There was no comparison to a similarly depressed or anxious group who were not prescribed medication.
Also, due to the "criminalization of mental illness," many children/adolescents come to the attention of the mental health system due to violent behavior, so there is a strong selection bias for those who are aggressive to enter treatment vs those who suffer in silence. Further, depression/anxiety are also risk factors for drug/alcohol use, which are well established as contributors to violence, and those don't appear to be controlled for with a comparison group either.
It's possible (though in my view unlikely) that SSRIs could meaningfully increase the risk of violence, but this is not a study that can make that conclusion.
Source: I'm an MD, board certified in general, child, and forensic psychiatry who works with mentally ill offenders.