Associations of Diverse Internet Device Use and Activities with Depression in Chinese Adolescents: Gender and Geographical Differences

doi:10.21203/rs.3.rs-2961689/v1

Download PDF

Research Article

Associations of Diverse Internet Device Use and Activities with Depression in Chinese Adolescents: Gender and Geographical Differences

https://doi.org/10.21203/rs.3.rs-2961689/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background After the COVID-19 outbreak, Chinese adolescents are increasingly dependent on the Internet. They use multiple devices and engage in various Internet activities. This study explored the associations of mobile/desktop device use and five popular activities with depression in different subgroups of Chinese adolescents. The classification of subgroups was based on gender and geographic disparities in digital technology use.

Methods Data were from China Family Panel Studies (CFPS) in 2020 and included 2,877 primary and secondary school students aged 10-19 years. We employed the ordinary least squares regression models with interaction terms to observe the gender and geographical differences in the relationship between diverse Internet device use and activities and depression.

Results Only prolonged Internet use, whether on all devices or either one, was found to be positively associated with their risk of depression; using the desktop device for less than 1 hour was found to have the opposite effect. Online learning and gaming occasionally, shopping, and posting WeChat Moments frequently were positively linked with depression. Subsequently, mobile device usage time and the frequency of viewing short videos were positively associated with girls’ risk of depression, while in boys, opposite associations were observed. Furthermore, for rural adolescents, despite the adverse effects of prolonged desktop device use, using mobile devices for 1 to 3 hours or desktop devices for less than 1 hour was associated with their reduced risk of depression. Less frequency of online learning, games, and posting WeChat Moments, or increased frequency of watching short videos, could mitigate their risk of depression. These Internet use variables exhibited different associations with depression risk among adolescents in provincial capitals, prefecture-level cities, and counties.

Conclusion Our study demonstrates that using mobile/desktop devices and engaging in various online activities can lead to disparate benefits and risks for Chinese adolescents, depending on their gender and geographic location. Findings guide parents and caregivers in helping children develop healthy and balanced Internet usage habits, considering children’s gender and residential area.

Mobile devices

desktop devices

online activities

depression

digital inequality

Chinese adolescents

The “China Youth Development Report” reported that about 30 million children under the age of 17 years in China suffer from various emotional disorders and behavioral problems [1]. Depression is a prevalent concern, with a detection rate of 24.6% among adolescents, of which 7.4% were found to be severe cases. The detection rate of depression is higher among rural and female adolescents compared to their urban and male counterparts, respectively [1]. The consequences of failing to address depressive symptoms can extend into adulthood, leading to illness, disability, and suicide [2].

Researchers have attributed the increase in depression cases among adolescents to excessive Internet use [3, 4]. Compared with adolescents in developed countries, Chinese adolescents are relatively late in popularizing smartphones and using the Internet daily [5]. In the early years, Chinese students are more inclined towards using computers for online gaming, while students in Western countries are more likely to use them for study [5, 6]. However, in recent years, especially after the COVID-19 outbreak, mandatory physical distancing and the loss of offline social connection have resulted in a greater reliance on the Internet among Chinese adolescents [7]. In 2020, their Internet penetration rate was 94.9%. Compared with the figures in 2019, the proportions of more than 30 minutes on weekdays and more than 1 hour on holidays increased by 6.9% and 8.9%, respectively. Taking online classes became the norm, and online social media became a way for them to cope with stress and poor mental health [8, 9]. Overall, the positive function of Internet use is gaining prominence [10-13], and therefore, we doubt whether daily Internet use is a poor behavior for Chinese adolescents. Most studies on Internet use among Chinese adolescents used data from a time when only a minority of young people were online. They focused on Internet addictive behaviors or measured Internet use time as a continuous variable without observing the health effects of short durations or moderate use [14-17].

Chinese adolescents use various Internet devices, including smartphones, tablets, desktops, and laptops. The most popular purpose of using the Internet is learning, listening to music, watching video clips, gaming, and online communication [8, 9]. Research suggests that the relationship between Internet use time and depression varies by device and online activity [17-22]. However, they did not consider the impact of short-duration or moderate use of various devices and some popular activities (e.g., posting Moments on WeChat). Most importantly, the previous studies did not further explore whether the relationships between different Internet device use and activities and adolescent depression varied by individual characteristics.

Previous research suggests that the relationship between Internet use and depression could depend on gender and geographic location [15, 20, 23, 24]. One critical reason is the gender and geographical location differences in access to digital technologies (the first-level digital divide) and the extent and pattern of use of digital technologies (the second-level digital divide) among adolescents [25, 26]. Girls and adolescents from areas of low socioeconomic resources often have less access to technology and the Internet [8, 9]. Digital inequalities have real consequences on the daily lives of children and young people and may impact their development across a wide range of areas [27]. Few studies have been conducted on whether digital inequalities can be detrimental or beneficial to the mental health of Chinese girls or adolescents from areas of less socioeconomic resources. This study will use data from a national survey collected after the outbreak of COVID-19 to explore the relationships between different Internet device use and activities and depression among adolescents across gender and geographical location.

Relationship between Internet use time and depression

Researchers have proposed the hypothesis of displacement and stimulation [10]. The displacement hypothesis assumes that the negative impact of Internet use on mental health manifests itself primarily as time displacement and social interaction displacement. Internet activity is primarily conducted in solitude and often recreationally. It displaces individuals in more meaningful daily activities such as sleep, physical exercise, school attendance, and offline interactions [10, 28-30]. Adolescents who spend large amounts of time on the Internet may have conflicts with their parents and guardians and have an increased risk of developing mood disorders, including loneliness, distress, anger, loss of control, and social withdrawal [31]. Moreover, when used for communication purposes, online social interaction is primarily with weak relationships rather than with close family and friends, and thus has little benefit to the psychosocial well-being of individuals [32]. Meanwhile, the time spent online displaces face-to-face social engagement, reducing the quality of social relationships. The stimulation hypothesis holds that Internet use may serve as a coping mechanism for depressive feelings and can stimulate well-being by helping people avoid boredom and cope with a lack of stimuli in everyday situations, making them aware of interesting events, enhancing social connectedness, providing social support, and enabling individuals to express thoughts and feelings [10-12].

Research on adolescents in developed countries suggests an inconsistent relationship between daily Internet use time and depression. Some studies support the displacement hypothesis, suggesting that daily Internet use time or frequency is positively associated with the risk of depression [15, 33, 34]. Some studies found a nonlinear U-shaped relationship [35, 36]. For instance, Moreno and colleagues reported that compared to low daily use (for example, <30 minutes), adolescents who use the Internet regularly (30 minutes to 3 hours) are at a lower health risk, while those who use the Internet excessively (>3 hours) are at high risk [35].

Research on Chinese adolescents also examined the effect of Internet use time on the risk of depression among Chinese adolescents and found that adolescents with longer online durations reported higher levels of depression [15, 16]. However, they generally focused on long-duration Internet use and did not show the health effects of low or moderate daily use. Furthermore, the data used in most of these studies were collected before COVID-19. Digital technologies provided various benefits in addressing people’s mental health concerns during COVID-19 [37]. Affected by the epidemic, children and young people became dependent on the Internet for learning, entertainment, and social interaction [8, 9]. When adolescents use the Internet more for learning and spend less time on recreational activities, it can promote the quality of their relationships with their parents and teachers and benefit their mental health. This means that the relationship between daily Internet use time and depression may no longer be a positive linear relationship, which needs to be re-examined.

The associations of diverse Internet device use and activities with depression: Differences in gender and geographical location

Researchers suggest that measuring Internet use should consider the frequency of use, variation/range of use, and autonomy of use [38, 39]. Therefore, we measured the time spent on different Internet devices that can reflect the autonomy of use. Internet activities were measured by combining the frequency and variation of use, expressed as the frequency adolescents engaged in online learning, gaming, watching short video clips, shopping, and posting WeChat Moments.

Compared to desktop devices, mobile devices are more portable and user-friendly [40]. They are more effective in improving adolescents’ access to information, enhancing their learning and knowledge, helping to avoid boredom, adding fun and enjoyment to daily life, and providing more opportunities to interact with family and friends [41]. However, adolescents are easily distracted by the plethora of choices provided by smartphones [17, 42], which makes adolescents more likely to experience a range of adverse outcomes, such as sleep deprivation, anxiety, mood disorders and reduced academic performance [43, 44]. A study by Ma and Gu, which can be seen as a precursor to this study, simultaneously investigated the effect of mobile and desktop device use on the risk of depression among Chinese adolescents [17]. However, as they measured time consumption on different devices as a continuous variable, we still cannot determine which specific combination of durations on various devices could help reduce the risk of depression.

