Establishing consistencies in mental illnesses is necessary for a better general understanding, diagnosis and treatment. The early efforts to systemize mental illness were based on common causes, or the similarity of prognoses, but these efforts were soon replaced by descriptions in terms of observable symptoms (Kihlstrom, 2002). Throughout the different editions and revisions of the DSM, characteristic sets of symptoms have remained the basis to categorize disorders and differentiate between them. In case of the DSM, the selection of relevant symptoms and the step from symptom sets to the classification of mental illnesses are based on expert consensus. However, even if a diagnosis is defined in terms of a required number out of a set of specific symptoms, such a definition does not imply that the underlying disorder is necessarily categorical. It is possible that the liability to develop a disorder is normally distributed, with higher values leading to an increasingly severe symptomatology (Falconer, 1981). Alternatively, it is possible that there are affected and unaffected groups in the population that are categorically distinct.
