| | Neighborhood Deprivation Is Associated with Lower Levels of Serum Carotenoids among Adults Participating in the Third National Health and Nutrition Examination SurveyAbstract ObjectiveThis study tested the hypothesis that neighborhood deprivation will be associated with lower levels of serum carotenoids in comparison with wealthy residential areas. DesignCross-sectional, nationally representative survey data were used to assess the relationship between neighborhood level socioeconomic status and serum carotenoids. SubjectsSeventeen thousand two participants aged 17 years and older from the Third National Health and Nutrition Examination Survey were linked with 1990 census data. Main outcome measuresSerum levels of lycopene, β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. Statistical analysisMultivariate linear regression was used to model the association of serum carotenoids and neighborhood deprivation, which is a summary index of 11 indicators for tract level socioeconomic status. Adjustments are made for individual level age, sex, years of education, household income, employment, race/ethnicity, body mass index, serum cotinine, alcohol use, physical activity, and serum cholesterol. ResultsMultivariate analysis revealed a negative and statistically significant association between high levels of neighborhood deprivation and β-carotene (β=−2.98 μg/dL [−0.06 μmol/L], P=0.00), α-carotene (β=−1.28 μg/dL [−0.02 μmol/L], P=<0.0001), lutein/zeaxanthin (−1.69 μg/dL [−0.03 μmol/L], P=0.00, β-cryptoxanthin (β=−1.34 μg/dL [−0.02 μmol/L], P<0.0001), and total carotenoids (β=−8.20 μg/dL, P=<0.0001). Lycopene was not related to neighborhood deprivation. Adjusted mean levels of carotenoids for high deprivation neighborhoods were lower than neighborhoods with low deprivation: β-carotene=8.72 μg/dL [0.16 μmol/L] vs 20.64 μg/dL [0.38 μmol/L], α-carotene=0.44 μg/dL [0.008 μmol/L] vs 5.56 μg/dL [0.10 μmol/L], lutein/zeaxanthin=13.79 μg/dL [0.24 μmol/L] vs 20.55 μg/dL [0.36 μmol/L], β-cryptoxanthin=4.57 μg/dL [0.08 μmol/L] vs 9.93 μg/dL [0.18 μmol/L], lycopene=22.07 μg/dL [0.41 μmol/L] vs 25.63 μg/dL [0.48 μmol/L], and total=49.56 μg/dL vs 82.36 μg/dL. ConclusionsNeighborhood deprivation was associated with lower serum levels of carotenoids. There was a substantial disparity between low deprivation and high deprivation residential areas with respect to fruit and vegetable intake. Several studies have revealed an association between poor and minority residential environments and higher prevalence of death and illness (1, 2). Nutrition may be one mechanism that is in part responsible for spatial patterns of morbidity and mortality (3, 4). Fruit and vegetable consumption has received attention in recent years for its potential role in chronic disease prevention through the antioxidant action of carotenoids (5, 6, 7, 8). At the individual level, there is some evidence that socioeconomic status is related to poor nutrition. Observational studies have demonstrated that individuals of lower socioeconomic status (SES) have poorer diet quality (9, 10, 11, 12), higher rates of food insufficiency (13), and higher obesity rates (12, 14) than individuals of higher SES, which may contribute to these health disparities. Research focusing specifically on fruit and vegetable consumption has also shown disadvantage among those of lower SES. For example, education may be positively associated with fruit and vegetable consumption (11, 15, 16). However, other factors that may play a role in the relationship between SES and fruit and vegetable consumption such as residential environment have not received much attention. The availability and cost of quality food in residential environments may be a mechanism leading to poor nutrition (17). Residents of poor and minority communities do not have sufficient access to supermarkets that have a wide variety of food choices at lower prices, particularly fresh produce (18, 19, 20, 21). Furthermore, some authors found a positive association between neighborhood deprivation and the mean number of fast-food restaurants per 1,000 people (22). Although it seems plausible that the structure of access to healthful food in poor and minority neighborhoods may be related to poor nutrition, few studies have analyzed spatial patterns of nutrition behavior and these studies have used self-reported food frequency questionnaires to assess the nutritional profile of residents (4, 23, 24). No study has been able to utilize objective indicators of dietary behavior like serum levels