Child maltreatment, adulthood adversities, and grey matter volume by biological sex and age in the UK Biobank

Marie-Céline Schulte, PhD¹,³, Nicholas Magill, PhD², Charlotte Warren-Gash, PhD³, Karen Devries, PhD¹
Faculty of Epidemiology & Population Health, London School of Hygiene & Tropical Medicine, UK
¹Department of Population Health, Child Protection Research Group ²Department of Medical Statistics
³Department of Non-communicable Diseases Epidemiology, Brain Health Group, Electronic Health Records Research Group

Exposure to ≥2 types of adulthood adversity, but not child maltreatment alone, was associated with lower sex-specific, age-adjusted residual total grey matter volume (GMV), of the largest magnitudes in mid-life (ages 47 to <65 years) males and females, compared to participants who reported no adversities. Findings support adversity-informed, sex- and age-stratified models of residualised GMV that distinguish mid-life (before age 65), consistent with young-onset dementia susceptibility, from later-life (after age 65).

INTRODUCTION

Child maltreatment (CM) associated with lower grey matter volume (GMV) in cognition and stress-regulation regions,^1-3 and to young-onset dementia (YOD) risk,⁴ compared to the unexposed.
Adulthood adversities (violence, housing insecurity, socioeconomic inequity, life-threatening illness or accidents, and social isolation) also correlate with lower GMV,⁵ independently of CM.⁶
Prior neuroimaging findings are inconsistent; studies rarely model CM and adulthood adversity jointly, and most omit residualisation of GMV and sex- and age-stratification, obscuring differences in females and males before versus after age 65.
Aim
Test whether CM and levels of adulthood adversity are associated independently and jointly with sex-specific, age-adjusted residual whole-brain GMV, stratified by sex and age at MRI (47 to <65 vs ≥65 years).

METHODS

Sample
UK Biobank sub-cohort: 28,298 adults with brain MRI + Mental Health Questionnaire (MHQ) data: 12,987 males (46%), 15,311 females (54%).
Outcome
Total GMV (T1-weighted MRI, head-size normalised, 2015–2024), residualised on age and age² within females and males separately for sex-specific, age-adjusted residual total GMV.
Exposures (from MHQ)
CM: 0 vs ≥1 type (5-item Childhood Trauma Screener).
Adulthood adversity: 0, 1, or ≥2 (11 types).
Analysis
Linear regression stratified by sex × age group (4 strata); adjusted marginal means (Stata 18). Tested main effects and CM × adulthood adversities interaction.
Adjusted for ethnicity, Townsend Deprivation Index, education, maternal & paternal dementia, Alzheimer’s disease and schizophrenia polygenic risk scores. Complete-case analysis (~7% missing).

RESULTS

Adulthood adversity showed a negative exposure–response pattern within each sex-age stratum in residual total GMV. ≥2 adulthood adversities correlated with the largest GMV differences below the mean in males and females before age 65 years compared to no adversity (chart above).
CM alone: no evidence of association with residual GMV in any stratum (all 95% CIs spanned zero).
Interaction: borderline CM × adulthood adversity in mid-life females (p=0.053); no evidence in other strata (females ≥65 p=0.273; males 47–<65 p=0.312; males ≥65 p=0.227).

CONCLUSIONS

Exposure to ≥2 adulthood adversities, not CM alone, was associated with lower sex-specific, age-adjusted residual total GMV, particularly in mid-life.
Negative differences from the mean (−4,673 to −2,782 mm³) are comparable in magnitude to hippocampal volume, a marker of dementia risk.⁷
Sex differences (mid-life males > females) may reflect sex-specific cardiometabolic and neuroendocrine ageing.⁸
Supports adversity-informed, sex-age stratified models of residualised GMV suggesting potentially elevated YOD risk before age 65.
Hypothesis-generating: routine screening of asymptomatic adults is not currently indicated, but adulthood violence and adversity burden warrants attention in mid- and later-life brain health equity research.⁹
Clinicians and patient support services may benefit from integrating violence survivor-centred practices into brain health promotion education and dementia early detection.

REFERENCES

1. Yang W, et al. Psychol Med. 2023;53:1681-99.
2. Orellana SC, et al. Proc Natl Acad Sci USA. 2024; 121(13):e2304704121.
3. Teicher MH, et al. Nat Rev Neurosci. 2016;17(10):652-66.
4. Schulte MC, et al. 2026. Under review.
5. Ringwald KG, et al. Hum Brain Mapp. 2022;43(11):3577-84.
6. See CRZ, et al. Eur Psychiatry. 2025;68(1):e18.
7. Fotenos AF, et al. Neurology. 2005;64(6):1032-9.
8. Mosconi L, et al. Neurology. 2017;89(13):1382–90.
9. Murchland AR, et al. Alzheimers Dement. 2025;21(5):e70029.

CONTACT

Marie-Céline Schulte, PhD (LinkedIn)
London School of Hygiene & Tropical Medicine
Faculty of Epidemiology and Population Health
Child Protection Group
Brain Health Group
UK Biobank Application 521978
Funding: AAUW Fellowship (M-C. Schulte); Wellcome CDA 225868/Z/22/Z (C. Warren-Gash).