Ji-Hong Liu, Sc.D.,1,2 Sonya J. Jones, Ph.D.,3 Han Sun, MSPH,1 Janice C. Probst, Ph.D.,2,4
Anwar T. Merchant, Sc.D.,1 and Philip Cavicchia, Ph.D.1
Background: The aim of this study was to examine the differences in diet, physical activity, and weight status among children living in rural and urban America and to study the roles of obesity-related behaviors in residence-based differences in childhood obesity.
Methods: We performed cross-sectional analysis of the 1999–2006 National Health and Nutrition Examination Survey data, restricted to 14,332 children aged 2–19 years old (2771 rural, 13,766 urban). Residence was measured at the census tract level using Rural-Urban Commuting Areas. Age-specific questions were used to assess physical activity, and the 24-hour diet recall was used to measure dietary intake.
Results: Among 2- to 11-year-olds, rural children consumed 90 more kcal/day on average than urban children (p < 0.05) and were more likely to consume the recommended two to three cups of dairy per day (p < 0.05). More 2- to 11-year-old rural children also reported participating in exercise five or more times per week than urban children of the same age (79.7% vs. 73.8%). Among 12- to 19-year-olds, rural children were less likely to consume any fruit or meet the recommendation of two cups of fruit. Using measured height and weight data, proportionately more rural than urban children were overweight/obese (35.4% v. 29.3%) and obese (18.6% v. 15.1%). Rural children had 30% higher odds of being overweight and/or obese even after adjustment for socio demographics, health, diet, and exercise behaviors.
Conclusions: The persistently higher odds of overweight and obesity among rural children even after adjustment suggest that rural environments may be “obesogenic” in ways that a person-level analysis cannot discern. Future research should examine disparities in the accessibility and affordability of healthy food and beverage choices and safe physical activity locales in rural areas.
Introduction Nearly 20% of the U.S. population resides in rural areas.1 National evidence suggests that rural adults are more likely to be obese than urban adults (23.0% vs. 20.5%).2,3 Recent studies suggest that obesity is also more prevalent in rural children and adolescents than their urban counterparts.4–6 Davis et al. reported that in the 2003–2006 National Health and Nutrition Examination Survey (NHANES) the prevalence of obesity in 2- to 18-year-olds was about 5% higher in rural compared to urban children (21.8% vs. 16.9%).6 If residing in rural areas is associated with a greater risk of obesity in children, then a better understanding of why children are more likely to become obese in rural areas is warranted. Previous studies have shown that people of low socioeconomic status and those living in the southeastern states are disproportionately affected by the obesity epidemic.7,8 Compared to urban residents, rural residents usually have lower incomes and rural heads of household are less likely to have completed high school or college.4,9 The southern region of the country is disproportionately more rural.4,9 Thus, the higher prevalence of obesity among rural residents might be related to the
1Department of Epidemiology and Biostatistics, Arnold School of Public Health, University of South Carolina, Columbia, SC.
2South Carolina Rural Health Research Center, University of South Carolina, Columbia, SC.
3 Center for Research in Nutrition and Health Disparities and Department of Health Promotion, Education and Behavior, Arnold School of Public Health, University of South Carolina, SC.
4Department of Health Services Policy and Management, Arnold School of Public Health, University of South Carolina, Columbia, SC.
sociodemographic characteristics of rural populations.
However, the limited number of studies in this area have shown that even after adjusting for differences in sociodemographics, the urban and rural differences in overweight status persist.2–6,10 A second commonly hypothesized reason that rural people may have higher obesity rates is that cultural and structural differences in rural and urban areas lead to different diet and activity patterns.6,11 Access to healthy foods12,13 and opportunities to be physically active14,15 may differ between rural and urban areas. Rural residents living in the South frequently consume fried foods, fast foods, sugary and carbonated beverages, and processed, high-fat and high-sodium foods. In addition, their intake of fruits and nonstarchy vegetables is low.16
Disproportionately more rural adults are physically inactive compared to their urban peers (62.8% vs. 59.3%).2 A solid body of research demonstrates that risky behaviors such as sexual activity, weapon carrying, and tobacco, alcohol, and substance use,17 increase during adolescence, and rural adolescents are at greater risk than urban adolescents.
