Published online October 1, 2007
PEDIATRICS Vol. 120 No. 4 October 2007, pp. e1043-e1050 (doi:10.1542/peds.2007-0089)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tudor-Locke, C.
Right arrow Articles by Kuby, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tudor-Locke, C.
Right arrow Articles by Kuby, M.
Related Collections
Right arrow Nutrition & Metabolism

ARTICLE

A Geographical Comparison of Prevalence of Overweight School-aged Children: The National Survey of Children's Health 2003

Catrine Tudor-Locke, PhDa, Jennie J. Kronenfeld, PhDb, Sam S. Kim, MAb, Mary Benin, PhDb and Michael Kuby, PhDc

a Walking Research Laboratory, Department of Exercise and Wellness
b School of Social and Family Dynamics
c School of Geographical Sciences, Arizona State University, Mesa, Arizona


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 CONCLUSIONS
 REFERENCES
 
OBJECTIVES. This study presents a geographical comparison of state-specific prevalence estimates of children who are at risk of overweight and/or overweight using the 2003 National Survey of Children's Health.

METHODS. Using the 2003 National Survey of Children's Health, we computed prevalence estimates of children who are at risk of overweight and/or overweight among a nationally representative sample of 69000 children between 5 and 17 years old.

RESULTS. Overall, 36.4% of the children (39.8% of the boys and 32% of the girls) in the sample were in the combined category of at risk of overweight or overweight, representing an estimated 17 million US children. We found geographic variation at the state and the regional levels. The southeastern states, especially those west of the Appalachians and in the lower Mississippi region, had the highest prevalence of children who are at risk of overweight and/or overweight. The central Rocky Mountain states of Colorado, Utah, and Wyoming had the lowest prevalence, followed by the northwestern quadrant of the lower 48 states and New England.

CONCLUSIONS. These National Survey of Children's Health data provide clinicians and public health professionals with useful data required for policy and planning related to childhood obesity at state levels. These data also serve as important baseline indicators and can be used to track changes over time.

Key Words: child • geography • growth • body composition • obesity

Abbreviations: NHANES—National Health and Nutrition Examination Survey • CDC—Centers for Disease Control and Prevention • NSCH—National Survey of Children's Health

Direct height and weight measures from the National Health and Nutrition Examination Survey (NHANES) indicate that the prevalence of overweight (defined as ≥95th percentile of age- and gender-specific BMI derived from the 2000 Centers for Disease Control and Prevention [CDC] growth charts1) of 6- to 19-year-old American youth has tripled, from ~5% in 1988–1994 to ~16% in 1999–2002.2,3 Although NHANES was used to monitor national youth prevalence of overweight and at risk for overweight (defined as ≥85th but <95th percentile of age- and gender-specific BMI1), it was not designed to provide prevalence estimates at the state level. In adults, state level data of overweight (BMI: 25–29.9 kg/m2) and obesity (BMI ≥30 kg/m2) patterns were determined by using self-reported data from the Behavioral Risk Factor Surveillance System; the infamous associated maps vividly depict geographic variation in the prevalence of obesity in adults.4 Although nationally representative self-reported data are also available for adolescents through the Youth Risk Behavior Surveillance System,5 these are not available at the state level. Sherry et al6 previously reported prevalence of overweight for 30 states in 2- to 4-year-old children from low-income families determined through the CDC's Pediatric Nutrition Surveillance System; no geographic predominance was noted. The National Survey of Children's Health (NSCH) was designed to address the need for additional state-level data on a number of variables relevant to the health of children and adolescents,7 including parent-reported height and weight from which BMI could be determined. The US Department of Health and Human Services previously released a report focused on national and state estimates of overweight among 10- to 17-year-olds from this survey,8 with no direct geographic comparison. The purpose of this analysis is to extend these early estimates to a fuller range of school-aged children (ie, 5–17 years of age) and present a geographic comparison of US children who are at risk for overweight, overweight, and a third combined category (ie, either at risk for overweight or overweight, defined as age- and gender-specific BMI ≥85th percentile, derived from the 2000 CDC growth charts).


