 |
|
Trends in the Association of Poverty With Overweight Among US Adolescents, 1971-2004
Richard A. Miech, PhD, MPH;
Shiriki K. Kumanyika, PhD, MPH;
Nicolas Stettler, MD, MSCE;
Bruce G. Link, PhD;
Jo C. Phelan, PhD;
Virginia W. Chang, MD, PhD
JAMA. 2006;295:2385-2393.
ABSTRACT
| |
Context Prevalence of adolescent overweight in the United States has increased substantially during the past 3 decades. Whether socioeconomic disparities in adolescent overweight increased, decreased, or remained constant during this period is not known.
Objective To examine trends in adolescent overweight from 1971 to 2004 by family poverty status, as well as trends in potentially relevant eating and physical activity behaviors.
Design, Setting, and Participants Four cross-sectional, nationally representative surveys (US National Health and Nutrition Examination Surveys [NHANES] of 1971-1974, 1976-1980, 1988-1994, and 1999-2004) were examined for trends in the prevalence of overweight among adolescents aged 12 to 17 years by family poverty status.
Main Outcome Measures Prevalence of adolescent overweight, defined as body mass index at or above the 95th percentile for age and sex in the 2000 Centers for Disease Control and Prevention growth charts. Intermediate outcomes were physical inactivity in the past 30 days, proportion of caloric intake from sweetened beverages (24-hour recall), and whether respondent skipped breakfast (24-hour recall).
Results Trends in the association of adolescent overweight with family poverty differed by age stratum (P = .01). In 12- to 14-year-old adolescents, prevalence did not significantly differ by family poverty status in any of the surveys; however, among non-Hispanic black adolescents, overweight prevalence increased faster in nonpoor vs poor families. In contrast, a widening disparity that disfavored adolescents from poor families was present in the 15- to 17-year-old adolescents. This trend was similar among male, female, non-Hispanic white, and non-Hispanic black adolescents, resulting in an overall prevalence of overweight in 1999-2004 more than 50% higher among adolescents in poor vs nonpoor families (23.3% vs 14.4%, respectively; P<.001). Additional analyses suggest that physical inactivity, sweetened beverage consumption, and skipping breakfast may contribute to these disparities.
Conclusions Trends of increasing overweight showed a greater impact in families living below the poverty line vs not living below the poverty line among older (15-17 years) but not younger (12-14 years) adolescents. Furthermore, physical inactivity, high consumption of sweetened beverages, and breakfast skipping may be candidate targets for prevention programs aimed at reducing this recently emerged disparity.
INTRODUCTION
The prevalence of overweight status among adolescents in the United States has more than doubled during the past 3 decades.1-3 This increase has been noted among male and female adolescents, and across different racial and ethnic groups.1-2,4 Recent data from 1999 through 2004 show no evidence that the trend is reversing.1 As the prevalence of adolescent overweight increases, so too will its associated consequences, including type 2 diabetes mellitus, hypertension, obstructive sleep apnea, poor quality of life, and increased morbidity and mortality in adulthood.5-8
Whether the increasing prevalence of adolescent overweight is characterized by larger, smaller, or unchanged disparities in overweight status across socioeconomic strata is not known but is important for evaluating the success of recent efforts by the Department of Health and Human Services to reduce health disparities.9 To address this question, we examined trends in the prevalence of overweight among adolescents aged 12 to 17 years by family poverty status in 4 nationally representative, cross-sectional data sets spanning from 1971 to 2004.
Our primary goal was to test the hypothesis that the disparity in overweight has widened among US adolescents. Individuals in the upper social strata are generally better able to protect themselves from diseases and conditions that are highly preventable10-11 and may have been better able to resist the factors responsible for the current obesity epidemic. Because environmental influences on adolescent overweight differ substantially across older and younger adolescents,12-13 we considered the possibility that any widening in the disparity may vary by age among adolescents. A secondary goal of our study was to examine the potential role of behaviors related to energy intake and energy expenditure in disparities of adolescent overweight.
METHODS
Surveys
Data were from the US National Health and Nutrition Examination Survey (NHANES) program of the National Center for Health Statistics, Centers for Disease Control and Prevention (CDC). Each NHANES cross-sectional survey provides national estimates for the US population at the time of the survey, enabling examination of trends over time. In each survey, the NHANES program selected a nationally representative sample of the US civilian noninstitutionalized population by using a complex, stratified, multistage probability cluster sampling design. The analyses of our study use data from NHANES I (1971-1974), NHANES II (1976-1980), and NHANES III (1988-1994). Beginning in 1999, NHANES became a continuous survey without a break between cycles; therefore, our study also uses data from the first 6 years of the continuous NHANES (1999-2004). NHANES III and NHANES 1999-2004 have been approved by institutional review boards and include written informed consent by study participants or their legal guardians. Institutional review board approval using current standards was not obtained for NHANES I or II, but internal human subjects review was conducted.
Outcome Measures
Adolescents' weight and height were measured by trained technicians in mobile examination centers who used standardized procedures and equipment. Overweight status was defined as recommended by the CDC. Unlike the definitions for adults, the CDC uses the term overweight rather than the term obesity to refer to the highest body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) category for children and adolescents. Overweight in adolescents is identified as BMI at or above the 95th percentile for sex and age, according to the 2000 CDC growth charts for the United States.14 The BMI growth charts are based on national data collected between 1963 and 1994; therefore, the prevalence of overweight may exceed 5% in recent surveys. Additional analyses were performed substituting the outcome of adolescent overweight status with the following: at risk for overweight or overweight, defined as a BMI at or above the 85th percentile for age and sex in the 2000 CDC growth charts; unadjusted BMI; and BMI, adjusted for age and sex using standard deviation scores (z scores).
