While not all correlation is causation, all causation, if it exists, creates correlation.
Thus, correlation is a good flag to indicate where to investigate further.
If you proposed a clinical trial, today, to induce temporary sterility in teenaged females by giving them synthetic hormones without measuring pituitary function, GI function, and mental/emotional state over say, ten years, you would probably be arrested.
Such a trial would, of course, never be allowed.