[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
August 2003

Is Physical Activity Differentially Associated With Different Types of Sedentary Pursuits?

Author Affiliations

From the Groupe de recherche interdisciplinaire en santé, École de Réadaptation, Université de Montréal (Dr Feldman), Centre for Clinical Epidemiology and Community Studies, Sir Mortimer B. Davis Jewish General Hospital (Drs Feldman, Shrier, Rossignol, and Abenhaim), and Direction de la Santé Publique, Montréal Centre (Ms Barnett and Dr Rossignol), Montréal, Québec.

Arch Pediatr Adolesc Med. 2003;157(8):797-802. doi:10.1001/archpedi.157.8.797

Objective  To determine whether there is a relationship between the time adolescents spend in physical activity and time they spend in different sedentary pursuits: watching television, playing video games, working on computers, doing homework, and reading, taking into account the effect of part-time work on students' residual time.

Design  Cross-sectional cohort design.

Participants and Setting  Seven hundred forty-three high school students from 2 inner-city public schools and 1 private school.

Methods  Students completed a self-administered questionnaire that addressed time spent in physical activity, time spent in sedentary pursuits, musculoskeletal pain, and psychosocial issues and were also measured for height and weight.

Main Outcome Measure  Level of physical activity (low, moderate, high).

Results  There were more girls than boys in the low and moderate physical activity groups and more boys than girls in the high activity group. Ordinal logistic regression showed that increased time spent in "productive sedentary behavior" (reading or doing homework and working on computers) was associated with increased physical activity (odds ratio, 1.7; 95% confidence interval, 1.2-2.4), as was time spent working (odds ratio, 1.3; 95% confidence interval, 1.2-1.4). Time spent watching television and playing video games was not associated with decreased physical activity.

Conclusions  Physical activity was not inversely associated with watching television or playing video games, but was positively associated with productive sedentary behavior and part-time work. Some students appear capable of managing their time better than others. Future studies should explore the ability of students to manage their time and also determine what characteristics are conducive to better time management.

PHYSICAL ACTIVITY is an integral part of a healthy lifestyle and has been associated with many health benefits.1,2 A well-documented decline in sports participation occurs during adolescence,3-5 and since physical activity may track from adolescence to adulthood,5-7 there is concern that morbidity and mortality associated with physical inactivity will increase.8-10

Determinants of physical activity have been studied in adolescents and include behavioral attributes such as sensation seeking, history of physical activity, involvement in community sports, and sedentary pursuits after school and on weekends (negative association); physical environment factors such as opportunities for exercise; and demographic and biological factors such as sex, age, and ethnicity.11 Although one would expect sedentary behavior to be at one end of the physical activity continuum, it appears that the determinants of "activity" and of "inactivity" differ substantially.12,13

Determinants of sedentary behavior have been studied far less but appear to be related to sociodemographic factors (eg, socioeconomic status [SES]) and possible psychosocial variables (eg, body image dissatisfaction in females, and quality of life and school performance in males)14 as well as depressive symptoms.13 Most investigations of the relationship between physical activity and sedentary behavior have focused solely on television watching, occasionally including video game playing.11 While the association between physical activity and television watching is equivocal,14-19 that between physical activity and other types of sedentary behaviors is largely unexplored. Although Strauss et al20 reported that time spent watching television and doing computer activities were both inversely correlated with moderate-level physical activity, these were only examined separately univariately; in multivariate regression analyses, a combined "sedentary behavior time" was created and was independently correlated with moderate activity level. While an all-encompassing indicator of "sedentary time" is intuitively appealing, different types of sedentary behaviors should likely not be collapsed into a single variable. In fact, there is evidence that not all sedentary activities are equal in value,21 and it is conceivable that children are far more willing to forfeit time for physical activity to watch television than they are to do their homework. In addition, time spent watching television or playing video games may be subject to restrictions or limited access, while time spent reading or doing homework is more likely to be entirely volitional in nature. Because of possible differences in the preference for and access to sedentary behaviors, we investigated the associations between physical activity and explicit types of behaviors independently.

Adolescents have reported time constraints22,23 to be the most important barrier to increasing their participation in physical activity. While adolescents have also reported having a job to be an important barrier to exercise,23 the association between work and physical activity is not clear. Notwithstanding work-related physical activity, time spent working after school and on weekends detracts from time spent in physical activity, but it may also increase access to physical activity, either by exposure to available facilities at the workplace, or by earning money to pay for activities that could not otherwise be afforded.

