Acute alcohol impairs human goal-directed action

https://doi.org/10.1016/j.biopsycho.2012.02.016Get rights and content

Abstract

There are two forms of motivated behaviour. Goal-directed action is mediated by knowledge of the consequences whereas habitual action is elicited directly by stimuli associated with the action. Alcohol may impair goal-directed control, favouring habit. To evaluate this proposal, participants were administered with 0.4 g/kg of alcohol or placebo before acquiring separate instrumental responses for chocolate and water points. Chocolate was then fed to satiety to devalue this outcome before choice between the two responses was tested in extinction. Any reduction in chocolate choice must be mediated by knowledge of the current incentive value of this outcome, i.e. must be goal-directed. Alcohol attenuated the devaluation effect on choice in extinction, but had no effect on reacquisition performance, the hedonic appraisal of rewards or acquisition of the instrumental contingencies. Acute alcohol impaired goal-directed control of action selection, favouring habit, which may mediate alcohol effects on under-controlled behaviour more broadly.

Highlights

► Motivated behaviour is the sum of goal-directed choice and automatic habits. ► Alcohol may impair goal-directed control in favour of habit. ► Alcohol impaired human goal-directed action in the outcome devaluation protocol. ► This impairment may drive under-controlled behaviour more broadly.

Introduction

Acute alcohol intoxication produces a broad range of effects. These include physiological changes, for example hypothermia (Crowell et al., 1981) and analgesia (Woodrow and Eltherington, 1988). Alcohol also changes emotional state, specifically, elevating positive and negative mood as blood alcohol ascends and descends respectively (King et al., 2002, Sutker et al., 1983), reduces anxiety (Josephs and Steele, 1990, Pohorecky, 1981) and increases aggression (Giancola, 2000). Alcohol also impairs learning, for instance, extinction (Baum, 1971) and suppression (Conger, 1951, Vogel-Sprott, 1967) in rats and cognitive performance in humans (Fillmore, 2007, Schweizer et al., 2005). Effects of acute alcohol have also been found on risky behaviour (Lane et al., 2004), impulsivity (Reynolds et al., 2006) and susceptibility to accidents (Fillmore et al., 2008, Li et al., 1997). Finally, acute alcohol enhances ingestive behaviour, including alcohol desire and self-administration (Fillmore, 2001, Fillmore and Rush, 2001, Rose and Grunsell, 2008), reinstatement and relapse to drug use (Lê et al., 1998, Shaham et al., 2003, Shiffman et al., 1996), food consumption (Caton et al., 2004) and sexual behaviour (Wilson, 1977).

A perennial question addressed by alcohol research concerns the psychological process that mediates the increase in risky, injurious and ingestive behaviour. Candidate mechanisms include changes in attention, disinhibition, executive function, impulsivity, reward sensitivity, etc. Whilst these candidates are all plausible mediators, there is little empirical basis to favour one over another, perhaps because these constructs are not consistently defined and operationalised across experiments and research groups. In contrast, associative learning theory has recently substantiated a model of motivated behaviour with which the psychological effect of alcohol may be specified. On this account, motivated behaviour is controlled by two dissociable learning processes (Balleine and O’Doherty, 2010, de Wit and Dickinson, 2009, Killcross and Coutureau, 2003). Goal-directed or intentional instrumental behaviour is mediated by explicit knowledge of the instrumental contingency between response (R) and receipt of the outcome (O), combined with knowledge of the incentive value of the O, which provides evaluative feedback to determine the propensity to perform the R. By contrast, habitual instrumental behaviour is elicited directly by stimuli (S) which have become associated with the response (R) through contingent reinforcement (S–R/reinforcement), without retrieving a representation of the outcome. These two forms of learning, goal-directed vs. habit, can be regarded as largely synonymous with the controlled-automatic, implicit–explicit, or executive/non-executive dual-process theories common to cognitive psychology (Shanks, 2007).