Furthermore, research suggests that different Internet use creates unequal effects on adolescent mental health [17-22]. For instance, Ma and Gu found that adolescents who used the Internet for gaming, shopping, and viewing short videos had a higher risk of depression, while online learning frequency was not significantly linked to their depression risk [17]. They did not examine the health effect of using chat apps, such as WeChat. WeChat Moments is the most popular platform on which people can convey their current status to all or a selected group of contacts by sharing feelings, photos, and short videos of their daily lives. Meanwhile, users can also easily view the latest status of all their contacts. Some researchers have claimed that status updates, an active form of social networking sites (SNS) use, can positively predict adolescents’ well-being, as they can enhance connections with family and friends, thereby increasing their social support perceptions [45, 46]. However, when users post WeChat Moments, they will inevitably browse the statuses posted by their friends, which is a passive form of SNS use. As per social comparison theory, users have an innate drive to compare themselves with their friends. When engaged in upward comparison behaviors (i.e., comparing with someone they think is better), they may develop various adverse outcomes attributable to their reluctance to solicit help when needed [45, 47, 48]. These complicated features lead to an inconstant relationship between the behavior of posting WeChat Moments and depression risk in Chinese adolescents. Furthermore, during the epidemic, the positive and negative effects of posting WeChat Moments may be strengthened because WeChat Moments is a critical channel for gaining information, seeking help, and emotional release. Therefore, this relationship may become more complex. Prior studies on Chinese teenagers concentrated on the health consequences of social networking addiction and did not specifically analyze the effect of WeChat Moments [49, 50].

Gender differences

Gender schema theory and social role theory posit that boys and girls develop gender-appropriate cognitive schemas in early childhood through social learning, which largely influence individuals’ thought processes and behaviors, enabling them to perform different social roles depending on the specific social and cultural environment [51]. Research reported that women were less likely to use the Internet and had less access to opportunities for Internet devices, device diversity, and peripheral diversity [25, 52]. Men were found to have a more confident attitude towards technology use and were more motivated to acquire digital knowledge than women [53], thus, they develop less digital anxiety and have higher self-efficacy [54]. These imply that when forced to use the Internet frequently, compared to boys, girls may have a heavier psychological burden. Some studies have confirmed that among those with high levels of Internet use, women were at a higher risk of depression than men [15, 55, 56]. However, existing research does not provide an answer to whether both desktop and mobile device use would pose a higher threat to girls’ mental health and which combinations of Internet device use are associated with a lower risk of depression for boys and girls.

Studies have found that males prefer to use the Internet for information, games, and entertainment, while females are more likely to use communication tools [44, 57, 58]. Research on adolescents in developed countries suggests that girls are more deeply affected by certain screen activities, such as listening to music and using the Internet, and are at a higher risk of depression compared to boys [20, 21]. However, the relationships between playing electronic games and adolescent mental health were found to not vary by gender [21]. Due to the differences in online activities between adolescents in China and developed countries [58, 59], we cannot determine whether various online activities, such as online learning, gaming, watching short videos, shopping, and posting WeChat Moments, lead to a higher risk of depression in Chinese girls than in boys.

Geographical location differences

The disparities in the levels of socioeconomic development in different administrative divisions are replicated in the digital world. Schools in underdeveloped regions generally lag regarding information and communication technology (ICT) resources and teacher capital, such as teachers’ limited beliefs about technology. This study divided geographical locations into four tiers: provincial capital, prefecture, county, and rural area, rather than just urban and rural areas.

Urban adolescents use multiple devices, with a particular focus on applications such as search engines, social networking sites, news, and shopping, which can help them accumulate more capital and provide them with opportunities and resources to improve their learning and create friendships and social groups. In contrast, rural adolescents primarily use mobile phones for their Internet needs, and they prefer leisure and entertainment applications like short videos, animations, and comics for instant gratification [8, 9]. Internet use has both positive and negative effects on individuals’ mental health [10-12, 29, 30]. This means that the impact patterns of Internet use on urban and rural adolescents’ depression may be different. To our knowledge, only two studies compared urban-rural differences in the impact of time spent using the Internet on Chinese adolescents’ mental health or well-being [15, 24]. Zhou and Ding found that prolonged use can widen the depression gap between adolescents in developed and underdeveloped regions [15], while Long and Han found that Internet use can help narrow the subjective well-being gap between urban and rural adolescents [24]. They did not distinguish the desktop and mobile device use and employed data collected before the epidemic or earlier.

Furthermore, only Long and Han differentiated some online activities, including learning, socializing, and entertainment [24]. They found that online learning and entertainment were positively associated with the subjective well-being of rural but not urban adolescents. It can be inferred that using the Internet for learning and entertainment may be beneficial in reducing the risk of depression among rural adolescents, which remains to be verified. Since they did not distinguish various popular entertainment activities, the respective effects of frequent participation in watching short videos, gaming, posting WeChat Moments, and shopping on adolescent depression in rural areas and different tiers of cities are uncertain [15, 24].

The current study

The current study serves three purposes. First, using data collected during the pandemic, this study re-examined the relationship between Internet use time and the risk of depression in Chinese adolescents, taking into account the effects of low and moderate daily Internet use. Second, it explored the association of different time durations on mobile and desktop devices with adolescent depression; and the relationship between the frequency of online activities (primarily posting WeChat Moments) and depression. Finally, this study further investigated whether there are gender and location differences in the relationships between different Internet device use and online activities and depression, thus discussing the impact of digital inequalities on adolescent mental health.

Survey procedure and participants

We used data from China Family Panel Studies (CFPS), a national longitudinal survey conducted by Peking University since 2010. The CFPS sample was drawn from the population of 25 provinces, which account for approximately 95% of the total population of mainland China. CFPS adopted the probability proportionate to size sampling (PPS). According to the administrative divisions and socioeconomic levels, the organizers selected 16 districts and counties from Liaoning, Henan, Gansu, and Guangdong (“large provinces”), respectively, a total of 64; sampled 32 subdistricts (townships) from Shanghai and 80 districts and counties from other 20 provinces (“small provinces”). Then they selected two villages from each subdistrict (township) in Shanghai and four villages from each district and county in the other 24 provinces, having a total of 640 villages. After that, 28-42 households were sampled from each village. More details can be found in the introduction by Xie and Hu [61].

The data used in this study were collected from June to August 2020, known as CFPS 2020. This study focused on primary and secondary school students aged 10-19 years. After excluding the samples that did not meet the research objectives, the total number was 2,877. The sample characteristics are shown in Table 1.

Measures

Dependent variables

Depression was measured by using a short form of the Center for Epidemiologic Studies Depression Scale (CES-D), known as CES-D 8. CES-D, developed by Radloff, containing 20 items, is a self-rated scale to measure depressive symptoms in the past week prior to the survey, involving depressed mood, feelings of guilt and worthlessness, helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance [62]. Each item is scored from 1 ‘rarely (less than one day)’ to 4 ‘most of the time (5-7 days)’. CFPS initially adopted CES-D 20, but the 2012 survey reported 20 items to be overloaded and not well accepted, and therefore starting in 2016, they employed CES-D 8. The research has suggested that CES-D-8 has good reliability and validity in measuring depression in various populations [63]. In the sample of this study, a 100% response rate and a Cronbach alpha of 0.90 indicate its good reliability. The total score, the dependent variable, was obtained by adding the eight items after inverting the scores for the two items measuring positive emotions. It ranged from 9 to 31 in the sample of this study with higher scores associated with a higher frequency of depressive complaints.

Independent variables

Mobile device use time was measured by asking ‘On average, how much time per day do you surf the Internet on your mobile devices?’. Possible responses ranged from ‘0 to 1440’ minutes and were categorized into four time periods in hours: 0 hour, ≤1 hour, 1 to 3 hours, and >3 hours.

Desktop device use time was measured by asking ‘On average, how much time per day do you surf the Internet on your computer devices?’. Due to the small number of respondents who responded more than 180 minutes, we divided the possible answers into three time periods, measured in hours: 0 hour, ≤1 hour, and >1 hour.

Daily Internet use time was obtained by adding the use minutes of the above two devices and was divided into four groups: 0 hour, ≤1 hour, 1 to 3 hours, and >3 hours.

Internet usageactivities were measured by the frequency of five types of online activities. Starting with ‘During the past week, have you’ (1) ‘used e-learning, including watching or listening to various courses on platforms such as MOOC, or participating in online training?’ and ‘used e-learning almost daily?’ (2) ‘watched short videos or live shows on platforms such as Douyin, Kuaishou, Weishi, Douyu?’ and ‘watched short videos or live shows almost daily?’ (3) ‘played online games, including mobile games such as Honor of Kings; computer games such as World of Warcraft, Tian Long Ba Bu; and other small games such as Dou Dizhu, Happy Farm, QQ Games?’ and ‘played online games almost daily?’ (4) ‘shopped online, including using online shopping platforms or apps such as Taobao, Weidian, JD.com?’. All questions had two response choices, ‘yes’ and ‘no’. For the first three activities, we combined the answers to each set of two questions into three: 0 = ‘no’, 1 = ‘sometimes (yes, but not almost daily)’, and 2 = ‘almost daily’. Posting WeChat Moments were measured by asking ‘During the past year, how often did you share your work or life on WeChat Moments?’, with seven possible answers: never, every few months, once a month, 2-3 times a month, 1-2 times a week, 3-4 times a week, almost daily. We grouped them into three categories: 0 = ‘never’, 1 = ‘sometimes (including every few months, once a month, and 2-3 times a month)’, 2 = ‘often (including 1-2 times a week, 3-4 times a week, and almost daily)’.