of carotenoids because contextual information linked with these data have not been available. Recently, the Third National Health and Nutrition Examination Survey (NHANES III) was geocoded to census tracts using 1990 census data. An advantage of using contextual linked NHANES data is the availability of nationally representative data on objective blood markers of nutrition with sufficient variation at the neighborhood level to capture a mixture of residential environments with high and low SES. Serum carotenoid levels correlate significantly with intake of dietary carotenoids and studies have demonstrated that serum carotenoid concentrations can adequately serve as biomarkers of fruit and vegetable consumption (25, 26, 27, 28, 29). Evaluating fruit and vegetable consumption by use of serum biomarkers may serve as a way to bypass biases that are associated with self-reported dietary recall of fruit and vegetable intake. Therefore, this study used NHANES III data linked with census tracts to examine the association of neighborhood deprivation and five serum carotenoids: lycopene, β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. We hypothesized that neighborhood deprivation would be associated with lower levels of serum carotenoids in comparison to wealthy residential areas. Methods  Sample Data files from the 1990 US Census were merged with NHANES III to form a contextual file that included information at both the census tract and individual level. NHANES III is a nationally representative sample of 39,695 community-dwelling respondents from 1988 through 1994. The survey protocol for NHANES includes an extensive survey instrument covering prevalent medical conditions, diet, health-related behaviors, a medical examination, and laboratory assays of blood and other physical specimens. Further information about the sample design, operations, and indicators have been published elsewhere (30). Information on Census tracts is not available in the public version of NHANES. Investigators can apply for special permission from the National Center for Health Statistics to link contextual data with NHANES. The Research Data Center at the National Center for Health Statistics merged our Census data file, which contained only tract level data, with individual level data in NHANES by using a tract identifier, and then provided remote access to the merged analytic file. After matching the data from the survey, laboratory, and examination files and excluding respondents aged 0 to 16 years or those who did not have laboratory data for carotenoids, the final analysis sample consisted of 17,002 respondents aged 17 years and older. Dependent Variables Serum carotenoids are well established as biomarkers for fruit and vegetable intake (25, 26, 27, 28, 29). Mobile examination centers collected blood samples and analyzed the samples for a large range of nutrition markers, including five carotenoids: lycopene, β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. Reversed-phase high-performance liquid chromatography was used to quantify carotenoid levels (31, 32). This procedure does not differentiate between lutein and zeaxanthin. Therefore, the combination of these carotenoids is presented as a single value. Carotenoid levels below detection limits were set to zero and each carotenoid was measured continuously (micrograms per deciliter). Independent Variable Neighborhood deprivation measures SES within census tracts by using 11 indicators found in the 1990 US Census: percent low education (<9 years), percent high education (college educated), percent managerial/professional occupations of all with reported occupations, median family income, ratio of income for highest and lowest quintiles, median house value, median gross rent, percent of persons in labor force who are unemployed, percent of persons for whom poverty status is determined who live in households with income ≤100% of the federal poverty level by census definition, percent of households with no telephone, percent of household with no plumbing (31). The sum of these 11 indicators formed an additive index that was then divided into four equal categories of deprivation with higher values indicating more deprivation. National Center for Health Statistics staff were unable to geocode all addresses to each case, which resulted in 2,433 missing cases, which tended to be rural, nonwhite, and have low education and incomes. Control Variables Several demographic- and individual-level SES variables are included as covariates. Age is categorically