Depression is a multifactorial clinical syndrome with a low pharmacological treatment response rate. Therefore, identifying predictors of treatment response capable of providing the basis for future developments of individualized therapies is crucial. Here, we applied model-free and model-based measures of whole-brain turbulent dynamics in resting-state functional magnetic resonance imaging (fMRI) in healthy controls and unmedicated depressed patients. After eight weeks of treatment with selective serotonin reuptake inhibitors (SSRIs), patients were classified as responders and non-responders according to the Hamilton Depression Rating Scale 6 (HAMD6). Using the model-free approach, we found that compared to healthy controls and responder patients, non-responder patients presented disruption of the information transmission across spacetime scales. Furthermore, our results revealed that baseline turbulence level is positively correlated with beneficial pharmacological treatment outcomes. Importantly, our model-free approach enabled prediction of which patients would turn out to be non-responders. Finally, our model-based approach provides mechanistic evidence that non-responder patients are less sensitive to stimulation and, consequently, less prone to respond to treatment. Overall, we demonstrated that different levels of turbulent dynamics are suitable for predicting response to SSRIs treatment in depression.

抑郁症是一种多因素临床综合征，药物治疗反应率较低。因此，确定能够为个体化治疗的未来发展提供基础的治疗反应预测因子至关重要。在这里，我们在健康对照组和未接受药物治疗的抑郁症患者的静息态功能磁共振成像（fMRI）中应用了无模型和基于模型的全脑湍流动力学测量。使用选择性血清素再摄取抑制剂 (SSRI) 治疗八周后，根据汉密尔顿抑郁量表 6 (HAMD6) 将患者分为有反应者和无反应者。使用无模型方法，我们发现与健康对照和有反应的患者相比，无反应的患者表现出跨时空尺度的信息传输中断。此外，我们的结果显示基线湍流水平与有益的药物治疗结果呈正相关。重要的是，我们的无模型方法可以预测哪些患者将成为无反应者。最后，我们基于模型的方法提供了机械证据，表明无反应患者对刺激不太敏感，因此不太容易对治疗产生反应。总体而言，我们证明了不同水平的湍流动力学适合预测 SSRIs 治疗抑郁症的反应。