[Analysis on risk factors of screen time among Chinese primary and middle school students in 12 provinces].

OBJECTIVE

To investigate the characteristics of screen time and its risk factors in Chinese primary and middle school students.

METHODS

During April 2012 and June 2012, according to the geographical distribution, the stratified random cluster sampling method was used to select 4 provinces from eastern, central and western China, respectively. The convenience sampling method was used to select 2 primary and middle schools from urban, 2 primary and middle schools from rural in each province. In each school, all grades were included, and 2 classes were selected in each grade. A total of 51 866 students or parents were selected as study participants, and 43 771 questionnaires were valid. Information on demographics, academic performance, screen time (TV, computer and cellphone) at weekdays and weekends and the prevalence of the high screen time were compared, multivariate logistic regression was used to analyze the association between screen time >2 h/d and potential influential factors.

RESULTS

The percentage of students with screen time >2 h/d at weekdays and weekends were 16.2% (7 082/43 771) and 41.5% (18 141/43 771) (χ(2)=6 280.14, P<0.001), respectively. The distribution of P50 (P25-P75) for screen time at weekdays and weekends were 0.9(0.4-1.6) and 1.8(1.0-3.0) (Z=-131.26, P<0.001), respectively. The results of multinomial logistic regression analysis showed that, at weekdays, subjects characterized as primary school students, boys, urban area, living in western area and sufficient vigorous physical activity ≤2 d/w had higher risk for screen time >2 h/d than those characterized as elementary school students, girls, rural area, living in eastern area and sufficient vigorous physical activity >3 d/w, odds ratio were 2.01, 1.54, 1.21, 1.09, and 1.07, respectively (P<0.05 for all); subjects characterized as a normal or worse self rating academic performance had higher risk for screen time >2 h/d than those characterized as a good self rating academic performance, odds ratioes were 1.24 and 1.73, respectively (P<0.05 for all); subjects characterized as paternal education level as elementary school, middle school, high school or secondary school had higher risk for screen time >2 h/d than those characterized as paternal education level as college school or high, odds ratioes were 1.41, 1.47 and 1.52, respectively (P<0.05 for all); subjects characterized as maternal education level as elementary school, middle school and high school or secondary school had higher risk for screen time >2 h/d than those characterized as maternal education level as college, odds ratioes were 1.40, 1.52 and 1.47, respectively (P<0.05 for all). At weekends, subjects characterized as primary school students, boys, urban area and sufficient vigorous physical activity ≤2 d/w had higher risk for screen time >2 h/d than those characterized as elementary school students, girls, rural area and sufficient vigorous physical activity >3 d/w, odds ratioes were 2.11, 1.51, 1.20 and 1.05, respectively (P<0.05 for all). At weekends, subjects characterized as a normal or worse self rating academic performance had higher risk for screen time >2 h/d than those characterized as a good self rating academic performance, odds ratioes were 1.09 and 1.26, respectively (P<0.05 for all); subjects characterized as paternal education level as elementary school, middle school, high school or secondary school had higher risk for screen time >2 h/d than those characterized as paternal education level as college school or high, odds ratioes were 1.29, 1.30 and 1.19, respectively (P<0.05 for all); subjects characterized as maternal education level as elementary school, middle school had higher risk for screen time >2 h/d than those characterized as maternal education level as college school or high, odds ratioes were 1.19 and 1.16 and, respectively (P<0.05 for all).

CONCLUSION

The prevalence of screen time >2 h/d is high; screen time at weekdays is longer than weekends, and there are significant differences among different sexes, urban or rural areas, living areas, self rating academic performance, parents education levels and physical activity groups.