Should substance use disorders be considered as categorical or dimensional?

Area of Science:

• Psychometrics
• Statistical modeling
• Mental health diagnostics

Background:

• Conventional statistical models for diagnostic criteria, such as latent class analysis and item response theory (IRT) analysis, have limitations.
• Representing diagnostic criteria effectively is crucial for accurate mental disorder classification.

Purpose of the Study:

• To introduce and evaluate novel hybrid statistical models for representing diagnostic criteria.
• To compare the performance of these new models against conventional approaches.

Main Methods:

• Development of hybrid mixture models integrating both categorical and dimensional representations.
• Application and comparison of conventional (latent class, IRT) and hybrid models using National Epidemiologic Survey on Alcohol and Related Conditions data for DSM-IV alcohol use disorders.
• Comparison of hybrid model classifications with the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criterion count method.

Main Results:

• Hybrid mixture models demonstrate superior suitability for analyzing diagnostic criteria compared to latent class and IRT models.
• The new models offer a more nuanced representation of diagnostic criteria.

Conclusions:

• Hybrid models provide a valuable framework for representing diagnostic criteria, offering both categorical and dimensional insights.
• Findings have implications for future diagnostic systems like DSM-V, suggesting integrated reporting of categorical and dimensional results.