You are here

Article Text

Article menu

Download PDFPDF

Original article
Behavioural and social correlates of sedentary time in young people
  1. E M F van Sluijs1,
  2. A Page2,
  3. Y Ommundsen3,
  4. S J Griffin1
  1. 1Medical Research Council Epidemiology Unit, Cambridge, UK
  2. 2Exercise, Nutrition & Health Sciences, University of Bristol, Bristol, UK
  3. 3Norwegian School of Sport Sciences, Oslo, Norway
  1. Correspondence to Dr E M F van Sluijs, MRC Epidemiology Unit, Institute of Metabolic Science, Box 285, Addenbrooke's Hospital, Hills Road, Cambridge CB2 0QQ, UK; esther.vansluijs{at}mrc-epid.com.ac.uk

Abstract

Objective In this study, behavioural and social correlates of objectively measured sedentary time in young people were identified.

Design Cross-sectional analysis of data from the European Youth Heart Study (EYHS).

Setting Schools in Denmark, Estonia, Portugal and Norway.

Participants Respondents were invited using a twostage cluster sampling procedure. Analyses include 2107 children (9–10 years old) and adolescents (14–15 years old).

Assessment of independent variables Seven behavioural and 15 social variables assessed by parental and computerised child questionnaires.

Main outcome measure Sedentary activity as assessed by accelerometry (10 min blocks at <200 counts/min). Analyses were stratified by country, and interactions with grade and sex were investigated.

Results Adolescents were more sedentary than children (335.4 (90.4 min/day) vs 217.2 (75.6 min/day), p<0.001). Patterns of associations differed across countries. High computer use and no television viewing before school in Norway, and being sedentary during school breaks in Estonia were positively associated with sedentary time. No behavioural variables were associated with sedentary time in the Danish and Portuguese models. Socioeconomic position was positively associated with sedentary time in Portugal and Estonia, father's body mass index negatively in the Estonian model. Norwegian participants with a game console at home and Portuguese participants with a television in their bedroom were more sedentary.

Conclusions A single strategy aimed at reducing sedentary behaviour is unlikely to be effective across Europe as the target populations and behaviours of focus differ between countries. Targeting high socioeconomic groups in Portugal and Estonia or focusing on reducing computer use in Norway might be effective intervention strategies to reduce overall sedentary time.

https://doi.org/10.1136/bjsm.2008.049783

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Regular physical activity in young people is believed to be associated with various physical and mental health benefits, some of which extend into adulthood.1,–,3 Alongside engaging in too little physical activity, it has also been suggested that young people spend too much time in sedentary activities, which further increases their risk of future health problems.4 5 Although TV viewing is a frequently used proxy measure for overall sedentary activity, the appropriateness of this approach has been questioned.4,–,6 Research has shown that only low to moderate correlations exist between sedentary activities,7 8 suggesting that sedentariness among young people is multifaceted and cannot accurately be represented by one proxy measure. Moreover, no association to weak associations have been reported between TV viewing and physical activity,7 9 indicating that the behaviours are not two sides of the same coin. There is also a lack of evidence of an association between TV viewing and health indicators, such as body fatness.10 However, recent research shows that overall time spent in sedentary activities is associated with health outcomes, such as weight gain11 12 in children and adults, independent of physical activity level,13 14 whereas interruptions to prolonged periods of sedentary time are beneficially associated with metabolic risk in adults.15 These results indicate that targeting reductions in overall sedentary time instead of a single behaviour might be more effective.

