McCullough, Mary Beth MA; Marks, Amy K. PhD
Childhood and adolescent obesity is widely recognized as a significant and growing public health concern in the United States. The overall rate of childhood and adolescent obesity has nearly tripled over the last 30 years and is estimated at approximately 18%.1 Although obesity in adolescents in the United States seems to be leveling off, particularly among females, federal initiatives aimed to decrease rates of obesity in the United States have not been met.2 The high rates of obesity among US youth, particularly among adolescents, are undoubtedly a major concern because of the range of adverse obesity-related health outcomes. Compared with children and adolescents of healthy weight, youth who struggle with excess weight are at a heightened risk for deleterious health outcomes (e.g., orthopedic complications, Type 2 diabetes, hypertension, asthma, sleep apnea) and psychosocial concerns (e.g., lower health-related quality of life, social isolation, victimization, psychological distress, poorer educational outcomes3,4). Furthermore, studies have revealed that adolescent-onset obesity is a stronger predictor of obesity and associated adverse health conditions in adulthood than both child- and adult-onset obesity.5
Not only are adolescents as a whole at-risk for obesity, but certain groups of adolescents have been recognized as more vulnerable than others. Particularly concerning is the rapid rise of obesity among second- and third-generation immigrant adolescents in the United States. Research has shown that second-generation (US-born children with one or both immigrant parents) and third-generation (US-born children with both US-born parents) immigrant youth are demonstrating higher rates of overweight and obesity than first-generation youth (foreign-born children with both immigrant parents)6; a finding referred to as the “immigrant paradox”.7 Since most traditional assimilation models would predict better outcomes as immigrants spend more time in the United States (e.g., increased access to resources, English language proficiency),8 the results showing that first-generation immigrants are doing better than later generations is referred to as a paradox.9 That is, despite having fewer economic and social resources, first-generation immigrant adolescents suffer lower rates of obesity than their later-generation peers. This theoretical orientation of the paradox does not discount the importance of economic and social correlates of health for the well-being of immigrants. On the contrary, poverty, lack of structural and material resources, and low education/low-wage job rates are just as concerning as risk factors for newcomers as they are for US-born youth. Instead, the paradox posits that, even after accounting for such economic and social risk factors, first-generation immigrant youth seem to be faring better than their circumstances would predict, which may provide insight into how to better serve second- and third-generation immigrant communities.10 The paradox can therefore be understood as a population-level pattern important to understand and explain, because it may provide key insights into some of the protective cultural or social practices that may be lost because of acculturation in immigrant communities.
This pattern of results underscores the importance of examining the physical and mental health of immigrant families in the United States. Immigrant youth are the fastest growing population of children, and are leading the ethnic and racial transformation that is occurring in the United States.11 The population of children and adolescents in immigrant families has grown nearly 7 times faster than the population of children of US-born parents.11 In fact, as of 2005, nearly one fourth (23%) of children lived in immigrant families, with the majority of these children (79%) born in the United States.11 These numbers are only expected to increase. The Census Bureau projects for 2030 that 54% of children in the United States will live in immigrant families.11
To better understand the physical health of this growing population, many researchers have conducted studies that focus specifically on immigrant generation status and body mass index (BMI).6,7,12 These studies are difficult to interpret because they often use inconsistent methodology and have produced conflicting results. For example, although some studies show that Latino youth are experiencing higher levels of obesity as they spend more time in the United States, other research reveals that obesity rates among Latino youth are similar or even decreasing across generations.7,12 Generational differences in obesity also vary across subpopulations of the “same” ethnic group. Although Cuban and Mexican-American children are both classified as Latino youth, obesity rates are increasing for Cuban immigrants and, in some studies, decreasing or staying the same across generations for Mexican immigrants.6 It is unclear whether this variation in obesity rates is due to inconsistency in measurement across studies or true differences in rates of obesity across generations for some ethnic minority immigrant groups. This highlights the importance of using standard measures across studies and continuing to examine this trend among immigrant groups.
