Abstract
Background:
In recent decades, China has experienced an exponential growth in the number of internet users, especially among the youngest population, as well as a rapid proliferation of Western-type fast food restaurants. The health consequences of internet availability and fast food consumption among youth have been largely studied in Western countries, but few studies have focused on China.
Objectives:
This paper has two goals. The first is to evaluate the differences in new media exposure and preferences for fast foods between rural and urban areas. The second goal is to test the association between new media exposure and fast food consumption. The targets of this analysis are Chinese children and adolescents aged 6–18 attending school at the time of the interview.
Methods:
Research hypotheses were tested using mean-groups comparisons for differences between rural urban sub-samples, and logistic regressions with odds ratios to estimate the relationship between media exposure and preferences towards fast foods. Cross-sectional data from the 2009 China Health and Nutrition Survey were employed.
Results:
Watching online videos and playing computer games are behaviors associated with higher probabilities of eating at fast food restaurants in both rural and urban young residents, with higher odds in rural areas. Surfing the internet is associated with higher odds of being overweight in both rural and urban settings. Results also show that children living in rural areas spend significantly more time playing computer games, watching TV and videotapes, but less time doing homework than their urban peers.
Conclusions:
This paper suggests that monitoring the nutritional effects of new media exposure in China is of key importance in order to develop adequate health promotion policies, in both rural and urban areas.
Introduction
In the last 20 years, China has undergone massive economic and societal changes. The rapid increase of the middle income class has contributed to the introduction of new forms of consumption habits already popular in many Western countries, such as frequently eating at fast food restaurants, excessive amounts of time spent playing computer games, and surfing on the Internet. These habits are associated with higher rates of being overweight and obesity, which can lead to serious health concerns such as cardiovascular diseases or diabetes later in life. In particular, eating at a fast food restaurant increases the intake of high-caloric foods, while computer-related activities reduce the amount of time an individual spends doing physical activities.
These fast-paced changes pose new challenges for the Chinese public health system. Lee (1) recognized, for example, that the increase of non-communicable and mental diseases in China, such as obesity and other cardiovascular diseases, requires particular attention. The exponential increase of internet users, especially among the youngest segments of the population, opens the possibility to the investigation of research questions about the effects of new media on health-related behaviors.
In general, a prolonged exposure to new media can affect both the physical and mental health of individuals. In terms of physical health, children and adolescents who spend a lot of time in front of the computer are more likely to reduce the amount of time spent playing outdoor and exercising, thus increasing the likelihood of becoming overweight and obese. Video-game users have also a greater food intake than non-game users due to a sedentary physiological effect. In an experimental setting, Chaput et al. provided evidence that video-game players tended to increase their calorie intake more than non-players (2), with an average surplus of 163 kcal. Internet users and online game players are also more likely to be exposed to high-caloric food advertising. The effects of food advertising on health have already been documented for traditional media in both Western and Eastern societies (3–7), while the consequences of Internet food marketing strategies are still unknown. Food companies use websites to promote high-caloric products to children and adolescents using several marketing techniques like ‘advergames’ (8,9). Food marketing has been associated with increasing overweight and obesity in children and adolescents (10), but to the authors’ knowledge, its relationship with Internet exposure and health consequences has not yet been tested. With regard to mental health, an excessive consumption of online video games can lead to or exacerbate addiction, aggressive behaviors, and social isolation. In fact, the effects of online games on mental health have been largely studied in the academic literature (11–16), and currently there are many worldwide institutions that specifically treat online game addiction. Conversely, new media can also have beneficial effects on mental health, offering innovative ways of learning and teaching with positive outcomes for both educational attainment and health. For example, Chan and Fang (17) found that, among Hong Kong students, the Internet was the first medium used to search for new information on various topics, including health education.
