Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Obesity (Silver Spring). Author manuscript; available in PMC 2010 November 1.
Published in final edited form as:
PMCID: PMC2861148

Television viewing is not predictive of Body Mass Index in Black and Hispanic young adult females


Previous studies have observed that television viewing is predictive of obesity and weight gain. We examined whether the cross-sectional association between television viewing and body mass index (BMI) varied by racial/ethnic subgroups among young women in Wave III (collected in 2001–2002) of the National Longitudinal Study of Adolescent Health. We used multivariate linear regression to examine the relationship between TV viewing and BMI among 6,049 females while controlling for socio-demographic and health attributes. We stratified the sample by race/ethnicity to better understand the association between TV viewing and BMI across different groups. Black and Hispanic females had higher BMIs (Black: 28.5 kg/m2, Hispanic: 27.3 kg/m2, White: 26.0kg/m2) than White females while Black females reported higher numbers of hours spent watching TV (Black: 14.7 hrs/wk, Hispanic: 10.6 hrs/wk, White: 11.2 hrs/wk) when compared to their White and Hispanic peers. TV viewing was positively associated with BMI (β=0.79, p=0.003 for 8–14 v. ≤7 hrs/wk; β=1.18, p=0.01 for >14 v. ≤ 7 hrs/wk) independent of race/ethnicity, age, maternal education, history of pregnancy, parental obesity, and household income. However, in models stratified by race/ethnicity, increased TV viewing was associated with increased BMI only among White females. TV viewing was not predictive of higher BMI in Black or Hispanic young adult females. Among Black and Hispanic females, counseling to decrease TV viewing may be important but insufficient for promoting weight loss.

Keywords: Obesity, TV viewing, race/ethnicity


Obesity in adolescents and young adults is a major public health problem with numerous consequences in adolescence and adulthood. During the past three decades the prevalence of obesity among adolescents has increased markedly, with recent estimates as high as 17% for adolescents and 28% for young adults (1, 2). There is evidence that being obese in childhood and adolescence at least doubles the risk of obesity in adulthood, (3) and some studies have shown that the most obese adolescents have a seventeen-fold greater chance of becoming obese adults when compared to their non-obese peers (4, 5). Moreover, obese adolescents are 30% more likely to die prematurely compared to their normal-weighted peers, although this is largely explained by their increased risk for adult obesity (5).

The burden of obesity is currently greater within certain racial and ethnic groups in the United States. Among females, this disparity is most pronounced in African-Americans and some Hispanic subpopulations (2, 68). In the most recent National Health and Nutrition Examination Survey (NHANES), Ogden et al. found non-Hispanic Black girls were twice as likely, and Mexican American girls were 1.5 times as likely, as non-Hispanic White girls to be obese (1). In another NHANES study, 50% of Black women ages 20–39 were obese, compared to 36% of Mexican American women and 24% of White women (2). There is evidence that this disparity starts early (9). Kimm, et al found an increase in the difference in median Body Mass Index (BMI) between White and Black girls from 0.4 to 2.3 kg/m2 during the ages of 9 and 19. (10) This race difference is particularly worrisome since Black girls who are obese in childhood and adolescence are more likely than obese White girls to remain obese as adults (9, 11).

Health behaviors, such as dietary intake, physical activity, and inactivity, have been a major area of focus in trying to understand the origins of obesity. Television viewing has been one of the health behaviors most consistently associated with obesity (12, 13). In their seminal paper linking television and obesity in children and adolescents, Dietz and Gortmaker note that this relationship may be mediated by a direct displacement of physical activity, as well as an increase in caloric consumption induced by food advertisements and snacking time (14). Although television viewing is often associated with lower levels of physical activity among girls (12, 13, 15, 16), it appears to have an independent relationship with weight status even after controlling for levels of physical activity (17). Furthermore, several studies show a stronger correlation between obesity and inactivity, including television viewing, than obesity and physical activity (1822). While the association between television viewing and weight status has been found in both genders, the relationship appears stronger and more consistent in females (17, 19, 20, 23) (18).

There is evidence that adolescents and young adults from racial and ethnic minority groups watch more television (7, 16, 18, 2426) and are less physically active(7, 25, 27, 28) than their white peers. Despite these consistent findings, several studies of pre-adolescent to young adult females have found that television viewing and weight are not statistically related in Black girls. (16, 25, 26, 28) In fact, the lack of a relationship between television viewing and BMI among Black girls may be one reason why some studies of television and obesity using racially heterogeneous populations have had null findings (16). The reasons why television and weight are not as consistently related in females from racial and ethnic minority groups as they are in White females is not well understood.