    To develop interventions aimed at reducing sedentary time, an understanding of the determinants of this behaviour is required. These are likely to differ from determinants of physical activity behaviour.16 17 Research into correlates of sedentary behaviour should be conceptualised within an ecological framework to identify a broad range of intrapersonal, interpersonal and environmental influences.18 To date, research has mainly focused on TV viewing, with recent reviews showing consistent associations with body mass index (BMI) and sociodemographic factors (eg, ethnicity, socioeconomic status (SES)), but few associations with modifiable factors.18 19 Previous studies of the correlates of overall sedentary behaviour have relied on self-reported measures of the behaviour. In a British sample of 11–12-year-olds, sedentary behaviour (defined as screen time) was greater among those from minority ethnic backgrounds, living in more deprived neighbourhoods and having psychological problems.20 Sedentary behaviour was also shown to increase during puberty, independent of ethnicity or SES.21 In a cross-sectional sample of American 11–15-year-olds, where sedentary behaviour was self-reported but more widely defined (including screen time and activities such as listening to the radio and talking on the phone), age was also positively associated with overall sedentary behaviour.22 Moreover, sedentary behaviour-specific psychological factors (such as selfefficacy, and perceived pros and cons), family environmental factors and hills in the neighbourhood were associated with high levels of sedentary time (>240 min) in girls, whereas only self-efficacy and barriers to change were associated in boys. No studies have investigated correlates of objectively measured sedentary time, although accelerometry may be a suitable method to assess overall sedentary time,23 and objectively measured sedentary time has been shown to be associated with metabolic outcomes in children.12 Use of a more precise measure of sedentary time also reduces measurement error, increasing the likelihood of detecting associations.24 The aim of the current paper is to investigate associations of objectively measured sedentary time with self-reported sedentary behaviours and sociocultural factors in a large population-based sample of European children.

    Methods

    Study design and participant recruitment

    The European Youth Heart Study (EYHS) was designed to study the personal, environmental and lifestyle influences on cardiovascular disease risk factors in European children. A detailed description of the study rationale, aims, design, selection criteria and sample size is provided elsewhere.25 A brief overview is provided below.

    A defined population of children was identified in each study region (Odense, Denmark; Tartu, Estonia; Oslo, Norway; Madeira, Portugal). Children were randomly selected using a two-staged cluster sampling procedure with schools as the primary sampling unit and classes being the secondary sampling unit. Schools were stratified by the sociodemographic characteristics of their local area, and a minimum of 20 schools were randomly selected from local authority lists using probability proportional to school size. Children from grades 3 and 9 (aged 9 and 15 years, respectively) were then randomly sampled within schools using the school register. The overall response to invitations was 72% and was similar across age and sex groups (total n = 4072). Written and verbal explanations of study procedures were provided, and written informed consent was obtained from the child's parent or legal guardian. Ethical approval was obtained from the local ethics committee in each study region.

    Data collection

    The present cross-sectional analyses uses baseline data collected during school visits between 1997 and 2000. A standardised research protocol was followed in each country in collecting, processing and analysing the data.25 Child-reported variables were assessed using a validated computer-based questionnaire,26 and physical activity was assessed for 4 days following the school visit. Parent-reported questionnaires were completed separately by both parents (no parent data was collected for Norwegian grade 9 participants).

    Outcome variable—sedentary time

    Data on overall activity levels was collected using the MTI accelerometer (model 7164). The MTI is a hip-worn electronic motion sensor comprising a single plane (vertical) accelerometer, which has been validated in both children and adolescents.25 Verbal and written instructions were given to both the children and their parents regarding its use. Epoch time was set to 60 s, and children were asked to wear the monitor for four consecutive days, including two weekend days, during all waking hours (excluding water-based activities). Upon return, data were downloaded into a computer and stored for further analyses. A Microsoft Excel-based macro was used for data reduction (including exclusion of sequences of 10 or more consecutive minutes with zero counts) and to derive relevant variables. Children who did not manage to record at least 600 min/day for at least 3 days were excluded from analyses. A total of 2107 children (1608 nine-year-olds and 1131 fifteen-year-olds) provided valid accelerometer data and were included in the current analyses.

    To assess sedentary time, all recorded activity between 23:00 and 05:00 was excluded to avoid overestimation caused by children sleeping with the monitor. A cut-off of <200 counts/min was used to identify sedentary activity. To date, cut points of 100 counts/min,27 500 counts/min,12 28 and 80029 have been identified and used to define sedentary activity. A relatively low cut point was selected in this analysis to avoid including minutes of “very light to light activity” and because time spent below this threshold has activity” and because time spent below this threshold has year-old children.30 To minimise the influence of incidental low values, we used the number of sedentary minutes in bouts of at least 10 min, allowing for 1 min exceptions, as our main outcome variable.

    Exposure variables

    Where relevant, variables were dichotomised based on frequency distributions to reduce the number of dummy variables included in the model and maximise statistical power.

    General variables

    Height and weight were measured using standard procedures,25 and BMI was calculated by dividing weight in kilograms by height in metres squared. Child's ethnicity was reported by parents and collapsed into a dichotomous variable (white vs other ethnic background).

