Abstract
BACKGROUND: Few studies have shown that relatively younger children and adolescents (those born later in the same school year) were less likely to engage in physical activity in a phenomenon termed the relative age effect. Although these studies mainly targeted elementary and middle school students, no study has reported on the relative age effect on physical activity in ordinary high school students. Moreover, the relative age effect on sedentary behavior might show an opposite association with physical activity. Therefore, we aimed to clarify the relative age effects on physical activity and sedentary behavior across different school stages in Japanese children and adolescents. METHODS: A cross-sectional questionnaire survey was conducted with 21,491 children and adolescents (elementary, middle, and high school students aged 10-18 years) in various Japanese regions from January 2018 to July 2019. Overall, 18,281 children and adolescents (10,299 boys and 7,982 girls) were finally included in our analysis. Data on the birth month, frequency and duration of their physical activity (vigorous, moderate, and moderate-to-vigorous intensity), and sedentary behavior (weekday and weekend total sedentary time, television viewing, playing video games, and Internet use) were obtained. We utilized two-part model regression analyses. Statistically significant association with birth month indicated that a relative age effect was observed. RESULTS: The relatively younger individuals were less likely to engage in physical activity (especially vigorous physical activity (VPA)); this association was observed in middle and high school students. Conversely, the relatively younger boys spent more time during weekends with sedentary behaviors and Internet use. Based on school stage, the relative age effect on VPA time was observed in middle school boys (average margin effect (AME) = -0.74, p < 0.01) and high school students (boys: AME = -0.69, p < 0.01; girls: AME = -0.53, p < 0.05). The relative age effect on sedentary behavior was observed during weekends with sedentary behaviors (AME = 3.55, p < 0.01), playing video games (AME = 0.72, p < 0.05) and Internet use (AME = 2.46, p < 0.01) for male high school students and on television viewing (AME = 1.12, p < 0.05) for female middle school students. CONCLUSION: The relative age effect on physical activity can persist after middle school, and the relatively younger individuals are more likely to replace physical activity with sedentary behavior.