Most researchers have used acculturation13 to explain these findings, such that immigrants lose the values specific to their native culture and fully adopt the behaviors, beliefs, practices, and values of the United States (e.g., increased fast-food consumption, involvement in sedentary activities7). This explanation is limited because it is merely descriptive at the population level, and does not specify the individual behaviors or mechanisms that contribute to elevated rates of obesity among later generations of immigrant youth. Moreover, acculturation does not fully explain why some ethnic minority groups seem to be “protected” from this pattern of increased obesity rates. To gain a comprehensive understanding of this concerning health outcome, we need to focus on identifying specific mechanisms that are driving the relation between BMI and generation status.
Therefore, in this study, we provide more specificity for this concerning health trend by investigating the possible mediating role of several health behaviors on the relation between generation status and BMI. Based on research showing that exposure to obesogenic environments in the United States fosters unhealthy behaviors among immigrant children,14 we propose that sedentary behaviors, fast-food consumption, and low rates of physical activity may serve to explain the obesity immigrant paradox. We focus on Latino youth, because this is an immigrant group that has demonstrated high rates of obesity as well as high levels of sedentary behaviors, physical inactivity, and fast-food consumption.7 We are also hoping to resolve some inconsistencies in the literature related to acculturation, physical activity, and dietary intake among Latino immigrants. Although some studies show that rates of physical activity decrease as immigrants spend more time in the United States,15 others show that rates of physical activity increase and amount of sedentary behaviors decrease.16,17 Additionally, though many studies have linked higher rates of fast-food consumption to second- and third-generation immigrants,18 other studies have not revealed any notable differences in fast-food consumption between first and later generations.19 Understanding the mechanisms underlying the decline in health for specific immigrant ethnic groups as they acculturate to the United States could provide an important empirical basis for future culturally sensitive obesity intervention and prevention efforts.
Participants and Procedures
The study sample included 2292 Latino male (n = 1138) and female (n = 1154) participants (M [mean] age = 22.29; range = 16–27) enrolled in wave III (2001) of the National Longitudinal Study of Adolescent Health (Add Health), a longitudinal, nationally representative school-based study of US adolescents. Wave III data were used because of the detailed health behavior information available (see measures below). First-generation (n = 514), second-generation (n = 841) and third-generation (n = 937) participants were included in the analyses. Variables used in the analyses were taken from the in-home interview, a series of questionnaires administered to adolescents at their residence. The survey design and sampling frame have been described in detail in other studies.20
Immigrant Generation Status
Immigrant generation was based on participant reports of their own nativity and their parents' nativity as recorded in the in-home interview. Adolescents were coded as “first generation” if neither they nor their resident parents were born in the United States, “second generation” if they were born in the United States but one or both of their resident parent was not born in the United States, and “third generation and above” if both they and their resident parents were born in the United States. Thus, children born third generation and above were collapsed into one category as is typical in generational research.21
Ethnicity for first- and second-generation participants was determined by resident mother's nativity. First- and second-generation adolescents and emerging adults were coded as Latino if their resident mother endorsed any of the following as her birth country or region: Colombia, Cuba, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Nigeria, Panama, Peru, Puerto Rico, Central and Northern South America, or Southern South America. Since mothers of third-generation participants were born in the United States, they were coded as Latino if they marked “yes” to the question: “Are you of Hispanic origin?”.
Moderate-to-vigorous physical activity was assessed using standard 7-day recall (times per week) questionnaire methodology.22 The Add Health adolescents were asked about the times/week spent in various physical activities (e.g., “During the past week, how many times did you go rollerblading, bicycling, skateboarding?”). Based on previous work,6 a physical activity summary variable was created by summing together participation in 8 moderate-to-vigorous physical activities each week.23 Responses for each question could range from 0 (none) to 7 (7 or more times); therefore, responses for the physical activity summary variable could range from 0 to 56.
