Background
Screen-based sedentary behavior (SSB), including many new forms of digital media, is ubiquitous and is an important part of daily activity in the current information era. A longer time spent in SSB was significantly associated with lower fitness, overweight or obesity, depression, inattention, shortened sleep duration, and cardiovascular disease risk factors in children or adolescents [
1‐
6], independent of meeting or not meeting the recommended level of physical activity [
7,
8]. Moreover, the habits of SSB developed in childhood more easily persist into adulthood compared with the habits of moderate or vigorous physical exercise [
9]. The “Health Behavior in School-aged Children” project (World Health Organization) indicated that students aged 11–15 years spent more than 2 h per day viewing television (an indicator of SSB) in 30 countries, and screen time (ST) has been growing along with further popularization of new screen devices (e.g., personal computer, tablet, smart phone, etc.) [
10]. A study conducted in a city in Eastern China also showed that 51.5% of adolescents spend more than 2 h on ST per day [
11].
Limited previous studies have indicated that increased ST was related to male sex, being older, weekends, low parental education, and high media accessibility (such as having a television [TV] or a computer in the bedroom) [
12‐
15]. Other studies on environmental correlates of SSB, however, have not been very clear, and some of the results were inconsistent (such as the presence of a television set in the bedroom) [
16,
17]. Meanwhile, environmental correlates, such as media accessibility and parental social factors (rules or viewing accompanied by parents), play a key role in evidence-based interventions on restricting prolonged ST in school-aged children [
16]. To our knowledge, few studies have explored and confirmed the modifiable correlates of prolonged ST, especially for new screen devices, such as cell phones and pads/tablets, in children in developing countries [
18].
The aim of the present study was therefore to examine the correlates, including individual and environmental factors, of ST based on a validated self-reported SSB questionnaire in 8–19-year old Chinese girls and boys.
Methods
Participants
A total of 1164 students aged 8–19 years from five elementary, junior high, or senior high schools participated in the project from September 2015 to May 2016 in Zhejiang Province, China. We analyzed 1063 students after excluding subjects with missing variables for age or sex (n = 26), no informed consent (n = 26), or missing variables of other correlates (n = 49). All students and their parents provided signed informed consent forms, and the study protocols were approved by the institutional review board of Zhejiang Financial College.
Questionnaire of SSB
The self-administered SSB questionnaire was designed based on the Adolescent Sedentary Activity Questionnaire [
19] and included four items: TV or video viewing, playing for recreation or leisure on a computer, computer use for homework, and playing for recreation or leisure on a cell phone or tablet. ST, representing SSB, was defined using the same definition for sedentary screen time [
20], not including physically active behavior. One question, “Do you participate in this activity?”, was changed from an answer of “daily” to include two columns: “Monday to Friday” and “Saturday to Sunday”. The intra-class correlation coefficient of the revised SSB questionnaire (Chinese version) was > 0.8 (boys: 0.81, girls: 0.85), which is considered excellent according to the recommendations of Landis et al. [
21], based on an analysis of test-retest reliability.
Correlates measurement
Environmental factors were divided into three questions of media accessibility (“Do you have a personal computer?”, “Do you have a TV in your bedroom?”, and “Do you have a cell phone/tablet?”) and two questions of parent/others social factors (“Do you have a parent or others present when you play on the computer?” and “Do you have a parent or others present when you are watching TV?”). The options of media accessibility questions were binary including “Yes” (encoded as 1) and “No” (encoded as 0). A four-point Likert scale, including “Strongly disagree” (encoded as 1), “Disagree” (encoded as 2), “Agree” (encoded as 3), and “Strongly agree” (encoded as 4), was used to measure the effects of parents or other persons. Individual information, such as sex, age, grade, and parental education level (in part), was also assessed.
Statistical analysis
Correlates and ST are described and sex differences of variables were evaluated using the chi-squared test for categorical variables or T-test for continuous variables. Three questions of media accessibility were added as categorical variables (no screen, one screen, two screens, and three screens were 0, 1, 2, and 3, respectively) or binary variables (the values of low and high were 0–1 and 2–3, respectively) and two questions of the presence of parents/others were also added as categorical variables (strongly disagree, disagree, agree, and strongly agree were 2–3, 4, 5–6, and 7–8, respectively) or binary variables (low and high were 2–4 and 5–8, respectively). Prolonged ST was defined as ≥2 h per day, and prolonged weekend ST was defined as ≥5 h/day [
22‐
24]. Leisure ST was defined as ST excluding time spent using the computer for school-related studying. A mixed effects model was applied to explore correlates of and interactions with ST, which included fixed effects (age, grade, media accessibility, and presence of parents/others) and random effects (schools and classes). Because of the high requirement for normality in mixed effects models, zero-mean normalization was used to transform continuous variables of age and ST previously for their non-normality. Odds ratios (OR) and 95% confidence intervals (CI) were used to describe the effects of environmental factors on ST (binary variables) based on the logistic regression model. All data were evaluated using Epidata 3.0 (double entry) and IBM SPSS 20.0 (statistical analysis); the significance level was
p < 0.05.