18,19 Consistent with this general trend of increasing health risk behaviors in adolescence, older adolescents are less likely to meet physical activity recommendations than younger adolescents.20
To advance the literature on factors associated with rural and urban difference in childhood obesity, this study had two objectives. First, differences in dietary, physical activity, and sedentary behaviors by age group and residential location were examined. Second, the risk of overweight and obesity among rural and urban adolescents controlling for dietary, physical activity, and sedentary behaviors was estimated. The goals of this study were to understand whether the cultural and structural differences that lead to dietary, physical activity, and sedentary behavior differences between rural and urban children also explain the higher risk of obesity in rural and urban children. Population and Methods
Data came from the 1999–2006 continuous NHANES, an ongoing, nationally representative study conducted by the National Center for Health Statistics (NCHS). NHANES uses a complex, stratified, multistage probability ampling procedure designed to provide prevalence estimates describing the health and nutritional status of the civilian, noninstitutionalized U.S. population. To allow for accurate estimates for subgroups, NHANES oversamples adolescents, blacks, and Hispanics. NHANES reaches about 5000 persons each year in 15 selected counties across the United States, with a new sample of counties implemented each year. Households were selected randomly from within each county for participation. The response rate for the years 1999– 006 for home interviews was 81%, and 95% of respondents interviewed at home had a follow-up examination at mobile examination centers (MEC). The study methods for NHANES are described in detail on the study website (www.cdc.gov/nchs/nhanes.htm/). Data from NHANES are broadly used in standard setting and tracking of Healthy People objectives.21
The overweight and obesity prevalence estimates were based on 16,537 children aged 2–19 years, excluding 673 children who were not interviewed at MEC visits, 177 pregnant children, and 420 children with missing values for weight and height. For analyses addressing obesityrelated behavioral factors, 14,332 children were included, after additional exclusions for children with incomplete 24-hour dietary recalls (1054) and those with missing values for total energy intake (300) and the main covariates (851). Because the percentage of missing values was relatively high for the household poverty variable (6.0%), a dummy variable for “missing” was created to retain these children in the analysis. Excluded children were more likely to fall in the 2- to 11-year-old age group, be female, report excellent health, and have a household income at or below 130% of the poverty level than included children. They were not different in urban and rural residence, race, education, and region. The study was approved by a local institutional review board. Measures
Residence. In the health policy and research arena, multiple definitions have been used to define rural areas in the United States.22 On the basis of the availability of data, this study defined urban and rural residence using the Rural-Urban Commuting Area (RUCA) definition developed by the University of Washington’s Rural Health Research Center and the Economic Research Service at the USDA.22 RUCA categories are based on the size of settlements and towns as delineated by the Census Bureau and the functional relationships between places as measured by tract-level work-commuting data.22 This definition has the strength of identifying the rural portions of metropolitan counties and the urban portions of nonmetropolitan counties; thus, it has been widely used for policy and research purposes.22 Urban areas were defined as tracts with RUCA codes between 1 and 3 and rural areas as tracts with RUCA codes between 4 and 10.23 The protected census tract data in NHANES were accessed through the Research Data Center at NCHS.
Anthropometry measures. At MEC visits, trained examiners took height and weight measurements using digital scales and stadiometers. Because of differences in children’s development by gender and age during childhood, obesity and overweight were defined based on the sex-specific BMI (kg/m2) for age growth charts from the CDC.24 Children whose BMI-for-age and -gender was in the 85th percentile or greater were classified as overweight or obese (including children over the 95th percentile, the indicator for obesity).
Physical activity. In NHANES, age-specific questions were used to examine children's physical activity. For 2- to 11-year-old children, NHANES only asked one question to proxy respondents about the child’s physical activity. Specifically, they were asked the number of times per week the child played or exercised enough to make him/her sweat and breathe hard. Thus, children were considered inactive if they reported this level of activity less than five times per week and active if they reported this level of activity five or more times per week. A similar grouping has been used in prior studies.25 Children aged 12–19 years self-reported their participation in moderate and vigorous leisure time activities that lasted for at least 10 minutes in the past 30 days. Those who reported any moderate or vigorous activities were further asked to report their participation in 24 types of vigorous activities [the metabolic equivalent (MET) value ≥6] and 32 types of moderate activities (MET values between 3 and 6) and provide the corresponding frequencies and duration of each selected activity. Using this information, a measure of meeting the 2008 physical activity recommendations was defined as participation in 60 minutes or more of moderate to vigorous physical activity daily.26
Sedentary behaviors. Sedentary behaviors included the daily number of hours spent sitting and watching TV or videos, using a computer, or playing computer games.