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 CONCLUSIONS
 REFERENCES
 
Data
Data for this study were obtained from the 2003 NSCH, which was sponsored and funded by the Maternal and Child Health Bureau of the Health Resources and Services Administration.9 The purpose of the NSCH was to produce national and state prevalence estimates for a variety of health indicators and measures of children's experiences with the health care system. In particular, the NSCH examined the physical and emotional health of children ages 0 to 17 years of age, with special emphasis on a variety of physical, emotional, and behavioral factors that may relate to the well-being of children (ie, family interactions, parental health, school and after-school experiences, and respondents' perceptions of the neighborhoods).7 The NSCH used the State and Local Area Integrated Telephone Survey (SLAITS) program. The SLAITS program, jointly conducted by the CDC and NCHS, is an ongoing surveillance system available for tracking and monitoring the health and well-being of children and adults. Applying the same sampling frame as the National Immunization Survey, the NSCH used a multistage cluster design based on a random-digit-dialed sample of households with children <18 years of age selected from each of the 50 states and the District of Columbia. Thus, the data allow computation of reliable state and national estimates. The NSCH was conducted from January 2003 to July 2004, and a total of 102353 interviews were completed. The public-use data and documentation for the NSCH can be accessed at www.cdc.gov/nchs/about/major/slaits/nsch.htm.

The 2003 NSCH included questions on the respondent child's gender, height, weight, and age. To protect the confidentiality of individual children, extreme heights and weights were suppressed.9 Because this process might hinder calculation of BMIs, the NSCH also provided the age- and gender-specific weight status designation (ie, underweight, normal weight, at risk of overweight, and overweight) derived from original information on respondent children's age, weight, and height and based on the CDC's BMI-for-age gender-specific growth charts (www.cdc.gov/growthcharts). These growth charts have been used by pediatricians, nurses, and parents to track the growth of American infants, children, and adolescents in the United States since 1977. Herein, NSCH children aged 2 to 17 years were classified as either underweight (ie, gender-specific BMI for age is in the ≤5th percentile), normal weight (BMI for age 5th to <85th percentile), at risk for overweight (ie, gender-specific BMI for age is in the ≥85th percentile but <95th percentile), or overweight (ie, gender-specific BMI for age is in the ≥95th percentile).9 It should be noted that the 2003 NSCH parent-reported heights and weights were compared by survey administrators to independent direct measures taken in the National Health and Nutrition Examination Survey.10 Their analysis indicated that height was generally underreported and weight was generally overreported for children under 10 years of age. Taken together, these biases would lead to an overestimate of BMI classification, at least in children <10 years of age.

Given the caveat, we estimated national and state prevalences of the full range of school-aged children (ie, 5- to 17-year-olds, subgrouped into 5- to <10-year-olds and ≥10- to 17-year-olds) at risk of overweight, overweight, and the combined category based on 74264 children with complete BMI data. Although there were 102353 children aged 0 to 17 in the 2003 NCHS, BMI is not available for those under 2 years and 16.3% or 2602 of 2- to 4-year-olds contained missing BMI data. For children aged 5 to 17 years old, 7% of BMIs were missing, primarily for younger children. Overall, we lost an additional 3229 children aged 5 to <10 years and 2305 children ages ≥10 to 17 years because of missing BMI data, leaving a sample size of 69000 for analysis herein. Children aged 5 to 17 missing BMI data are more likely to be Hispanic (26.4% lost) than non-Hispanic white (3.9% lost), non-Hispanic black (6.9% lost), or non-Hispanic other (5.42% lost). Because Hispanic children are more likely to be at risk for overweight or overweight than non-Hispanic white children (see Table 1), the fact that we are disproportionately missing BMI class data on Hispanic children suggests that our state estimations (Table 2) of at risk for overweight and overweight are underestimated for states with large Hispanic populations (ie, New Mexico, California, and Texas). Missing BMI data varied from a low of 3% in South Dakota to a high of 20% from California, most likely because of the ethnic balances of those states. There was no difference in missing BMI data by the gender of the child.


View this table:
[in this window]
[in a new window]

  		TABLE 1 Children Who Are at Risk of Overweight, Overweight, and Combined in the United States According to Selected Characteristics: NSCH 2003

 

View this table:
[in this window]
[in a new window]

  		TABLE 2 NSCH: Prevalence Estimates of Children Who Are at Risk of Overweight, Overweight, and Combined in the United States

 
Analytical Methods
For each state and for the entire United States, prevalence estimates were calculated for at risk for overweight, overweight, and the combined category. Stata (StataCorp, College Station, TX) was used to conduct all analyses. Standard errors were adjusted by using Taylor linearization to take into account weighting, clustering, stratification, and increased variability, a result of NSCH's sampling design complexity. Specifically, because of the clustered sampling design feature, all analyses were conducted using Stata's survey procedures, and the subpopulation option allowed all cases with complete data to be used to estimate standard errors, although we focused only on children aged 5 to 17 years. This is the procedure recommended by NCHS for dealing with subgroup analyses with clustered sampling designs.9 Tests for significance for proportions of boys and girls in each state who were at risk for overweight or overweight were conducted using the adjusted Wald test which corrects for the clustered sampling design.