    Behavioural variables

    All variables were child-reported, and previously validated questionnaires were used to assess break-time activity and active travel.31 A composite variable representing activity level during breaks was constructed from reported activity at morning and lunch break (see appendix 1). Travel mode to school was dichotomised into active travel (walking, cycling) and non-active travel (car, bus, train). Frequency of eating meals while watching TV was reported in five categories and dichotomised into nearly everyday (everyday, most days) or less frequent. The four categories of reported TV viewing before school and computer use were dichotomised into less than 1 or 1 h or more. TV viewing after school was reported in five categories and dichotomised at 2 h. Previous EYHS analyses have shown these measures of TV viewing to be associated with waist circumference32 and metabolic syndrome.33

    Parental variables

    All parental variables were self-reported by mothers and fathers separately, including height and weight. Parental sports participation was measured by asking parents whether they participated in regular physical activity or sports (yes/no). In each country, parental education was classified into four categories (basic/primary; secondary/trade apprentice; higher vocational qualifications; university) and income into eight categories. Socioeconomic position was assessed by summing the score for mother's and father's educational level and the mean score for mother's and father's income.

    Sedentary opportunity variables

    Children self-reported the number of TVs in the household and if they had a games computer or a video at home or a TV in their bedroom (yes vs maybe/no).

    Psychosocial variables

    Children self-reported on a four-point scale how often (1) they could choose what TV programme to watch; (2) the TV was on when they returned home and (3) they felt that they were not allowed to play outside because of parental safety concerns (hardly ever or never, once or twice a week, most days, everyday). All scales were dichotomised by combining the first two and the last two categories.

    Data analyses

    Country distinct relationships in this data set have been reported previously,34 therefore all analyses (conducted in 2006) were stratified by country. Initial analyses were conducted using univariate linear regression analyses with sedentary time as dependent variable. Differences in effect between sexes or grades were investigated by creating two-way interaction terms and separately including these in the univariate model. Analyses including BMI were adjusted for grade to take into account that BMI increases with age during youth. All single-term exposures and interaction terms achieving a probability of less than 0.20 were subsequently entered into multivariable models. A manual backward stepwise regression method was used to build the final models, with decision on removal of variables based on the p-value of the exposure variable. Where relevant, non-significant interaction terms were removed before single terms. The significance level for remaining in the model was set at 0.05. Initially, a country-specific model including sex, grade and other remaining significant general variables was built. The subsequent four models were built by adding to the base model: model 2, plus parental variables; model 3, plus sedentary opportunity variables; model 4, plus psychosocial variables; model 5, plus behavioural variables.

    Results

    Descriptive characteristics

    On average, EYHS participants spent 262 (99.7 minp/day) in sedentary time (just under 4.5 h) (fig 1). Danish participants were the most sedentary, compared with participants from the other three countries (all p<0.05). No other differences were observed. In all countries, adolescents were more sedentary than children (335.4 (90.4 min/day) vs 217.2 (75.6 min/day), p<0.001). Table 1 shows the frequency and distribution of the exposure variables across the different countries. Betweencountry differences were observed for almost all variables. Remarkable observations include the low level of sports participation in Portuguese mothers and fathers and the higher proportion of children in Norway that have a game console at home and are not allowed to play outside because of safety reasons.

    Figure 1
    Figure 1

    Average sedentary time per day for grade 3 and grade 9 children in the EYHS cohorts in Denmark, Estonia, Norway and Portugal (measured with Actigraph accelerometer). Grade differences are significant within countries at p<0.001.

    View this table:
    Table 1

    Descriptive data for exposure variables, per country. Values are percentages within country or mean (SD)

    Univariate analyses

    The results of the univariate analyses are reported in table 2. A significant difference between sexes was only observed in Estonia, with boys being less sedentary than girls. Gradeadjusted BMI was positively associated with sedentary activity in Portugal, negatively in Estonia and not associated in other countries. Out of the 19 behavioural and social factors considered,ineachcountry, lessthanhalfweresignificantly associated with sedentary time in the univariate analyses (seven in Denmark, six in Estonia, nine in Norway and six in Portugal).

