PROJECT ABSTRACT
Alcohol use disorder (AUD) rates have doubled in the past five years. Despite having effective treatments, over
50% of patients relapse within the first year. Theories suggest that a primary reason for high relapse rates is
the development of anxiety during abstinence that causes continued alcohol use. That is, individuals with AUD
use alcohol for its negative reinforcing property, which removes anxiety that occurs during alcohol abstinence.
Chronic alcohol use impacts negative reinforcement neural circuits, which contain the bed nucleus of the stria
terminalis (BNST) at its core. The BNST is a region that contributes to negative affect, mediates stress-induced
reinstatement (relapse), and was recently implicated in negative reinforcement. The BNST has extensive
inputs from regions implicated in reward and emotion processing—specifically, the amygdala, ventral and
dorsal striatum, and dorsal anterior cingulate cortex (dACC). In alcohol research, the primary emphasis has
been on positive reinforcers (i.e., alcohol), providing us with extensive knowledge about the effects of positive
reinforcement on behavior and brain function in AUD. In contrast, the alcohol field has a limited understanding
of the effects of negative reinforcement on behavior and brain function. Considering that conceptual models of
AUD cite negative reinforcement as a driver of relapse, this represents a major gap in knowledge. Human
studies can leverage findings from animal models to provide a better understanding of negative reinforcement
in AUD. Specifically, cross-species translation will be foundational for determining underlying behavioral and
brain mechanisms of AUD and relapse. To address cross-species translation and underlying mechanisms of
negative reinforcement, we formed a collaboration with a preclinical researcher to translate a well-validated
animal paradigm of negative reinforcement into humans. The proposed project will investigate negative
reinforcement behaviors and connectivity in a negative reinforcement neural network in adults with AUD who
are in early abstinence (EA). Specific Aim 1 will characterize negative reinforcement learning and bias in EA
adults compared to controls (CON). Specific Aim 2 will determine connectivity strength of a BNST negative
reinforcement neural network using intrinsic (‘resting state’) functional connectivity and diffusion tensor imaging
(DTI). Both aims will test for group differences (EA, CON) and exploratory analyses will determine whether
negative reinforcement network connectivity is associated with negative reinforcement behaviors in the task.
The hypotheses are that EA participants will (1) learn negative reinforcement faster and have a bias for
negative reinforcement and (2) display stronger negative reinforcement network connectivity than controls. The
results of this study will fill a critical gap in knowledge to better understand the behavioral and neural
mechanisms that underly AUD and relapse, representing a critical target to guide future prevention and
treatment of AUDs.