RESUMO
BACKGROUND: The present study examines (i) if the level of vigorous physical activity (VPA) at age 15 predicts low VPA at ages 19 and 27 and (ii) whether the observed prediction pattern differs by childhood socio-economic position (SEP). In this way, prediction analyses are applied to study tracking behaviour. METHODS: Data are from The Danish Longitudinal Health Behaviour Study. The baseline survey was conducted in 1990 at age 15, the first follow-up in 1994 at age 19, and the second follow-up in 2002 at age 27, n = 561. The study population was a random sample of the Danish population selected from the National Civic Registration System, and data were collected by anonymous postal questionnaires. The indicator of childhood SEP was family occupational social class. Prediction analyses are conducted by stratified logistic regression analyses. RESULTS: There was a significant and marked predictive power of low levels of VPA in mid adolescence (aged 15) for low VPA in late adolescence (age 19) [odds ratio (OR)=4.95 (2.83-8.66)], from late adolescence (age 19) into early adulthood (age 27) [OR = 2.71 (1.61-4.55)] and also over the full study period from age 15 to age 27 [2.91 (1.72-4.94)]. Analyses stratified by SEP revealed that the predictive power of VPA at age 19 for low VPA at age 27 was only significant among participants from low SEP. CONCLUSION: These findings suggest that low VPA tracks through adolescence while tracking into adulthood only occurs among individuals with low childhood SEP.
