Depression, suicidality leading to substance (ab)use also seems like a plausible direction for causality to flow. It's striking how the specific wording of the headline can cause you to make assumptions even if it is technically accurate (the assumption being that recreational use causes depression).
My first thought on reading the headline was that I suspected the arrow was backwards. Self medication for mental health issues is common, not seeing a relationship would be suspect!
Published study here, looks like source data is survey: https://jamanetwork.com/journals/jamanetworkopen/fullarticle...