As used in previous research,23 the frequency of fast-food consumption was assessed using a standard 7-day recall (times per week) questionnaire methodology.22 Participants were asked 1 question about the times per week spent eating fast food (e.g., “On how many of the past 7 days, did you eat food from a fast-food place, or a local fast-food restaurant?”) Responses could range from 0 (none) to 7 (7 or more times).
Sedentary behaviors were also assessed using standard 7-day recall (hours per week) questionnaire methodology.22 Frequency of sedentary behavior was based on questions about the amount of adolescents and emerging adults' participation in sedentary behaviors each week (“On average, how many hours do you spend watching TV, playing video games?”). Based on previous research,6 a sedentary behavior summary variable was created by summing together participation in 3 sedentary activities each week. Responses for each question could range from 0 to 168 hours; therefore, responses for the sedentary behavior summary variable could range from 0 to 504 hours.
Body Mass Index
Height and weight were directly measured by field interviewers in Wave III during in-home surveys. Body mass index (BMI; kg/m2) stratified by age and gender was used to classify individuals as underweight (BMI < fifth percentile), normal weight (BMI = 5 to < 85th percentile), overweight (BMI = 85 to < 95th percentile), or obese (BMI > 95th percentile).2
Before analyses, variables were inspected for missing data and normality of distribution. No deviations were observed, which would preclude the use of parametric tests. As a first step in analyses, Pearson bivariate correlations were calculated to quantify relations among health behaviors and our outcome variable, BMI. A χ2 test examined whether immigrant generation status differed by categorical overweight designation from established BMI cutoff levels (Fig. 1). A series of analyses of variance (ANOVAs) were then conducted to examine generational differences in our health behavior and BMI variables of interest. Note that we used an analysis of covariance when testing for the paradox on BMI by immigrant generation status, while controlling for parent income (a control condition necessary to demonstrate the paradox). The health behavior ANOVAs were also used to identify predictor variables to be used in our mediation analysis; any health behaviors that demonstrated significant generation differences after controlling for family income were considered as potential mediators to explain generational differences in BMI. For health behaviors with a significant immigrant generation effect, a series of ordinary least squares regression models were calculated to test the following mediation conditions on BMI (dependent variable [DV]): (1) the independent variable (IV) must be significantly associated with the DV, (2) the IV must be significantly associated with the mediator variable (MV), (3) the MV must be significantly associated with the DV, and (4) the impact of the IV on the DV is less after controlling for the MV.24 Note that the final mediation model (4) also included family income as a control variable. The results presented below follow this data analytic plan.
Bivariate correlations were first examined to identify relations among physical activity, sedentary behaviors, fast-food consumption, and body mass index (BMI). Frequency of sedentary behavior was positively correlated with BMI (r = .12, p < .05) and fast-food consumption (r = .15, p < .01), such that as frequency of sedentary behaviors increased, so did BMI and frequency of fast-food consumption. No other significant correlations emerged.
We next ran a 1-way analysis of covariance to identify any differences in our outcome variable by generation status (while controlling for family income). We documented the immigrant paradox by observing a significant difference by immigrant generation on our outcome variable, BMI (F(1,2275) = 10.94, p < .01). Post hoc analyses showed that second (MBMI = 29.09) and third-generation (MBMI = 30.97) immigrant youth had significantly higher BMI levels than first-generation youth (MBMI = 25.88), after controlling for family income. Furthermore, first-generation youth had a significantly lower percentage of overweight classifications (19.8%) than second (37.2%) and third-generation adolescents (43.0%) (χ2(1,N=2278) = 14.53, p < .01) (Figure 1).