    View this table:
    Table 2

    Results of univariate analyses between general, behavioural, parental, sedentary opportunity and psychosocial exposures and sedentary activity (unstandardised b coefficient and 95% CI)

    Multivariable analyses

    Table 3A–D presents the results of the country-specific multivariable models.

    View this table:
    Table 3

    Results of the final multivariable analyses between general, behavioural, parental, sedentary opportunity, and psychosocial exposures and sedentary activity, stratified by country (presenting general model and additionally statistically significant single terms and interactions)

    Parental, opportunity and psychosocial correlates

    Only one social correlate was retained in the Danish and Norwegian models. In the Danish model, this was a significant interaction between father's reported sports participation and sex, although stratified analyses showed that the association was non-significant in both boys (β = 15.35, 95% confidence interval (CI) −4.25 to 34.95) and girls (β = −17.26, 95% CI −35.21 to 0.69). The model for Estonia retained two social variables (a positive association with father's BMI and a negative association with socioeconomic position), while four were retained in the Portuguese model (positive associations with mother's sport participation, socioeconomic position and having a TV in the bedroom and a negative association with finding the TV on when arriving home).

    Behavioural correlates

    In both the Danish and Portuguese behavioural models, none of the variables or interaction terms was retained. In the Estonian model, one behavioural correlate was retained (negative association with activity during break), as well as a significant interaction between sex and TV viewing before school. Stratified analyses, however, did not show a significant association for either girls (β = 14.17, 95% CI −12.32 to 40.66) or boys (β = −28.41, 95% CI −59.40 to 2.58). Two significant correlates were retained in the Norwegian behavioural model (positive association with TV viewing before school and negative association with hours of computer use per day).

    Discussion and conclusion

    The aim of the current study was to investigate behavioural and social correlates of objectively measured sedentary time in Danish, Estonian, Norwegian and Portuguese youth. The results indicated that participants from the four countries spent different amounts of time in sedentary activities, with Danish participants being the most sedentary. On average, participants spent 262 min/day in sedentary activity, which is lower than previously reported in 12-year-old American girls.35 Whether this is because of a true difference or differences in data collection and processing is difficult to ascertain. For example, the American study not only used a lower cut point to define sedentary time (≤100 counts/min) but also imputed missing data and did not exclude accidental night-time wearing. Consistent with previous reports, sedentary time was higher in adolescents than in children,4,20 22 suggesting that early intervention may be needed to prevent this increase. While Estonian boys were less sedentary than Estonian girls in both age groups, no sex differences in sedentary time were observed in the three remaining countries. This is in contrast to physical activity research, which consistently shows boys being more active than girls.36,–,38 However, previous research has shown no difference between sexes in self-reported total sedentary time either,22 although the children did engage in different behaviours.

    Despite considering a wide range of potential correlates of sedentary time, only a few significant associations were observed in the multivariable analyses, as most of the univariate observations were attenuated after controlling for age and sex. This was mainly due to the fact that the prevalence of many of the exposure variables differed per age group. For example, adolescents reported going to bed later than children, explaining the observed positive association with bed time in the univariate analyses. Nevertheless, the observed associations were consistent across age groups and sex.

    To our knowledge, this is the first study aiming to identify which specific behaviours contribute to overall sedentary time. We considered seven self-reported behaviours ranging from various screen-based behaviours to active travel. Results were inconsistent across countries. No associations with behavioural factors were observed in Denmark or Portugal. In Estonia, activity during school breaks was associated with sedentary time, suggesting that decreasing sedentary activities during school break times might be a target for intervention. This might be achieved by environmental changes to the playground or by providing more equipment.39 In Norway, both high computer use during leisure time and having a games console were associated with higher sedentary time, independent of age and sex. An intervention targeting reductions in computer use during leisure time might therefore be an effective strategy to reduce sedentary time in Norway. In contrast, Norwegian participants reporting television viewing for more than 1 h before school were less sedentary than those who watched less, independent of computer usage, age and sex. Only 9% of Norwegian participants reported longer periods of TV viewing before school, and it might reflect behaviour of highly active participants. Previous work has shown that high sedentary time can coexist with high levels of physical activity.7 8

