Email updates

Keep up to date with the latest news and content from IJBNPA and BioMed Central.

Open Access Highly Accessed Research

Socio-ecological predictors of participation and dropout in organised sports during childhood

Stewart A Vella*, Dylan P Cliff and Anthony D Okely

Author Affiliations

Early Start Research Institute, Faculty of Social Sciences, Northfields Avenue, University of Wollongong, Wollongong 2522, Australia

For all author emails, please log on.

International Journal of Behavioral Nutrition and Physical Activity 2014, 11:62  doi:10.1186/1479-5868-11-62

The electronic version of this article is the complete one and can be found online at: http://www.ijbnpa.org/content/11/1/62


Received:18 November 2013
Accepted:2 May 2014
Published:13 May 2014

© 2014 Vella et al.; licensee BioMed Central Ltd.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Abstract

Background

The purpose of this study was to explore the socio-ecological determinants of participation and dropout in organised sports in a nationally-representative sample of Australian children.

Methods

Data were drawn from Waves 3 and 4 of the Longitudinal Study of Australian Children. In total, 4042 children aged 8.25 (SD = 0.44) years at baseline were included, with 24-months between Waves. Socio-ecological predictors were reported by parents and teachers, while cognitive and health measures were assessed by trained professionals. All predictors were assessed at age 8, and used to predict participation and dropout by age 10.

Results

Seven variables at age 8 were shown to positively predict participation in organised sports at age 10. These included: sex (boy); fewer people in household; higher household income; main language spoken at home (English); higher parental education; child taken to a sporting event; and, access to a specialist PE teacher during primary school. Four variables predicted dropout from organised sports by age 10: lower household income; main language spoken at home (non-English); lower parental education; and, child not taken to a sporting event.

Conclusions

The interplay between child sex, socioeconomic indicators, and parental support is important in predicting children’s participation in organised sports. Multilevel and multicomponent interventions to promote participation and prevent dropout should be underpinned by the Socio-Ecological Model and targeted to high risk populations using multiple levels of risk.

Keywords:
Youth sport; Attrition; Physical activity; Public health

Background

Participation in organised sports during childhood and adolescence has important benefits for physical, psychological, and social health [1,2]. Pertinent outcomes include higher rates of physical fitness [3], greater involvement in physical activity over time [4,5] lower rates of sedentary behaviour [6], higher rates of self-esteem, more positive social interactions, decreased levels of depressive symptoms [2], and lower rates of obesity [7]. These important benefits may be underpinned by the substantial contribution that participation in organised sports makes to overall levels of physical activity during childhood and adolescence [8-10]. However, there is some evidence to suggest that the health benefits of participation in organised sport go beyond those attributable to physical activity and may be underpinned by the social nature of sports participation. This is consistent with the greater health benefits associated with team sports over individual sports [2].

Given the high prevalence of sports participation in developed countries such as Australia [11], the health benefits of participation in organised sports are potentially significant at a national level. As a consequence, the alarmingly high rate of dropout from organised sports may also be of significance at the population level. Although two-thirds of all Australian children and adolescents participate in organised sports in any given year [11], participation rates start to decline in late childhood and continue to decline with age [12,13]. The statistics on dropout are ambiguous, however, it has been estimated that up to 35% of youth sport participants drop out of organised sports every year. This figure would also include those participants who transfer between sports, however, it has typically been very hard to distinguish between the two and as such the real prevalence and predictors of dropout are unknown [14]. Nonetheless, the high level of dropout throughout late childhood and early adolescence is associated with corresponding declines in physical activity [13], and increasingly sedentary lifestyles [15].

