Mapping associations between early life adversity and alcohol use disorder through human epidemiology and rat neurobiology

Early life adversity (ELA) and adolescent alcohol use are both recognized contributors to the risks of developing Alcohol Use Disorder (AUD) in adulthood. However, the extent and nature of their interactions remains unclear. In this thesis, I combine epidemiological data and pre-clinical experiments to test the hypothesis that adolescent drinking lies on the causal pathway linking ELA to risk of AUD in adulthood.

I used data from the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort to test the hypothesis that Adverse Childhood Events (ACEs) are associated with AUD at age 24, and that this effect is mediated by heavy adolescent drinking. Using multiple imputation by chained equations combined logistic regression and g-formula estimation, I show that both ACEs and heavy adolescent drinking are strong independent predictors of AUD, though heavy adolescent drinking is only a weak mediator of the association between ACEs and AUD.

To model ACEs preclinically, I developed a within home cage operant system for alcohol administration to test the effect of maternal separation (MS) on alcohol-related behaviour in male (n=22) and female (n=26) Sprague-Dawley rats. The operant system successfully enabled longitudinal quantification of alcohol consumption. MS was not associated with changes in any measures of alcohol-related behaviour in adolescence or adulthood, or with anxiety-like behaviour on the elevated plus maze. In contrast, adolescent alcohol intake was associated with changes in adulthood drinking behaviour, whose scale and nature varied with sex and MS.

A subset of rats (n=15) from the operant study were implanted with local field potential (LFP) electrodes targeting neural circuits central to addiction. Spectral analysis of prefrontal cortex LFP demonstrated increased beta (15-30Hz) and decreased low gamma (30-48Hz) power associated with adolescent alcohol intake, and increased theta (6-10Hz) and decreased low and high gamma (52-90Hz) power associated with adulthood alcohol preference. Directed functional connectivity measures evidence complex changes in cortico-limbic and cortico-striatal circuits related to MS, adolescent alcohol intake, and adulthood alcohol consumption. 

While this thesis did not identify strong evidence for the mediation model, or an effect of MS on drinking behaviour in rats, the studies presented converge on the important role of adolescent alcohol use in the development of AUD. I highlight the potential importance of reducing heavy adolescent drinking via alcohol policy and demonstrate how application of the pre-clinical model could contribute to discovering targeted pharmacological and neuromodulation treatments for AUD.