Answer options ranged from less than 1 hour to ≥5 hoursper day, with 1-hour increments. Across 1999–2006, children aged 16 or older were asked about their typical sedentary behaviors over the past 30 days. Children aged 15 or younger were asked different questions across the period. In 1999– 000, these children were asked about computer and television use for the day prior to the interview. In 2001–2006, they were asked about typical daily use over the past 30 days. On the basis of the information provided, the total number of hours per day that the child spent in television and computer use was estimated. Using national guidelines,27,28 excessive total screen time was defined as ≥2 hours of screen time per day. Dietary intake. During the MEC visit, trained diet technicians used a “multiple pass” 24-hour dietary interview format to collect detailed dietary intake information including the types and amounts of all foods and beverages consumed (excluding plain drinking water) in the 24-hour period prior to the visit (1999–2002). If respondents seemed unable to recall what they ate in the previous 24-hour period (or for a particular eating occasion), a prompt was used. Beginning in 2003, NHANES used 2-day 24-hour dietary recalls. To maintain equivalence in the measure across time, only the first-day 24-hour dietary recall data for NHANES 2003–2006 were used. Prior studies have concluded that 24-hour recalls provide adequate validity and reliability for population-level estimations.
29–31 Previous intervention and observational studies have shown that energy-adjusted sweetened beverage intake, whole grain, fruit and vegetables intake, and dairy product intake are the most important dietary factors in the development of obesity for children.32–35 Thus, these dietary variables were included as potential risk factors for childhood obesity. For diet questionnaires, proxy interviews were conducted for survey participants less than 6 years of age, and assisted interviews and self-report were used for the participants 6–11 years of age and 12–19 years of age, respectively. In NHANES, the preferred proxy respondent is the person who is the most knowledgeable about the child (usually the mother or father).
Because obesity prevalence and obesity-related risk factors vary by age and the measurement of physical activity differed for young children (2–11 years old) and adolescents (12–19 years old), the proportions of overweight/obese and obesity for the total sample as well as the subsamples based on age were calculated. Within each age group, the differences in sample characteristics, overweight and/or obesity, and obesity-related behaviors by urban and rural residence were compared using chi-squared tests. Logistic regression was used to examine the relationship between obesity-related risk factors and dichotomous weight status. An incremental modeling approach was used to understand the independent contributions of risk factors to urban/ rural differences in weight status. First, crude models with only residence or other obesity-related behaviors as covariates were run. Second, sociodemographic variables including age (continuous), race, gender (total samples only), perceived health status, the reference person's education, poverty, region, and survey year were added. The reference person was the first household member who was included in the study, was aged 18 years or older, and who owned or rented the dwelling unit. Without any other parental information, this variable was used to estimate education level of the household. Then obesity-related behaviors were sequentially added into the models with sociodemographic factors until all three obesity-related risk factors had been added. Interaction terms between gender, residence, and obesity-related behaviors were examined and none of them were statistically significant at the 0.05 level. Statistical analyses were conducted using SAS-Callable SUDAAN to account for the weighted sampling structure and complex survey design of NHANES. Sample weights accounted for differential nonresponse, noncoverage, and planned oversampling of certain groups. NHANES analytic guidelines were followed to calculate the appropriate 8-year sample weights.36
Sample Characteristics by Urban and Rural Residence
Overall, 22.4% of children aged 2–19 years lived in rural America. A larger proportion of rural children were white (76.1%) compared to urban children (55.3%). Rural children were more likely to live in households under 130% of Federal Poverty Level and in households where the adults' education level was equal to or less than 12 years. Rural children were not different from urbanchildren in terms of sex, health status, and region in our sample. Similar sample characteristics were observed for 2- to 11- and 12- to 19-year-olds (Table 1).Overweight and Obesity by Urban and Rural Residence
In 1999–2006, among children aged 2–19 years old, the prevalence of overweight/obese was higher among rural children than urban children (urban 29.3%, rural 35.4%, p < 0.05). Likewise, a greater proportion of rural children aged 2–19 years old were obese than their urban counterparts (urban, 15.1%; rural, 18.6%; p < 0.05). By age group, the higher prevalence of overweight/obese or obesity among rural children was significant among adolescents (12–19 years old), but not among younger children (2–11 years old) (Figs. 1 and 2). When the younger children (2–11 years old) were divided into preschool children (2–5 years old) and school-aged children (6–11 years old), both preschool and school-aged rural children had a non-statistically significant higher prevalence of overweight/obesity or obesity than urban children of the same age (data not shown).
