A cross-sectional study was designed using date from a cohort study—the Generation R Study. The Generation R Study is an ongoing prospective cohort study from fetal life onward in Rotterdam, The Netherlands. Detailed information on the Generation R Study has been published elsewhere [28]. Briefly, all pregnant women with an expected delivery date from April 2002 through January 2006 living in Rotterdam were invited to participate. A total of 6842 children participated in the assessment at age 13 years (2015-2019). Children with missing data on the number of social media platforms used or the time spent on social media were excluded (n=2300), as were children with missing data on HRQoL (n=945). To avoid data clustering, second (n=192) and third children (n=8) of the same mother were excluded, leaving a study population of 3397 children (Figure 1). Written informed consent was obtained from all participants.

The Medical Ethics Committee of the Erasmus University Medical Center approved the study (MEC 217.595/2002/202).

Descriptive analyses were applied to characterize the study population. Distribution of the reported problems in the 5 dimensions of HRQoL by the number of social media platforms used and the time spent on social media were assessed by chi-square tests and shown in Multimedia Appendix 1. The VAS score by the number of social media platforms used and the time spent on social media were evaluated by 1-way analysis of variance and shown in Multimedia Appendix 1. The relationship between the number of social media platforms used and HRQoL was investigated using logistics regression for the 5 dimensions and linear regression for VAS score. All models were adjusted for the child’s age, sex, ethnic background, family composition, maternal educational level, and net household income and additionally adjusted for time spent on social media. The relationship between the time spent on social media and HRQoL was investigated using logistics regression for 5 dimensions and linear regression for VAS score and shown in Multimedia Appendix 2. All models were adjusted for the covariates. All models were run separately with weekday and weekend social media use.

Missing data on covariates were imputed using multiple imputation methods, and 10 imputed data sets were generated. Pooled effect estimates (odds ratios [ORs] and β coefficients) and confidence intervals from these 10 imputed data sets were reported, and P<.05 was used to indicate statistical significance. Statistical analyses were performed using SPSS for Windows (version 24.0, IBM Corp).

Children with missing data on HRQoL (n=945) were compared with children without missing data (n=3397) using chi-square tests. Data were more often missing for children from mothers with a low educational level, a low household income, or a 1-parent family (all P<.05). No statistical difference was found in the number of social media platforms used and time spent on social media between children with or without data on HRQoL.

To examine the robustness of our results, the relationship between the number of social media platforms used and HRQoL was examined using the continuous variable of the number of social media platforms used and shown in Multimedia Appendix 3.

Additionally, one variable-parental supervision was added as a confounder into the regression models and shown in Multimedia Appendices 4 and 5. This variable was obtained by a parent-reported questionnaire when the child was aged 13 years. Parents answered the question “I sit at the computer together with my child if they are on social media.” Response options were never, rarely, sometimes, often, and always.

In total, 3397 children were included in this study. The mean age of the children was 13.5 (SD 0.4) years; 46.9% (1594/3397) were boys. More than two-thirds (2466/3397, 72.6%) of the children used 3 or more social media platforms. Of these, 37.7% (1234/3276) of the children used social media more than 2 hours on a weekday and 58.3% (1911/3277) did on a weekend day. From all 5 dimensions in HRQoL, children reported the most problems in “having pain or discomfort” (876/3385, 25.9%), followed by “feeling worried, sad, or unhappy” (557/3388, 16.4%). The average VAS score was 83.5 (SD 14.8; Table 1).

Multimedia Appendix 1 shows the distribution of reported problems in the 5 dimensions and VAS score by the number of social media platforms used and the time spent on social media. Significant differences were found in “mobility,” “doing usual activities,” “having pain or discomfort,” and “feeling worried, sad or unhappy” dimensions between children who spent different time on social media (all P<.05). The VAS score decreased with more social media platforms used and with more time spent on social media.

Compared with children who used 0 to 2 platforms, those who used 5 to 6 platforms or 7 or more platforms were more likely to report having some or a lot of problems on “having pain or discomfort” dimension (5-6 platforms: OR 1.41, 95% CI 1.12 to 1.78; 7 or more platforms: OR 1.55, 95% CI 1.20 to 1.99). Compared with children who used 0 to 2 platforms, those who used more platforms were more likely to report having some or a lot of problems on “feeling worried, sad or unhappy” dimension (3-4 platforms: OR 1.42, 95% CI 1.09 to 1.85; 5-6 platforms: OR 1.61, 95% CI 1.25 to 2.08; 7 or more platforms: OR 1.99, 95% CI 1.52 to 2.60). Furthermore to using 7 or more social media platforms was related to lower VAS score (β –3.81, 95% CI –5.54 to –2.09; Table 2).

