Psychosocial and Environmental Correlates of Adolescent Sedentary Behaviors
http://www.100md.com
《小儿科》
Department of Family and Preventive Medicine, University of California, San Diego, La Jolla, California
Department of Psychology, San Diego State University, San Diego, California
ABSTRACT
Objective. To determine correlates of sedentary behaviors in adolescents through the examination of psychosocial and environmental variables.
Method. The study used a cross-sectional design to evaluate an ethnically diverse clinic-based sample of 878 adolescents who were 11 to 15 years old. Bivariate and multivariate analyses were stratified by gender to assess correlates of sedentary behaviors occurring on the most recent nonschool day (television viewing, computer video games, sitting listening to music, and talking on the phone).
Results. For girls, age, family support, television/video rules, and hills in the neighborhood were associated with sedentary behaviors. Furthermore, psychological constructs such as self-efficacy, enjoyment, change strategies, and pros and cons of change emerged as correlates of sedentary behaviors. A moderator effect revealed that the proportion of girls in the low-BMI group decreased with increased self-efficacy, whereas the proportion of girls in the high-BMI group did not vary significantly by self-efficacy. For boys, age, ethnicity, BMI, cons of change, and self-efficacy were associated with sedentary behaviors.
Conclusions. This study provides evidence of factors associated and not associated with adolescent sedentary behaviors. Similar to physical activity, measures of specific psychosocial constructs of sedentary behavior demonstrated important associations. The results highlight the need for additional examination of the correlates of sedentary behavior to determine which correlates are mechanisms of behavior change.
Key Words: determinants behavioral theory television sedentary behavior adolescents
Abbreviations: TV, television
Children in the United States spend an estimated 75% of the day being inactive.1 However, the relationships among sedentary behaviors, physical activity levels, and overweight are only beginning to be understood for the adolescent population. Sedentary behaviors, particularly watching television (TV) and videos, have been found to be related to higher body mass index (BMI) for children and adolescents.2–6 The literature linking physical activity levels with risk of overweight in adolescents is not consistent,7, 8 but physical activity is an important component of effective obesity treatments.9 The relationship between sedentary behaviors and physical activity has been inconsistent across studies.4, 5, 10 Determining specific correlates of sedentary behavior may help to clarify these relationships.
Experimental studies provide compelling evidence that reducing time spent in sedentary behaviors can be an effective weight-loss strategy for youth independent of changes in physical activity level.6, 9, 11–13 Sedentary behaviors, particularly TV viewing and playing computer games, are likely independent risk factors for children's overweight status. The likely mechanism for TV viewing's effect on overweight is its effect on energy balance. That is, TV viewing affects both energy expenditure (less time being physically active) and energy intake (overeating while watching TV and exposure to commercials promoting less-than-healthy foods).2, 14
Individual and population-based interventions that target modifiable correlates of behaviors are more likely to lead to behavior change.15 However, although much is known about the correlates of physical activity for adolescents,16 there has been little research on correlates of sedentary behaviors. One study of a large sample of seventh-graders found that higher depressive symptoms and lower perceived academic rank were associated with more sedentary-behavior time for boys and girls.17 A study of university students found that psychosocial correlates of physical activity were poorly correlated with TV viewing,18 suggesting that sedentary-specific psychological variables are needed. Gordon-Larsen et al19 concluded that physical activity and sedentary behaviors were associated with different determinants. They found that physical activity related most to environmental factors, whereas sedentary behaviors related more to sociodemographic factors.
The available evidence indicates that sedentary behaviors are not the mirror image of physical activity and that the 2 types of behaviors are relatively uncorrelated.20 Excessive sedentary behavior is a health risk, but the correlates of adolescent sedentary behaviors are not clear. The present study examined theory-derived psychological, social, and environmental factors associated with sedentary behaviors in adolescents. Determining the correlates of sedentary behaviors is the first step in understanding the mechanisms that influence these behaviors that can then be incorporated into intervention programs.21
METHODS
Participants
Adolescents between the ages of 11 and 15 years were recruited through their primary care providers as part of a health-promotion intervention trial. A total of 45 primary care providers from 6 clinic sites in San Diego County, California, agreed to participate. Over a 13-month period, trained recruiters attempted to contact 3366 households (including wrong numbers, those not eligible, and refusals), from which 878 adolescents (64% of eligible contacts) were enrolled onto the study after signing consent forms and completing baseline measurements. Adolescents received $10 for completing all measurements and were entered into a lottery drawing for 1 of 10 cash prizes ranging between $10 and $50. All study procedures were approved by university and clinic institutional review boards. Table 1 presents the sample characteristics for boys and girls. The sample was ethnically diverse, with 42% of participants from minority backgrounds.
Procedure
Items analyzed in the current study were part of a larger battery of survey questions completed at the baseline assessment. Adolescents completed measures on a desktop computer by using a Web-based interface. Responses to items were made by clicking on radio buttons using a computer mouse. Items were displayed on the computer screen with the directions and response scales presented at the top of each screen. The parent who accompanied the adolescent completed several survey instruments about the adolescent's home and neighborhood. These surveys were completed in a paper-and-pencil format, and response values were entered into a database by trained research assistants.
Measures
Adolescents completed measures pertaining to their behavior, attitudes, and beliefs related to being sedentary. The constructs assessed were based on social cognitive theory,22 the transtheoretical model,23 and environmental variables derived from ecological models.24 Parents completed surveys regarding home and neighborhood environments and their support for their child's physical activity. Unless otherwise stated below, reported internal consistency coefficients were calculated in the current study. All of the survey measures are available (through G.J.N.) at www.paceproject.org.
Height and Weight
An Accu-Hite wall stadiometer model 216 (Seca Accu-Hite, Hanover, MD) measured standing height. Weight was measured with the digital Body Comp scale from American Weights & Measures (Rancho Santa Fe, CA). Each measure was taken twice, and the average of the 2 readings was calculated. BMI was calculated as kilograms per square meter. BMI-for-age percentile was determined from Centers for Disease Control and Prevention national norms using age to the nearest month and gender-specific median, standard deviation (SD), and power of the Box-Cox transformation.25 BMI-for-age percentiles were split at the 85th percentile, which is defined as at risk for overweight.
Sedentary Behaviors
Self-reported sedentary behaviors were assessed with a survey adapted from Robinson.13 Participants were asked how much time they spent doing the following leisure-time sedentary behaviors: watching TV (including videos on VCR/DVD); playing computer or video games (such as Nintendo or Sega); sitting and listening to music on the radio, audiotapes, or CDs; and sitting and talking on the telephone. Questions were asked first for "most recent day when you were not in school" and then for the "most recent school day." Participants responded to each item by using a 9-point scale with anchors of "none," "15 minutes or less," "30 minutes," "1 hour," "2 hours," "3 hours," "4 hours," "5 hours," or "6 hours or more."
An index of sedentary-behavior time was computed by summing the 4 items for nonschool days. Item responses correlated highly between nonschool day and school day (r = 0.83). Because a nonschool day was likely to include more unstructured time for an adolescent, allowing for more time engaging in sedentary behaviors, the sum of the nonschool-day responses was used as the outcome variable. Internal consistency for this scale was moderate, with = .59. Test-retest reliability was determined from a separate sample to be sufficient (intraclass correlation coefficient: .77).
We dichotomized sedentary time into (0) 240 minutes and (1) >240 minutes. Given that Healthy People 201026 and the American Academy of Pediatrics27 recommend <2 hours of TV viewing per day, we reasoned that 4 hours per day would be a logical extension when considering several more types of sedentary behaviors. This cut point was also the median split for the sample, which maximizes the variable's potential to correlate with the predictor variables. Validity for the dichotomized sedentary-behavior index was demonstrated with significant coefficients for percent body fat (r = 0.13; P < .001), aerobic fitness (r = –0.11; P < .001), and average daily minutes of vigorous physical activity measured with accelerometers (r = –0.17; P < .001).
