2 ms·

The article links to this Lancet study, which does start to go beyond just correlation. https://www.thelancet.com/article/S2215-0366(19)30048-3/fulltext https:

• ejstronge 3y ago
  It is a compelling study, but retrospective trials, in general, aren't the correct design to answer this kind of question. I am not saying the results aren't accurate - just that in medicine, there have been instances where a prospective trial offers a different conclusion that that of prior, retrospective trials. Next, I'm no statistician, but I have never heard of this kind of trial being proof of causality (it can only provide supportive data - but as above, we know that the results may prove incorrect in a stronger study design). There are a lot of confounding factors that make causality a difficult thing to prove in humans. There is actually a way to prove causality in humans, and I suspect someone will do this for marijuana since it is a hot-button issue: https://archive.cdc.gov/#/details?url=https://www.cdc.gov/genomics/events/precision_med_pop.htm https://archive.cdc.gov/#/details?url=https://www.cdc.gov/ge...