Critical issues in ALS case-control studies: the case of the Euro-MOTOR study