Behavior-Change Strategies
Fifteen items were created that reflect thoughts, activities, and feelings that people may use when making a behavior change. These change strategies were similar to constructs described as processes of change in the transtheoretical model28 and were based on items developed by Saelens et al.29 Item examples included "thinking about how surroundings affect the amount of sedentary time" and "overcoming hassles to reduce sedentary time." The response format assessed how often each strategy was used, with a 5-point Likert scale ranging from 1 (never) to 5 (many times). Norman et al30 determined test-retest reliability for the scale to be .75. Internal consistency for the 15 items was = .91.
Pros and Cons of Change
Decisional balance consists of 2 constructs labeled the "pros" and "cons" of change that address cognitive and motivational aspects of human decision-making.31 Norman et al32 developed a 4-item pro and 5-item con scale of reducing sedentary behaviors and determined that the 2-factor structure demonstrated good internal validity. Pros included: "my parents would be pleased if I spent less time playing computer/video games" and "playing computer/video games hurts my eyes." Cons of change included: "I enjoying playing computer/video games for many hours" and "watching TV or playing computer games is my way to escape from the world." Participants rated the importance of each item by using a 5-point Likert scale with anchors of 1 (not at all important) to 5 (extremely important). Test-retest reliability for the pro and con scales were .86 and .77, respectively. Internal consistency was = .60 for the pros and = .76 for the cons.
Self-Efficacy
Situational self-efficacy represents a person's confidence about meeting a behavioral criterion in situations that may present barriers to the behavior. A 7-item sedentary-behaviors self-efficacy scale assessed adolescents' confidence to reduce the amount of time that they spend being sedentary (eg, plan ahead of time what TV shows you will watch during the week).32 Participants responded to each item on a 5-point Likert scale ranging from 1 (I'm sure I can't) to 5 (I'm sure I can). Test-retest reliability for this scale was .81,30 and internal consistency was = .81.
Family Support
Four questions assessed adolescents' perceptions of how often a household member gave different types of support: (1) encourages you to spend less time being sedentary; (2) discusses with you how sedentary habits can be unhealthy; (3) helps you think of ways to reduce the time you spend on sedentary habits; and (4) tells you that you are doing a good job reducing your sedentary habits. The 5-point scale ranged from "never" to "every day." These items were adapted from a family-support scale that was developed by Prochaska et al33 that had good internal consistency and validity with self-reported physical activity. Test-retest reliability for the 4-item sedentary-behaviors family-support scale was found to be good (intraclass correlation coefficient: .72),30 and internal consistency was = .86.
Enjoyment of Sedentary Behaviors
Adolescents rated a single item, "I enjoy doing sedentary habits like watching TV or playing computer/video games," on a 5-point scale ranging from "strongly disagree" to "strongly agree." Similar measures were used in large population-based studies and were found to have good reliability and validity.34
TV and Video Household Rules
Two items assessed limits set by parents on (1) hours of TV/Video viewing and (2) hours of video/computer games. The 4-point response format ranged from "never" to "always." Internal consistency for the scale was = .87.
Parent-Reported Support for Physical Activity
Items from the Amherst Health and Activity Study Questionnaire35 assessed parents' reported level of support for their child's physical activity during a typical week. Four items formed a composite measure with an internal consistency of = .75: (1) watching your child participating in physical activity; (2) encouraging your child to do physical activity; (3) doing physical activity with child; and (4) providing transportation so your child can get to a place where he or she can do physical activity. The 5-point scale for these items ranged from "never" to "every day."
Home Environment
Sixteen items assessed the presence of items in the home that would facilitate physical activity. Sallis et al36 found this scale to have a high test-retest reliability (.89) and to be correlated with self-reported physical activity. Adolescent parents responded "yes" or "no" for whether each item was in the home, yard, apartment complex, or community. The number of "yes" responses was totaled to form a single score. Items included bicycle, dog, swimming pool, sports equipment, skis, and toning devices. Internal consistency for the scale in the present study was = .59.
Neighborhood Environment Variables
Parents responded to questions about their neighborhoods on the Neighborhood Environment Walkability scale, which was developed recently by Saelens et al.37 All of the summated scales were found to have test-retest reliability and, with the exception of the Safety From Crime scale, differentiated individuals in a highly walkable neighborhood compared with those in a neighborhood that is not particularly walkable. The Neighborhood Environment Walkability scale subscales used in the present study are described below.
Land-Use Mix/Diversity
We used 14 of the 16 items from this subscale that assess the extent to which stores and facilities are in walking distance of a participant's home. Participants rated each location on a 5-point scale from "1–5 minutes" to "31+ min." Nine items were used to form a "retail in neighborhood" scale that included grocery store, supermarket, hardware store, fruit/vegetable market, laundry/dry cleaner, clothing store, other stores, post office, and library. Five items formed a "recreation in neighborhood" scale and included elementary school, other schools, park, recreation center, gym, and fitness facility. Internal consistency for these scales was = .90 and .79, respectively.
Walking and Cycling Facilities
Six items assessed proximity and quality of neighborhood trails and sidewalks for walking and bicycling. The 4-point response format for these items ranged from "strongly disagree" to "strongly agree." Internal consistency for this scale was = .74.
Hills in Neighborhood
Two items assessed the difficulty of getting around in the neighborhood by bicycle or on foot, considering the hilliness of the area. The 4-point response format for these items ranged from "strongly disagree" to "strongly agree." Internal consistency was = .76.
Pedestrian/Traffic Safety
Eight items assessed traffic speed and crosswalks in regard to impeding walking in the neighborhood. The 4-point response format ranged from "strongly disagree" to "strongly agree." Internal consistency for this scale was = .67.
Crime Safety
Six items assessed crime rate and street lighting in the neighborhood. The mean of 6 items on a 4-point scale ranging from "strongly disagree" to "strongly agree" was calculated. Internal consistency for this scale was = .68.
Analyses
All analyses were conducted separately for girls and boys. First, bivariate analyses of each independent variable with the dichotomous outcome variable sedentary behaviors determined unadjusted odds ratios. For these exploratory analyses, P values were considered statistically significant at .05 and were not adjusted for multiple tests. Second, multivariate logistic-regression models were built and estimated. Model building occurred in several blocks. The first block of variables forced into the model consisted of demographic variables including adolescents' age, ethnicity, BMI percentile, and highest adult household education level. For the next block, predictor variables were allowed to step into the model with the entry level set at P .10. In the third block, all first-order interaction terms of demographic x predictor variables already in the model were allowed to step into the model with an entry level set at P .05. The Hosmer-Lemeshow goodness-of-fit statistic assessed overall model fit. A small test statistic and a large P value (P > .10) indicated a model that provided a good fit to the data. The Nagelkerke R2 statistic estimated the percentage of variance accounted for by the model.
Additional analyses were conducted with school-day sedentary time and nonschool-day sedentary behavior as continuous variables. These analyses resulted in essentially the same set of statistically significant predictors and multivariate models and, in the interest of brevity, are not reported here.
RESULTS
Description of the Outcome Variable
Relationships Among Predictor Variables
Correlations among predictor variables were in the low to moderate range, with a similar pattern of relationships found for girls and boys. For the 120 correlations among the 16 predictor variables, there were only 3 pairs that were above r = 0.50. The highest correlations were between neighborhood recreation and neighborhood retail (girls: r = 0.65; boys: r = 0.64), change strategies and family support (girls: r = 0.57; boys: r = 0.62), and transportation and encouragement for physical activity (girls: r = 0.56; boys: r = 0.57) (all P < .01). Twelve correlation pairs were between r = 0.30 and 0.50, and the remaining correlations were below r = 0.30. This pattern of relationships among the predictors indicated that colinearity would not be a problem in the multivariate analyses.
