When Habits Are Dangerous: Alcohol Expectancies and Habitual Decision Making Predict Relapse in Alcohol Dependence

doi:10.1016/j.biopsych.2017.04.019

Biological Psychiatry

Volume 82, Issue 11, 1 December 2017, Pages 847-856

https://doi.org/10.1016/j.biopsych.2017.04.019 Get rights and content

Abstract

Background

Addiction is supposedly characterized by a shift from goal-directed to habitual decision making, thus facilitating automatic drug intake. The two-step task allows distinguishing between these mechanisms by computationally modeling goal-directed and habitual behavior as model-based and model-free control. In addicted patients, decision making may also strongly depend upon drug-associated expectations. Therefore, we investigated model-based versus model-free decision making and its neural correlates as well as alcohol expectancies in alcohol-dependent patients and healthy controls and assessed treatment outcome in patients.

Methods

Ninety detoxified, medication-free, alcohol-dependent patients and 96 age- and gender-matched control subjects underwent functional magnetic resonance imaging during the two-step task. Alcohol expectancies were measured with the Alcohol Expectancy Questionnaire. Over a follow-up period of 48 weeks, 37 patients remained abstinent and 53 patients relapsed as indicated by the Alcohol Timeline Followback method.

Results

Patients who relapsed displayed reduced medial prefrontal cortex activation during model-based decision making. Furthermore, high alcohol expectancies were associated with low model-based control in relapsers, while the opposite was observed in abstainers and healthy control subjects. However, reduced model-based control per se was not associated with subsequent relapse.

Conclusions

These findings suggest that poor treatment outcome in alcohol dependence does not simply result from a shift from model-based to model-free control but is instead dependent on the interaction between high drug expectancies and low model-based decision making. Reduced model-based medial prefrontal cortex signatures in those who relapse point to a neural correlate of relapse risk. These observations suggest that therapeutic interventions should target subjective alcohol expectancies.

Section snippets

Participants

All data were collected as part of the Learning and Alcohol Dependence study, a bicentric German study hosted at Universitätsklinikum Dresden/Technische Universität Dresden and Charité–Universitätsmedizin Berlin. Two hundred two subjects (106 alcohol-dependent patients, 96 healthy control subjects [HCs]) completed the two-step task (39) to disentangle habitual from goal-directed decision making and the brief German version of the AEQ (27). Patients fulfilled diagnostic criteria for AD according

Sample Characteristics

Compared to HCs, abstainers and relapsers reported significantly higher symptoms in almost all clinical characteristics, increased deficits in neuropsychological testing, and increased blood parameters related to alcohol consumption (Table 1).

Matching of HCs and alcohol-dependent patients was successful in all variables of interest (gender, school education, smoking status, and age). At baseline, there were no significant differences between abstainers and relapsers, except that the patients in

Discussion

The main findings of our study are 1) a reduction in mPFC activation during model-based behavior in relapsers and that 2) an interaction between alcohol expectancies and goal-directed control distinguishes relapsers from abstainers and HCs. Reductions in goal-directed behavior per se were not significantly associated with AD or relapse. Instead, relapsers had high alcohol expectancies in association with low goal-directed behavior and vice versa, suggesting that the interaction between alcohol

Acknowledgments and Disclosures

This work was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft Forschergruppe 1617 Grant Nos. HE2597/14-1 and HE2597/14-2 [to AH], RA1047/2-1 and RA1047/2-2 [to MAR], SM 80/7-1 and SM 80/7-2 [to MNS], and ZI1119/3-1 and ZI1119/3-2 [to USZ]).