Moderators

The gender was coded as 0 = ‘female’ and 1 = ‘male’. Geographic location was measured by the school location at which the adolescent was attending. There were four tiers: provincial capital (including municipalities), prefecture-level city, county, and rural area.

Control variables

The control variables included the respondent’s education, boarding school student (1 = yes; 0 = no), family dinner, and father’s and mother’s education. The respondent’s education also represented the age, which was coded as 1 = ‘primary’, 2 = ‘junior secondary’, and 3 = ‘senior secondary’. Family dinner was measured by asking ‘How many days per week do you eat dinner or supper with your family?’ Possible responses included ‘0 to 7’ and were collapsed into categorical variables: 0 to 2, 3 to 5, and 6 to 7 days a week. Father’s and mother’s education was coded as 1 = ‘primary and below’, 2 = ‘junior secondary’, and 3 = ‘senior secondary and above’.

Analysis

First, we compared differences in depression scores and Internet use time and activities among adolescents between the two genders and four locations using independent sample t-tests, ANOVA, and cross-tabulation with chi-square tests.

We then employed ordinary least squares (OLS) regression and created a series of models with depression scores as the dependent variable to answer the seven questions. Specifically, Models 1a and 1b were created to examine the associations between daily Internet use time and adolescent depression. Model 1a only involved daily Internet use time as an independent variable, while Model 1b added control variables, including gender, geographic location, the respondent’s education, boarding school student, family dinner, and father’s and mother’s education. Models 2a and 2b were designed to examine the relationship between different time durations on mobile and desktop devices and adolescent depression. Models 3a and 3b were developed to examine the associations between online activities and adolescent depression. These two groups performed steps similar to Models 1a and 1b.

Based on Models 2b and 3b, Models 4 and 5, respectively, included the interactions between mobile/desktop device use time and gender, between online activities and gender to examine the gender differences in the relationships between mobile/desktop device use time and each type of activity and adolescent depression. Likewise, Models 6 and 7 contained the interactions between mobile/desktop device use time and geographic location, between online activities and geographic location to examine the geographical differences.

Adolescent depression and diverse Internet device use and activities by gender and geographic location

Table 2 showed that girls had slightly higher mean depression scores than boys. Differences were found in mean depression scores for adolescents in the four localities, in descending order of rural area, county and prefecture-level city, and provincial capital.

The proportions of boys who used the Internet every day (79.83%) and spent more than 3 hours online (31.25%) were significantly higher than those of girls (x² = 11.10, p < 0.05). Specifically, compared to girls, boys had a higher proportion of using mobile and desktop devices and a higher proportion of time spent on these two devices for more than 1 hour per day. The proportions of adolescents who used the Internet every day and spent more than 1 hour a day were similar in provincial and prefecture cities, higher than those of adolescents in counties and much higher than those of rural adolescents (x² = 168.24, p < 0.001). Specifically, the proportions of adolescents using mobile and desktop devices, and using each type of device for more than 1 hour a day, were close between the provincial and prefecture cities, which were higher than the counties and much higher than rural areas.

Among the five types of online activities, the proportion of adolescents watching video clips was the highest. Boys and girls did not differ significantly in the frequency of watching video clips and online learning. Meanwhile, boys reported playing online games ‘sometimes’ and ‘almost daily’ in much higher proportions than girls. Girls reported a higher percentage of online shopping (32.09%), as well as significantly higher proportions of ‘sometimes’ (25.40%) and ‘often’ WeChat posts (8.96%) than boys. The participation frequencies of rural adolescents in five types of online activities were significantly lower than those of urban adolescents. Among urban adolescents, those in provincial capitals had the highest proportion of frequent learning, playing games, shopping, and posting WeChat Moments, followed by adolescents in prefecture cities and counties. The proportion of adolescents in prefecture cities who watched video clips every day was much higher than that of adolescents in provincial capitals or counties.

Relationship between daily Internet use time and depression

Models 1a and 1b in Table 3 showed that after adjusting for control variables, the relationship between daily Internet use time and depression tended to be U-shaped. The coefficient of more than 3 hours was significantly positive; compared to adolescents who spent 0 hour a day online, those who spent more than 3 hours had much higher mean depression scores.

The associations of diverse Internet device use and activities with depression

In Table 3, Models 2a and 2b showed that after adjusting for control variables, the coefficient of using mobile devices for more than 3 hours was positive and statistically significant, while the coefficients of less than 1 hour and 1 to 3 hours were also positive, but not significant. That is, adolescents who spent more than 3 hours per day had far higher depression scores than those who did not use mobile devices. Additionally, the relationship between desktop device use time and depression scores had a U-shaped trend. The coefficients of less than 1 hour and more than 1 hour were negative and positive, respectively. That is, compared to adolescents who did not use desktop devices, adolescents who spent 1 hour or less had significantly lower depression scores, while those who used more than 1 hour scored higher.

Model 3a showed that online gaming and shopping were positively correlated with depression. Specifically, adolescents who played games ‘sometimes’ had higher depression scores than those who did not (0.30, p < 0.1). The coefficient of gaming ‘every day’ was not statistically significant. Adolescents who shopped online scored significantly higher on depression than those who didn’t (0.46, p < 0.01). After adjusting for control variables, model 3b showed that the regression coefficient of gaming ‘sometimes’ increased with a nominal significance level ≤ 0.05, while the online shopping coefficient decreased but was still significant at the 5% level. Meanwhile, the coefficients of ‘sometimes’ learning online and ‘often’ posting WeChat Moments increased and were statistically significant. The depression scores of adolescents who ‘sometimes’ participated in online learning or ‘often’ posted WeChat Moments were significantly higher than those who did not.

Gender differences

Model 4 in Table 4, Model 4 showed that the relationship between mobile device usage time and depression differed by gender (-0.93, p < 0.05; -0.90, p < 0.05; -1.16, p < 0.01). The regression coefficients for using mobile devices for less than 1 hour, 1 to 3 hours, and more than 3 hours were 0.73, 0.51, and 1.12, respectively. Compared to the depression scores of boys who did not use mobile devices, the scores of boys who spent less than 1 hour, 1 to 3 hours, and more than 3 hours were lower by 0.20 (0.73-0.93), 0.39 (0.51-0.90), and 0.04 (1.12-1.16), which means that their risk of depression decreased with increasing mobile use time. The depression scores of girls who used mobile devices online for less than 1 hour, 1 to 3 hours, and more than 3 hours were higher by 0.73, 0.51, and 1.12 than those of girls who did not use, indicating that the depression of girls increased with the use time. However, the effect of desktop use time on depression was not significantly different between genders.

Model 5 showed that the relationship between the frequency of watching video clips and depression varied by gender (-0.80, p < 0.05; -0.90, p < 0.05). The regression coefficients for watching short videos ‘sometimes’ and ‘almost daily’ were 0.16 and 0.67. Compared to boys who did not watch video clips, the depression scores of boys who watched sometimes or almost daily were lower by 0.64 (0.16-0.80) and 0.23 (0.67-0.90). This suggested that their risk of depression decreased as the frequency of watching increased. The depression scores of girls who watched short videos sometimes or almost daily were higher by 0.16 and 0.67 than those of girls who did not, indicating that the depression of girls increased with the frequency of watching.

Geographic location differences

Model 6 in Table 5 showed that the effect of using mobile devices for 1 to 3 hours a day on depression was significantly different between adolescents in counties and rural areas (0.97, p < 0.05). The regression coefficient for using mobile devices for 1 to 3 hours a day was -0.14. The depression scores of adolescents in counties who used mobile devices for 1 to 3 hours a day were higher by 0.83 (-0.14+0.97) than those who did not use. Compared to the rural adolescents who did not use mobile devices, the depression scores of those who spent 1 to 3 hours were lower by 0.14. Additionally, the relationship between desktop device use time and depression scores differed significantly between adolescents in rural areas and provincial capitals. The regression coefficients for desktop device use for less than 1 hour and more than 1 hour were -0.70 and 0.22, respectively. The depression scores of adolescents in provincial capitals who used desktop devices for less than 1 hour and more than 1 hour were higher by 0.85 (-0.70+1.55) and 2.26 (0.22+2.04) than those who did not use. This suggests that the risk of depression among adolescents in provincial capitals increased substantially with increasing desktop device use time. Compared to the depression scores of rural adolescents who did not use desktop devices, the scores of rural adolescents who spent less than 1 hour and more than 1 hour were lower by 0.70 and higher by 0.22, respectively. That is, as rural adolescents’ time on the desktop device increased, their risk of depression decreased first and then increased.

