Real-time aggression in youth and its connection to region-specific structural brain alterations

Aggression is expressed across psychiatric diagnoses. Literature displays associations between aggression and grey matter volume (GMV) in frontal, limbic, and striatal areas. Aggression is commonly assessed using self-report. Real-time aggressive incidents may provide a distinctive representation of the neural correlates of aggression. Structural MRI data were collected from 185 youth: 97 (32F/65 M) in the aggression group (AG) [mean age: 16.22 (SD = 1.10)] and 88 (36F/52 M) in the non-aggressive group (NG) [mean age: 16.29 (SD = 1.31)]. Data were acquired through the Boys Town National Database. Youth were included in AG if they had at least one aggressive incident within the first three months of residential care at Boys Town. FreeSurfer was used to estimate region-specific volumetric parameters following whole brain parcellation into 68 cortical and 14 sub-cortical regions. A multivariate analysis of covariance (MANCOVA) was conducted on pre-hypothesized bilateral brain regions of interest (ROIs) [i.e., middle frontal cortex, orbitofrontal cortex (OFC), anterior cingulate cortex (ACC), pericalcarine, insula, amygdala, and striatum], including age and intracranial volume as covariates. Our MANCOVA showed significant differences in GMV [F(24,158) = 1.61, p ≤ 0.05]. AG showed lower GMV relative to NG in left lateral OFC [F(1,183) = 4.54, p ≤ 0.05], bilateral caudal ACC [F(1,183) = 5.21-6.37, ps ≤ 0.05], bilateral rostral ACC [F(1,183) = 3.80-5.54, ps ≤ 0.05], and bilateral caudate [F(1,183) = 4.60-5.61, ps < 0.05]. Remaining ROIs had no significant differences in GMV. Our results provide insight regarding the structural brain biomarkers of real-time aggression in youth. These findings provide consistent evidence for region-specific structural alterations in the adolescent brain that may aid in clinical applications across diagnoses.