UCLA

Subjective Response and Self-Administration of Alcohol: Implications for Alcoholism Etiology

by

Spencer Bujarski

Doctor of Philosophy in Psychology

University of California, Los Angeles, 2017

Professor Lara Allison Ray, Chair

The biobehavioral effects of alcohol have emerged as central components of alcoholism etiology in both clinical and preclinical research. To understand the biology of alcohol use disorder (AUD) preclinical researchers have developed animal models of drug and alcohol addiction, however, the translational applicability of these models is often assumed by face validity. The overarching aim of this dissertation is to advance a human laboratory framework for translating the preclinically derived Allostatic Model, and Incentive Salience Theory (IST) to clinical samples.

Despite the central role subjective responses to alcohol (SR) in several etiological theories, relatively little is known about its structure. Therefore, prior to leveraging SR as a translational phenotype, Paper I examined the factor structure of SR using data from a large and well-controlled alcohol challenge study. Results suggested a 4-factor solution to SR along the domains of Stimulation/Hedonia, Negative Affect, Sedation/Motor Intoxication, and Craving/Motivation capturing positive reward, negative reward, punishment and subjective motivation respectively. This paper offers a synthesis of multiple SR measures and provides recommendations for consolidating these measures into four meaningful and interrelated constructs.

In Paper II, predictions from the Allostatic Model regarding a transition from positive reinforcement to negative reinforcement in alcohol dependence were tested. Heavy drinking participants representing a range of AUD severity completed a novel alcohol administration paradigm designed to capture domains of alcohol reward and reinforcement via a standardized alcohol challenge and progressive ratio self-administration. Validating the laboratory paradigm, AUD predicted greater alcohol craving, and greater alcohol reinforcement. Furthermore, craving during the challenge robustly predicted subsequent reinforcement behavior. These data however, provided evidence for neither a transition from positively reinforced alcohol use, nor a transition to negative reinforcement. AUD severity did not predict stimulation/hedonia, and hedonic responses didn't predict self-administration regardless of AUD severity. While AUD severity was associated with greater basal negative affect, alleviation of negative affect by alcohol did not differ by AUD severity and negative affect did not predict self-administration at any level of AUD severity. Sedative response however did predict lower reinforcement behavior consistent with the Differentiator and Low-Level of Response models derived in human laboratory research. This study thus represented a novel approach to translating predictions from the Allostatic Model to a clinical sample. While the experimental paradigm did meaningfully capture domains of alcohol reward and reinforcement, hypotheses regarding allostatic processes were generally not supported.

In Paper III, hypotheses from the Incentive Sensitization Theory were tested in terms of the dissociation between alcohol liking and wanting, and the role of incentive salience in motivating alcohol consumption. Overall the hypotheses based on IST were not supported. AUD severity predicted neither subjective liking nor wanting during the alcohol challenge, and these variables remained highly correlated across AUD severity. Utilizing an alcohol dot probe measure of attentional bias, AUD severity did not predict greater incentive salience, and incentive salience variables did not predict reinforcement. Of note however, these data were consistent with recent evidence suggesting the dot probe tasks may be unreliable. Therefore, it is possible that these null results are a consequence of poor reliability for the incentive salience measure. Thus, by examining a behavioral measure of incentive salience in addition to subjective liking and wanting, this paper advanced a more complete framework for translating IST to clinical populations, and highlighted the need for accurate behavioral measures of incentive salience in humans.

In order to optimally benefit from the precision of preclinical research, it is necessary to establish that preclinical models accurately map onto key aspects of human AUD. Together these three papers advance a novel human laboratory framework for translating preclinical models of addiction to clinical samples. Factor analytic work validated SR as a translational phenotype capturing key aspects of reward and punishment. The novel experimental paradigm in Papers II and III captured variations in both alcohol reward and reinforcement and found results which were consistent with prominent human laboratory models of alcoholism risk. Conversely however, few of our hypotheses derived from the Allostatic Model or Incentive Sensitization Theory were supported highlighting the need for increased reverse translation efforts to promote consilience between preclinical and human subjects research. Ultimately, determining whether preclinical theories of alcoholism etiology accurately capture pathological processes in humans suffering from AUD is essential to fully understanding alcoholism etiology and developing efficacious treatments to alleviate the sizeable human cost of alcohol addiction.