Introduction

The growing global pandemic of childhood obesity, with its large variations among countries [1], has been estimated to include more than 155 million school-aged children (10% of the world’s 5- to 17-year-olds) who either are overweight or obese [2]. The rise in the prevalence of childhood obesity is likely due to the interactive effects of growing caloric consumption, increased sedentary lifestyle (e.g., watching television, playing videogames, and studying), and decreased physical activity, all precipitated by increasingly obesogenic environments [3–6].

Sedentary behaviors have been defined as a range of endeavors with an energy expenditure ≤1.5 times the resting energy expenditure [7]. Types of sedentary behavior include television viewing, reading, working at a computer, or talking with friends on the phone while sitting [8]. Evidence from prospective observational studies suggests that decreased sedentary behavior is protective against relative weight and fatness gains over childhood and adolescence [9]. Recommendations have been established that suggest children should not spend more than 2 h per day on electronic media for entertainment [10, 11]. Increasing trends in overweight and obesity are often ascribed to decreased physical activity and increased sedentary behaviors [12, 13]. From a population perspective, although the obesity pandemic has risen at an alarming rate globally, especially in developed countries, studies suggest that the proportion of US students grades 9–12 who watched television three or more hours per day actually decreased from 1991 to 2007 [14], and one study found no clear pattern of increasing trends in television watching in some European countries [15].

Overweight (hereafter, including overweight and obesity) is becoming a significant public health problem in China and has been increasing among Chinese youth during recent years [16–19]. Given that about 90% of Chinese households now own televisions [20], the potential for television watching to contribute to the increasing prevalence of overweight among Chinese children and adolescents (in the same way it has in other countries) needs to be assessed. Another noteworthy modern phenomenon is that Chinese youth are under great stress to perform well in school and that heavy homework loads are typical [21]. Parents and society exert tremendous pressure on children and adolescents to succeed in school so they can be admitted to a university that might enable them to pursue an intellectual career rather than one of manual labor. Thus, Chinese students tend to spend a great amount of time studying before and after school [20]. Shi et al. [22] found that adolescents in the junior high schools of Jiangsu Province aged 12–14 spent a total of 10.8 h every day studying in school and at home.

Previous studies suggest a positive association between television viewing and BMI among Chinese adolescents [23, 24]. The purpose of this study was to examine relationships between sedentary behavior and BMI and trends in the level of sedentary behavior among Chinese school children and adolescents residing in seven diverse provinces. The results of this study will help better understand the increasing prevalence of overweight among Chinese youth.

Methods

Survey Design

Designed to monitor the health and nutritional status of all ages of the Chinese population, the China Health and Nutrition Surveys (CHNS) primarily examined the effects of national, provincial, and local government health and family planning policies and programs [24, 25]. A multistage random cluster sampling strategy was used to select households in each of the chosen provinces. The survey design has been described in greater detail elsewhere [19, 26, 27]. The School of Public Health of the University of North Carolina at Chapel Hill and the Chinese Centers for Disease Control and Prevention reviewed and approved the procedures for data collection. The current study was approved (protocol no. 07-12141) by the Indiana University Human Subjects Committee.

Sample

Overall, 3,840 households were randomly selected from eight provinces during each survey year. CHNS measurement specialists surveyed the members of each chosen household. For this study, CHNS data from the seven provinces—Jiangsu, Shandong, Henan, Hubei, Hunan, Guangxi, and Guizhou—that consistently participated in all survey years (1997, 2000, and 2004) were selected for analysis (N = 15,119). Participants younger than 6 years, older than 18 years, or who were not in school were excluded from the analysis. The total size of the multiple cross-sectional samples meeting criteria for this study was 5,497. The mean age of the sample was 12.3 years (SD = 3.2), and 47% of the sample was female.

Measurement

A wide range of information about health and nutrition was collected by the CHNS, which included data about each person’s dietary intake, body composition, blood pressure, health history, and health-related behaviors [28]. Height and weight were measured directly by trained health workers who followed standard protocols similar to the National Health and Nutrition Examination Survey protocol developed by the National Center for Health Statistics in the USA [20]. Each study participant wore lightweight, indoor clothing while her or his weight was measured to the nearest tenth of a kilogram with a balance beam scale. Height was measured without shoes to the nearest tenth of a centimeter using a portable stadiometer. BMI was calculated as weight in kilograms divided by square of height in meters. The international age- and gender-specific BMI reference standard proposed by the International Obesity Task Force was used to define overweight, including obesity, in this study [29].

