Distinctive lexical profiles in Korean-speaking patients with schizophrenia: Analysis of verbal productivity, lexical diversity, and part-of-speech distributions

초록

Background: Language disturbance is a core feature of schizophrenia, yet most linguistic research has been restricted to Indo-European languages. This study investigated lexical characteristics in Korean-speaking patients to determine if reported linguistic deficits are trans-linguistic or language-specific. Methods: We analyzed speech samples from 249 patients with schizophrenia spectrum disorders and 159 healthy controls across eight tasks, ranging from daily life narratives to projective tasks. Using a morphological analyzer validated for Korean, we computed total content-word count, part-of-speech proportions, and lexical diversity using length-robust indices: the Moving-Average Type-Token Ratio (MATTR), Measure of Textual Lexical Diversity (MTLD), and Hypergeometric Distribution D (HD-D). Group differences were examined using Rank Analysis of Covariance, controlling for relevant covariates. Results: Patients exhibited a pervasive reduction in verbal productivity across all contexts. Lexical diversity was also diminished, but with index-specific patterns: MTLD and MATTR revealed widespread deficits across nearly all tasks, whereas reductions in HD-D were exclusively observed in visually constrained tasks. Regarding morphosyntactic profiles, a significant reduction in general adverbs was consistently observed across almost all tasks. Conclusion: Reduced verbal output, diminished lexical diversity, and a selective impairment of adverbial modifiers appear to be robust, trans-linguistic markers of schizophrenia, independent of grammatical typology. By validating these deficits in a non-Indo-European language, this study supports the utility of lexical analysis as an objective, cross-culturally applicable tool for precision psychiatry.

키워드