That's because 99% of the time when a study says "we found an effect in Y even after we controlled for X" it just means that they included X in their multi-variate regression and error analysis in their study of Y. "controlled for" sounds like it has something to do with ruling out the effect of something from causality, but it doesn't.