PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Phys Act Health. Author manuscript; available in PMC 2010 April 27.
Published in final edited form as:
J Phys Act Health. 2010 January; 7(1): 26–36.
PMCID: PMC2860451
NIHMSID: NIHMS175410

The Association of Employment and Physical Activity among Black and White 10th and 12th Grade Students in the United States

Abstract

Background

Evidence of an association between employment and physical activity (PA) in youth has been mixed, with studies suggesting both positive and negative associations. We examined the association between employment and PA among U.S. high school students as measured by self-reported overall PA, vigorous exercise, and participation in school athletic teams.

Methods

We employed a secondary analysis using weighted linear regression to a sample of black and white 10th grade (n=12073) and 12th grade students (n=5500) drawn from the nationally representative cross-sectional 2004 Monitoring the Future Study.

Results

Overall, 36.5% of 10th and 74.6% of 12th grade students were employed. In multivariable analyses, 10th graders working >10 hours a week reported less overall PA and exercise and those working >20 hours a week reported less participation in team sports. Among 12th graders, any level of employment was associated with lower rates of team sports; those working >10 hours a week reported less overall PA; and those working >20 hours reported less exercise.

Conclusions

Employment at and above 10 hours per week is negatively associated with PA. Increasing work intensity may shed light on the decline of PA as adolescents grow older and merits further attention in research.

Keywords: adolescence, employment, physical activity, team sports, time use

Background

Children’s participation in regular physical activity (PA) provides many health benefits, including better mental health and emotional well-being related outcomes, improved musculoskeletal health, and prevention of chronic diseases such as hypertension, type 2 diabetes, obesity and cardiovascular disease.1 As children who are active during childhood are more likely to be active as adults,24 childhood may be an important developmental stage for establishing a lifelong physically active lifestyle. Unfortunately, participation in regular physical activity tends to decline as children age into adolescence.5, 6 In the United States, the majority of high school students (64.2%) do not meet current recommended levels of moderate or vigorous physical activity.7

One potential explanation for the low engagement in recommended levels of physical activity is the transition of adolescents into the workforce. Nearly 75% of high school seniors in the United States were employed in the labor force at some point during the school year, of which 56% worked 21 or more hours per week year according to the U.S. Department of Labor.8 Because time constraints and competing activities such as paid employment have been reported by adolescents as important barriers to physical activity, 9, 10 assessing the impact of employment on youth participation in physical activity may provide insights into the low numbers of youth who meet current physical activity recommendations.

Evidence of an association between employment and physical activity in school-aged youth has been mixed, with some studies reporting a positive association,11, 12 others finding youth participation in part-time work to be negatively associated with physical activity1316, and others reporting no statistically significant association11, 17. Cariere et. al conducted two separate analyses using independent data sources—one cross-sectional and demonstrating a positive association, and one prospective—which did not demonstrate any association.11 The contrasting findings of this literature to date may be due to variations in the populations studied—four of these studies were conducted in the United States,1315, 17 three are from Canada11, 12, and one is from Iceland.16 Other factors that may explain the contrasting findings include the use of differing measures of employment and physical activity, the varying number and type of covariates controlled for in analyses, or other conceptual and methodologic differences.

This study contributes to the current body of research by examining the association of student employment with three indicators of physical activity (overall physical activity, vigorous exercise, and sports team participation) among a nationally representative sample of black and white U.S. high school students. In addressing limitations of past studies, this study tests the assumption of a linear relationship between employment and physical activity, explores potential gender differences, and examines the effect of adjusting for multiple sociodemographic, school achievement, and leisure time use covariates.

Methods

Data for the present study were collected as part of the 2004 Monitoring the Future: A Continuing Study of American Youth survey. We used cross-sectional self-report survey data obtained in the publically available 2004 1) 8th and 10th grade survey18 and 2) 12th grade survey19. The Monitoring the Future survey (MTF) uses a multi-stage random sampling procedure to select a nationally representative sample of high school students each year. The overall response rates for the 10th and 12th grade data were 88% and 82% respectively20. More detailed description of the survey is available elsewhere20.

