Fig 1.
Directed acyclic graph demonstrating the relations between measured and unmeasured confounding variables, osteoarthritis, and cardiovascular disease.
Table 1.
Description of the study sample by exposure status and sex (N = 1,838,434)a.
Fig 2.
Main results.
Fig 3.
Cumulative incidence curves of cardiovascular disease for men and women with and without hip or knee osteoarthritis.