On the psychometric properties and genomic etiology of the general factor of psychopathology.

There is a high degree of comorbidity in psychiatry. This implies that there might be a set of underlying common factors that influence multiple conditions. A decade ago, researchers suggested that one way to estimate these is by applying hierarchical factor models, which partition that which is shared among all psychiatric conditions (commonly labeled general psychopathology) from that which is shared by subsets of conditions (commonly labeled specific factors). Whereas the specific factors were relatively uncontroversial, the general factor has been the subject of debate. On the one hand, the general factor appears to predict clinically relevant outcomes; on the other hand, critics have questioned its psychometric foundations and genomic etiolology. In this review, I present evidence that the general factor appears to have sound psychometric properties, and that it is at least possible that it could have a genomic etiology. With the caveat that it is very difficult to identify the data generating processes underlying multivariate data, I end by offering a cautious amendment to the suggestion that the general factor might measure impulsive reactions to emotions.