In order to maximise the public health benefits of organised sports participation it is necessary to understand the determinants of both participation and dropout in organised sports. For example, research has shown that Australian girls of a low socioeconomic position experience a disproportionately high number of barriers to sport participation when compared with their peers [16]. The Socio-Ecological Model has been used as a framework to examine the predictors of children’s health and health behaviours [17-19]. According to this model the predictors of sport participation are multidimensional in nature and operate across individual, interpersonal, community and societal levels [19,20]. Eime and colleagues have developed a conceptual model of Health through Sport [2] whereby socio-ecological constructs at all levels predict participation in organised sports. For example, children with strong physical and social functioning, strong parental support for sport, or access to affordable sports programs may be more likely to participate in organised sports. In turn, participation in organised sports has benefits for physical, psychological, and social health. These health outcomes subsequently feed back into the Socio-Ecological Model as determinants of participation in organised sports, whereby higher rates of health predispose young people to maintain sports participation. For example, sports participation leads to greater levels of physical activity and the development of social skills [5,21], which in turn predispose a child to continuing participation in sports. This model provides a theoretical foundation for the study of participation and dropout in organised sports as it predicts that both social (e.g., parental support) and individual (e.g., adiposity) health-related variables derived from the Socio-Ecological Model will underpin an individual’s participation in organised sports.

The purpose of this study was to explore the socio-ecological determinants of participation and dropout in organised sports in a nationally-representative sample of Australian children. Data from the Longitudinal Study of Australian Children offer a unique opportunity to examine the true predictors of dropout from organised sports by distinguishing between those who had transferred between sports and those who had dropped out. Furthermore, it is important to simultaneously investigate the impact of variables at multiple levels of the Socio-Ecological Model in order to account for the multidimensional influences that interact to determine participation in organised sports. Such analyses are important during childhood as they represent a substantial opportunity to understand the factors which may underpin participation in organised sports at a time prior to the onset of high rates of dropout, and where targeted interventions and public policy may be most beneficial. Importantly, constructs informed by the Conceptual Model of Health through Sport as well as Socio-Ecological Models of childhood obesity and health behaviour are important areas of investigation because these health-related variables are purported to predict children’s participation in organised sports [18,19]. This includes individual, interpersonal, community, and societal level variables.

Method

Study design and participants

Data were obtained from the Kindergarten (K) cohort of the Longitudinal Study of Australian Children (LSAC) [22]. LSAC is a longitudinal examination of the social, environmental and economic impacts on development and wellbeing in a nationally-representative sample of children. A two-stage sampling design was used whereby Australian postal codes were the primary sampling unit and were stratified by state location (rural/urban). In the second stage, children were randomly selected from the Medicare database which is a national health insurance database and is the most comprehensive database of the Australian population. Wave 1 of data was collected in 2004, and data have been collected every two years following. Trained professionals use face-to-face interviews with the children’s primary parent (in more than 96% of cases this is the child’s mother), parental self-report questionnaires, parent-reported time-use diaries, and teacher self-report questionnaires to collect the range of data. For the K-Cohort, there were 4,983 children aged 4-5 included at Wave 1. This represented a total response rate of approximately 50%. The current study used data obtained from the K-Cohort of LSAC during Waves 3 (2008) and 4 (2010), when information on the children’s sport participation was first collected. Thus, it wasn’t possible to assess sports participation or drop out prior to Wave 3. In total, 4,164 children were included in Wave 4, representing an attrition rate of 16.4% across the 4 Waves. However, only children who had complete sports participation data were included in the study. This resulted in a total of 4,042 children from both Waves 3 and 4 when they were aged 8-9 and 10-11 years, respectively. This represents 97% of total available cases. Ethics approval for the LSAC study was given by the Australian Institute of Family Studies Ethics Committee.

Measures

Sports Participation

Sports participation was measured using two items from the LSAC parent-report questionnaire pertaining to their child’s regular participation in team and individual sports at both Wave 3 and Wave 4. Parents were asked “In the last 12 months, has (your) child regularly participated in team sport (e.g. football, cricket or netball)?”, and, “In the last 12 months, has (your) child regularly participated in individual sport (e.g. tennis, karate or gymnastics)?” In order to qualify as participating ‘regularly’, children were required to have participated at least once per week for a duration of three months or more (e.g., a sporting season). Parents could answer either “yes” or “no” for each item. Children were defined as participating in sports if parents answered “yes” to at least one of the items.

