Childhood adversity and late-life cognitive and brain health in a diverse cohort

Abstract

Introduction: Childhood adversity harms neurodevelopment. Literature on late-life brain health is limited, and findings on late-life cognition are mixed.

Methods: Pooling data from Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) and Study of Healthy Aging in African Americans (STAR) cohorts, we assessed the impact of childhood adversity (factor score from seven self-reported items) on (a) executive function and verbal memory decline using linear mixed effects models (n = 2447), (b) structural magnetic resonance imaging (MRI) using linear regression (n = 618), and (c) amyloid positron emission tomography (PET) using generalized linear models (n = 331), all adjusting for early-life demographic and socioeconomic confounders.

Results: Childhood adversity was not associated with cognition except for a slightly faster decline in verbal memory ( $\hat{\beta }$ = -0.013 SD/year, 95% confidence interval [-0.025, -0.001]). Among neuroimaging outcomes, childhood adversity was associated with only larger temporal lobe volumes ( $\hat{\beta }$ = 0.092 SD [0.012, 0.173]).

Discussion: More research evaluating sources of resilience, heterogeneity, and bias is needed to explain inconsistent findings across studies.

Highlights: We developed measurement models to capture childhood adversity in a diverse cohort. Childhood adversity was associated with a slightly faster verbal memory decline. We examined childhood adversity's effect on structural MRI and amyloid PET measures. Higher childhood adversity was associated with larger temporal lobe volumes.

Keywords: Alzheimer's disease; amyloid; childhood adversity; cognition; executive function; neuroimaging; verbal memory.

FIGURE 1

Estimated effect (SD units) of the childhood adversity factor z‐score on z‐scored late‐life brain volumes residualized for intracranial volume. Model 1: unadjusted; Model 2: adjusted for age at scan, sex/gender, southern US birth, race and ethnicity, and parental education; Model 3: additionally adjusted for childhood SES measures (family financial status, ever skipped meal/went hungry because insufficient money for food, and family standing in the community). SES, socioeconomic status.