Maartje Boer

SMU AND MENTAL HEALTH 179 6 which group of adolescents high SMU intensity turns into developing SMU problems, is considered as an important direction for future research. At the same time, high SMU intensity may also be beneficial, as our findings showed that adolescents whose SMU intensity increased reported increased face-to-face peer contact within the same year (in T1 and T2). Although we cannot derive directionality from this correlation, the finding refutes the idea that time spent on social media replaces time spent with friends offline, as frequently proposed (Twenge, Joiner, et al., 2018; Twenge & Campbell, 2018; Underwood & Ehrenreich, 2017; Wallsten, 2013). In line with our finding, other researchers also reported a positive association between SMU intensity and offline social interaction with friends or perceived friends support (Boer, Van den Eijnden, et al., 2020; De Looze et al., 2019; Valkenburg & Peter, 2007). High SMU intensity may be used to maintain contact with existing friends and may thereby be indicative of social involvement with peers, rather than neglecting friendships. Furthermore, research has shown that the more adolescents socialize with peers on social network sites, the less lonely they feel, which supports the idea that SMU may be used to strengthen and maintain friendships (Apaolaza et al., 2013). In contrast, findings of the present study underline the potential harmful effect of SMU problems to adolescents’ mental health. Moreover, SMU problems predicted increased levels of upward social comparisons and cybervictimization over time. These increases, in turn, did not decrease mental health one year later, which implies that upward social comparison and cybervictimization did not mediate the negative effect of SMU problems on mental health. However, adolescents’ increases in upward social comparisons and cybervictimization co-occurred with decreases in mental health within the same year. Therefore, and because of previously found effects of social comparisons and cybervictimization on mental health (Feinstein et al., 2013; Roeder et al., 2016), there may have been a mediating effect, but the measurements were possibly too far apart to observe it. For example, research shows that while adolescents experience increased emotional arousal shortly after posting on Facebook, this effect does not persist in the long run (Bayer et al., 2018). Correspondingly, when adolescents experience cybervictimization on social media or increased levels of upward social comparisons due to viewing social media content, they may experience more depressive symptoms or less