Identifying stable and consistent resting-state functional connectivity patterns across illness trajectories has the potential to be considered fundamental to the pathophysiology of schizophrenia. We aimed to identify consistent resting-state functional connectivity patterns across heterogeneous schizophrenia groups defined based on treatment response. In phase 1, we used a cross-sectional case-control design to characterize and compare stable independent component networks from resting-state functional magnetic resonance imaging scans of antipsychotic-naïve participants with first-episode schizophrenia ( = 54) and healthy participants ( = 43); we also examined associations with symptoms, cognition, and disability. In phase 2, we examined the stability (and replicability) of our phase 1 results in 4 groups ( = 105) representing a cross-sequential gradation of schizophrenia based on treatment response: risperidone responders, clozapine responders, clozapine nonresponders, and clozapine nonresponders following electroconvulsive therapy. Hypothesis-free whole-brain within- and between-network connectivity were examined. Phase 1 identified posterior and anterior cerebellar hypoconnectivity and limbic hyperconnectivity in schizophrenia at a familywise error rate–corrected cluster significance threshold of < .01. These network aberrations had unique associations with positive symptoms, cognition, and disability. During phase 2, we replicated the phase 1 results while comparing each of the 4 schizophrenia groups to the healthy participants. The participants in 2 longitudinal subdatasets did not demonstrate a significant change in these network aberrations following risperidone or electroconvulsive therapy. Posterior cerebellar hypoconnectivity (with thalamus and cingulate) emerged as the most consistent finding; it was replicated across different stages of treatment response (Cohen’s range −0.95 to −1.44), reproduced using different preprocessing techniques, and not confounded by educational attainment. Posterior cerebellar-thalamo-cingulate hypoconnectivity is a consistent and stable state-independent neural marker of schizophrenia.

中文翻译：

---

小脑后部静息态功能性低连接：精神分裂症不同治疗反应阶段的神经标志物

识别疾病轨迹中稳定且一致的静息态功能连接模式有可能被认为是精神分裂症病理生理学的基础。我们的目的是确定基于治疗反应定义的异质精神分裂症组之间一致的静息态功能连接模式。在第一阶段，我们使用横断面病例对照设计来表征和比较来自未接受过抗精神病药物的首发精神分裂症参与者 (= 54) 和健康参与者 (= 43）；我们还检查了与症状、认知和残疾的关联。在第 2 阶段，我们检查了 4 组 (= 105) 的第 1 阶段结果的稳定性（和可重复性），这些组代表了基于治疗​​反应的精神分裂症的交叉顺序分级：利培酮反应者、氯氮平反应者、氯氮平无反应者和氯氮平无反应者电休克疗法。对网络内和网络间的无假设全脑连接进行了检查。第一阶段以家族错误率校正聚类显着性阈值 < 0.01 确定了精神分裂症患者的小脑后部和前部低连接性和边缘系统高连接性。这些网络畸变与阳性症状、认知和残疾具有独特的关联。在第二阶段，我们复制了第一阶段的结果，同时将 4 个精神分裂症组与健康参与者进行比较。两个纵向子数据集中的参与者在利培酮或电休克治疗后并未表现出这些网络畸变的显着变化。小脑后部连接性低下（与丘脑和扣带回）是最一致的发现；它在治疗反应的不同阶段（科恩范围-0.95至-1.44）进行了复制，使用不同的预处理技术进行了复​​制，并且不会受到教育程度的影响。小脑后部-丘脑-扣带回连接低下是精神分裂症的一致且稳定的状态独立神经标志物。