There may be a correlation vs. causation mix-up involved here. Look up the "self-medication hypothesis" (or variants). It posits that many people, with pre-existing psychotic tendencies, are self-medicating with marijuana. IOW, the psychosis is the cause, not the effect of the marijuana usage. Pot might actually be reducing the incidences of psychotic episodes.
The full story is not known yet here.
What seems to be happening is that kids who are predisposed by genetic factors to become psychotic use marijuana, which then triggers the psychosis to manifest. It could be the latent psychotic disorder that makes them more prone to use. What researchers are seeing is that the marijuana users have earlier and more severe manifestations of psychotic disorders. For kids who are at risk of psychosis (based on family history), it would be better for them to avoid any marijuana use, and for treatment to address the underlying genetic disorder, perhaps even to prevent it from ever manifesting.
Once the brain is fully developed (about age 25), there is no more risk of developing a psychosis related to marijuana use. The risk is highest in teenagers.