The objectives of this study were to determine whether there was a differential relationship between the time spent during physical activity and the time spent during different types of sedentary activity, including watching television, playing video games, working on computers, doing homework, and reading, taking into account the effect that part-time work may have on students' residual time. Other factors included were mental health score and musculoskeletal pain, as these may also be associated with physical activity and sedentary pursuits.13,24,25


The study was a cross-sectional cohort design. The potential sample included 948 seventh- to 10th-grade high school students in Montreal, Quebec, from 3 schools—2 public inner-city schools and 1 smaller private school. Students completed a self-administered questionnaire addressing lifestyle, health, and psychosocial issues. They were also measured for height and weight. This study was approved by the Research and Ethics Committee of the Sir Mortimer B. Davis Jewish General Hospital, McGill University, Montreal, and the Research Committee of the Montreal Catholic School Commission. In accordance with the laws of the province of Quebec, all students and parents of those younger than 14 years provided a signed informed consent before entry into the study.

Information on the outcome, physical activity, was collected by having the students mark any activity in which they had participated in the past 6 months, using a predetermined list. They indicated how long, on average, they spent in the activity (<5, 5-10, or >10 h/wk for each activity). These activities included badminton, baseball, basketball, bicycling, diving, football, handball, hockey, gymnastics, martial arts, racquetball, ringuette, in-line skating, ice skating, skiing, squash, soccer, swimming, tennis, volleyball, dance, and a category for "other (specify)." We classified physical activity into 3 appropriately sized categories based on number of subjects and face validity: low activity, moderate activity, and high activity. Subjects who engaged in low activity included those who participated in 0 to 2 activities, each less than 5 h/wk. Subjects who engaged in moderate activity included those who participated in 1 or 2 activities, 5 to 10 h/wk in each activity, or those who were active in 3 or more activities less than 5 h/wk for each of the 3 or more activities. Highly active individuals were those who participated in at least 1 activity more than 10 h/wk for each activity or in 3 or more activities 5 to 10 h/wk for each activity. To control for different intensities of exercise, a second classification of activity was used in which we categorized sport activities according to their respective metabolic equivalent levels.26 We used the mean number of hours for a particular category as the duration (eg, 5-10 h/wk was considered 7.5 h/wk).

The independent variables were defined as the average number of hours per day, during the past 6 months, spent in sedentary pursuits: watching television, playing video games, working on the computer, and doing homework or reading. This information was collected separately for weekdays and weekends. We also dichotomized these variables into "productive" (increasing knowledge that will help students improve their education and awareness, ie, working on the computer, reading or doing homework) and "leisure" (lacking the gains noted for productive activity, ie, watching television and playing video games) activities.

Covariates included time spent working, having substantial musculoskeletal pain, mental health status, and type of school (public vs private). For work during the past 6 months, a score was given for each type of work depending on how many hours per week. A score of 1 meant 1 to 10 h/wk, a score of 2 meant 11 to 20 h/wk, and a score of 3 meant more than 20 h/wk. The scores for each type of work were then summed into one continuous variable for work. Musculoskeletal pain was defined as having pain in the neck, shoulder, arm, hand, back, hip, knee, leg, foot, or ankle at a frequency of at least once a week for some period within the past 6 months.27,28 Mood and anxiety during the previous week were determined by the 5-item Mental Health Index from the Medical Outcomes Study 36-Item Short-Form Health Survey.29 The variable school type (public vs private) was used as a proxy for SES.

Analysis consisted of descriptive statistics and regression modeling (ordinal and linear). We used ordinal regression when we categorized the outcome variable as ordered levels of physical activity, and used the covariates discussed in the previous paragraph in the model. This method assumes equal steps between categories and the assumption can be tested. To test the sensitivity of our analysis, we also defined our outcome as continuous (using total hours per week of activity and total metabolic equivalents per week) and performed a linear regression using the same covariates as the ordinal regression model. All analyses were done with the SAS for Windows version 8 software (SAS Institute Inc, Cary, NC).


Of the 948 students who were eligible to participate in the study, 743 (78.4%) consented and were present on the day of testing. Mean age was 15.1 years (SD, 1.2); 48.3% were female, and 43.7% had experienced musculoskeletal pain in the past 6 months (45.4% in girls and 42.2% in boys). Smokers composed 30.3% of the group (equal for girls and boys), and 62.2% had worked in the past 6 months (75.2% of girls and 5.00% of boys). Median body mass index (calculated as weight in kilograms divided by the square of height in meters) was 20.9 (mean [SD] was 21.6 [4.1] overall, 21.9 [4.2] in girls, and 21.4 [3.9] in boys).

