Comment by WhompingWindows

These studies control for such variables, confounding effects are easy to think of, but 9/10 times here on HN, the authors DID control and adjust for that confounder, the commenters just didn't read the methods.