A new neuroimaging study suggests that more severe childhood abuse is linked to harsher discipline of teenagers among mothers whose brains react more strongly to social rejection than to social approval. The research, published in Social Cognitive and Affective Neuroscience, indicates that the way adults process interpersonal feedback may help identify families at greater risk of passing mistreatment to the next generation.
Among mothers who did not show heightened brain reactivity to negative social feedback, the severity of abuse did not predict harsher parenting. The findings point to a neural and cognitive mechanism that could illuminate why some families are more prone to intergenerational cycles of mistreatment than others.
Why childhood abuse does not always lead to harsh parenting
Childhood maltreatment encompasses physical, emotional, or sexual abuse as well as neglect—the failure to provide expected care and stimulation. Children who experience mistreatment are two to three times more likely to mistreat their own children as parents, often through harsh discipline such as shouting, verbal hostility, physical punishment or punishment without explanation. A 2019 meta-analysis of 142 studies indicates that abuse transmits across generations more consistently than neglect.
Because abuse involves interpersonal threat, the processing of social feedback—signals of approval, criticism, or rejection—may shape intergenerational risk. Adults who survived childhood abuse tend to display a negativity bias, for example reading neutral faces as angry or interpreting ambiguous child behaviour as defiant. A 2011 longitudinal study of 499 mothers and their infants found that maternal childhood physical abuse, unlike neglect, was linked to next-generation mistreatment and related to social isolation and aggressive response biases.
Nevertheless, many adults who endured childhood abuse raise their children without hostility. Individual differences in social responsiveness may moderate this link, meaning they can strengthen or weaken the association, with sensitivity to negative cues increasing vulnerability and responsiveness to positive cues offering protection.
Prior neuroimaging work has pointed to circuits where these differences might emerge. A 2019 systematic review of 109 brain-imaging studies of youth found abuse-related threat consistently related to structural differences and heightened threat reactivity in emotion-processing regions, while neglect affected frontal and parietal areas supporting self-control. Altered social processing also appears in adulthood: an earlier neuroimaging study linked childhood maternal hostility and rejection to dampened reward-network activity when anticipating social rewards, particularly in PTSD.
Whether parents’ brain reactions to approval versus rejection moderate the link between childhood abuse and teenage discipline—and whether that pattern differs from responses to monetary outcomes—had not been tested. Camille R. Johnston, Ronan M. Cunningham and colleagues at Temple University and the University of Oregon investigated this using matched social and monetary tasks.
Scanning how mothers respond to social and monetary feedback
The researchers studied 56 female primary caregivers from the greater Philadelphia area, averaging 46 years old, who were raising adolescents aged 11 to 18. The sample included 32 identifying as White, 16 as Black or African American, 5 as Asian or Pacific Islander and 3 as multiracial, with 7 identifying as Hispanic and a median yearly household income of $55,500. To preserve statistical power, only women were recruited. Data from three other volunteers were excluded (two for excessive head motion and one for falling asleep in the scanner). Study data and code were posted on the Open Science Framework.
Mothers completed questionnaires on childhood maltreatment and current parenting. The Childhood Trauma Questionnaire assessed physical, emotional and sexual abuse plus physical and emotional neglect across five-item subscales; using clinical cutoffs, 60 per cent met the threshold for substantial abuse. Mothers also rated harsh discipline, including verbal hostility, physical punishment and punishing without explanation, on a scale from 1 (“never”) to 5 (“always”), with an average of 1.48.
During a prior video call, researchers photographed each mother and told her that age-matched peers from earlier sessions would rate whether they “liked” or “disliked” her. Later, mothers underwent functional magnetic resonance imaging, lying in a scanner and viewing a mirrored screen while pressing buttons as the machine tracked brain activity. All participants received equal positive, negative and neutral feedback; 54 believed the cover story, and results were unchanged if the remaining two were omitted.
In separate scanner runs, caregivers completed matched social and monetary tasks. In the social task, across 42 trials, mothers viewed pairs of gender-matched faces and guessed which peer liked them or disliked them. Feedback signals used green and red circles to indicate a peer’s liking or disliking, while a white circle signified an incorrect guess with no peer rating. The monetary task mirrored the social version, but participants chose between doors concealing $2 or $1 losses, with one random trial added to their payoff.
Researchers compared brain activity during positive versus negative feedback across seven regions linked to childhood abuse and social processing. Three lie deep beneath the cortex—the amygdala, the anterior hippocampus and the putamen—while four sit along the outer cortical surface: the medial prefrontal cortex, the posterior cingulate cortex, the dorsal anterior cingulate cortex and the anterior insula.
Brain sensitivity to peer rejection tracks with harsher discipline
Across all 56 mothers, positive feedback elicited stronger activity than negative feedback only in the putamen for social cues and in the medial prefrontal cortex for money. However, individual differences in the social task consistently moderated the abuse–parenting link. This pattern held across a combined six-region network designed to limit multiple testing, and in six of the seven target regions examined individually.
Among the six moderating regions were the amygdala, anterior hippocampus, putamen, dorsal anterior cingulate cortex, posterior cingulate cortex and medial prefrontal cortex. The anterior insula did not show a statistically significant interaction. In the monetary task, there were no statistically significant moderation effects in the network or in any region. An exploratory direct comparison of social and monetary moderation effects was not statistically significant.
With 56 participants, the study had power for medium-sized interactions within a single task but was not precise enough to distinguish correlated social and monetary responses head-to-head. Sensitivity checks found no statistically significant moderation in two nearby cortical areas associated with general social perception and theory of mind, nor in a motor area. Replacing abuse with neglect in the analysis produced no significant results either.
The authors note that because the neural measure contrasted negative against positive feedback, a general negativity bias could reflect heightened sensitivity to rejection, dampened sensitivity to approval, or both. They suggest that brief stranger feedback in a scanner may not capture the lasting impact of supportive relationships that foster resilience, but that heightened neural reactivity to rejection may signal vulnerability during challenging parent–teen exchanges.
Caveats and directions for future research
The researchers emphasise several caveats. Brain activity, parenting behaviour and abuse recollections were measured at a single time point, so the study cannot establish causation. It remains possible that heightened sensitivity to rejection promotes harsh discipline, or that family conflict itself sensitises the brain to negative social cues.
The sample also limits generalisability: 60 per cent met clinical thresholds for childhood abuse, higher than in a large community sample, yet their average harsh parenting score was below community averages and did not exceed 2.3 on the scale used. The findings may not extend to fathers or to broader partner dynamics.
Additionally, the scanner task delivered one-time feedback from strangers rather than a longitudinal view of real-world interactions or reactions to one’s own child, and combining physical, emotional and sexual abuse into a single score leaves open whether distinct abuse types relate differently to brain function. The researchers call for larger, preregistered longitudinal studies that track families over time and verify social specificity. If confirmed, they suggest interventions aimed at helping parents regulate emotions and reframe threatening social cues to interrupt cycles of abuse.
The study, “Neural Bias to Social Feedback Moderates Link Between Childhood Abuse and Subsequent Harsh Parenting,” was authored by Camille R. Johnston, Ronan M. Cunningham, Isabelle Moger, OgheneTejiri V. Smith, Katie Dow, Anna Kilmer, William J. Mitchell, Vishnu P. Murty and Johanna M. Jarcho.