Physical activity levels are illustrated in Figure 1. There were 244 students in the low-activity category, 325 in the moderate-activity group, and 174 who reported being highly active. There were more girls than boys in the low- and moderate-activity groups and significantly more boys than girls in the high-activity group. Mean time spent per day in sedentary pursuits was similar for weekdays and weekends with the exception of reading or doing homework (Table 1). We constructed models for weekday and weekends separately. Because our results were similar for both, we report only on weekday sedentary values.

Figure 1.
Level of physical activity by sex (±SE).

Level of physical activity by sex (±SE).

Table 1. 
Descriptive Statistics
Descriptive Statistics

Unadjusted comparisons suggest that time spent in all sedentary pursuits under study actually increased slightly as activity level increased (Figure 2). Boys spent more time playing video games but less time doing homework than girls (data not shown).

Figure 2.
Time spent in sedentary pursuits by activity level (±SE).

Time spent in sedentary pursuits by activity level (±SE).

The multivariate model (ordinal logistic regression, outcome: high, moderate, and low physical activity; explanatory variables: sedentary behaviors entered individually) showed that neither musculoskeletal pain nor mental health was associated with physical activity. Working was associated with increased physical activity for both girls and boys. Students who attended private school tended to be more physically active than those in the public schools (Table 2). Only 95 students in the study attended the private school. Physical activity was higher in the private school students (51.6% vs 42.7% in the moderate physical activity category, and 29.5% vs 22.8% in the highly active group), but there were no differences in time spent working (P = .60).

Table 2. 
Ordinal Logistic Regression for Physical Activity, Stratified by Sex*
Ordinal Logistic Regression for Physical Activity, Stratified by Sex*

When the sedentary variables were categorized as leisure vs productive in a logistic regression model, the productive sedentary behavior variable was associated with increased physical activity (odds ratio, 1.7; 95% confidence interval, 1.2-2.4), whereas the leisure sedentary behavior variable was not (odds ratio, 1.1; 95% confidence interval, 0.8-1.5) (Table 3). In this model, musculoskeletal pain was associated with physical activity (odds ratio, 1.5; 95% confidence interval, 1.1-2.0). These associations were similar for both sexes. Private school students spent more time in productive sedentary behavior than public school students did (P<.001) and less time in leisure sedentary behavior (P<.001).

Table 3. 
Ordinal Logistic Regression for Physical Activity
Ordinal Logistic Regression for Physical Activity

Linear regression analyses using continuous outcomes of activity, either total activity per week or total metabolic equivalent values of activity per week, yielded the same conclusions (data not shown).


We found that physical activity was not associated with watching television or playing video games. However, time spent working on the computer was positively associated with physical activity.

The literature is divided on the issue of television watching and its association with physical activity. Pate et al30 found that low activity was associated with greater television watching; however, this association appeared to be a reflection of the association between television watching with ethnocultural factors. Andersen et al15 did not find an association between television viewing time and physical activity, but reported that lower physical activity was associated with higher body mass index. Katzmarzyk et al17 reported no association between time spent watching television and physical activity in teenaged boys and girls, although they noted that the prevalence of obesity was very low in their sample. Fotheringham and associates31 reported that young adults in the highest tertile of computer use tended to be inactive physically; however, their sample population was considerably older than the adolescents in the present study.

When individuals were categorized as engaging in "productive sedentary behavior" vs "leisure sedentary behavior," we found that the former was significantly associated with physical activity, while the latter was not. These results were the same for both sexes and may reflect similarity to the results of a recent American study, where higher perceived academic rank was associated with higher levels of physical activity.13 This suggests that adolescents who spend time "learning" and engaging in "academic" pursuits are more likely to make time for physical activity. In contrast, time spent in more "nonacademic" pursuits, like watching television and playing video games, was not associated at all with time spent doing physical activity. Because some adolescents clearly do manage to "do it all," a true lack of time as a barrier to physical activity22 is likely more a perception or an excuse than a reason. Rather, it may be that teens who make time for doing homework and reading (ie, are more productive during their leisure time) are better able to manage time or less likely to lose time by sleeping in late, talking on the telephone, etc. This may also be supported by the fact that those who worked tended to be more physically active as well. Another possibility is that those who are not physically active choose not to be, regardless of whether they watch television. As such, motivational strategies are necessary to encourage adolescents to participate in physical activity. More research is needed to identify reasons for these discrepancies.

