions: Lester R. Curtin, PhD, National Center for Health Statistics, provided statistical support, and Yang Yu, MS, Harris Corporation, Hyattsville, Maryland, provided computer programming support. Neither received compensation.