coded as 17 to 45 years, 46 to 64 years, and 65 to 90 years. Sex (men as the reference category) and race/ethnicity (non-Hispanic white as the reference category) are also included as demographic covariates. Individual-level SES is captured by three measures: years of education (<12 years=1), household income (<$20,000=1), and employment status (employed=1). Health behaviors are also controlled using four measures. Smoking status is indicated by serum levels of cotinine ≥14.1 ng/mL (80.09 nmol/L). Current alcohol use and number of physical activities (zero activities in the past month=1) are both self-reported indicators. Body mass index is converted into two categories: 0 to 29 and ≥30. Total cholesterol is divided into two categories; serum levels <240 mg/dL (6.24 mmol/L) is used as the reference group. Statistical Analysis The first analysis presents descriptive statistics for serum levels of lycopene, β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. Estimates for the association of neighborhood deprivation on each serum carotenoid are derived using multivariate linear regression and include controls for individual level demographic, SES, and health behaviors. Multilevel modeling is not recommended for this study design. Multilevel modeling is an excellent analytic strategy for nested data where multiple members of the sample live in the same census tract. However, our data are not nested because there are only about two respondents per tract. Some scholars have argued that use of multilevel modeling in this situation may lead to incorrect estimates and instead suggest that analysts apply the appropriate survey weights and account for design effects (32, 33). Data that were missing for the dependent variable were excluded from the analysis yielding an analysis file with 17,002 observations. Multiple imputation procedure using five iterations of the Markov Chain Monte Carlo method was used to impute missing data for neighborhood deprivation, education, income, body mass index, serum cotinine, physical activity, and serum cholesterol. All models were adjusted for sample weights, population sampling units, and strata and estimated with the Statistical Analysis System using the SURVEYREG and PROC MI procedures (34). Results Table 1 presents descriptive statistics for serum carotenoids by study variables. Mean levels of carotenoids decrease with increasing levels of neighborhood deprivation. The control variables follow expected patterns. Overall, higher levels of carotenoids are found among adults aged 65 years and older, women, non-Hispanic whites, employed, nonsmokers, former/never consumers of alcohol, physically active, nonobese, high cholesterol, incomes >$20,000, those with ≥12 years of education. In most cases lycopene followed the opposite pattern. | | |  | | n | Lycopenea | β-Carotenea | α-Carotenea | Lutein/zeaxanthinb | β-Cryptoxanthinc | Total | |
|---|
| | | ←mean±SD→ | | | Area level deprivation | | | | | | | | | | 1st Quartile (low) | 3,118 | 23.19±11.21 | 23.66±22.84 | 6.13±6.71 | 24.07±11.68 | 11.09±8.21 | 88.14±41.37 | | | 2nd Quartile | 3,089 | 22.99±11.47 | 19.39±19.21 | 4.63±4.75 | 22.37±11.85 | 10.44±8.04 | 79.83±37.29 | | | 3rd Quartile | 3,005 | 21.70±11.20 | 18.42±19.26 | 4.20±4.28 | 21.83±11.16 | 10.30±7.63 | 76.45±36.54 | | | 4th Quartile (high) | 5,357 | 20.25±11.08 | 18.23±19.07 | 3.90±4.72 | 23.14±14.04 | 11.04±8.90 | 76.55±38.73 | | | Missing | 2,433 | 20.27±11.61 | 19.34±20.07 | 3.83±4.32 | 22.11±14.29 | 8.84±6.85 | 74.39±39.06 | | | Age | | | | | | | | | | 17-45 y | 9,134 | 24.30±10.81 | 15.50±14.61 | 3.92±4.97 | 20.92±10.60 | 10.46±7.81 | 75.09±33.21 | | | 46-64 y | 3,724 | 20.90±11.54 | 21.09±20.38 | 4.96±5.22 | 24.64±14.56 | 10.69±8.89 | 82.27±41.80 | | | 65-90 y | 4,144 | 16.05±10.17 | 27.42±26.73 | 5.31±5.10 | 25.26±14.83 | 10.40±8.24 | 84.44±46.26 | | | Sex | | | | | | | | | | Female | 9,025 | 21.02±10.94 | 21.78±21.25 | 4.94±5.41 | 22.58±12.69 | 10.80±8.36 | 81.13±40.43 | | | Male | 7,977 | 22.13±11.76 | 17.20±18.43 | 3.97±4.66 | 23.03±12.97 | 10.15±7.91 | 76.48±36.97 | | | Education | | | | | | | | | | <12 y | 7,185 | 23.47±11.44 | 19.91±20.41 | 4.71±5.26 | 22.23±11.25 | 10.03±7.20 | 77.00±40.28 | | | ≥12 y | 9,704 | 18.97±10.69 | 19.24±19.75 | 4.17±4.84 | 23.51±14.64 | 11.11±9.27 | 80.35±37.85 | | | Missing | 113 | 19.81±10.91 | 20.49±15.28 | 5.29±5.17 | 24.86±13.76 | 11.09±7.26 | 81.54±36.39 | | | Household income | | | | | | | | | | <$20,000 | 7,399 | 19.76±11.05 | 18.86±20.41 | 4.10±5.40 | 22.54±13.68 | 10.36±8.33 | 75.63±40.05 | | | ≥$20,000 | 7,972 | 23.52±11.31 | 20.24±20.11 | 4.81±4.79 | 22.86±11.98 | 10.49±7.82 | 81.92±37.76 | | | Missing | 1,631 | 20.01±11.26 | 20.11±18.50 | 4.66±5.01 | 23.58±12.76 | 11.08±8.96 | 79.44±38.09 | | | Employment status | | | | | | | | | | Employed | 9,437 | 23.86±11.22 | 17.91±17.88 | 4.38±5.17 | 22.29±11.49 | 10.43±7.77 | 78.87±36.43 | | | Unemployed | 7,565 | 18.65±10.83 | 21.77±22.39 | 4.61±5.00 | 23.42±14.30 | 10.58±8.62 | 79.04±41.81 | | | Race | | | | | | | | | | Non-Hispanic white | 7,036 | 21.49±11.53 | 21.59±20.69 | 4.83±4.57 | 21.38±11.81 | 8.74±6.07 | 78.02±37.75 | | | African American | 4,617 | 22.29±12.11 | 19.42±22.47 | 3.61±6.15 | 24.31±13.71 | 8.93±6.03 | 78.58±41.51 | | | Hispanic | 4,665 | 20.99±10.16 | 16.47±15.52 | 4.36±3.59 | 22.81±12.56 | 14.55±10.54 | 79.19±36.40 | | | Other | 684 | 20.84±11.49 | 22.38±21.95 | 7.72±8.42 | 26.89±16.02 | 11.37±10.61 | 89.20±46.98 | | | Body mass index | | | | | | | | | | <30 | 11,981 | 21.82±11.38 | 20.84±21.30 | 4.76±5.34 | 23.09±12.72 | 10.80±8.33 | 81.30±39.95 | | | ≥30 | 4,167 | 21.00±11.28 | 16.11±15.84 | 3.69±4.22 | 21.85±12.98 | 9.53±7.46 | 72.18±34.92 | | | Missing | 854 | 20.33±10.96 | 19.83±19.46 | 4.57±5.15 | 23.22±13.26 | 10.91±8.59 | 78.86±38.83 | | | Serum cotinine (ng/mL)d | | | | | | | | | | <14.1 | 11,690 | 22.02±11.24 | 21.50±21.23 | 5.11±5.55 | 23.72±13.04 | 11.61±8.70 | 83.95±40.21 | | | ≥14.1 | 4,782 | 21.00±11.45 | 14.54±15.60 | 2.89±3.13 | 20.58±12.04 | 7.88±6.00 | 66.89±32.10 | | | Missing | 530 | 16.02±11.04 | 24.30±22.29 | 5.19±5.71 | 22.30±12.58 | 9.57±7.33 | 77.38±42.02 | | | Alcohol use | | | | | | | | | | Current | 7,327 | 23.45±11.23 | 16.79±17.43 | 4.12±4.65 | 22.17±11.49 | 10.13±7.73 | 76.65±35.60 | | | Former/never | 9,675 | 20.10±11.22 | 21.78±21.67 | 4.77±5.39 | 23.26±13.73 | 10.77±8.46 | 80.69±41.17 | | | Physical activity (past month) | | | | | | | | | | 0 | 11,757 | 22.76±11.35 | 19.61±20.29 | 4.51±4.94 | 22.62±12.59 | 10.36±7.76 | 76.90±39.90 | | | ≥1 | 5,236 | 18.84±10.85 | 19.67±19.69 | 4.43±5.44 | 23.17±13.33 | 10.79±8.98 | 79.86±38.44 | | | Missing | 9 | 13.67±4.30 | 18.56±10.77 | 3.67±2.06 | 24.22±9.42 | 12.67±10.21 | 72.78±24.55 | | | Serum cholesterol (mg/dL)e | | | | | | | | | | <240 | 13,651 | 20.81±10.81 | 17.98±18.24 | 4.22±4.91 | 21.23±10.47 | 9.96±7.41 | 74.20±34.68 | | | ≥240 | 3,328 | 24.56±13.30 | 26.39±25.31 | 5.58±5.68 | 29.22±18.38 | 12.72±10.42 | 98.46±48.14 | | | Missing | 23 | 22.96±13.03 | 17.22±20.17 | 3.30±3.95 | 19.04±13.55 | 8.39±4.03 | 70.91±41.71 | | | | |
| a To convert μg/dL lycopene, β-carotene, and α-carotene to μmol/L, multiply μg/dL by 0.01863. bTo convert μg/dL lutein/zeaxanthin to μmol/L, multiply μg/dL by 0.01758. cTo convert μg/dL β-cryptoxanthin to μmol/L, multiply μg/dL by 0.01809. dTo convert ng/mL serum cotinine to nmol/L, multiply ng/mL by 5.68. eTo convert mg/dL serum cholesterol to mmol/L, multiply mg/dL by 0.026. |
Results from the multivariate linear regression for serum carotenoids are presented in Table 2 and Table 3. The analyses used poststratification weights and standard errors and statistical tests were calculated with adjustments for the effects of the complex sample design. The multivariate models explained 13% to 17% of the variation for serum carotenoids. The relationship between neighborhood deprivation and serum carotenoids is statistically significant except for lycopene although the pattern of results for lycopene is consistent with the other carotenoids. Each quartile of neighborhood deprivation is negatively associated with serum carotenoids. However, the pattern of results does not suggest a dose–response relationship because the effect size does not uniformly increase as the level of deprivation increases. Rather, the results may be suggestive of a threshold effect because residents of low deprivation neighborhoods appear to have significantly better levels of serum carotenoids than residents of high deprivation neighborhoods. Total carotenoids were analyzed by forming an additive index from the five carotenoids as a means for testing total fruit and vegetable intake. The pattern of results, shown in Table 3, is similar to β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. | | | | | Total Carotenoids | |
|---|
| | βb±SEc | P value | |
|---|