Unadjusted Relationships for Girls
For girls, 8 bivariate associations between predictor variables and leisure-time sedentary behavior were statistically significant (Table 3). Older girls (13–15 years old) were more likely to be in the >240-minutes of sedentary time group than younger girls (OR: 1.69; 95% CI: 1.18–2.44). Higher scores on change strategies, pros, self-efficacy, and TV/video rules were related to decreased likelihood of being in the high-sedentary-behavior group. High scores on cons, enjoyment of sedentary behaviors, and hills in the neighborhood were related to increased likelihood of being in the high-sedentary-time group.
Unadjusted Relationships for Boys
For boys, 7 bivariate associations were also found between predictor variables and sedentary behavior (Table 4). Boys who were older, nonwhite, and in the >85th percentile BMI category were more likely to be in the high-sedentary-behavior group than boys who were younger, white, and in the <85th percentile BMI category. Higher scores on self-efficacy and TV/video rules were associated with decreased likelihood of being in the high-sedentary-behavior group. Higher scores on the cons and enjoyment of sedentary behaviors were associated with decreased likelihood of being in the high-sedentary-behavior group.
Final Multivariate Model for Girls
The final multivariate logistic model for girls included all of the variables that were associated with the outcome from the unadjusted bivariate analyses. The model-building process resulted in the inclusion of 1 interaction for BMI percentile and self-efficacy. To simplify presentation of the results, Table 5 presents the main-effects model parameters. The R2 for the main-effects model was 0.25, and the inclusion of the interaction term increased the R2 to 0.28. The Hosmer-Lemeshow test indicated that the fit of the model was good (P = .25). Figure 1 graphically displays the interaction effect by dividing self-efficacy into quartiles. Girls in the low-BMI category (<85th percentile) demonstrated a fairly linear relationship with increasing confidence scores related to decreased percentages of girls in the high-sedentary-behavior category. However, for girls in the high-BMI (85th percentile) group, there was little relationship between confidence and sedentary behavior.
Final Multivariate Model for Boys
The final multivariate logistic model for boys included age, BMI percentile, cons, and self-efficacy as significant correlates of sedentary time (Table 6). Ethnicity was forced into the model but no longer reached statistical significance (P = .088). The variables enjoyment of sedentary behaviors, TV/video home rules, and transportation for physical activity, which were significantly related to sedentary time in the unadjusted analyses, did not enter the model. The final model's R2 was 0.22, and the fit of the model was good (P = .35).
DISCUSSION
Sedentary behaviors are likely to continue to be pervasive in technologically developed societies, and the results of this study contribute to a greater understanding of the correlates of these behaviors in adolescence. Knowledge about nonmodifiable factors (eg, age) can be useful in developing an overall behavioral profile for an individual or population subgroup. Knowledge about modifiable correlates can be used to tailor health messages delivered by pediatricians and public health professionals. Some of the most meaningful findings are summarized below.
Demographic and Anthropometric Factors
Gender, age, ethnicity, and BMI were found to be related to sedentary-behavior time to varying degrees. Overall, although nearly half of all girls and boys reported spending >4 hours on a nonschool day engaged in sedentary behaviors, what contributed to the amount of sedentary-behavior time was different for each group. Girls reported more time spent sitting and listening to music and sitting talking on the phone, whereas boys reported more time playing computer games. These data have implications for what clinicians might discuss with parents who are concerned about their child's behavior, as well as for school and community-based groups developing interventions to address adolescent health.
Similar to findings from previous studies showing that physical activity decreases with increasing age in adolescence,16 older adolescents reported more sedentary time than younger adolescents. As adolescents move into their teenage years, they likely increase the time that they spend engaging in multiple sedentary behaviors that compete with physical activities. Older adolescents have more time that is not supervised by a parent or caregiver, which creates opportunities to increase TV-viewing time and other sedentary behaviors. Encouraging alternative physical activities or teaching ways to engage in physical activity while engaging in pastimes such as listening to music, talking on the phone, and watching TV may help adolescents to develop more active lifestyles.
Hispanic and other minority boys were more likely to be in the high-sedentary-time group than non-Hispanic white boys. This pattern is consistent with previous research that found large ethnic differences for inactivity and found boys to engage in more hours of sedentary behaviors than girls.38 The largest ethnic group represented in the sample was Hispanic adolescents, suggesting that there may be cultural factors contributing to these findings among Hispanic families.39 In any case, these findings suggest that practitioners hoping to effect changes in sedentary behavior among adolescents will need to target population approaches and tailor individual-level strategies based on localized ethnic and cultural factors.
BMI was expected to be related to sedentary time based on previous studies.6, 13 However, BMI status was only associated with sedentary behaviors for boys. In previous studies it was common to find inconsistent associations across subgroups,6 which perhaps reflects measurement error or sample-specific characteristics. A recent meta-analysis also found only a weak relationship between TV viewing and body fatness.40 The evidence suggests that simply trying to reduce adolescent TV viewing may not be the best intervention strategy and must be combined with strategies to reduce overall sedentary-behavior time and caloric intake.
Psychosocial Factors
Several psychosocial constructs were related to sedentary behavior, including the cons of reducing sedentary behaviors and self-efficacy for girls and boys and change strategies, pros, and family support for girls. These associations identify important modifiable factors that can be targeted in health-promotion interventions. For example, a pediatrician can ask about an adolescent's self-efficacy to reduce sedentary behaviors and then use this knowledge to provide brief, tailored counseling to help shape beliefs and provide encouragement to change.22
Environmental Factors
There in increasing interest in how environmental factors help shape physical activity and other health behaviors.24, 41 However, there was no evidence in this group of adolescents that environmental neighborhood factors such as land-use mix or safety from crime and traffic were associated with time spent engaging in sedentary behaviors, which is a finding that is consistent with others19 indicating that environmental factors were more related to physical activity and not sedentary behavior. We did find that girls living in more hilly neighborhoods reported more sedentary-behavior time. Living in a hilly neighborhood may discourage adolescent girls from outdoor activities such as walking and bicycling during a time in their lives when sedentary alternatives such as talking on the telephone are more desirable. This finding with adolescents contrasts with a positive association between hilly neighborhood and physical activity in older women,42 which highlights the need for additional studies focused on physical environment correlates in diverse populations.
Strengths of this study include the use of a large, ethnically diverse sample of adolescents, the inclusion of variables from multiple domains (from demographic to environmental levels), and the use of variables measured with established psychometric properties. Study limitations include the use of a cross-sectional design that limits the variable associations to the "correlates" level of evidence; however, this is a necessary preliminary step for understanding mechanisms of behavior change.21 The composite measure of sedentary behavior precluded the assessment of correlates of individual sedentary behaviors and the potential of engaging in multiple sedentary or active behaviors at the same time. However, the measure was appropriate for estimating cumulative sedentary time, which may be more important than looking at separate sedentary behaviors.40
The reliance on self-report data raises the possibility of response-set biases such as social desirability. Being sedentary may be considered an undesirable behavior. Ensuring confidentiality and anonymity of responses can help to minimize response-set biases. Using mainly summated scales rather than individual items in the analyses also helped to minimize bias and measurement error. The generalizability of the study results is limited to the age range of adolescents between 11 and 15 years old and adolescents whose families have medical insurance. These findings may not persist for older or younger samples of youth or for adolescents in families in low socioeconomic segments that do not have medical coverage.
This study provides initial evidence of factors associated and not associated with adolescent sedentary behavior. Examining potential correlates from intrapersonal, interpersonal, and environmental domains indicated that sedentary behavior was mainly associated with intrapersonal factors such as demographic characteristics and psychosocial constructs that specifically addressed sedentary behavior. The findings suggest that adolescent sedentary behavior has correlates that are distinct from correlates found for physical activity, and these correlates may provide the basis for programs that can be delivered through primary care settings43, 44 or in the public health arena. For researchers, the correlates found in this study can be useful for generating hypotheses to be tested in future studies. Additional studies with prospective designs are needed to determine the temporal ordering and dose-response relationship of modifiable correlates that are potential determinants of sedentary behavior.