To determine which potential mediators (i.e., physical activity, fast-food consumption, sedentary activities) differed by immigrant generation, we ran 3 additional 1-way analyses of variances. Sedentary behaviors emerged as the only health behavior significantly related to generation status (F(1,457) = 4.84, p < .05). First-generation Latino immigrants demonstrated significantly lower rates of sedentary behaviors (M = 18.15 hr/wk, SD = 11.2) than their third-generation peers (M = 22.04 hr/wk, SD = 16.73). First generation also showed lower rates of sedentary activities than second-generation immigrant youth (M = 19.45, SD = 16.47), though these results were not significant (Figures 2 and 3). On average, first-generation Latino adolescents spent approximately 4 fewer hours engaging in sedentary activities than third-generation Latinos, and 3 fewer hours than second-generation Latinos. Therefore, we proceeded to test for sedentary behaviors only in our final mediation models. See Table 1 for descriptive information for the variables of interest by generation status.
Using multiple linear regression and controlling for household income, the first model established the immigrant paradox (F(1,2276) = 13.46, p < .01), such that generation status significantly predicted BMI (β = 3.101, t(2277) = 3.67, p < .01), with first- and second-generation youth demonstrating lower BMIs than third-generation youth. The second model also was significant (F(1,457) = 4.59, p < .05), such that generation status predicted frequency of sedentary behaviors (β = 3.129, t(458) = 2.14, p < .05), with first- and second-generation youth demonstrating lower rates of sedentary behavior than third generation. The third and final model indicated that sedentary behaviors predicted BMI levels (β = .157, t(458) = 2.38, p < .05), such that higher rates of sedentary behaviors predicted greater levels of BMI. The relation between generation status and BMI was no longer significant when accounting for sedentary behaviors. These results reveal that sedentary behaviors partially mediate the relation between generation status and BMI (Table 2). This final model predicted 16% of the variability in adolescent BMI.
To the best of our knowledge, this study is the first to demonstrate that specific health behaviors may be linked to elevated rates of obesity among second- and third-generation Latino immigrant adolescents and emerging adults, and account for the population “paradox” observed for this group. Over the last 3 decades, the United States has experienced an alarming increase in rates of childhood and adolescent obesity, particularly among immigrant ethnic minorities; although this trend is beginning to level off, the average body mass index (BMI) is greater than Centers for Disease Control recommendations. Research such as this study, which offers evidence regarding specific behaviors that may be placing Latino adolescents and emerging adults at increased risk for obesity as they acculturate, is therefore of timely importance.
Unlike other studies, this research goes beyond using the term “acculturation” as a description of this obesity concern to identify specific health behaviors that may serve to explain this health outcome. Our findings revealed that higher rates of sedentary behaviors common among adolescents and emerging adults in the United States, such as playing video games, watching TV, and playing on the computer, play a role in explaining the higher rates of obesity among third-generation Latino immigrant youth. Recent work in the health behavior literature has established that sedentary behavior, or sitting for prolonged periods of time, is a distinct health behavior from engaging in low levels of exercise.25 In fact, engaging in sedentary activities has been shown to be an independent predictor of premature mortality risk.25 Individuals who meet the public health guidelines for physical activity are still at risk for metabolic conditions if they engage in prolonged periods of sedentary behaviors.25 Our findings, which show the unique role of sedentary behaviors in predicting obesity risk, are concerning given the powerful association between sedentary activity and numerous adverse health conditions.
Interestingly, our results showed that the only significant differences observed in sedentary behaviors were between first-generation and third-generation youth, such that third-generation immigrants engaged in significantly more sedentary activities than first generation. While second-generation youth demonstrated higher rates of sedentary activities than first generation, these rates were not statistically significant. These findings indicate a shift that may occur in health behaviors that is most pronounced from the second to the third generation. When compared with white nonimmigrants, research has shown that third-generation Latino immigrants have higher rates of sedentary behavior as well, particularly higher rates of TV viewing and videogame playing.26 These findings identify sedentary behaviors as an important consideration for intervention efforts among Latino immigrant families, particularly for third-generation Latino-Americans.