Model 7 showed that the relationship between online learning ‘almost daily’ and depression scores differed significantly between adolescents in rural areas and counties (-1.00, p < 0.1). The regression coefficient for online learning ‘almost daily’ was 0.71. County adolescents who learned online ‘almost daily’ had depression scores lower by 0.29 (0.71-1.00) than those who did not use the Internet for learning. Conversely, the depression scores of rural adolescents who learned online almost daily were higher by 0.71 than those who did not study online. Model 7 also showed that the associations between the frequency of online gaming and depression differed significantly between rural areas and provincial capitals (-1.25, p < 0.05) and between rural areas and counties (-0.83, p < 0.1). The regression coefficients for playing games sometimes and almost daily were 0.70 and 0.59, respectively. The scores of adolescents in provincial capital cities who sometimes played games were lower by 0.55 (0.70-1.25) than those who did not. Rural adolescents who sometimes played online games scored higher by 0.70 than those who did not play. The depression scores of county adolescents who played games almost daily were lower by 0.24 (0.59-0.83) than those who did not play games, while rural adolescents who played games nearly every day scored higher by 0.59 than those who did not play. That is, rural adolescents who played games were at higher risk of depression. Additionally, the effect of the frequency of watching video clips on depression scores differed significantly between rural areas and different levels of cities. The regression coefficients for watching video clips sometimes and almost daily were -0.79 and -0.25, respectively. Compared with rural adolescents who did not watch clips, the depression scores of those who watched sometimes and almost daily were lower by 0.79 and 0.25, indicating that the risk of depression was low. Conversely, the risk of depression among urban adolescents increased as the frequency of watching short videos increased. Compared to adolescents in provincial capitals who did not watch video clips, those who watched sometimes and almost daily had depression scores higher by 0.77 and 1.15 (1.56, p < 0.05; 1.40, p < 0.05). Adolescents in prefecture-level cities who watched sometimes had depression scores higher by 0.08 than those who did not watch (0.87, p < 0.1). County adolescents who watched sometimes and almost daily scored lower by 0.02 and higher by 0.67 than those who did not (0.77, p < 0.1; 0.92, p < 0.05), respectively.

There were significant differences in the relationship between the frequency of posting WeChat Moments and depression among adolescents in rural areas and prefecture-level cities (0.99, p < 0.05; -1.50, p < 0.05). The regression coefficients of sometimes and often posting WeChat Moments were -0.46 and 0.90, respectively. Compared to rural adolescents who didn't post WeChat Moments, the depression scores for those who sometimes and often posted were lower by 0.46 and higher by 0.90, meaning that their depression risk first decreased and then increased sharply. Conversely, among adolescents in prefecture-level cities, the scores of those who sometimes and often posted WeChat Moments were higher by 0.53 and lower by 0.60 than those who did not, indicating that their risk of depression first increased and then decreased.

Different from prior research that generally treated Internet use as a singular behavior, this study disaggregated Internet behavior and explored the relationship between behavioral granularity and depression in different subgroups of Chinese adolescents in the context of pervasive Internet usage. Regarding the first purpose of this study, Table 3 suggested that using the Internet for less than an hour and one to three hours was not associated with the risk of depression among Chinese adolescents. A higher risk of depression was only found among adolescents who used the Internet more than three hours per day. This does not follow the existing findings [15, 24]. For example, Zhou and Ding found that an increase in Internet use time was associated with a higher risk of depression among Chinese junior middle school students, i.e., a linear relationship [15]. Our results suggest that it is inappropriate to maintain a consistently negative attitude towards low and moderate Internet use.

As for the second purpose of this study, the results suggested that both prolonged mobile device use (e.g., >3 hours per day) and desktop use (e.g., >1 hour per day) could increase the risk of depression among adolescents. Using desktop devices for less than 1 hour was negatively associated with the risk of depression. This means that if adolescents must be online for a long time, combining two types of devices, such as controlling mobile device use within two hours and desktop device use within one hour, is an effective strategy to protect their mental health because it can help mitigate reliance on a single device and prevent them from becoming addicted. This finding differs from that of Ma and Gu, who suggested that adolescents should refrain from using mobile devices but are free to use desktop devices as desired [17]. This discrepancy can be attributed to their failure to consider the potential consequences of low or moderate use.

Consistent with Ma and Gu, this study showed that adolescents who engaged in online gaming and shopping could have a higher risk of depression [17]. Differently, we found that adolescents who played online games less than every day had higher depression scores than those who did not or who played almost daily, which contradicts previous research in other countries that found frequent gamers had the highest risk of depression [64, 65]. The reason may be related to the motivation to play [66]. In the summer vacations during the pandemic, the primary purpose of most adolescents playing games might be to have fun and socialize. Frequent gaming can enable them to maintain close contact and interaction with peers, which has a positive impact on their psychological well-being [64, 66]. The social benefits obtained from playing games occasionally may be relatively limited. Furthermore, playing occasionally may exacerbate adolescents’ desire, particularly when forced to do so, which can be distracting and lead to a bad mood. Griffiths summarized that gaming can positively affect individuals when it adds to life but can have negative consequences when it takes from life [67]. This means that the effect of the new policy that “youths are to play games for no more than three hours per week and only on weekends and holidays” needs to be re-examined.

Furthermore, after adjusting for control variables, it was found that the adolescents who studied online less frequently than every day were more depressed than those who did not. This finding differs from that of Ma and Gu, who reported no significant correlation between the frequency of online learning and depression levels [17]. The main reason is that they divided the depression scores into two grades: above 14 and below 14. Adolescents may experience a multitude of challenges during their participation in online learning, including lack of engagement, inadequate proficiency in using technological tools, and limited preparedness and self-efficacy for online learning [68]. They may be susceptible to distractions, frustration, anxiety, and confusion, which can be exacerbated by poor self-control [68]. However, learning online almost daily can acquaint them with online learning and reduce the associated pressures, while also bringing a sense of achievement and academic competitiveness [69]. Another finding of this study that contradicts the research conducted by Ma and Gu is that the relationship between the frequency of watching short videos and adolescent depression tends to be U-shaped but not statistically significant. The reason is that Ma and Gu compared the depression difference between adolescents who watched short videos less than daily and almost daily [17], while this study took the adolescents who did not watch short videos as a reference group to report their relative risk of depression. Taken together, it can be concluded that adolescents can watch short videos without increasing their risk of depression unless watching every day.

Frequently posting WeChat Moments is associated with an increased risk of depression. It implies that the detrimental health effects of frequent postings outweigh the potential benefits. In addition to the adverse outcomes of upward social comparison [47, 48], individuals who frequently update their status can expose themselves to privacy breaches and defamation. Meanwhile, their endless checking of updates and notifications can also lead to internal distraction, unable to focus, and a sense of emptiness [70].

Regarding the third purpose of this study, girls’ low access to mobile devices seems to be a protective factor for their mental health since we found that their risk of depression became much higher as time spent on mobile devices increased. Viewing video clips could significantly increase girls’ risk of depression. Possibly because Internet content may be more conducive to the interest of men. Most victims of cyberbullying are women [71]. A disturbing wave of “gender wars” and misogyny culture has emerged on popular short-video platforms. In numerous social news videos addressing the topic of ‘gender’, comments made by misogynistic groups frequently express dissatisfaction with the increasing prominence of women’s voices while also attacking women’s positions, belittling, stigmatizing, and oppressing them [72]. Moreover, in life-sharing or beauty-related videos, any behavior, speech, or habits exhibited by the female protagonists that deviate from the “good woman” standard dictated by the patriarchal culture become reasons for criticism or even slander, often accompanied by sexism and linguistic violence in the comments section [72, 73]. Unfortunately, apps based on incomplete data and badly-designed algorithms are further amplifying these problems [73].

The moderate utilization of mobile devices (1 to 3 hours per day) and desktop devices (≤1 hour per day) can mitigate the risk of depression for rural adolescents, suggesting that low access to multiple Internet devices is somewhat detrimental to their mental health. Prolonged use of desktop devices can increase their risk of depression. This result differs from the linear relationship found in previous studies because of variations in measurement [15]. We also observed a positive correlation between moderate mobile use and depression scores among adolescents in counties, as well as between time spent on desktop devices and depression scores among adolescents in provincial capitals. These geographical differences may be related to the different effects of various online activities. Specifically, daily online learning could increase the risk of depression among rural adolescents. The reason may be that rural adolescents face more difficulties and pressures when learning online, such as difficulties in network connectivity and technical operation of learning software, since their parents or caregivers (such as grandparents) are usually less educated and unable to offer sufficient technical support. Additionally, parents or caregivers of rural adolescents may not be able to supervise them frequently, resulting in poor learning outcomes that can affect their self-confidence and self-efficacy, leading to negative emotions such as anxiety and frustration [15]. Engaging in sometimes and daily gaming was negatively associated with the risk of depression among adolescents in provincial capitals and counties, respectively, whereas it was linked to a heightened risk among rural counterparts. The possible explanation may be that games serve as a socializing tool and an effective means of relaxation and stress relief for urban adolescents who are frequently quarantined during the epidemic [64]. However, for rural adolescents with ample physical activity spaces, frequent gaming can lead to addiction, poor academic performance, and increased conflicts with parents or caregivers who generally have a negative perception of gaming [74].

