This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).
Data on the national and state levels is often used to inform policy decisions and strategies designed to reduce racial disparities in obesity. Obesityrelated health outcomes are realized on the individual level, and policies based on state and nationallevel data may be inappropriate due to the variations in health outcomes within and between states. To examine countylevel variation of obesity within states, we use a smallarea analysis technique to fill the void for countylevel obesity data by race. Five years of Behavioral Risk Factor Surveillance System data are used to estimate the prevalence of obesity by county, both overall and racestratified. A modified weighting system is used based on demographics at the county level using 2010 census data. We fit a multilevel reweighted regression model to obtain countylevel prevalence estimates by race. We compare the distribution of prevalence estimates of nonHispanic Blacks to nonHispanic Whites. For 25 of the 26 states included in our analysis there is a statistically significant difference between withinstate countylevel average obesity prevalence rates for nonHispanic Whites and nonHispanic Blacks. This study provides information needed to target disparities interventions and resources to the local areas with greatest need; it also identifies the necessity of doing so.
Obesity is known to drastically increase the risk of chronic diseases and is associated with excess morbidity and mortality [
The policies and strategies to reduce racial health disparities are often implemented on the national and state levels, informed by national and statelevel data. These levels are often too far removed from the individual level where health outcomes are realized; a shift in focus to the local level may be necessary to accelerate progress in reducing racial health disparities. The lack of lowerlevel data hinders the effective evaluation of public health policy, programs, and interventions that occur at the local level [
Survey data from the Behavioral Risk Factor Surveillance System (BRFSS) is commonly used to estimate the prevalence of chronic disease. The BRFSS telephone survey collects data on preventative health practices and risk behaviors for chronic diseases in adults. The data is collected in a uniform manner across all states [
In 2010, when stratified by race, 37% of nonHispanic Black adults in the United States were obese, as compared to 26% of nonHispanic Whites [
We combine five years (2006–2010) of BRFSS survey data to obtain the prevalence estimates for obesity, using the following two survey questions: “About how much do you weigh without shoes?” and “About how tall are you without shoes?”. If necessary, we convert the reported weight to kilograms and the reported height to meters. Body mass index (BMI) is calculated as:
To determine obesity status, we used the National Institute of Health recommendation of a BMI greater than or equal to 30 as the cutoff [
Study sample schema.
Analysis was completed using SAS/STAT version 9.3 (SAS Institute Inc., Cary, NC, USA). In a previous study, it was established that multilevel logistic regression models show the least amount of discrepancy when estimating countylevel prevalence of chronic disease by race [
In Equation (2),
After calculating the regression parameter estimates, we estimated the countylevel prevalence rates by race. The countylevel agebyracebysex estimated prevalence is calculated from the regression model predictors, Equation (3):
The countylevel prevalence rates by race are then calculated with the following Formula (4):
Here,
We compared the distribution of prevalence estimates of nonHispanic Blacks to nonHispanic Whites. After determining that the samples were normally distributed, we performed a twosample test for proportions to compare the mean difference in estimated countylevel prevalence of obesity and an Ftest to examine the equality of the variances from the two distributions. We calculated the betweenstate variance in obesity prevalence using 2006–2010 BRFSS data. We calculated two variances, one using the 26 states included in this analysis and a second using all 50 states and the District of Columbia. We then compared the variance of obesity prevalence withinstates to these betweenstate variance estimates and subsequently the racestratified obesity prevalence estimates to the betweenstate estimates in order to examine the difference in variability within states as compared to between states. Here, for racestratified variance we calculate the countylevel prevalence estimate for the nonHispanic White population and subsequently for the nonHispanic Black population and include both estimates for each county within each state to produce a “racestratified withinstate” variance estimate.
After exclusion criteria, 806,154 individual respondents were included in the analysis (
Study sample.