The outcome devaluation protocol is the main empirical tool for studying goal-directed and habitual action (Dickinson, 1985). In a common instantiation of this method, rats first learn that two instrumental responses earn distinct outcomes (e.g. pellets and sucrose, respectively), before one outcome is devalued by specific satiety or taste aversion conditioning in a separate context. In the test phase that follows, rats again have the opportunity to perform the two instrumental responses, but this time in extinction, that is, where the two responses no longer produce their respective outcomes. This extinction test is critical to ensure that choice cannot be influenced by direct feedback from the outcomes (i.e. S–R/reinforcement). Instead, a selective reduction of responding for the devalued outcome indicates that the response is goal-directed, i.e. is controlled by knowledge of the R–O contingencies established in training combined with knowledge of the differential incentive value of the two Os which determine the propensity to select the two Rs. By contrast, equivalent responding for the devalued and non-devalued outcome indicates that these responses have become autonomous of the current incentive value of the outcomes, that is, they have become S–R habits elicited directly by the instrumental context without retrieving knowledge of the consequences.

Impairments in goal-directed control of action selection in the outcome devaluation protocol has been found in a number of ‘abnormal’ populations, including impulsive humans (Hogarth et al., 2011), children under 2.5 years of age (Klossek et al., 2008), patients with Parkinson's disease (de Wit et al., 2011), as well as normal participants exposed to conflict (de Wit et al., 2007) and stress (Schwabe et al., 2011). Furthermore, in animals, goal-directed action is impaired by chronic psychostimulant exposure (Nelson and Killcross, 2006, Zapata et al., 2010), lesions of the prelimbic cortex (Corbit and Balleine, 2003, Killcross and Coutureau, 2003) and dorsomedial striatum (Yin et al., 2005).

Three publications have suggested that alcohol may impair goal-directed control of action selection. First, Dickinson et al. (2002) trained rats to acquire two instrumental responses, one for alcohol and one for food pellets, before one of these outcomes was devalued by pairing it with lithium chloride induced sickness. When the rats were again given the opportunity to respond for these outcomes in extinction, it was found that performance of the food-seeking response was reduced by the devaluation treatment, indicating that food-seeking was goal-directed. By contrast, performance of the alcohol-seeking response was insensitive to devaluation suggesting that alcohol-seeking had become a habit; that is, had come to be elicited directly by the instrumental context without the animal retrieving a representation of the current incentive value of alcohol. Thus, alcohol-seeking appeared to be particularly prone to habit formation. Additionally, Corbit and Janak (in press) have measured the time course of the transition from goal-directed to habitual alcohol self-administration across training, and found that an intact dorsomedial and dorsolateral striatum are required the two forms of behavioural control, respectively, revealing the neural basis of dual-process addiction psychology.

Third, Ostlund et al. (2010) trained rats on a classic outcome devaluation protocol with natural rewards. However, the extinction test was conducted in either an alcohol- or placebo-paired chamber in separate sessions. The important result was that rats showed no evidence of goal-directed control over action selection in the alcohol paired chamber, but showed goal-directed control in the placebo paired chamber. The implication is that alcohol cues interfered with the retrieval of current goals, thereby abolishing the capacity to use goal-directed control of action selection in the extinction test.

The purpose of the current experiment was to test whether acute alcohol administration would attenuate goal-directed control in a human outcome devaluation procedure. Twenty four participants were randomly assigned to receive 0.4 g/kg of alcohol or placebo. Following a 10 min absorption period, participants were trained on a concurrent schedule in which pressing one key earned chocolate points whereas pressing a second key earned water points (see Hogarth et al., 2011). All participants then consumed chocolate to devalue this outcome before choice between the two keys was tested in extinction. Any reduction in choice of the chocolate key in this extinction test must be mediated by an integration of knowledge about the R–O contingencies acquired in concurrent training with knowledge of the differential incentive value of the two outcomes (chocolate vs. water) which determines the propensity to select the associated response, i.e. action selection must be goal-directed. On the basis of the aforementioned animal alcohol-devaluation studies (Corbit and Janak, in press, Dickinson et al., 2002, Ostlund et al., 2010), we would anticipate that acute alcohol will impair goal-directed control of action selection in this extinction test.