Interestingly, we did not find any generational differences in engagement in physical activity or fast-food consumption of adolescents and emerging adults. Participants across all generations reported that they engaged in little physical activity and fast-food consumption. Compared with white nonimmigrant youth, though, other studies have shown that Latino immigrant youth have lower rates of physical activity and higher consumption of unhealthy foods26; indicating the potential role of other health behaviors in the development and persistence of obesity in the Latino immigrant community. In our study, physical activity and fast-food consumption were measured in “times per week” whereas sedentary behaviors were measured in “hours per week.” This floor effect observed for the physical activity and fast-food consumption scales could be one reason why we did not find generational differences with these 2 health behaviors. Future studies would benefit from using the same measurement for all health behaviors as to be able to make meaningful comparisons. Additionally, using summary scores for these health behaviors may not have fully captured the day-to-day variability in health behavior patterns. More accurate assessments of health behaviors, such as accelerometers to measure daily levels of physical activity, should be used in future studies to represent engagement in physical activity health behaviors. Though the Add Health data set is comprehensive and includes a nationally representative sample of adolescents, the year of data collection (2001) may have also limited our understanding of more current health behavior patterns among immigrant youth.
Overall, our findings suggest that interventions focused on decreasing sedentary behaviors may be one way to reduce obesity among third-generation Latino immigrant adolescents and emerging adults in the United States. One of the main strengths of the study is that it represents an important first step in better understanding weight gain among Latino immigrant populations and sets the stage for future studies to further examine this concerning population health pattern. Since sedentary behaviors emerged as only partially mediating the relation between generation status and BMI, future research could examine other culturally specific health behaviors not captured in the Add Health study that also may be playing a role in the intergenerational variation in obesity rates. Additionally, prospective studies could measure behaviors in participants' culture of origin and behaviors in the United States to see how changes in culturally reinforced behaviors from one culture to another may impact the development of unhealthy behaviors. Though this study adds important information to this topic area, there is still much more variability remaining to more robustly explain BMIs of Latino adolescents and emerging adults by immigrant generation. Future studies that take into account important immigration-related contextual factors (such as immigrant's legal status) and culture-specific eating and physical activity beliefs and behaviors, which may provide greater detail needed to support intervention and prevention efforts among Latino communities. Additionally, although most literature on the immigrant paradox phenomenon is cross-generational in nature,10 future research should consider the important acculturation-related construct of time since arrival in the United States. It will be very important moving forward for researchers to track acculturation within subjects (i.e., using longitudinal designs) to determine whether acculturation within individuals over time also indicates the “declines” in health behaviors and outcomes observed across immigrant generations.
1. Ogden CL, Carroll MD, Kit BK, et al.. Prevalence of obesity and trends in body mass index among US children and adolescents, 1999-2010. JAMA. 2012;307:483–490.
2. Center for Disease Control and Prevention (CDC): National Center for Health Statistics. 2009–2010. NHANES data on the prevalence of overweight among children and adolescents: United States. Available at: http://www.cdc.gov/healthyyouth/obesity/facts.htm
. Accessed November 10, 2013.
3. Daniels SR, Arnett DK, Eckel RH, et al.. Overweight in children and adolescents: pathophysiology, consequences, prevention, and treatment. Circulation. 2005;111:1999–2002.
4. Cunningham PB, Ellis DA, Naar-King S. Treating pediatric obesity using an empirically supported treatment: a case report. J Clin Child Adolesc Psychol. 2010;39:141–150.
5. Guo SS, Roche AF, Chumlea WC, et al.. The predictive value of childhood body mass index values for overweight at age 35 years. Am J Clin Nutr. 1994;59:810–819.
6. Gordon-Larsen P, Harris KM, Ward DS, et al.. Acculturation and overweight-related behaviors among Hispanic immigrants to the US: the national longitudinal study of adolescent health. Social Sci Med. 2003;57:2023–2034.
