Physical activity in adolescents: examining influence of the best friend dyad.

PURPOSE The aim of this study was to determine the influence of dyadic friendship relationships on adolescent physical activity (PA) behavior. Unique to this study was examination of the "best friend" dyad. METHODS Participants were 268 adolescents (13-18 years of age) of both sexes (boys n = 140, girls n = 128), constituting 134 best friendship dyads; 84 were reciprocal best friends and 50 were not. PA was evaluated with the short version of the International Physical Activity Questionnaire. The Actor-Partner Interdependence Model was tested using a hierarchical linear model. First, a null model was run to calculate intraclass correlation coefficient for each type of PA. Next, sex, age, and reciprocity in friendship were included in the model as PA predictors. RESULTS The null models revealed significant intraclass correlation coefficient values for vigorous PA (VPA) (.32) moderate PA (MPA) (.31), and sitting behavior (.21), but not for walking. Age was the only predictor for VPA, whereas for MPA the predictors were sex and age of the partner. Reciprocity in friendship (best friend) was not a significant predictor for PA. CONCLUSIONS These results indicate a high degree of similarity between best friend dyad in PA, except for walking. The age and sex of the best friend (partner) were important predictors of MPA. The reciprocity in friendship (being reciprocal best friends) was not a relevant variable.