Such studies can prove only that the correlation falls below a certain threshold---it's in the nature of statistical analysis that they can never prove "no link."
as well as the studies which have analyzed the research which allegedly found a link, and spelled out exactly how their methodology was flawed.
I'd love to. Have any links?