Objectives
While many studies on the relationships between new media and health-related behaviors are available for developed countries, few have been conducted in China. By using cross-sectional data from the China Health and Nutritional Survey, we investigate if and how Internet-related activities and fast food consumption—both having an effect on health—interact with each other. In order to account for a health-specific outcome, we also evaluate the association between Internet use and overweight. This paper additionally aims at evaluating the differences between urban and rural residents of the youngest Chinese population, with respect to these associations.
New media in China
International data from the World Bank reported that the number of Internet users per 100 people has increased from 7.3 to 45.8 between 2004 and 2014, and for the same time period, the total number of fixed broadband Internet subscribers went from 24,939,000 to 188,909,000 (18), representing an average yearly increment of 65%. According to the International Telecommunication Union (ITU), China ranked 86 out of 166 countries in the Internet Development Index (IDI), and between 2002 and 2014 China went up four places (19). The 2009 ITU report stressed how Internet penetration in China, measured by access, use, and skill indicators, had largely improved in recent years. National data from the China Internet Network Information Center showed that, in 2013, the average per week amount of time spent online reached 25 hours, which represents an increase of 4.5 hours with respect to the previous year. The number of mobile online game users in China had reached 215 million by the end of December 2013, representing a growth of 75.94 million, or 54.5%, over the end of December 2012. Of all the available mobile online games, users had played with 43.1% of them, representing an increase of 9.9% between 2011 and 2012. Students are the largest population among Internet users, accounting for 25.5%. Mobile lnternet consumption is also significantly increasing in China. By the end of December 2013, there were 500 million mobile Internet users, which represents a growth of 80.09 million compared with that at the end of 2012. i
Methods
Data
This study used secondary data from the 2009 dataset of the China Health and Nutrition Survey (CHNS), which is a large longitudinal survey representative of nine Chinese provinces. The CHNS collects a broad range of information on both adult and youth populations, including information about new media usage and length of exposure, and preferences towards fast foods. For the purpose of this study, we considered only the sub-sample of the youth population aged 6–18 years old. All the analyses considered in this paper were conducted for children in school and, within this group, we also observed differences between rural and urban sub-samples. The China Health and Nutrition Survey questionnaire included a dummy question asking whether the respondent lived in an urban or rural area. Specifically, the survey sample design followed a multistage random cluster process so that urban sites included neighborhoods of the capital and the cities of the provinces selected in the sample, while rural sites included villages and townships. Since 2000, the primary sampling has been composed as follows: 36 urban neighborhoods, 36 suburban neighborhoods, 36 towns, and 108 villages. ii
Too few observations for children and adolescents not in school were available in the 2009 CHNS dataset, so no statistical analysis was performed for this population. Young individuals who are in school allocate their time differently from children and adolescents not in school because they spend more time in school-related activities, such as doing homework and studying, with respect to children who are not in school.
New media exposure was measured by dummy and continuous variables. In particular, respondents were asked if during weekends or before and after school they participated in activities like watching movies and videos online, playing video games, surfing on the Internet, participating in chat rooms, and playing computer games. iii They were also asked how much time they spend on each activity. The total amount of minutes per week was calculated and used in the statistical comparisons. The variables measuring fast food consumption were the number of times the respondent had eaten at a fast food restaurant in the past three months, whether the respondent liked or did not like fast food restaurants, and whether the respondent liked high-energy foods typically consumed in fast foods, like salty snacks and sugary drinks. We also included gender and age measures and a variable indicating whether the child or adolescent perceived himself/herself as overweight or not.
Statistical analysis
We first used t-tests and z-tests to evaluate the different length of time of media exposure among normal and high new media users and among urban and rural residents for children and adolescents in school. To determine whether an individual was a normal or a high new media user, a continuous variable summing up all the internet-related activities was created. Normal users were defined as respondents spending two or fewer hours in front of the computer every day, high users as respondents spending more than two hours. This definition followed the guidelines of Medline Plus, US National Library of Medicine and National Institutes of Health. iv Besides measures of new media, we also included the daily time spent watching traditional media (TV, DVD, and videotapes), the time spent in extracurricular reading, writing, and drawing and the time spent playing with toys. Logistic regression models were then employed to analyze the associations of different activities on fast food preferences’ outcomes. Odds ratios were also calculated. To account for the complex sample design of the CHNS, strata at county level were used in the regression analysis. Statistical analyses were performed using Stata version 12.