Watching short videos helps to reduce the risk of depression among rural adolescents, while the risk of depression among urban adolescents increases with the frequency of video consumption. The reason may be that popular short video platforms employ automatic video recommendation algorithms based on users’ geographic location and viewing history. During COVID-19, the video-sharing apps of rural adolescents might show less epidemic-related information and more entertainment and leisure content since they experienced fewer control and preventive measures. In contrast, urban areas are subject to more frequent and stringent pandemic control measures. The video-sharing apps of urban adolescents may be inundated with information on local pandemic situations, which may exacerbate their susceptibility to anxiety and depression [75].

Furthermore, intermittent posting of WeChat Moments could mitigate the likelihood of depression among rural adolescents, while frequent posting could elevate their risk. Conversely, frequent posting of WeChat Moments was beneficial to adolescents in prefecture-level cities. The positive and negative effects of posting WeChat Moments are dependent upon various factors, including the number of friends on WeChat, the quality of interaction and emotional content, self-disclosure, communication skills, and social comparison with friends [45, 46]. For rural adolescents, frequent self-presentation may lead to a higher risk of rumors and cyberbullying [76]. For adolescents in prefecture-level cities, the frequent posting of WeChat Moments may serve as a means of emotional and stress release or a channel for ostentation, offering them the opportunity to acquire additional social support or a sense of psychological superiority [45, 46, 77].

This study has some limitations. First, although the study examined the health implications of specific online activities, the measurements of various Internet activities lacked sufficient granularity. For example, due to restrictions in the original questionnaire and data, this study did not measure the precise duration of each online activity, making the potential impact of time spent on a weekly or daily basis across different activities and the threshold for time duration at which the risk of depression may be increased uncertain. This study also failed to investigate the video-sharing platforms and different kinds of content, leaving uncertain explanations for the observed gender and geographic location differences in the association between the frequency of video clip viewing and the risk of depression. Second, due to sampling constraints, the number of observations from provincial capital cities was relatively low in this study, potentially leading to biased results when comparing such cities with rural areas. Third, the cross-sectional design employed in this study couldn’t establish causal relationships between diverse Internet device use and online activities and depression. However, the longitudinal studies in Western countries confirmed causal relationships, such as between more frequent use of mobile phones and higher levels of depressive symptoms [20] and between limited social media use and low well-being [4, 78]. At least that means that we need to highlight the health consequences of high levels of Internet use, especially mobile phone use. Future investigations should consider more comprehensive and detailed measures of various online activities and conduct longitudinal studies to discern causal relationships between diverse Internet device use and online activities and depression and to explore individual differences.

To sum up, the relationship between daily Internet use time and adolescent depression is inclined to be nonlinear; only prolonged use (>3 hours per day) is associated with a higher risk of depression. This relationship was found to depend on different types of Internet device usage and various activities. Prolonged mobile device use (>3 hours per day) and desktop use (>1 hour per day) are positively associated with the risk of depression; moderate use of desktop devices (≤1 hour per day) shows the opposite effect. Meanwhile, intermittent online learning and gaming, shopping, and frequently posting WeChat Moments are positively related to depression. Further examination found that the relationships between different types of Internet device usage or online activities and depression varied by gender and geographic location. Limited availability of mobile devices can help reduce girls’ risk of depression, while viewing video clips significantly increases their risk. In boys, an opposite effect was observed. Furthermore, despite the adverse effects of prolonged Internet use on rural adolescents, low access to mobile and desktop devices remains a significant contributor to their increased risk of depression. The high accessibility of desktop devices is an unfavorable factor for adolescent depression in provincial capital cities, and moderate mobile use is detrimental to adolescents in counties. The low frequency of Internet usage for studying, gaming, and posting WeChat Moments can help protect the mental health of rural adolescents; however, it is detrimental to adolescents in some urban localities. Not watching short videos would be disadvantageous for rural adolescents but advantageous for urban adolescents. These findings offer specific guidance for parents and caregivers of adolescents across gender and geographical location on monitoring and regulating children’s Internet use behaviors, fostering a healthy and balanced use pattern.

Ethics approval and consent to participate

CFPS was approved by t the Peking University Biomedical Ethics Committee (approval on Project No. IRB00001052-14010). Participation was voluntary.

Consent for publication

Not applicable.

Availability of data and materials

The data analysed from the database is a publicly available database. http://www.isss.pku.edu.cn/cfps/sjzx/gksj/index.htm

Competing interests

The authors declare no competing interests.

Funding

This research was funded by the National Social Science Foundation of China (20CRK021).

Authors' contributions

The first author performed data analysis, wrote the main manuscript text, and prepared all tables. The second and third authors contributed to the design and data interpretation and critically revised the manuscript for important intellectual content.

Acknowledgements

We would like to thank the CFPS project team for providing us with data for free. CFPS is a nationally representative survey.