Predictor variables

Variables used as potential predictors of participation and dropout in organised sports are shown in Table 1. An extensive number of variables were selected according to the Socio-Ecological models presented by Davison and Birch [18], Sallis and Owen [19], and Eime et al. [2]. Individual level variables included measures of physical (e.g., BMI, waist circumference), social (e.g., social functioning), and emotional health (e.g., emotional functioning, mental health), as dictated by the Conceptual Model of Health through Sport. In addition, individual level health behaviours which may displace (e.g., sedentary behaviours) or promote (e.g., physical activity) sports participation, and measures of children’s temperament (e.g., sociability, persistence, temper) which may predispose children to participate or dropout in organised sports were also included. Interpersonal variables included measures of parental support for physical activity and sport (e.g., whether parents took the child to a sporting event, or whether parents were physically active with the child), as well as relevant parenting styles which may underpin a child’s participation (e.g., warm parenting) or dropout (e.g., harsh parenting). Relevant demographic variables which have been shown to be associated with sports participation at the community level (e.g., cultural background, household income, parental education) may also be important predictors and were included [6]. Finally, relevant societal level variables such as measures of societal infrastructure (e.g., access to parks and playgrounds, neighbourhood safety) and school support (e.g., access to specialist physical education teachers and time spent in physical education) may also promote and facilitate participation or dropout in organised sports and were also included. Information presented in Table 1 gives details of the measure that was used, number of items, response format, source of report, and reliability and validity information where appropriate and available. All predictor variables were assessed at Wave 3 when participants were aged 8/9 for their ability to predict sports participation over the subsequent two year period.

Table 1. Measures used as socio-ecological predictors of organised sports participation and dropout

Statistical analyses

All data were analysed using IBM SPSS statistical software (version 19, IBM, New York, United States). Selected demographic characteristics and predictor variables were used to summarise the characteristics of the sample, including weight status [31]. We then conducted two separate analyses to examine (i) predictors of dropout in organised sports, and (ii) predictors of participation in organised sports. In order to assess predictors of dropout we examined all children who were participating in sport at age 8 to determine which variables predicted dropout by age 10. In order to assess predictors of participation in organised sports we conducted analyses on all children at age 8 to examine which variables predicted sports participation at age 10. Both analyses followed the same strategy. At the first stage all potential predictor variables were entered individually into an unadjusted binary logistic regression model. Significance at this stage was set at p < 0.05. For each comparison pair potential collinearity was assessed amongst all significant predictor variables using the estimated individual variance inflation factor (VIF), tolerance for each predictor, and the mean VIF. Items with a VIF of >10 and a tolerance of <0.1 were to be removed from subsequent analyses [32,33]. Where the mean VIF was significantly greater than 1, but no predictor variable had an individual VIF >10, the variable with the highest VIF was to be removed. However, no collinearity was detected and no items were removed from secondary analyses. In the second stage of analyses all predictor variables were entered into fully adjusted binary logistic regression models in order to examine potentially significant predictors after adjusting for the effect of all other significant predictor variables. The significance level at this stage was set at p < 0.05. Lastly, the Hosmer and Lemeshow test was used to assess goodness-of-fit for each of the final binary logistic regression models.

Results

Participants

There were 4,164 participants who provided data at Wave 4 of the LSAC study. Of those, a total of 4,042 (97%) had complete sports participation data at both Wave 3 and Wave 4, and were included in analyses. Participants with complete data in Waves 3 and 4 (n = 122) were more likely to have a parent that was older, hold a tertiary education degree, and speak English at home. Complete data were also more likely for non-Indigenous children, children enrolled in Government schools (vs. Independent schools), and in areas where higher proportions of residents had completed high school [34,35]. Selected demographic characteristics of the entire sample and by grouping are shown in Table 2. At age 10, organised sports participants were more likely to be boys, speak English as a main language, were of a higher socioeconomic position (all p < .001), and were more physically active (p = .011) than nonparticipants. There were no differences according Indigenous status, BMI, or sedentary time. Children who dropped out between the ages of 8 and 10 were more likely to be girls (p < .001), non-English speaking (p = .002), and Indigenous (p = .026), were of a lower socioeconomic position (p < .001), and were less physically active at age 8 (p = .024). There were no differences according to BMI or sedentary time.