Discussion
Based on data collected using a validated questionnaire, male sex, junior high school students, weekends, presence of parents/others, and media accessibility were significantly associated with prolonged ST. These results are important for developing a series of effective strategies against prolonged ST in 8–19-year-old students in the future.
The percentage of prolonged weekday and weekend leisure ST (≥2 h/d) were 11.7% and 83.1%, respectively; this was lower for weekdays and greater for weekends compared with the results of another study in Chinese children [
25]. The findings that boys have more leisure ST than girls do and that there is more ST on weekends than on weekdays are also consistent with previous studies [
10,
25]. Another study indicated that age was positively associated with ST [
10]. In our participants, age and grade were nonlinearly associated with ST (leisure ST decreased after 15 years of age/junior high school grade). This might be because there is more pressure in regards to studying and preparing for college entrance examinations in Chinese senior high school students than is seen for senior high school students in Western culture; this may push students to place more attention and time on study-related behavior, resulting in a “crowding-out effect” [
26,
27].
In addition, correlations between environmental factors and leisure ST were not significantly changed after additional adjustment for parental education levels. In contrast with previous studies [
28], low parental education was not a risk of prolonged ST in students in this study. In China, this might because parents with higher education level (high school or above) do not restrict their children’s SSB as well as because they might have greater media accessibility owing to a higher socioeconomic status [
29].
Previous studies have shown that parents viewing TV with their children or parental role modeling of TV viewing were positively associated with the duration of TV viewing [
30‐
32]. In our results, however, the presence of parents/others was negatively associated with leisure ST. Parents of students in the present study might play a supervisory role while watching TV with their children, rather than merely being present or accompanying their children. The results also potentially indicate that parental rules concerning screen viewing or parental expectations could mediate SSB (especially for TV viewing) in this population [
26,
33]. Accordingly, media accessibility, especially having a TV set in the bedroom, seems a more important target than the accompaniment of parents/others for preventing prolonged leisure ST. These results differ from those of Carson et al. [
34], who indicated that social environments might be more important than physical environmental factors, such as neighborhoods. Upon further exploration of the data, the presence of a TV set in the bedroom was not associated with leisure ST in our participants. Students with a TV in the bedroom did not have a significantly greater amount of time spent watching TV, but did have a greater amount of time using a phone/tablet and computer. Therefore, this does not appear to be a causal relationship, but rather just a relationship observed between the presence of a TV set in the bedroom and leisure ST. Most students with a TV in the bedroom were older and were more likely to be alone (not accompanied by parents or others) while watching TV. It was deduced that these students might have a greater amount of individual space, independence, and insensitivity to parental control, which could increase leisure ST.
Furthermore, our results indicate that cell phone/tablet-based leisure ST has become a main component of ST, although Jiang et al. reported that playing on a mobile phone was less prevalent before 2014 in Chinese adolescents [
25]. In our data, 91.0% students with one type of screen and 96.7% students with two types of screens were the owners of a cell phone/tablet. This observation shows that cell phone/tablet-based leisure ST has dramatically increased in recent years and now comprises the largest proportion of ST, as was observed with Australian children [
35]. These results also suggest a call for future studies that focus on exploring the correlates of small/mobile leisure ST because of its potentially different correlates.
This study has some limitations. First, caution should be noted in generalizing our results or inferring any causal relationships, because this was a cross-sectional study performed in the north of Zhejiang Province (a region with a relative high economic status in China), even though the students were from different grades and schools. Second, the presence of parents/others has not been defined clearly; for example, it is unclear whether there were efforts to restrict or encourage students’ SSB; moreover, the study lacked a direct investigation of parents. Previous studies indicated that parental ST, role modeling, and familial rules for SSB were related to leisure ST of students or children [
36,
37]. The strength of our study was measuring time spent using cell phones/tablets, which was highly prevalent in Chinese students. Previous studies in this field mainly focused on descriptions of sociodemographic characteristics or correlates of TV viewing and computer use.
Conclusions
Greater media accessibility and less presence of parents/others are associated with prolonged leisure ST in 8–19-year-old Chinese students, based on data collected using our reliable questionnaire. Additional attention should be paid to mobile/small screen-based devices, such as cell phones and pads/tablets, to avoid prolonged leisure ST, especially in boys and/or on weekends. Development of effective strategies for intervention of small/mobile device-based ST will be required in future studies.
Acknowledgements
We gratefully acknowledge the teachers, students, and parents for their participation in our study and the ongoing support of the schools in Hangzhou and in Huzhou, Zhejiang Province, China.