Generally speaking you should never, ever change your behavior based on the results of a single study -- even a controlled, double-blind study, much less an epidemiological survey. You should wait for a comprehensive literature review paper in a high-impact peer reviewed journal before you consider a result reliable.
That said, correlation is still quite valuable -- to researchers. Science doesn't have the resources to come up with quick, definitive answers on a question like this, involving a complex system that is expensive and ethically tricky to monkey with. So science spends a lot of time doing safer, more affordable stuff like looking for epidemiological correlations, until it can justify spending a lot of rare research dollars on something more probative. And those dollars are about to get a lot rarer too.