Information theory methods for quantifying diagnostic heterogeneity in psychopathology

Abstract

Objectives: Although specifiers for a major depressive disorder (MDE) are supposed to reduce diagnostic heterogeneity, recent literature challenges the idea that the atypical and melancholic features identify more homogenous or coherent subgroups. We attempt to replicate these findings and explore whether symptom heterogeneity is reduced in depression subgroups using novel data-analytic techniques. Methods: Using data derived from the National Epidemiological Survey on Alcohol and Related Conditions (NESARC Wave I; N = 5,749) and Sequenced Treatment Alternatives to Relieve Depression (STAR*D; N = 2,498) we computed the Hamming and Manhattan distance ratios comparing within and between individuals for the melancholic and atypical specifier subgroups. Results: In neither of the datasets was the heterogeneity between subgroups higher than the heterogeneity within subgroups, suggesting that the melancholic and atypical specifiers do not create more coherent (i.e., more homogeneous) subgroups. Conclusion: Replicating prior work, melancholic and atypical depression subtypes appear to have limited utility in reducing heterogeneity. The current study does not support the claim that symptom and course specifiers create more coherent subgroups as operationalized by similarity in symptoms and their severity.

Rethinking trauma‐related psychopathology in the Hierarchical Taxonomy of Psychopathology (HiTOP)

“…However, subtyping within a criterion set necessarily limits the breadth of potential syndromes to those anchored by current PTSD criteria. Moreover, although subtypes and specifiers are an appealing solution to within‐diagnosis heterogeneity, they tend to be unsuccessful in actually resolving these issues (Buss et al., 2023; Watson, 2003).…”

To fully leverage fine-grained clinical phenomena, we have to think beyond DSM-based concepts and the presumption of diagnostic kinds.

“…This practice can also obscure individual differences in dynamic, within-person processes, which may have little or nothing to do with diagnostic group status (Kullar et al, 2023). Unless a purported syndrome can be justified as a discrete and valid entity, efforts to resolve within-disorder heterogeneity via subtyping are, therefore, unlikely to be successful in their stated aim (Buss et al, 2023; Watson, 2003). The use of diagnostic kinds as the foundation for explanation ultimately overstates the stability and distinctiveness of group status and causal structure and perpetuates existing reification of diagnostic categories rather than reimagining psychopathology nosology beyond DSM (Berenbaum, 2013; Böhnke & Croudace, 2015).…”