A parent or caretaker of the participating adolescent was asked to sum the income of all family members (eg, earnings, unemployment compensation, or social security). The poverty income ratio is the ratio of a family's income to the US Census Bureau's poverty threshold, which varies with the number and ages of family members and is revised yearly.15-16 For example, the US poverty threshold for a family of 4 with 2 dependent children in 2004 was $19 157.17 The poverty income ratio allows comparison across time because it is continually revised to account for inflation. In our analysis, family poverty is a dichotomous variable that indicates a poverty income ratio of less than 1. In the 4 NHANES surveys, the percentage of adolescents in poverty was 16%, 18%, 22%, and 19% in chronological order. Poverty is a strategic measure of socioeconomic status in studies of overweight because poverty is linked to both poor diet and physical inactivity.18-20 We did not include education because educational categories have shifted substantially in meaning and value during the 3 decades covered in our study.
Proportion of caloric intake from sweetened beverages was derived from the 24-hour dietary recall interviews of the NHANES 1999-2002 (currently the most recently released NHANES dietary recall) and the NHANES 1988-1994, which collected dietary information using similar procedures and codes. At the mobile examination centers, participants reported all food eaten in the prior 24-hour period (midnight to midnight) by using a structured interview procedure.21-24 For each respondent, the amount of kilocalories derived from sweetened beverages (as indexed by foods with prefixes 924-927 inclusive in the Nutrient Database System25) divided by the total kilocalories consumed was calculated. Breakfast skipping was defined as a dichotomous variable coded 1 for respondents who reported 0 calories from breakfast in the 24-hour recall data. Analyses that use the 24-hour recall data excluded respondents with missing information (3% of respondents with information on BMI in each of the 2 NHANES surveys). Physically inactive was defined as a dichotomous measure indicating respondents who reported both no vigorous and also no moderate physical activity during the 30 days before the questionnaire interview in NHANES 1999-2004. Analyses that used this variable excluded respondents with missing information (4% of respondents with information on BMI). A comparable measure of physical inactivity was not included in earlier NHANES surveys.
Age and sex were self-reported at the time of examination in the mobile examination center. Race was based on interviewer observation in NHANES I and II, and on self-report in later NHANES surveys. Hispanic ethnicity was based on self-report in NHANES II and later surveys; NHANES I did not collect sufficient information to identify Hispanic respondents. Year of NHANES survey was coded as 0 for NHANES 1971-1974, 6 for NHANES 1976-1980, 19 for NHANES 1988-1994, and 29 for NHANES 1999-2004 to indicate the time between surveys.
Data Analysis and Statistical Methods
Data were analyzed using STATA version 8.0.26 Pregnant female adolescents were excluded from the analysis. All surveys used a complex sampling design and drew individuals from preselected strata and primary sampling units. All analyses used the NHANES-provided sampling weights that were calculated to take into account unequal probabilities of selection resulting from the sample design, nonresponse, and planned oversampling of selected subgroups. Standard errors were calculated with STATA version 8.0 using Taylor series linearization. P<.05 was considered statistically significant.
Analyses included tests of 2-way and 3-way multiplicative interaction terms in regression equations.27-28 All regressions with interactions included the necessary subcomponents. For example, a test of a 3-way interaction included the 3 necessary 2-way interaction terms and the 3 main components.
RESULTS
Survey Respondents
Table 1 shows the unweighted sample sizes for each analytic stratum. For NHANES II (1976-1980), the number of Hispanic and non-Hispanic other race respondents was too small to support statistical analysis; therefore, the analysis did not include separate analyses for these groups. Limited sample sizes also precluded meaningful analyses of the association with poverty by sex within race/ethnicity.
|
|
|
|
Table 1. Number of Survey Respondents by Family Poverty, Sex, Race/Ethnicity, and NHANES Survey*
|
|
|
Prevalence of Adolescent Overweight
Figure 1 shows the prevalence of adolescent overweight by 3 levels of family income, age, and NHANES survey. These results show that differences in the prevalence of adolescent overweight were most evident among respondents aged 15 to 17 years in the last 2 NHANES surveys. Figure 1A illustrates the lack of large disparities in adolescent overweight, at any age, in the first 2 NHANES surveys. Figure 1B shows that in the last 2 NHANES surveys, differences in the prevalence of adolescent overweight between family income levels was substantially larger among respondents aged 15 to 17 years than it was among respondents aged 12 to 14 years. The results also indicate that the largest discrepancy in adolescent overweight was for families below poverty level compared with other income levels, in which the prevalence of adolescent overweight was similar. Subsequent analyses used the dichotomous variable below poverty level as a measure of family income and examined potential differences in trends between adolescents aged 12 to 14 years and 15 to 17 years.
|
|
|
|
Figure 1. Adolescent Overweight by Poverty Status, NHANES Survey, and Age of Respondent
NHANES indicates National Health and Nutrition Examination Survey. Overweight is defined as at least the 95th percentile body mass index for age, as defined by the 2000 Centers for Disease Control and Prevention growth chart.
|
|
|
To test whether a disparity in overweight by family poverty widened in recent decades, the analysis included all adolescents aged 12 to 17 years in 1 analysis pool that combined all NHANES surveys and tested whether the prevalence of adolescent overweight increased more for adolescents in poverty compared with those not in poverty. These results did not support a widening or a narrowing disparity for the combined group of adolescents. The empirical model consisted of the 3 variables of family poverty, year of NHANES, and the multiplicative interaction of these 2 variables as predictors of adolescent overweight. The interaction term was not large or statistically significant, indicating that for the group of adolescents aged 12 to 17 years disparities in overweight did not change over the 4 NHANES surveys (unexponentiated  = .003; 95% confidence interval [CI], –.01 to .02; P = .70). However, this model did not take into account the possibility that disparities in adolescent overweight may differ by age group.
