Degree Name

Doctor of Philosophy


School of Health and Society


Background Physical activity (PA) participation among youth tends to decline with age. Information about longitudinal changes in domains of PA may provide a more nuanced understanding of how this decline occurs, which may lead to more specific behavioural targets for intervention. The overall aim of this thesis was to explore changes in PA domains during the transition from childhood to adolescence and the factors that may influence those changes, with a particular focus on non-organised PA. Non-organised PA includes unstructured and freely-chosen activities that occur for their own sake (e.g. active playground games or informal sports).

Methods A mixed-method sequential explanatory research design was adopted. Firstly, a systematic literature review investigated longitudinal trends in organised PA, non-organised PA, active transport and active chores/work during childhood and adolescence. Following this, a quantitative research phase used national data from the Longitudinal Study of Australian Children (LSAC). Domain-specific PA participation was measured at 11y (2010), 13y (2012) and 15y (2014) using a 24-hour time-use diary. Analyses included multilevel mixed-effects models, regression models, and two-stage cluster analysis. Finally, a qualitative research phase was conducted with young adults (18-22y) who reported withdrawing from non-organised PA and overall PA between 11y and 15y. Interview participants were recruited from undergraduate/vocational classes and churches in the Illawarra region, Australia. Recruitment was stratified by sex and socioeconomic status to promote diversity of voices. Data analysis was supported by concept and pattern coding.

Results The systematic review included 23 studies which cumulatively reported data from 27,231 participants. Few studies had a low risk of bias (n=6). Only two studies had reported data for non-organised PA in the transition from childhood to adolescence, and the overall synthesis of these studies was inconclusive.

The quantitative phase revealed a large quadratic decline in non-organised PA between 11y, 13y and 15y (-48 min/d, p<0.001) and this was similar to the decline in overall PA (n=4,108). Participation in other domains of PA either increased slightly or remained stable between 11y and 15y. Non-organised PA at 13y was positively predicted by PA enjoyment (OR=1.36, p=0.007) and number of siblings (OR=1.11, p<0.001), and negatively predicted by sex (females) (OR=0.66, p<0.001) and home computer use at 11y (OR=0.98, p=0.002) (n=3,193). Non-organised PA was more likely to decline between 11y and 13y among those from low socioeconomic backgrounds (OR=0.92, p=0.047) and those who consumed more sugary drinks at 11y (OR=1.06, p=0.033). Finally, two segments were identified among youth who reduced their non-organised PA between 11y and 13y (κ = 0.66, n=1,043). The ‘Social Screens’ segment (n=143) had large increases in texting, emailing, social media use (+56 min/day, p<0.001) and other internet use (+32 min/day, p<0.001) between 11y and 13y. Conversely, ‘the Mainstream’ segment (n=900) had smaller increases in a wider range of activities and were more likely to participate in PA.

In the qualitative phase (n=22), themes included adult modelling/influence, concern about being childish, puberty, identity, adult choices and responsibilities, and changing life circumstances. Barriers to non-organised PA included peer judgement and rejection. Enablers of non-organised PA included safe people and places, accessible games and, for girls, having an identity that supported challenging gender norms.

Discussion The domain of non-organised PA has been scarcely investigated among adolescents, and this thesis fills several research gaps. This thesis includes the first simultaneous longitudinal exploration of four domains of PA (organised PA, non-organised PA, active transport, active chores/work) between childhood and adolescence. Other research gaps that have been filled include testing a number of potential predictors of non-organised PA that had not been explored in this context before; identifying segments of youth based on longitudinal changes in activities that may compete with non-organised PA; and exploring age-related norms connected with nonorganised PA during adolescence.

The findings of this thesis may guide social marketing strategies to promote nonorganised PA in the transition from childhood to adolescence in Australia. Future studies may seek to explore and develop potential intervention strategies to promote non-organised PA in this context. Particular intervention approaches may include ‘reframing’ childhood play activities for adolescent audiences and developing ‘identity congruent’ types of PA. These strategies may be targeted at particular population groups such as girls, adolescents without siblings or those with higher computer usage. Such approaches may support youth to participate in intrinsically enjoyable and motivating types of PA.

This thesis is unavailable until Thursday, December 01, 2022



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.