Well at least it was stated early in the article that "The results don't prove there's a connection". Making causal assertions based on observational studies th

Well at least it was stated early in the article that "The results don't prove there's a connection". Making causal assertions based on observational studies that are so wide open and uncontrolled might soothe our sensitivies but usually they are just appealing to whatever bias we already have.