Cognitive fusion-modulated structural covariance networks predict treatment response in drug-naive OCD: A multiscale neuroimaging-transcriptomic study
Xingbo Suo, Yang Zhang, Xinming Li, Jiejun Liu, Zihe Xu, Luyao Zeng, Lijuan Lei, Menglin Zhang, Zhijun Luo, Yan Min, Xianjun Zeng, Maorong HuAbstract
Background
Cognitive fusion (CF), a core maladaptive process in obsessive-compulsive disorder (OCD), describes overidentification with intrusive thoughts. While implicated in symptom maintenance, its neuroanatomical basis and treatment prediction utility remain poorly understood.
Methods
We enrolled 80 first-episode, drug-naïve OCD patients and 70 matched healthy controls. All participants underwent structural magnetic resonance imaging and clinical assessment. Patients received 8-week selective serotonin reuptake inhibitors treatment with endpoint reevaluation. Using multimodal neuroimaging, we examined group differences in gray matter volume (GMV) and structural covariance network (SCN), tested whether CF levels modulated connectivity, and evaluated baseline features for predicting treatment response. Transcriptomic analysis linked the findings to the Allen Human Brain Atlas data.
Results
OCD patients showed GMV reductions in frontal-limbic regions and increases in parietal/inferior frontal areas. SCN analysis revealed enhanced structural covariance within fronto-parieto-cerebellar circuits. Higher CF scores were associated with stronger structural coupling between the left angular gyrus and the right middle temporal/medial superior frontal gyri. Baseline GMV in the left dorsolateral superior frontal gyrus (SFGdl) positively correlated with CF improvement, remaining significant after controlling for overall symptom reduction. Transcriptomic analysis indicated the left SFGdl GMV patterns spatially associated with glutamatergic synapse and calcium signaling pathways.
Conclusions
CF in OCD links to altered structural integration between large-scale networks, with the left SFGdl emerging as a potential CF-specific treatment predictor. This multiscale study highlights combining neuroimaging with psychological constructs to understand treatment-relevant changes in OCD.