Biologically Annotated Heterogeneity of Depression Through Neuroimaging Normative Modeling

Depression is not a unitary disorder; it is heterogeneous in nature. Likewise, no 2 individuals with depression are entirely alike, and therefore their associated symptoms are highly personalized. Over the past decade, numerous approaches have been developed to identify neuroimaging-derived biomarkers to advance our understanding of the neurobiology of patients with depression at the group level. However, substantial clinical heterogeneity among individuals with depression hinders the development of biomarkers for personalized interventions. Recently, publicly available resources have enabled researchers to investigate precision neuromarkers for depression using integrative multineuroimaging approaches. In this review, we systematically revisit previous findings and discuss the advances in data-driven neuroimaging analyses for depression heterogeneity, including the disentangling of dimensional and overlapping strategies, individual-specific abnormal patterns based on normative modeling frameworks, and associations between multiscale organizations. We also discuss the limitations, challenges, and future directions for depression heterogeneity. A summary of these advances is crucial for enhancing the understanding of the neurobiology of depression and will facilitate more accurate diagnoses and personalized interventions.