Preliminary analysis (Figure 1) led us to test whether a widening disparity in adolescent overweight by family poverty was specific to older adolescents (15-17 years). A widening disparity specific to older age groups would be indicated by a 3-way interaction of age group (15-17 vs 12-14 years), poverty, and year of survey. This 3-way interaction term was statistically significant (unexponentiated = .033; 95% CI, .007-.059; P=.01) and indicated that analysis of trends in poverty and adolescent overweight should consider separately the age groups 12 to 14 years and 15 to 17 years. Subsequent analyses stratified the sample by these age categories.
Trends in the distribution of overweight by family poverty among adolescents aged 12 to 14 years are shown in Table 2 and Figure 2A. In general, at no year of NHANES surveys did the prevalence of overweight status significantly differ across poor and nonpoor families among all adolescents aged 12 to 14 years, or within any of the sex and racial/ethnic subsamples. Table 2 also shows the change in the disparity over the period covered by the surveys and in almost all cases the change was small. An exception was found among non-Hispanic black adolescents aged 12 to 14 years, for whom a disparity in adolescent overweight reversed over time so that the prevalence of overweight was higher among nonpoor families compared with poor families by the last years of the NHANES survey.
|
|
|
|
Table 2. Prevalence of Overweight Status for Adolescents Aged 12 to 14 Years for Each NHANES Survey by Poverty, Sex, and Race/Ethnicity
|
|
|
|
|
|
|
Figure 2. Trend in Adolescent Overweight by Poverty Status and NHANES Surveys, 1971-2004
NHANES indicates National Health and Nutrition Examination Survey. Overweight is defined as at least the 95th percentile body mass index for age, as defined by the 2000 Centers for Disease Control and Prevention growth chart. Error bars indicate 95% confidence intervals of observed values.
|
|
|
Table 3 shows trends in the distribution of overweight by family poverty for adolescents aged 15 to 17 years. Results for all adolescents aged 15 to 17 years are shown in Table 3 and depicted graphically in Figure 2B. In general, the prevalence of overweight increased over the period of the surveys at a faster rate among adolescents from poor families compared with nonpoor families. In the first 2 NHANES surveys (1971-1974 and 1976-1980), no difference in the prevalence of overweight between poor and nonpoor families was apparent, but in the last 2 NHANES surveys (1988-1994 and 1999-2004), adolescents in poor families had a higher prevalence of overweight than adolescents in nonpoor families. This trend was apparent in all the sex and racial/ethnic subgroups, although it was not always statistically significant. The absolute difference between the prevalence of overweight for adolescents from poor vs nonpoor families increased by approximately 11% during the course of the survey. About the same magnitude in the increase of disparity over time was shown in each sex and racial/ethnic subgroup, despite different initial levels of disparities in overweight.
|
|
|
|
Table 3. Prevalence of Overweight Status for Adolescents Aged 15 to 17 Years for Each NHANES Survey by Poverty, Sex, and Race/Ethnicity
|
|
|
In confirmatory analyses using alternative outcomes of at risk for overweight or overweight, BMI, or BMI z score, results were similar or stronger than they were in the analyses of the outcome of adolescent overweight. Among all adolescents aged 15 to 17 years, a significant increase in disparity for each of these outcomes by family poverty during the study period was present, as well as in the subgroups of female, male, non-Hispanic white, and non-Hispanic black adolescents. These trends were not present among adolescents aged 12 to 14 years.
Behaviors Related to Energy Expenditure and Dietary Intake
We next examined the extent to which the trends in adolescent overweight documented could plausibly result from comparable major trends in behaviors related to energy expenditure and consumption. These analyses centered on the older adolescents (15-17 years) in the most recent NHANES surveys and examined whether any disparity in these trends by family poverty were more pronounced in comparison both with younger adolescents (12-14 years) in the same NHANES survey and older adolescents (15-17 years) in the previous NHANES surveys (data permitting). We first considered self-reported physical inactivity in NHANES 1999-2004, which was significantly associated with overweight among older adolescents aged 15 to 17 years (odds ratio [OR], 1.71; 95% CI, 1.12-2.62), although not among younger adolescents aged 12 to 14 years (OR, 0.77; 95% CI, 0.47-1.25).