ArcGIS (Environmental Systems Research Institute, 2005) was used to render 2 sets of maps displaying the percentages of children by state who were (1) overweight and (2) the combined category of at risk of overweight or overweight. Each set consists of 3 maps, 1 of each representing boys, girls, and the total population of children. Each map in each set of 3 uses the same class break points for easy comparison among the boys, girls, and total. Furthermore, break points are set at round numbers (multiples of 5) for simplicity as well as for comparison to the well-known maps of adult obesity,4,11 which used 4 classes with break points at 10%, 15%, and 20%. Whereas other characteristics of children such as race and ethnicity are likely influential, we did not tabulate or map state prevalences by race or ethnicity because some states have too few relevant cases, which is a potential source of bias. Thus, we report overweight status by race and ethnicity for the overall US population only.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
CONCLUSIONS
 REFERENCES
 
Table 1 presents prevalence estimates for US children who are at risk of overweight, overweight, and combined by gender, age grouping, and race/ethnicity. Boys were significantly more likely to be overweight than girls; there was no statistically significant gender difference in at risk of overweight. Boys were significantly more likely to be in the combined category of at risk for overweight or overweight. In terms of age, there was no significant difference between the 2 age groups (ie, 5- to <10-year-olds and ≥10- to 17-year-olds) with respect to being at risk of overweight, but the younger group was significantly more likely to be overweight or in the combined category than the older age group. With regards to race/ethnicity, black children were most likely, Hispanic children the next most likely, others the third most likely, and white children the least likely to be overweight. All these differences were statistically significant. The racial/ethnic groups did not differ significantly in terms of being at risk for overweight. The combined category showed the same significant differences as the overweight sample: black and Hispanic children were much more likely and others slightly more likely than white children to be in the combined category.

National and state prevalence of children aged 5 to 17 years at risk of overweight, overweight, and the combined category of either at risk of overweight or overweight are presented by gender in Table 2. Overall, these data indicate that ~15.5% of American children are at risk of overweight. Delaware had the highest prevalence of children who are at risk of overweight (18.9%) followed by Louisiana (17.7%) and Alabama (17.6%). Oregon had the lowest prevalence (13.1%).

Twenty percent of school-aged children in the United States are overweight. The District of Columbia had the highest prevalence of overweight children (29.8%) followed by Mississippi (27.7%) and Louisiana (26.8%). Colorado had the lowest prevalence of overweight children (13%). Overall, 36.4% of the children in the sample were either at risk of overweight or overweight. Based on the 2000 Census, this represents an estimated 17 million school-aged American children. The District of Columbia had the highest prevalence of children either at risk of overweight or overweight (46.1%), followed by Mississippi (44.9%) and Louisiana (44.5%). Utah had the lowest prevalence of either at risk of overweight or overweight (27.2%).

Nationally, 39.8% of boys and 32% of girls were either at risk of overweight and/or overweight (ie, they were classified in the combined category). Among states, West Virginia had the highest prevalence of boys in the combined category (48%), followed by Louisiana (47.9%) and Mississippi (47.7%). Utah had the lowest prevalence of boys in the combined category (30.8%). For girls, the District of Columbia had the highest (47.2%) followed by Mississippi (42%) and Louisiana (40.9%). Colorado had the lowest prevalence of girls in the combined category (22.4%). Boys were significantly more likely than girls to be overweight in 35 of the states, and in no states were girls significantly more likely to be overweight than boys. However, there were far fewer differences in being at risk for overweight. Only 6 states showed a gender difference in being at risk for overweight and in each of these cases, boys were at greater risk.

Figure 1 maps the percentage of school-aged children who were overweight in 2003 for (1) boys, (2) girls, and (3) the total sample. Figure 2 displays the same for the combined category of children who are at risk of overweight and/or overweight. For the total school-aged sample, both maps show similar regional patterns. Specifically, Figs 1C and 2C both show high percentages in the southeast, especially along a northeast–southwest axis from West Virginia to Louisiana on the western side of the Appalachian Mountains and in the lower Mississippi region. In both total maps, the next highest category includes the rest of the southeast, south central states from New Mexico to Arkansas, lower Midwestern states (Illinois to Ohio), and most of the Midatlantic states. Also in both total maps, the central Rocky Mountain states of Colorado, Utah, and Wyoming have the lowest percentages, with New England, the upper Midwest, the Great Plains, and the Northwest in the second-lowest category.


