Brain signatures in children who contemplate suicide: learning from the large-scale ABCD study

Background Suicide is the second-leading cause of death in youth. Understanding the neural correlates of suicide ideation (SI) in children is crucial to ongoing efforts to understand and prevent youth suicide. This study characterized key neural networks during rest and emotion task conditions in an epidemiologically informed sample of children who report current, past, or no SI. Methods Data are from the adolescent brain cognitive development study, including 8248 children (ages 9–10; mean age = 119.2 months; 49.2% female) recruited from the community. Resting-state functional connectivity (RSFC) and activation to emotional stimuli in the salience (SN) and default mode (DMN) networks were measured through fMRI. Self-reported SI and clinical profiles were gathered. We examined the replicability of our model results through repeated sub-sample reliability analyses. Results Children with current SI (2.0%), compared to those without any past SI, showed lower DMN RSFC (B = −0.267, p < 0.001) and lower DMN activation in response to negative as compared to neutral faces (B = −0.204, p = 0.010). These results were robust to the effects of MDD, ADHD, and medication use. Sub-sample analysis further supported the robustness of these results. We did not find support for differences in SN RSFC or in SN activation to positive or negative stimuli for children with or without SI. Conclusions Results from a large brain imaging study using robust statistical approaches suggest aberrant DMN functioning in children with current suicide ideation. Findings suggest potential mechanisms that may be targeted in suicide prevention efforts.