Comment by Dan_-

Exactly. Being unable to account for this covariate (tobacco use) pretty much invalidates this analysis. The odds ratio for tobacco use is basically the same (3x).

Also, title needs a 2025.