When we stratified on school type (public vs private), it appeared that those who attended private school were more physically active despite the fact that they also spent more time working on the computer and doing homework or reading than their public school counterparts. Because there is no evidence that these students enjoy more residual free time, some adolescents seem to be able to manage their time better than others, or perhaps have a greater desire to enjoy a variety of both active and more sedentary pursuits.

All students included in the study participated in physical education, and although some had musculoskeletal pain, it did not prevent them from participating in physical education class. Our findings of an association between pain and physical activity in the model that combined sedentary behaviors into 2 categories may indicate that this pain may not have deterred students from participating in voluntary extracurricular physical activities either. The lack of association between mental health score and physical activity may reflect the fact that the students in the present study were all slightly active by virtue of the fact that they participated in physical education class at school, and perhaps those with poorer mental health scores were the ones who did not participate in physical education class.

Among its limitations, this study was cross-sectional. The intent was to explore associations between sedentary pursuits and physical activity in a cohort of high school students. It is not possible to determine causation under such a design.

There may have been some selection bias due to nonparticipation. The proportion of students who reported being physically active in at least 1 sport was higher in our cohort than in 2 other studies involving adolescents28,32; however, we measured teens during their physical education class. Those who did not attend physical education or who declined to participate in our study may have been those who were more inactive.

Another possible source of bias is misclassification. Data were based on self-reports. Although it is possible that individuals with a poor or inflated sense of time would be a source of spurious associations, this would not explain the differential associations of physical activity with sedentary behaviors found in this study. If it is present, such a bias would act on all reported behaviors; there is little reason to suspect that this bias would be present for some sedentary activities and not for others. The questionnaire obtained information about such factors as physical activity and employment that occurred within the previous 6 months. The validity of physical activity surveys in adolescents that addressed past-year physical activity has been shown to be good.33 In that study, the average of four 7-day recalls of activity was used as the gold-standard comparison with the "past-year" questionnaire. Spearman correlation coefficients ranged between 0.55 and 0.83, indicating a fairly good correlation between 7-day recall of activity and past-year recall. The present study's use of a 6-month recall time frame should be no worse (and possibly better) than the past-year scenario. Alternative methods of measuring physical activity include accelerometers and heart rate monitors. Overall, we believed that a survey was the better method for our study, given that our objective was to measure all types of physical activity (including swimming and cycling that would not be measured by conventional accelerometers) in a cohort of 743 inner-city adolescents during several months. As for employment, the proportion of adolescents who reported that they had worked was similar to that found by a study in Minnesota, where 58% of 10th graders reported working, with an average age at first employment of 14.7 years.34

We used a 5-item Mental Health Index for mood and anxiety. Although the scale contains only 5 items, this score is as good as the 18-item Mental Health Inventory and the 30-item General Health Questionnaire, and better than the 28-item Somatic Symptom Inventory, for the detection of depression, affective disorders, and anxiety.35

It is possible that some unmeasured confounding variable could bias the results. One possible confounder is SES.13 Although school type is an imperfect proxy variable for SES, the public schools were in lower-SES neighborhoods, whereas the private school was in a high-SES district. Moreover, a stratified analysis based on school (public vs private) as a proxy for SES did not affect our conclusions.

Finally, we did not measure all types of nonphysical activity (eg, talking on the telephone, going to movies, shopping). Although adolescents spend time on these types of activities as well, our objective was to describe the relationship between physical activity and time spent in the 4 specific sedentary pursuits (watching television, playing video games, working on computers, and doing homework or reading) most commonly hypothesized as being associated with physical activity.


Our findings indicate that physical activity was not inversely associated with watching television or playing video games. On the other hand, students who spent more time in productive sedentary behaviors (working on computers and reading or doing homework) tended to be more physically active. This may imply that certain teens are more capable of managing their time to include both physical activity and sedentary pursuits, or that those who are not physically active choose not to be, regardless of whether they watch television. As such, reducing television viewing may not be enough with respect to increasing physical activity. Time management skills or motivational strategies may be necessary to encourage participation in physical activity. Future research should address time management skills of students, developing profiles suitable for screening those at risk for poor time management and interventions to improve these skills in the hope of promoting increased participation in physical activity.


Corresponding author: Debbie Ehrmann Feldman, PhD, École de Réadaptation, Université de Montréal, C.P. 6128 Succ. Centre Ville, Montréal, Québec, Canada H3C 3J7 (e-mail: Debbie.Feldman@umontreal.ca).

