Abstract
Importance: Childhood maltreatment is associated with serious developmental consequences that may be different depending on the form of maltreatment. However, relatively little research has investigated this issue despite implications for understanding the development of psychiatric disorders after maltreatment. Objective: To determine the association of childhood maltreatment and potential differential associations of childhood abuse or neglect with neural responsiveness within regions of the brain implicated in emotional responding and response control. Design, Setting, and Participants: In this cross-sectional study, participants aged 10 to 18 years with varying levels of prior maltreatment as indexed by the Childhood Trauma Questionnaire (CTQ) were recruited from a residential care facility and the surrounding community. Blood oxygen level-dependent response data were analyzed via 2 analyses of covariance that examined 2 (sex) × 3 (task condition [view, congruent, incongruent]) × 3 (valence [negative, neutral, positive]) with Blom-transformed covariates: (1) total CTQ score; and (2) abuse and neglect subscores. Data were collected from April 1, 2016, to June 30, 2018. Data analyses occurred from June 10, 2018, to October 31, 2018. Main Outcomes and Measures: Blood oxygenation level-dependent signals in response to an Affective Stroop task were measured via functional magnetic resonance imaging. Results: The sample included 116 youths (mean [SD] age, 15.0 [2.2] years; 70 [60.3%] male). Fifteen participants reported no prior maltreatment. The remaining 101 participants (87.1%) reported at least some prior maltreatment, and 55 (54.5%) reported significant maltreatment, ie, total CTQ scores were greater than the validated CTQ score threshold of 40. There were significant total CTQ score × task condition associations within the bilateral postcentral gyrus, left precentral gyrus, midcingulate cortex, middle temporal gyrus, and superior temporal gyrus (left postcentral gyrus: F = 11.73; partial η2 = 0.14; right postcentral and precentral gyrus: F = 9.81; partial η2 = 0.10; midcingulate cortex: F = 12.76; partial η2 = 0.12; middle temporal gyrus: F = 13.24; partial η2 = 0.10; superior temporal gyrus: F = 10.33; partial η2 = 0.11). In all examined regions of the brain, increased maltreatment was associated with decreased differential responsiveness to incongruent task trials compared with view trials (left postcentral gyrus: r = -0.34; 95% CI, -0.17 to -0.51; right postcentral and precentral gyrus: r = -0.31; 95% CI, -0.14 to -0.49; midcingulate cortex: r = -0.36; 95% CI, -0.18 to -0.53; middle temporal gyrus: r = -0.35; 95% CI, -0.17 to -0.52; superior temporal gyrus: r = -0.37; 95% CI, -0.20 to -0.55). These interactions were particularly associated with level of abuse rather than neglect. A second analysis of covariance revealed significant abuse × task condition (but not neglect × task) interactions within the midcingulate cortex (F = 13.96; partial η2 = 0.11), right postcentral gyrus and inferior parietal lobule (F = 15.21; partial η2 = 0.12), left postcentral and precentral gyri (F = 11.16; partial η2 = 0.12), and rostromedial frontal cortex (F = 10.36; partial η2 = 0.08)). In all examined regions of the brain, increased abuse was associated with decreased differential responsiveness to incongruent task trials compared with view trials (midcingulate cortex: partial r = -0.33; P < .001; right postcentral gyrus and inferior parietal lobule: partial r = -0.41; P < .001; left postcentral and precentral gyri: partial r = -0.40; P < .001; and rostromedial frontal cortex: partial r = -0.40; P < .001). Conclusions and Relevance: These data document associations of different forms of childhood maltreatment with atypical neural response. This suggests that forms of maltreatment may differentially influence the development of psychiatric pathology.