They were controlled for in the statistical analyses, however, residual confounding can not be ruled out. The data reported are for associations, however, associations can not establish causality.

Your name:*Required

Your email address:*Required

Pick a user name:*Required