Hou J, Chen Z. Interannual evolution of adolescent mental health status from 2009 to 2020. In Fu X, Zhang K, Chen X, Chen Z. editors. Report on national mental health development in China (2019-2020). Beijing, China: Social Science Academic Press; p. 188-202. (In Chinese)
Fergusson DM, Horwood LJ, Ridder EM, Beautrais AL. Subthreshold depression in adolescence and mental health outcomes in adulthood. Arch Gen Psychiatry. 2005; 62:66-72.
Augner C, Hacker GW. Associations between problematic mobile phone use and psychological parameters in young adults. Int J Public Health. 2012; 57:437-41.
Shakya HB, Christakis NA. Association of Facebook use with compromised well-being: A longitudinal study. Am J Epidemiol. 2017; 185:203-11.
Mak KK, Lai CM, Watanabe H, Kim DI, Bahar N, Ramos M, et al. Epidemiology of internet behaviors and addiction among adolescents in six Asian countries. Cyberpsychol Behav Soc Netw. 2014;17(11):720-8.
Li N, Kirkup G. Gender and cultural differences in Internet use: A study of China and the UK. Comput Educ. 2007; 48:301-17.
Liu Y, Yue S, Hu X, Zhu J, Wu Z, Wang J, et al. Associations between feelings/behaviors during COVID-19 pandemic lockdown and depression/anxiety after lockdown in a sample of Chinese children and adolescents. J Affect Disord. 2021; 284:98-103.
Communist Youth League Central Committee, China Internet Network Information Center. (2020). National Research Report on Internet Use among Minors in 2019. http://pg.jrj.com.cn/acc/Res/CN_RES/INDUS/2020/5/15/f9820326-b918-44c1-a9c4-a343f6293268.pdf. Accessed 25 Dec 2022. (In Chinese)
Communist Youth League Central Committee, China Internet Network Information Center. (2020). National Research Report on Internet Use among Minors in 2020. https://pic.cyol.com/img/20210720/img_960114c132531c521023e29b6c223e438461.pdf. Accessed 25 Dec 2022. (In Chinese)
Valkenburg PM, Peter J. Online communication and adolescent well-being: Testing the stimulation versus the displacement hypothesis. J Comput-Mediat Commun. 2007; 12:1169-82.
Bessiere K, Kiesler S, Kraut R, Boneva BS. Effects of Internet use and social resources on changes in depression. Inf Commun Soc. 2008; 11:47-70.
Keles B, McCrae N, Grealish A. A systematic review: the influence of social media on depression, anxiety and psychological distress in adolescents. Int J Adolesc Youth. 2020; 25:79-93.
Guo X, Zhang Q. (2023). Internet Use and Physical and Mental Health of Youth Group: A Mediating Effect Analysis Based on Social Interaction and Internet Dependence. Contemp Youth Stud. 2023; 382:87-97. (In Chinese)
Wang P, Liu S, Zhao M, Yang X, Zhang G, Chu X, et al. How is problematic smartphone use related to adolescent depression? A moderated mediation analysis. Child Youth Serv Rev. 2019; 104:104384.
Zhou M, Ding X. Internet use, depression, and cognitive outcomes among Chinese adolescents. J Community Psychol. 2023; 51:768-87.
Wu H, Ji JM, Qian Y, Jin XH, Yu HR, Liu XM, et al. Relationship between depressive symptoms and internet usage time among adolescents: Results from a nationwide survey. Psychiatry Res. 2022; 313:114603.
Ma Y, Gu J. Internet and depression in adolescents: Evidence from China. Front Psychol. 2023; 14:
Selfhout MH, Branje SJ, Delsing M, ter Bogt TF, Meeus WH. Different types of Internet use, depression, and social anxiety: The role of perceived friendship quality. J Adolesc. 2009; 32:819-33.
Hökby S, Hadlaczky G, Westerlund J, Wasserman D, Balazs J, Germanavicius A, et al. Are mental health effects of internet use attributable to the web-based content or perceived consequences of usage? A longitudinal study of European adolescents. JMIR Ment Health. 2016; 3: e5925.
Twenge JM, Farley E. Not all screen time is created equal: associations with mental health vary by activity and gender. Soc Psychiatry Psychiatr Epidemiol. 2021; 56:207-17.
Werneck AO, Hoare E, Stubbs B, van Sluijs EM, Corder K. Association of mentally-active and mentally-passive sedentary behaviour with depressive symptoms among adolescents. J Affect Disord. 2021; 294:143-50.
de Calheiros Velozo J, Stauder JE. Exploring social media use as a composite construct to understand its relation to mental health: A pilot study on adolescents. Child Youth Serv Rev. 2018; 91:398-402.
Pujazon-Zazik M, Park MJ. To tweet, or not to tweet: gender differences and potential positive and negative health outcomes of adolescents’ social internet use. Am J Mens Health. 2010; 4:77-85.
Long C, Han J. Does Internet use narrow the happiness gap between urban and rural adolescents? World Econ Papers. 2022; 1:52-68. (In Chinese)
Van Deursen AJ, Van Dijk JA. Toward a multifaceted model of internet access for understanding digital divides: An empirical investigation. Inf Soc. 2015; 31:379-91.
Joiner R, Stewart C, Beaney C. Gender digital divide: does it exist and what are the explanations? In Rosen LD, Cheever NA, Carrier LM. editors. The Wiley handbook of psychology, technology, and society. Wiley-Blackwell; 2015. p. 74-88.
Livingstone S, Mascheroni G, Stoilova M. The outcomes of gaining digital skills for young people’s lives and wellbeing: A systematic evidence review. New Media Soc. 2023; 25:1176-202.
Mota J, Ribeiro J, Santos MP, Gomes H. Obesity, physical activity, computer use, and TV viewing in Portuguese adolescents. Pediatr Exerc Sci. 2006; 18:113-21.
Zhou R, Fong PS, Tan P. Internet use and its impact on engagement in leisure activities in China. PLoS One. 2014; 9: e89598.
Chen Y, Zheng Z, Yi J, Yao S. Associations between physical inactivity and sedentary behaviors among adolescents in 10 cities in China. BMC Public Health. 2014; 14:1-9.
Cerniglia L, Zoratto F, Cimino S, Laviola G, Ammaniti M, Adriani W. Internet Addiction in adolescence: Neurobiological, psychosocial and clinical issues. Neurosci Biobehav Rev. 2017; 76:174-84.
Winstone L, Mars B, Haworth C, Kidger J. Social media use and social connectedness among adolescents in the United Kingdom: A qualitative exploration of displacement and stimulation. BMC Public Health. 2021; 21:1-15.
Kircaburun K. Self-Esteem, Daily Internet Use and Social Media Addiction as Predictors of Depression among Turkish Adolescents. J Educ Pract. 2016; 7:64-72.
Lim CH, Kim EJ, Kim JH, Lee JS, Lee Y, Park SH. The correlation of depression with Internet use and body image in Korean adolescents. Korean J Pediatr. 2017; 60:17-23.
Moreno MA, Jelenchick L, Koff R, Eickhoff J. Depression and internet use among older adolescents: An experience sampling approach. Psychology. 2012; 3:743-8.
Kim JY. The nonlinear association between Internet using time for non-educational purposes and adolescent health. J Prev Med Public Health. 2012; 45:37-46.
Spanakis P, Peckham E, Mathers A, Shiers D, Gilbody S. The digital divide: amplifying health inequalities for people with severe mental illness in the time of COVID-19. Br J Psychiatry. 2021; 219:529-31.
Hohlfeld TN, Ritzhaupt AD, Dawson K, Wilson ML. An examination of seven years of technology integration in Florida schools: Through the lens of the Levels of Digital Divide in Schools. Comput Educ. 2017; 113:135-61.
Zhong B, Zhu F, Xia L. Is there a digital divide between urban students and migrant students in China? SAGE Open. 2021; 11:21582440211016387.
Odgers CL, Jensen MR. Annual research review: Adolescent mental health in the digital age: Facts, fears, and future directions. J Child Psychol Psychiatry. 2020; 61:336-48.
Harwood J, Dooley JJ, Scott AJ, Joiner R. Constantly connected–The effects of smart-devices on mental health. Comput Human Behav. 2014; 34:267-72.
David P, Kim JH, Brickman JS, Ran W, Curtis CM. Mobile phone distraction while studying. New Media Soc. 2015; 17:1661-79.
Kwok SW, Lee PH, Lee RL. Smart device use and perceived physical and psychosocial outcomes among Hong Kong adolescents. Int J Environ Res Public Health. 2017; 14:205.
Buabbas AJ, Hasan H, Buabbas MA. The associations between smart device use and psychological distress among secondary and high school students in Kuwait. PLoS One. 2021; 16: e0251479.
Frison E, Eggermont S. Exploring the relationships between different types of Facebook use, perceived online social support, and adolescents’ depressed mood. Soc Sci Comput Rev. 2016; 34:153-71.
Gilmour J, Machin T, Brownlow C, Jeffries C. Facebook-based social support and health: A systematic review. Psychol Pop. 2020; 9:328-46.
Barthorpe A, Winstone L, Mars B, Moran P. Is social media screen time really associated with poor adolescent mental health? A time use diary study. J Affect Disord. 2020; 274:864-70.
Skogen JC, Hjetland GJ, Bøe T, Hella RT, Knudsen AK. Through the looking glass of social media. Focus on self-presentation and association with mental health and quality of life. A cross-sectional survey-based study. Int J Environ Res Public Health. 2021; 18:3319.
Li JB, Mo PK, Lau JT, Su XF, Zhang X, Wu AM, Mai JC, Chen YX. Online social networking addiction and depression: The results from a large-scale prospective cohort study in Chinese adolescents. J Behav Addict. 2018; 7:686-96.
Wang P, Lei L, Yu G, Li B. Social networking sites addiction and materialism among Chinese adolescents: A moderated mediation model involving depression and need to belong. Front Psychol. 2020; 11:581274.
Baloğlu M, Özteke Kozan Hİ, Kesici Ş. Gender differences in and the relationships between social anxiety and problematic internet use: Canonical analysis. J Med Internet Res. 2018; 20: e33.
Girl Effect and Vodafone Foundation. Real Girls, Real Lives, Connected. https://static1.squarespace.com/static/5b8d51837c9327d89d936a30/t/5bbe7cbe9140b7d43f282e21/1539210748592/GE_VO_Executive+Summary+Report.pdf (2018). Accessed 15 Jan 2023.
Cai Z, Fan X, Du J. Gender and attitudes toward technology use: A meta-analysis. Comput Educ. 2017; 105:1-13.
Cussó-Calabuig R, Farran XC, Bosch-Capblanch X. Effects of intensive use of computers in secondary school on gender differences in attitudes towards ICT: A systematic review. Educ Inf Technol. 2018; 23:2111-39.
Mars B, Gunnell D, Biddle L, Kidger J, Moran P, Winstone L, et al. Prospective associations between internet use and poor mental health: A population-based study. PLoS One. 2020; 15: e0235889.
Yang SY, Lin CY, Huang YC, Chang JH. Gender differences in the association of smartphone use with the vitality and mental health of adolescent students. J Am Coll Health. 2018; 66:693-701.
Dong H, Yang F, Lu X, Hao W. Internet addiction and related psychological factors among children and adolescents in China during the coronavirus disease 2019 (COVID-19) epidemic. Front Psychiatry. 2020; 11:00751.
Su W, Han X, Yu H, Wu Y, Potenza MN. Do men become addicted to internet gaming and women to social media? A meta-analysis examining gender-related differences in specific internet addiction. Comput Human Behav. 2020; 113:106480.
Tang CS, Wu AM, Yan EC, Ko JH, Kwon JH, Yogo M,et al. Relative risks of Internet-related addictions and mood disturbances among college students: A 7-country/region comparison. Public Health. 2018; 165:16-25.
Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas. 1977; 1:385-401.
Xie Y, Lu P. The sampling design of the China family panel studies (CFPS). Chin J Sociol. 2015; 1:471-84.
Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas. 1977; 1:385-401.
Van de Velde S, Levecque K, Bracke P. Measurement equivalence of the CES-D 8 in the general population in Belgium: a gender perspective. Arch Public Health. 2009; 67:1-15.
Jones C, Scholes L, Johnson D, Katsikitis M, Carras MC. Gaming well: links between videogames and flourishing mental health. Front Psychol. 2014; 5:260.
Santos IK, Medeiros RC, Medeiros JA, Almeida-Neto PF, Sena DC, Cobucci RN, et al. Active video games for improving mental health and physical fitness—An alternative for children and adolescents during social isolation: an overview. Int J Environ Res Public Health. 2021; 18:1641.
Hellström C. Online gaming in relation to negative consequences and ill health among adolescents Västerås: Tryckeriet Västmanlandssjukhus; 2013.
Griffiths M. Online video gaming: what should educational psychologists know? Educ Psychol Pract. 2010; 26:35-40.
Van NT, Irum S, Abbas AF, Sikandar H, Khan N. Online Learning--Two Side Arguments Related to Mental Health. Int J online biomed Eng. 2022; 18:131-43.
Zhang Q, Gao Y. Competition and game-playing: The academic returns and mental health costs of private tutoring. Chin J Sociol. 2022; 42:159-94. (In Chinese)
Zhang W, Yang XY, Yang T, Bottorff JL, Peng S, Yu L, et al. Association of Excessive WeChat Use with Mental Disorders: A Representative Nationwide Study in China. Am J Health Behav. 2021; 45(6):1002-15.
Marcum CD, Higgins GE, Freiburger TL, Ricketts ML. Battle of the sexes: An examination of male and female cyber bullying. Int J Cyber Criminol. 2012 ;6: 904-11.
Xu Z, Gao S. The Internalization of Gender Discrimination in Internet Female Autonomous Region: A Study on the Female Sexism Discourse in Self-Media Beauty Videos. Chin J Journalism Commun. 2019; 41:145-63. (In Chinese)
Lu X, Shan P. Visible and Invisible: Research on Rural Women's Body Narrative in Short Video Platform. News and Writing. 2022; 11:42-50. (In Chinese)
Li L, Abbey C, Wang H, Zhu A, Shao T, Dai D, et al. The Association between Video Game Time and Adolescent Mental Health: Evidence from Rural China. Int J Environ Res Public Health. 2022; 19:14815.
Hou F, Bi F, Jiao R, Luo D, Song K. Gender differences of depression and anxiety among social media users during the COVID-19 outbreak in China: a cross-sectional study. BMC Public Health. 2020; 20:1-11.
Kwan GC, Skoric MM. Facebook bullying: An extension of battles in school. Comput Hum Behav. 2013; 29:16-25.
Duan W, He C, Tang X. Why do people browse and post on WeChat moments? Relationships among fear of missing out, strategic self-presentation, and online social anxiety. Cyberpsychol Behav Soc Netw. 2020; 23:708-14.
Shakya HB, Christakis NA. Association of Facebook use with compromised well-being: A longitudinal study. Am J Epidemiol. 2017; 185:203-11.

