I don't buy this "rigorous" experiment. So from 37 self identified gluten sensitive patients most of them have nocebo effect? Great. It clearly doesn't prove that gluten sensitivity doesn't exist, it proves that if you pick people from the self identified group than most of them are pretending (also can't actual sensitives have nocebo effects?). Surprise, surprise, it's already predicted that only 1% of the local population is gluten sensitive yet 18% buys gluten free products, probably many of them self identified gluten sensitives.