Even if NF1 kids are more likely to have pre-existing coeliac disease identified, wouldn't that just push up the rate of identification, rather than changing the interaction between the diseases? So having one of the conditions is independent of having the other, but having one diagnosed is not independent of having the other?