Multimodal Brain Imaging and Plasma Proteins Reveal Links Between Depression and Physical Diseases

Brisbane Convention & Exhibition Centre 

Depression impacts both the brain and body, with peripheral pathological changes increasingly recognized as integral to its pathophysiology. Plasma proteins serve as key indicators of peripheral changes; however, their relationship with depression mediated by brain structure and function remains underexplored. Leveraging data from 3,966 UK Biobank participants, we identified a multimodal neuroimaging-plasma protein component of depression (MNI-PPC-Dep) through a constrained multimodal fusion approach (MCCAR+jICA). The brain modality reveals hippocampal atrophy, reduced sensorimotor network connectivity, and structural deficits in the default mode network. Additionally, abnormality in the subcallosal cingulate is identified, a region linked to metabolic dysfunction in depression and targeted for deep brain stimulation in resistant cases. The plasma protein modality, identified through fusion with brain imaging features, is primarily enriched in metabolic pathways and associated with genetic risks for type 2 diabetes, in contrast to those identified by traditional approaches. Notably, MNI-PPC-Dep demonstrates generalizability by reliably predicting depression symptom severity across datasets, underscoring its clinical potential. Environmental and lifestyle factors, such as air pollution and alcohol use, are linked to MNI-PPC-Dep, which, in turn, predicts the onset of future physical diseases, including cardiovascular and kidney-related conditions. Together, this study highlights metabolic dysfunction as a potential bridge between brain changes, depression, and physical diseases, providing a novel biomarker and valuable insights to inform depression treatment strategies.