UCLA

Alcohol use disorders, a cluster of symptoms centered around a loss of control over alcohol use resulting in negative consequences on health and well-being, continue to cause worldwide suffering and economic loss. While decades of research have advanced our understanding of the genetic and brain-based causes of alcohol use disorders, the complexity of the symptom cluster and the corresponding intricacy of the implicated neural circuitry complicates the development of interventions. Thus, instead of focusing on the disorder as a whole, psychiatric research has placed research emphasis on intermediate phenotypes, or `endophenotypes' that lie between the genetic basis and the disorder as a whole (Ducci & Goldman, 2008, 2012; D. Goldman, Oroszi, & Ducci, 2005; Manji, Gottesman, & Gould, 2003).

A number of endophentypes have been explored, most notably craving or incentive sensitization (Robinson & Berridge, 2001, 1993), affect or mood (Ahmed & Koob, 2005; G. F. Koob & Le Moal, 1997), and aspects of executive function (R. Goldstein & Volkow, 2011) includ- ing impulse control (Jentsch & Taylor, 1999). Related most closely to executive function, this dissertation examines biases in risk-related decision-making as having explanatory value for understanding alcohol dependence at its various stages: from alcohol use initiation, to acquisition of the clinical syndrome, to recovery and abstinence. Data presented herein was drawn from both human clinical populations, and animal models, leveraging a translational approach.

First, the neurobiology and genetics of addiction are examined, as is the methods used to measure risk-related decision-making, and how this phenotype might conceptually relate to alcohol use disorders. Next, two analyses of risk-taking data from a clinical and sub- clinical sample demonstrate that, counter to initial predictions, risk-taking as assessed by the Balloon Analogue Risk Task [BART; (C. Lejuez et al., 2002)] is negatively related to clinical symptomatology. The nature of this relationship is further probed, revealing that differences in alcohol problem severity predict differences in loss reactivity, such that those with greater severity take less risk after a big loss than those with lesser severity.

Next, the effects of acute and sub-chronic alcohol dosing regimens are explored in an rodent analogue of the BART. Results indicated that acute alcohol dose dependently de- creased risk-taking in the rat-BART, an effect partially consistent with the human literature where alcohol and other drugs are shown to have few effects on behavior in the human task (Peacock, Bruno, Martin, & Carr, 2013). Furthermore, six weeks of sub-chronic admin- istration of alcohol (versus a saline control) did not significantly alter performance of the rat-BART. Suggestions for future studies are explored.

Finally, one component of the definition of an endophenotype is that it is heritable, and thus regulated by genes (Ducci & Goldman, 2008, 2012; D. Goldman et al., 2005; Manji et al., 2003). In order to assess the heritability of risk-related decision-making, a panel of inbred strains was phenotyped, revealing that about 55% of the variability in performance is attributable to genetic effects. An attempt is made to identify regions of the rat chromosomes that are linked with rat-BART performance using an F2 intercross strategy and quantitative trait loci (QTL) analysis. Preliminary results indicated a candidate region on Chromosome 1 (between approximately 90.99 Mb and 129.99 Mb). However, these results should be interpreted with caution as higher density trait mapping is required to more narrowly define the QTL region, and this significant result was only found in one of six variants of the rat-BART subject to analyses.

In sum, the data presented here, in the context of the broader literature, indicated that future research must address differences in behavioral metrics of risk-related decision-making, as results obtained can be contradictory depending on the task used. Additionally, there are areas where future research is more likely to be clinically influential than others. In particular, more work should be done to examine risk-taking during adolescence, at recovery and during treatment. Interventions at those stages appear most likely to improve outcomes for alcohol dependence.