ACKNOWLEDGMENTS
This project was supported by National Cancer Institute grants R01 CA81495 and R01 CA85873.
FOOTNOTES
Accepted Jan 4, 2005.
No conflict of interest declared.
REFERENCES
Strauss RS, Rodzilsky D, Burack G, Colin M. Psychosocial correlates of physical activity in healthy children. Arch Pediatr Adolesc Med. 2001;155 :897 –902
Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M. 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
Utter J, Neumark-Sztainer D, Jeffery R, Story M. Couch potatoes or french fries: are sedentary behaviors associated with body mass index, physical activity, and dietary behaviors among adolescents J Am Diet Assoc. 2003;103 :1298 –1305
Eisenmann JC, Bartee RT, Wang MQ: Physical activity, TV viewing, and weight in US youth: 1999 Youth Risk Behavior Survey. Obes Res. 2002;10 :379 –385
Saelens BE. Helping individuals reduce sedentary behavior. In: Andersen RE, ed. Obesity: Etiology, Assessment, Treatment, and Prevention. Champaign, IL: Human Kinetics; 2003:217 –238
Riddoch C. Relationships between physical activity and physical health in young people. In: Biddle S, Sallis JF, Cavill NA, eds. Young and Active Young People and Health Enhancing Physical Activity: Evidence and Implications. London, England: Health Education Authority; 1998:17 –48
Patrick K, Norman G, Sallis J, et al. The relative contributions of diet, physical activity, and sedentary behaviors as risk factors for overweight in adolescence. Arch Pediatr Adolesc Med. 2004;158 :385 –390
Epstein LH, Myers MD, Raynor HA, Saelens BE. Treatment of pediatric obesity. Pediatrics. 1998;101 :554 –570
Feldmann 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 :797 –802
Epstein LH, Valoski AM, Vara LS, et al. Effects of decreasing sedentary behavior and increasing activity on weight change in obese children. Health Psychol. 1995;14 :109 –115
Epstein LH, Paluch RA, Gordy CC, Dorn J. Decreasing sedentary behaviors in treating pediatric obesity. Arch Pediatr Adolesc Med. 2000;154 :220 –226
Robinson TN. Reducing children's television viewing to prevent obesity: a randomized controlled trial. JAMA. 1999;282 :1561 –1567
Dietz WH, Gortmaker SL. Preventing obesity in children and adolescents. Annu Rev Public Health. 2001;22 :337 –353
Baranowski T, Anderson C, Carmack C. Mediating variable framework in physical activity interventions. How are we doing How might we do better Am J Prev Med. 1998;15 :266 –297 [published correction appears in Am J Prev Med. 1999;17:98]
Sallis JF, Prochaska JJ, Taylor WC. A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc. 2000;32 :963 –975
Schmitz KH, Lytle LA, Phillips GA, Murray DM, Birnbaum AS, Kubik 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;34 :266 –278
Williams CD, Sallis JF, Calfas KJ, Burke RS. Psychosocial correlates of television viewing. Am J Health Promot. 1999;13 :207 –214
Owen N, Leslie E, Salmon J, Fotheringham MJ. Environmental determinants of physical activity and sedentary behavior. Exerc Sport Sci Rev. 2000;28 :153 –158
Bauman AE, Sallis JF, Dzewaltowski DA, Owen N. Toward a better understanding of the influences on physical activity: the role of determinants, correlates, causal variables, mediators, moderators, and confounders. Am J Prev Med. 2002;23 (2 suppl):5–14
Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall; 1986
Prochaska JO, DiClemente CC, Norcross JC. In search of how people change: applications to addictive behaviors. Am Psychol. 1992;47 :1102 –1114
Sallis JF, Owen N. Ecological models of health behavior. In: Glanz K, Rimer BK, Lewis FM, eds. Health Behavior and Health Education: Theory, Research, and Practice. 3rd ed. San Francisco, CA: Jossey-Bass; 2002
Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC Growth Charts: United States Advance Data From Vital and Health Statistics. Hyattsville, MD: National Center for Health Statistics; 2000:462 –484
US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, DC: US Government Printing Office; 2000
American Academy of Pediatrics. Children, adolescents, and television. Pediatrics. 2001;107 :423 –426
Prochaska JO, Velicer WF, DiClemente CC, Fava J. Measuring processes of change: application to the cessation of smoking. J Consult Clin Psychol. 1998;56 :520 –528
Saelens BE, Gehrman CA, Sallis JF, Calfas KJ, Sarkin JA, Caparosa S. Use of self-management strategies in a 2-year cognitive-behavioral intervention to promote physical activity. Behav Ther. 2000;31 :365 –379
Norman GJ, Sallis JF, Gaskins R. Comparability and reliability of paper- and computer-based measures of psychosocial constructs for adolescent physical activity and sedentary behaviors. Res Q Exerc Sport. 2005; In press
Velicer WF, DiClemente CC, Prochaska JO, Brandenburg N. A decisional balance measure for assessing and predicting smoking status. J Pers Soc Psychol. 1985;48 :1279 –1289
Norman GJ, Vaughn AA, Roesch SC, Sallis JF, Calfas KJ, Patrick K. Development of decisional balance and self-efficacy measures for adolescent sedentary behaviors. Psychol Health. 2004;19 :561 –575
Prochaska JJ, Rodgers MW, Sallis JF. Association of parent and peer support with adolescent physical activity. Res Q Exerc Sport. 2002;73 :206 –210
Salmon J, Owen N, Crawford D, Bauman A, Sallis JF. Physical activity and sedentary behavior: a population-based study of barriers, enjoyment, and preference. Health Psychol. 2003;22 :178 –188
Sallis JF, Taylor WC, Dowda M, Freedson PS, Pate RR. Correlates of vigorous physical activity for children in grades 1 through 12: comparing parent-reported and objectively measured physical activity. Pediatr Exerc Sci. 2002;14 :30 –44
Sallis JF, Johnson MF, Calfas KJ, Caparosa S, Nichols JF. Assessing perceived physical environmental variables that may influence physical activity. Res Q Exerc Sport. 1997;68 :345 –351
Saelens BE, Sallis JF, Black JB, Chen D. Neighborhood-based differences in physical activity: an environment scale evaluation. Am J Public Health. 2003;93 :1552 –1558
Gordon-Larson P, McMurray RG, Popkin BM. Adolescent physical activity and inactivity vary by ethnicity: the National Longitudinal Study of Adolescent Health. J Pediatr. 1999;135 :301 –306
Shelton BA, John D. Race, ethnicity, and difference: a comparison of white, black, and Hispanic men's household labor time. In: Hood JC, ed. Men, Work and Family. Newbury Park, CA: Sage; 1993
Marshall SJ, Biddle SJH, Gorely T, Cameron N, Murdey IM. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord. 2004;28 :1238 –1246
Sallis JF, Bauman A, Pratt M. Environmental and policy interventions to promote physical activity. Am J Prev Med. 1998;15 :379 –397
King AC, Castro C, Wilcox S, Eyler AA, Sallis JF, Brownson RC. Personal and environmental factors associated with physical inactivity among different racial-ethnic groups of U.S. middle-aged and older-aged women. Health Psychol. 2000;19 :354 –364
Patrick K, Sallis J, Lydston D, et al. Preliminary evaluation of a multi-component program for nutrition and physical activity change in primary care: PACE+ for Adolescents. Arch Pediatr Adolesc Med. 2001;155 :940 –946
Saelens BE, Sallis J, Wilfley D, Patrick K, Cella J, Buchta R. A randomized trial of a multi-componential behavioral weight control treatment initiated in primary care versus typical care for overweight adolescents. Obes Res. 2002;10 :22 –32(Gregory J. Norman, PhD, B)
Department of Psychology, San Diego State University, San Diego, California
ABSTRACT
Objective. To determine correlates of sedentary behaviors in adolescents through the examination of psychosocial and environmental variables.