Physical Activity Trends
A disparity in physical inactivity by family poverty was more pronounced for older compared with younger adolescents in the cross-sectional NHANES 1999-2004 (Figure 3A). The predicted scores come from a model in which physical inactivity was modeled as a function of family poverty status, age, and an age-squared term (for which P<.01). A multiplicative interaction of age and family poverty status was not statistically significant. However, the disparity in physical inactivity by poverty status increased with age, measured by the absolute difference in the prevalence of physical inactivity. This absolute difference in the predicted score was 6.6% among respondents aged 12 years (13.0% vs 6.5%), and it increased by age 17 years to 11.2% (23.8% vs 12.6%). The relative difference in the predicted score was similar across all ages and indicated that adolescents in poor families had about twice the level of physical inactivity compared with adolescents from nonpoor families. Among adolescents aged 12 to 17 years, girls reported significantly more physical inactivity than boys (13% vs 8%, P<.001), but the percentage increase in inactivity from ages 12 to 17 years was similar across sex. The earlier NHANES surveys did not contain similar measures of physical inactivity to allow historical comparison.
|
|
|
|
Figure 3. Trends in Risk Factors for Adolescent Overweight by Age and Poverty Status
NHANES indicates National Health and Nutrition Examination Survey. Predicted results are from regression equations of outcome on year of age and poverty status. A, Equation includes an age-squared term. B, Equation includes a multiplicative interaction of poverty status and age. C, Equation includes only year of age and poverty status as predictor variables.
|
|
|
Dietary Trends
We next examined whether major trends in dietary behaviors in recent decades have disproportionately affected older adolescents in poverty. We considered the major dietary trends that have been identified in an extensive literature review29 and for which good measures existed in the NHANES surveys. These included sweetened beverage consumption, breakfast skipping, percentage of calories consumed while eating away from home, and percentage of calories from snacking between meals. The first 2 factors were significantly associated with overweight among adolescents aged 12 to 17 years in NHANES 1999-2002 (P<.001 for both, respectively), but the second 2 factors were not associated with overweight and therefore were not considered any further.
A disparity across poverty status in percentage of calories from sweetened beverages was greater for older vs younger adolescents in the cross-sectional NHANES 1999-2002 (Figure 3B). Predicted results are shown from an analysis in which proportion of calories from sweetened beverages was modeled as a linear function of age (an age-squared term that would have indicated a curvilinear function was not significant), family poverty status, and the multiplicative interaction of these 2 variables (P = .07). The borderline statistically significant interaction suggests that the disparity in this risk factor across poor and nonpoor families (adolescents) became larger with increasing age.
Because the last 2 NHANES surveys (1988-1994 and 1999-2002) contain similar dietary measures, they allowed comparison of caloric intake from sweetened beverages during the past decade. The analyses indicated that a disparity across poverty status in proportion of calories from sweetened beverage consumption significantly increased across the 2 surveys, and this increase was concentrated among adolescents aged 15 to 17 years in poor families. For the combined age-group 12 to 17 years, the percentage of caloric intake from sweetened beverages significantly increased from 10.6% in NHANES 1988-1994 to 12.2% in NHANES 1999-2002 (P = .01). This increase was concentrated among respondents aged 15 to 17 years (from 10.7% to 13.2%, P = .003) and was not present among respondents aged 12 to 14 years (from 10.6% to 11.3%, P = .28). Among respondents aged 15 to 17 years, the increase across the last 2 NHANES surveys (1988-1994 and 1999-2002) was significantly larger in poor families (from 9.2% to 15.4%) compared with nonpoor families (from 11.1% to 12.6%). This finding was indicated by a positive, multiplicative interaction (P = .002) of NHANES survey and poverty status in a model predicting proportion of calories from sweetened beverages that combined adolescents aged 15 to 17 years from both NHANES surveys into 1 analysis pool.
Figure 3C shows a disparity across poverty status in the prevalence of breakfast skipping in NHANES 1999-2002 that had a similar magnitude across all adolescents aged 12 to 17 years. This disparity (P<.001) appears to have emerged recently because, in the NHANES 1988-1994 survey, breakfast skipping was not significantly associated with poverty status for the combined group of adolescents aged 12 to 17 years (P = .10). Among older adolescents aged 15 to 17 years, the absolute difference in the disparity increased from 1.1% to 16.2% across the last 2 NHANES surveys (the prevalence for poor vs nonpoor families was 32.7% vs 31.7% in NHANES 1988-1994 and 45.6% vs 29.4% in NHANES 1999-2002, respectively). This increase was significant (P = .02) in analyses that tested the interaction of NHANES survey and poverty status in a combined pool of respondents from NHANES 1999-2002 and NHANES 1988-1994. Among younger adolescents aged 12 to 14 years, the absolute difference in the disparity did not significantly change across the last 2 NHANES surveys (the prevalence for poor vs nonpoor families was 33.5% vs 21.7% in NHANES 1988-1994 and 33.6% vs 23.5% in NHANES 1999-2002, respectively).
COMMENT
We examined whether socioeconomic disparities in adolescent overweight status increased, decreased, or remained constant during the past 3 decades and whether these trends varied by age among US adolescents. Our results indicated an increasing disparity in adolescent overweight status by family poverty among adolescents aged 15 to 17 years but not among adolescents aged 12 to 14 years. For the older adolescents, the magnitude of the increase in the disparity was similar across the demographic subgroups of male, female, non-Hispanic white, and non-Hispanic black respondents. For the younger adolescents, no disparity in overweight by family poverty was apparent for any demographic group at any of the 4 survey years; although among non-Hispanic black adolescents, the overall rate of overweight during the course of the surveys increased faster in nonpoor vs poor families, a finding consistent with trends in the adult population.30
To our knowledge, our study is the first to document an increasing disparity in overweight by poverty status in any US population group. These results challenge the current conclusion that disparities in overweight by poverty status have not increased in the United States, a conclusion based on analyses of adults31 and also from data from the mid-1990s that combined younger and older adolescents into a single group.32 Our findings that later adolescence is a life stage with a unique association between poverty and overweight is plausible because both food choices and physical activity levels in adolescence differ considerably from those earlier in childhood33-34 and adulthood.35-36 Our results therefore point to the utility of a life-course perspective that takes into account findings specific to life stages.
