Objectives: The DSM-5 Level 1 Cross-Cutting Symptom Measure (DSM-XC) is a transdiagnostic mental health symptom measure that has shown promise in informing clinical diagnostic evaluations and as a screening tool for research. However, few studies have assessed the latent dimensionality of the DSM-XC. We examined the factor structure of the DSM-XC in a large convenience sample of participants with varying degrees of psychological health.
Methods: Participants (n=3533) enrolled in a protocol conducted at the National Institute of Mental Health (NCT04339790). We used a factor analytic framework to evaluate an existing two-factor solution (Lace & Merz, 2020) and two additional candidate solutions.
Results: The Lace and Merz solution had acceptable fit. Exploratory factor analysis yielded two candidate solutions: a six-factor (characterized as mood, worry, activation, somatic, thoughts, and substance use) and a bifactor (general factor of non-specific psychopathology, residual factors characterized as internalizing and thought disorder), which both had good fit and full measurement invariance across age, sex, and enrollment date.
Conclusions: Our findings confirm that the DSM-XC may be conceptualized as a multidimensional instrument and provide a scoring solution for researchers who wish to measure distinct constructs. Future research on the psychometric profile of the DSM-XC is needed, focused on the validity of these candidate solutions and their performance across research populations and settings.
Keywords: COVID-19; clinical research; factor analysis; mental health; transdiagnostic.