Table 2. Selected demographic characteristics of the sample and by participation groupa

Unadjusted analyses

Table 3 presents the results of the unadjusted analyses used to assess the significant Socio-Ecological predictors of participation and dropout in organised sports. In total, 24 of the 48 variables entered as potential predictors of participation in organised sports at age 10 were significant in unadjusted analyses and were retained for the fully adjusted model. This included a range of individual, interpersonal/family, and community level predictors. Of note, Indigenous status, BMI, waist circumference, and general health at age 8 did not predict sports participation at age 10 in unadjusted analyses.

Table 3. Socio-Ecological predictors of participation and dropout in organised sports in unadjusted analysesa

Twenty-four variables were significantly predictors of dropout from organised sports by age 10 in unadjusted analyses, including a range of individual, interpersonal/family, and community level variables. As with predictors of participation, BMI, waist circumference, and general health were not significant predictors. However, in contrast with participation, Indigenous status was a significant predictor of dropout in unadjusted analyses. All 24 significant predictors were entered into the fully adjusted model.

Predictors of participation

Results of the adjusted model used to predict participation in organised sports at age 10 are presented in Table 4. In total, 7 of the variables entered were found to predict sports participation at age 10 after adjusting for the effect of all 24 predictors found to be significant in unadjusted analyses. At an individual level, sports participants were more likely to be boys. Measures of physical and mental health did not predict subsequent sports participation, nor did physical activity. At an interpersonal/family level, sports participants were more likely to live in houses with fewer people and with a greater standardised household income, and who took them to sporting events. Sports participants were also more likely to have more highly educated parents, and to speak English as a main language at home. Parent’s level of physical activity and parenting style were not predictive of subsequent sports participation. At a community level, children who had access to a specialist Physical Education teacher at age 8 were more likely to participate in organised sports at age 10. Socioeconomic position and neighbourhood safety were not predictive of subsequent sports participation. Overall, the fully adjusted model proved to be a good predictor of participation in organised sports (χ2(8) = 9.14, p = .331), successfully predicting 82.6% of cases.

Table 4. Socio-ecological predictors of participation and dropout in organised sports in adjusted analysesa

Predictors of dropout

Results of the fully adjusted model used to predict dropout from organised sports between ages 8 and 10 are presented in Table 4. Of the 24 variables entered into the fully adjusted model, 4 predicted dropout from organised sports after adjusting for the effects of all other variables. No individual level variables were found to predict dropout. At the interpersonal/family level, children who dropped out of sport were less likely to have been taken to a sporting event in the last month, resided in a household with a lower standardised income, had parents with lower educational attainment, and were more likely to be non-English speakers at age 8. Notably, child sex, Indigenous status, various physical and mental health measures, and all community level variables were not predictive of dropout from organised sports. Overall, the fully adjusted model proved to be a good predictor of dropout from organised sports (χ2(8) = 5.41, p = .219), and successfully predicted 89.6% of cases.

Discussion

In this large and nationally representative sample of Australian children, prospective analyses between the ages of 8 and 10 years revealed 7 socio-ecological predictors of participation in organised sports during childhood, and 4 predictors of dropout. Of these, there was a distinct overlap between the predictors of participation and dropout in organised sports where indicators of greater social disadvantage predicted both nonparticipation and dropout in organised sports. This is consistent with previous research which shows that differences in children’s sports participation are best accounted for by sociocultural and socioeconomic indicators [6,36]. This study also found that children being taken to a sporting event at age 8 predicted participation and maintenance of sports participation at age 10. This may provide an indication that, consistent with the correlates of physical activity [37], parental support is an important facilitator of sports participation among children. This may be especially so because parents or caregivers typically play a key role in children’s enrolment and continued participation in organised sports by way of finance, time, and transportation. Parents’ investment in adolescents’ sporting endeavours has been shown be highly predictive of dropout from organised sport [38]. Together, these results show that sports participation during childhood can best be predicted by a combination of socioeconomic, cultural, and parental support variables.