Accepted for publication January 23, 2003.

Dr Feldman is the recipient of a network scholar research award by the Canadian Arthritis Network, Toronto, Ontario. Dr Shrier is the recipient of a clinician researcher award by the Fonds de la Recherche en Santé du Québec, Montréal, Québec. Ms Barnett is the recipient of a doctoral award by the Canadian Institutes of Health Research, Ottawa, Ontario.

What This Study Adds

Physical activity in adolescence is important for a healthy lifestyle and may positively affect adult health. Many adolescents reportedly spend too much time watching television and playing video games and not enough time on physical activity. In this study, the results indicate that physically active adolescents were not less involved in these sedentary pursuits. Moreover, those who spent more time on "productive sedentary behavior" (doing homework, reading, and working on the computer) were actually more physically active. This may imply that certain teens are more capable of managing their time to include both physical activity and sedentary pursuits or that those who are not physically active choose not to be, regardless of whether they watch television. As such, reducing television viewing may not be enough with respect to increasing physical activity. Time management skills or motivational strategies may be necessary to encourage participation in physical activity.

Fletcher  GFBalady  GBlair  SN  et al.  Statement on exercise: benefits and recommendations for physical activity programs for all Americans: a statement for health professionals by the Committee on Exercise and Cardiac Rehabilitation of the Council on Clinical Cardiology, American Heart Association.  Circulation. 1996;94857- 862PubMedGoogle ScholarCrossref
Steptoe  AButler  N Sports participation and emotional wellbeing in adolescents.  Lancet. 1996;3471789- 1792PubMedGoogle ScholarCrossref
Caspersen  CJPereira  MACurran  KM Changes in physical activity patterns in the United States, by sex and by cross-sectional age.  Med Sci Sports Exerc. 2000;321601- 1609PubMedGoogle ScholarCrossref
Kimm  SYGlynn  NWKriska  AM  et al.  Longitudinal changes in physical activity in a biracial cohort during adolescence.  Med Sci Sports Exerc. 2000;321445- 1454PubMedGoogle ScholarCrossref
Raitakari  OTPorkka  KVKTaimela  STelama  RRasanen  LViikari  JSA Effects of persistent physical activity and inactivity on coronary risk factors in children and young adults.  Am J Epidemiol. 1994;140195- 205PubMedGoogle Scholar
Kelder  SHPerry  CLKlepp  KILytle  LL Longitudinal tracking of adolescent smoking, physical activity and food choice behaviors.  Am J Public Health. 1994;841121- 1126PubMedGoogle ScholarCrossref
Telama  RXiaolin  YLaakso  LViikari  J Physical activity in childhood and adolescence as predictor of physical activity in young adulthood.  Am J Prev Med. 1997;13317- 323PubMedGoogle Scholar
Andersen  LBSchnohr  PSchroll  MHein  HO All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work.  Arch Intern Med. 2000;1601621- 1628PubMedGoogle ScholarCrossref
Manson  JEHu  FBRich-Edwards  JW  et al.  A prospective study of walking as compared to vigorous exercise in the prevention of coronary disease in women.  N Engl J Med. 1999;341650- 658PubMedGoogle ScholarCrossref
Paffenbarger Jr  RSHyde  RTWing  ALHsieh  CC Physical activity, all-cause mortality, and longevity of college alumni.  N Engl J Med. 1986;314605- 613PubMedGoogle ScholarCrossref
Sallis  JFProchaska  JJTaylor  WC A review of correlates of physical activity of children and adolescents.  Med Sci Sports Exerc. 2000;32963- 975PubMedGoogle ScholarCrossref
Gordon-Larsen  PMcMurray  RGPopkin  BM Adolescent physical activity and inactivity vary by ethnicity: the National Longitudinal Study of Adolescent Health.  J Pediatr. 1999;135301- 306PubMedGoogle ScholarCrossref
Schmitz  KHLytle  LAPhillips  GAMurray  DMBirnbaum  ASKubik  MY Psychosocial correlates of physical activity and sedentary leisure habits in young adolescents: the Teens Eating for Energy and Nutrition at School Study.  Prev Med. 2002;34266- 278PubMedGoogle ScholarCrossref
