Background: We examined the covarying patterns of physical activity and sedentary behaviors among adolescents and their long-term maintenance.
Methods: Data came from the National Longitudinal Study of Adolescent Health (1995-2002). We used latent class analysis to identify distinct covarying patterns in adolescence. Logistic regression models were used to predict odds of meeting moderate-to-vigorous physical activity (MVPA) recommendations (> or = 5 bouts/week) and exceeding screen time guidelines (> 2 hours/day) 6 years later based on their adolescent class profile.
Results: Five classes for each gender were identified and labeled as low physical activity (PA)/low sedentary behaviors (SED), moderate (Mod) PA/high (HI) SED, Mod PA/low SED, HI PA/low SED, and HI PA (except skating/biking)/low SED. Compared with low PA/low SED, males and females in Mod PA/low SED, HI PA/low SED, and HI PA (except skating/biking)/low SED classes had increased odds of meeting MVPA recommendations in young adulthood. Mod PA/HI SED had higher odds of exceeding screen time guidelines in young adulthood (adjusted odds ratio [AOR] for females: 1.67, 95% CI: 1.00-2.81; AOR for males: 3.31, 95% CI: 1.80-6.09).
Conclusions: Findings are useful to aid the development of multifactorial interventions that promote physical activity and reduce screen time among adolescents transitioning to adulthood.