We restricted our analyses to 10th and 12th grade black and white students who provided a response for our primary independent variable (employment) and at least one of the 3 dependent variables measuring physical activity. We did not analyze data for 8th grade students as we rationalized that 8th grade students are less likely to hold legal employment. We limited the analysis to data on black and white students due to the restricted range of responses to race and ethnicity questions in the publicly available dataset.

Physical Activity

Three items measured the respondent’s physical activity: 1) “How often do you do each of the following…actively participate in sports, athletics, or exercising?” answered on a 1–5 scale ranging from never to almost every day (defined in this study as “overall physical activity”); 2) “How often do you…exercise vigorously (jogging, swimming, calisthenics, or any other active sports)?” answered on a 1–6 point scale ranging from never to every day (“vigorous exercise”); and 3) “To what extent have you participated in the following school activities during this school year…athletic teams?” answered on a 1–5 scale ranging from not at all to great (“school athletic team participation”).

Employment

Tenth grade students were asked “On the average over the school year, how many hours per week do you work in a paid job?” Twelfth grade students were asked about hourly employment in paid and unpaid jobs (e.g. work done without remuneration in the service of a family business). Eight possible response options ranged from “none” to “>30 hours per week”. In order to test the assumption of linearity and to determine how best to model the shape of the association between employment and physical activity, we examined bivariate descriptive statistics and ran multiple linear regression models. We used Wald tests in the linear regression models to test the equality of the coefficients for the original 8 response categories of employment. Categories that were not significantly different from each other were combined, resulting in the following 4 mutually exclusive categories: none, ≤10 hours, >10 and ≤ 20 hours, and >20 hours per week.

Covariates

We include multiple demographic and socioeconomic covariates, including sociodemographic variables (race, sex, a race by sex interaction term, age, parental education [a stable and reliable measurement that captures child socioeconomic status21], and urbanity, a factor associated with youth physical activity in some studies.22 Age was defined as <16 or 16+ year in the 10th grade sample and <18 or 18+ years in the 12th grade sample, due to the restricted range of responses to age in the publicly available dataset. Sex was male or female. Race was black or white. Parent’s education level was defined as the highest level of schooling completed by the parent with the highest level of education who lives in the same home as the respondent on a continuous 1–6 scale ranging from 1=grade school to 6=graduate school; if neither parent lived with the child, the mother’s educational level was assigned. Urbanicity was available only for the 10th grade sample and was defined according to where the respondent currently lives: rural (farm or country) or urban (city or town). We include an interaction term to capture the race-gender interaction found in previous studies (wherein black boys report greater PA than white boys while black girls report less PA than white girls).23, 24

We also covaried for self-reported grade point average (GPA) and leisure time use. Grade average is associated with work intensity 17, 25 and physical activity 26 and thus are an obvious covariate. Grade average was defined as a continuous variable measuring the respondent’s average grades so far in high school on a 1–9 scale (1=“D [69 or below],” 2=“C− [70–72],” 3=“C [73–76],” 4=“C+ [77–79],” 5=“B− [ 80–82],” 6=“B [83–86],” 7=“B+ [87–89],” 8=“A− [90–92],” 9=”A [93–100].”) Because leisure time use is associated with the dependent variable, physical activity27 and the independent variable, employment,15, 17 these variables are also included as covariates in this study. We selected leisure time use variables that were answered by the full sample and demonstrated consistent bivariate associations with employment and at least two of the three measures of physical activity. The time use variables were based on respondents’ ranking on a 1(never) to 5 (daily) scale of how often they did each of the following activities: go to movies, participate in community affairs or volunteer work, get together with friends informally (in your free time), spend at least an hour of leisure time (free time) alone, go to parties or other social affairs, or watch TV (12th graders only). For 10th graders, television viewing was measured with the question: “How much TV do you estimate you watch on an average weekday?” measured on a 1 (none) to 7 (5+ hours) scale. Two additional time use variables were available for the 12th grade sample: going out and dating. Going out was measured on a 1 (<1 week) to 6 (6–7 times week) point scale with the question, “During a typical week, on how many evenings do you go out for fun and recreation?” Dating was measured on a 1 (never) to 6 (>3 times week) point scale with the question, “On the average, how often do you go out with a date (or your spouse, if you are married)?”