Obesity-Related Behaviors by Urban and Rural Residence
Among children aged 2–11 years old, rural children were less likely to be inactive than urban children (20.3% versus 26.2%, p < 0.05; Table 2). Rural children were not much different from urban children in terms of the percentages of excessive total screen time (≥2 hours per day) and the\consumption of sweetened beverages, whole grains, fruits, and vegetables.
However, rural children aged 2–11 years old averaged a higher total energy intake than urban children (1934 versus 1844 kcal, p < 0.05). More rural children reported consuming the recommended cups of dairy per day (2–3 cups per day)37 than urban children (Table 2). Among children aged 12–19 years old, there was no significant difference in physical activity and sedentary behavior by residence (Table 2).
Few dietary differences were found by residence among adolescents except fruit consumption. Fewer rural adolescents reported eating two or more cups of fruits per day than urban adolescents (12.2% versus 16.5%). In general, these data showed that most 2- to 11-year- lds and 12- to 19-year-olds in the United States were watching more than 2 hours of TV per day, eating less than 1 cup of vegetables per day, less than 2 cups of fruit per day, less than 2 cups of dairy per day, and consuming more than 24 ounces of sweetened beverages, but were not meeting activity recommendations or consuming whole grains, regardless of where they lived.
Residence and Obesity-Related Behaviors as Risk Factors for Overweight or Obesity After controlling for sociodemographic and health variables, additional adjustments for obesity-related behaviors did not change the estimates for the residence variable and the other behavioral variables. Thus, only the results from the full models are presented. Rural children aged 2–11 years old had higher odds of overweight or obesity than urban children [adjusted odds ratio (AOR) = 1.32, 95% confidence interval (CI) 1.03, 1.69] (Table 3).
Children aged 2–11 years old who drank 24 ounces or more of sweetened beverages (two or more regular cans) a day had higher odds of being overweight or obese than those who drank 0 to 8 ounces sweetened beverages (AOR = 1.23, 95% CI 0.99, 1.53). Likewise, 2- to 11-year-olds who consumed three or more cups of dairy products a day had higher odds of being overweight or obese than those who consumed less than 1 cup a day (AOR = 1.33, 95% CI 1.03, 1.73).
The difference in obesity in 2- to 11-year-olds by urban and rural residence was not significant in either crude or adjusted models (AOR = 1.24, 95% CI 0.96, 1.60) (Table 3). Among 12- to 19-year-olds, rural adolescents were more likely to be overweight/obese than urban adolescents (AOR = 1.33, 95% CI 1.08, 1.64) (Table 3). Similar differences were observed for obesity (AOR = 1.32, 95% CI 1.05, 1.67). Among adolescents, no consumption of whole grains and excessive screen time were associated with increased odds of being overweight or obese and obese (Table 3). Discussion The existing literature on childhood obesity in rural America is very limited.4–6,10,38 Using measured weight and height data from a nationally representative sample, this study found that both overweight and obesity were more prevalent among rural children.
These findings are consistent with published studies, although the prevalence of obesity was higher using measured data (18.6% rural vs. 15.1% urban) compared to the findings from reported data in the 2003 National Survey of Children’s Health (16.5% rural vs. 14.3% urban).4 Using 8 years worth of data from NHANES, this study had a larger rural sample than another study using 4 years worth of NHANES data,6 which allowed for the identification of a significant urban and rural difference in overweight and obesity among 12- to 19-year-old children.
The present study is unique in that it examined whether residence-based differences in childhood obesity may be explained by obesity-related behavioral factors. Among children aged 2–11 years, rural children were found to be slightly more active than urban children. Balancing this positive finding, a higher proportion of rural children spent 2 or more hours daily on screen activities, consumed more total calories, and reported a higher sweetened beverage intake than urban children.
After adjusting for sociodemographic, health, and obesity-related behavioral factors, the rural disadvantage in the prevalence of overweight status persisted. This suggests that the excessive risk associated with rural residence may not be fully explained by the risk factors considered in this study. More research is needed to understand residence-based differences in obesity and obesity-related behaviors in this age group by further separating preschool children (2–5 years old) from school-aged children (6–11 years old).
