Nonetheless, despite these limitations, the results are in accordance with previous research that has examined comorbidity using dimensional models in general community or twin samples [1, 2, 33]. In parallel with the models specified in these studies, distinct patterns of comorbidity were evident among individual participants in the current study, reflecting systematic coherency among fear-related disorders (social and specific phobia, agoraphobia, and to a lesser extent panic disorder), among disorders entailing high levels of distress and dysphoria (GAD, dysthymia, and major depression), and among disorders involving deficient impulse control (conduct disorder, alcohol dependence, and drug dependence). Notably, the fear, distress, and externalizing classes were each characterized by high levels of a core group of two to three disorders. This pattern of results indicates that these latent classes do not reflect subsets of individuals who differ simply in overall severity of mental illness, but rather distinct groups of individuals prone to differing combinations of disorders. Additionally, and more importantly, a class emerged that has not been identified in dimensional models—namely, the “multimorbid” class. Across both samples, this class emerged as the