Statistical Analysis

Linear regression models were used to examine the associations of hours of employment with three physical activity behaviors: 1) overall activity, 2) vigorous exercise, and 3) school athletic teams. For each outcome, model building proceeded sequentially: Model 1 included only the bivariate association of work and physical activity; Model 2 controlled for demographic and socioeconomic correlates; and Model 3 controlled for grade point average and leisure time use variables. All analyses were adjusted for the sampling design of the MTF survey using the sample weights calculated by the MTF researchers for use with the publically available dataset. More detail about the sampling procedures and sample weights is available elsewhere28. Because independent samples are drawn for the 10th and the 12th grade students, all analyses are conducted separately by grade level to take into account the separate sampling designs and different weighting procedures for these students. Stata version 9.1 was used for all analyses. (Stata Statistical Software: Release 10.0. College Station, TX: StataCorp LP.)

Results

Sample characteristics by hours worked are presented in Table 1. Among 10th grade students, 63.5% do not work, 20.6% work ≤ 10 hours a week, 10.6% work >10 and ≤ 20 hours, and 5.3% work >20 hours per week. Among 12th grade students, 25.4% do not work, 21.0% work ≤10 hours, 27.4% work >10 and ≤20 hours, and 26.8% work >20 hours per week.

Table 1
Sample Characteristics by Grade and Employment: Number (Survey Adjusted Percent) or Survey Adjusted Mean (Linearlized Standard Error)

Results from regression models demonstrating associations between employment and physical activity for 10th and 12th grade students are presented in Table 2. Due to space considerations relating to our estimation of 18 separate regression models, we are unable to provide the full models and instead display only the estimates related to the primary independent variable of interest, hours of employment.

Table 2
Main Effects of Employment from Regression Analysis Predicting Physical Activity for 10th and 12th Grade Students

In Models1 and 2, some positive associations between work hours and physical activity were evident in 10th grade sample. These associations demonstrated that, compared to their unemployed peers, 10th graders working ≤10 hours a week reported higher rates of overall physical activity and team sports. Negative associations were also evident, demonstrating that 10th graders working >10 reported less exercise and overall activity and those reporting >20 hours reported less participation in team sports. In Model 2, all but one of these negative associations remained significant. After controlling for the full set of covariates in Model 3 analyses, only negative associations remained significant: 10th graders working >10 hours a week reported less overall physical activity and exercise and those working >20 hours a week reported less participation in team sports.

Among 12th graders, bivariate associations in Model 1 demonstrated less participation in team sports for those working >10 hours a week and less overall activity and exercise for those working >20 hours a week. In Model 2 those associations persisted, and additionally, 12th graders working >10 hours demonstrated less overall physical activity. After accounting for the full set of covariates for 12th graders in Model 3, any level of employment was associated with lower rates of team sports, those working >20 hours reported less exercise, and those working >10 hours a week reported less overall physical activity.

Discussion

This study examined the association of employment with three indicators of physical activity among a large, nationally representative sample of 10th and 12th grade black and white high school students in the United States. Overall, we found students who reported working more than 10 or 20 hours per week during the school week engaged in less physical activity- a finding that persisted after controlling for several key sociodemographic and leisure time use covariates. Our results provide some context for the mixed findings of previous studies on this topic by suggesting a threshold effect for number of hours worked on PA engagement and by demonstrating the importance of adequate control for sociodemographic, school achievement, and time use covariates.