What factors explain an overweight disparity by poverty status that is specific both to older adolescents and also more recent NHANES surveys? The observed differences across older vs younger adolescents are consistent with the greater autonomy that comes with increasing age. Adolescents aged 15 to 17 years vs those aged 12 to 14 years have more opportunities to purchase their own food and determine their own leisure time pursuits and also have more discretionary income with which to act on their preferences.
After identifying trends in poverty and adolescent overweight among older adolescents, we then examined ex post facto trends in specific behaviors that may play a role in this disparity. Our analysis investigated dietary behaviors with documented increases in recent years,29 including sweetened beverage consumption, eating out, and snacking between meals. We also examined physical inactivity37-38 and breakfast skipping, which observational studies show to be associated with adolescent overweight.39 A behavior involved in the emerging disparity identified in our study should fulfill a specific and detailed set of 4 expectations. This disparity should have higher prevalence for those families in poverty, higher prevalence among adolescents who are overweight, show a greater disparity by poverty status among older vs younger adolescents, and have developed in recent years.
Proportion of calories from sweetened beverages followed this exact pattern of findings, suggesting that recent trends in sweetened beverage consumption among older adolescents may, at least in part, explain the disparity in overweight status. Physical inactivity also followed this pattern of findings, although measures were restricted to the most recent NHANES survey and historical comparisons were not possible. Other studies have noted that physical inactivity among adolescents has been increasing in recent years.40 These results suggest that policy and prevention efforts that target these behaviors for overweight41-42 can be further justified and motivated by the potential of these efforts to reduce an emerging socioeconomic disparity in adolescent overweight.
Breakfast skipping also came close to fulfilling all 4 criteria, with the exception that in the most recent NHANES survey the recently emerged disparity had the same magnitude for younger and older adolescents. These results point to breakfast skipping as a factor that warrants more attention in the study of adolescent overweight. To our knowledge, there has been no randomized controlled trial of breakfast promotion for overweight prevention or treatment, and many of the intervening mechanisms remain speculative. One longitudinal study43 showed that adolescents with normal weight who skipped breakfast were more likely to experience BMI increases over time. This result supports possible mechanisms, such as rebound overeating at lunch and dinner, a poor dietary pattern characterized by soda and snacking in place of regular meals, or possibly a biological effect associated with efficient calorie utilization after long periods of not eating. At the same time, breakfast skipping may also be used as a dieting strategy among overweight individuals.
Our study was limited to analysis of risk factors that were measured in the NHANES surveys, and these are likely only a subset of a much larger pool that are at work in the emerging disparity in adolescent overweight by poverty. Availability of energy-dense food,44-45 perceived dangerousness of neighborhoods,46 and limited access to supermarkets that sell nutritious, low-calorie food47 are a few examples of additional circumstances that are currently linked to adolescent overweight and warrant more attention in analysis of disparities in overweight. At the same time, current influences that have fostered a socioeconomic disparity in adolescent overweight have emerged recently and new factors may emerge in the near future. Ultimately, successful intervention and prevention of the emerging disparity will require constant surveillance of potential intervening mechanisms and, ideally, attention to the upstream influences that set these intervening mechanisms into motion.11
Our study has some limitations. The BMI-based definition for adolescent overweight is recommended by the CDC, but the best way to identify overweight and obesity in children is still a subject of debate.3 Another limitation is that data from Hispanic adolescents could not be meaningfully analyzed due to differences in sampling strategies and in measurement of Hispanic ethnicity across the NHANES surveys. Furthermore, the sample sizes were too small to allow examination of sex differences within racial/ethnic groups over time, which are present among adults.30-31,48-49 The cross-sectional design and methods used for NHANES limit investigation to only broad, major risk factors for overweight and do not have the precision to measure small, cumulative energy imbalances of only dozens of calories that can lead to overweight if experienced over protracted periods.50 Finally, our measurement of family socioeconomic status is limited to self-reported family income, because the NHANES data lack consistent measures of other socioeconomic status components, such as family education, wealth, neighborhood conditions, and past socioeconomic status experience. Socioeconomic status is a multidimensional construct and single measures of socioeconomic status should be accompanied by measures of other socioeconomic status dimensions when data are available.51
Counterbalancing these limitations are the strengths of our study. The main study finding of an increasing socioeconomic disparity in overweight among adolescents aged 15 to 17 years was robust across 4 different weight status indices and across sex and racial subgroups. Additional strengths include the NHANES measurement of height and weight by trained technicians, which removes the results from the potential biases associated with self-reports,52-53 and that the sample is representative of all US adolescents.
In conclusion, a widening disparity in overweight that disadvantages adolescents in poor families has emerged in the 15- to 17-year-old age group in recent years. These results suggest that efforts to reduce health disparities in the United States require monitoring of population health, so that emergent disparities and their underlying causes can be detected and addressed at early stages of their development.