Results
Table 1 reports different percentages between normal and high new media users in relation to fast food consumption variables, gender and overweight. The great majority of respondents were normal users. However, we can observe that in the sub-sample of children, the percentages of respondents that liked fast foods and salty snacks, were higher than percentages of normal users (75% vs. 62.78%; 80% vs. 77%) but the differences were not statistically significant. In the sub-sample of adolescents, high new media users liked fast foods and salty snacks significantly more than normal users (82.35% vs. 63.66% and 88.88% vs. 69.27%, p < 0.05).
Differences between low and high new media users.
Table 2 shows the descriptive statistics of the variables considered in the analysis, and also tests for the mean differences between the rural and urban sub-samples. In particular, the variables included were the weekly time spent in front of traditional and new media, the time spent doing homework and playing with toys, the number of times the respondents ate at a fast food restaurant in the past three months, the proportions of respondents who reported liking fast foods, salty snacks, sugary drinks, and who perceived themselves as overweight. Gender and age were also included.
Descriptive statistics and mean comparisons between urban and rural residents for children and adolescents (mean ± SE; z or t-score).
CHNS: China Health and Nutrition Survey 2009.
Significance: *p < 0.1, p **p <0.05, ***p < 0.01.
In the sub-sample of children we found that rural residents spent per week a greater amount of time in front of traditional and new media than do urban residents. For example, the rural group spent on average two hours more in front of the TV (p < 0.01) and four hours more watching online videos (p < 0.01). Children in rural areas also spent significantly more time playing video games (p < 0.1), chatting online (p < 0.01), and playing (p < 0.05). However, urban children spent more time in extracurricular reading (p < 0.05) and doing homework (p < 0.1), suggesting that the pressure coming from school systems and parents was greater in urban than in rural settings. Children living in urban areas ate at fast food restaurants more often, and they liked fast foods, salty snacks and sugary drinks more than the rural sub-sample. These results were not surprising because the density of fast food restaurants in China is greater in urban areas. Furthermore, the proportion of respondents defining themselves as overweight was 15% in urban and 8% in rural areas (p < 0.05). Urban adolescents instead spent more time in front of the TV (p < 0.01) and chatting online (p < 0.05) and more time doing homework (p < 0.01). They also eat at fast food restaurants more often than their rural counterparts (p < 0.01).
Logistic regressions were performed next in order to evaluate the associations between exposure to different new media and fast food consumption. Odds ratios for the urban sub-sample are reported in Table 3. Results show that children who played computer games had higher odds of eating at a fast food restaurant (p < 0.05) and liking sugary drinks (p < 0.01) than children who didn’t play computer games. Children spending time on the Internet were also more likely to be overweight (p < 0.05). Adolescents who watched online videos were more likely to like fast food (although the significance of the p-value is only p < 0.1), and respondents who spent time doing extracurricular reading presented lower odds of liking sugary drinks (p < 0.05).
Odds ratios from the logistic regression for the urban sub-sample.
SE in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.
Blank cells: variable omitted from the model.
Table 4 reports the odds ratios from the logistic regressions of respondents living in rural areas. With regard to the sub-sample of children, watching online videos (p < 0.01), extracurricular reading, surfing on the Internet and playing online games (p < 0.05) were all activities associated with higher odds of eating at fast food restaurants. Surfing the Internet was also associated with higher odds of being overweight. We also found that watching videotapes and spending time in chat rooms were positively related to the variable ‘like fast food’ (p < 0.05 and p < 0.01, respectively). Not many significant findings emerged from the analysis of the sub-sample of adolescents. The only noticeable results were that spending more time in chat rooms and doing homework increased the odds of liking fast food (p < 0.05), while playing computer games increased the odds of obesity (p < 0.1).