Previous studies among school-aged youth have had mixed results, with some studies demonstrating positive associations that persisted following control for a limited set of sociodemographic11, 12 and time use variables.12 Studies of young adult populations have also demonstrated some positive associations.2931 In our study, while two of the 10th grade models demonstrated positive associations between working ≤ 10 hours a week and physical activity, both of these positive associations dropped to insignificance after controlling for our more comprehensive set of sociodemographic and time use covariates.

Other studies have demonstrated negative associations between employment and physical activity among adolescents.1316 In our study, we demonstrate a number of negative associations, all but one of which were evident at higher levels of work intensity—at >10 or >20 hours of employment per week.

Our finding of threshold effects suggests that the relationship between work intensity and physical activity may not be as straightforward as suggested by previous studies. Several of the aforementioned studies on employment and physical activity, while concluding that the association between employment and physical activity follows a general linear trend11, 13, 15 or alternatively, demonstrates no significant linear trend,17 did show some limited evidence of potential nonlinear or threshold effects. By utilizing continuous or dichotomous measures of employment, displaying only a limited set of the data, graphing the mean values of physical activity by work hours without testing for variations in linearity, these studies effectively masked any existing curvilinear or threshold effects. However, unlike these studies, we demonstrated statistically significant positive and negative associations in our bivariate models. In our fully adjusted models, however, only negative associated remained, demonstrating that, in general, only work at or above 10 or 20 hours a week was associated with decreased PA.

The contrasting findings in previous research may be the result of other factors as well, including the wide variation in the type and number of measures of physical activity, employment, and covariates. For example, one recent study demonstrates the potential impact of employment on measurement of physical activity. In a study of 12th grade girls, employed girls reported higher total physical activity, however, after activity related to work was discounted, the employed girls reported less physical activity compared to the non-employed girls.14 Previous research has demonstrated that different operationalizations of employment may alter findings regarding the association of work and problem behaviors among adolescents and adults, demonstrating the importance of choice between competing measures of work.32 Notably, only one article that we found in this area provided any justification for a decision regarding the treatment of employment as a continuous or categorical variable15 and no studies provided any justification behind the number or cutpoints of categories. Moreover, not all of these studies controlled for covariates; of those that did, only a limited set of sociodemographic variables were included and just one study controlled for any type of leisure activities,12 opening the potential for residual confounding.

Our study is unique in the literature on employment and PA because it controls for a comprehensive set of sociodemographic and time use covariates. After controlling for these covariates, we found that the few positive associations between employment and PA were ameliorated while the negative associations remained significant. These findings suggest that sociodemographic characteristics, school achievement, and time use covariates play some important role in physical activity engagement and/or may confound the association of interest. For example, time use variables may mediate the relationship between employment and physical activity. On the other hand, sociodemographic variables such as gender and race may simply confound the association. Alternatively, employment itself may simply be a mediator of underlying relationships between sociodemographic variables such as race and socioeconomic status and PA. However, because our data is cross-sectional, we are unable to appropriately test any potentially mediating pathways.

The positive association of higher grades and greater time spent in leisure-time pursuits with PA—with the exception of time spent watching television and time spent alone, both of which were negatively associated with PA [data not shown], underscore the importance of other developmental assets and time use behaviors that may influence PA engagement.

We also found gender, race/ethnicity and socioeconomic associations with engagement in the three PA indicators. Similar to other studies,23, 24 we uncovered statistically significant race by gender interactions. Because the magnitude and direction of the primary association of interest were similar in males and females across all models and the small number of blacks in many of the categories of employment limited our ability to explore race-specific effects in stratified models, we opted not to present race-or gender-stratified models. Notably, previous research on high school students has demonstrated similar associations of employment and PA between girls and boys 12 and in a study of girls only, between blacks and whites.14

