Poster #60 - Lifetime Alcohol Use and Cortical Structure: Potential Targets for Developmental Work on Alcohol Use

• In Event: Poster Session 17

Abstract

Despite the devastating and global issues caused by alcohol use, treatments are often ineffective (Filbey, 2019). To find better interventions, work must be done to update and improve our models of the etiology and consequences of alcohol use and addiction. The associations between alcohol use and the brain are one promising route towards this goal. Via meta-analysis, Buhler and Mann (2011) found that alcohol use is associated with structural changes in the brain that partially reverse after abstinence. Evidence also suggests that women drink less and may be more sensitive to the effects of alcohol than men (Erol and Karpyak, 2015). However, research into these associations has been limited by analyses that average over large regions of the cortex (Clark et al., 2017; Yang et al., 2020). Given that the human cortex is estimated to consist of 300-400 functional areas (Van Essen et al., 2011), focusing on these large regions limits our ability to link associations to potential mechanisms. Recent methods allow us to divide the brain into finer-grained yet related regions and label these within a framework of associated resting-state networks (Schaefer et al., 2017). Analyses of finer-grained regions run into the problem of multiple comparisons, but regularization methods allow researchers to deal with high-dimensional data and address this issue in a statically principled way. In this study, we examined regions of the cortex associated with alcohol use, cumulatively assessed from late childhood, through adolescence, and into adulthood, in a large sample of adult twins using machine learning methods. Alcohol use was assessed using indicators of quantity, intensity, and frequency of drinking. Anatomical MRI data, collected in adulthood, will be analyzed using 1,000 runs of 10-fold cross validation with elastic nets to predict lifetime alcohol use using cortical thickness, sex, and interactions with sex. Regions with nonzero weight in these cross-validation models will be followed up with mixed-effect models to calculate effect sizes for associations between cortical thickness and lifetime alcohol use. Finally, we will conduct co-twin control analyses to assess whether cortical thickness in these regions is related to between-family differences (consistent with pre-existing genetic and shared environmental liability underlying associations) or within-family differences (consistent with an exposure-related effect) in lifetime alcohol use. All data for these planned analyses have been collected. MRI data, alcohol use data, and data for other covariates of interest have been prepared. Cross-validation models have been run, and final analyses are now ready to be completed. Identifying finer-grained regions of the cortex associated with alcohol use will provide future targets for further investigation related to their role in brain function.

Authors

• Felix Pichardo, University of Minnesota - Twin Cities
  Presenting Author

• Stephen M Malone, University of Minnesota

• Jeremy Harper, University of Minnesota

• William G Iacono, University of Minnesota

• SYLIA WILSON, University of Minnesota - Twin Cities