AUTHOR INFORMATION
Corresponding Author: Richard A. Miech, PhD, MPH, Department of Mental Health, Bloomberg School of Public Health, Johns Hopkins University, 624 N Broadway, Hampton House #853, Baltimore, MD 21205 (rmiech{at}jhsph.edu).
Author Contributions: Dr Miech had full access to all of the data in the study and takes full responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Miech, Kumanyika, Link, Phelan, Chang.
Acquisition of data: Miech.
Analysis and interpretation of data: Miech, Kumanyika, Stettler, Link, Phelan, Chang.
Drafting of the manuscript: Miech, Kumanyika, Stettler, Chang.
Critical revision of the manuscript for important intellectual content: Miech, Kumanyika, Stettler, Link, Phelan, Chang.
Statistical analysis: Miech, Stettler, Chang.
Administrative, technical, or material support: Miech.
Financial Disclosures: None reported.
Funding/Support: Dr Miech is supported in part by grant 1K01DA015089 from the National Institutes of Health (NIH). Dr Kumanyika is supported in part by project EXPORT grant P60 MD000209 from the National Center for Minority Health and Health Disparities at the NIH. Dr Stettler is supported in part by grant 5K23RR016073 from the NIH. Drs Link and Phelan have been supported in part by a Robert Wood Johnson Foundation Health Policy Investigator Award. Dr Phelan is also supported by grant K02MH65330 from the NIH. Dr Chang is supported in part by grant K12HD043459 from the NIH.
Role of the Sponsors: The funding organizations listed above were not involved in the design or conduct of the study, data management, analysis, or interpretation of the data, or in the preparation, review, or approval of the manuscript.
Author Affiliations: Department of Mental Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Md (Dr Miech); Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia (Drs Kumanyika and Stettler); Division of Gastroenterology, Hepatology, and Nutrition, The Children's Hospital of Philadelphia, Philadelphia, Pa (Dr Stettler); Departments of Epidemiology (Dr Link) and Sociomedical Sciences (Drs Link and Phelan), Mailman School of Public Health, Columbia University, New York, NY; New York State Psychiatric Institute, New York (Dr Link); and Veterans Affairs Center for Health Equity Research and Promotion, Department of Medicine, University of Pennsylvania School of Medicine, and Department of Sociology, University of Pennsylvania, Philadelphia (Dr Chang).
REFERENCES
| |
1. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA. 2006;295:1549-1555.
FREE FULL TEXT
2. 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.
FREE FULL TEXT
3. Troiano RP, Flegal KM. Overweight children and adolescents: description, epidemiology, and demographics. Pediatrics. 1998;101:497-504.
FREE FULL TEXT
4. Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents: the National Health and Nutrition Examination Surveys 1963 to 1991. Arch Pediatr Adolesc Med. 1995;149:1085-1091.
FREE FULL TEXT
5. Daniels SR, Arnett DK, Eckel RH, et al. Overweight in children and adolescents: pathophysiology, consequences, prevention, and treatment. Circulation. 2005;111:1999-2012.
FREE FULL TEXT
6. Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public-health crisis, common sense cure. Lancet. 2002;360:473-482.
FULL TEXT
|
ISI
| PUBMED
7. Williams J, Wake M, Hesketh K, Maher E, Waters E. Health-related quality of life of overweight and obese children. JAMA. 2005;293:70-76.
FREE FULL TEXT
8. Schwimmer JB, Burwinkle TM, Varni JW. Health-related quality of life of severely obese children and adolescents. JAMA. 2003;289:1813-1819.
FREE FULL TEXT
9. US Department of Health and Human Services. Healthy People 2010. Washington, DC: US Dept of Health and Human Services; 2000.10. Phelan JC, Link BG, Diez-Roux AV, Kawachi I, Levan B. "Fundamental causes" of social inequalities in mortality: a test of the theory. J Health Soc Behav. 2004;45:265-285.
ISI
| PUBMED
11. Link BG, Phelan J. Social conditions as fundamental causes of disease. J Health Soc Behav. 1995(special issue):80-94.
FULL TEXT
| PUBMED
12. Dowda M, Ainsworth BE, Addy CL, Saunders R, Riner W. Environmental influences, physical activity, and weight status in 8- to 16-year-olds. Arch Pediatr Adolesc Med. 2001;155:711-717.
FREE FULL TEXT
13. French SA, Story M, Neumark-Sztainer D, Fulkerson JA, Hannan P. Fast food restaurant use among adolescents: associations with nutrient intake, food choices and behavioral and psychosocial variables. Int J Obes Relat Metab Disord. 2001;25:1823-1833.
FULL TEXT
|
ISI
| PUBMED
14. 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.
15. US Census Bureau. How the Census Bureau measures poverty. http://www.census.gov/hhes/poverty/povdef.html. Accessed September 6, 2005.16. Fisher GM. The development of the Orshansky Poverty Thresholds and their subsequent history as the official US poverty measure. http://www.census.gov/hhes/poverty/povmeas/papers/orshansky.html. Accessed September 6, 2005.17. US Census Bureau. Current Population Survey 2005 annual social and economic supplement. http://pubdb3.census.gov/macro/032005/pov/new35_000.htm. Accessed October 14, 2005.18. Gordon-Larsen P, McMurray RG, Popkin BM. Determinants of adolescent physical activity and inactivity patterns. Pediatrics. 2000;105:E83.
