In Reply In our study,1 we used 3 waves of data from the longitudinal Population Assessment of Tobacco and Health study to show that adolescents who report more frequent social media use report greater symptoms of psychopathology 1 year later. In their letter, Keyes and Kreski replicated our findings. They then showed that, after further adjusting for psychopathology measured concurrently with social media use, the prospective association we identified disappears, except for among the most frequent users of social media use.