    We studied social correlates of sedentary time in three main groups of influence; factors related to the parents, factors providing opportunities for engaging in sedentary behaviour and psychosocial factors. Parental factors were only retained in the Estonian and Portuguese multivariable models with higher socioeconomic status being associated with more sedentary time. This is in contrast with previous work, which tends to show that children from lower socioeconomic backgrounds spend more time on sedentary activity20 or watching TV.18 40 This evidence, however, mostly comes from work in Western (Anglo-Saxon) countries, with different cultural and socioeconomic backgrounds than Portugal and Estonia. Previous work has shown that in rapidly developing countries, health problems such as childhood obesity are more likely to occur in children from higher socioeconomic backgrounds than in those from lower socioeconomic backgrounds.41 Previous EYHS analyses also showed that higher family income and parental educational attainment were associated with increased insulin resistance in Portugal and Estonia, while the opposite was true for Danish participants.34 These results indicate that the role of socioeconomic status in influencing health behaviours across countries is complex and warrants further investigation.

    Child's BMI was not associated with sedentary time in three of the four countries, in contrast to previous work consistently showing positive associations.18 22 Unexpected and independent inverse associations with child's and father's BMI were observed in the Estonian participants. BMI is known to be a fairly crude measure of body composition, influenced by fat mass as well as fatfree mass.42 It could be hypothesised that both these children and their fathers are highly active, but also engage in substantial amounts of sedentary behaviour. This is supported by the observation that among Estonian participants, no correlation was observed between minutes spent in at least moderate intensity physical activity and sedentary time (r =−0.03; data not shown).

    Sedentary time in Portuguese participants was influenced by factors from all three categories. Unexpectedly, children with active mothers and those who reported that the TV was only on twice a week or less when they came home, spent more time in sedentary activities. This might indicate that Portuguese children who have more autonomy to decide what sedentary activities they do, such as watching TV after school, are more likely to choose to spend more of their time being sedentary. Portuguese children who reported having a TV in their bedroom were also more sedentary than those who did not, a finding not observed in the other countries. Mixed results on the association between having access to your own television and TV viewing have previously been reported,18 40 43 44 and the current analyses seem to support the idea that these associations might be culturally dependent.

    The analyses presented in this paper were conducted in a large data set and allowed for investigating differences in associations between various European countries. In addition, we included a large number of exposure variables as potential correlates of sedentary time and studied effect modification by age and sex. However, we conducted a large number of individual statistical tests, which might have led to type 1 errors. Due to the cross-sectional nature of the design, no conclusion regarding direction of associations can be drawn. Half of the recruited sample did not provide valid data on sedentary time, with a higher proportion of dropout in Portugal. Those not included did not differ on BMI but were more likely to be adolescent boys (data not shown), limiting the generalisability for male adolescent populations. As mentioned previously, there is no consensus regarding cut-off points to use for defining sedentary activity in children. As a sensitivity analysis, we re-ran our final models using a cut-off point of 500 counts/min. In general, results were consistent with those presented here. A few significant associations disappeared, however, possibly indicating that this higher cut-point also includes behaviours with an intensity level higher than sedentary. A lack of behaviour-specific validity of the exposure measures may explain why only few associations were observed. The questionnaire administered was validated, but is primarily aimed at assessing physical activity correlates.26

    As in previous studies on correlates of sedentary behaviour,18 most of the variance in objectively measured sedentary time was explained by the general demographic variables included in the models rather than the behavioural or sociocultural factors. This is consistent with studies investigating correlates of objectively measured physical activity, in which only small amounts of variance could be explained.45 46 Sedentary time clearly represents a wide range of independent behaviours which are culturally distinct. A more diverse range of behaviours should be considered in future studies to accurately capture sedentary time, such as computer use for homework versus gaming, mobile phone use and reading, aiding the identification of useful intervention targets.

    Conclusion

    The aim of the current paper was to study behavioural and social correlates of overall sedentary time. Sedentary time increased substantially with increasing age, indicating that early intervention is needed to prevent the development of health problems associated with sedentary behaviour. Culturally distinct profiles of sedentary behaviours and correlates emerged from the analyses, indicating that target populations and behaviours should differ between the countries. Future work should aim to use objective measures of sedentary time to study larger populations that are homogeneous on key non-modifiable exposures (such as age, sex and country) to identify suitable targets for intervention within these populations. In addition, they should focus on increasing the precision of the exposure measures and on identifying the most appropriate behaviours that represent time spent in sedentary activity. This strategy should help identify suitable targets for intervention efforts and how to promote positive changes to reduce time spent in sedentary activity in youth.

