Identifying Transdiagnostic Mechanisms in Mental Health Using Computational Factor Modeling

Area of Science:

• Neuroscience
• Psychiatry
• Computational modeling

Background:

• Psychiatric disorders frequently co-occur, and symptom dimensions are not confined to single diagnostic categories.
• Current treatment development and allocation strategies, heavily reliant on diagnostic categories, result in suboptimal outcomes for 25-40% of individuals.

Purpose of the Study:

• To review efforts in characterizing psychiatric symptom dimensions using a dimensional, mechanistic approach.
• To highlight the utility of computational factor modeling in understanding transdiagnostic symptom dimensions and their underlying neurocognitive mechanisms.

Main Methods:

• Leveraging large-scale, remote, online, and citizen science studies to collect data on unselected samples.
• Employing computational factor modeling to formally specify, test, and falsify candidate mechanisms underlying transdiagnostic symptom dimensions.
• Validating identified symptom dimensions against computationally defined neurocognitive processes.

Main Results:

• Computational factor modeling has identified specific associations between cognitive processes (e.g., planning, metacognition) and transdiagnostic symptom dimensions.
• This approach has revealed previously obscured relationships and demonstrated generalizability to smaller clinical and nonclinical samples.
• The method is adaptable and being optimized for broader application in psychiatric research.

Conclusions:

• Computational factor modeling provides a powerful framework for understanding the dimensional and mechanistic underpinnings of psychiatric symptoms.
• Bridging basic research with clinical application through direct investigations of treatment response is a critical next step.
• This dimensional approach holds promise for improving treatment development and allocation in psychiatry.