FULL TEXT
| PUBMED
19. Lee RE, Cubbin C. Neighborhood context and youth cardiovascular health behaviors. Am J Public Health. 2002;92:428-436.
FREE FULL TEXT
20. Guenther PM, Jensen HH, Batres-Marzues SP, Chen C-F. Sociodemographic, knowledge, and attitudinal factors related to meat consumption in the United States. J Am Diet Assoc. 2005;105:1266-1274.
PUBMED
21. National Center for Health Statistics. The NHANES 2002 MEC. In: Person Dietary Interviewers Procedure Manual. Hyattsville, Md: National Center for Health Statistics; 2000.22. National Center for Health Statistics. The NHANES 1999-2001 Dietary Interviewers Procedure Manual. Hyattsville, Md: National Center for Health Statistics; 2000.23. McDowell M, Briefel RR, Warren RA, Buzzard IM, Feskanich D, Gardner SN. The Dietary Data Collection System: an automated interview and coding system for NHANES III. Paper presented at: 14th National Nutrient Databank Conference; June 19-21, 1989; Ithaca, NY.24. Westat Inc. NHANES III dietary interviewer's manual. In: NHANES III Reference Manuals and Reports (CD-ROM). Hyattsville, Md: National Center for Health Statistics; 1992.25. USDA Food and Nutrient Database for Dietary Studies 1.0. In: U.S. Department of Agriculture Survey Nutrient Database System. Beltsville, Md: Agricultural Research Service of the Food Surveys Research Group; 2004.26. StataCorp. Stata Statisical Software: Release 8.0. College Station, Tex: StataCorp; 2004.27. Aiken LS, West SG. Multiple Regression: Testing and Interpreting Interactions. Thousand Oaks, Calif: Sage Publications; 1991.28. Cohen J, Cohen P. Applied Multiple Regression/Correlation Analysis for the Behaviorial Sciences. Hillsdale, NJ: Erlbaum Associates; 1983.29. Briefel RR, Johnson CL. Secular trends in dietary intake in the United States. Annu Rev Nutr. 2004;24:401-431.
FULL TEXT
|
ISI
| PUBMED
30. Chang VW, Lauderdale DS. Income disparities in body mass index and obesity in the United States, 1971-2002. Arch Intern Med. 2005;165:2122-2128.
FREE FULL TEXT
31. Zhang Q, Wang Y. Trends in the association between obesity and socioeconomic status in U.S. adults: 1971 to 2000. Obes Res. 2004;12:1622-1632.
ISI
| PUBMED
32. Gordon-Larsen P, Adair LS, Popkin BM. The relationship of ethnicity, socioeconomic factors, and overweight in U.S. adolescents. Obes Res. 2003;11:121-129.
ISI
| PUBMED
33. Demory-Luce D, Morales M, Nicklas T, Baranowski T, Zakeri I, Berenson G. Changes in food group consumption patterns from childhood to young adulthood: the Bogalusa Heart Study. J Am Diet Assoc. 2004;104:1684-1691.
FULL TEXT
| PUBMED
34. Janz KF, Dawson JD, Mahoney LT. Tracking physical fitness and physical activity from childhood to adolescence: the muscatine study. Med Sci Sports Exerc. 2000;32:1250-1257.
ISI
| PUBMED
35. Bertheke Post G, de Vente W, Kemper HC, Twisk JW. Longitudinal trends in and tracking of energy and nutrient intake over 20 years in a Dutch cohort of men and women between 13 and 33 years of age: the Amsterdam Growth and Health Longitudinal Study. Br J Nutr. 2001;85:375-385.
ISI
| PUBMED
36. Nelson MC, Gordon-Larsen P, Adair LS, Popkin BM. Adolescent physical activity and sedentary behavior: patterning and long-term maintenance. Am J Prev Med. 2005;28:259-266.
FULL TEXT
|
ISI
| PUBMED
37. Levin S, Lowry R, Brown DR, Dietz WH. Physical activity and body mass index among US adolescents. Arch Pediatr Adolesc Med. 2003;157:816-820.
FREE FULL TEXT
38. US Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, Ga: US Dept of Health and Human Services; 2004.39. Delva J, Johnston L, O’Malley PM. Obesity and Lifestyle Habits Among American Adolescents: A Study of SES, Gender, and Racial/Ethnic Differences 1986-2003. Ann Arbor, Mich: Robert Wood Johnson Foundation; 2005.40. Brownson RC, Boehmer TK, Luke DA. Declining rates of physical activity in the United States: what are the contributors? Annu Rev Public Health. 2005;26:421-443.
FULL TEXT
|
ISI
| PUBMED
41. Health Policy Tracking Service—A Thomson West Business. State actions to promote nutrition, increase physical activity, and prevent obesity: a legislative overview. http://www.rwjf.org/files/research/July%202005%20-%20Report.pdf. Accessed April 27, 2006.42. World Health Organization. Global Strategy on Diet, Physical Activity, and Health. Geneva, Switzerland: World Health Organization; 2004.43. Berkey CS, Rockett HR, Gillman MW, Field AE, Colditz GA. Longitudinal study of skipping breakfast and weight change in adolescents. Int J Obes Relat Metab Disord. 2003;27:1258-1266.
FULL TEXT
|
ISI
| PUBMED
44. Drewnowski A. Fat and sugar: an economic analysis. J Nutr. 2003;133:838S-840S.
FREE FULL TEXT
45. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr. 2004;79:6-16.