There are well-documented socioeconomic disparities in physical activity among adolescents, with adolescents of lower socioeconomic status reporting less participation in physical activity than those of higher status.33, 34 In our study, higher parental education was significantly and positively related to physical activity in each of the models [data not shown]. Moreover, children with parents who had less than a college education were more likely to work than children with parents who graduated college and to work longer hours [data not shown]. As such, work intensity may simply be a proxy measure of socioeconomic status or an important factor in a larger causal web of disadvantage. For example, employment may mediate the relationship between socioeconomic status and PA such that low socioeconomic status results in higher hours worked during adolescence, which in turn decreases the opportunity for engagement in PA due to higher time constraints and lower financial ability. Alternatively, as a proxy measure of socioeconomic status, work intensity may simply reflect unmeasured aspects of low socioeconomic status that have been associated with lower PA such as higher concerns about neighborhood safety35 and lower neighborhood access to recreational facilities and parks.36 Future research should utilize longitudinal data to examine the associations between socioeconomic status, employment, and physical activity among adolescents.

Two theoretical perspectives that may help to explain the association between employment and physical activity are the “Time Trade-Off” perspective and the “Contextual Competence” perspective. The “Time Trade-Off” perspective, described by Safron and colleagues posits that time spent in one activity leads to less time spent in other developmentally beneficial activities.15 Findings of this study provide some support for a Time-Trade Off hypothesis as threshold effects were found in which increased hours of work was negatively associated with engagement in three PA indicators.

However, in contrast to the literature that has framed youth activities as either positive or negative and simplifies the complexity of adolescent lives into bivariate associations, research building on a model of contextual competence may provide a richer understanding of the social and contextual influences on PA engagement. A contextual competence perspective suggests that competence among adolescents, or positive social and psychological functioning, develops in multiple proximal social contexts.37 In a study of urban youth, engagement in multiple vs. single youth developmental contexts (e.g. peer, school, athletics, employment, etc.) predicted more positive psychological wellbeing.37 Along these lines, other research has suggested that time management and the juggling of multiple leisure time activities may be as relevant to adolescent PA participation as the types or number of hours spent on specific activities.38 In another study of adolescents and young adults, complex patterns between gender, age, school enrollment, employment, and different types of PA39 also support the thesis that behavior is shaped in an environment inclusive of multiple contexts, activities, and demands on time. Seen under this perspective, our results may suggest that in moderation, employment may form part of a larger clustering of positive behaviors and experiences that include physical activity, while excessive hours spent in work to the exclusion of engagement in other activities may be associated with more negative developmental outcomes. Future research under this perspective could examine participation in PA among workers who are engaged in multiple contexts vs. more disengaged workers who have limited engagement in other developmental contexts.

While our study provides evidence of an inverse association between higher work intensity and physical activity engagement among youth, we need to be cautious with our interpretations of the role work plays in physical activity engagement based on these data. In the bivariate models, work intensity explained only a small percent of the variance in physical activity (0.3%–5.2%), while the multivariable models controlling for the full set of covariates explained a much greater percentage of the variance (11.5%–16.0%). These finding suggest that adolescent employment is one of a constellation of adolescent attributes, behaviors and social circumstances associated with physical activity. While policy and intervention suggestions for working youth need to be made with caution, given the paucity of research on this population, the potential negative effects of working >10 or >20 hours a week is troubling. Workplace tobacco control interventions for adolescents have been implemented successfully and the feasibility and effectiveness of workplace interventions for PA also deserves exploration.40 Parents and teachers may also consider monitoring the effect of work intensity on students’ engagement in physical activity.