Table 1 Sample characteristics

	All	Missings
Characteristics	n = 2,877	%
Age (years): mean (SD)	13.78 (2.63)	0
Gender: n (%) Girl Boy	1,360 (47.27%) 1,517 (52.73%)	0
Educational level: n (%) Primary Junior secondary Senior secondary	1,189 (41.33%) 1,001 (34.79%) 687 (23.88%)	0
Geographic location: n (%) Rural Provincial capital Prefecture-level city County	225 (7.82%) 524 (18.21%) 918 (31.91%) 1,202 (41.78%)	0.28
Boarding at school: n (%) Yes No	1,674 (58.19%) 1,203 (41.81%)	0

Table 2 Depression and Internet use behavior among Chinese adolescents by gender and geographic location

Gender

Geographic location

Girls

Boys

t test / x²

Provincial capital

Prefecture-level city

County

Rural

F test / x²

Depression

12.43 (3.61)

12.28 (3.27)

1.17

12.00 (3.58)

12.24 (3.47)

12.27 (3.31)

12.52 (3.49)

2.10+

Daily Internet use time

0 hour

≤1 hour

1-3 hours

>3 hours

321 (23.74%)

352 (26.04%)

326 (24.11%)

353 (26.11%)

304 (20.17%)

384 (25.48%)

348 (23.09%)

471 (31.25%)

11.10*

20 (8.97%)

59 (26.46%)

55 (24.66%)

89 (39.91%)

57 (10.94%)

130 (24.95%)

140 (26.87%)

194 (37.24%)

163 (17.95%)

232 (25.55%)

220 (24.23%)

293 (32.27%)

384 (32.00%)

313 (26.08%)

257 (21.42%)

246 (20.50%)

168.24***

Mobile device use time

0 hour

≤1 hour

1-3 hours

>3 hours

341 (25.18%)

403 (29.76%)

310 (22.90%)

300 (22.16%)

333 (22.05%)

440 (29.14%)

352 (23.31%)

385 (25.50%)

6.45+

23 (10.22%)

72 (32.00%)

56 (24.89%)

74 (32.89%)

63 (12.07%)

164 (31.42%)

133 (25.48%)

162 (31.03%)

184 (20.18%)

259 (28.40%)

227 (24.89%)

242 (26.54%)

403 (33.64%)

345 (28.80%)

244 (20.37%)

206 (17.20%)

154.61***

Desktop device use time

0 hour

≤1 hour

>1 hour

1,039 (76.57%)

241 (17.76%)

77 (5.67%)

1,049 (69.56%)

306 (20.29%)

153 (10.15%)

25.10***

124 (55.61%)

66 (29.60%)

33 (14.80%)

317 (60.61%)

150 (28.68%)

56 (10.71%)

657 (72.04%)

175 (19.19%)

80 (8.77%)

986 (82.24%)

152 (12.68%)

61 (5.09%)

129.45***

Learning

Sometimes

Almost daily

894 (65.74%)

311 (22.87%)

155 (11.40%)

999 (65.85%)

336 (22.15%)

182 (12.00%)

0.39

129 (57.33%)

57 (25.33%)

39 (17.33%)

322 (61.45%)

129 (24.62%)

73 (13.93%)

576 (62.75%)

234 (25.49%)

108 (11.76%)

861 (71.63%)

224 (18.64%)

117 (9.73%)

37.22***

Gaming

Sometimes

Almost daily

901 (66.25%)

362 (26.62%)

97 (7.13%)

564 (37.18%)

577 (38.04%)

376 (24.79%)

283.60***

101 (44.89%)

72 (32.00%)

52 (23.11%)

221 (42.18%)

194 (37.02%)

109 (20.80%)

453 (49.35%)

303 (33.01%)

162 (17.65%)

689 (57.32%)

364 (30.28%)

149 (12.40%)

49.66***

Table 2 (continued)

Shopping

Yes

923 (67.87%)

437 (32.13%)

1,208 (79.63%)

309 (20.37%)

51.66***

153 (68.00%)

72 (32.00%)

358 (68.32%)

166 (31.68%)

648 (70.59%)

270 (29.41%)

965 (80.28%)

237 (19.72%)

43.24***

Watching short videos

Sometimes

Almost daily

516 (37.94%)

468 (34.41%)

376 (27.65%)

525 (34.61%)

549 (36.19%)

443 (29.20%)

3.65

76 (33.78%)

80 (35.56%)

69 (30.67%)

142 (27.10%)

193 (36.83%)

189 (36.07%)

312 (33.99%)

325 (35.40%)

281 (30.61%)

509 (42.35%)

414 (34.44%)

279 (23.21%)

50.86***

Posting WeChat Moments

Sometimes

Often

892 (65.59%)

346 (25.44%)

122 (8.97%)

1138 (75.02%)

283 (18.66%)

96 (6.33%)

30.75***

126 (56.00%)

67 (29.78%)

32 (14.22%)

305 (58.21%)

159 (30.34%)

60 (11.45%)

618 (67.32%)

237 (25.82%)

63 (6.86%)

976 (81.20%)

164 (13.64%)

62 (5.16%)

139.59***

+p < 0.10, *p < 0.05, ***p < 0.001.