FREE FULL TEXT
46. Molnar BE, Gortmaker SL, Bull FC, Buka SL. Unsafe to play? neighborhood disorder and lack of safety predict reduced physical activity among urban children and adolescents. Am J Health Promot. 2004;18:378-386.
ISI
| PUBMED
47. Morland K, Wing S, Diez Roux A, Poole C. Neighborhood characteristics associated with the location of food stores and food service places. Am J Prev Med. 2002;22:23-29.
FULL TEXT
|
ISI
| PUBMED
48. Jahns L, Baturin A, Popkin BM. Obesity, diet, and poverty: trends in the Russian transition to market economy. Eur J Clin Nutr. 2003;57:1295-1302.
FULL TEXT
|
ISI
| PUBMED
49. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevelance of overweight and obesity among US children, adolescents, and adults, 1999-2002. JAMA. 2004;291:2847-2850.
FREE FULL TEXT
50. Goran MI, Shewchuk R, Gower BA, Nagy TR, Carpenter WH, Johnson RK. Longitudinal changes in fatness in white children: no effect of childhood energy expenditure. Am J Clin Nutr. 1998;67:309-316.
ABSTRACT
51. Braveman PA, Cubbin C, Egerter S, et al. Socioeconomic status in health research: one size does not fit all. JAMA. 2005;294:2879-2888.
FREE FULL TEXT
52. Nawaz H, Chan W, Abdulrahman M, Larson D, Katz DL. Self-reported weight and height: implications for obesity research. Am J Prev Med. 2001;20:294-298.
FULL TEXT
|
ISI
| PUBMED
53. Stewart AW, Jackson RT, Ford MA, Beaglehole R. Underestimation of relative weight by use of self-reported height and weight. Am J Epidemiol. 1987;125:122-126.
FREE FULL TEXT
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati
What's this?
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
|
Epi + demos + cracy: Linking Political Systems and Priorities to the Magnitude of Health Inequities--Evidence, Gaps, and a Research Agenda
Beckfield and Krieger
Epidemiol Rev 2009;0:mxp002v2-mxp002.
ABSTRACT
| FULL TEXT
'Ready. Set. ACTION!' A theater-based obesity prevention program for children: a feasibility study
Neumark-Sztainer et al.
Health Educ Res 2009;24:407-420.
ABSTRACT
| FULL TEXT
Preventive Care for Adolescents: Few Get Visits and Fewer Get Services
Irwin et al.
Pediatrics 2009;123:e565-e572.
ABSTRACT
| FULL TEXT
Children's Nutritional Health: Running on Empty?
Terre
AMERICAN JOURNAL OF LIFESTYLE MEDICINE 2009;3:115-118.
ABSTRACT
Changes in macronutrient intake among HIV-infected children between 1995 and 2004
Sharma et al.
Am. J. Clin. Nutr. 2008;88:384-391.
ABSTRACT
| FULL TEXT
Population-Based Prevention of Obesity: The Need for Comprehensive Promotion of Healthful Eating, Physical Activity, and Energy Balance: A Scientific Statement From American Heart Association Council on Epidemiology and Prevention, Interdisciplinary Committee for Prevention (Formerly the Expert Panel on Population and Prevention Science)
Kumanyika et al.
Circulation 2008;118:428-464.
ABSTRACT
| FULL TEXT
Breakfast Eating and Weight Change in a 5-Year Prospective Analysis of Adolescents: Project EAT (Eating Among Teens)
Timlin et al.
Pediatrics 2008;121:e638-e645.
ABSTRACT
| FULL TEXT
Assessment and Management of Patients with Obesity Hypoventilation Syndrome
Mokhlesi et al.
Proc Am Thorac Soc 2008;5:218-225.
ABSTRACT
| FULL TEXT
Expert Committee Recommendations Regarding the Prevention, Assessment, and Treatment of Child and Adolescent Overweight and Obesity: Summary Report
Barlow and and the Expert Committee
Pediatrics 2007;120:S164-S192.
ABSTRACT
| FULL TEXT
Why Do Children from Socioeconomically Disadvantaged Families Suffer from Poor Health When They Reach Adulthood? A Life-Course Study
Melchior et al.
Am J Epidemiol 2007;166:966-974.
ABSTRACT
| FULL TEXT
Recent Advances in Obesity Hypoventilation Syndrome
Mokhlesi and Tulaimat
Chest 2007;132:1322-1336.
ABSTRACT
| FULL TEXT
Using concentration index to study changes in socio-economic inequality of overweight among US adolescents between 1971 and 2002
Zhang and Wang
Int J Epidemiol 2007;36:916-925.
ABSTRACT
| FULL TEXT
The Obesity Epidemic in the United States--Gender, Age, Socioeconomic, Racial/Ethnic, and Geographic Characteristics: A Systematic Review and Meta-Regression Analysis
Wang and Beydoun
Epidemiol Rev 2007;0:mxm007v1.
ABSTRACT
| FULL TEXT
Obesity and public health law.
Koga et al.
NEJM 2006;355:1283-1283.
FULL TEXT
Increase in Overweight Status Among Poor Older Adolescents
JWatch Emergency Med. 2006;2006:8-8.
FULL TEXT
What's new in the other general journals.
Fister
BMJ 2006;332:1383-1384.
FULL TEXT
|
