Descriptive characteristics of the participants.

In Japan, physical activity duration in junior high schools is substantially higher than that in elementary schools. Using longitudinal data, this study examined the association between changes in physical activity duration and depressive symptoms in 1225 adolescents without depressive symptoms at baseline (51.2% female) from a rural city in Japan. Adolescents in the group that changed from "Active" (physical activity duration ≥7 h /week) in the fifth-grade of elementary school (age 11 years) to "Inactive" in the second-grade of junior high school (age 14 years) had a higher odds ratio of being rated as having depressive tendencies at age 14 years than adolescents in the Active–Active group. Additionally, the results of the cross-lagged effects model analysis suggested that depressive symptoms may be a factor in the reduced physical activity duration.