Table 3 Relationship between Internet use behavior and depression among Chinese adolescents

	Model 1a	Model 1b	Model 2a	Model 2b	Model 3a	Model 3b
Daily Internet use time (Ref. 0 hour) ≤1 hour 1-3 hours >3 hours	0.11 (-0.26, 0.47) -0.13 (-0.51, 0.24) 0.50** (0.14, 0.86)	0.24 (-0.14, 0.62) -0.06 (-0.46, 0.34) 0.57** (0.16, 0.98)
Mobile device use time (Ref. 0 hour) ≤1 hour 1-3 hours >3 hours			0.18 (-0.18, 0.54) 0.04 (-0.34, 0.42) 0.60** (0.22, 0.98)	0.25 (-0.12, 0.62) 0.05 (-0.35, 0.45) 0.55* (0.13, 0.98)
Desktop device use time (Ref. 0 hour) ≤1 hour >1 hour			-0.47** (-0.80, -0.13) 0.24 (-0.25, 0.72)	-0.32+ (-0.66, 0.03) 0.45+ (-0.04, 0.95)
Learning (Ref. No) Sometimes Almost daily					0.26 (-0.06, 0.58) -0.06 (-0.46, 0.35)	0.35* (0.03, 0.68) 0.08 (-0.33, 0.49)
Gaming (Ref. No) Sometimes Almost daily					0.30+ (-0.01, 0.61) 0.03 (-0.35, 0.42)	0.39* (0.06, 0.71) 0.24 (-0.18, 0.65)
Shopping (Ref. No) Yes					0.46** (0.15, 0.77)	0.39* (0.05, 0.72)
Watching short videos (Ref. No) Sometimes Almost daily					-0.19 (-0.53, 0.15) 0.28 (-0.09, 0.64)	-0.25 (-0.59, 0.09) 0.22 (-0.16, 0.60)
Posting WeChat Moments (Ref. No) Sometimes					-0.06 (-0.39, 0.27)	-0.03 (-0.37, 0.32)

Table 3 (continued)

Often					0.41 (-0.09, 0.90)	0.48+ (-0.03, 0.99)
Gender (Ref. Girl) Boy		-0.21 (-0.47, 0.04)		-0.21 (-0.47, 0.04)		-0.21 (-0.48, 0.07)
Geographic location (Ref. Rural) Provincial capital Prefecture-level city County		-0.37 (-0.91, 0.17) -0.23 (-0.63, 0.16) -0.35* (-0.68, -0.03)		-0.36 (-0.90, 0.18) -0.21 (-0.60, 0.19) -0.34* (-0.66, -0.01)		-0.38 (-0.92, 0.16) -0.28 (-0.67, 0.11) -0.37* (-0.69, -0.05)
Education (Ref. Primary) Junior secondary Senior secondary		-0.00 (-0.33, 0.33) -0.08 (-0.51, 0.35)		0.03 (-0.30, 0.36) -0.07 (-0.50, 0.36)		-0.11 (-0.43, 0.22) -0.25 (-0.69, 0.19)
Boarding school student (Ref. No) Yes		0.30+ (-0.01, 0.61)		0.28+ (-0.03, 0.59)		0.33* (0.02, 0.64)
Family dinner (Ref. 0-2) 3-5 6-7		0.29 (-0.26, 0.85) -0.38* (-0.73, -0.02)		0.30 (-0.26, 0.86) -0.38* (-0.73, -0.02)		0.22 (-0.34, 0.77) -0.40* (-0.75, -0.05)
Father’s education (Ref. Primary and below) Junior secondary Senior secondary and above		-0.06 (-0.38, 0.25) -0.15 (-0.56, 0.26)		-0.07 (-0.38, 0.25) -0.13 (-0.54, 0.28)		-0.06 (-0.37, 0.25) -0.10 (-0.51, 0.30)
Mother’s education (Ref. Primary and below) Junior secondary Senior secondary and above		-0.33* (-0.64, -0.02) -0.52* (-0.96, -0.08)		-0.33* (-0.64, -0.03) -0.51* (-0.95, -0.06)		-0.32* (-0.62, -0.01) -0.52* (-0.95, -0.09)
R²	0.005	0.022	0.008	0.023	0.013	0.029
Adj R²	0.004	0.016	0.006	0.017	0.009	0.022

Data in Models 1–4 was presented as regression coefficients (95% CI).

+p < 0.10, *p < 0.05, **p < 0.01.

Table 4 Gender differences in the relationship between Internet use behavior and depression

	Model 4	Model 5
Mobile device use time × Gender ≤1 hour × Male 1-3 hours × Male >3 hours × Male	-0.93* (-1.66, -0.21) -0.90* (-1.66, -0.14) -1.16** (-1.93, -0.40)
Desktop device use time × Gender ≤1 hour × Male >1 hour × Male	-0.19 (-0.86, 0.49) 0.05 (-0.97, 1.07)
Learning × Gender Sometimes × Male Almost daily × Male		-0.23 (-0.88, 0.42) 0.38 (-0.44, 1.20)
Gaming × Gender Sometimes × Male Almost daily × Male		-0.05 (-0.71, 0.60) -0.13 (-1.06, 0.80)
Shopping × Gender Yes × Male		0.22 (-0.42, 0.87)
Watching short videos × Gender Sometimes × Male Almost daily × Male		-0.80* (-1.48, -0.11) -0.90* (-1.65, -0.14)
Posting WeChat Moments × Gender Sometimes × Male Often × Male		-0.25 (-0.92, 0.41) -0.64 (-1.66, 0.37)
R²	0.025	0.035
Adj R²	0.016	0.024

Data in Models 4 and 5 was shown as regression coefficients (95% CI).

*p < 0.05, **p < 0.01.

Models 4 and 5 included Internet use variables and control variables in Table 3.

Table 5 Geographical differences in the relationship between Internet use behavior and depression

Model 6

Model 7

Mobile device use time × Geographic location

≤1 hour × Provincial capital

1-3 hours × Provincial capital

>3 hours × Provincial capital

≤1 hour × Prefecture-level city

1-3 hours × Prefecture-level city

>3 hours × Prefecture-level city

≤1 hour × County

1-3 hours × County

>3 hours × County

-0.61 (-2.39, 1.16)

-0.70 (-2.55, 1.14)

-0.78 (-2.61, 1.04)

-0.88 (-2.03, 0.27)

-0.44 (-1.63, 0.76)

0.42 (-0.78, 1.63)

0.48 (-0.36, 1.31)

0.97* (0.08, 1.85)

0.60 (-0.32, 1.52)

Desktop device use time × Geographic location

≤1 hour × Provincial capital

>1 hour × Provincial capital

≤1 hour × Prefecture-level city

>1 hour × Prefecture-level city

≤1 hour × County

>1 hour × County

1.55* (0.33, 2.77)

2.04* (0.39, 3.69)

0.41 (-0.51, 1.33)

-0.19 (-1.55, 1.17)

0.51 (-0.34, 1.35)

0.06 (-1.17, 1.29)

Learning × Geographic location

Sometimes × Provincial capital

Almost daily × Provincial capital

Sometimes × Prefecture-level city

Almost daily × Prefecture-level city

Sometimes × County

Almost daily × County

-0.63 (-1.87, 0.61)

-1.08 (-2.52, 0.37)

-0.18 (-1.09, 0.72)

-0.75 (-1.87, 0.37)

-0.44 (-1.21, 0.33)

-1.00+ (-1.99, 0.00)

Gaming × Geographic location

Sometimes × Provincial capital

Almost daily × Provincial capital

Sometimes × Prefecture-level city

Almost daily × Prefecture-level city

Sometimes × County

Almost daily × County

-1.25* (-2.50, -0.01)

0.23 (-1.17, 1.63)

-0.60 (-1.47, 0.27)

-0.37 (-1.46, 0.72)

-0.22 (-0.97, 0.53)

-0.83+ (-1.79, 0.13)

Shopping × Geographic location

Yes × Provincial capital

Yes × Prefecture-level city

Yes × County

0.90 (-0.26, 2.05)

0.48 (-0.39, 1.35)

0.43 (-0.34, 1.20)

Watching short videos × Geographic location

Sometimes × Provincial capital

Almost daily × Provincial capital

Sometimes × Prefecture-level city

Almost daily × Prefecture-level city

Sometimes × County

Almost daily × County

1.56* (0.28, 2.84)

1.40* (0.02, 2.78)

0.87+ (-0.10, 1.83)

0.35 (-0.69, 1.39)

0.77+ (-0.04, 1.58)

0.92* (0.02, 1.82)

Table 5 (continued)

Posting WeChat Moments × Geographic location

Sometimes × Provincial capital

Often × Provincial capital

Sometimes × Prefecture-level city

Often × Prefecture-level city

Sometimes × County

Often × County

0.47 (-0.79, 1.72)

0.08 (-1.61, 1.77)

0.99* (0.08, 1.91)

-1.50* (-2.85, -0.14)

0.34 (-0.49, 1.17)

-0.11 (-1.42, 1.19)

R²

0.030

0.044

Adj R²

0.018

0.027

Data in Models 6 and 7 was presented as regression coefficients (95% CI).

+p < 0.10, *p < 0.05.

Models 6 and 7 included Internet use variables and control variables in Table 3.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Associations of Diverse Internet Device Use and Activities with Depression in Chinese Adolescents: Gender and Geographical Differences

Status:

Version 1

Abstract

Background

Relationship between Internet use time and depression

The associations of diverse Internet device use and activities with depression: Differences in gender and geographical location

Gender differences

Geographical location differences

The current study

Methods

Survey procedure and participants

Measures

Dependent variables

Independent variables

Moderators

Control variables

Analysis

Results

Adolescent depression and diverse Internet device use and activities by gender and geographic location

Relationship between daily Internet use time and depression

The associations of diverse Internet device use and activities with depression

Gender differences

Geographic location differences

Discussion

Conclusions

Declarations

Ethics approval and consent to participate

Consent for publication

Availability of data and materials

Competing interests

Funding

Authors' contributions

Acknowledgements

References

Tables

Additional Declarations

Status:

Version 1