Prior studies of how race and ethnicity affect the relationship between television and obesity have often been small, localized to a few cities, and/or focused only on adolescents. The National Longitudinal Study of Adolescent Health (Add Health) offers a unique opportunity to add to the current literature on ethnic and racial differences in television viewing and obesity because it contains data from over 7,000 young women from a variety of race/ethnic groups surveyed both in adolescence and in young adulthood. The study includes a wealth of information on personal and socioeconomic backgrounds and health behaviors. Although Add Health has information about additional forms of inactivity, such as listening to the radio and/or playing computer games, we opted to focus exclusively on TV because of its potential to impact weight status through multiple pathways such as involving the participant in a sedentary behavior or inducing increased snacking through exposure to advertisments for calorie dense foods. We also felt that TV watching was less likely than other forms of media to be done concurrently with other activities (i.e. running while listening to a portable music player, playing an active video game such as Dance Dance Revolution).

We used Add Health to address the following hypotheses: 1) Increased television viewing is associated with higher average BMI in young adult women; and 2) The association between TV viewing and BMI will be different in White young adult females when compared to both Black and Hispanic young adult females.


Study population

This research uses data from the third of four Waves of the National Longitudinal Study of Adolescent Health, a nationally representative school-based study of adolescents enrolled in grades 7 through 12 at initial recruitment. We used two baseline variables from the first Wave to capture the socioeconomic status of the participants’ family of origin. Wave I data were collected in 1994–1995 and Wave III data in 2001–2002. Because our a priori hypotheses focused on differences between Hispanic, Black, and White females, we limited our sample to females who self-identified as White, non-Hispanic; Black, non-Hispanic; or Hispanic. We excluded those with height <4 feet or ≥7 feet, as well as those with weight ≤50 pounds or ≥500 pounds, due to concerns about plausibility. The distribution of reported television viewing was markedly skewed and thus we excluded those with self-reported television viewing above the 99.5th percentile (>80hours/week (n=39). We also excluded those who were currently pregnant (n=256) and/or disabled (n=298) because we felt that their television viewing and/or weight status might be influenced by factors different than those in the non-disabled, non-pregnant population.

In addition to the above exclusions, we excluded those who had missing data for either the dependent variable or for any of our key independent variables. However, prior to this final exclusion, we addressed a high non-response rate (approximately 10%) for the two variables from Wave I measuring socioeconomic status—parent reported maternal education and household income. In an effort to avoid selection bias and inaccurate inferences resulting from listwise deletion, we imputed these variables by best-subset regression (29, 30). After this imputation and all exclusions (total excluded=1053), our final sample contained 6,049 adolescent/young adult females.

Study Variables

Outcome variable

Body Mass Index (BMI) [weight (kg)/height (m2)] was our outcome variable. BMI was calculated from measured weight and height when available (95%) and from self-report in the small number of participants missing measured weight and height (5%).

Primary predictor variable

Our primary predictor variable of interest was self-reported television viewing. Participants were asked, “On average, how many hours a week do you spend watching television?” to which they responded with a number ranging from 0–168 hours/week. Because of the skewed distribution of responses we dropped those who were beyond the 99.5 percentile or 80 hours/week of reported TV viewing (n=39) We then modeled television as a categorical rather than continuous variable and grouped respondents into three categories: 0–7 hrs/week; 8–14 hrs/week; ≥14 hrs/week. We based our categories on an average number of hours per day extrapolated to total number of hours reported per week. We chose the highest category based on recommendations from professional organizations to limit TV viewing to less than 2 hours/day. Participants were also asked, “In the past 7 days, how many times did you watch television?” and responses ranged from 1 to 7. We used this variable for sensitivity analyses and found similar results to those using self-reported hours/week spent watching TV. Therefore, we present only findings using the report of hours/week of television viewing.

Additional independent variables

Race/ethnicity was constructed from two questions, one that asked participants to indicate if they were of Hispanic/Latino origin and a second that asked them to choose a category of race that best describes them. We constructed 6 categories: Hispanic, Black or African-American (not Hispanic), Asian/Pacific Islander, White (not Hispanic), Native American/American Indian, and Other. As described above, we limited our population to those who identified as Hispanic, White, or Black. We then constructed subpopulations within the Hispanic population based on responses to the question, “What is your Hispanic/Latino background?” (Mexican, Chicano/a, Cuban, Puerto Rican, Central/South American, and Other Hispanic). We combined Mexican-American and Chicano due to the small number who self-identified as Chicano. Due to additional small population numbers, we also combined Central/South American with Other Hispanic and anyone who chose more than one category.