We thank the Learning and Alcohol Dependence study teams in Berlin and Dresden for behavioral and neuroimaging data acquisition and Claudia Haegele, Katharina Scholz, and Anna-Maria Walter for collection of the follow-up data in

References (76)

S.N. Katner et al.
Reinstatement of alcohol-seeking behavior by drug-associated discriminative stimuli after prolonged extinction in the rat
Neuropsychopharmacology
(1999)
S.B. Flagel et al.
Antecedents and consequences of drug abuse in rats selectively bred for high and low response to novelty
Neuropharmacology
(2014)
R.W. Wiers et al.
Automatic and controlled processes and the development of addictive behaviors in adolescents: A review and a model
Pharmacol Biochem Behav
(2007)
H. Gundersen et al.
Separating the effects of alcohol and expectancy on brain activation: An fMRI working memory study
Neuroimage
(2008)
N.D. Daw et al.
Model-based influences on humans’ choices and striatal prediction errors
Neuron
(2011)
R.C. Oldfield
The assessment and analysis of handedness: The Edinburgh inventory
Neuropsychologia
(1971)
S.W. Lee et al.
Neural computations underlying arbitration between model-based and model-free learning
Neuron
(2014)
J. Glascher et al.
States versus rewards: Dissociable neural prediction error signals underlying model-based and model-free reinforcement learning
Neuron
(2010)
J. Ashburner et al.
Voxel-based morphometry—The methods
Neuroimage
(2000)
J.L. Gowin et al.
Individualized relapse prediction: Personality measures and striatal and insular activity during reward-processing robustly predict relapse
Drug Alcohol Depend
(2015)

E. Koya et al.

Role of ventral medial prefrontal cortex in incubation of cocaine craving

Neuropharmacology

(2009)

V. Vengeliene et al.

The alcohol deprivation effect model for studying relapse behavior: A comparison between rats and mice

Alcohol

(2014)

B.J. Everitt et al.

Neural systems of reinforcement for drug addiction: from actions to habits to compulsion

Nat Neurosci

(2005)

B.J. Everitt et al.

Drug addiction: Updating actions to habits to compulsions ten years on

Annu Rev Psychol

(2016)

K.C. Berridge

From prediction error to incentive salience: Mesolimbic computation of reward motivation

Eur J Neurosci

(2012)

N.D. Volkow et al.

Drug addiction: The neurobiology of behaviour gone awry

Nat Rev Neurosci

(2004)

A. Heinz et al.

Correlation between dopamine D(2) receptors in the ventral striatum and central processing of alcohol cues and craving

Am J Psychiatry

(2004)

T.H. McKim et al.

Addiction history associates with the propensity to form habits

J Cogn Neurosci

(2016)

V. Voon et al.

Disorders of compulsivity: A common bias towards learning habits

Mol Psychiatry

(2014)

K.D. Ersche et al.

Carrots and sticks fail to change behavior in cocaine addiction

Science

(2016)

M. Sebold et al.

Model-based and model-free decisions in alcohol dependence

Neuropsychobiology

(2014)

Z. Sjoerds et al.

Behavioral and neuroimaging evidence for overreliance on habit learning in alcohol-dependent patients

Transl Psychiatry

(2013)

S.T. Tiffany et al.

A cognitive processing model of alcohol craving and compulsive alcohol use

Addiction

(2000)

R.W. Wiers et al.

Retraining automatic action tendencies changes alcoholic patients’ approach bias for alcohol and improves treatment outcome

Psychol Sci

(2011)

S. Killcross et al.

Coordination of actions and habits in the medial prefrontal cortex of rats

Cereb Cortex

(2003)

S. de Wit et al.

Differential engagement of the ventromedial prefrontal cortex by goal-directed and habitual behavior toward food pictures in humans

J Neurosci

(2009)

J.P. O’Doherty

Contributions of the ventromedial prefrontal cortex to goal-directed action selection

Ann N Y Acad Sci

(2011)

S.B. Ostlund et al.

Lesions of medial prefrontal cortex disrupt the acquisition but not the expression of goal-directed learning

J Neurosci

(2005)

W.H. Alexander et al.

Medial prefrontal cortex as an action-outcome predictor

Nat Neurosci

(2011)

R.Z. Goldstein et al.

Dysfunction of the prefrontal cortex in addiction: Neuroimaging findings and clinical implications

Nat Rev Neurosci

(2011)

S.M. Grüsser et al.

Cue-induced activation of the striatum and medial prefrontal cortex is associated with subsequent relapse in abstinent alcoholics

Psychopharmacology

(2004)

A. Beck et al.

Effect of brain structure, brain function, and brain connectivity on relapse in alcohol-dependent patients

Arch Gen Psychiatry

(2012)

K. Charlet et al.

Increased neural activity during high working memory load predicts low relapse risk in alcohol dependence

Addict Biol

(2014)

J.M. Barker et al.