    What is already known on this topic

    Sedentary activity appears to be associated with health outcomes, independent of physical activity. Previous work have focussed on studying correlates of TV viewing, whereas two studies identified various correlates of self-reported sedentary activity.

    To date, no study has studied correlates of objectively measured sedentary activity.

    What this study adds

    Sedentary activity is associated with a variety of behavioural and social factors, and associations are different for different countries. A single strategy aimed at reducing sedentary behaviour is therefore unlikely to be effective across Europe as the target populations and behaviours of focus differ between countries.

    Acknowledgments

    The authors thank the children and their parents for their participation in the European Youth Heart Study, and those involved in the collecting of the data for their effort.

    Appendix 1

    Description of how data on activity during morning break and lunch break were combined to create the new variable “activity during breaks” and new categories “sedentary”, “variable” and “active”

    View this table:
    Table 1

    Previous treatments of participants

    References

    1. ↵
      1. Telama R,
      2. Yang X,
      3. Viikari J,
      4. et al
      . Physical activity from childhood to adulthood: a 21-year tracking study. Am J Prev Med 2005 Apr;28:267–73.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Twisk JW,
      2. Kemper HC,
      3. van Mechelen W
      . Tracking of activity and fitness and the relationship with cardiovascular disease risk factors. Med Sci Sports Exerc 2000;32:1455–61.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      Global strategy on diet, physical activity and health. WHO 2004. http://www.who.int/dietphysicalactivity/strategy/eb11344/en/index.html (accession date 1 Jun 2009)
    4. ↵
      1. Spanier PA,
      2. Marshall SJ,
      3. Faulkner GE
      . Tackling the obesity pandemic: a call for sedentary behaviour research. Can J Public Health 2006 May–Jun;97:255–7.
      OpenUrlPubMedWeb of Science
    5. ↵
      1. Biddle SJ,
      2. Gorely T,
      3. Marshall SJ,
      4. et al
      . Physical activity and sedentary behaviours in youth: issues and controversies. J R Soc Health 2004 Jan;124:29–33.
      OpenUrlCrossRefWeb of Science
    6. ↵
      1. Marshall SJ,
      2. Gorely T,
      3. Biddle SJH
      . A descriptive epidemiology of screen-based media use in youth: a review and critique. J Adolesc 2006;29:333–49.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Marshall SJ,
      2. Biddle SJ,
      3. Sallis JF,
      4. et al
      . Clustering of sedentary behaviors and physical activity among youth: a cross-national study. Pediatr Exerc Sci 2002;14:401–17.
      OpenUrlWeb of Science
    8. ↵
      1. Zabinski MF,
      2. Norman GJ,
      3. Sallis JF,
      4. et al
      . Patterns of sedentary behavior among adolescents. Health Psychol 2007 Jan;26:113–20.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Robinson TN,
      2. Hammer LD,
      3. Killen JD,
      4. et al
      . Does television viewing increase obesity and reduce physical activity? Cross-sectional and longitudinal analyses among adolescent girls. Pediatrics 1993 Feb;91:273–80.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Marshall SJ,
      2. Biddle SJ,
      3. Gorely T,
      4. et al
      . Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord 2004 Oct;28:1238–46.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Proctor MH,
      2. Moore LL,
      3. Gao D,
      4. et al
      . Television viewing and change in body fat from preschool to early adolescence: the Framingham Children's Study. Int J Obes Relat Metab Disord 2003 Jul;27:827–33.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Ekelund U,
      2. Anderssen SA,
      3. Froberg K,
      4. et al
      . Independent associations of physical activity and cardiorespiratory fitness with metabolic risk factors in children: the European youth heart study. Diabetologia 2007 Sep;50:1832–40.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Salmon J,
      2. Bauman A,
      3. Crawford D,
      4. et al
      . The association between television viewing and overweight among Australian adults participating in varying levels of leisure-time physical activity. Int J Obes Relat Metab Disord 2000 May;24:600–6.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Healy GN,
