Exploring Novel Neuroanatomical Biomarkers for Alcohol Use Disorder: Considerations of Hippocampal and Amygdalar Subregions, Sulcal Morphology, and Fractal Dimensionality

Abstract

Objective: Alcohol use disorder (AUD) remains a leading cause of worldwide mortality and morbidity. The development of neuroanatomical biomarkers offers the potential of novel clinical indicators to guide prevention, early diagnosis, and treatment. Methods: In 76 participants with DSM-5 diagnosed AUD (Mage = 35.75; 51.3% female) and 79 controls (Mage = 34.71; 59.5% female), we utilized magnetic resonance imaging (MRI) to investigate four novel measures: hippocampal and amygdalar subregion volumes, sulcal morphology (SM), and fractal dimensionality (FD). MRI processing, segmentation, and SM and FD quantification were completed using FreeSurfer v6.0 and v7.0, and MATLAB toolboxes, respectively. A significance value of p < .05 was employed for analysis and sex, age, and intracranial volume were included as covariates. Results: Volumes of the right presubiculum, subiculum, and molecular layer head; left lateral and accessory basal nuclei; and corticoamygdaloid transition area were significantly lower in AUD participants relative to healthy controls. Widths of the left occipito-temporal, right middle occipital and lunate, and right marginal part of the cingulate sulci and depth of the post-central sulci were significantly increased in AUD participants relative to controls. Finally, decreased left caudate, left thalamus, right putamen and right pallidum FD and greater inferior lateral and third ventricle FD were observed in AUD participants relative to controls. Each novel measure’s reliability was assessed using test-retest data from the Human Connectome Project and indicated high reliability with median intraclass correlations of .93, .91, .88, and .93 for the hippocampal subfields, amygdalar nuclei, SM, and FD, respectively. Conclusion: These results indicate selectively decreased hippocampal and amygdala subregion volume, increased sulcal depth and width, and differences in FD as promising neuroanatomical biomarkers for AUD.