About the linked study, it's important to emphasize that it's retrospective -- it tracks the behavior of people who applied for and received opioid prescriptions. This means the stated results may not apply to a person not inclined to use opioids.

A hypothetical prospective study, obviously impractical on multiple grounds -- one that chooses study subjects from the population at random -- might produce a completely different result.

Or the outcome might be the same -- for ethical reasons that deserve to exist, we may never know.