Figure 1
View larger version (38K):
[in this window]
[in a new window]

  		FIGURE 1 Prevalence of children 5 to 17 years old who are overweight: 2003 NSCH. A, Boys; B, girls; C, total.

 

Figure 2
View larger version (40K):
[in this window]
[in a new window]

  		FIGURE 2 Prevalence of children 5 to 17 years old who are at risk of overweight or overweight: 2003 NSCH. A, Boys; B, girls; C, total.

 
The maps comparing genders clearly show the higher prevalence of boys who are overweight or in the combined category compared with girls. The boys' map of overweight (Fig 1A) has no states in the lowest category <15%, whereas the girls' map (Fig 1B) has no states in the highest category >25%. Likewise, for the combined category, the boys' map (Fig 2A) has no states in the lowest category under 30%, whereas the girls' map (Fig 2B) has but 2 states (Louisiana and Mississippi) in the highest category over 40%. The geographic patterns depicted by the maps of boys and girls are more fragmented than on the total maps, but certain patterns continue to stand out on all 4 maps. Specifically, for both boys and girls, there are consistently higher prevalences in the southeast and Illinois, and lower rates in Colorado, Utah, and Minnesota.


   CONCLUSIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 CONCLUSIONS
REFERENCES
 
A previous study using the 1999–2002 NHANES data reported that 31% of US children were either at risk for overweight or overweight and 16% were overweight.2 In contrast, the 2003 NSCH data analyzed herein showed higher estimates; 36.4% were either at risk for overweight or overweight and 21% were specifically overweight. Differences in the magnitudes of these study findings may be explained by methodologic differences in the way data were collected. In the NHANES data, BMI was derived from directly measured height and weight. In the NSCH, information was gathered through telephone interviews with children's parents or guardians. There are some problems of misreporting and of inaccuracy in parental reports of children's weight that have been acknowledged previously.12–14 For example, in a study of Nutrition Program for Women, Infants, and Children participants, inaccuracy and misperception of the child's weight was more common in mothers with less education and overweight mothers.12 In a more recent study, 35% of mothers surveyed underestimated, and 5% overestimated, their child's weight.14

Despite the shortcomings inherent with such proxy-report methods, the NSCH is a large and nationally representative survey designed to provide state-level data. It is currently the best available resource that permits computation of prevalence estimates of children who are at risk of overweight and overweight at both national and state levels. In addition, the subset of NSCH data that we used in this study represents a much larger sample population (ie, 69000 children) than that previously reported (4018 children in the NHANES study1). Moreover, the potential biases discussed above suggest that the prevalence estimates of children who are at risk of overweight or overweight are underestimated, not overestimated.

The most recently published map of adult obesity by Mokdad et al4 focused on 2001 data. Acknowledging the differences in how obesity is defined in adults and overweight is defined in children, the national average for adult obesity in 2001 was 20.9% vs 21.0% for children's overweight in 2003. Despite the apparent similarities in these national averages, at least 1 notable difference is that the variability of state averages is greater for the children's data. Specifically, there are 6 states (District of Columbia, West Virginia, Kentucky, Tennessee, Mississippi, and Louisiana) above the 25% break point for overweight in the children's data, compared with only 1 state (Mississippi) above the 25% break point for obese in the adult data.4 Likewise, there are 3 states (Colorado, Utah, and Wyoming) averaging below the 15% break point for overweight in the children's data versus only 1 (Colorado) at the same break point for obesity in adults. Finally, comparing the adult obesity maps4 with our maps of children's overweight reveals consistently high prevalences in the southeast and parts of the Midwest and low prevalences in New England and parts of the Rocky Mountains for both adults and children.

Healthy People 2010 includes objective 19–3: "reduce the proportion of children who are overweight or obese."15 Their term "obese" was used in this early document, although now the terms "at risk of overweight" and "overweight" are considered more appropriate when discussing weight status in children.1 The Healthy People 2010 baseline data of 11% for the objective was based on data from 1988–1994. The "better than best" 2010 target is for 5% of US children to be classified as overweight or obese. Unfortunately, it is apparent from iterative collections of the NHANES data that the prevalence estimates of US children either at risk of overweight and overweight are not decreasing,2,3 and we are moving farther away from baseline measures instead of toward national targets.

In this study we did not aim to point out the candidate factors amenable to intervention. In 2005, the Institute of Medicine published an extensive exploration of the social, environmental, and behavioral factors responsible for the current childhood epidemic nature.16 Their ultimate product was a blueprint for responsive action by a range of stakeholders including the government, industry, media, communities, schools, and families. The prevalence estimates presented herein are important, however, to determine likelihood of at risk of overweight and overweight in children and to estimate associated social and economic burdens for planning and policy purposes. In concert with other national data, these data continue to indicate that the health of our nation's children is in peril and increasing burden can be anticipated.