Low prefrontal PSA-NCAM confers risk for alcoholism-related behavior

Nat Neurosci

(2012)

A.R. Otto et al.

The curse of planning: Dissecting multiple reinforcement-learning systems by taxing the central executive

Psychol Sci

(2013)

Q.J.M. Huys et al.

Bonsai trees in your head: How the Pavlovian system sculpts goal-directed choices by pruning decision trees

PLoS Comput Biol

(2012)

S.A. Brown

Expectancies versus background in the prediction of college drinking patterns

J Consult Clin Psychol

(1985)

B.C. Leigh

Attitudes and expectancies as predictors of drinking habits: A comparison of three scales

J Stud Alcohol

(1989)

Cited by (121)

The role of decision-making and impulsivity in beverage consumption
2024, Appetite
Soft drink and alcohol consumption have become significant public health issues. This study aimed to explore the mediating role of decision-making processes in the relationship between trait impulsivity and calorie dense beverage consumption. Participants comprised a community sample of 300 adults (aged 19–75). They completed self-report measures assessing impulsivity (SUPPS-P), reward sensitivity (RST-PQ), and participated in decision-making tasks related to risk propensity (BART), short-term strategy preference (IGT), and delay discounting rate. Beverage consumption was calculated using the BEVQ-15. Impulsivity was conceptualised within the framework of the two-factor model as consisting of rash impulsivity and reward sensitivity. Both facets of impulsivity were positively associated with both alcohol and soft drink consumption, and each independently predicted consumption of these beverages. Additionally, there was a significant interaction between rash impulsivity and reward sensitivity on soft drink consumption. Importantly, there were significant indirect effects of both rash impulsivity and reward sensitivity on soft drink consumption via delay discounting. The results support the logic of the two-factor model of impulsivity in the prediction of consumption of unhealthy beverages. Furthermore, the mediating role of delay discounting supports the hypothesis that personality traits can pass through to behaviour via decision-making processes. Further research should extend these findings to other consumption domains in both clinical and non-clinical populations.
Doing what's not wanted: Conflict in incentives and misallocation of behavioural control can lead to drug-seeking despite adverse outcomes
2023, Addiction Neuroscience
Despite being aware of negative consequences and wanting to quit, long-term addicts find it difficult to quit seeking and consuming drugs. This inconsistency between the (often compulsive) behavioural patterns and the explicit knowledge of negative consequences represents a cognitive conflict which is a central characteristic of addiction. Neurobiologically, differential cue-induced activity in distinct striatal subregions, as well as the dopamine connectivity spiraling from ventral striatal regions to the dorsal regions, play critical roles in compulsive drug seeking. The focus of this work is to illustrate the mechanisms that lead to a cognitive conflict and it’s impact on actions taken i.e. addictive choices. We propose an algorithmic model that captures how the action choices that the agent makes when reinforced with drug-rewards become impervious to the presence of negative consequences that often follow those choices. We advance the understanding of having a decision hierarchy in representing “cognitive control” and how lack of such control at higher-level in the hierarchy could potentially lead to consolidated drug-seeking habits. We further propose a cost-benefit based arbitration scheme, which mediates the allocation of control across different levels of the decision-making hierarchy. Lastly, we discuss how our algorithmic model could help us understand how addictive drugs progressively hijack the dopamine-spiralling circuit at the neural implementation level.
Subjective response to alcohol: Interactive effects of early life stress, parental risk for mood and substance use disorders, and drinking context
2023, Pharmacology Biochemistry and Behavior
Early life stress, specifically childhood maltreatment, and parental risk for mood and substance use disorders (SUDs) are associated with increased risk for alcohol use disorder (AUD). There is limited data on how these factors interact to contribute to alcohol-related outcomes. Prior work has suggested early life stress may increase sensitivity to psychostimulants and that subjective response to alcohol is heritable. It is unclear if early life stress alters sensitivity to alcohol and interacts with parental risk for mood/SUDs, which in turn may act as a risk factor for AUD. The current study uses within-subjects placebo-controlled alcohol administration methods to investigate the effects of childhood maltreatment on subjective response to alcohol in young adults with and without parental risk of mood/SUDs. Additionally, we explored interactions with drinking context (i.e., drinking in a bar vs. non-bar context). Within individuals with parental risk for mood/SUDs, there was a positive relation between total Childhood Trauma Questionnaire (CTQ) score and how drunk individuals reported feeling across both alcohol and placebo conditions (parental risk group-by-CTQ interaction p = .01; main effect of CTQ within individuals with parental risk for mood/SUDs p = .005). When exploring interactions with drinking context (bar vs. non-bar context), we observed a significant drinking context-by-parental risk-by-CTQ interaction (p = .03), with CTQ score positively associated with greater positive valence/positive arousal feelings in the parental risk group if they consumed their beverages in the bar context (p = .004) but not if they consumed their beverages in the non-bar context. Results suggest childhood maltreatment may contribute to variation in subjective response to the positive effects of alcohol—possibly mediated by alcohol cues and/or expectancies—in young adults with parental risk for mood/SUDs.