| Area level deprivation | | | | | 1st Quartile (low)d | | | | | 2nd Quartile | −6.77±1.53 | <0.0001 | | | 3rd Quartile | −8.91±1.64 | <0.0001 | | | 4th Quartile (high) | −8.20±1.88 | <0.0001 | | | Age | | | | | 17-45 yd | | | | | 46-64 y | 6.47±1.38 | <0.0001 | | | 65-90 y | 10.58±1.50 | <0.0001 | | | Sex | | | | | Maled | | | | | Female | 4.65±0.74 | <0.0001 | | | Education | | | | | ≥12 yd | | | | | <12 y | −3.45±0.96 | 0.00 | | | Race | | | | | Non-Hispanic whited | | | | | African American | 8.82±1.57 | <0.0001 | | | Hispanic | 11.10±1.61 | <0.0001 | | | Other | 11.10±2.33 | <0.0001 | | | Employment status | | | | | Unemployedd | | | | | Employed | 2.73±1.05 | 0.01 | | | Household income | | | | | ≥$20,000d | | | | | <$20,000 | −5.21±0.95 | <0.0001 | | | Body mass index | | | | | <30d | | | | | ≥30 | −12.72±0.99 | <0.0001 | | | Serum cotininee | | | | | <14.1 ng/mLd | | | | | ≥14.1 ng/mL | −14.59±0.88 | <0.0001 | | | Alcohol use | | | | | Former/neverd | | | | | Current | −0.92±0.99 | 0.36 | | | Physical activity | | | | | ≥1 | | | | | 0 | −6.36±0.98 | <0.0001 | | | Serum cholesterolf | | | | | <240 mg/dLd | | | | | ≥240 mg/dL | 20.79±1.34 | <0.0001 | | | Intercept | 82.36±1.89 | <0.0001 | | | R2 | 0.17 | | | | |
| a Total carotenoids is a summary index of five carotenoids: lycopene, β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. Models are adjusted for sample weights, strata, and population sampling units. bUnstandardized regression coefficient. cSE=standard error. dReference category. eTo convert ng/mL serum cotinine to nmol/L, multiply ng/mL by 5.68. fTo convert mg/dL serum cholesterol to mmol/L, multiply mg/dL by 0.026. |
Figure 1 graphically depicts the predicted mean level of carotenoids with adjustment for individual level characteristics for low and high deprivation neighborhoods. There is a significant disparity between low and high deprivation neighborhoods with regard to an objective marker of fruit and vegetable consumption. Adjusted mean levels of carotenoids for high deprivation neighborhoods were lower than neighborhoods with low deprivation: β-carotene=8.72 μg/dL [0.16 μmol/L] vs 20.64 μg/dL [0.38 μmol/L], α-carotene=0.44 μg/dL [0.008 μmol/L] vs 5.56 μg/dL [0.10 μmol/L], lutein/zeaxanthin=13.79 μg/dL [0.24 μmol/L] vs 20.55 μg/dL [0.36 μmol/L], β-cryptoxanthin=4.57 μg/dL [0.08 μmol/L] vs 9.93 μg/dL [0.18 μmol/L], lycopene=22.07 μg/dL [0.41 μmol/L] vs 25.63 μg/dL [0.48 μmol/L], and total=49.56 μg/dL vs 82.36 μg/dL. Discussion This study examined the association of neighborhood deprivation and serum carotenoids using data from NHANES III linked with census tracts from the 1990 US Census. After adjustment for individual level demographic factors, SES, and health behaviors, neighborhood deprivation was negatively associated with four of the five carotenoids: β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin. Disparities in carotenoids were striking between low and high deprivation residential neighborhoods, which suggests that the nutritional profile of low-income areas may be more carotenoid deficient than high-income areas. In particular, the findings suggest that the economic effect of residential areas on nutrition may be interpreted as a difference between wealthy areas and residential districts characterized by any level of deprivation, including middle-class communities. This study was the first we know of to analyze the neighborhood distribution of an objective marker for fruit and vegetable intake. Previous studies that used self-reported data on food intake found mixed evidence for an association between low-income areas and poor nutrition (4, 23, 24). Other reports have produced impressive findings regarding the distribution of food stores and restaurants, leading to the conclusion that low-income areas may have disproportionately more fast-food restaurants and be extremely disadvantaged with respect to access to supermarkets, which generally have a wider variety of quality foods at lower prices (17, 18, 19, 20, 21, 22). The authors of these reports argued that poor access to quality food was a mechanism for the clustering of illness and death in low-income residential areas. However, the evidence base for this assertion was underdeveloped because it was unclear if low access translates into poor nutrition as measured through low levels of objective biomarkers for fruit and vegetable consumption. The findings of this report seem to support the assertion that low-income neighborhoods have poor nutrition and the likely mechanism is the structural conditions of food access. In this study neighborhood deprivation was significantly associated with serum carotenoid levels, with the exception of lycopene. Studies that have explored the correlation between serum carotenoid levels and fruit/vegetable