Structural connectomic signatures of childhood maltreatment across affective and psychotic disorders.

Abstract

Exposure to childhood maltreatment (CM) is a crucial risk factor for affective disorders and schizophrenia, which are linked to structural brain network alterations. We hypothesized a significant association between CM and brain structural connectivity in a transdiagnostic sample. Participants included patients with major depressive disorder (n = 827), bipolar disorder (n = 134), schizophrenia spectrum disorders (n = 118), and healthy controls (n = 932) aged 18-65 from the Marburg-Muenster Affective Disorders Cohort Study. Structural brain networks were reconstructed from structural and diffusion-weighted MRI. CM was assessed using the Childhood Trauma Questionnaire (CTQ) based on established cut-offs. Analyses of covariance investigated the main effect of CM and the interaction of diagnosis and CM on global network metrics. In addition, network-based statistic (NBS) analyses were performed to identify the underlying subnetworks. History of CM was positively associated with the number of connections (p_FDR=0.01; CM: 95 %-CI=[533.9; 540.6], No CM: 95 %-CI=[529.1; 535.0]), normalized (p_FDR=0.02; CM: 95 %-CI=[0.860; 0.863], No CM: 95 %-CI=[0.857; 0.860]) and non-normalized network efficiency (p_FDR=0.01; CM: 95 %-CI=[0.405; 0.407], No CM: 95 %-CI=[0.403; 0.405]) and negatively associated with normalized network clustering (p_FDR=0.045; CM: 95 %-CI=[4.976; 5.065], No CM: 95 %-CI=[5.100; 5.150]). There were no significant interaction effects of CM and diagnosis. NBS analyses revealed a large network of edges, in which individuals with CM exhibited higher structural connectivity compared to those without, which persisted after correcting for diagnosis. These findings indicate a connectome signature of CM, marked by hyperconnectivity, across major mental disorders and in healthy individuals, suggesting that CM affects connectome architecture independent of present psychiatric diagnoses.