Psychopathology symptoms exist on a continuum from health to impairment, suggesting the existence of dimensions rather than categories. Dimensional models are useful because they take into account subthreshold symptoms that might otherwise be disregarded in a case-control design. One avenue of our work focuses on measuring the dimensional rather than categorical nature of psychopathology symptoms.
Importance of Comorbidity
Case-control designs often compare a particular disorder to a control group while excluding those with comorbidities. However, given the high comorbidity that naturally exists between disorders, individuals with a single diagnosis are less common and may not be representative of the presentation of symptoms in the general population. We are interested in studying brain-behavior relationships in samples with naturally occurring comorbidities.
Research suggests that psychopathology symptoms may be better represented as a hierarchy, with a general psychopathology factor (p factor) representing the symptoms shared across disorders. Part of our research uses statistical modeling to identify the hierarchical classification of psychopathology symptoms.
Common and Dissociable Neurobiology
We also relate hierarchical classification models to neuroimaging measures to examine common and dissociable neurobiological indices of psychopathology. Such work shows that many disorders are more common than they are different in terms of the brain, bringing into question the conventional demarcation between categorical diagnoses.
There is also considerable heterogeneity within psychopathology in terms of both symptoms and neurobiology. Two individuals with the same diagnosis may have widely different symptom presentations and potentially different underlying neurobiology. We are interested in using machine-learning tools to better understand the neurobiological heterogeneity that exists within psychopathology.