After adjusting for all variates, higher social media use both on weekdays and weekends were related to more problems reported in each dimension and lower VAS score (Multimedia Appendix 2). On a weekday, compared to children who used social media less than 30 minutes per day, children who used social media between 2 to 4 hours, 4 to 6 hours, or more than 6 hours were more likely to report having some or a lot of problems on 3 dimensions (“doing usual activities,” “having pain or discomfort” and “feeling worried, sad, or unhappy” dimensions; all P<.05). Of those 3 dimensions, children who used social media more than 6 hours reported the highest odds ratio of reporting problems on “doing usual activities” dimension compared to children who used social media less than 30 minutes per day (OR 4.00, 95% CI 1.84 to 8.71). On the “mobility” dimension, only children who used social media more than 6 hours reported significantly more problems compared to those who used social media less than 30 minutes (OR 2.21, 95% CI 1.05 to 4.65).

With regard to the VAS score, higher social media use was related to lower VAS score. Children who used social media more than 6 hours per day had the lowest VAS scores compared with children who used less than 30 minutes per day (β –7.70, 95% CI –11.82 to –3.59). The results of the relationship between the time spent on social media and HRQoL during weekends were comparable to the results of weekdays, although effect estimates (ORs and β) were smaller.

The results showed that using more social media platforms was related to reported problems on the “having pain or discomfort” (OR 1.10, 95% CI 1.04 to 1.17) and “feeling worried, sad or unhappy” (OR 1.11, 95% CI 1.03 to 1.18) dimensions and lower VAS score (β –0.42, 95% CI –0.81 to –0.03) when regarding the number of social media platforms used as the continuous variable (Multimedia Appendix 3). Additionally adjusting for the confounder parental supervision, all the results were comparable to previous analyses, although effect estimates (ORs and β) were smaller (Multimedia Appendices 4 and 5).

This study investigated the relationship between the number of social media platforms used, time spent on social media, and HRQoL in children aged 13 years. The findings show that 72.6% of the children aged 13 years used 3 or more social media platforms. Around 40% of the children used social media at least 2 hours per day during weekdays (37.7%) and over half of the children on weekends (58.3%). Our findings also show that using more social media platforms and spending more time on social media were related to lower HRQoL.

A higher number of social media platforms used was related to lower HRQoL, which is in line with our hypothesis. Children who used more social media platforms had a higher odds of reporting having some or a lot of problems on the 5 HRQoL dimensions and reported lower self-rated health than their counterparts. An increased number of social media platforms used may elevate the stress of meeting the expectations to check for updates and respond on time on several social media platforms [34]. Moreover, being bombarded with information and communication from numerous social media platforms may result in media multitasking problems [35]. Media multitasking is a specific type of media use behavior in which users simultaneously perform at least 2 media activities or frequently change from one media activity to another. The act may decrease children’s psychological well-being, self-esteem, and overall HRQoL [36].

However, it is important to note that decreasing the number of social media platforms used for children may be challenging in today’s world. Children may be reluctant to give up any platform because they use different platforms for different reasons [37]. For example, a child can have a Facebook account to keep in touch with friends, Pinterest for cooking, Twitter for news, and Instagram for blogging. Therefore, educational intervention may help adolescents, especially those beginning to use social media, to better understand which platforms are truly necessary to their lives and valuable for their goals and which ones may not be [38].

Our findings suggest that more social media use (on both weekdays and weekends) was related to a lower HRQoL, which is in line with the hypothesis. Children who spent more time on social media were more likely to report having some or a lot of problems on the 5 HRQoL dimensions, and report lower self-rated health than those who spent less time on social media. These findings are in line with previous studies demonstrating that social media use among children aged 13 years and older was inversely associated with HRQoL outcomes such as physical health status and psychological well-being [39,40].

Previous studies have shown that children who spend more time on social media might have a greater chance of being the victim of cyberbullying or direct attacks from others on their sense of well-being [41]. Sampasa et al [42] reported that social media use was associated with cyberbullying victimization and that the risk of being cyberbullied increased in a dose-response manner in children. Children who suffer from cyberbullying are more likely to have lower HRQoL [43]. On the other hand, studies also reported that children who had limited activities and/or high levels of feeling worried, sad or unhappy may perceive online social communication via social media as a possibility to gain social support and relieve themselves of negative feelings through the ease of online self-disclosure [44]. Future studies are recommended to study intermediate variables to better understand the pathway between adolescents’ social media use and HRQoL.

In this study, social media use was assessed with the number of social media platforms used and time spent on social media. Further, this study was conducted among children aged 13 years, a sensitive developmental period during which little is known about social media use and its impact on HRQoL. However, the following methodological considerations need to be taken into account. First, for our cross-sectional analyses data from children aged 13 years were used. During this measurement, data on time spent on social media and the type of social media used were collected. The next measurement was performed on child aged 18 years. We recommended follow-up studies to explore the longitudinal associations between social media use and HRQoL. Second, social media use was captured via self-reported questionnaire, which may lead to overestimating or underestimating the time spent on social media. More accurate measuring method such as passive sensing (eg, apps installed on the devices that track real-time use [45]), could be adopted in future studies.

This study captured the number of social media platforms used and time spent on social media among children aged 13 years. Children using more social media platforms and spending more time on social media were more likely to report a lower HRQoL, including reporting having some or a lot of problems on the 5 HRQoL dimensions and lower self-rated health. Better understanding of how social media platforms contribute to children’s health and well-being may contribute to more support in adolescent social media use.