Maartje Boer

CHAPTER 6 180 life satisfaction within the same time frame, but this decrement in mental health may not persist for a year. In other words, mediating processes whereby adolescents’ SMU problems decrease mental health through upward social comparisons and cybervictimization may emerge within a shorter time frame. More longitudinal research, using shorter time intervals, is required to verify this suggestion. In addition, our results suggest that the negative association between SMU problems and mental health was unidirectional, thus that decreases in mental health did not lead to increases in SMU problems. In our additional analysis on gender differences, the unidirectional finding remained stable, suggesting that this accounted both for boys and girls. Hence, we did not find support for the cognitive behavioral model of addiction, which posits that pre-existing psychopathology drive the development of addiction- like internet-related behaviors (Davis, 2001; Griffiths, 2013). However, other longitudinal research among adults showed that, in line with the cognitive behavioral model, decreased life satisfaction predicted increased social media self-control failure (Du et al., 2021), which is also an element of SMU problems. An explanation for the seemingly contrasting findings might be that adults with poor mental health may be more sensitive to developing SMU problems than adolescents with mental health impairments. It has been proposed that while adults with mental health problems may engage in addiction-like SMU to alleviate their problems, adolescents with poor mental health may refrain from social media because SMU further deteriorates their mental health, as they may be more sensitive to comparing themselves with others on social media than adults (Ho et al., 2017). Alternatively, the conclusion that mental health problems do not underlie adolescents’ SMU problems may be premature. Although the cognitive behavioral model postulates that pre- existing psychopathology is a necessary condition for the development of addiction symptoms, pre-existing psychopathology does not by definition lead to developing addiction-like behavior (Davis, 2001). To that end, mental health problems may pose a risk for developing SMU problems for a specific group of adolescents, which is possibly not detected in our analysis on a heterogeneous sample. Therefore, more longitudinal research on the effect of poor mental health on SMU problems is warranted, focusing on potential moderators of the effect (e.g., social anxiety, personality traits).