We chose to examine measures of SES from Wave I rather than Wave III. The participants in Wave III ranged in age from 18–27. This age range represents a developmental period of incredible flux with regards to SES; the income reported during this time may be misleading because some people are in school (undergraduate or graduate), while others are in the workforce. Thus, we used the Wave I parental report of the highest grade of education achieved by the participant’s mother and the total household income over the last year. After performing sensitivity analyses we collapsed the maternal education variable into a dichotomous variable (less than college degree v. college graduate or beyond). We transformed the household income variable into a continuous measure that was a ratio of household income relative to the poverty level (taking into consideration the household size then comparing it to poverty thresholds in 1995, the year the data were collected). As mentioned above, responses for maternal education and household income were lower than other response rates because they relied on the parental responses and thus, we present findings using the imputed values.

We included additional variables that we theorized may impact weight status. These included history of pregnancy (yes/no), familial obesity (defined as having at least one obese parent yes/no), work status (having a job and/or being a student v. reporting neither having a job nor being a student), U.S. native (born in the U.S.), physical activity (the number of times in the last 7 days participated in some activity) and smoking status (current smoker v. nonsmoker).


In all models, we accounted for the complex survey design using svy commands in STATA and applied weights to account for the unequal likelihood of being sampled for certain subpopulations. We performed bivariate analyses of our covariates of interest with our outcome variable to test for significant relationships. In our multivariate regression models, we initially looked at relationships in our overall sample. Because of our a priori hypotheses, we then looked at models stratified by race/ethnicity. We performed additional sensitivity analyses within our racial/ethnic groups stratified by ever having been pregnant as well as by age ≤22 v. age ≥23. Regression diagnostic procedures showed no evidence of multicollinearity, heteroscedasticity, or substantial influence from outliers.


Black and Hispanic females had significantly higher BMIs compared to Whites (Black: 28.5 kg/m2; Hispanic: 27.3 kg/m2; White: 26.0kg/m2) (Figure 1) and were more likely to be overweight or obese (57.2% of Blacks, 53.4% of Hispanics were overweight or obese compared to 42.7% of Whites). Black females reported watching TV more hours/week than either Hispanics or Whites (15 hrs/week for Black females v. 11 for both Hispanic and White females) (Figure 2). There were also significant differences between the socioeconomic status of the different races/ethnicities with Whites having higher parent-reported household income than either Blacks or Hispanics and being more likely to have a mother who completed college (Table 1). Black females were nearly twice as likely as White females to report ever having been pregnant. All three groups had high levels of participation in either work or school.

Figure 1
Average BMI by Race/Ethnicity
Figure 2
Amount of television viewed (hours per week) by Race/Ethnicity
Table 1
Demographic, familial and health factors by race/ethnicity

The results for our multivariate models with the 3 racial/ethnic groups (Blacks, Hispanics, and Whites) combined and with the subpopulations of Hispanics separated are shown in Table 2. In this model, BMI increased with greater time spent watching TV. Those who watch 8–14 hours/week had BMIs that were on average 0.8 kg/m2 higher and those who watch more than 14 hours/week had BMIs that were on average 1.2 kg/m2 higher than those who watched 7 hours or less/week. In this overall model, Black, Mexican-American, and Puerto Rican females were all significantly heavier than their white peers. Having at least one obese parent was associated with a 3.7kg/m2 increase in BMI relative to those without an obese parent.

Table 2
Characteristics and behaviors associated with BMI among 6049 young adult females in Add Health

We next stratified our results by race/ethnicity (Table 3). Among white females we observed a similar relationship between TV viewing and BMI as in the overall population. However, among both Black and Hispanic subpopulations we found no association between TV viewing and BMI. In an additional model, we examined the effect of TV viewing across subpopulations of Hispanics and found no difference in the magnitude of the association (p>0.05 for all interaction terms of Hispanic subpopulations × TV viewing; data not shown).

Table 3
Associations between TV viewing and BMI across racial/ethnic subpopulations

Because pregnancy rates varied by race/ethnicity, in a secondary analyses we assessed the impact of pregnancy on the relationship between TV viewing and BMI across racial/ethnic groups. We did not observe a significant interaction between history of pregnancy and TV viewing in any of the racial/ethnic groups. Moreover, models stratified on history of pregnancy did not yield materially different results. We also examined the effect of US nativity, smoking and physical activity in our models and found no association (data not shown).


