Bullying victimization, physical inactivity and sedentary behavior among children and adolescents: a meta-analysis

Background Physical activity and sedentary behavior are related with psycho-social variables among youth, however its relationship with bullying victimization is unclear. The aim of the study was to clarify the associations between physical activity and sedentary behaviors with bullying victimization among children and adolescents. Methods Two independent authors searched in four databases. The studies were selected/included only if participants were children and/or adolescents and the relationship between physical activity and/or sedentary behavior with bullying victimization was reported. Random-effects meta-analyses were used. Results A total of 18 cross-sectional studies (including 386,740 children and adolescents, 51.8% females) were reviewed. Our study found that not meeting the physical activity guidelines (Odds Ratio [OR] = 1.14, 95% confidence interval [CI], 1.04 to 1.23) and excessive sedentary behavior (i.e., 2 h per day or more of screen time) (OR = 1.21, 95% CI, 1.14 to 1.28) were associated with 14 and 21% higher bullying victimization, respectively. Consistent associations were also found when we analyzed specific forms of bullying for sedentary behavior, including traditional and cyberbullying. Conclusions The present study establishes the first quantitative framework for understanding the influence of physical activity and sedentary behavior on bullying victimization, and lays the groundwork for future studies and interventions aimed to its promotion. Trial registration CRD42018099388.


Introduction
Bullying victimization denotes to the process by which an individual is repeatedly and over time exposed to intentional harmful or negative actions by their counterparts [1]. Bullying can occur in different environments, including but not limited to workplace, community settings, schools and home [2]. Bullying victimization is prevalent across countries worldwide, reaching prevalence rates up to 35% for traditional bullying and 15% for cyberbullying [3]. According to the Centers for Disease Control and Prevention (CDC), bullied youths frequently report high levels of poor school performance, sleep difficulties, loneliness, anxiety, depression [4] and are more likely to commit suicide [5].
Physical inactivity among youth predicts a wide range of health problems that are detrimental to well-being [6]. International guidelines from organizations such as the US Department of Health and Human Services [7] and the World Health Organization [8] recommend that children and adolescents should have at least 1 h or more per day of moderate-to-vigorous physical activity. There is evidence that suggests that, when a child or adolescent spends a large amount of time in a sedentary behavior it relates to a poorer health outcome, being somewhat stronger for television (TV) viewing and other screen behaviors than for total sedentary time [9]. In recent years, the rapid adoption of digital media has displaced the consumption of legacy media such as reading a book or a magazine [10]. The American Academy of Pediatrics recommends no more than 2 h per day of any screen-based activity [11]. In this regard, there is evidence that has shown that spending an excessive amount of time in screen-based activities can have a direct relationship to the development of anti-social behavior and aggression [12] and more negative feelings [13]. Therefore, it is important to understand how bullying status relates to whether youth meet or do not meet the physical activity and sedentary behavior recommendations. Therefore, the purpose of this study was to provide a quantitative analysis on the associations of physical activity and sedentary behavior on bullying victimization among children and adolescents.

Methods
A systematic review and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [14]. The review was registered in PROSPERO (registration number: CRD42018099388).

Data sources and searches
Two independent authors (IH-A & AG-H) searched PubMed, EMBASE, ERIC and PsycARTICLES, from database inception to May 20th, 2020. Studies were identified by using all possible combinations of the following groups of search terms: (a) "bull" OR "victim" OR "peer relation"; (b) "physical activity" OR "exercise"; and (c) "sedentary" OR "screen time" OR "television" OR "video game" OR "computer". The complete search strategy is shown in Additional file. Only English articles were included. In addition, the reference lists and related links of retrieved articles were examined to detect studies potentially eligible for inclusion.

Study selection
Studies needed to meet the following criteria: (a) subjects: children and adolescents aged 6 up to 18 years old; (b) type of study: cross-sectional and prospective cohort studies; (c) exposure: objective or subjective measured of physical activity and/or sedentary behavior; and (d) outcomes: bullying victimization as dependent variable. The first and second author (IH-A & AG-H) independently assessed the electronic search results. Reasons for exclusion of identified studies were recorded.

