Obsessive-compulsive disorder (OCD) is a chronic psychiatric disorder characterized by uncontrollable compulsive behaviors and/or obsessive thoughts, with poor prognosis and a high disability rate (Abramowitz et al., 2009). Currently, the first-line treatment options for OCD are selective serotonin reuptake inhibitors (SSRIs) and cognitive behavioral therapy; however, there is still a significant portion of OCD patients who do not respond to treatment with SSRIs (Choi, 2009). Accordingly, the 5-HT hypothesis cannot fully explain the pathogenesis of OCD. Several disease models have been proposed to explain the etiology of OCD. These theories encompass a range of perspectives. Neurobiological theories suggest that cortico-striato-thalamo-cortical (CSTC) circuits play a significant role in the disorder's etiology. Learning-based models focus on the acquisition and extinction of fear. Related cognitive models emphasize the role of disorted beliefs, particularly those related to obsessional anxiety, in the pathogenesis of OCD (Jalal et al., 2023).

Despite the emergence of new disease models for OCD, the CSTC circuit remains recognized as the core pathological foundation of the disorder, with an increasing number of subcircuits now identified as playing a role (Robbins et al., 2024). The CSTC circuit mainly includes the orbitofrontal cortex (OFC), medial prefrontal cortex (mPFC)/anterior cingulate cortex (ACC), thalamus, and caudate nucleus (Brennan et al., 2013; Milad and Rauch, 2012). The CTSC circuits are organized into highly specialized loops that integrate cortical, thalamic, and striatal regions to regulate motor, cognitive, and emotional functions. Within these circuits, the basal ganglia play a central role in modulating information flow through a combination of inhibitory and disinhibitory mechanisms mediated by GABAergic and glutamatergic neurotransmission. Importantly, the caudate nucleus, as part of the striatum, is a key node in these circuits, integrating cortical inputs and modulating thalamic activity through its GABAergic projections. The thalamus, being a critical hub for relaying sensory and motor information to the cortex, is densely interconnected with both the basal ganglia and cortical regions (Halassa and Kastner, 2017). The thalamus provides excitatory glutamatergic feedback to the cortex (Robbins et al., 2019). This intricate balance of excitatory and inhibitory signaling within CTSC circuits underscores the importance of studying both Glutamate (Glu) and Gamma-Aminobutyric Acid (GABA) dynamics to understand their functional roles.

However, the CSTC model also has notable limitations. First, increasing evidence suggests that other brain regions, such as the parietal cortex, occipital cortex, and cerebellum, are also implicated in OCD, indicating that the CSTC circuit alone cannot fully account for the disorder's complex mechanisms (Rauch, 2000). Second, the CSTC model is a simplified framework that may not capture the heterogeneity of OCD, which likely involves multiple neural circuits and neurotransmitter systems. Future research should focus on elucidating the interactions between the CSTC circuit and other brain regions, as well as integrating multidisciplinary approaches to develop a more comprehensive understanding of OCD (Pauls, D. L. et al., 2014).

Consistent with the CSTC model, studies utilizing various neuroimaging and genomic techniques have provided robust evidence supporting the involvement of specific brain regions in the pathophysiology of OCD. The 1H-magnetic resonance spectroscopy (1H-MRS) and resting-state functional MRI (rs-fMRI) studies have consistently demonstrated altered metabolic abnormalities and functional connectivity within the CSTC circuit, particularly involving the caudate nucleus and thalamus (Parmar et al., 2019; Zhao et al., 2021). Furthermore, genomic studies have identified genetic variants associated with these regions, further implicating their role in OCD (Pauls, D. L. et al., 2014). These findings collectively reinforce the importance of the caudate nucleus and thalamus in the CSTC model, while also highlighting the need to explore how these regions interact with other neural networks in the broader context of OCD pathology (Alcaraz et al., 2018; Naaijen et al., 2015; Piantadosi et al., 2021; Starck et al., 2008; Xu et al., 2023).