Previous Australian research has shown that socioeconomic variables and parental support may not be independent of one another [16]. Dollman and Lewis found that children of a lower socioeconomic status reported a greater number of barriers to sports participation and lower levels of both instrumental and affective parental support for playing sport than children of a higher socioeconomic status. Furthermore, these differences were more pronounced in girls. This is consistent with the results of the current study where child sex, socioeconomic indicators, and parental support predicted sports participation, and is supported by the results of Toftegaard-Støckel and colleagues [36] who found that socioeconomic and sociocultural factors were the greatest determinants of adolescent’s sport participation in a sample of Danish children. It is possible that there exists an important interplay between child sex, socioeconomic variables, and parental support in the prediction of children’s participation in organised sports. In particular, girls from lower socioeconomic areas who receive little parental support are most at risk of nonparticipation.

The current results reinforce the importance of using the Socio-Ecological Model to study participation and dropout in organised sports because it enables simultaneous adjustment for the influence of factors at multiple levels of influence [20]. It is imperative, therefore, that interventions to promote participation and prevent dropout in organised sports are appropriately targeted at populations who are at highest risk by accounting for risk at multiple levels (individual, familial, and societal). This is important because dropout from organised sports and nonparticipation may have short- and long-term health consequences for children, including higher rates of obesity and physical inactivity [5,7]. This is particularly problematic for children from a low socioeconomic position and minority cultural backgrounds who are already at greater risk of health conditions such as obesity [39,40]. Furthermore, to maximise their effectiveness these interventions should include multiple components aimed at intervening at multiple levels of influence. This may include issues such as access and affordability, parental support, as well as children’s motivation for organised sports. Such factors have been prioritised in both government and non-government reports in Australia which aim is to increase national participation in organised sports and physical activity [41,42].

A particularly novel finding of this study is that children who had access to a specialist physical education teacher at school at the age of 8 were more likely to participate in organised sports at age 10. This was over and above the impact of socioeconomic status, and it is therefore unlikely that this is due to the possibility that schools of a higher socioeconomic status being more likely to employ a specialist physical education teacher. Therefore, it may be that access to a specialist physical education teacher predisposes children to participation in organised sports. For example, specialist physical education teachers in primary schools have been shown to facilitate greater sport-related outcomes such as higher levels of physical activity, greater fitness, and greater levels of fundamental movement skills when compared with untrained teachers [43-45]. Thus, it is possible that access to a specialist physical education teacher can predispose children to participate in organised sports by increasing skills and behaviours that are conducive to successful participation in organised sports.

This study found no evidence to support the hypothesis that physical, social, or psychological health are predictors of organised sports participation during childhood. Given that overweight and obese Australian youth participate in organised sports at the same rate as their normal weight peers [6], it is possible that some Australian children are being enrolled in organised sports to improve their health, and thus, measures of physical health are not predictors of participation. This may also be true for children with lower social or psychological health who may be enrolled in sport in order to gain social or psychological benefits. More research is needed to investigate the directionality of relationships between participation in organised sports and measures of physical, social and psychological health. In particular, potential bi-directionality should be investigated in light of the Conceptual Model of Health through Sport [2] which proposes a feedback loop that operates between the physical, social, and psychological health benefits of sports participation and the likelihood that children will commence or maintain participation in sports.

A strength of this study was to distinguish between children who dropped out of all organised sports and children who transferred between sports, thus allowing the prediction of dropout. To our knowledge, this is the first study to do so. This study also examined a large number of socio-ecological predictors of participation and dropout in organised sports during childhood, thus allowing a comprehensive analysis of potential predictors. However, potentially important sport-specific predictors were not included. Research shows that a child’s sporting experience is highly predictive of dropout, and the omission of sport-specific predictors limits the application of the study findings. It is currently unclear what the respective influence of sport-specific factors is when compared with socio-ecological predictors such as those presented in this study. Furthermore, while the large range of socio-ecological predictors included in the study is a strength, the validity and reliability of some measures used is unknown. In addition, parental or teacher report of some child variables such as physical activity level and gross motor coordination are subjective measures and may not provide a reliable measure of these important variables.