Data extraction
The following information was extracted from the studies meeting the selection criteria: country of study, participants (e.g., sex, age), type of bullying (i.e., traditional or cyberbullying), physical activity and sedentary behavior assessment of each study and study results.

Risk of bias in individual studies
The risk of bias was assessed by The Quality Assessment Tool for Observational Cohort and Cross-sectional Studies [15]. This methodological tool is composed of fourteen items rated as "yes", "no" or "not reported".

Data synthesis and analysis
Odds Ratio (OR) and 95% confidence interval (CI) were pooled using random effect models (DerSimonian and Laird) [16] to account for anticipated between-study heterogeneity. This heterogeneity (Cochran's Q-statistic) was estimated using I 2 , considering I 2 values of < 25%, 25-50, and > 50% as small, medium, and large amounts of heterogeneity respectively [17].
Publication bias was determined by visual examination of funnel plots (only used for overall adiposity). Also, small-study effects bias was assessed using Egger's test [18].
Finally, subgroup moderator analyses were conducted to determine whether results differed according to sex, specifics form of bullying and type of sedentary behavior.
All analyses were carried out using the STATA 13.1 Software (Stata Corporation LP, College Station, TX, USA).

Literature search
The electronic search strategy identified 521 studies and, after screening for duplicates, 43 full-text studies were assessed for inclusion after checking titles and abstracts. Finally, 18 studies met the inclusion criteria and were included in the systematic review and 14 in the metaanalysis. The flow diagram showing the number of articles excluded at each stage of the systematic review and meta-analysis is shown in Fig. 1. Table 1 summarizes the characteristics of the 18 included studies. All of them were cross-sectional observational studies, and were published from 2007 to 2020. The studies included 386,740 children and adolescents.
All included studies used a self-reported questionnaire to assess physical activity and sedentary behavior, except two which assessed physical activity with accelerometers [22,36]. Most studies defined physical inactivity as less than 60 min of moderate to vigorous-intensity physical activity per day on at least 7 days per week [19,21,23,24,26,34]. Regarding sedentary behavior, studies analyzed this behavior as hours per day of television, computer/ video game, and screen time use or sitting time; also, most studies defined excessive sedentary behavior as at least 3 h average per school day.
School bullying victimization was assessed using a questionnaire. While a limited number of the included studies used a validated bullying questionnaire [20,22,25,29,31,32,35], the majority of the studies used a single item to assess bullying (e.g., "During the past 12 months, have you ever been bullied on school property?"). Moreover, most studies focused on traditional bullying victimization (i.e. referring to physical abuse behaviors, verbal or weight-teasing, psychological abuse and social exclusion) and five in cyberbullying (i.e. texting, emails, social network sites) [22,26,30,33,34].

Risk of bias within studies
All 18 studies met at least 6 criteria and were considered to have moderate methodological quality. The average score was 7.5/14.0 (Table 1 and Additional file 1).

Meta-analysis of the association between physical activity and bullying victimization
Overall, there was evidence that not meeting the current physical activity guidelines was associated with higher bullying victimization (OR = 1.14, 95% CI, 1.04-1.23; p < 0.001; I 2 = 73.8%) (Fig. 2).

Sensitivity analysis and publication bias
When the impact of individual studies was examined by removing studies from the analysis one at a time, we observed that the pooled results estimate remained consistent. Both funnel plot asymmetry (Additional file 1) and Egger test showed no significant publication bias for physical activity (bias = 1.99; 95%CI, 0.38 to 3.60; p = 0.018) and sedentary behavior (bias = − 1.14; 95%CI, − 2.89 to 0.61; p = 0.194), indicating no evidence of publication bias.

Discussion
To the best of our knowledge, the current review is the first to assess the associations of physical activity and sedentary behavior with bullying victimization among children and adolescents. Our study found that not meeting the physical activity guidelines and excessive  sedentary behavior were associated with 14 and 21% higher bullying victimization, respectively. When looking at specific forms of bullying, we also found a consistent association between sedentary behavior with both traditional and cyberbullying.