Method. The study used a cross-sectional design to evaluate an ethnically diverse clinic-based sample of 878 adolescents who were 11 to 15 years old. Bivariate and multivariate analyses were stratified by gender to assess correlates of sedentary behaviors occurring on the most recent nonschool day (television viewing, computer video games, sitting listening to music, and talking on the phone).
Results. For girls, age, family support, television/video rules, and hills in the neighborhood were associated with sedentary behaviors. Furthermore, psychological constructs such as self-efficacy, enjoyment, change strategies, and pros and cons of change emerged as correlates of sedentary behaviors. A moderator effect revealed that the proportion of girls in the low-BMI group decreased with increased self-efficacy, whereas the proportion of girls in the high-BMI group did not vary significantly by self-efficacy. For boys, age, ethnicity, BMI, cons of change, and self-efficacy were associated with sedentary behaviors.
Conclusions. This study provides evidence of factors associated and not associated with adolescent sedentary behaviors. Similar to physical activity, measures of specific psychosocial constructs of sedentary behavior demonstrated important associations. The results highlight the need for additional examination of the correlates of sedentary behavior to determine which correlates are mechanisms of behavior change.
Key Words: determinants behavioral theory television sedentary behavior adolescents
Abbreviations: TV, television
Children in the United States spend an estimated 75% of the day being inactive.1 However, the relationships among sedentary behaviors, physical activity levels, and overweight are only beginning to be understood for the adolescent population. Sedentary behaviors, particularly watching television (TV) and videos, have been found to be related to higher body mass index (BMI) for children and adolescents.2–6 The literature linking physical activity levels with risk of overweight in adolescents is not consistent,7, 8 but physical activity is an important component of effective obesity treatments.9 The relationship between sedentary behaviors and physical activity has been inconsistent across studies.4, 5, 10 Determining specific correlates of sedentary behavior may help to clarify these relationships.
Experimental studies provide compelling evidence that reducing time spent in sedentary behaviors can be an effective weight-loss strategy for youth independent of changes in physical activity level.6, 9, 11–13 Sedentary behaviors, particularly TV viewing and playing computer games, are likely independent risk factors for children's overweight status. The likely mechanism for TV viewing's effect on overweight is its effect on energy balance. That is, TV viewing affects both energy expenditure (less time being physically active) and energy intake (overeating while watching TV and exposure to commercials promoting less-than-healthy foods).2, 14
Individual and population-based interventions that target modifiable correlates of behaviors are more likely to lead to behavior change.15 However, although much is known about the correlates of physical activity for adolescents,16 there has been little research on correlates of sedentary behaviors. One study of a large sample of seventh-graders found that higher depressive symptoms and lower perceived academic rank were associated with more sedentary-behavior time for boys and girls.17 A study of university students found that psychosocial correlates of physical activity were poorly correlated with TV viewing,18 suggesting that sedentary-specific psychological variables are needed. Gordon-Larsen et al19 concluded that physical activity and sedentary behaviors were associated with different determinants. They found that physical activity related most to environmental factors, whereas sedentary behaviors related more to sociodemographic factors.
The available evidence indicates that sedentary behaviors are not the mirror image of physical activity and that the 2 types of behaviors are relatively uncorrelated.20 Excessive sedentary behavior is a health risk, but the correlates of adolescent sedentary behaviors are not clear. The present study examined theory-derived psychological, social, and environmental factors associated with sedentary behaviors in adolescents. Determining the correlates of sedentary behaviors is the first step in understanding the mechanisms that influence these behaviors that can then be incorporated into intervention programs.21
METHODS
Participants
Adolescents between the ages of 11 and 15 years were recruited through their primary care providers as part of a health-promotion intervention trial. A total of 45 primary care providers from 6 clinic sites in San Diego County, California, agreed to participate. Over a 13-month period, trained recruiters attempted to contact 3366 households (including wrong numbers, those not eligible, and refusals), from which 878 adolescents (64% of eligible contacts) were enrolled onto the study after signing consent forms and completing baseline measurements. Adolescents received $10 for completing all measurements and were entered into a lottery drawing for 1 of 10 cash prizes ranging between $10 and $50. All study procedures were approved by university and clinic institutional review boards. Table 1 presents the sample characteristics for boys and girls. The sample was ethnically diverse, with 42% of participants from minority backgrounds.
Procedure
Items analyzed in the current study were part of a larger battery of survey questions completed at the baseline assessment. Adolescents completed measures on a desktop computer by using a Web-based interface. Responses to items were made by clicking on radio buttons using a computer mouse. Items were displayed on the computer screen with the directions and response scales presented at the top of each screen. The parent who accompanied the adolescent completed several survey instruments about the adolescent's home and neighborhood. These surveys were completed in a paper-and-pencil format, and response values were entered into a database by trained research assistants.
Measures
Adolescents completed measures pertaining to their behavior, attitudes, and beliefs related to being sedentary. The constructs assessed were based on social cognitive theory,22 the transtheoretical model,23 and environmental variables derived from ecological models.24 Parents completed surveys regarding home and neighborhood environments and their support for their child's physical activity. Unless otherwise stated below, reported internal consistency coefficients were calculated in the current study. All of the survey measures are available (through G.J.N.) at www.paceproject.org.
Height and Weight
An Accu-Hite wall stadiometer model 216 (Seca Accu-Hite, Hanover, MD) measured standing height. Weight was measured with the digital Body Comp scale from American Weights & Measures (Rancho Santa Fe, CA). Each measure was taken twice, and the average of the 2 readings was calculated. BMI was calculated as kilograms per square meter. BMI-for-age percentile was determined from Centers for Disease Control and Prevention national norms using age to the nearest month and gender-specific median, standard deviation (SD), and power of the Box-Cox transformation.25 BMI-for-age percentiles were split at the 85th percentile, which is defined as at risk for overweight.
Sedentary Behaviors
Self-reported sedentary behaviors were assessed with a survey adapted from Robinson.13 Participants were asked how much time they spent doing the following leisure-time sedentary behaviors: watching TV (including videos on VCR/DVD); playing computer or video games (such as Nintendo or Sega); sitting and listening to music on the radio, audiotapes, or CDs; and sitting and talking on the telephone. Questions were asked first for "most recent day when you were not in school" and then for the "most recent school day." Participants responded to each item by using a 9-point scale with anchors of "none," "15 minutes or less," "30 minutes," "1 hour," "2 hours," "3 hours," "4 hours," "5 hours," or "6 hours or more."
An index of sedentary-behavior time was computed by summing the 4 items for nonschool days. Item responses correlated highly between nonschool day and school day (r = 0.83). Because a nonschool day was likely to include more unstructured time for an adolescent, allowing for more time engaging in sedentary behaviors, the sum of the nonschool-day responses was used as the outcome variable. Internal consistency for this scale was moderate, with = .59. Test-retest reliability was determined from a separate sample to be sufficient (intraclass correlation coefficient: .77).
We dichotomized sedentary time into (0) 240 minutes and (1) >240 minutes. Given that Healthy People 201026 and the American Academy of Pediatrics27 recommend <2 hours of TV viewing per day, we reasoned that 4 hours per day would be a logical extension when considering several more types of sedentary behaviors. This cut point was also the median split for the sample, which maximizes the variable's potential to correlate with the predictor variables. Validity for the dichotomized sedentary-behavior index was demonstrated with significant coefficients for percent body fat (r = 0.13; P < .001), aerobic fitness (r = –0.11; P < .001), and average daily minutes of vigorous physical activity measured with accelerometers (r = –0.17; P < .001).
