Physical activity and associated factors in high-school adolescents in Southern Brazil

OBJECTIVE: To estimate the prevalence of physical activity in adolescents and to identify associated factors. METHODS: A cross-sectional study was conducted with a representative sample (n=1,518, 59.2% females) of students aged between 14 and 18 years, enrolled in the public school network of the city of Curitiba, Southern Brazil, in 2006. Physical activity practice was self-reported, according to the number of days per week when they perform moderate to vigorous physical activity lasting >= 60 minutes. This practice was analyzed in two distinct models. In the first model, the variable was dichotomized into "0 day" and ">= 1 day"; in the second, into "<= 4 days" and ">= 5 days". Independent variables were as follows: biological-demographic (sex, age, body mass index); socioeconomic (parents' level of education, number of cars); behavioral (number of hours spent watching television, number of hours spent using a computer); and sociocultural (social support from family and friends and the perception of barriers to the practice of activities), tested with Poisson regression. RESULTS: In the first model of analysis, the prevalence of physical activity was 58.2% (75.1% in males; 46.5% in females; p<0.001), while, in the second, it was 14.5% (22.3% and 9.1%, respectively; p<0.001). In the first model, the variables associated with physical activity were: male sex (PR=1.63, 95% Cl: 1.48;1.78), social support from family (PR=1.14, 95% CI: 1.05;1.23), social support from friends (PR=1.52, 95% Cl: 1.31;1.78) and high perception of barriers (PR=0.54, 95% Cl: 0.46;0.62). In the second model, only male sex (PR=2.45, 95% Cl: 1.73;3.46) and high perception of barriers (PR=0.24, 95% CI: 0.15;0.38) were associated with physical activity. CONCLUSIONS: More than half of adolescents practice physical activity at least one day of the week, although 14.5% achieved the current recommendations. The recommended levels are associated with a lower number of factors. Gender and perception of barriers were consistently associated with physical activity levels.