Scholarlink

<p>Inhibitory control deficits are consistently observed in both schizophrenia (SCZ) and major depressive disorder (MDD), yet whether these impairments reflect shared tansdiagnostic mechanisms or disorder-specific dysfunction remains poorly understood. We recruited <i>N</i> = 259 participants (<i>n</i> = 87 SCZ, <i>n</i> = 86 MDD, <i>n</i> = 86 healthy controls) who completed <i>Stroop</i> and <i>Go/No-Go</i> tasks assessing interference and response inhibition, respectively. Hierarchical drift-diffusion modeling (HDDM) was used to decompose behavioral performance into latent computational parameters. Both patient groups exhibited significant behavioral impairments across most measures, with SCZ showing greater severity. At the computational level, reduced drift rate—reflecting less efficient evidence accumulation—emerged as a core transdiagnostic deficit across both inhibitory subcomponents and both disorders, with more pronounced reductions in SCZ. By contrast, prolonged non-decision time was disorder-differentiating, being consistently elevated across both tasks in SCZ but, in MDD, confined to response inhibition and not robust to processing-speed adjustment. Within-group clinical correlates further diverged, relating to negative-symptom severity and illness duration in SCZ and to current depressive symptom severity in MDD, although these diagnosis-specific associations were not directly compared across disorders. Machine learning classification corroborated these findings, with drift rate serving as the primary shared contributor to patient–control discrimination (area under the receiver operating characteristic curve [AUC] = 0.762–0.846), and non-decision time on the interference task the top-ranked feature in the SCZ–MDD classifier (AUC = 0.718). Together, these findings delineate shared versus disorder-specific computational signatures of inhibitory control impairment in SCZ and MDD, and suggest that computational phenotyping of inhibitory control may serve as a quantitative adjunct in precision psychiatry. <b>Preregistration:</b> This study was preregistered with the Chinese Clinical Trial Registry prior to data collection (registration number: ChiCTR2500111117; <a href="https://www.chictr.org.cn/showproj.html?proj=290089">https://www.chictr.org.cn/showproj.html?proj=290089</a>).</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Shared and disorder-specific computational mechanisms of interference and response inhibition in schizophrenia and major depressive disorder

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月31日

查看全文
  • Tongyi Zhang,
  • Sirui Wang,
  • Haoyang Shi,
  • Jia Li,
  • Xiaolong Yang,
  • Jin Jiao,
  • Xiaoning Huo,
  • Xin Zhao

摘要

Inhibitory control deficits are consistently observed in both schizophrenia (SCZ) and major depressive disorder (MDD), yet whether these impairments reflect shared tansdiagnostic mechanisms or disorder-specific dysfunction remains poorly understood. We recruited N = 259 participants (n = 87 SCZ, n = 86 MDD, n = 86 healthy controls) who completed Stroop and Go/No-Go tasks assessing interference and response inhibition, respectively. Hierarchical drift-diffusion modeling (HDDM) was used to decompose behavioral performance into latent computational parameters. Both patient groups exhibited significant behavioral impairments across most measures, with SCZ showing greater severity. At the computational level, reduced drift rate—reflecting less efficient evidence accumulation—emerged as a core transdiagnostic deficit across both inhibitory subcomponents and both disorders, with more pronounced reductions in SCZ. By contrast, prolonged non-decision time was disorder-differentiating, being consistently elevated across both tasks in SCZ but, in MDD, confined to response inhibition and not robust to processing-speed adjustment. Within-group clinical correlates further diverged, relating to negative-symptom severity and illness duration in SCZ and to current depressive symptom severity in MDD, although these diagnosis-specific associations were not directly compared across disorders. Machine learning classification corroborated these findings, with drift rate serving as the primary shared contributor to patient–control discrimination (area under the receiver operating characteristic curve [AUC] = 0.762–0.846), and non-decision time on the interference task the top-ranked feature in the SCZ–MDD classifier (AUC = 0.718). Together, these findings delineate shared versus disorder-specific computational signatures of inhibitory control impairment in SCZ and MDD, and suggest that computational phenotyping of inhibitory control may serve as a quantitative adjunct in precision psychiatry. Preregistration: This study was preregistered with the Chinese Clinical Trial Registry prior to data collection (registration number: ChiCTR2500111117; https://www.chictr.org.cn/showproj.html?proj=290089).