Due to a small sample of rural children, it was not possible to do this here. Among adolescents aged 12–19 years, obesity-related behaviors did not differ by residence with one exception. Rural adolescents consumed less fruit than urban adolescents. After adjusting for sociodemographic, health, and obesity-related behavioral factors, residence-based differences in overweight/obesity and obesity remained significant.
Excessive screen hours and no intake of whole grains were significant predictors of overweight and obesity status among adolescents. The persistently higher odds of overweight and obesity among rural children, even after controlling for sociodemographic, health, physical activity, and diet characteristics, suggests that rural environments may be “obesogenic” in ways that a person-level analysis cannot discern. Future studies are needed to explicate these contributors to childhood obesity. Environmental factors are promising areas for further examination. For example, lack of sidewalks and public transportation, long distance to recreational facilities, fewer destinations within walking distance, low population density, and perceived unsafe neighborhoods have been linked to higher obesity in rural communities among adults.39 The presence of recreational facilities has been associated with increased moderate-vigorous physical activity and reduced odds for overweight among adolescents.
40 Moreover, fewer grocery outlets, limited selection, and increased costs of fruits and vegetables may also lead to unhealthy diets in rural settings.41 Furthermore, nutritionists tend to be less available in rural areas, and fewer school and community nutrition education opportunities exist, compared to urban communities.42,43 Unfortunately, no data on the above-mentioned environmental factors were collected in NHANES. Compared to the published studies in this area, a strength of this study was the use of objectively measured weight and height data1–5 from a nationally representative survey. Using 8 years worth of NHANES data and a larger sample of rural children, this study was able to find a significantly higher prevalence of overweight/obese in rural children compared to urban children, which was not significant in another recent study using 4 years worth of NHANES data.27
This study also directly compared the Urban and rural differences in obesity-related risk behaviors such as diet, physical activity, and sedentary behaviors and examined how these behavioral factors might contribute to the urban and rural differences in childhood obesity, which was not considered in another study.27 Limitations of this study should be noted. First, the study’s cross-sectional design did not permit inferences about the causal direction of the relationship. It is possible that those children who were already overweight or obese, particularly adolescents, changed their diet and physical activity behaviors, limiting the usefulness of current physical activity and dietary variables in explaining overweight status.
Second, proxy-report (among children 2- to 11-year-olds) and self-report (among adolescents) of physical activity are less accurate than objective measures, such as pedometers or accelerometers. Third, although dietary information was obtained using the multiple pass 24-hour recall, the best approach available for estimating population level dietary trends,44 proxy interviews or assisted interviews were used to collect obesity-related behaviors among younger children (<12 years old). All recall methods suffer from limitations related to memory, social desirability, and quality of nutrition data matched to the recalls. Parental reports may slant toward “healthy” answers, regardless of actual behavior, allowing for confusing findings. Inaccurate diet measurements might bias our results to the null. Also the use of 1-day 24-hour recall might not provide a good overview of the children’s usual dietary intake. Fourth, while the RUCA definition offers a finer grain of observation than larger units, such as counties, it still may fail to get at differences in neighborhoods within each RUCA. Finally, as noted earlier, the sample was slightly biased by the need to exclude children with missing values on key covariates.
The sample included fewer young children, females, and poor children; the effect of this bias on the relationship between residence and obesity status is not known. Conclusion This study found that rural children experienced higher prevalence of overweight and obesity than urban children. Some evidence that rural disadvantages were associated with obesity-related behaviors among adolescents, but not among younger children was also found. Because excess weight in children and adolescents has significant ramifications for their immediate and long-term physical health,45–47 a higher prevalence of obesity in rural children can have lifelong implications for rural communities. Future research is much needed to understand the mechanisms for the rural disadvantage in childhood obesity at the individual level as well as at a higher structural level. Policies and programs are also needed to reduce gaps between rural and urban children’s risk of overweight and obesity.
Acknowledgment The study was funded by a grant award from the Office of Rural Health Policy, Health Resources and Services Administration (No 6 U1CRH03711-06-00). Special acknowledgment is due to Ms. Nusrat Harun for her assistance in data management in the early stage of this project. Author Disclosure Statement All authors have no competing financial interests to disclose.
References 1. Gamm LD, Huchison LL, Dabney BJ, et al. (eds). Rural Healthy People 2010: A Companion Document to Healthy People 2010. Texas A&M University System Health Services Center, School of Rural Public Health, Southwest Rural Health Research Center: College Station, TX, 2003.