This study demonstrates that increased television viewing is associated with increased body mass index in a nationally representative sample of young women, but that the association differs by racial and ethnic group. Among non-Hispanic Whites, after controlling for age, parental BMI, maternal education, income, pregnancy history, and education or job status, there was a significant association between television viewing and BMI. However, this study did not find an association between television viewing and BMI in Black or Hispanic women either before or after controlling for known and suspected confounders. This lack of an association is consistent with findings from earlier studies of younger participants (16, 25, 26, 28). Given that Blacks watched more hours of television than the Whites and Hispanics in the study, one possible explanation would be that the majority of Blacks are beyond some threshold effect for television exposure. Although mean hours of television viewing were higher in Blacks than in other racial/ethnic groups in this study, there was still considerable variation in viewing hours among Blacks and the distribution of their viewing hours overlapped considerably with those of Whites and Hispanics. Therefore, a threshold effect is unlikely to be the reason that an association was not identified.

Another possibility is that Blacks interact with television media differently than whites. While shows popular with Blacks have been documented as having more overweight characters and more food commercials,(31) it is possible that Blacks are less susceptible to the effects of television advertising than Whites. Although there is some self-reported data suggesting that Blacks are more likely than Whites to eat while watching television,(32) little is known about what individuals are actually doing while the television is on. Some have hypothesized that because the television is more likely to be on throughout the day in Black homes, reported television viewing hours may not actually be watched with full attention and therefore may have less of an impact. (26)

A third possibility is that television is too minor a factor in the lives of Black Americans relative to other “obesogenic” aspects of their environment. Blacks and Hispanics are more likely than Whites to live in neighborhoods with fewer supermarkets, fewer parks, more fast food establishments, higher crime, inadequate recreational areas, to be from single-parented, lower income households, and to skip breakfast.(33) Lower socioeconomic status is consistently associated with higher rates of obesity and this relationship is likely mediated by many of the factors noted above. A study by Singh and colleagues which did find an association between television viewing and BMI in Blacks noted that this association was stronger in those from higher socioeconomic groups, (23) perhaps because fewer other negative factors were present. Maternal education, a marker for socioeconomic status, was indeed significantly associated with BMI in the model for Blacks in our study, although television viewing was not.

This study also did not find an association between television viewing and BMI among Hispanic females. Hispanic participants have not been included in studies as often as non-Hispanic Blacks and Whites and thus the relationship between television viewing and BMI in Hispanic females is less clear. Results from prior studies of television viewing and BMI among nationally representative and ethnically diverse samples have been inconsistent. One reason for the inconsistent findings among Hispanics may be their ethnic diversity and levels of acculturation, which lead to different health beliefs, behaviors, and outcomes.(34) However, when we subdivided the Hispanics in this study by country of origin, we still found no relationship between television viewing and obesity. This is surprising given the heterogeneity of the Hispanic population, but may be related to the small numbers of individuals in each subgroup and the lack of adequate power to find differences among them.

There are several important limitations to this study. First, this is a cross-sectional study and therefore a causal relationship between television viewing and obesity in young women cannot be demonstrated by this analysis. It is possible that those young women who were already overweight or obese preferred to watch television because they were physically less able to be active. It is also possible that increased television viewing is a marker of another factor associated with obesity that we did not measure, such as depression. However, while the majority of studies examining the relationship between television viewing and obesity have been cross-sectional, a few longitudinal studies have demonstrated a link between early television use and obesity later in adolescence and young adulthood (19, 20, 35), suggesting that there is indeed a causal relationship.

Second, this study used self-reported measures of television viewing. Results are therefore subject to social desirability and social approval bias given that some individuals may have chosen to under-report their sedentary behaviors. Indeed, individuals with high scores on instruments measuring social desirability have been shown to over-report physical activity(36) and under-report calorie intake(37) when compared to objective measures of these behaviors. Furthermore, there is evidence that Blacks and Mexican Americans score higher on social desirability scales and use more extreme categories on Likert scales as compared to Whites.(38) Under-reporting television viewing would be expected to bias results toward the null, so it is possible that under-reporting of viewing time by Blacks and Hispanics contributed to our inability to find an association in these groups.