Three control tests were undertaken to exclude alternative explanations of any alcohol induced impairment of the devaluation effect in extinction. First, a reacquisition test was conducted in which the outcomes of the two responses were again delivered, identical to concurrent training. If the alcohol and placebo groups show an equivalent devaluation effect in this reacquisition test (in contrast to the extinction test), then demonstrably outcome delivery has ‘normalised’ the action control of the alcohol group. Such normalisation of action control could be achieved by experience of the outcomes enhancing retrieval of their relative values which facilitates goal-directed control, or experience of the outcomes could modify the capacity of contextual cues to prompt the two choices through S–R/reinforcement habit learning (see Klossek et al., 2008 experiment 2). Irrespective of these interpretational specifics, an equivalent devaluation effect in the reacquisition test clearly suggests that the two groups were equally sensitive to the devaluation treatment, and were equally engaged with the task. Second, participants reported their desire for and pleasantness of the outcomes before and after satiety to determine whether hedonic evaluation of the outcomes was equivalent between groups. Third, participants reported their knowledge of the R–O contingencies following training to assess whether acquisition of these contingencies was equivalent between the two groups. The question was whether alcohol would selectively impair goal-directed control in the extinction test, but not influence the devaluation effect in the reacquisition test, or the hedonic evaluation of outcomes, or acquisition of contingency knowledge. This pattern of results would suggest that alcohol produces a selective impairment in the use of knowledge of consequences to guide goal-directed choice, and a concomitant preponderance of habitual control. This shift in behavioural control could mediate the effect of alcohol on risk, injury and ingestion.

Section snippets

Participants

Twenty-four healthy participants (12 male) aged between 18 and 32 years (mean 21 years) were recruited. These participants were randomly allocated to the alcohol and placebo group, balancing for gender and response-outcome assignment within each group. Three participants were excluded because they failed to consume all of the chocolate in the satiety treatment (consuming only 63 g, 81 g and 94 g of the total 147 g of chocolate provided). Two were from the placebo group and one from the alcohol

Choice tests

Fig. 1 shows the percent choice of the chocolate vs. water key during concurrent training, extinction test and reacquisition test, in the alcohol and placebo groups. Visual inspection suggests that the groups differed selectively in the extinction test. This impression was confirmed by ANOVA which yielded a significant interaction between group (alcohol, placebo) and block (concurrent training, extinction test, reacquisition test), F(2,36) = 6.80, p = .003. When concurrent and extinction data were

Discussion

The key finding of this study was that acute alcohol impaired goal-directed control of action selection in the outcome devaluation protocol, but had no impact on choice in the reacquisition test, the hedonic evaluation of rewards or the acquisition of explicit instrumental contingency knowledge. To phrase this more specifically, the alcohol group failed to translate a decline in hedonic evaluation of chocolate into an intentional reduction in choice of this outcome in the extinction test. By

Acknowledgements

MRM is a member of the UK Centre for Tobacco Control Studies, a UKCRC Public Health Research Centre of Excellence. Funding from the Economic and Social Research Council, the British Heart Foundation, Cancer Research UK, the Department of Health and the Medical Research Council, under the auspices of the UK Clinical Research Collaboration, is gratefully acknowledged. This work was supported by Medical Research Council Grant (# G0701456 to Lee Hogarth).

References (57)

  • M.S. Stanford et al.

    Fifty years of the Barratt Impulsiveness Scale: an update and review

    Personality and Individual Differences

    (2009)
  • P.B. Sutker et al.

    Acute alcohol intoxication mood states and alcohol metabolism in women and men

    Pharmacology, Biochemistry, and Behavior

    (1983)
  • G.T. Wilson

    Alcohol and human sexual behavior

    Behaviour Research and Therapy

    (1977)
  • K.M. Woodrow et al.

    Feeling no pain: alcohol as an analgesic

    Pain

    (1988)
  • T.F. Babor et al.
  • B.W. Balleine et al.