Dietary data were collected at both the household and individual levels [27, 30]. Three-day 24-h recall was administered by asking each individual, except children under 12, to report all food consumed at home and away from home on a daily basis. For children under 12, his or her mother or the person who prepared the food was asked to recount each child’s food consumption. The 1991 Chinese Food Composition Table was used to calculate nutrient values for the dietary data [27]. Given the difference in the recommended dietary intake for children and adolescents by age and gender [31], the percentage of energy intake relative to the Chinese dietary recommended intake was calculated. The percentage of energy intake from fat, carbohydrate, and protein was also calculated.

In addition, since 1997, questionnaires have been used to collect data about self-reported physical activity and sedentary behavior from children and adolescents. Field workers interviewed children and adolescents about the amount of time they spent during the previous week on listed physical activities (i.e., gymnastics, track and field, swimming, badminton, tennis, soccer, basketball, ping pong, board games, boxing, martial arts, or others). For the 6- to 11-year-olds, commuting on foot to and from school counted as part of a physical activity [20]. It is not typical for Chinese children under 12 to ride a bicycle by themselves. Thus, those under 12 who reported commuting to school by bicycle likely rode on a bicycle pedaled by one of their parents and, therefore, were not given active commuting credit [20]. For the 12- to 18-year-olds, commuting to and from school on foot and by bicycle counted as physically active time. Weekly duration of physical activity was calculated by totaling weekly time spent on physical activities in school, physical activities outside school, and active commuting to and from school.

Respondents were also asked about the amount of weekly time spent on the following sedentary behavior: watching TV/videos; reading; writing; drawing; building blocks; listening to the radio; and playing with sedentary games, toy vehicles, puppets, and board games. Weekly duration of sedentary behavior was calculated by totaling weekly time spent on all sedentary behavior above.

Data Analysis

All statistical analyses were conducted using Stata 10.0. Because the prevalence of overweight was higher in the younger group (6- to 11-year-olds) than the older group (12- to 18-year-olds, p < 0.05), analyses were conducted separately for these two age groups.

The median and interquartile range (percentile 25 to percentile 75) of time reported engaging in total sedentary behavior per day (hours), television/videos watching per day (hours), and studying before and after school per day (hours) were calculated by gender, age group, and survey year. Differences in time reported engaging in total sedentary behavior per day, television/videos watching per day, and studying before and after school per day between the younger and older age groups were tested using nonparametric (two-sided Wilcoxon rank sum) tests (α = 0.05). Differences in time reported engaging in total sedentary behavior between 1997 and 2004 were also tested (α = 0.05).

The proportion of those who watched television/videos more than 2 h/day and those who studied before and after school more than 2 h/day were calculated by gender, age group, and survey year. Then, the average annual change in each proportion was computed by dividing the proportion difference between two different time points (2004 and 1997) by the number of years between the two time points (2004–1997 = 7) [32]. Average relative change in each proportion was calculated by dividing the average annual change by the baseline proportion in 1997 [32]. Because data on physical activity and sedentary behavior were only collected since 1997, the analysis for this paper focused on the three time points, 1997, 2000, and 2004. Sequential multiple regression analyses were run in the following sequence: each individual’s biological factors (age and sex), the environmental factor (region), family income factor (household income), behavioral dietary factors (total energy intake and percentage of energy from fat), behavioral physical activity factor (physical activity), and behavioral sedentary behavior factors (watching TV/videos and studying before and after school). Regression coefficients for predictors in the final model of sequential multiple regressions are reported in the results. Analyses have been stratified by gender.

Results

In Table 1, we present the selected descriptive characteristics of the population sample by gender, age group, and survey year. In all the survey years (1997, 2000, and 2004), the mean age was similar in males and females between the two age groups. In all the survey years, compared with females, males reported higher levels of physical activity and higher caloric intake, whereas both genders reported a similar level of time spent on sedentary behaviors. In all the survey years, males and females reported a similar level of time spent watching TV, while females reported more time spent studying before and after school than males in the recent survey years of 2000 and 2004.