Future studies and interventions should consider participation and dropout through the lens of the Socio-Ecological Model [20]. In particular, attention should be payed to the interplay of child sex, sociocultural and socioeconomic indicators, and parental support. Given the consistency of studies in this area, culminating in these prospective and nationally-representative data, interventions to promote sports participation and prevent dropout from organised sports are urgently required and must be targeted to populations at highest risk of dropout and nonparticipation. These include girls, children of lower socioeconomic backgrounds, and children who receive low parental support. Preferential allocation should be given to children who meet all three criteria. Interventions should include multiple components and intervene at multiple levels of influence.

Abbreviations

BMI: Body mass index; LSAC: Longitudinal study of Australian children; PA: Physical activity; PedsQL: Pediatric quality of life scale; SDQ: Strengths and difficulties questionnaire; VIF: Variance inflation factor.

Competing interests

The authors have no competing interests to disclose.

Authors’ contributions

SV conceptualised and designed the study, carried out the initial analyses, drafted the initial manuscript, and approved the final manuscript as submitted. DC conceptualised and designed the study, reviewed and revised the manuscript, and approved the final manuscript as submitted. AO conceptualised and designed the study, reviewed and revised the manuscript, and approved the final manuscript as submitted. All authors read and approved the final manuscript.

Authors’ information

SV is a postdoctoral research fellow at the Early Start Research Institute, University of Wollongong. DC is a National Heart Foundation of Australia postdoctoral research fellow at the Early Start Research Institute, University of Wollongong. AO is a Professor at the Early Start Research Institute, University of Wollongong.

Acknowledgements

The Longitudinal Study of Australian Children is conducted in partnership between the Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA), the Australian Institute of Family Studies (AIFS) and the Australian Bureau of Statistics (ABS). They were responsible for the design and conduct of the study, and the collection and management of data. The findings and views reported in this paper are those of the authors and should not be attributed to FaHCSIA, AIFS or the ABS. Dylan Cliff is funded by a National Heart Foundation of Australia - Postdoctoral Research Fellowship (PH 11S 6025). Anthony Okely is supported by a National Heart Foundation of Australia Career Development Fellowship (CR 11S 6099).

References

  1. Janssen I, LeBlanc AG: Systematic review of the health benefits of physical activity and fitness in school-aged children and youth.

    Int J Beh Nutr Phys Activ 2010, 7:40. BioMed Central Full Text OpenURL

  2. Eime RM, Young JA, Harvey JT, Charity MJ, Payne WR: A systematic review of the psychological and social benefits of participation in sport for children and adolescents: informing development of a conceptual model of health through sport.

    Int J Beh Nutr Phys Activ 2013, 10:98. BioMed Central Full Text OpenURL

  3. Hoffman JR, Kang J, Faigenbaum AD, Ratamess NA: Recreational sports participation is associated with enhanced physical fitness in children.

    Res Sports Med Int J 2005, 13:149-161. Publisher Full Text OpenURL

  4. Barnett TA, O’Loughlin JO, Paridis G: One- and two-year predictions of decline in physical activity among inner-city schoolchildren.

    Am J Prev Med 2002, 23:121-128. PubMed Abstract | Publisher Full Text OpenURL

  5. Nelson TF, Stovitz SD, Thomas M, LaVoi NM, Bauer KW, Neumark-Sztainer D: Do youth sports prevent pediatric obesity? a systematic review and commentary.

    Curr Sports Med Rep 2011, 10:360-370. PubMed Abstract | Publisher Full Text OpenURL

  6. Vella SA, Cliff DP, Okely AD, Scully M, Morley B: Associations between organized sports participation and obesity-related health behaviors in Australian adolescents.

    Int J Beh Nutr Phys Activ 2013, 10:113. BioMed Central Full Text OpenURL

  7. Drake K, Beach ML, Langacre MR, MacKenzie T, Titus LJ, Rundle AG, Dalton MA: Influence of sports, physical education, and active commuting to school on adolescent weight status.

    Pediatrics 2012, 130:e296-e304. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  8. Booth M, Okely AD, Chey T, Bauman A: Patterns of activity energy expenditure among Australian adolescents.