intake have consistently found that serum carotenoids are highly correlated with dietary intake from fruit/vegetables, with the exception of lycopene (26, 35, 36, 37). Lycopene does not appear to be as widely distributed among varieties of fruits and vegetables as other carotenoids, with its major source being tomatoes (38). Dietary lycopene has been shown, in US samples, to be primarily obtained from food items made with tomato-based products, such as pizza and pasta sauce, and not from consumption of raw tomatoes themselves (39, 40). Furthermore, the bioavailability of lycopene found in tomato-based foods may be higher than the bioavailability from raw tomatoes (41, 42, 43). Often, investigation of fruit and vegetable consumption does not account for tomato consumption in mixed dishes such as pizza, pasta, or tomato-based soups. Therefore, serum lycopene may not be as directly related to fruit/vegetable consumption as other carotenoids, because it is measured by dietary recall surveys. Serum lycopene levels may be more adequate than other carotenoids in deprived neighborhoods, not because of fruit/vegetable consumption, but because of consumption of other foods that contain tomato-based products. This may explain why serum lycopene concentrations were not significantly different by neighborhood deprivation. Although there are few studies that have examined the sociodemographic correlates of serum carotenoids, our results are consistent with other findings for education and ethnicity. For example, previous research using data from NHANES III (35) found that mean lycopene and lutein/zeaxanthin were highest for African Americans, mean β-carotene was highest for whites, and mean β-cryptoxanthin was highest for Mexican Americans. α-Carotene was highest for Mexican Americans in this sample, but the value was not statistically different than that of whites. One limitation of this study is the use of cross-sectional data. Although we acknowledge that longitudinal data would be better able to rule out that the association we found was not simply an aggregation of individuals with similar eating preferences, a strength of these data are the scope of the sample to provide sufficient variation with respect to both the independent and dependent variables. Our use of census tracts to describe neighborhood deprivation may be a limitation because these areas generally contain about 4,000 people, which could lead to a mixture of socially advantaged individuals with social disadvantaged individuals. Future research could benefit from better defined neighborhoods although this would require substantial resources. Overall, we found evidence that neighborhood level deprivation was associated with lower serum levels of β-carotene, α-carotene, lutein/zeaxanthin, and β-cryptoxanthin, and that there is a substantial disparity between low SES and high SES residential areas with respect to fruit and vegetable intake. Given mounting evidence that this disparity may be related to the economic conditions of residential areas, policymakers and public health officials should target this issue as a means to address the clustering of morbidity and mortality in low-income and minority communities. This research was supported by the National Institute of Aging (grant no. T32AG000270) and the National Cancer Institute (grant no. 1 P50 CA105631-02). References 1. 1Diez Roux AV, Merkin SS, Arnett D, Chambless L, Massing M, Nieto FJ, et al. Neighborhood of residence and incidence of coronary heart disease. N Engl J Med. 2001;345:99–106. MEDLINE |
CrossRef
2. 2Pickett KE, Pearl M. Multilevel analyses of neighborhood socioeconomic context and health outcomes: A critical review. J Epidemiol Community Health. 2001;55:111–122.
CrossRef
3. 3Morland K, Diez Roux AV, Wing S. Supermarkets, other food stores, and obesity: The atherosclerosis risk in communities study. Am J Prev Med. 2006;30:333–339. Abstract | Full Text |
Full-Text PDF (95 KB)
|
CrossRef
4. 4Diez Roux AV, Nieto FJ, Caulfield L, Tyroller HA, Watson RL, Szklo M. Neighbourhood differences in diet: The atherosclerosis risk in communities (ARIC) study. J Epidemiol Community Health. 1999;53:55–63.
CrossRef
5. 5Pavia M, Pileggi C, Nobile C, Angelillo IF. Association between fruit and vegetable consumption and oral cancer: A meta-analysis of observational studies. Am J Clin Nutr. 2006;83:1126–1134. MEDLINE 6. 6Dauchet L, Amouyel P, Dallongeville J. Fruit and vegetable consumption and risk of stroke: A meta-analysis of cohort studies. Neurology. 2005;65:1193–1197.