    Human and rodent homologies in action control: corticostriatal determinants of goal-directed and habitual action

    Neuropsychopharmacology

    (2010)
  • J.J. Conger

    The effects of alcohol on conflict behavior in the albino rat

    Quarterly Journal of Studies on Alcohol

    (1951)
  • Corbit, L.H., Janak, P.H. Habitual alcohol seeking: time course and the contribution of subregions of the dorsal...
  • C.R. Crowell et al.

    The role of conditional drug responses in tolerance to the hypothermic effects of ethanol

    Psychopharmacology

    (1981)
  • S. de Wit et al.

    Habitual versus goal-directed action control in Parkinson disease

    Journal of Cognitive Neuroscience

    (2011)
  • S. de Wit et al.

    Associative theories of goal-directed behaviour: a case for animal–human translational models

    Psychological Research

    (2009)
  • S. de Wit et al.

    Stimulus-outcome interactions during instrumental discrimination learning by rats and humans

    Journal of Experimental Psychology: Animal Behavior Processes

    (2007)
  • A. Dickinson

    Actions and habits – the development of behavioral autonomy

    Philosophical Transactions of the Royal Society of London Series B – Biological Sciences

    (1985)
  • A. Dickinson et al.

    Alcohol seeking by rats: action or habit

    Quarterly Journal of Experimental Psychology

    (2002)
  • J.F. Etter et al.

    A self-administered questionnaire to measure dependence on cigarettes: the Cigarette Dependence Scale

    Neuropsychopharmacology

    (2003)
  • M.T. Fillmore

    Cognitive preoccupation with alcohol and binge drinking in college students: alcohol-induced priming of the motivation to drink

    Psychology of Addictive Behaviors

    (2001)
  • M.T. Fillmore

    Acute alcohol-induced impairment of cognitive functions: past and present findings

    International Journal on Disability and Human Development

    (2007)
  • M.T. Fillmore et al.

    Alcohol effects on inhibitory and activational response strategies in the acquisition of alcohol and other reinforcers: priming the motivation to drink

    Journal of Studies on Alcohol

    (2001)
  • Cited by (53)

    • Reduced Glutamate Turnover in the Putamen Is Linked With Automatic Habits in Human Cocaine Addiction

      2021, Biological Psychiatry
      Citation Excerpt :

      Other paradigms, such as the two-stage decision-making task or pavlovian-to-instrumental transfer procedures, may not be comparable, thus explaining some apparent inconsistencies in the literature. For example, increased habitual responding following outcome devaluation has been reported both in healthy individuals under the acute influence of alcohol (87) and in patients with alcohol use disorder (63) but not using the two-stage decision-making task (88) or a pavlovian-to-instrumental transfer paradigm (89). A variety of measures and experimental paradigms is thus warranted to understand the interplay between goal-directed and habitual control in health and disease.

    • Behavioral and neural mechanisms underlying habitual and compulsive drug seeking

      2018, Progress in Neuro-Psychopharmacology and Biological Psychiatry
      Citation Excerpt :

      Alcohol dependence was also associated with a bias toward habit learning, and functional imaging revealed a decreased recruitment of goal-directed brain areas and increased recruitment of habit-associated brain areas (Sjoerds et al., 2013). Furthermore, acute alcohol exposure in humans impaired goal-directed behavior and caused a bias for habit learning (Hogarth et al., 2012). The majority of animal studies investigating instrumental response strategies following a history of drug exposure have utilized instrumental tasks that involve food reward, with only a few studies investigating the influence on drug reward.

    • Understanding the balance between goal-directed and habitual behavioral control

      2018, Current Opinion in Behavioral Sciences
      Citation Excerpt :

      Investigation of the ways in which things like drug exposure alter the neural circuits controlling actions and habits may also yield insights into how these circuits interact normally. As an example, alcohol consumption and exposure causes habitual control to emerge more quickly [17,40,41,47]. Electrophysiological studies have shown that acute alcohol exposure decreases inhibitory transmission onto the medium spiny neurons of the DLS effectively disinhibiting them [48].

    View all citing articles on Scopus
    View full text