EXPOSOME SIGNATURES IN BRAIN STRUCTURE AND FUNCTION: EVIDENCE FROM MONOZYGOTIC TWINS

Understanding how environmental exposures relate to the human brain is challenging because genetic and shared-family factors can confound observed associations. We leverage a monozygotic (MZ) co-twin control design: because MZ twins share near-identical genomes and much of their early family environment, within-pair differences provide a stringent test of whether exposure differences covary with neural phenotypes. In 154 MZ twin pairs (age 27.8 ± 6.68 years; 54.3% female at birth), we collected multi-source exposome data (~10,000 variables) alongside structural MRI (n=308), resting-state fMRI (n=304), and mental health and cognitive measures. Exposome assessment included psychosocial questionnaires, wearable-derived sleep and activity phenotyping, air pollution indices, and ecological momentary assessment. We analyse within-pair difference scores, derive polyenvironmental risk scores (PERS) to summarise cumulative exposure burden, and apply sparse partial least squares (sPLS) to identify multivariate exposome–brain patterns.
Despite shared genetics, twins showed substantial within-pair variability in exposome domains (twin similarity ICCs=0.34–0.58), brain structure (ICCs=0.30–0.96), and resting-state functional connectivity (ICCs=0.00–0.27). Functional connectivity exhibited lower twin similarity than structural measures, consistent with greater non-shared contributions and/or higher state and measurement variability in fMRI phenotypes. Candidate exposome domains emerging from initial analyses include sleep/circadian patterns, air quality, and physical activity, highlighting potentially modifiable exposures to prioritise for exposome–brain modelling in this genetically controlled design.