That's always the case with study vs experiment. If you didn't control for all the variables but one, you can't make a conclusion about cause.

I think the honest thing to do is to include alternate hypotheses; or at least mention some possible confounding circumstances.