The differential neural substrates for reward choice under gain-loss contexts and risk in alcohol use disorder: Evidence from a voxel-based meta-analysis
2023, Drug and Alcohol Dependence
Making a risky decision is a complex process that involves the evaluation of both the values of the options and the associated risk level; this process is distinct from reward processing in gain versus loss contexts. Although disrupted reward processing in mesolimbic dopamine circuitry is suggested to underlie pathological incentive processing in patients with alcohol use disorder (AUD), the differential neural processes subserving these motivational tendencies for risk situations or gain/loss choices in decision-making have not been identified.
To examine the common or distinct neural mechanisms in the evaluation of risk versus outcomes for AUD, we conducted two separate coordinate-based meta-analyses of functional neuroimaging studies by using Seed-Based d Mapping software to evaluate 13 studies investigating gain and loss processing and 10 studies investigating risky decision-making.
During gain and loss processing, relative to healthy controls, AUD patients showed reduced activation in the mesocortical-limbic circuit, including the orbital prefrontal cortex (OFC), dorsal striatum, insula, hippocampus, cerebellum, cuneus cortex and superior temporal gyrus, but hyperactivation in the inferior temporal gyrus and paracentral lobule (extending to the middle cingulate cortex (MCC) and precuneus). During decision-making under risk, AUD patients exhibited hypoactivity of the prefrontal and cingulate cortices, including the posterior cingulate cortex (extending to the MCC), middle frontal gyrus, medial prefrontal cortex, dorsolateral prefrontal cortex, OFC and anterior cingulate cortex.
Our results extend existing neurological evidence by showing that a reduced response in the mesocortical-limbic circuit is found in gain versus loss processing, with decreased responsivity in cortical regions in risk decision-making. Our results implicate dissociable neural circuit responses for gain-loss processing and risk decision-making, which contribute to a better understanding of the pathophysiological mechanism underlying nondrug incentive and risk processing in individuals with AUD.
Individual differences in computational psychiatry: A review of current challenges
2023, Neuroscience and Biobehavioral Reviews
Bringing precision to the understanding and treatment of mental disorders requires instruments for studying clinically relevant individual differences. One promising approach is the development of computational assays: integrating computational models with cognitive tasks to infer latent patient-specific disease processes in brain computations. While recent years have seen many methodological advancements in computational modelling and many cross-sectional patient studies, much less attention has been paid to basic psychometric properties (reliability and construct validity) of the computational measures provided by the assays. In this review, we assess the extent of this issue by examining emerging empirical evidence. We find that many computational measures suffer from poor psychometric properties, which poses a risk of invalidating previous findings and undermining ongoing research efforts using computational assays to study individual (and even group) differences. We provide recommendations for how to address these problems and, crucially, embed them within a broader perspective on key developments that are needed for translating computational assays to clinical practice.
Goal-directed and habitual decision making under stress in gambling disorder: An fMRI study
2023, Addictive Behaviors
The development of addictive behaviors has been suggested to be related to a transition from goal-directed to habitual decision making. Stress is a factor known to prompt habitual behavior and to increase the risk for addiction and relapse. In the current study, we therefore used functional MRI to investigate the balance between goal-directed ‘model-based’ and habitual ‘model-free’ control systems and whether acute stress would differentially shift this balance in gambling disorder (GD) patients compared to healthy controls (HCs). Using a within-subject design, 22 patients with GD and 20 HCs underwent stress induction or a control condition before performing a multistep decision-making task during fMRI. Salivary cortisol levels showed that the stress induction was successful. Contrary to our hypothesis, GD patients did not show impaired goal-directed ‘model-based’ decision making, which remained similar to HCs after stress induction. Bayes factors provided three times more evidence against a difference between the groups or a group-by-stress interaction on the balance between model-based and model-free decision making. Similarly, no differences were found between groups and conditions on the neural estimates of model-based or model-free decision making. These results challenge the notion that GD is related to an increased reliance on habitual (or decreased goal-directed) control, even during stress.

