Volumetric trajectories of hippocampal subfields and amygdala nuclei influenced by adolescent alcohol use and lifetime trauma

Alcohol use and exposure to psychological trauma frequently co-occur in adolescence and share many risk factors. Both exposures have deleterious effects on the brain during this sensitive developmental period, particularly on the hippocampus and amygdala. However, very little is known about the individual and interactive effects of trauma and alcohol exposure and their specific effects on functionally distinct substructures within the adolescent hippocampus and amygdala. Adolescents from a large longitudinal sample (N = 803, 2684 scans, 51% female, and 75% White/Caucasian) ranging in age from 12 to 21 years were interviewed about exposure to traumatic events at their baseline evaluation. Assessments for alcohol use and structural magnetic resonance imaging scans were completed at baseline and repeated annually to examine neurodevelopmental trajectories. Hippocampal and amygdala subregions were segmented using Freesurfer v6.0 tools, followed by volumetric analysis with generalized additive mixed models. Longitudinal statistical models examined the effects of cumulative lifetime trauma measured at baseline and alcohol use measured annually on trajectories of hippocampal and amygdala subregions, while controlling for covariates known to impact brain development. Greater alcohol use, quantified using the Cahalan scale and measured annually, was associated with smaller whole hippocampus (β = −12.0, p(FDR) = 0.009) and left hippocampus tail volumes (β = −1.2, p(FDR) = 0.048), and larger right CA3 head (β = 0.4, p(FDR) = 0.027) and left subiculum (β = 0.7, p(FDR) = 0.046) volumes of the hippocampus. In the amygdala, greater alcohol use was associated with larger right basal nucleus volume (β = 1.3, p(FDR) = 0.040). The effect of traumatic life events measured at baseline was associated with larger right CA3 head volume (β = 1.3, p(FDR) = 0.041) in the hippocampus. We observed an interaction between baseline trauma and within-person age change where younger adolescents with greater trauma exposure at baseline had smaller left hippocampal subfield volumes in the subiculum (β = 0.3, p(FDR) = 0.029) and molecular layer HP head (β = 0.3, p(FDR) = 0.041). The interaction also revealed that older adolescents with greater trauma exposure at baseline had larger right amygdala nucleus volume in the paralaminar nucleus (β = 0.1, p(FDR) = 0.045), yet smaller whole amygdala volume overall (β = −3.7, p(FDR) = 0.003). Lastly, we observed an interaction between alcohol use and baseline trauma such that adolescents who reported greater alcohol use with greater baseline trauma showed smaller right hippocampal subfield volumes in the CA1 head (β = −1.1, p(FDR) = 0.011) and hippocampal head (β = −2.6, p(FDR) = 0.025), yet larger whole hippocampus volume overall (β = 10.0, p(FDR) = 0.032). Cumulative lifetime trauma measured at baseline and alcohol use measured annually interact to affect the volume and trajectory of hippocampal and amygdala substructures (measured via structural MRI annually), regions that are essential for emotion regulation and memory. Our findings demonstrate the value of examining these substructures and support the hypothesis that the amygdala and hippocampus are not homogeneous brain regions.