7. Singh GK, Kogan MD, Yu SM. Disparities in obesity and overweight prevalence among US immigrant children and adolescents by generational status. J Community Health. 2009;34:271–281.
8. Alba RD, Logan JR, Stults BJ. The changing neighborhood contexts of the immigrant metropolis. Social Forces. 2000;79:587–621.
9. Rumbaut RG. Assimilation and its discontents: between rhetoric and reality. Int Migration Rev. 1997;31:923–960.
10. Garcia Coll C, Marks AK. The Immigrant Paradox in Children and Adolescents: Is Becoming American a Developmental Risk? Washington, DC: American Psychological Association; 2012.
11. Hernandez DJ, Denton NA, Blanchard VL. Children in the United States of America: a statistical portrait by race-ethnicity, immigrant origins, and language. Ann Am Acad Polit Soc Sci. 2011;633:102–127.
12. Popkin BM, Udry JR. Adolescent obesity increases significantly in second and third generation U.S. immigrants: the national longitudinal study of adolescent health. J Nutr. 1998;128:701–706.
13. Berry JW. Social and cultural change. In Triandis HC, Brislin RW, eds. Handbook of Cross-cultural Psychology: Social Psychology. Vol 5. Boston, MA: Allyn and Bacon; 1980:211–279.
14. Sussner KM, Lindsay AC, Greaney ML, et al.. The influence of immigrant status and acculturation on the development of overweight in Latino families: a qualitative study. J Immigrant Minor Health. 2008;10:497–505.
15. Esparza J, Fox C, Harper I, et al.. Daily energy expenditure in Mexican and USA Pima Indians: low physical activity as a possible cause of obesity. Int J Obes. 2000;24:55–59.
16. Crespo CJ, Smit E, Carter-Pokras O, et al.. Acculturation and leisure-time physical inactivity in Mexican American adults: results from NHANES III, 1998–1994. Am J Public Health. 2001;91:1254–1257.
17. Wingo PA, Kulkarni A, Borrud LG, et al.. Health disparities among Mexican American women aged 15–44 years: National Health and Nutrition Examination Survey, 1999-2004. Am J Public Health. 2009;99:1300–1307.
18. Creighton MJ, Goldman N, Pebley AR, et al.. Durational and generational differences in Mexican immigrant obesity: is acculturation the explanation? Soc Sci Med. 2012;75:300–310.
19. Ayala GX, Baquero B, Klinger S. A systematic review of the relationship between acculturation and diet among Latinos in the United States: implications for future research. J Am Diet Assoc. 2008;108:1330–1344.
21. Pena JB, Wyman PA, Brown CH, et al.. Immigration generation status and its association with suicide attempts, substance use, and depressive symptoms among Latino adolescents in the USA. Prev Sci. 2008;9:299–310.
22. Andersen RE, Crespo CH, Bartlett SJ, et al.. Relationship of physical activity and television watching with body weight and level of fatness among children: results from the Third National Health and Nutrition Examination Survey. JAMA. 1998;279:938–942.
23. Niemeier HM, Raynor HA, Lloyd-Richardson EE, et al.. Fast food consumption and breakfast skipping: predictors of weight gain from adolescence to adulthood in a nationally representative sample. J Adolesc Health. 2008;39:842–849.
24. Baron M, Kenny D. Moderator-mediator variables distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51:1173–1182.
25. Owen N, Healy GN, Matthews CE, et al.. Too much sitting: the population health science of sedentary behavior. Exerc Sport Sci Rev. 2010;38:105–113.
26. Allen ML, Elliott MN, Morales LS, et al.. Adolescent participation in preventive health behaviors, physical activity, and nutrition: differences across immigrant generations for Asians and Latinos compared with Whites. Am J Public Health. 2007;97:337–343.
Index terms: sedentary behaviors; immigrants; obesity; immigrant paradox