Williams  CDSallis  JFCalfas  KJBurke  R Psychosocial and demographic correlates of television viewing.  Am J Health Promot. 1999;13207- 214PubMedGoogle ScholarCrossref
Andersen  RECrespo  CJBartlett  SJCheskin  LJPratt  M Relationship of physical activity and television watching with body weight and level of fatness among children.  JAMA. 1998;279938- 942PubMedGoogle ScholarCrossref
Bungum  TLVincent  ML Determinants of physical activity among female athletes.  Am J Prev Med. 1997;13115- 122PubMedGoogle Scholar
Katzmarzyk  PTMalina  RMSong  TMKBouchard  C Television viewing, physical activity and health-related fitness of youth in the Quebec Family Study.  J Adolesc Health. 1998;23318- 325PubMedGoogle ScholarCrossref
O'Loughlin  JParadis  GKishchuk  NBarnett  TRenaud  L Prevalence and correlates of physical activity behaviors among elementary schoolchildren in multiethnic, low income, inner-city neighbourhoods in Montreal, Canada.  Ann Epidemiol. 1999;9397- 407PubMedGoogle ScholarCrossref
Sallis  JF Age-related decline in physical activity: a synthesis of human and animal studies.  Med Sci Sports Exerc. 2000;321598- 1600PubMedGoogle ScholarCrossref
Strauss  RSRodzilsky  DBurack  GColin  M Psychosocial correlates of physical activity in healthy children.  Arch Pediatr Adolesc Med. 2001;155897- 902PubMedGoogle ScholarCrossref
Epstein  LHRoemmich  JN Reducing sedentary behavior: role in modifying physical activity.  Exerc Sport Sci Rev. 2001;29103- 108PubMedGoogle ScholarCrossref
Allison  KRDwyer  JJMMakin  S Perceived barriers to physical activity among high school students.  Prev Med. 1999;28608- 615PubMedGoogle ScholarCrossref
Tappe  MKDuda  JLEhrnwald  PM Percieved barriers to exercise among adolescents.  J Sch Health. 1989;59153- 155PubMedGoogle ScholarCrossref
Norris  RCarroll  DCochrane  R The effects of physical activity and exercise training on psychological stress and well-being in an adolescent population.  J Psychosom Res. 1992;3655- 65PubMedGoogle ScholarCrossref
Newcomer  KSinaki  M Low back pain and its relationship to back strength and physical activity in children.  Acta Paediatr. 1996;851433- 1439PubMedGoogle ScholarCrossref
Ainsworth  BEHaskell  WLWhitt  MC  et al.  Compendium of physical activities: an update of activity codes and MET intensities.  Med Sci Sports Exerc. 2000;32(9, suppl)S498- S516PubMedGoogle ScholarCrossref
Brattberg  G Back pain and headache in Swedish schoolchildren: a longitudinal study.  Pain Clin. 1993;6157- 162Google Scholar
Salminen  JJ The adolescent back: a field survey of 370 Finnish schoolchildren.  Acta Paediatr Scand Suppl. 1984;3158- 122PubMedGoogle Scholar
McHorney  CAWare  JERaczek  AE The MOS 36-Item Short-Form Health Survey (SF-36), II: psychometric and clinical tests of validity in measuring physical and mental health constructs.  Med Care. 1993;31247- 263PubMedGoogle ScholarCrossref
Pate  RRHeath  GWDowda  MTrost  SG Associations between physical activity and other health behaviors in a representative sample of US adolescents.  Am J Public Health. 1996;861577- 1581PubMedGoogle ScholarCrossref
Fotheringham  MJWonnacott  RLOwen  N Computer use and physical inactivity in young adults: public health perils and potentials of new information technologies.  Ann Behav Med. 2000;22269- 275PubMedGoogle ScholarCrossref
Allison  KRAdlaf  EM Age and sex differences in physical inactivity among Ontario teenagers.  Can J Public Health. 1997;88177- 189PubMedGoogle Scholar
Aaron  DJKriska  AMDearwater  SRCauley  JAMetz  KFLaPorte  RE Reproducibility and validity of an epidemiologic questionnaire to assess past year physical activity in adolescents.  Am J Epidemiol. 1995;142191- 201PubMedGoogle Scholar
Parker  DLCarl  WRFrench  LRMartin  FB Nature and incidence of self-reported adolescent work injury in Minnesota.  Am J Ind Med. 1994;26529- 541PubMedGoogle ScholarCrossref
Berwick  DMMurphy  JMGoldman  PAWare  JEBarsky  AWeinstein  MC Performance of a five-item mental health screening test.  Med Care. 1991;29169- 176PubMedGoogle ScholarCrossref