N  Mean  Median  SD  Minimum  Maximum  


806,154  
Observations per county  1,512.5  922.0  1,904.4  83.0  16,463.0  

533  
Counties per state  20.5  12.0  17.0  5.0  57.0 
Obesity prevalence rates were calculated by race (nonHispanicBlack and nonHispanic White) for each county. All 26 states had a higher mean prevalence rate of obesity for nonHispanic Blacks than nonHispanic Whites. A ttest was used to examine the racial differences in prevalence rates of obesity for states with normal distributions of countylevel prevalence rates. Five of the 26 states had nonnormal distributions of countylevel prevalence rates. Wilcoxon signedrank test was used in place of a ttest for these states. An Ftest was conducted to compare variances; for all states included in the analysis, there is no evidence of unequal variances. For 25 of the 26 states, nonHispanic Blacks had a statistically significantly higher mean prevalence of obesity as compared to nonHispanic Whites (
Withinstate variances for obesity prevalence rates were then calculated using countylevel estimates. We compared these variances to the betweenstate variance in obesity prevalence rates. States with less than five counties reporting data were excluded; 26 states (51%) are included in the analysis of withinstate variability. We compared this withinstate variability to both the betweenstate variance calculated using the 26 states included in this analysis and the variance calculated using all 50 states and the District of Columbia. Fifteen out of 26 states (58%) included in the analysis had a greater withinstate variance estimates of obesity prevalence rates than the betweenstate variance for these 26 states (10.35, 95%CI 6.36–19.72). Sixteen out of 26 states (62%) included in the analysis had a greater withinstate variability of obesity prevalence rates than the betweenstate variability for all 50 states and the District of Columbia (9.23, 95%CI: 6.49, 14.32). Two of these are statistically significantly different with nonoverlapping confidence bounds. The withinstate variability of obesity prevalence rates and 95% confidence limits are shown in
Comparison of racestratified prevalence means by state.
White Prevalence of Obesity (%)  Black Prevalence of Obesity (%)  TTest  

State  N  Mean  SD  Mean  SD  T Statistic  
Alabama  34  24.46  4.62  31.60  4.83  6.23  <0.0001 
Arkansas  12  23.17  3.90  29.22  3.72  3.89  0.0008 
California  14  18.71  3.77  25.14  4.16  4.28  0.0002 
Colorado  5  15.69  1.98  22.07  2.25  7.50 *  0.0625 
Florida  57  21.22  3.87  27.58  4.73  7.86  <0.0001 
Georgia  28  18.58  3.08  25.53  3.53  7.85  <0.0001 
Illinois  8  18.17  1.38  24.61  2.37  6.63  <0.0001 
Indiana  5  20.49  1.89  28.17  1.78  6.63  0.0002 
Kansas  6  19.61  3.47  26.93  3.81  3.48  0.0059 
Kentucky  7  22.00  2.89  29.73  2.38  14.00 *  0.0156 
Louisiana  43  22.40  3.74  28.94  4.29  473.00 *  <0.0001 
Maryland  21  21.46  4.33  28.09  4.63  115.50 *  <.0001 
Massachusetts  8  17.43  2.46  24.42  2.68  5.42  <.0001 
Michigan  11  21.56  3.33  29.35  3.46  5.38  <0.0001 
Mississippi  56  23.57  3.13  29.77  3.38  10.06  <0.0001 
New Jersey  20  18.57  2.97  25.60  2.72  7.82  <0.0001 
New York  12  17.07  3.55  23.97  2.88  5.23  <0.0001 
North Carolina  55  22.62  3.85  30.13  4.04  9.99  <0.0001 
Ohio  10  21.54  2.16  29.61  2.26  8.17  <0.0001 
Oklahoma  7  22.25  2.10  30.45  2.46  6.72  <0.0001 
Pennsylvania  8  20.43  3.21  27.23  3.35  4.15  0.0010 
South Carolina  43  22.95  3.75  30.37  3.94  8.94  <0.0001 
Tennessee  12  21.45  2.43  28.23  1.84  7.70  <0.0001 
Texas  20  20.25  3.68  27.48  4.06  5.90  <0.0001 
Virginia  24  18.30  4.56  25.97  5.80  5.09  <0.0001 
Washington  7  20.50  1.89  27.54  2.38  14.00 *  0.0156 
Note: * Obesity Prevalence Rates had nonnormal distributions. Sstatistics were calculated in place of Tstatistics using Wilcoxon signedrank test.