CrossRef
7. 7Coyne T, Ibiebele TI, Baade PD, Dobson A, McClintock C, Dunn S, et al. Diabetes mellitus and serum carotenoids: Findings of a population-based study in Queensland, Australia. Am J Clin Nutr. 2005;82:685–693. MEDLINE 8. 8Conner SL, Ojeda LS, Sexton G, Weidner G, Conner WE. Diets lower in folic acid and carotenoids are associated with the coronary disease epidemic in Central and Eastern Europe. J Am Diet Assoc. 2004;104:1793–1799. Abstract | Full Text |
Full-Text PDF (121 KB)
|
CrossRef
9. 9Leiyu S. Sociodemographic characteristics and individual health behaviors. South Med J. 1998;91:933–941. MEDLINE 10. 10Dubois L, Girard M. Social position and nutrition: A gradiant relationship in Canada and the USA. Eur J Clin Nutr. 2001;55:366–373. MEDLINE |
CrossRef
11. 11Xie B, Gilliland FD, Yu-Fen L, Rockett H. Effects of ethnicity, income, and education on dietary intake among adolescents. Prev Med. 2003;36:30–40. MEDLINE |
CrossRef
12. 12Neumark-Sztainer D, Story M, Hannan PJ, Croll J. Overweight status and eating patterns among adolescents: Where do youths stand in comparison with the Healthy People 2010 objectives?. Am J Public Health. 2002;92:844–851. MEDLINE |
CrossRef
13. 13Alaimo K, Briefel RR, Frongillo EA, Olson CM. Food insufficiency exists in the United States: Results from NHANES III. Am J Public Health. 1998;88:419–426. MEDLINE |
CrossRef
14. 14Drewnowski A, Specter SE. Poverty and obesity: The role of energy density and energy costs. Am J Clin Nutr. 2004;79:6–16. MEDLINE 15. 15Trudeau E, Kristal A, Li S, Patterson R. Demographic and psychosocial aspects of fruit and vegetable intakes differ: Implications for dietary interventions. J Am Diet Assoc. 1998;98:1412–1417. Abstract | Full Text |
Full-Text PDF (702 KB)
|
CrossRef
16. 16Thompson B, Demark-Wahnefried W, Taylor G, McClelland JW, Stables G, Havas S, et al. Baseline fruit and vegetable intake among adults in seven 5-A-Day study centers located in diverse geographic areas. J Am Diet Assoc. 1999;99:1241–1248. Abstract | Full Text |
Full-Text PDF (824 KB)
|
CrossRef
17. 17Jetter KM, Cassady DL. The availability and cost of healthier food alternatives. Am J Prev Med. 2006;30:38–44. Abstract | Full Text |
Full-Text PDF (85 KB)
|
CrossRef
18. 18Algert SJ, Agrawal A, Lewis DS. Disparities in access to fresh produce in low-income neighborhoods in Los Angeles. Am J Prev Med. 2006;30:365–370. Abstract | Full Text |
Full-Text PDF (736 KB)
|
CrossRef
19. 19Morland 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. Abstract | Full Text |
Full-Text PDF (76 KB)
|
CrossRef
20. 20Moore LV, Diez Roux AV. Associations of neighborhood characteristics with the location and type of food stores. Am J Public Health. 2006;96:325–331. MEDLINE |
CrossRef
21. 21Morland K, Wing S, Diez Roux AV. The contextual effect of the local food environment on residents diets: The atherosclerosis risk in communities study. Am J Public Health. 2002;92:1761–1767. MEDLINE |
CrossRef
22. 22Cummins SCJ, McKay L, MacIntyre S. McDonald’s restaurants and neighborhood deprivation in Scotland and England. Am J Prev Med. 2005;29:308–310. Abstract | Full Text |
Full-Text PDF (54 KB)
|
CrossRef
23. 23Diehr P, Koepsell T, Cheadle A, Psaty BM, Wagner E, Curry S. Do communities differ in health behaviors?. J Clin Epidemiol. 1993;46:1141–1149. MEDLINE |
CrossRef
24. 24Ecob R, Macintyre S. Small area variations in health related behaviours: Do these depend on the behaviour itself, its measurement, or on personal characteristics?. Health Place. 2000;6:261–274. MEDLINE |
CrossRef
25. 25Al-Delaimy WK, Silmani N, Ferrari P, Key T, Spencer E, Johansson I, et al. Plasma carotenoids as biormarkers of intake of fruits and vegetables: Ecological-level correlations in the European Prospective Investigation into Cancer and Nutrition (EPIC). Eur J Clin Nutr. 2005;59:1397–1408. MEDLINE 26. 26Coyne T, Ibiebele TI, McNaughton S, Rutishauser IH, O’Dea K, Hodge AM, et al. Evaluation of brief dietary questions to estimate vegetable and fruit consumption using serum carotenoids and red-cell folate. Public Health Nutr. 2005;8:298–308. MEDLINE 27. 27Michaud DS, Giovannucci EL, Ascherio A, Rimm EB, Forman MR, Sampson L, et al. Associations of plasma carotenoid concentrations and dietary intake of specific carotenoids in samples of two prospective cohort studies using a new carotenoid database. Cancer Epidemiol Biomarkers Prev. 1998;7:283–290. MEDLINE 28. 28Campbell DR, Gross MD, Martini MFC, Grandits GA, Slavin JL, Potter JD. Plasma carotenoids as biomarkers of vegetable and fruit intake. Cancer Epidemiol Biomarkers Prev. 1994;3:493–500. MEDLINE 29. 29Block G, Norkus E, Hudes M, Mandel S, Helzlsouer K. Which plasma antioxidants are most related to fruit and vegetable consumption?. Am J Epidemiol. 2001;154:1113–1118. MEDLINE |