Behavior-Change Strategies
Fifteen items were created that reflect thoughts, activities, and feelings that people may use when making a behavior change. These change strategies were similar to constructs described as processes of change in the transtheoretical model28 and were based on items developed by Saelens et al.29 Item examples included "thinking about how surroundings affect the amount of sedentary time" and "overcoming hassles to reduce sedentary time." The response format assessed how often each strategy was used, with a 5-point Likert scale ranging from 1 (never) to 5 (many times). Norman et al30 determined test-retest reliability for the scale to be .75. Internal consistency for the 15 items was = .91.
Pros and Cons of Change
Decisional balance consists of 2 constructs labeled the "pros" and "cons" of change that address cognitive and motivational aspects of human decision-making.31 Norman et al32 developed a 4-item pro and 5-item con scale of reducing sedentary behaviors and determined that the 2-factor structure demonstrated good internal validity. Pros included: "my parents would be pleased if I spent less time playing computer/video games" and "playing computer/video games hurts my eyes." Cons of change included: "I enjoying playing computer/video games for many hours" and "watching TV or playing computer games is my way to escape from the world." Participants rated the importance of each item by using a 5-point Likert scale with anchors of 1 (not at all important) to 5 (extremely important). Test-retest reliability for the pro and con scales were .86 and .77, respectively. Internal consistency was = .60 for the pros and = .76 for the cons.
Self-Efficacy
Situational self-efficacy represents a person's confidence about meeting a behavioral criterion in situations that may present barriers to the behavior. A 7-item sedentary-behaviors self-efficacy scale assessed adolescents' confidence to reduce the amount of time that they spend being sedentary (eg, plan ahead of time what TV shows you will watch during the week).32 Participants responded to each item on a 5-point Likert scale ranging from 1 (I'm sure I can't) to 5 (I'm sure I can). Test-retest reliability for this scale was .81,30 and internal consistency was = .81.
Family Support
Four questions assessed adolescents' perceptions of how often a household member gave different types of support: (1) encourages you to spend less time being sedentary; (2) discusses with you how sedentary habits can be unhealthy; (3) helps you think of ways to reduce the time you spend on sedentary habits; and (4) tells you that you are doing a good job reducing your sedentary habits. The 5-point scale ranged from "never" to "every day." These items were adapted from a family-support scale that was developed by Prochaska et al33 that had good internal consistency and validity with self-reported physical activity. Test-retest reliability for the 4-item sedentary-behaviors family-support scale was found to be good (intraclass correlation coefficient: .72),30 and internal consistency was = .86.
Enjoyment of Sedentary Behaviors
Adolescents rated a single item, "I enjoy doing sedentary habits like watching TV or playing computer/video games," on a 5-point scale ranging from "strongly disagree" to "strongly agree." Similar measures were used in large population-based studies and were found to have good reliability and validity.34
TV and Video Household Rules
Two items assessed limits set by parents on (1) hours of TV/Video viewing and (2) hours of video/computer games. The 4-point response format ranged from "never" to "always." Internal consistency for the scale was = .87.
Parent-Reported Support for Physical Activity
Items from the Amherst Health and Activity Study Questionnaire35 assessed parents' reported level of support for their child's physical activity during a typical week. Four items formed a composite measure with an internal consistency of = .75: (1) watching your child participating in physical activity; (2) encouraging your child to do physical activity; (3) doing physical activity with child; and (4) providing transportation so your child can get to a place where he or she can do physical activity. The 5-point scale for these items ranged from "never" to "every day."
Home Environment
Sixteen items assessed the presence of items in the home that would facilitate physical activity. Sallis et al36 found this scale to have a high test-retest reliability (.89) and to be correlated with self-reported physical activity. Adolescent parents responded "yes" or "no" for whether each item was in the home, yard, apartment complex, or community. The number of "yes" responses was totaled to form a single score. Items included bicycle, dog, swimming pool, sports equipment, skis, and toning devices. Internal consistency for the scale in the present study was = .59.
Neighborhood Environment Variables
Parents responded to questions about their neighborhoods on the Neighborhood Environment Walkability scale, which was developed recently by Saelens et al.37 All of the summated scales were found to have test-retest reliability and, with the exception of the Safety From Crime scale, differentiated individuals in a highly walkable neighborhood compared with those in a neighborhood that is not particularly walkable. The Neighborhood Environment Walkability scale subscales used in the present study are described below.
Land-Use Mix/Diversity
We used 14 of the 16 items from this subscale that assess the extent to which stores and facilities are in walking distance of a participant's home. Participants rated each location on a 5-point scale from "1–5 minutes" to "31+ min." Nine items were used to form a "retail in neighborhood" scale that included grocery store, supermarket, hardware store, fruit/vegetable market, laundry/dry cleaner, clothing store, other stores, post office, and library. Five items formed a "recreation in neighborhood" scale and included elementary school, other schools, park, recreation center, gym, and fitness facility. Internal consistency for these scales was = .90 and .79, respectively.
Walking and Cycling Facilities
Six items assessed proximity and quality of neighborhood trails and sidewalks for walking and bicycling. The 4-point response format for these items ranged from "strongly disagree" to "strongly agree." Internal consistency for this scale was = .74.
Hills in Neighborhood
Two items assessed the difficulty of getting around in the neighborhood by bicycle or on foot, considering the hilliness of the area. The 4-point response format for these items ranged from "strongly disagree" to "strongly agree." Internal consistency was = .76.
Pedestrian/Traffic Safety
Eight items assessed traffic speed and crosswalks in regard to impeding walking in the neighborhood. The 4-point response format ranged from "strongly disagree" to "strongly agree." Internal consistency for this scale was = .67.
Crime Safety
Six items assessed crime rate and street lighting in the neighborhood. The mean of 6 items on a 4-point scale ranging from "strongly disagree" to "strongly agree" was calculated. Internal consistency for this scale was = .68.
Analyses
All analyses were conducted separately for girls and boys. First, bivariate analyses of each independent variable with the dichotomous outcome variable sedentary behaviors determined unadjusted odds ratios. For these exploratory analyses, P values were considered statistically significant at .05 and were not adjusted for multiple tests. Second, multivariate logistic-regression models were built and estimated. Model building occurred in several blocks. The first block of variables forced into the model consisted of demographic variables including adolescents' age, ethnicity, BMI percentile, and highest adult household education level. For the next block, predictor variables were allowed to step into the model with the entry level set at P .10. In the third block, all first-order interaction terms of demographic x predictor variables already in the model were allowed to step into the model with an entry level set at P .05. The Hosmer-Lemeshow goodness-of-fit statistic assessed overall model fit. A small test statistic and a large P value (P > .10) indicated a model that provided a good fit to the data. The Nagelkerke R2 statistic estimated the percentage of variance accounted for by the model.
Additional analyses were conducted with school-day sedentary time and nonschool-day sedentary behavior as continuous variables. These analyses resulted in essentially the same set of statistically significant predictors and multivariate models and, in the interest of brevity, are not reported here.
RESULTS
Description of the Outcome Variable
Relationships Among Predictor Variables
Correlations among predictor variables were in the low to moderate range, with a similar pattern of relationships found for girls and boys. For the 120 correlations among the 16 predictor variables, there were only 3 pairs that were above r = 0.50. The highest correlations were between neighborhood recreation and neighborhood retail (girls: r = 0.65; boys: r = 0.64), change strategies and family support (girls: r = 0.57; boys: r = 0.62), and transportation and encouragement for physical activity (girls: r = 0.56; boys: r = 0.57) (all P < .01). Twelve correlation pairs were between r = 0.30 and 0.50, and the remaining correlations were below r = 0.30. This pattern of relationships among the predictors indicated that colinearity would not be a problem in the multivariate analyses.