An increasing number of studies indicate that, in addition to 5-HT, neurotransmitters such as glutamate and GABA are also involved in the pathophysiological processes of OCD (Moon et al., 2018). Previous studies have predominantly focused on the cortical regions within the CSTC circuit of OCD patients using MRS, such as the ventromedial prefrontal cortex (vmPFC) and ACC, reporting relatively consistent findings (Brennan et al., 2013). Three studies showed that glutamate complex (Glx) or Gln concentration in ACC was decreased in OCD patients (Kosová et al., 2023; Rosenberg et al., 2004; Yücel et al., 2008). Another study of 59 OCD patients and 42 healthy controls (HCs) showed that Glu/Cr was decreased in vmPFC in OCD patients (Batistuzzo et al., 2021). However, there is limited research on subcortical nuclei, and the reported results are inconsistent. Several MRS studies investigated changes of Glu/GABA in the caudate nucleus/thalamus of OCD patients. Rosenberg et al. conducted a 1H-MRS study of 11 OCD children and 11 matched HCs, finding that the concentration of Glu and Glx of the left caudate nucleus (Caudate_L) head in OCD patients was higher than in HCs and the concentration of Glx of the Caudate_L decreased after taking medicine, suggesting the pathological hyper-Glu in the caudate nucleus of OCD patients (Rosenberg et al., 2000). Whiteside et al. found that cognitive behavior therapy is associated with a decrease in Glx in the right caudate in OCD children (Whiteside et al., 2012). However, another research of adult OCD patients detected no significant differences of Glx in the caudate compared to matched HCs (Whiteside et al., 2006). As for the thalamus, one study showed that the thalamic Glu concentration was decreased in OCD patients (Yücel et al., 2008), while the other three showed no change or increase in Glu concentration (Gnanavel et al., 2014; Parmar et al., 2019; Simpson et al., 2012). Briefly, findings on Glu and GABA in caudate nucleus and thalamus of OCD patients have shown inconsistencies, which may be attributed to the heterogeneity among OCD patients, necessitating further research to validate these conclusions.

Previous studies showed that functional connectivity of the caudate nucleus/thalamus was associated with the pathophysiology of OCD. A meta-analysis of 47 seed-based resting-state functional connectivity (rsFC) studies (1863 OCD patients; 1795 HCs) revealed increased rsFC between caudate and fronto-limbic network regions and decreased rsFC between caudate and frontoparietal network regions (Liu et al., 2022). The same meta-analysis also detected increased rsFC between the thalamus and the right middle temporal gyrus (MTG) extending to the parahippocampus and amygdala, as well as decreased rsFC between the thalamus and the left cerebellum/superior frontal gyrus (SFG) in OCD patients compared to HCs (Liu et al., 2022). In addition to abnormal rsFC in the CSTC circuit, increasing evidence suggests that the rsFC between the CSTC circuit and the occipital lobe may be related to the pathogenesis of OCD (Chen et al., 2019). Chen et al. found altered rsFCs beyond the CSTC circuit, encompassing the occipital region and limbic and motor systems (Chen et al., 2019). In a rs-fMRI study that compared OCD patients with their healthy first-degree relatives, increased rsFC was observed within the CSTC circuit among OCD patients, alongside decreased rsFC in the cerebellum, occipital cortex, and temporal cortex (Hou et al., 2014).

Glu functions as a crucial excitatory neurotransmitter in the central nervous system, facilitating neuronal firing and modulating metabolic reactions while promoting increased resting-state activity in HCs (Delgado et al., 2022; Enzi et al., 2012). Consequently, there may be a coupling alteration of rsFC and metabolite levels in OCD patients. It is helpful to combine rs-fMRI and 1H-MRS to explore the pathogenesis of OCD. Currently, only one study explored the coupling of rsFC alterations and metabolite alterations in OCD patients. Researchers performed a combined MRS and fMRI study with the prefrontal cortex and bilateral thalamus as ROIs and found that rsFCs between Thalamus_R and right dorsal anterior cingulate cortex (dACC)/middle occipital gyrus were correlated with Glu in the Thalamus_R in unmedicated OCD patients (Chen et al., 2019). However, no studies explored the relationship between the altered rsFCs and neurochemical alterations of the caudate nucleus in OCD patients.

In this study, we utilized a combined approach of 1H-MRS and rs-fMRI to explore the coupling features between rsFC and metabolic abnormalities in unmedicated OCD patients. The MRS data were collected from the bilateral thalamus and caudate nucleus, as they are core subcortical nuclei within the CSTC circuit (Brennan et al., 2013). We recruited adult untreated OCD patients, ruling out the effects of medication and age. Based on neurobiological CSTC models,we hypothesized that: (1) OCD patients would exhibit higher Glu and lower GABA in thalamus compared to controls. (2) Increased rsFC between the striatum (caudate/putamen) and the thalamus in patients with OCD compared to controls, reflecting disinhibition of the thalamus and excessive cortical feedback. Decreased rsFC between the prefrontal cortex (OFC/DLPFC) and the thalamuis in patients with OCD, reflecting impaired cognitive control over compulsive behaviors. (3) Patients with OCD will show a stronger positive association between thalamic Glu concentration and rsFC between the thalamus and cortical regions (e.g., OFC, ACC), reflecting hyperexcitability and excessive feedback in thalamocortical loops. Patients with OCD will show a weaker or negative association between thalamic GABA concentration and rsFC between the thalamus and cortical regions, reflecting reduced inhibitory control over thalamocortical activity. Controls will show no significant association or a weaker association between thalamic Glu/GABA concentrations and rsFC, reflecting a more balanced and regulated thalamocortical system.