It's a garbage study. The confidence interval is a literal order of magnitude. And the data is a 300 person self-reported questionaire asking questions that can go back decades. Do you remember how many diet cokes you drank back in the 80s?
It's a shame this kind of science is allowed publication.