Unadjusted Relationships for Girls
For girls, 8 bivariate associations between predictor variables and leisure-time sedentary behavior were statistically significant (Table 3). Older girls (13–15 years old) were more likely to be in the >240-minutes of sedentary time group than younger girls (OR: 1.69; 95% CI: 1.18–2.44). Higher scores on change strategies, pros, self-efficacy, and TV/video rules were related to decreased likelihood of being in the high-sedentary-behavior group. High scores on cons, enjoyment of sedentary behaviors, and hills in the neighborhood were related to increased likelihood of being in the high-sedentary-time group.
Unadjusted Relationships for Boys
For boys, 7 bivariate associations were also found between predictor variables and sedentary behavior (Table 4). Boys who were older, nonwhite, and in the >85th percentile BMI category were more likely to be in the high-sedentary-behavior group than boys who were younger, white, and in the <85th percentile BMI category. Higher scores on self-efficacy and TV/video rules were associated with decreased likelihood of being in the high-sedentary-behavior group. Higher scores on the cons and enjoyment of sedentary behaviors were associated with decreased likelihood of being in the high-sedentary-behavior group.
Final Multivariate Model for Girls
The final multivariate logistic model for girls included all of the variables that were associated with the outcome from the unadjusted bivariate analyses. The model-building process resulted in the inclusion of 1 interaction for BMI percentile and self-efficacy. To simplify presentation of the results, Table 5 presents the main-effects model parameters. The R2 for the main-effects model was 0.25, and the inclusion of the interaction term increased the R2 to 0.28. The Hosmer-Lemeshow test indicated that the fit of the model was good (P = .25). Figure 1 graphically displays the interaction effect by dividing self-efficacy into quartiles. Girls in the low-BMI category (<85th percentile) demonstrated a fairly linear relationship with increasing confidence scores related to decreased percentages of girls in the high-sedentary-behavior category. However, for girls in the high-BMI (85th percentile) group, there was little relationship between confidence and sedentary behavior.
Final Multivariate Model for Boys
The final multivariate logistic model for boys included age, BMI percentile, cons, and self-efficacy as significant correlates of sedentary time (Table 6). Ethnicity was forced into the model but no longer reached statistical significance (P = .088). The variables enjoyment of sedentary behaviors, TV/video home rules, and transportation for physical activity, which were significantly related to sedentary time in the unadjusted analyses, did not enter the model. The final model's R2 was 0.22, and the fit of the model was good (P = .35).
DISCUSSION
Sedentary behaviors are likely to continue to be pervasive in technologically developed societies, and the results of this study contribute to a greater understanding of the correlates of these behaviors in adolescence. Knowledge about nonmodifiable factors (eg, age) can be useful in developing an overall behavioral profile for an individual or population subgroup. Knowledge about modifiable correlates can be used to tailor health messages delivered by pediatricians and public health professionals. Some of the most meaningful findings are summarized below.
Demographic and Anthropometric Factors
Gender, age, ethnicity, and BMI were found to be related to sedentary-behavior time to varying degrees. Overall, although nearly half of all girls and boys reported spending >4 hours on a nonschool day engaged in sedentary behaviors, what contributed to the amount of sedentary-behavior time was different for each group. Girls reported more time spent sitting and listening to music and sitting talking on the phone, whereas boys reported more time playing computer games. These data have implications for what clinicians might discuss with parents who are concerned about their child's behavior, as well as for school and community-based groups developing interventions to address adolescent health.
Similar to findings from previous studies showing that physical activity decreases with increasing age in adolescence,16 older adolescents reported more sedentary time than younger adolescents. As adolescents move into their teenage years, they likely increase the time that they spend engaging in multiple sedentary behaviors that compete with physical activities. Older adolescents have more time that is not supervised by a parent or caregiver, which creates opportunities to increase TV-viewing time and other sedentary behaviors. Encouraging alternative physical activities or teaching ways to engage in physical activity while engaging in pastimes such as listening to music, talking on the phone, and watching TV may help adolescents to develop more active lifestyles.
Hispanic and other minority boys were more likely to be in the high-sedentary-time group than non-Hispanic white boys. This pattern is consistent with previous research that found large ethnic differences for inactivity and found boys to engage in more hours of sedentary behaviors than girls.38 The largest ethnic group represented in the sample was Hispanic adolescents, suggesting that there may be cultural factors contributing to these findings among Hispanic families.39 In any case, these findings suggest that practitioners hoping to effect changes in sedentary behavior among adolescents will need to target population approaches and tailor individual-level strategies based on localized ethnic and cultural factors.
BMI was expected to be related to sedentary time based on previous studies.6, 13 However, BMI status was only associated with sedentary behaviors for boys. In previous studies it was common to find inconsistent associations across subgroups,6 which perhaps reflects measurement error or sample-specific characteristics. A recent meta-analysis also found only a weak relationship between TV viewing and body fatness.40 The evidence suggests that simply trying to reduce adolescent TV viewing may not be the best intervention strategy and must be combined with strategies to reduce overall sedentary-behavior time and caloric intake.
Psychosocial Factors
Several psychosocial constructs were related to sedentary behavior, including the cons of reducing sedentary behaviors and self-efficacy for girls and boys and change strategies, pros, and family support for girls. These associations identify important modifiable factors that can be targeted in health-promotion interventions. For example, a pediatrician can ask about an adolescent's self-efficacy to reduce sedentary behaviors and then use this knowledge to provide brief, tailored counseling to help shape beliefs and provide encouragement to change.22
Environmental Factors
There in increasing interest in how environmental factors help shape physical activity and other health behaviors.24, 41 However, there was no evidence in this group of adolescents that environmental neighborhood factors such as land-use mix or safety from crime and traffic were associated with time spent engaging in sedentary behaviors, which is a finding that is consistent with others19 indicating that environmental factors were more related to physical activity and not sedentary behavior. We did find that girls living in more hilly neighborhoods reported more sedentary-behavior time. Living in a hilly neighborhood may discourage adolescent girls from outdoor activities such as walking and bicycling during a time in their lives when sedentary alternatives such as talking on the telephone are more desirable. This finding with adolescents contrasts with a positive association between hilly neighborhood and physical activity in older women,42 which highlights the need for additional studies focused on physical environment correlates in diverse populations.
Strengths of this study include the use of a large, ethnically diverse sample of adolescents, the inclusion of variables from multiple domains (from demographic to environmental levels), and the use of variables measured with established psychometric properties. Study limitations include the use of a cross-sectional design that limits the variable associations to the "correlates" level of evidence; however, this is a necessary preliminary step for understanding mechanisms of behavior change.21 The composite measure of sedentary behavior precluded the assessment of correlates of individual sedentary behaviors and the potential of engaging in multiple sedentary or active behaviors at the same time. However, the measure was appropriate for estimating cumulative sedentary time, which may be more important than looking at separate sedentary behaviors.40
The reliance on self-report data raises the possibility of response-set biases such as social desirability. Being sedentary may be considered an undesirable behavior. Ensuring confidentiality and anonymity of responses can help to minimize response-set biases. Using mainly summated scales rather than individual items in the analyses also helped to minimize bias and measurement error. The generalizability of the study results is limited to the age range of adolescents between 11 and 15 years old and adolescents whose families have medical insurance. These findings may not persist for older or younger samples of youth or for adolescents in families in low socioeconomic segments that do not have medical coverage.
This study provides initial evidence of factors associated and not associated with adolescent sedentary behavior. Examining potential correlates from intrapersonal, interpersonal, and environmental domains indicated that sedentary behavior was mainly associated with intrapersonal factors such as demographic characteristics and psychosocial constructs that specifically addressed sedentary behavior. The findings suggest that adolescent sedentary behavior has correlates that are distinct from correlates found for physical activity, and these correlates may provide the basis for programs that can be delivered through primary care settings43, 44 or in the public health arena. For researchers, the correlates found in this study can be useful for generating hypotheses to be tested in future studies. Additional studies with prospective designs are needed to determine the temporal ordering and dose-response relationship of modifiable correlates that are potential determinants of sedentary behavior.
