My gut instinct is that they reported a correlation, and the media ran with it and said it was causation. That is what happens 90% of the time anyway.
As in, they reported a finding as innocent as that children with little exposure to foods from other cultures then were more likely to not be exposed to other cultures and be racist. That does make sense; lack of exposure to something is often a reason to not like it.
The media twisted this into them saying that children who don't like spicy foods are racist, because that's a better headline and because most people don't know how science works, including media.