Maartje Boer

CHAPTER 6 182 insight into potential long-term effects over time, such time intervals also have drawbacks. More specifically, behaviors may influence each other within a shorter time frame (Orben, 2020a), which could not be captured with the research design of the present study. Therefore, more longitudinal research replicating our study using more intensive longitudinal data, such as daily measures of SMU and mental health, is considered as an important direction for future research. Fourth, as is typical for longitudinal studies, the present study dealt with considerable dropout of participants. Although the dropout in the present study was not selective and the associations between our study variables and dropout were small, dropout remained a limitation. We aimed to limit the bias that is typically associated with dropout by retaining all adolescents in the analyses. Conclusion Most adolescents spend considerable time on social media, which raises concerns among many. Findings of the present study emphasize the importance of considering SMU intensity and SMU problems as two distinct behaviors, because our results suggest that particularly SMU problems pose a risk toadolescents’mental health. The reversepatternwas not observed, which suggests that poor mental health does not lead to developing SMU problems. Moreover, SMU problems were found to increase upward social comparisons and cybervictimization, which implies that adolescents with SMU problems face multiple adversities. High SMU intensity and mental health were not associated in any direction. The findings imply that policies and guidelines aimed at identifying, preventing, or informing about unhealthy SMU should focus on SMU problems rather than on high SMU intensity. More longitudinal research replicating our findings is warranted and may advance our insight into the origins of SMU problems as well as the underlying mechanisms explaining the link between SMU problems and low mental health.