      2. Dunstan DW,
      3. Salmon J,
      4. et al
      . Objectively measured light-intensity physical activity is independently associated with 2-h plasma glucose. Diabetes Care 2007 Jun;30):1384–9.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Healy GN,
      2. Dunstan DW,
      3. Salmon J,
      4. et al
      . Breaks in sedentary time: beneficial associations with metabolic risk. Diabetes Care 2008 Apr;31:661–6.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Epstein LH,
      2. Kilanowski CK,
      3. Consalvi AR,
      4. et al
      . Reinforcing value of physical activity as a determinant of child activity level. Health Psychol 1999 Nov;18:599–603.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Owen N,
      2. Leslie E,
      3. Salmon J,
      4. et al
      . Environmental determinants of physical activity and sedentary behavior. Exerc Sport Sci Rev 2000 Oct;28:153–8.
      OpenUrlPubMed
    18. ↵
      1. Gorely T,
      2. Marshall SJ,
      3. Biddle SJ
      . Couch kids: correlates of television viewing among youth. Int J Behav Med 2004;11:152–63.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Van Der Horst K,
      2. Paw MJ,
      3. Twisk JW,
      4. et al
      . A brief review on correlates of physical activity and sedentariness in youth. Med Sci Sports Exerc 2007 Aug;39:1241–50.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Brodersen NH,
      2. Steptoe A,
      3. Williamson S,
      4. et al
      . Sociodemographic, developmental, environmental, and psychological correlates of physical activity and sedentary behavior at age 11 to 12. Ann Behav Med 2005 Feb;29:2–11.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Brodersen NH,
      2. Steptoe A,
      3. Boniface DR,
      4. et al
      . Trends in physical activity and sedentary behaviour in adolescence: ethnic and socioeconomic differences. Br J Sports Med 2007 Mar;41:140–4.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Norman GJ,
      2. Schmid BA,
      3. Sallis JF,
      4. et al
      . Psychosocial and environmental correlates of adolescent sedentary behaviors. Pediatrics 2005 October 1;116:908–16.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Reilly JJ,
      2. Coyle J,
      3. Kelly L,
      4. et al
      . An objective method for measurement of sedentary behavior in 3- to 4-year olds. Obes Res 2003 October 1;11:1155–8.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Wareham NJ,
      2. Rennie KL
      . The assessment of physical activity in individuals and populations: why try to be more precise about how physical activity is assessed? Int J Obes Relat Metab Disord 1998 Aug;22(Suppl 2):S30–8.
      OpenUrl
    25. ↵
      1. Riddoch CJ,
      2. Edwards D,
      3. Page AS,
      4. et al
      . The European Youth Heart Study— cardiovascular disease risk factors in children: rationale, aims, study design, and validation of methods. J Phys Act Health 2005;2:115–29.
      OpenUrl
    26. ↵
      1. Ommundsen Y,
      2. Page A,
      3. Ku PW,
      4. et al
      . Cross-cultural, age and gender validation of a computerised questionnaire measuring personal, social and environmental associations with children's physical activity: the European Youth Heart Study. Int J Behav Nutr Phys Act 2008;5:29.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Treuth MS,
      2. Schmitz K,
      3. Catellier DJ,
      4. et al
      . Defining accelerometer thresholds for activity intensities in adolescent girls. Med Sci Sports Exerc 2004 Jul;36:1259–66.
      OpenUrlPubMedWeb of Science
    28. ↵
      1. Eston RG,
      2. Rowlands AV,
      3. Ingledew DK
      . Validity of heart rate, pedometry, and accelerometry for predicting the energy cost of children's activities. J Appl Physiol 1998 Jan;84:362–71.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Puyau MR,
      2. Adolph AL,
      3. Vohra FA,
      4. et al
      . Validation and calibration of physical activity monitors in children. Obes Res 2002 Mar;10:150–7.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Blair SN,
      2. Leary SD,
      3. Ness A,
      4. et al
      . Sedentary behavior and body fat mass in adolescents. Med Sci Sports Exerc 2007;39(Suppl):S379.
      OpenUrl
    31. ↵
      1. Kowalski K,
      2. Crocker P,
      3. Faulkner GE