2. Patterson PD, Moore CG, Probst JC, et al. Obesity and physical inactivity in rural America. J Rural Health 2004;20:151–159.
3. Jackson JE, Doescher MP, Jerant AF, et al. A national study of obesity prevalence and trends by type of rural county. J Rural Health 2005;21:140–148.
4. Liu J, Bennett KJ, Harun N, et al. Urban-rural differences in overweight status and physical inactivity among US children aged 10–17 years. J Rural Health 2008;24:407–415.
5. Lutfiyya MN, Lipsky MS, Wisdom-Behounek J, et al. Is rural residency a risk factor for overweight and obesity for U.S. Children? Obesity (Silver Spring) 2007;15:2348–2356.
6. Davis AM, Bennett KJ, Befort C, et al. Obesity and related health behaviors among urban and rural children in the United States: Data from the National Health And Nutrition Examination Survey 2003–2004 and 2005–2006. J Pediatr Psychol 2011;36:669–676.
7. Wang Y, Beydoun MA. The obesity epidemic in the United States—gender, age, socioeconomic, racial/ethnic, and geographic characteristics: A systematic review and meta-regression analysis. Epidemiol Rev 2007;29:6–28.
8. S ingh GK, Kogan MD, van Dyck PC. A multilevel analysis of state and regional disparities in childhood and adolescent obesity in the United States. J Community Health 2007;30:206–216.
9. Probst JC, Moore CG, Glover SH, et al. Person and place: The compounding effects of race/ethnicity and rurality on health. Am J Public Health 2004;94:1695–1703.
10. Lewis RD, Meyer MC, Lehman SC, et al. Prevalence and degree of childhood and adolescent overweight in rural, urban, and suburban Georgia. J Sch Health 2006;76:126–132.
11. T ai-Seale T, Chandler C, editors. Nutrition and overweight concerns in rural areas: A literature review. The Texas A&M University System Health Science Center, School of Rural Public Health, Southwest Rural Health Research Center, College Station, TX, 2003.
12. S harkey JR, Johnson CM, Dean WR, et al. Association between proximity to and coverage of traditional fast-food restaurants and non-traditional fast-food outlets and fast-food consumption among rural adults. Int J Health Geogr 2011;10:37.
13. Creel JS, Sharkey JR, McIntosh A, et al. Availability of healthier options in traditional and nontraditional rural fast-food outlets. BMC Public Health 2008;8:395.
14. S cott AJ, Wilson RF. Social determinants of health among African Americans in a rural community in the Deep South: An ecological exploration. Rural Remote Health 2011;11:1634.
15. B evans KB, Fitzpatrick LA, Sanchez BM, et al. Physical education resources, class management, and student physical activity levels: A structure- process-outcome approach to evaluating physical education effectiveness. J Sch Health 2010;80:573–580.
16. B ovell-Benjamin A, Dawkins N, Pace R, et al. Dietary consumption practices and cancer risk in African Americans in the rural South. J Health Care Poor Underserved 2010;21:57–75.
17. B rener ND, Collins JL. Co-occurrence of health-risk behaviors among adolescents in the United States. J Adolesc Health 1998;22:209–213.
18. Atav S, Spencer GA. Health risk behaviors among adolescents attending rural, suburban, and urban schools: A comparative study. Fam Community Health 2002;25:53–64.
19. Fahs PS, Smith BE, Atav AS, et al. Integrative research review of risk behaviors among adolescents in rural, suburban, and urban areas. J Adolesc Health 1999;24:230–243.
20. Pearson N, Atkin AJ, Biddle SJ, et al. Patterns of adolescent physical activity and dietary behaviours. Int J Behav Nutr Phys Act 2009;6:45.
21. Centers for Disease Control and Prevention/National Center for Health Statistics. About National Health and Nutrition Examination Survey. Available at www.cdc.gov/nchs/nhanes/about_nhanes.htm/.
Last accessed August 10, 2012.
22. Hart LG, Larson EH, Lishner DM. Rural definitions for health policy and research. Am J Public Health 2005;95:1149–1155.
23. E conomic Research Service. Measuring Rurality: Rural-Urban Commuting Area Codes. United States Department of Agriculture, 2005.
24. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States. Adv Data 2000;314:1–27.