In conclusion, television viewing is related to body mass index in a diverse sample of young adult women, but the relationship is complex and inconsistent across racial/ethnic subgroups. We believe providers should continue to recommend limiting television viewing when counseling young women on healthy habits, given that television viewing is a modifiable behavior while other factors associated with BMI such as socioeconomic status are less quickly and easily changed. But they should understand that women from certain racial and ethnic minority groups may derive less of a benefit on weight status from reductions in television viewing time. Policy initiatives and public education campaigns aimed at changing the content of food advertising on television may also be expected to yield mixed results. More work is needed to identify the factors driving the obesity epidemic in young Black and Hispanic women, as these groups are particularly impacted by obesity.


Tracy Richmond was supported by the Charles H. Hood Foundation Child Health Research Award and the National Institutes of Health Career Development Award (1K01HD058042-01A1).


1. Ogden C, Caroll M, Flegal K. High body mass index for age among US children and adolescents, 2003–2006. Journal of the American Medical Association. 2008;299(20):2401–2405. [PubMed]
2. Ogden C, Caroll M, Curtin L, McDowell M, Tabak C, Flegal K. Prevalence of overweight and obesity in the United States. JAMA. 2006;295(13):1549–1555. [PubMed]
3. Singh A, Mulder C, Twisk J, Chinapaw M. Tracking of childhood overweight into adulthood: a systematic review of the literature. Obesity Reviews. 2008 Sep;:474–488. [PubMed]
4. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. The New England Journal of Medicine. 1997;337(13):869–873. [PubMed]
5. Engeland A, Bjorge T, Tverdal A, Sogaard AJ. Obesity in adolescence and adulthood and the risk of adult mortality. Epidemiology. 2004;15(1):79–85. [PubMed]
6. Hedley A, Ogden C, Johnson C, Caroll M, Curtin L, Flegal K. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA. 2004;291(23):2847–2840. [PubMed]
7. Delva J, O'Malley PM, Johnston LD. Racial/ethnic and socioeconomic status differences in overweight and health-related behaviors among American students: national trends 1986–2003. Journal of Adolescent Health. 2006;39:536–554. [PubMed]
8. Delva J, Johnston LD, O'Malley PM. The epidemiology of overweight and related lifestyle behaviors: racial/ethnic and socioeconomic status differences among American youth. American Journal of Preventative Medicine. 2007;33:S178–S186. [PubMed]
9. Freedman DS, Khan LK, Serdula MK, Dietz WH, Srinivasan SR, Berenson GS. Racial differences in the tracking of childhood BMI to adulthood. Obesity Research. 2005;13(5):928–935. [PubMed]
10. Kimm SY, Barton BA, Obarzanek K, McMahon RP, Sabry ZI, Waclawiw MA, et al. Racial divergence in adiposity during adolescence: the NHLBI Growth and Health Study. Pediatrics. 2001;107(3):1–7. [PubMed]
11. Gordon-Larsen P, Adair LS, Nelson MC, Popkin BM. Five-year obesity incidence in the transition period between adolescence and adulthood: the National Longitudinal Study of Adolescent Health. American Journal of Clinical Nutrition. 2004;80:569–575. [PubMed]
12. Hancox RJ, Milne BJ, Poulton R. Association between child and adolescent television viewing and adult health: a longitudinal birth cohort study. The Lancet. 2004;364:257–262. [PubMed]
13. Marshall S, Biddle S, Gorely T, Cameron N, Murdey I. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. International Journal of Obesity. 2004;28:1238–1246. [PubMed]
14. Dietz WH, Gortmaker SL. Do we fatten our children at the television set? Obesity and television viewing in children and adolescents. Pediatrics. 1985;75(5):807–812. [PubMed]
15. Robinson TN, Hammer LD, Killen JD, Kraemer HC, Wilson DM, Hayward C, et al. Does television viewing increase obesity and reduce physical activity? Cross-sectional and longitudinal analyses among adolescent girls. Pediatrics. 1993;81(2):273–280. [PubMed]
16. Lowry R, Wechsler H, Galuska DA, Fulton JE, Kann L. Television viewing and its associations with overweight, sedentary lifestyle, and insufficient consumption of fruits and vegetables among US high school students: differences by race, ethnicity, and gender. Journal of School Health. 2002;72(10):413–421. [PubMed]