In summary, these NSCH data provide slightly elevated estimates of US children who are at risk of overweight and overweight compared with the NHANES data. These data and maps do provide clinicians and public health professionals with useful data required for policy and planning related to childhood obesity at population levels in their home state. The states with the highest percentage of school-aged children who are at risk for overweight or overweight were found in the southeastern United States and neighboring states, especially for boys. These data also serve as important baseline indicators of childhood obesity and can be used to track changes over time.


   ACKNOWLEDGMENTS
 
We acknowledge the contributions of Barbara Trapido-Lurie, cartographer for the Arizona State University School of Geographical Sciences.


   FOOTNOTES
 
Accepted Mar 14, 2007.

Address correspondence to Catrine Tudor-Locke, PhD, Walking Research Laboratory, Department of Exercise and Wellness, Arizona State University, 7350 E Unity Ave, Mesa, AZ 85212. E-mail: tudor-locke{at}asu.edu

The authors have indicated they have no financial relationships relevant to this article to disclose.


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 CONCLUSIONS
 REFERENCES
 

  1. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat 11. 2002;(246) :1 –190
  2. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA. 2004;291 :2847 –2850[Abstract/Free Full Text]
  3. Ogden CL, Flegal KM, Carroll MD, Johnson CL. Prevalence and trends in overweight among US children and adolescents, 1999–2000. JAMA. 2002;288 :1728 –1732[Abstract/Free Full Text]
  4. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA. 2003;289 :76 –79[Abstract/Free Full Text]
  5. Lowry R, Galuska DA, Fulton JE, Burgeson CR, Kann L. Weight management goals and use of exercise for weight control among U.S. high school students, 1991–2001. J Adolesc Health. 2005;36 :320 –326[CrossRef][ISI][Medline]
  6. Sherry B, Mei Z, Scanlon KS, Mokdad AH, Grummer-Strawn LM. Trends in state-specific prevalence of overweight and underweight in 2- through 4-year-old children from low-income families from 1989 through 2000. Arch Pediatr Adolesc Med. 2004;158 :1116 –11124[Abstract/Free Full Text]
  7. van Dyck P, Kogan MD, Heppel D, Blumberg SJ, Cynamon ML, Newacheck PW. The National Survey of Children's Health: a new data resource. Matern Child Health J. 2004;8 :183 –188[CrossRef][ISI][Medline]
  8. US Department of Health and Human Services, Health Resources and Services Administration, Maternal and Child Health Bureau. The National Survey of Children's Health 2003. Rockville, MD: US Department of Health and Human Services; 2005
  9. Blumberg SJ, Olson L, Frankel MR, Osborn L, Srinath KP, Giambo P. Design and operation of the National Survey of Children's Health, 2003. Vital Health Stat 1. 2005;(43) :1 –131
  10. Child and Adolescent Health Measurement Initiative. National Survey of Children's Health. Available at: www.nschdata.org/Content/Default.aspx. Accessed August 2, 2007
  11. Mokdad AH, Bowman BA, Ford ES, Vinicor F, Marks JS, Koplan JP. The continuing epidemics of obesity and diabetes in the United States. JAMA. 2001;286 :1195 –1200[Abstract/Free Full Text]
  12. Baughcum AE, Chamberlin LA, Deeks CM, Powers SW, Whitaker RC. Maternal perceptions of overweight preschool children. Pediatrics. 2000;106 :1380 –1386[Abstract/Free Full Text]
  13. Maynard LM, Galuska DA, Blanck HM, Serdula MK. Maternal perceptions of weight status of children. Pediatrics. 2003;111 :1226 –1231[Abstract/Free Full Text]
  14. Boutelle K, Fulkerson JA, Neumark-Sztainer D, Story M. Mothers' perceptions of their adolescents' weight status: are they accurate? Obesity Res. 2004;12 :1754 –1757[ISI][Medline]
  15. US Department of Health and Human Services. Healthy People 2010: Conference Edition. Washington, DC: US Department of Health and Human Services; 2000
  16. Koplan JP, Liverman CT, Kraak VA, eds. Preventing Childhood Obesity: Health in the Balance. Washington, DC: Institute of Medicine, National Academies Press; 2005
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tudor-Locke, C.
Right arrow Articles by Kuby, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tudor-Locke, C.
Right arrow Articles by Kuby, M.
Related Collections
Right arrow Nutrition & Metabolism