Do children's expectations about future physical activity predict their physical activity in adulthood?

Benedetta Pongiglione,Neelaabh Gupta,H Andrew Schwartz,Alissa Goodman,J D Carpentieri,Margaret L Kern

doi:10.1093/ije/dyaa131

Abstract

BackgroundMuch of the population fails to meet recommended physical activity (PA) levels, but there remains considerable individual variation. By understanding drivers of different trajectories, interventions can be better targeted and more effective. One such driver may be a person’s physical activity identity (PAI)—the extent to which a person perceives PA as central to who they are.MethodsUsing survey information and a unique body of essays written at age 11 from the National Child Development Study (N = 10 500), essays mentioning PA were automatically identified using the machine learning technique support vector classification and PA trajectories were estimated using latent class analysis. Analyses tested the extent to which childhood PAI correlated with activity levels from age 23 through 55 and with trajectories across adulthood.Results42.2% of males and 33.5% of females mentioned PA in their essays, describing active and/or passive engagement. Active PAI in childhood was correlated with higher levels of activity for men but not women, and was correlated with consistently active PA trajectories for both genders. Passive PAI was not related to PA for either gender.ConclusionsThis study offers a novel approach for analysing large qualitative datasets to assess identity and behaviours. Findings suggest that at as young as 11 years old, the way a young person conceptualizes activity as part of their identity has a lasting association with behaviour. Still, an active identity may require a supportive sociocultural context to manifest in subsequent behaviour.

Highlights

Regular physical activity (PA) contributes to a range of positive physical and mental health outcomes,[1,2,3,4] yet a large proportion of the population does not meet PA recommendations
We take advantage of: (i) essays written in childhood; (ii) machine learning techniques; and (iii) longitudinal data collected from a large, nationally representative cohort across five decades, to examine the extent to which physical activity identity (PAI) expressed in childhood predicts PA levels and PA trajectories across adulthood
Participants were drawn from the National Child Development Study (NCDS), a UK-based study that has followed a cohort of over 17 000 individuals prospectively across their lives

Summary

Introduction

Regular physical activity (PA) contributes to a range of positive physical and mental health outcomes,[1,2,3,4] yet a large proportion of the population does not meet PA recommendations. From a lifespan epidemiological approach,[19] a person’s past experiences, perceptions, cognitions and habits contribute to subsequent behaviours, including receptiveness towards and engagement in programmes and interventions One such individual aspect may be a person’s physical activity identity (PAI), or the extent to which a person perceives PA as central to who they are.[20] Identity includes the mindset, beliefs and interpretations that a person or group has around different behaviours, cognitions and emotions.[21] In addition to giving meaning and value to past and current behaviour, identity can help shape expectations for the future as well as direct future behaviours.[22]. We take advantage of: (i) essays written in childhood; (ii) machine learning techniques; and (iii) longitudinal data collected from a large, nationally representative cohort across five decades, to examine the extent to which PAI expressed in childhood predicts PA levels and PA trajectories across adulthood

Participants

Results

Discussion