I think this study...
...falls into the common trap of declaring a cause and an effect based on a study that only showed a correllation between the two properties.
Hard to tell how exactly the study was done from this article, so I may be wrong. Science News will probably clear up the mystery in a few days.