CrossRef
30. 30National Center for Health Statistics. Plan and operation of the Third National Health and Nutrition Examination Survey, 1988-94. Vital Health Stat. 1994;1:27. 31. 31Gunter EW, Lewis BG, Koncikowski SM. Laboratory procedures used for the Third National Health and Nutrition Survey (NHANES III), 1998-1994. Atlanta, GA: National Center for Environmental Health, Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services; 1996;. 32. 32Sowell AL, Huff DL, Yeager PR, Caudill SP, Gunter EW. Retinol, α-tocopherol, lutein/zeaxanthin, β-cryptoxanthin, lycopene, alpha carotene, trans-beta carotene, and four retinyl esters in serum determined simultaneously by reversed-phase HPLC with multiwavelength detection. Clin Chem. 1994;40:411–416. MEDLINE 33. 33Singh GK, Miller BA, Hankey BF, Feuer EJ, Pickle LW. Changing area socioeconomic patterns in US cancer mortality, 1950-1998: Part I-all cancers among men. J Natl Cancer Inst. 2002;94:904–915. MEDLINE 34. 34Winkleby MA, Cubbin C. Influence of individual and neighbourhood socioeconomic status on mortality among black, Mexican-American, and white women and men in the United States. J Epidemiol Community Health. 2003;57:444–452.
CrossRef
35. 35Duncan C, Jones K, Moon G. Context, composition and heterogeneity: Using multilevel models in health research. Soc Sci Med. 1998;46:97–117. MEDLINE |
CrossRef
36. 36SAS Procedures Guide. Cary, NC: SAS Institute Inc; 2004;. 37. 37Ford ES, Gillepsie C, Ballew C, Sowell A, Mannino DM. Serum carotenoid concentrations in US children and adolescents. Am J Clin Nutr. 2002;76:818–827. MEDLINE 38. 38Campbell DR, Gross MD, Martini MC, Grandits GA, Slavin JL, Potter JD. Plasma carotenoids as biomarkers of vegetable and fruit intake. Cancer Epidemiol Biomarkers Prev. 1994;3:493–500. MEDLINE 39. 39van Kappel AL, Steghens JP, Zeleniuch-Jacquotte A, Chajes V, Toniolo P, Riboli E. Serum carotenoids as biomarkers of fruit and vegetable consumption in the New York Women’s Health Study. Public Health Nutr. 2001;4:829–835. MEDLINE |
CrossRef
40. 40Mangels AR, Holden JM, Beecher GR, Forman MR, Lanza E. Carotenoid content of fruit and vegetables: An evaluation of analytic data. J Am Diet Assoc. 1993;93:284–296. Abstract | Full Text |
Full-Text PDF (1248 KB)
|
CrossRef
41. 41Ganji V, Kafai MR. Population determinants of serum lycopene concentrations in the United States: Data from the third National Health and Nutrition Examination Survey, 1988-1994. J Nutr. 2005;135:567–572. MEDLINE 42. 42Mayne ST, Cartmel B, Silva F, Kim CS, Fallon BG, Briskin K, et al. Plasma lycopene concentrations in humans are determined by lycopene intake, plasma cholesterol concentrations and selected demographic factors. J Nutr. 1999;129:849–854. MEDLINE 43. 43Dewanto V, Wu X, Adom KK, Liu RH. Thermal processing enhances the nutritional value of tomatoes by increasing total antioxidant activity. J Agric Food Chem. 2002;50:3010–3014. MEDLINE |
CrossRef
J. P. Stimpson is an assistant professor, Department of Social and Behavioral Sciences, University of North Texas Health Science Center, Fort Worth; at the time of the study, he was a postdoctoral fellow, Sealy Center on Aging, University of Texas Medical Branch, Galveston. A. C. Nash is a doctoral student, Department of Preventive Medicine and Community Health, H. Ju is a biostatistician, and K. Eschbach is an associate professor, Department of Demography and Organization Studies, University of Texas at San Antonio. At the time of the study, K. Eschbach was associate professor, Department of Internal Medicine, University of Texas Medical Branch, Galveston.  Address correspondence to: Jim P. Stimpson, PhD, Department of Social and Behavioral Sciences, University of North Texas Health Science Center, 3500 Camp Bowie Blvd, Fort Worth, TX 76107-2699.
PII: S0002-8223(07)01626-4 doi:10.1016/j.jada.2007.08.016 © 2007 American Dietetic Association. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT01626-4/journalimage?img=PIIS0002822307016264.gr1.lrg.gif&fig=fig1&kwhquery=&issn=0002-8223&ishighres=false&allhighres=false&free=yes','journalimage');)