The results of this study are subject to several limitations, including reliance on cross-sectional self-report data and the use of single-item measures of PA. Of the 3 PA items used in this analysis, 2 do not have any available published information on their validity or reliability. The item measuring participation in athletic teams has previously been shown to correlate moderately (r=.40–.64, depending on grade and gender) with school representatives’ estimated reports on the percentage of students participating in athletic teams in their schools.33 One possible bias related to the PA measures, particularly the measure of vigorous exercise, is that these measures do not capture employment- vs. non-employment related PA. While employed teenagers may obtain much of their total PA while on the job,14 if work-related activity was captured in the 3 self-reported PA measures in this study, it would result in estimates biased toward the null. An additional limitation of this study is our sample consisting of only black or white students and our limited information on age. The publically available dataset provides restricted information about age and race in order to avoid threats to confidentiality. Due to the large sample size, it is possible that weak effects reached statistical significance; additionally, the multiple tests run in this study increase the chances of type I error. Finally, approximately half of students in this sample reported having one or more parents with a college education or beyond, representing a relatively highly educated sample. Studies have demonstrated that students of more educated parents report greater participation in physical activity and sports teams.33 Future studies examining the association of youth employment and physical activity should consider including additional measurements of socioeconomic status to guard against erroneous reporting of parental education by students and should also explore whether the association of adolescent employment and PA varies by socioeconomic status.

These limitations notwithstanding, the large, nationally representative sample of black and white students analyzed in this study provides some basis for generalizing these results to high school students of similar racial/ethnic and socioeconomic status backgrounds in the United States. Unlike many studies that examine engagement in a single type of physical activity, we analyzed three PA indicators representing different aspects of physical activity (overall physical activity, vigorous exercise, and sports team participation). Additionally, we improve upon the existing studies in this area by explicitly testing the assumption of a linear relationship between employment and physical activity, exploring the possibility of differential effects by gender and race, and examining the effect of adjusting for comprehensive sets of sociodemographic, school achievement, and time use covariates.

Conclusions

In conclusion, we found that longer hours of work (>10 or >20 hours/week) are associated with less PA among U.S. black and white adolescents. Our findings suggest that work intensity is a small but significant piece of a complex puzzle explaining adolescent engagement in physical activity. Our results provide a basis for future prospective research on the association of employment and physical activity by demonstrating potential threshold effects for employment and by providing some preliminary data on the complex relationships between work hours, leisure time use, and physical activity. Given the limited body of research on adolescent employment and physical activity, more research is needed to confirm these findings and to better understand the role work plays in physical activity engagement among youth. Future studies should utilize validated measures of physical activity, prospective data, and should test for a possible mediating effect of leisure time use on the association between employment and physical activity. Furthermore, studies should explore the effects of the type and timing of adolescent employment—including job type, paid, unpaid, school-day, weekend, school-year or summer work—and should explore the complex interrelations of race, gender, socioeconomic status, adolescent time use, and employment with adolescents’ engagement in physical activity.

Acknowledgments

The secondary data analysis described was supported by Grant Number R25-CA-057712 from the National Cancer Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.