    J Phys Activ Health 2004, 1:246-258. OpenURL

  9. Katzmarzyk P, Malina R: Contribution of organized sports participation to estimated daily energy expenditure in youth.

    Pediatr Exerc Sci 1998, 10:378-386. OpenURL

  10. Wickel EE, Eisenmann JC: Contribution of youth sport to total daily physical activity among 6- to 12-year old boys.

    Med Sci Sports Exerc 2007, 39:1493-1500. PubMed Abstract | Publisher Full Text OpenURL

  11. Australian Bureau of Statistics: Children’s participation in sport and leisure time activities 2003-2012. Canberra: Australian Bureau of Statistics; 2012. OpenURL

  12. Zimmerman-Sloutskis D, Wanner M, Zimmerman E, Martin BW: Physical activity levels and determinants of change in young adults: a longitudinal panel study.

    Int J Beh Nutr Phys Activ 2010, 7:2. BioMed Central Full Text OpenURL

  13. Department of Health and Ageing: Australian National Children’s Nutrition and Physical Activity Survey: Main Findings. Canberra: Department of Health and Ageing; 2007. OpenURL

  14. Hedstrom R, Gould D: Research in youth sports: Critical issues status. Kansas City, Mo: Institute for the Study of Youth Sports; 2004.

  15. Gebremariam MK, Bergh IH, Andersen LF, Ommundsen Y, Totland TH, Bjelland M, Grydeland M, Lien N: Are screen-based sedentary behaviors longitudinally associated with dietary behaviors and leisure-time physical activity in the transition into adolescence?

    Int J Beh Nutr Phys Activ 2013, 10:9. BioMed Central Full Text OpenURL

  16. Dollman J, Lewis NR: The impact of socioeconomic position on sport particiation among South Australian youth.

    J Sci Med Sport 2010, 13:318-322. PubMed Abstract | Publisher Full Text OpenURL

  17. Hinkley T, Salmon J, Okely AD, Hesketh K, Crawford D: Correlates of preschool children’s physical activity.

    Am J Prev Med 2012, 43:159-167. PubMed Abstract | Publisher Full Text OpenURL

  18. Davison KK, Birch LL: Childhood overweight: a contextual model and recommendations for future research.

    Obes Rev 2001, 2:159-171. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  19. Sallis J, Owen N: Ecological models of health behavior. In Health Behavior and Health Education: Theory, research, and practice. Edited by Glanz K, Rimer B, Lewis F. San Francisco: Jossey-Bass; 2002:462-485. OpenURL

  20. Bronfenbrenner U: The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press; 1979. OpenURL

  21. Fraser-Thomas JL, Côté J, Deakin J: Youth sport programs: an avenue to foster positive youth development.

    Phys Educ Sport Pedagogy 2005, 10:19-40. Publisher Full Text OpenURL

  22. Sanson A, Nicholson J, Ungerer J, Zubrick S, Wlison K, Ainley J, Berthelsen D, Bittman M, Broom D, Harrison L, Rodgers B, Sawyer M, Silburn S, Strazdins L, Vimpani G, Wake M: Introducing the Longitudinal Study of Australian Children. Melbourne, Australia: Australian Institute of Family Studies; 2002.

  23. Varni JW, Seid M, Kurtin PS: PedsQL4.0: reliability and validity of the Pediatric Quality of Life Inventory Version 4.0 generic core scales in healthy and patient populations.

    Med Care 2001, 39:800-812. PubMed Abstract | Publisher Full Text OpenURL

  24. Goodman R: The strengths and difficulties questionnaire: a research note.

    J Child Psychol Psyc 1997, 38:581-586. Publisher Full Text OpenURL

  25. McClowry SG: The development of the school-age temperament inventory.

    Merrill-Palmer Q 1995, 41:271-285. OpenURL

  26. Australian Bureau of Statistics: AusStats 6523.0 Household income and income distribution, Australia 2003-04 (Appendix 3). Canberra, Australia: Australian Bureau of Statistics; 2005. OpenURL

  27. Paterson G, Sanson A: The association of behavioral adjustement to temperament, parenting and family characteristics among 5-year-old children.