17. Eisenmann JC, Bartee R, Smith D, Welk GJ, Fu Q. Combined influence of physical activity and television viewing on the risk of overweight in US youth. International Journal of Obesity. 2008;32:613–618. [PubMed]
18. Crespo CJ, Smit E, Troiano RP, Bartlett SJ, Macera CA, Andersen RE. Television watching, energy intake, and obesity in US children. Archives of Pediatrics and Adolescent Medicine. 2001;135:360–365. [PubMed]
19. Hancox R, Poulton R. Watching television in associated with childhood obesity: but is it clinically important? International Journal of Obesity. 2005;30:171–175. [PubMed]
20. Boone JE, Gordon-Larsen P, Adair LS, Popkin BM. Screen time and physical activity during adolescence: longitudinal effects on obesity in young adulthood. International Journal of Behavioral Nutrition and Physical Activity. 2007;4(26):5868. [PMC free article] [PubMed]
21. Mark AE, Janssen I. Relationship between screen time and metabolic syndrome in adolescents. Journal of Public Health. 2008;30(2):153–160. [PubMed]
22. Vicente-Rodriquez G, Rey-Lopez JP, Martin-Matillas M, Moreno LA, Wamberg J, Redondo C, et al. Television watching, videogames, and excess of body fat in Spanish adolescents: the AVENA study. Nutrition. 2008;24:654–662. [PubMed]
23. Singh GK, Kogan MD, Van Dyck PC, Siapush M. Racial/ethnic, socioeconomic, and behavioral determinants of childhood and adolescent obesity in the United States: analyzing independent and joint associations. Annals of Epidemiology. 2008;18:682–695. [PubMed]
24. Nelson MC, Gordon-Larsen P, Adair LS, Popkin BM. Adolescent physical activity and sedentary behavior: patterning and long-term maintenance. American Journal of Preventative Medicine. 2005;28(3):259–266. [PubMed]
25. Hanson MD, Chen E. Socioeconomic status, race, and body mass index: the mediating role of physical activity and sedentary behaviors during adolescence. Journal of Pediatric Psychology. 2007;32(3):250–259. [PubMed]
26. Henderson VR. Longitudinal associations between television viewing and body mass index among white and black girls. Journal of Adolescent Health. 2007;41:544–550. [PMC free article] [PubMed]
27. Gordon-Larsen P, McMurray RG, Popkin BM. Adolescent physical actvitiy and inactivity vary by ethnicity: the National Longitudinal Study of Adolescent Health. The Journal of Pediatrics. 1999;135:301–306. [PubMed]
28. Gordon-Larsen P, Adair LS, Popkin BM. Ethnic differences in physical activity and inactivity patterns and overweight status. Obesity Research. 2002;10(3):141–149. [PubMed]
29. King G, Honaker J, Joseph A, Scheve K. List-wise deletion is evil: what to do about missing data in political science. Annual Meetings of the American Political Science Association; 1998; Philadelphia. 1998.
30. impute. SShf. Stata help for impute. In.
31. Tirodkar MA, Jain A. Food messages on African American television shows. American Journal of Public Health. 2003;93(3):439–441. [PubMed]
32. Mcnutt SW, Hu Y, Schreiber GB, Crawford PB, Obarzanek E, Mellin L. A longitudinal study of the dietary practices of black and white girls 9 and 10 years old at enrollment: the NHLBI Growth and Health Study. Journal of Adolescent Health. 1997;20:27–37. [PubMed]
33. Kumanyika S, Grier S. Targeting inverventions for ethnic minority and low income populations. The Future of Children. 2006;16(1):187–207. [PubMed]
34. Hajat A, Lucas J, Kington R. Health outcomes among Hispanic subgroups: data from the National Health Interview Survey, 1992–1995. Health Data. 2000;25(310):1–14. [PubMed]
35. Davison KK, Marshall SJ, Birch LL. Cross-sectional and longitudinal associations between TV viewing and girls' body mass index, overweight status, and percentage of body fat. Journal of Pediatrics. 2006;149(1):32–37. [PMC free article] [PubMed]
36. Adams SA, Matthews CE, Ebbeling CB, Moore CG, Cunningham JE, Fulton J, et al. The effect of social desirability and social approval on self-reports of physical activity. American Journal of Epidemiology. 2005;161(4):389–398. [PMC free article] [PubMed]
37. Herbert JR, Ma Y, Clemow L, Ockene IS, Saperia G, Stanek E, et al. Gender differenceces in social desirability and social approval bias in dietary self-report. American Journal of Epidemiology. 1997;146(12):1046–1055. [PubMed]
38. Warnecke RB, Johnson TP, Chavez N, Sudman S, O'Rourke DP, Lacey L, et al. Improving question wording in surveys of culturally diverse populations. Annals of Epidemiology. 1997;7:334–342. [PubMed]