ACKNOWLEDGMENTS
This project was supported by National Cancer Institute grants R01 CA81495 and R01 CA85873.
FOOTNOTES
Accepted Jan 4, 2005.
No conflict of interest declared.
REFERENCES
Strauss RS, Rodzilsky D, Burack G, Colin M. Psychosocial correlates of physical activity in healthy children. Arch Pediatr Adolesc Med. 2001;155 :897 –902
Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M. 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
Utter J, Neumark-Sztainer D, Jeffery R, Story M. Couch potatoes or french fries: are sedentary behaviors associated with body mass index, physical activity, and dietary behaviors among adolescents J Am Diet Assoc. 2003;103 :1298 –1305
Eisenmann JC, Bartee RT, Wang MQ: Physical activity, TV viewing, and weight in US youth: 1999 Youth Risk Behavior Survey. Obes Res. 2002;10 :379 –385
Saelens BE. Helping individuals reduce sedentary behavior. In: Andersen RE, ed. Obesity: Etiology, Assessment, Treatment, and Prevention. Champaign, IL: Human Kinetics; 2003:217 –238
Riddoch C. Relationships between physical activity and physical health in young people. In: Biddle S, Sallis JF, Cavill NA, eds. Young and Active Young People and Health Enhancing Physical Activity: Evidence and Implications. London, England: Health Education Authority; 1998:17 –48
Patrick K, Norman G, Sallis J, et al. The relative contributions of diet, physical activity, and sedentary behaviors as risk factors for overweight in adolescence. Arch Pediatr Adolesc Med. 2004;158 :385 –390
Epstein LH, Myers MD, Raynor HA, Saelens BE. Treatment of pediatric obesity. Pediatrics. 1998;101 :554 –570
Feldmann 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 :797 –802
Epstein LH, Valoski AM, Vara LS, et al. Effects of decreasing sedentary behavior and increasing activity on weight change in obese children. Health Psychol. 1995;14 :109 –115
Epstein LH, Paluch RA, Gordy CC, Dorn J. Decreasing sedentary behaviors in treating pediatric obesity. Arch Pediatr Adolesc Med. 2000;154 :220 –226
Robinson TN. Reducing children's television viewing to prevent obesity: a randomized controlled trial. JAMA. 1999;282 :1561 –1567
Dietz WH, Gortmaker SL. Preventing obesity in children and adolescents. Annu Rev Public Health. 2001;22 :337 –353
Baranowski T, Anderson C, Carmack C. Mediating variable framework in physical activity interventions. How are we doing How might we do better Am J Prev Med. 1998;15 :266 –297 [published correction appears in Am J Prev Med. 1999;17:98]
Sallis JF, Prochaska JJ, Taylor WC. A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc. 2000;32 :963 –975
Schmitz KH, Lytle LA, Phillips GA, Murray DM, Birnbaum AS, Kubik 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;34 :266 –278
Williams CD, Sallis JF, Calfas KJ, Burke RS. Psychosocial correlates of television viewing. Am J Health Promot. 1999;13 :207 –214
Owen N, Leslie E, Salmon J, Fotheringham MJ. Environmental determinants of physical activity and sedentary behavior. Exerc Sport Sci Rev. 2000;28 :153 –158
Bauman AE, Sallis JF, Dzewaltowski DA, Owen N. Toward a better understanding of the influences on physical activity: the role of determinants, correlates, causal variables, mediators, moderators, and confounders. Am J Prev Med. 2002;23 (2 suppl):5–14
Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall; 1986
Prochaska JO, DiClemente CC, Norcross JC. In search of how people change: applications to addictive behaviors. Am Psychol. 1992;47 :1102 –1114
Sallis JF, Owen N. Ecological models of health behavior. In: Glanz K, Rimer BK, Lewis FM, eds. Health Behavior and Health Education: Theory, Research, and Practice. 3rd ed. San Francisco, CA: Jossey-Bass; 2002
Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC Growth Charts: United States Advance Data From Vital and Health Statistics. Hyattsville, MD: National Center for Health Statistics; 2000:462 –484
US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, DC: US Government Printing Office; 2000
American Academy of Pediatrics. Children, adolescents, and television. Pediatrics. 2001;107 :423 –426
Prochaska JO, Velicer WF, DiClemente CC, Fava J. Measuring processes of change: application to the cessation of smoking. J Consult Clin Psychol. 1998;56 :520 –528
Saelens BE, Gehrman CA, Sallis JF, Calfas KJ, Sarkin JA, Caparosa S. Use of self-management strategies in a 2-year cognitive-behavioral intervention to promote physical activity. Behav Ther. 2000;31 :365 –379
Norman GJ, Sallis JF, Gaskins R. Comparability and reliability of paper- and computer-based measures of psychosocial constructs for adolescent physical activity and sedentary behaviors. Res Q Exerc Sport. 2005; In press
Velicer WF, DiClemente CC, Prochaska JO, Brandenburg N. A decisional balance measure for assessing and predicting smoking status. J Pers Soc Psychol. 1985;48 :1279 –1289
Norman GJ, Vaughn AA, Roesch SC, Sallis JF, Calfas KJ, Patrick K. Development of decisional balance and self-efficacy measures for adolescent sedentary behaviors. Psychol Health. 2004;19 :561 –575
Prochaska JJ, Rodgers MW, Sallis JF. Association of parent and peer support with adolescent physical activity. Res Q Exerc Sport. 2002;73 :206 –210
Salmon J, Owen N, Crawford D, Bauman A, Sallis JF. Physical activity and sedentary behavior: a population-based study of barriers, enjoyment, and preference. Health Psychol. 2003;22 :178 –188
Sallis JF, Taylor WC, Dowda M, Freedson PS, Pate RR. Correlates of vigorous physical activity for children in grades 1 through 12: comparing parent-reported and objectively measured physical activity. Pediatr Exerc Sci. 2002;14 :30 –44
Sallis JF, Johnson MF, Calfas KJ, Caparosa S, Nichols JF. Assessing perceived physical environmental variables that may influence physical activity. Res Q Exerc Sport. 1997;68 :345 –351
Saelens BE, Sallis JF, Black JB, Chen D. Neighborhood-based differences in physical activity: an environment scale evaluation. Am J Public Health. 2003;93 :1552 –1558
Gordon-Larson P, McMurray RG, Popkin BM. Adolescent physical activity and inactivity vary by ethnicity: the National Longitudinal Study of Adolescent Health. J Pediatr. 1999;135 :301 –306
Shelton BA, John D. Race, ethnicity, and difference: a comparison of white, black, and Hispanic men's household labor time. In: Hood JC, ed. Men, Work and Family. Newbury Park, CA: Sage; 1993
Marshall SJ, Biddle SJH, Gorely T, Cameron N, Murdey IM. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord. 2004;28 :1238 –1246
Sallis JF, Bauman A, Pratt M. Environmental and policy interventions to promote physical activity. Am J Prev Med. 1998;15 :379 –397
King AC, Castro C, Wilcox S, Eyler AA, Sallis JF, Brownson RC. Personal and environmental factors associated with physical inactivity among different racial-ethnic groups of U.S. middle-aged and older-aged women. Health Psychol. 2000;19 :354 –364
Patrick K, Sallis J, Lydston D, et al. Preliminary evaluation of a multi-component program for nutrition and physical activity change in primary care: PACE+ for Adolescents. Arch Pediatr Adolesc Med. 2001;155 :940 –946
Saelens BE, Sallis J, Wilfley D, Patrick K, Cella J, Buchta R. A randomized trial of a multi-componential behavioral weight control treatment initiated in primary care versus typical care for overweight adolescents. Obes Res. 2002;10 :22 –32(Gregory J. Norman, PhD, B)