(PS11-C64) Brain-predicted Aging Associates with Socioeconomic Status and Behavior in Childhood and Adolescence

Introduction: An important link between low socioeconomic status (SES) and poor psychological outcomes is their shared association with brain measures across child development. Different components of SES associate with different psychiatric outcomes and differences in brain morphology and connectivity. The task of synthesizing those findings into clinically useful information is difficult given the number of regions involved, and heterogeneous patterns of development across the brain. We applied a single-subject level prediction of brain aging to examine the dissociable effects of components of childhood SES, including parent education, parent occupation, household income-to-needs ratio (INR) and use of public assistance, on children's brain and behavior outcomes. How specific components of SES affect brain maturation and ultimately mental health outcomes remains understudied but is important for developing targeted prevention and intervention programs that will improve developmental outcomes for children living in low SES contexts. 

Methods: Prior work calculated Relative Brain Age (RBA) using covariation patterns for multiple cortical shape measures, white matter volume, and subcortical gray matter volume using the Joint and Individual Variation Explained method. RBA, phenotypic assessment (child behavior checklist [CBCL]), and SES components were available for 470 youth (aged 5-17) in the Healthy Brain Network Study. Multiple linear regression examined associations between SES components and brain age, including sex, parent psychiatric diagnosis, race and ethnicity as covariates. Subsequent analyses focused on the association between RBA and CBCL subtest scores, controlling for sex and scan location. Last, the association between SES components and CBCL subtest scores was evaluated, including sex, parent psychiatric diagnosis, race and ethnicity as covariates.

Results: RBA was significantly associated with public assistance (B = 0.12,  t = 2.15, p = 0.03), parent occupation (B = 0.15, t = 2.43, p = 0.02), and parent psychiatric diagnoses (B = -0.17,  t = -2.34, p = 0.02), but not with INR and parent education (p > 0.4).  These associations were not moderated by age (p > 0.3). RBA was significantly associated with CBCL anxiety/depression (B = -0.09, t = -2.05, p = 0.04), but not with aggressive behavior, attention problems, somatic complaints, withdrawn behavior, or rule breaking behavior (p’s > 0.2). CBCL anxiety/depression was additionally associated with parent occupation (B = -0.14, t = -2.21, p = 0.03) and parent psychiatric diagnoses (B = 0.16, t = 2.18, p = 0.03) but not with public assistance, INR or parent education.

Conclusions: Distinct components of SES are associated with brain aging (RBA) and internalizing problems in youth. Low parental occupational prestige was associated with child anxiety/depression scores and delayed brain aging, which in turn was also associated with child anxiety/depression. These findings suggest that delayed brain aging may represent a biological mechanism underlying known links between SES and mental health risk. Single subject measures of brain aging may prove useful for internalizing disorder risk assessment and diagnosis in youth and should be explored in future work.