References

1. Institute of Medicine. Committee on Prevention of Obesity in Children and Youth; Food and Nutrition Board; Board on Health Promotion and Disease Prevention. In: Koplan JP, Liverman CT, Kraak VI, editors. Preventing childhood obesity: Health in the balance. Washington, D.C: National Academies Press; 2005.
2. Kelder SH, Perry CL, Klepp KI, Lytle LL. Longitudinal tracking of adolescent smoking, physical activity, and food choice behaviors. Am J Public Health. 1994;84(7):1121–1126. [PubMed]
3. Telama R, Yang X, Laakso L, Viikari J. Physical activity in childhood and adolescence as predictor of physical activity in young adulthood. Am J Prev Med. 1997;13(4):317–323. [PubMed]
4. Telama R, Yang X, Viikari J, Valimaki I, Wanne O, Raitakari O. Physical activity from childhood to adulthood: a 21-year tracking study. Am J Prev Med. 2005;28(3):267–273. [PubMed]
5. Aaron DJ, Storti KL, Robertson RJ, Kriska AM, LaPorte RE. Longitudinal study of the number and choice of leisure time physical activities from mid to late adolescence: implications for school curricula and community recreation programs. Arch Pediatr Adolesc Med. 2002;156(11):1075–1080. [PubMed]
6. Trost SG, Pate RR, Sallis JF, et al. Age and gender differences in objectively measured physical activity in youth. Med Sci Sports Exerc. 2002;34(2):350–355. [PubMed]
7. Centers for Disease Control and Prevention. Youth risk behavior surveillance--2005. Surveillance summaries. MMWR. 2006;55(22–5):1–108.
8. Bureau of Labor Statistics. Work activity of high school students: Data from the national longitudinal survey of youth 1997. Washington D.C: U.S. Department of Labor; 2005. USDL 05–732.
9. Allison KR, Dwyer JJ, Makin S. Perceived barriers to physical activity among high school students. Prev Med. 1999;28(6):608–615. [PubMed]
10. Tappe MK, Duda JL, Ehrnwald PM. Perceived barriers to exercise among adolescents. J Sch Health. 1989;59(4):153–155. [PubMed]
11. Carriere G. Weekly work hours and health-related behaviours in full-time students. Health Rep. 2005;16(4):11–22. [PubMed]
12. Feldman DE, Barnett T, Shrier I, Rossignol M, Abenhaim L. Is physical activity differentially associated with different types of sedentary pursuits? Arch Pediatr Adolesc Med. 2003;157(8):797–802. [PubMed]
13. Bachman JG, Schulenberg J. How Part-Time Work Intensity Relates to Drug-Use, Problem Behavior, Time Use, and Satisfaction Among High-School Seniors - Are These Consequences Or Merely Correlates. Developmental Psychology. 1993;29(2):220–235.
14. Dowda M, Pfeiffer KA, Dishman RK, Pate RR. Associations among physical activity, health indicators, and employment in 12th grade girls. J Womens Health (Larchmt) 2007;16(9):1331–1339. [PubMed]
15. Safron DJ, Schulenberg JE, Bachman JG. Part-time work and hurried adolescence: the links among work intensity, social activities, health behaviors, and substance use. J Health Soc Behav. 2001;42(4):425–449. [PubMed]
16. Vilhjalmsson R, Thorlindsson T. Factors related to physical activity: a study of adolescents. Soc Sci Med. 1998;47(5):665–675. [PubMed]
17. Weller NF, Kelder SH, Cooper SP, Basen-Engquist K, Tortolero SR. School-year employment among high school students: effects on academic, social, and physical functioning. Adolescence. 2003;38(151):441–458. [PubMed]
18. Johnston LD, Bachman JG, O’Malley PM, Schulenberg JE. University of Michigan IfSRSRCIv. Monitoring the Future: A continuing study of American youth (8th- and 10th-grade surveys), 2004. [Computer file] Ann Arbor, MI: Inter-university Consortium for Political and Social Research; 2005.
19. Johnston LD, Bachman JG, O’Malley PM, Schulenberg JE. University of Michigan IfSRSRCIv. Monitoring the Future: A continuing study of American youth (12th-grade survey), 2004. [Computer file] Ann Arbor, MI: Inter-university Consortium for Political and Social Research; 2005.
20. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring the Future national survey results on drug use, 1975–2004: Volume 1, secondary school students. NIH Publication No. 05-5727. National Institute on Drug Abuse; Bethesda, MD: 2005.