    Soc Dev 1999, 8:293-309. OpenURL

  28. Statistics Canada: National Longitudinal Survey of Children and Youth (NLSCY) Cycle 3 Survey Instruments: Parent Questionnaire. Statistics Canada: Ottawa, Ontario, Canada; 2000. OpenURL

  29. Bayer J, Sanson A, Hemphill S: Parent influence on early childhood internalizing difficulties.

    J Appl Dev Psychol 2006, 27:542-559. Publisher Full Text OpenURL

  30. Commonwealth Department of Health and Ageing: Measuring remoteness: Accessibility/Remoteness Index of Australua (ARIA). Canberra: Commonwealth Government of Australia; 2001. OpenURL

  31. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH: Establishing a standard definition for child overwieght and obesity worldwide: international survey.

    BMJ 2000, 320:1-6. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  32. Hair JF, Anderson RE, Tatham RL, Black WC: Multivariate data analysis. Englewood Cliffs NJ: Prentice Hall; 1998. OpenURL

  33. Kutner MH, Nachtsheim CJ, Neter J, Li W: Applied linear statistical models. New York: McGraw-Hill Irwin; 2005. OpenURL

  34. Daraganova G, Sipthorp M: Longitudinal Study of Australian Children: LSAC technical paper no. 9: Wave 4 weights. Melbourne, Australia: Australian Institute of Family Studies; 2011.

  35. Sipthorp M, Misson S: The longitudinal study of Australian children: LSAC technical paper no. 6: Wave 3 weighting and non-response. Melbourne, Australia: Australian Institute of Family Studies; 2009.

  36. Toftegaard-Stockel J, Nielsen GA, Ibsen B, Andersen LB: Parental, socio and cultural factors associated with adolescents’ sports participation in four Danish municipalities.

    Scan J Med Sci Sport 2011, 21:606-611. Publisher Full Text OpenURL

  37. Van Der Horst K, Chin A, Paw MJ, Twisk JWR, Van Mechelen W: A brief review on correlates of physical activity and sedintariness in youth.

    Med Sci Sports Exerc 2007, 39:1241-1250. PubMed Abstract | Publisher Full Text OpenURL

  38. Boiche JCS, Sarrazin PG: Proximal and distal factors associated with dropout versus maintained participation in orgnized sport.

    J Sport Sci Med 2009, 8:9-16. OpenURL

  39. Morley BC, Scully ML, Niven PH, Okely AD, Baur LA, Pratt IS, Wakefield M: What factors are associated with excess body weight in Australian secondary school students?

    Med J Aust 2012, 196:189-192. PubMed Abstract | Publisher Full Text OpenURL

  40. Stamatakis E, Wardle J, Cole TJ: Childhood obesity and overweight prevalence trends in England: evidence for growing socioeconomic disparities.

    Int J Obes 2010, 34:41-47. Publisher Full Text OpenURL

  41. Commonwealth of Australia: The future of sport in Australia. Canberra: Commonwealth of Australia; 2009. OpenURL

  42. National Heart Foundation of Australia: Blueprint for an active Australia. Perth: National Heart Foundation of Australia; 2009. OpenURL

  43. McKenzie TL, Stone EJ, Feldman HA, Epping JN, Yang M, Strikmiller MS, Lytle LA, Parcel GS: Effects of the CATCH physical education intervention: teacher type and lesson location.

    Am J Prev Med 2001, 21:101-109. PubMed Abstract | Publisher Full Text OpenURL

  44. Sallis J, McKenzie TL, Alcaraz J, Kolody B, Faucette N, Hovell M: The effects of a 2-year physical education program (SPARK) on physical activity and fitness in elementary school students. Sports, Play and Active Recreation for Kids.

    Am J Pub Health 1997, 87:1328-1334. Publisher Full Text OpenURL

  45. Luctertini F, Spazzafumo L, De Lillo F, Centronze D, Valentini M, Federici A: Effectiveness of professional-guided physical education on fitness outcomes of primary school children.

    Eur J Sport Sci 2013, 13:582-590. PubMed Abstract | Publisher Full Text OpenURL