Physical activity and bullying victimization
Engaging in 1 h or more per day of moderate-tovigorous physical activity is normally recommended for better health and quality of life among youths [7,8]. The Physical Activity Guidelines for Americans recently suggested that meeting the physical activity recommendations is vital to the physical, psychological/social, and cognitive health of school-aged children and adolescents [7]. Our findings suggest that not meeting physical activity guidelines was associated with 14% higher bullying victimization among children and adolescents. This finding supports the work of other studies in this area [38,39]. For example, a recent randomized controlled trial seems to corroborate the role of physical activity on bullying victimization among Chilean children [39]. Several factors may underpin the relationship between physical inactivity and bullying victimization: (a) bullying victimization occurs frequently in activities not closely supervised and therefore youth tend to avoid these activities [40,41]; and (b) physically inactive youth may be at   an increased risk of being bullied due to factors such as poor motor skills [42], physical fitness [43], and low selfconfidence to engage in physical activities.

Sedentary behavior and bullying victimization
Among children and adolescents, television viewing and other screen-based forms of entertainment are the most prevalent leisure-time sedentary behaviors [44]. Recent studies of electronic media use and TV viewing have concluded that this type of sedentary behavior (i.e. screen-based) is associated with lower pro-social behavior, self-esteem [45] and subjective well-being [13] in children and adolescents. In this regard, our metaanalysis reveals that excessive sedentary behavior was related with 21% higher bullying victimization, and found that this relationship is consistent across both, traditional (22%) and cyberbullying (20%). Increased screenbased activities use may diminish experiences of personal interaction that helps youth develop interpersonal relationships and academic skills, pro-social behavior and conflict resolution [46], which could place youths at risk of multiple forms of victimization in relationships with dating partners, friends, and peers [47]. Furthermore, studies suggest that friendships can function as a protective factor against the negative adjustment often experienced by bullied youth [47]. In contrast, screen-based activities favor loneliness [12]. For example, more screen time such as play video game seems to compromise youth development of interpersonal skills hence making them vulnerable to all forms of bullying [33]. Regarding electronic bullying victimization, both male and female youths could be exposed to cyberbullying through social media consuming during computer use [33]. Socio-economic factors seem to be related with bullying prevalence. For example, it has been suggested that adolescents from lower social classes are more exposed to victimization compared to adolescents from high socio-economic status [24]. In this regard, it is interesting to highlight findings from Alfonso-Rosa et al. [19] who indicated that the odds of being bullied were lower for adolescents who exceeded sitting guidelines in high Human Development Index countries compared to those in low Human Development Index countries.

Limitations
There are several limitations to this meta-analysis that should be acknowledged. First, our results are based on cross-sectional studies and so causality cannot be inferred. Reverse causality could also be true, i.e., youths bullied could be less prone to follow healthy lifestyle habits [48,49]. Second, all of the studies except two [22,36] did not include a rigorous physical activity and sedentary behavior assessment. Therefore, self-reported questionnaires could potentially be subject to socially desirable reporting bias. Third, only five studies used validated bullying questionnaires [20,22,29,31,32,35]. In this sense, outcome measures for bullying are not standardized across the studies performed in different countries with likely variable cultural norms. Also, studies used different time frames to determine bullying (i.e. past month, past couple of months, current or past school year). Thus, the discrepancy in definitions might explain the heterogeneity of our results. Finally, physical activity and sedentary behavior were dichotomized, therefore, the current study cannot determine if the association varies across different levels of physical activity or sedentary intensity and participation.

Conclusions
This meta-analysis establishes the first quantitative framework for understanding the influence of physical activity on bullying victimization, showing that not meeting the physical activity guidelines and excessive sedentary behavior are associated with higher bullying victimization. Furthermore, it lays the groundwork for future studies and interventions aimed to bullying victimization prevention through promoting physical activity and reducing sedentary behaviors.
Additional file 1. Table. Items of Quality Assessment Tool for Observational Cohort and Cross-sectional studies. Figure. Funnel plot for physical activity on bullying victimization, with 95% confidence limits. Figure. Funnel plot for sedentary behavior on bullying victimization, with 95% confidence limits.
Abbreviations CDC: Centers for disease control and prevention; CI: Confidence interval; TV: Television; OR: Odds ratio