The lower variability estimates of states, such as Ohio, indicate that the obesity prevalence rates for each county within the state are similar. In these cases, a statewide intervention would be more appropriate than in a state such as Virginia where there is greater variability of obesity prevalence rates. Virginia, and other states with greater withinstate variability, has a wide variety of obesity prevalence rates among counties. Therefore, local targeted interventions would be the most useful in reducing obesity in these states.
WithinState variability of obesity prevalence rates compared to betweenstate variability.
The forest plot above shows the variance point estimates and 95% confidence intervals for all 26 states included in the analysis. The red squares represent the variance estimates while the horizontal red lines represent the confidence intervals for each state variance estimate. The blue diamond represents the betweenstate variance for all 50 states and the District of Columbia while the vertical, gray dotted lines represent the 95% confidence interval for this betweenstate variance. To the right, the column labeled “VAR” provides the numeric variance estimates for each state while the columns labeled “LCL” and “UCL” represent the lower and upper confidence limits, respectively, for the variance estimates.
Withinstate variance was then calculated again for each state with racespecific prevalence rates included for each county. This racestratified withinstate variability aims to quantify the racial disparity among obesity rates in each state. The calculated racestratified withinstate variances were higher than both the unstratified withinstate variance and the overall betweenstate variance for all 26 states included in the analysis (
Since withinstate variance distributions are normal for both races, a two sample ttest was used to compare mean variance between the two groups. The mean racestratified withinstate variance (24.22) is significantly higher than that of the unstratified withinstate variance (11.40) (
Racestratified withinstate variability of obesity prevalence rates.
Obesity prevalence rates on the state or national level are potentially misleading and overlook the information provided by a more local approach. Smallarea analysis of publically available data is a costeffective and useful way to produce locallevel prevalence estimates. Over half of the states included in our analysis have greater variability in obesity prevalence between their counties than the overall variability in obesity prevalence between states. Perhaps states are not an appropriate subunit to examine obesity prevalence in the United States. Counties with high prevalence rates might be overlooked because they are in a state with an overall low prevalence. For example, Virginia has the largest amount of countylevel variability and yet ranked 29th in overall obesity prevalence. By giving Virginia less attention due to its relatively low prevalence, Virginia counties with higher obesity prevalence may not receive the resources they need. When stratifying by race, the estimates are even more variable, further necessitating the use of racestratified countylevel data.
One limitation of this study is the nature of the available data. Due to the BRFSS sampling and the response rates by race, many counties had less than 30 nonHispanic Black respondents, and therefore could not be included in our analysis. This potentially led to a misestimation of withinstate variance. In order to control for this, we excluded states with less than 5 counties and included 95% confidence limits for the variance estimates. Another potential limitation is the assumption that selfreport bias does not vary across data levels. A recent study has shown that there is a geographic pattern in selfreported height and weight, causing estimates and rankings to be misleading [
The rising prevalence of obesity has major fiscal implications. In 1999, obesity attributed to 9.4% of the nation’s total health care expenditures [
The work of Melody Goodman and members of the Goodman Lab (Lucy D’AgostinoMcGowan, Renee Gennarelli, Sarah Lyons) is supported by the BarnesJewish Hospital Foundation, Siteman Cancer Center, National Institutes of Health, National Cancer Institute grant U54CA153460, Washington University School of Medicine (WUSM) and WUSM Faculty Diversity Scholars Program.
The authors declare no conflict of interest.