Plasma proteomic signatures in bipolar disorder and their implications for data-driven patient stratification.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
Bipolar disorder (BD) is a heterogeneous mood disorder that lacks robust biomarkers for biological stratification. We used high-throughput plasma proteomics to characterize case-control differences, identify proteomic subgroups within BD, and examine their associations with blood-based markers and clinical features. Plasma samples from the French I-GIVE (Immuno-Genetics, Inflammation, retroVirus, Environment) cohort, including 248 participants with BD and 76 healthy controls, were assessed using the Olink Reveal panel (1,034 proteins). Age-, sex-, and BMI-adjusted case-control analyses identified 527 differentially expressed proteins in the imputed analysis and 136 in the complete-case analysis, with high agreement in effect estimates and involvement of immune, metabolic, oxidative stress, and endothelial-related processes. Despite the large number of differentially expressed proteins, pathway enrichment was limited to B-cell receptor and chronic myeloid leukemia-related signaling in BD. Unsupervised clustering within BD identified two proteomic subgroups, with enrichment between clusters in longevity-regulating pathways, cellular senescence, FLT3 signaling, and transcription-related pathways. Among four measured inflammatory/endothelial markers, serum amyloid A was the only marker showing a significant difference between groups (FDR < 0.05). Clinical characteristics were similar across clusters, although childhood physical abuse and a greater somatic comorbidity burden were associated with the subgroup showing higher protein abundance. Our findings support biological heterogeneity within BD and suggest that proteomic subgrouping may capture variation not evident in pooled case-control analyses.