J Psychopathol Clin Sci. 2025 Oct 6. doi: 10.1037/abn0001069. Online ahead of print.
ABSTRACT
This article contends that many of the chief complaints about the Diagnostic and Statistical Manual of Mental Disorders can be obviated by a set-theoretic, combinatorial approach. Arbitrary cutoffs, polythetic criteria, and category heterogeneity can be avoided by using a data-driven approach that assesses whether particular symptom combinations represent sufficient conditions for clinical benchmarks at an acceptable level of conditional probability. Using data from the National Comorbidity Survey-Replication, this study employed generalized anxiety disorder, major depressive disorder, posttraumatic stress disorder, and the union of major depressive disorder and generalized anxiety disorder as exemplars and set a target probability threshold of p ≥ .90 for sufficiency. All possible symptom combinations were generated for each subsample, with sample sizes of N = 1,948, N = 2,285, N = 777, and N = 3,129, respectively. Sufficient sets were identified for diagnosis, clinical distress, and functional impairment. Establishing sufficiency reduced the number of possible symptom combinations by at least 94% (M = 98.7%, SD = 1.79%). Finally, in a large, randomly split-halved subsample (N = 6,656), sufficient sets were identified at p ≥ .90 and tested in the holdout data. Results yielded an average conditional probability of .91 (SD = .03), reinforcing the robustness and generalizability of the current methods. These results suggest that a large amount of the heterogeneity in symptom combinations in internalizing disorders may be nested and reducible. Thus, much of the combinatorial information in the symptom presentations of these disorders may be overlapping and there may be core features of psychopathology that are sufficient to produce fidelity without requiring additional complexity. (PsycInfo Database Record (c) 2025 APA, all rights reserved).
PMID:41051852 | DOI:10.1037/abn0001069