Odds ratios for the rural sub-sample.
SE in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1.
Blank cells: variable omitted from the model.
Discussion
While many differences exist between rural and urban areas in terms of education or healthcare access facilities, media penetration and fast food restaurants have reached even remote rural parts of China. It is thus important to discuss the differences between the two groups and the implications that can be drawn from the results.
When comparing rural and urban residents, we observed that rural children spent more time in front of traditional media than their urban peers. These results were also consistent with Chan and McNeal (20), who reported that the time spent in front of the television is lower in urban than in rural residents and that 90% of children in urban areas possess a book, while this percentage was 70% for rural residents. In rural settings, we found that children also spent more time playing video games. However, this result was counterbalanced by the longer time the urban young population spent doing homework and extracurricular reading. Urban residents were subject to a greater pressure for educational attainment, and they allocated most of their free time in doing homework. Furthermore, the educational quality was higher and more competitive in urban than in rural areas (21).
In both the urban and rural sub-samples we found that children who watch online videos were more likely to have eaten or like fast food restaurants and similar results were found for the variable ‘playing computer games’. Surfing on the Internet was positively associated with overweight for children both in rural and urban areas. The mechanism explaining this finding may be related to a higher exposure to online advertisements. Food advertising and marketing directed at young consumers make use of both traditional and new media to reach their target (22), but further evidence is needed in order to draw stronger conclusions. Another interesting result is that extracurricular reading was associated with higher preference for fast foods in both rural and urban settings.
Implications on health promotion policies
Although it is too early to determine whether the effects of new media exposure will be the same for urban and rural residents, it is very important that policy makers carefully monitor the consequences of these rapid societal changes, especially for the child population. Physical inactivity and poor eating habits have been identified as the leading risk factors for the global obesity epidemic and there is also evidence that exposure to the advertising of high-caloric foods contributes to childhood and adolescents’ obesity worldwide (10,23). In the long run, rural residents may try to imitate urban behaviors and experience those that they can afford, with potentially negative health outcomes. Chan and Cai reported that rural residents are disturbed that they are unable to afford the lifestyle of urban residents (24), and at the same time they admire affluent people. Studies on traditional media suggest that attitudes and behaviors towards advertising vary across these two groups. Chang argued that urban children are more skeptical towards advertising (23), and are thus more objective towards the veracity of the lifestyle promoted by advertising. TV commercials are based on idealized urban rather than rural consumption behaviors, and can thus be taken as realistic by children and adolescents without a direct experience of them (24). A similar study suggested that computer games would have become an escape from a stressful life (25), where adolescents struggle to build their own identity because of the opposition between the traditional and conservative values on the one hand, and the modern and consumer-oriented lifestyle on the other.
Developing adequate surveillance systems and collecting timely and updated data on health-related lifestyle behaviors is thus of key importance to better understanding the implications of the rapid expansion of these new forms of consumption and to develop adequate health promotion policies, as for example educational programs. This is particularly urgent in remote rural China where healthcare access and utilization is low (26).
Some limitations of this study have to be acknowledged. First, no causality can be inferred due to the cross-sectional dataset used in this analysis. It is likely that unobserved heterogeneity may affect the link between health-related behavior and new media consumption and this is confirmed by the non-significance of some of the logistic regression models. The analysis did not control for the role of parents and other socio-demographic variables. Parents play a fundamental role in deciding the time of exposure, in monitoring the type of content, and in deciding whether a child can have media equipment in his/her room. Secondly, the continuous variables of sedentary and active time were self-reported and thus statistics might be affected by measurement errors.
Conclusion
We found a significant relationship between new media exposure and fast food consumption, in both rural and urban areas. Both of these behaviors are potentially associated with negative mental and physical health outcomes and we suggest that further research and implementation of surveillance systems are needed in order to monitor the consequences and promote health promotion policies.
Footnotes
Conflict of interest
None declared.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.