>especially increased risk of psychotic symptoms [1]
AFAIK all studies on psychosis and marijuana have been cohort studies, which are essentially useless at determining the direction of causation. While the study you linked attempts to address the possibility of "reverse causality", it only addresses one specific hypothesis (self medication).
Let us repeat this to ourselves until we all remember: You cannot establish direction of causation with a cohort study.
AFAIK all studies on psychosis and marijuana have been cohort studies, which are essentially useless at determining the direction of causation. While the study you linked attempts to address the possibility of "reverse causality", it only addresses one specific hypothesis (self medication).
Let us repeat this to ourselves until we all remember: You cannot establish direction of causation with a cohort study.