Familial and Parental Predictors of Physical Activity in Late Adolescence: Prospective Analysis over a Two-Year Period.

Children's health behaviors are highly influenced by their parents and family. This study aimed to prospectively evaluate the parental/familial factors associated with physical activity levels (PALs) among older adolescents. The participants were 766 adolescents, who were prospectively observed at baseline (when they were 16 years of age), at first follow-up measurement (FU1; 17 years of age), and second follow-up measurement (FU2; 18 years of age). Sociodemographic factors (age, gender, socioeconomic status, and sport participation) and parental/familial variables were evaluated at baseline. PALs (evidenced by the Physical-Activity Questionnaire-for-Adolescents) were prospectively evidenced at baseline, FU1, and FU2. Factorial analysis of variance for repeated measurements showed a significant decrease in PALs during the study course (F = 83.05, p < 0.001). Sport participation and male gender were significant predictors of PALs at baseline, FU1, and FU2. Logistic regression, controlled for sport participation and male gender, evidenced paternal education as a significant predictor of baseline PALs. Parental conflict was a significant predictor of PALs in all three testing waves. The significant influence of paternal education on the children's PALs existed from younger adolescence until the age of 17 years. The association between parental conflict and PALs developed in older adolescence. These results should be used in the development of specific and targeted interventions aimed at the improvement of PALs and a reduction of sedentarism in youth.