Bullying at School, Cyberbullying, and Loneliness: National Representative Study of Adolescents in Denmark

Loneliness is a subjective feeling of isolation. It is often defined as a cognitive discrepancy between the social relations an individual wishes to have and those that one perceives to have, and the affective reactions of sadness and emptiness that follow [1]. The feeling of loneliness is common in adolescence [2,3,4,5], and many adolescents will experience loneliness for short periods. The reasons may be feeling left out among peers, a change of school, parental divorce, or other adverse life events [6,7]. However, some adolescents experience prolonged feelings of loneliness that result from repeated failure to reconnect with others, which is a serious threat to their quality of life [6,7,8] and academic performance [9]. A recent meta-analysis of longitudinal studies suggested that loneliness tended to remain stable from adolescence to adulthood [10]. Loneliness is also an important public health problem because it is associated with a range of health problems [1,6,11,12,13,14] and risk behaviors [15,16,17]. It is important to understand the precursors of loneliness to strengthen preventive efforts. The current study focuses on two potential precursors: bullying victimization at school and exposure to cyberbullying.

Bullying victimization at school is common among adolescents [2,18,19], although the prevalence has been diminishing over the past decades in Europe and North America [20,21]. There is abundant documentation for an association between exposure to bullying and adverse psychological consequences such as poor life satisfaction [18], mental health problems, and suicidal behavior [5,22,23,24,25,26,27,28,29]. A small number of cross-sectional [30,31,32,33,34,35] and prospective [35,36,37] studies confirm that there is an association between loneliness and bullying victimization at school. For instance, Due et al. [30] found a strong and graded association between loneliness and exposure to bullying at school, a finding which was consistent across twenty-eight countries. The effect sizes vary across studies, from small to large. The variation in effect sizes suggests a need for further studies.

Exposure to cyberbullying, sometimes labeled internet bullying, online victimization, or internet harassment, is the use of digital technologies to harass, threaten, embarrass, or target another person. This phenomenon is also common among adolescents [2,18,21,24,38,39], although the prevalence of exposure to cyberbullying is lower than exposure to bullying at school [21]. The studies that find an association between exposure to cyberbullying and loneliness [40,41,42] show considerable variations in effect sizes, from weak to strong. There is some doubt about the causal pathway since a few prospective studies show that loneliness is a precursor of cyberbullying rather than the reverse [43]. As with face-to-face, in-person bullying victimization, the variation in effect sizes across studies for the association between loneliness and cyberbullying highlights the need for further studies.

There are reasons why there might be differences in the associations between loneliness and face-to-face, in-person bullying and loneliness and cyberbullying. For example, Van den Eijnden et al., 2014 [44], emphasize that findings from research on bullying at school cannot automatically be transferred to cyberbullying because these two phenomena differ in important ways: Cyberbullying via the internet has much higher accessibility to the target than bullying during school hours. Cyberbullying can reach a much larger audience than bullying at school and may remain visible for a long time to the victim and the audience, potentially resulting in longer-lasting negative effects, e.g., loneliness. A study covering six North European countries showed little overlap between bullying at school and cyberbullying, suggesting that the two may be different phenomena [18]. It is, therefore, important to analyze which kind of exposure is more closely associated with loneliness and to analyze the association between loneliness and double exposure (to bullying at school and cyberbullying). Only a few studies focus on such combined effects, and they found higher rates of loneliness among adolescents who were exposed to bullying in both contexts [31]. Van den Eijnden et al. [44] suggest that the two kinds of bullying are mutually reinforcing, i.e., exposure to bullying in one context increases the risk of exposure in the other. Studies about the association between exposure to bullying and loneliness use different reference periods, which makes comparisons difficult [40]. A few studies suggest that the association between loneliness and exposure to bullying varies by sex and age group [33,43,45]. According to Cava et al. [45], older teenage girls might be more vulnerable to cyberbullying than boys, for instance when exposed to cyber-control from a romantic relationship. These girls reported more feelings of loneliness and assessed their social network as worse than those never victimized. Landstedt and Persson [24] suggest more focus on the gender issue.

The overall prevalence of loneliness was 9.0%. Table 1 shows that loneliness was significantly more prevalent among girls vs. boys, 15-year-olds vs. 11-year-olds, immigrants vs. Danish origin, and students from lower vs. higher OSC. The proportion exposed to bullying at school at least a couple of times per month was 6.3%, significantly more prevalent among girls vs. boys, 11-year-olds vs. 15-year-olds, immigrants vs. Danish origin, and students from lower vs. higher OSC. The proportion exposed to cyberbullying at least a couple of times per month was 4.8% and not significantly related to any of the socio-demographic variables. The variable that combined exposure to habitual bullying at school and cyberbullying classified the students into four categories: 4867 (90.4%) were not exposed to any bullying, 175 (3.3%) were exposed to cyberbullying but not bullying at school, 259 (4.8%) were exposed to bullying at school but not cyberbullying, and 81 (1.5%) were exposed to both kinds of bullying. These figures suggest that there is little overlap between exposure to the two types of bullying since most of the students who were exposed to bullying in one context were unexposed in the other context.

Table 2 shows a strong and graded association between loneliness and exposure to bullying. Even among students with low exposure to bullying (once or twice in the past couple of months), the odds ratio (OR) for loneliness was significantly elevated compared to non-exposed students. The OR (95% CI) for loneliness was 11.58 (7.21–18.61) among students exposed to bullying at school several times a week. The corresponding figure for exposure to cyberbullying was 5.79 (3.37–9.86). Finally, the OR for loneliness among the few students exposed to both kinds of bullying was 10.80 (6.87–16.97). Table 2 also shows that the OR estimates changed little when adjusted for socio-demographic control variables. Separate analyses for boys and girls and the three age groups showed a significant association between loneliness and the three measures of bullying in every sub-group (not shown in the table). The association between exposure to bullying at school and loneliness was steeper for boys than girls, manifested via a statistically significant interaction term between sex and exposure to bullying at school, p = 0.0165.