A new neuroimaging study published in European Neuropsychopharmacology has found that childhood abuse and neglect are linked to distinct differences in brain structure among young adults, with patterns varying by age, biological sex, and the specific type of adversity faced. The research identifies two separate brain-behaviour profiles: one associated with physical abuse and trauma denial, the other linked to emotional abuse in women and reductions in the left amygdala.
Childhood maltreatment encompasses physical, emotional and sexual abuse, as well as physical and emotional neglect. Individuals who experience maltreatment in childhood are more likely to develop mood and anxiety conditions at a younger age and often exhibit more severe symptoms than peers with the same psychiatric diagnoses who were not mistreated.
Scientists in the study examine how early stress relates to the brain’s physical structure, focusing on the cerebral cortex—the folded outer layer involved in complex thought and sensory processing—and the amygdala, a deep-brain region central to processing emotions and threats. The work adds to a growing body of evidence suggesting that trauma does not produce uniform brain changes across all individuals.
The researchers drew on data from 251 help-seeking individuals aged 17 to 35, with an average age of about 25, and a near-even split between 130 males and 121 females. Participants were recruited across six university hospitals and early detection centres in Germany as part of a project tracking people with early mood or anxiety symptoms, ADHD or other risk factors for bipolar disorder. All participants completed the Childhood Trauma Questionnaire and underwent structural MRI scans to measure cortical thickness and surface area, as well as the volumes of nine amygdala subregions and multiple hippocampal subregions.
The team applied a pattern-finding method, sparse partial least squares, to combine 28 trauma items (including minimisation or denial of hardship), age, sex and 182 brain measurements to identify clusters of life experiences that co-vary with brain structure. They validated the patterns by testing them on data held out from the model.
Two distinct brain-behaviour patterns emerge
The analysis revealed two distinct profiles. The first linked older age, severe physical abuse and pronounced trauma denial with thinning of the cerebral cortex across both hemispheres. This cortical-thinning pattern accounted for 18% of the shared variance between participants’ trauma profiles and their brain measurements, with cortical thickness contributing the vast majority of the pattern.
Follow-up checks showed that the physical abuse item—asking whether a person had been hit hard enough for a teacher, neighbour or doctor to notice—along with the denial item, independently predicted this cortical thinning. However, the cortical thinning pattern did not predict how well participants functioned in daily life one or two years later.
The second profile was female-specific. It associated female sex, higher levels of emotional abuse and smaller contributions from physical abuse, emotional neglect and sexual abuse with reduced volumes in the left amygdala. The reduction targeted the left amygdala as a whole and three internal subregions—the basal, paralaminar and accessory basal nuclei—while hippocampal subregions did not drive this pattern. This profile explained about 32% of the shared variance between the trauma and brain data and related most strongly to emotional trauma in women.
The female-specific pattern was linked to functioning over time. Across all female participants with follow-up data, average daily functioning rose from baseline but those with higher emotional abuse scores tended to have worse functioning at one year and at two years, even after accounting for baseline levels. Interestingly, among those with high emotional trauma, smaller left amygdala subregions initially predicted better functioning at the two-year mark, though this association was no longer statistically significant after adjusting for baseline functioning.
To better understand this amygdala pattern, researchers split the female participants with high emotional trauma into subgroups with smaller versus larger left amygdala volumes and trained a classifier on 17 clinical and diagnostic measures. Among the most extreme third of the female group (26 women per subgroup), the algorithm distinguished those with larger versus smaller left amygdala volumes with 69.2% balanced accuracy, while models for male participants did no better than chance.
Lead author Fabian Huth, of the Carl Gustav Carus University Hospital at Technische Universität Dresden, explained the study’s ambition: “What interests me most is the attempt to understand how different experiences of maltreatment relate to brain structure across the early stages of mental health difficulties.” He emphasised that developing an understanding of developmental history may reveal information that a current diagnosis alone cannot capture, noting that cortical thinning was not linked to functioning at follow-up.
External expert Zai-Fu Yao, an associate professor at National Tsing Hua University in Taiwan who was not involved in the study, welcomed the approach but urged caution. “The clinical meaning of an imaging finding still has to be demonstrated,” he said. “One finding deserves particular attention: the cortical-thickness pattern was not associated with functioning at follow-up. Identifying a brain difference and explaining a person’s everyday difficulties are separate scientific tasks.”
Yao described the study as “promising exploratory evidence,” stressing the need for replication in independent samples before drawing firm conclusions about sex- or age-specific brain mechanisms. He also pointed out that single-timepoint brain scans and reliance on self-reported trauma histories can limit interpretations about causation and the evolution of these patterns over time.
The authors acknowledge several caveats. The study relied on self-reported trauma, which can be influenced by memory distortions or current distress, and the single assessment means it cannot determine whether brain differences were caused by maltreatment, predated it, or emerged as temporary changes in early mood disorders. They call for longitudinal work that tracks brain development alongside objective records of adversity and cognitive testing to determine how these patterns evolve and whether they might inform personalised mental health care.
The study, titled Childhood maltreatment is linked to transdiagnostic age- and sex-specific cortical thinning and reduced amygdalar volume in young adults, was authored by Fabian Huth and colleagues, including researchers from Dresden, the Max Planck Institute for Psychiatry and Ludwig-Maximilian-University in Munich, among others.