25. Gordon-Larsen P, McMurray RG, Popkin BM. Adolescent physical activity and inactivity vary by ethnicity: The National Longitudinal Study of Adolescent Health. J Pediatr 1999;135:301–306.3
26. U.S. Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans Summary. Available at www. health.gov/PAGuidelines/. Last accessed August 10, 2012.
27. American Academy of Pediatrics: Children, adolescents, and television. Pediatrics 2001;107:423–426.
28. American Academy of Pediatrics. Media violence. Committee on Public Education. Pediatrics 2001;108:1222–1226.
29. Resnicow K, Odom E, Wang T, et al. Validation of three food frequency questionnaires and 24-hour recalls with serum carotenoid levels in a sample of African-American adults. Am J Epidemiol 2000;152:1072–1080.
30. Kahn HA, Whelton PK, Appel LJ, et al. Validity of 24-hour dietary recall interviews conducted among volunteers in an adult working community. Ann Epidemiol 1995;5:484–489.
31. Jonnalagadda SS, Mitchell DC, Smiciklas-Wright H, et al. Accuracy of energy intake data estimated by a multiple-pass, 24-hour dietary recall technique. J Am Diet Assoc 2000;100:303–308; quiz 9–11.
32. S lavin JL. Dietary fiber and body weight. Nutrition 2005;21:411– 418.
33. B erkey CS, Rockett HR, Willett WC, et al. Milk, dairy fat, dietary calcium, and weight gain: A longitudinal study of adolescents. Arch Pediatr Adolesc Med 2005;159:543–550.
34. Welsh JA, Cogswell ME, Rogers S, et al. Overweight among low-income preschool children associated with the consumption of sweet drinks: Missouri, 1999–2002. Pediatrics 2005;115:e223– e229.
35. James J, Thomas P, Cavan D, et al. Preventing childhood obesity by reducing consumption of carbonated drinks: Cluster randomised controlled trial. Br Med J 2004;328:1237.
36. Centers for Disease Control and Prevention / National Center for Health Statistics. NHANES Analytic and Reporting Guidelines. Available at www.cdc.gov/nchs/data/nhanes/nhanes_03_04/nhanes_analytic_guidelines_dec_2005.pdf/. Last accessed August 10, 2012.
37. Centers for Disease Control and Prevention. How many fruits and vegetables do you need? Available at www.fruitsandveggiesmattergov/ . Last accessed August 10, 2012.
38. McMurray RG, Harrell JS, Bangdiwala SI, et al. Cardiovascular disease risk factors and obesity of rural and urban elementary school children. J Rural Health 1999;15:365–374.
39. B oehmer TK, Lovegreen SL, Haire-Joshu D, et al. What constitutes an obesogenic environment in rural communities? Am J Health Promote 2006;20:411–421.
40. Gordon-Larsen P, Nelson MC, Page P, et al. Inequality in the built environment underlies key health disparities in physical activity and obesity. Pediatrics 2006;117:417–424.
41. Liese AD, Weis KE, Pluto D, et al. Food store types, availability, and cost of foods in a rural environment. J Am Diet Assoc 2007;107:1916–1923.
42. Leeper J, Hullett S, Wang L. Rural Alabama Health Professional Training Consortium: Six-year evaluation results. Fam Community Health 2001;24:18–26.
43. Lindseth G. Evaluating rural nurses for preparation in implementing nutrition interventions. J Rural Health 1990;6:231–245.
44. T hompson FE, Subar AF. Dietary Assessment Methodology. Nutrition in the Prevention and Treatment of Disease, 2nd ed. Elsevier: San Diego, CA, 2001.
45. Parsons TJ, Power C, Logan S, et al. Childhood predictors of adult obesity: A systematic review. Int J Obes Relat Metab Disord 1999;23:S1–S107.
46. Freedman DS, Dietz WH, Srinivasan SR, et al. The relation of overweight to cardiovascular risk factors among children and adolescents: The Bogalusa Heart Study. Pediatrics 1999;103:1175–1182.
47. Narayan KM, Boyle JP, Thompson TJ, et al. Lifetime risk for diabetes mellitus in the United States. JAMA 2003;290:1884–1890. Address correspondence to: Ji-Hong Liu, Sc.D. Department of Epidemiology and Biostatistics Arnold School of Public Health University of South Carolina 800 Sumter Street Columbia, SC 29208 E-mail: email@example.com