View all citing articles on Scopus

View full text

Archival ReportWhen Habits Are Dangerous: Alcohol Expectancies and Habitual Decision Making Predict Relapse in Alcohol Dependence

Abstract

Background

Methods

Results

Conclusions

Section snippets

Participants

Sample Characteristics

Discussion

Acknowledgments and Disclosures

Neuropsychopharmacology

Neuropharmacology

Pharmacol Biochem Behav

Neuroimage

Neuron

Neuropsychologia

Neuron

Neuron

Neuroimage

Drug Alcohol Depend

Neuropharmacology

Alcohol

Neural systems of reinforcement for drug addiction: from actions to habits to compulsion

Nat Neurosci

Drug addiction: Updating actions to habits to compulsions ten years on

Annu Rev Psychol

From prediction error to incentive salience: Mesolimbic computation of reward motivation

Eur J Neurosci

Drug addiction: The neurobiology of behaviour gone awry

Nat Rev Neurosci

Correlation between dopamine D(2) receptors in the ventral striatum and central processing of alcohol cues and craving

Am J Psychiatry

Addiction history associates with the propensity to form habits

J Cogn Neurosci

Disorders of compulsivity: A common bias towards learning habits

Mol Psychiatry

Carrots and sticks fail to change behavior in cocaine addiction

Science

Model-based and model-free decisions in alcohol dependence

Neuropsychobiology

Behavioral and neuroimaging evidence for overreliance on habit learning in alcohol-dependent patients

Transl Psychiatry

A cognitive processing model of alcohol craving and compulsive alcohol use

Addiction

Retraining automatic action tendencies changes alcoholic patients’ approach bias for alcohol and improves treatment outcome

Psychol Sci

Coordination of actions and habits in the medial prefrontal cortex of rats

Cereb Cortex

Differential engagement of the ventromedial prefrontal cortex by goal-directed and habitual behavior toward food pictures in humans

J Neurosci

Contributions of the ventromedial prefrontal cortex to goal-directed action selection

Ann N Y Acad Sci

Lesions of medial prefrontal cortex disrupt the acquisition but not the expression of goal-directed learning

J Neurosci

Medial prefrontal cortex as an action-outcome predictor

Nat Neurosci

Dysfunction of the prefrontal cortex in addiction: Neuroimaging findings and clinical implications

Nat Rev Neurosci

Cue-induced activation of the striatum and medial prefrontal cortex is associated with subsequent relapse in abstinent alcoholics

Psychopharmacology

Effect of brain structure, brain function, and brain connectivity on relapse in alcohol-dependent patients

Arch Gen Psychiatry

Increased neural activity during high working memory load predicts low relapse risk in alcohol dependence

Addict Biol

Low prefrontal PSA-NCAM confers risk for alcoholism-related behavior

Nat Neurosci

The curse of planning: Dissecting multiple reinforcement-learning systems by taxing the central executive

Psychol Sci

Bonsai trees in your head: How the Pavlovian system sculpts goal-directed choices by pruning decision trees

PLoS Comput Biol

Expectancies versus background in the prediction of college drinking patterns

J Consult Clin Psychol

Attitudes and expectancies as predictors of drinking habits: A comparison of three scales

J Stud Alcohol

Archival Report
When Habits Are Dangerous: Alcohol Expectancies and Habitual Decision Making Predict Relapse in Alcohol Dependence