Table 1 Descriptive results of variables by gender, survey year, and age group
Full size table

Table 2 illustrates trends in the median and interquartile range (percentile 25 to percentile 75) of the time spent engaging in total sedentary behavior per day by gender, age group, and survey year. From 1997 to 2004, the time spent engaging in total sedentary behavior per day significantly increased among both 6- to 11-year-olds and 12- to 18-year-olds (p < 0.05). The time spent engaging in watching TV/videos per day increased significantly in both age groups from 1997 to 2004 (p < 0.05). Among both age groups, the time spent engaging in studying before and after school also increased among both 6- to 11-year-olds and 12- to 18-year-olds (p < 0.05).

Table 2 Trends in the median (interquartile range) time reported engaging in sedentary behavior (hours/day) among Chinese school children and adolescents: 1997–2004
Full size table

Table 3 shows trends in the proportion of the combined 6- to 18-year-olds who reported watching TV/videos more than 2 h/day, which increased from 5.4% in 1997 to 23.1% in 2004, a fourfold increase with an average annual increase of 2.5% and an average relative increase of 46.8%. Overall, the proportion of the combined 6- to 18-year-olds who reported studying before and after school more than 2 h/day increased from 8.7% in 1997 to 38.7% in 2004, more than a fourfold increase. The proportion of studying before and after school more than 2 h/day experienced a notably higher average annual increase (4.3% vs. 2.5%) and a notably higher average relative increase (49.3% vs. 46.8%) than the proportion of watching TV/videos more than 2 h/day.

Table 3 Trends in the proportion of weight-related behavioral factors among Chinese school children and adolescents: 1997–2004
Full size table

Table 4 exhibits the regression analysis of BMI on potential explanatory variables among Chinese school children and adolescents by age group and gender. Watching TV/videos was positively associated with BMI in 12- to 18-year-old males, whereas such a significant association was not found in 6- to 18-year-old females. A significant association was found between BMI and time spent studying before and after school only in 6- to 11-year-old males.

Table 4 Results from multiple linear regression showing the standard coefficients b and associated p values that evaluated the association between BMI and potential explanatory variables
Full size table

Discussion

Our findings indicate a significantly rising trend in sedentary behavior, including watching TV/videos and studying before and after school, among Chinese school children and adolescents in seven Chinese provinces. This study found that in 2004, 24.9% of 6- to 11-year-olds and 21.5% of 12- to 18-year-olds did not meet the recommendation of limiting watching television to ≤2 h/day. Studying before and after school and watching TV/videos each accounted for about half the increase in sedentary behavior. Both watching TV/videos and studying before and after school were significantly and positively associated with BMI in males, but not in females

Chinese school children and adolescents were found to be twice as sedentary in 2004 as their counterparts in 1997. During the years assessed, the data confirm that Chinese youth did not watch television/videos to the extent observed in other countries despite the relatively high levels of household television ownership [20, 24]. However, we found increasing trends in watching television/videos between 1997 and 2004. While 94.5% of Chinese school children and adolescents met the recommended hours of watching television in 1997, only 76.9% did so in 2004, which is a 17.6% decrease in a short 7-year period. Alarmingly, the proportion of school children and adolescents who did not meet the recommendation increased with an average annual change rate of 2.5%. The worrisome trend in watching television may be partly attributed to the introduction of American-style satellite television. In 1997, almost no Chinese households had access to the contemporary television programs that might encourage children’s television viewing [20, 33]. American-style satellite television (e.g., 30 or more channels with programs that include game shows, reality-based shows, prime-time dramas, and product advertisements) has recently been introduced on a large scale in China [20, 34]. Additional research is needed to monitor whether television watching by Chinese youth will continue to increase, and effective interventions should be implemented to address television watching by youth.

Consistent with a previous study [33], our findings suggest that Chinese youth spend almost as much time studying before and after school as they do watching TV/videos. Our findings indicate that over the years, Chinese school children and adolescents were spending more time studying before and after school with an average annual increase of 4.3% and an average relative increase of 49.3%. After-school hours, often referred to as “critical hours,” is a time when young people engage in a considerable proportion of their daily leisure time activity and is predictive of the amount and pattern of their daily leisure time activity. Previous research indicates that the after-school hours should be targeted for physical activity promotion in young people [35]. Hence, although we should not discourage studying, the amount of time outside school that Chinese youth study in addition to TV viewing time should be considered when developing programs and policies to promote physical activity in China. In fact, there is an emerging body of evidence which suggests that physical activity does not appear to negatively affect academic performance and that evidence also indicates associations between physical activity and reduced incidence of depression, anxiety, and fatigue [36, 37]. This should be used to urge governments and parents to provide more physical activity opportunities for the youth.