      . Validation of the physical activity questionnaire for older children. Pediatr Exerc Sci 1997;9:174–86.
      OpenUrlWeb of Science
    32. ↵
      1. Ortega FB,
      2. Ruiz JR,
      3. Sjostrom M
      . Physical activity, overweight and central adiposity in Swedish children and adolescents: the European Youth Heart Study. Int J Behav Nutr Phys Act 2007;4:61.
      OpenUrlCrossRefPubMed
    33. ↵
      1. Ekelund U,
      2. Brage S,
      3. Froberg K,
      4. et al
      . TV viewing and physical activity are independently associated with metabolic risk in children: the European Youth Heart Study. PLoS Med 2006 Dec;3:e488.
      OpenUrlCrossRefPubMed
    34. ↵
      1. Lawlor DA,
      2. Harro M,
      3. Wedderkopp N,
      4. et al
      . Association of socioeconomic position with insulin resistance among children from Denmark, Estonia, and Portugal: cross sectional study. BMJ 2005 July 23;331:183.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Pate RR,
      2. Stevens J,
      3. Pratt C,
      4. et al
      . Objectively measured physical activity in sixthgrade girls. Arch Pediatr Adolesc Med 2006 December 1;160:1262–8.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Sallis JF,
      2. Prochaska JJ,
      3. Taylor WC
      . A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc 2000 May;32:963–75.
      OpenUrlPubMedWeb of Science
    37. ↵
      1. Riddoch CJ,
      2. Bo Andersen L,
      3. Wedderkopp N,
      4. et al
      . Physical activity levels and patterns of 9- and 15-yr-old European children. Med Sci Sports Exerc 2004 Jan;36:86–92.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Trost SG,
      2. Pate RR,
      3. Sallis JF,
      4. et al
      . Age and gender differences in objectively measured physical activity in youth. Med Sci Sports Exerc 2002 Feb;34:350–5.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. van Sluijs EM,
      2. McMinn AM,
      3. Griffin SJ
      . Effectiveness of interventions to promote physical activity in children and adolescents: systematic review of controlled trials. BMJ 2007 Oct 6;335:703.
    40. ↵
      1. Salmon J,
      2. Timperio A,
      3. Telford A,
      4. et al
      . Association of family environment with children's television viewing and with low level of physical activity. Obes Res 2005 Nov;13:1939–51.
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      1. Sakamoto N,
      2. Wansorn S,
      3. Tontisirin K,
      4. et al
      . A social epidemiologic study of obesity among preschool children in Thailand. Int J Obes Relat Metab Disord 2001 Mar;25:389–94.
    42. ↵
      1. McCarthy HD
      . Body fat measurements in children as predictors for the metabolic syndrome: focus on waist circumference. Proc Nutr Soc 2006 Nov;65:385–92.
      OpenUrlPubMedWeb of Science
    43. ↵
      1. Salmon J,
      2. Hume C,
      3. Ball K,
      4. et al
      . Individual, social and home environment determinants of change in children's television viewing: the Switch-Play intervention. J Sci Med Sport 2006 Oct;9:378–87.
      OpenUrlCrossRefPubMedWeb of Science
    44. ↵
      1. Delmas C,
      2. Platat C,
      3. Schweitzer B,
      4. et al
      . Association between television in bedroom and adiposity throughout adolescence. Obesity 2007 October 1;15:2495–503.
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      1. Trost SG,
      2. Saunders R,
      3. Ward DS
      . Determinants of physical activity in middle school children. Am J Health Behav 2002 Mar–Apr;26:95–102.
      OpenUrlPubMedWeb of Science
    46. ↵
      1. Sallis JF,
      2. Taylor WC,
      3. Dowda M,
      4. et al
      . Correlates of vigorous activity for children in grades 1 through 12: comparing parent-reported and objectively measured physical activity. Pediatr Exerc Sci 2002;14:30–44.
      OpenUrlWeb of Science

    Footnotes

    • Funding This study was supported by the following grants—Denmark: Danish Heart Foundation, Danish Medical Research Council, Health Foundation, Danish Council for Sports Research, Foundation of 17-12-1981, Foundation in Memory of Asta Florida Bolding née Andersen, and Faculty of Health Sciences, University of Southern Denmark; Estonia: Estonian Science Foundation grant nos. 3277 and 5209; Norway: Norwegian Council of Cardiovascular Diseases and Eckbo Legacy; Portugal: European Social Fund. None of the funders had an involvement in the study itself or its interpretation.

    • Competing interests None.