Network Analysis of DSM Symptoms of Substance Use Disorders and Frequently Co-Occurring Mental Disorders in Patients with Substance Use Disorder Who Seek Treatment.

Abstract

Background: Substance use disorders (SUD) often co-occur with other psychiatric conditions. Research on SUD and comorbid disorders generally flows from a categorical diagnostic or dimensional latent variable perspective, where symptoms are viewed as independent indicators of an underlying disorder. In contrast, the current study took a network analysis perspective to examine the relationships between DSM symptoms of SUD, ADHD, conduct disorder (CD), depression (MDD), and borderline personality disorder (BPD). In addition, we explored possible gender differences in the network structures of these symptoms. Method: In a sample of 722 adult treatment-seeking patients with SUD from the International ADHD in Substance Use Disorders Prevalence Study (IASP) we estimated the network structure for 41 symptoms of SUD, ADHD, CD, MDD, and BPD. We described the structure of symptom networks and their characteristics for the total sample, and we compared the symptom networks for males and females. Results: Network analyses identified seven clusters of symptoms, largely corresponding with the DSM diagnostic categories. There were some connections between clusters, mainly between some hyperactivity symptoms and CD and depressive symptoms. ADHD hyperactivity was most central in the symptom network. Invariance tests revealed no significant gender differences in the structure of symptom networks. Conclusions: The current findings support the categorical DSM classification of mental disorders in treatment-seeking patients with SUD. Future network analyses should include a broader range of symptoms and prospectively explore changes in the symptoms network of patients during treatment.