This study suggests that time spent watching TV/videos and studying before and after school are associated with BMI in males, but the association was not significant in females. This finding is in line with another study of 6,848 Chinese high school students aged 12–18 years [23] which observed that the positive relationship between TV time and BMI was significant among boys, but not among girls. However, studies conducted in other countries reported that such an association was stronger in girls than boys [38, 39]. This may be attributed to the different numbers of hours spent watching TV/videos between different countries. For example, the mean TV/video watching hours in the CYKIDS Study [38] were 2.42 h for boys and 2.32 h for girls per day. Over two thirds of boys and girls in the CYKIDS Study spent more than 2 h per day watching TV/videos, whereas less than one fourth of our Chinese sample did. Given the lower level of physical activity reported by females, many hours of TV/video watching among females in the CYKIDS study might have contributed to an aggravated energy balance in females, yielding a strong association between BMI and TV/video watching hours among females.

As shown in Table 3, it is notable that there is no significant difference in the mean time spent on sedentary behavior, including watching TV/videos and studying before and after school, between males and females. Thus, a possible explanation of the observation of the positive relationship of sedentary behavior with BMI only among males might have to do with the fact that boys tend to have a higher caloric intake than females (Table 3). Presumably, studying before and after school in combination with increased hours spent watching TV/videos may displace the amount of time available for physical activity, thus resulting in relatively lower levels of energy expenditure. Indeed, a formative study found that Chinese middle school students perceived not having enough time and having too many assignments as barriers to daily participation in physical activity before and after school [40]. In addition, typical Chinese students have to spend a large amount of compulsory study time at home on school evenings and on weekends. Thus, if they spend a considerable time watching TV/videos, they are likely to have to reduce the time for their physical activity [23]. As suggested by Owen et al. [7] and Wijndaele et al. [41], sedentary behavior may have independent effects on health outcomes. Perhaps, participating in sedentary behavior may lower the basal metabolic rate independently [42]. Additionally, watching TV may result in an increased consumption of high-fat snacks and sugar-sweetened beverages, higher exposure to food advertisements, and less frequent consumption of fruits and vegetables [43]. Previous research suggests that Chinese children and adolescents seem to be vulnerable to TV commercials, often resulting in actual consumption of snacks seen on TV [44].

The above reported results of the positive association between watching TV/videos, studying before and after school, and BMI suggest that the substantial increase in watching TV/videos in addition to the simultaneous considerable increase in studying may contribute to increasing BMI. Various studies [45, 46] have also recommended that an obesity prevention intervention emphasizing decreasing sedentary behaviors may be more effective than an intervention aiming only to increase regular physical activity. To contain the increasing epidemic of overweight in China, interventions should focus on decreasing excessive time watching television/videos and studying in addition to increasing physical activity.

A limitation of this study is that the CHNS data were not nationally representative. Although the CHNS was not designed to provide exact nationally representative data, findings about the patterns and trends in diet and body composition were almost identical with nationally representative data collected by the National State Bureau of Statistics [19, 47–49]. In addition, comparisons between different years of cross-sectional samples might have confounded the findings as different survey samples might not equally represent the populations. The physical activity and sedentary behavior were self-reported data, and this may have also confounded the results.

Chinese school children and adolescents have become significantly more sedentary over recent years. The combination of a rise in watching television and growth in studying before and after school contributes to the greater sedentary behavior among Chinese youth, which may play a substantial role in the increasing prevalence of overweight among them. With an increase of 1.3 kg/m2 or 8% in BMI between 1997 and 2004, sedentary behavior has nearly doubled in that short period of time. In the absence of policies and programs addressing the two examined barriers to physical activity, we expect a continued increase in overweight Chinese youth who will become overweight adults with increasing medical problems as a result of their weight. Treating these increasing health problems among such a large population will exact an unprecedented economic toll upon the people of China.