21. Hauser RM. Measuring Socioeconomic Status in Studies of Child Development. Child Development. 1994;65(6):5p. [PubMed]
22. Joens-Matre RR, Welk GJ, Calabro MA, Russell DW, Nicklay E, Hensley LD. Rural-urban differences in physical activity, physical fitness, and overweight prevalence of children. J Rural Health Winter. 2008;24(1):49–54. [PubMed]
23. Gordon-Larsen P, McMurray RG, Popkin BM. Determinants of adolescent physical activity and inactivity patterns. Pediatrics. 2000;105(6):E83. [PubMed]
24. Richmond TK, Hayward RA, Gahagan S, Field AE, Heisler M. Can school income and racial/ethnic composition explain the racial/ethnic disparity in adolescent physical activity participation? Pediatrics Jun. 2006;117(6):2158–2166. [PubMed]
25. Mortimer JT, Finch MD, Ryu S, Shanahan MJ, Call KT. The effects of work intensity on adolescent mental health, achievement, and behavioral adjustment: new evidence from a prospective study. Child Dev. 1996;67(3):1243–1261. [PubMed]
26. Trudeau F, Shephard RJ. Physical education, school physical activity, school sports and academic performance. Int J Behav Nutr Phys Act. 2008;5:10. [PMC free article] [PubMed]
27. Mota J, Santos MP, Ribeiro JC. Differences in leisure-time activities according to level of physical activity in adolescents. Journal of Physical Activity and Health. 2008;5 [PubMed]
28. Bachman JG, Johnston LD, O’Malley PM. Appendices A & B. Ann Arbor, Michigan: Survey Research Center, Institute for Social Research, The University of Michigan; 2005. Monitoring the Future: Questionnaire responses from the nation’s high school seniors, 2004; pp. 361–384.
29. Bell S, Lee C. Emerging adulthood and patterns of physical activity among young Australian women. Int J Behav Med. 2005;12(4):227–235. [PubMed]
30. Horn DB, O’Neill JR, Pfeiffer KA, Dowda M, Pate RR. Predictors of physical activity in the transition after high school among young women. Journal of Physical Activity and Health. 2008;5:275–285. [PubMed]
31. Leslie E, Owen N, Salmon J, Bauman A, Sallis JF, Lo SK. Insufficiently active Australian college students: perceived personal, social, and environmental influences. Prev Med. 1999;28(1):20–27. [PubMed]
32. Bauermeister JA, Zimmerman MA, Barnett TE, Caldwell CH. Working in High School and Adaptation in the Transition to Young Adulthood among African American Youth. Journal of Youth & Adolescence. 2007;36(7):14p.
33. Johnston LD, Delva J, O’Malley PM. Sports participation and physical education in American secondary schools: current levels and racial/ethnic and socioeconomic disparities. American Journal of Preventive Medicine Oct. 2007;33(4 Suppl):S195–208. [PubMed]
34. Starfield B, Riley AW, Witt WP, Robertson J. Social class gradients in health during adolescence. J Epidemiol Community Health. 2002;56(5):354–361. [PMC free article] [PubMed]
35. Molnar BE, Gortmaker SL, Bull FC, Buka SL. Unsafe to play? Neighborhood disorder and lack of safety predict reduced physical activity among urban children and adolescents. Am J Health Promot. 2004;18(5):378–386. [PubMed]
36. Gordon-Larsen P, Nelson MC, Page P, Popkin BM. Inequality in the built environment underlies key health disparities in physical activity and obesity. Pediatrics. 2006;117(2):417–424. [PubMed]
37. Pedersen S, Seidman E, Yoshikawa H, Rivera AC, Allen L, Aber JL. Contextual competence: Multiple manifestations among urban adolescents. American Journal of Community Psychology. 2005;35(1/2):65–82. [PubMed]
38. Rushovich BR, Voorhees CC, Davis C, et al. The relationship between unsupervised time after school and physical activity in adolescent girls. Int J Behav Nutr Phys Act. 2006;3:20. [PMC free article] [PubMed]
39. Zick CD, Smith KR, Brown BB, Fan JX, Kowaleski-Jones L. Physical activity during the transition from adolescence to adulthood. J Phys Act Health. 2007;4(2):125–137. [PubMed]
40. Fagan P, Stoddard AM, Hunt MK, Frazier L, Girod K, Sorensen G. The feasibility of evaluating a tobacco control intervention for working youth. Tob Control. 2003 Dec;12(Suppl 4):IV34–39. [PMC free article] [PubMed]