Cue Effects in Human Addiction: Pavlovian to Instrumental Transfer

Individuals with substance use disorders (SUD) have to cope with drug-related cues and contexts, which can affect instrumental drug seeking as shown with Pavlovian to instrumental transfer (PIT) paradigms in animals and humans. The investigators aimed to investigate the impact of acute and chronic stress on Pavlovian-to-instrumental transfer (PIT), how PIT it is associated with cognitive control abilities and whether such effects predict losing vs. regaining control in subjects with AUD. Moreover, the investigators aimed to develop a novel full transfer task that assesses both, general and specific PIT to investigate whether specific PIT differs between alcohol use disorder (AUD) and control subjects.

This project has four research aims: 1. The investigators want to assess the effects of acute as well as repeated stress exposure on PIT and its neuroimaging correlates, and the investigators will test whether modulation of PIT by stress predicts losing vs. regaining control of drug intake in mild to moderate AUD subjects. 2. The investigators will investigate how cognitive control abilities (e.g. interference control and response inhibition) are related to the costs of interference between Pavlovian and instrumental control in PIT. 3. The investigators will compare PIT, its relation to cognitive control, and the neural mechanisms between different drugs of abuse by investigating participants with AUD and also smokers. 4. The investigators will develop an outcome-specific PIT paradigm with alternative (food) and drug rewards to compare AUD and healthy controls, and the investigators will explore communalities and differences in behavioral and neural processes between general PIT and outcome-specific PIT. Therefore, the investigators conduct 3 Experiments with the following Hypotheses: Experiment 1: Here the investigators assess on a behavioral and neural level a general PIT task that consists of four parts: In the instrumental part, participants have to learn which shells to collect or leave in order to gain as much money as possible; the correct choice is probabilistically rewarded in 80% of all times. Absolutely unrelated to this instrumental part, during Pavlovian conditioning colorful fractals are presented in the back of the screen, which have been passively paired either with monetary gain, no change or monetary loss, i.e. they act as positively valued, neutral or negatively valued cues. These fractals thus act as independent, Pavlovian conditioned background stimuli and interact with instrumental behavior in the unrelated third transfer part of the PIT task. Also, the investigators assess the effects of drug-specific cues in the background on instrumental behavior during this transfer part. To avoid further learning, the transfer part will be conducted under nominal extinction. Finally, query trials, in which participants will have to choose one of two pictures will to assess the relative cue value. Acute stress will be induced by the Trier Social Stress Test (TSST), chronic stress will be assessed by questionnaires and hair cortisol. Hypothesis 1a: Acute and chronic stress exposure increase behavioral and neural PIT effects elicited by monetary Pavlovian conditioned stimuli (CS) and decrease inhibitory effects of alcohol Pavlovian CS. Functional correlates of PIT effects will be found in the amygdala and in the ventral striatum. Hypothesis 1b: Following stress exposure, high PIT effects elicited by monetary Pavlovian CS and low inhibitory effects elicited by alcohol Pavlovian CS predict higher subsequent alcohol intake in the follow-up period of 6 months. Methods: Between group analysis: The investigators will analyze on a behavioral and imaging level how acutely consuming AUD subjects and matched controls differ regarding money- and alcohol-related PIT. Within-group analysis: The investigators will analyze how acute stress affects PIT effects. This 2x2 design (1 within-group subject factor with two levels: STRESS vs. NO STRESS and 1 between-group subject factor with two levels: AUD vs. matched controls) will be tested on a behavioral and neural level. Experiment 2: In this experiment, the investigators will utilize a 2x3 factorial group design. Participants will be categorized into two main groups: those with AUD and those without (non-AUD). Each of these groups will be further divided into three subgroups: daily smokers, non-daily smokers and non-smokers. Each cell (individual subgroup) will consist of 35 subjects. In addition to the PIT task described above, the N = 210 participants will additionally perform a Counting Stroop task and a No-go Simon task. Participants will also perform a stop signal task during functional MRI (fMRI). Methods: Interference Costs: Interference costs will be calculated based on error rate (ER) and reaction time (RT) differences between task conditions, e.g., incongruent vs. congruent conditions. Between-group analysis: the investigators will examine group differences in the interference PIT effect at both the behavioral and neural levels using a two (AUD vs. non-AUD) by three (daily smokers vs. non-daily smokers vs. non-smokers) design. In addition, the investigators aim to investigate the associations between the interference PIT effect and other continuous measures: Alcohol Use Disorders Identification Test (AUDIT), Fagerström Test for Nicotine Dependence (FTND), along with quantity-frequency indices. Other analysis: In order to gain a better understanding of the mechanisms underlying PIT, the investigators will also fit a drift diffusion model to the behavioral data and explore the connectivity within the neural networks during PIT in relation to the connectivity during the stop signal task (response inhibition). Hypothesis 2a: Interference costs (i.e. increased error rate \[ER\] in instrumental responses) during PIT are associated with interference costs at the stimulus level (Stroop task) and at the response level (No-go Simon task), and these costs will also be correlated with the ER during response inhibition (No-go Simon task). Hypothesis 2b: The investigators hypothesize that the interference PIT effect will be more pronounced in the AUD group compared to the non-AUD group. The investigators also anticipate that the effect will be stronger in the smoker group. Furthermore, the investigators hypothesize a positive correlation between the PIT effect and AUDIT, FTND scores, as well as the quantity and frequency of use. As the behavioral PIT effect intensifies, the investigators expect corresponding neural responses to similarly increase. Experiment 3: The investigators here use a newly-developed full PIT paradigm that introduces primary reinforcers, i.e. alcoholic and soft drinks (that the investigators administer on a trial-by-trial basis). Furthermore, the investigators implemented additional Pavlovian stimuli to examine both general and outcome-specific PIT effects. General PIT effects occur when positively valued environmental Pavlovian stimuli enhance instrumental responding regardless of the reinforcer being used. Here, stimuli that are associated with e.g. monetary rewards can enhance instrumental responding for alcohol drinks, demonstrating a general motivational effect. In contrast, during outcome-specific PIT, this effect is limited for the same stimulus-reward association. For instance, Pavlovian stimuli associated with a soft drink promote choices for the same soft drink outcome only. Methods: Between-group Analysis: The investigators will assess the outcome-specific PIT effects by measuring the change in the proportion of alcohol choices when the alcohol cue is presented versus when the soft drink (juice) cue is shown. On the other hand, the investigators will define general PIT effects as the increase in button presses (i.e., slope) corresponding to the increased valence of the conditioned monetary cues. The investigators plan to examine the differences in both specific and general PIT effects at both the behavioral and neural levels between the AUD group and the non-AUD group, controlling for smoking status. Additionally, the investigators will investigate the relationship between specific and general PIT effects and AUDIT, along with quantity-frequency measures Comparison between the two PIT tasks: the investigators will explore similarities and differences in PIT effects between the full PIT paradigm and the previously-mentioned single-lever PIT paradigm. Stress: the investigators will test whether levels of chronic stress (assessed with questionnaires and hair cortisol as for Experiment 1) moderate outcome-specific and full PIT effects. Hypothesis 3a: The investigators expect that both outcome-specific and general PIT effects will be stronger in participants with AUD compared to those without AUD, evident at both behavioral and neural levels. Additionally, the investigators predict positive associations between PIT effects and both AUDIT scores and quantity-frequency measures. Hypothesis 3b: the investigators hypothesize that the PIT effect found in the single-lever PIT task is more strongly associated with the general PIT and less strongly associated with the outcome-specific PIT effect in the full PIT task. Hypothesis 3c: Increased chronic stress will enhance effects of both general and outcome-specific PIT.