PREDICTORS OF PHYSICAL ACTIVITY IN LATE ADOLESCENCE

Abstract

528 It is generally accepted that physical activity declines during adolescence. However, there is currently little known about factors that may influence physical activity during the adolescent years. In a 4-yr longitudinal study of 1245 adolescents, physical activity (PA) was measured by questionnaire. The purpose of this report was to identify factors in early adolescence (grades 7-9) that predict PA in late adolescence (grades 10-12). A total of 775 subjects had complete PA data for both yrs 1 and 4 of the study. Factors examined in this analysis included demographics, baseline measures of PA, physical fitness measures, body mass index (BMI), and risk taking behaviors. In males, univariate analyses indicted that race, socioeconomic status (SES), days participated in vigorous activity in past two weeks (vigorous PA), being on an athletic team, cigarette use, baseline PA (hrs/wk), BMI, time to complete a 1-mile run/walk, and flexibility were significant predictors of PA in late adolescence. In females, univariate analyses indicated age, race, SES, vigorous PA, being on an athletic team, TV viewing, marijuana use, baseline PA (hrs/wk), time to complete a 1-mile run/walk, and flexibility were significant predictors of PA in late adolescence. The stepwise linear regression model for males included baseline PA (hrs/wk), vigorous PA, being on an athletic team, and flexibility as positive predictors and mile run time, cigarette use and low SES as negative predictors (R2=.144). The stepwise linear regression model for females included baseline PA (hrs/wk), being on an athletic team, flexibility, race (white), and high SES as positive predictors and mile run time, age, and marijuana use as negative predictors (R2=.241). These results indicate that measures of PA and physical fitness in early adolescence are independent significant predictors of PA in late adolescence.