Alcohol-related learning (both instrumental and classical) appears to be an important factor in accounting for alcohol consumption variability. The way in which alcohol-related information, stored in memory, is processed is likely to be similar in individuals at different points of the continuum of consumption. (Indeed, it is a feature of social learning and social cognition theory that there is continuity of process: Maisto, Carey & Bradizza, 1999.) This means that there should be value in extending the use of the cue-reactivity paradigm from the pole of dependent (clinical) consumption, through the (subclinical) regions where there is no longer dependence but there are still consumption-related problems, to the problem-light (nonclinical) pole. There should be value that goes beyond simply providing a baseline control against which cue-reactivity at the dependent pole can be assessed. The current experiment is designed to add to the knowledge provided by those researchers who have recognized this (e.g. Walitzer & Sher, 1990; Greeley et al, 1993) by increasing both the sensitivity and the range of the dependent variables they have used.
Design: A 2 x 2 between-subjects design was used for the cue exposure experiment, with 'drink cue' (two levels: soft (S) and alcoholic (A)) and 'order' (two levels: 'desire' questionnaire first (DE) and 'expectancy' questionnaires first (ED)) as the between-subject factors and with random assignment of subjects to cells. This crossing generated four independent groups each with 22 participants. During cue exposure participants completed a 'taste preference' questionnaire, and were then assessed on the dependent variables, desire to drink alcohol and alcohol outcome expectancies.
Measures and cue exposure stimuli: For the cue exposure, participants had a choice of one from either three soft drinks (cans) or three alcoholic drinks (bottles) according to group allocation. The 'taste preference' questionnaire, applicable to both soft and alcoholic drinks, took participants systematically through the stages of cue exposure (sight, smell, taste), systematically directing the participants' attention to the drink cues by asking a range of questions about the drink. The cue exposure/ taste preference was designed to last 10 minutes. Desires/urges were assessed with a modified version of the Desire for Alcohol Questionnaire (DAQ: Love et al., 1998), which consists of 14 items on four subscales: (i) strong desires and intentions; (ii) negative reinforcement; (iii) control over drinking; and (iv) mild desires to drink. Responses were measured on a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Alcohol expectancies were assessed using the Negative Alcohol Expectancy Questionnaire (NAEQ: Jones & McMahon, 1994, 1995) and the positive Alcohol Expectancy Questionnaire (AEQ: Brown et al., 1987). The NAEQ has three subscales representing 60 items of negative expectancy that (a) surround the period of consumption (subscale 'same day'), that (b) relate to the following day (subscale 'next day') and that (c) relate to the longer term (months and years) should consumption continue at the current level (subscale 'continued drinking'). The AEQ has six subscales that represent 64 items of positive effects of drinking: 'global positive change', 'physical and social pleasure', 'sexual enhancement', 'social assertiveness', 'relaxation and tension reduction' and 'arousal and aggression'. Responses to NAEQ and AEQ items were measured on five-point Liken scales ranging from 1 ('highly unlikely') to 3 ('possible') to 5 ('highly likely'). An adaptation of the Timeline Follow-back procedure (Sobell & Sobell, 1992) was used to collect information on the previous week's alcohol consumption and derive the number of units of alcohol consumed in the week. Demographic measures were taken.
Procedure: Before the experiment, participants were informed that the 'taste preference' experiment may or may not involve drinking up to a unit of alcohol (8 g of absolute alcohol). They agreed to take part by signing a consent form.
Participants were assigned randomly to groups and were invited to make a choice of their preferred drink. They then completed the drinking diary and demographic questions, following which they were asked to open their drink, smell it, pour part of it into the provided glass, smell it again and then sample it while completing the stooge taste preference questionnaire. It was explained that repeatedly pouring, smelling and sampling the drink would help them fill in the taste questionnaire most appropriately. Once the taste preference was completed, participants were asked to complete the desire (DAQ) and expectancy questionnaires (NAEQ/AEQ). Half the participants filled in the DAQ first (approximately 3 minutes), the other half filled in the expectancy questionnaires first (approximately 20 minutes)--the order of NAEQ and AEQ was counterbalanced. There was no time limit set for completing these questionnaires and participants could consume as little or as much of their drink as they wished (up to I unit of alcohol for alcoholic drinks). On completion of the questionnaires, participants were debriefed and paid their fee.
Within the context of the current experiment, cue manipulation reveals alcohol cue-reactivity in social drinkers for desire for alcohol (DAQ total score). This adds to the findings of a small number of other studies that have shown alcohol cue-reactivity in social drinkers with respect to desire (Walitzer & Sher, 1990; Greeley et a., 1993; Schulze & Jones, 1999) and in response to similar cues used in the current study (sight, smell and taste of alcoholic drink). However, in conjunction with Schulze & Jones's earlier finding, the current study is unique in adopting a multi-dimensional representation of desire in exploring the alcohol cue response in social drinkers. The fact that the total desire score manifests cue-reactivity is not inconsistent with Glautier & Tiffany's (1995) claim that uni-dimensional representations are likely to be insensitive and relatively ineffectual for exploring desire reactivity, for the DAQ total score explicitly integrates data from a number of different dimensions of desire (i.e. the items of the DAQ subscales) onto a single representation-compared with a single subjective judgement cast onto either a numerical (e.g. McCusker & Brown, 1990) or analogue scale (e.g. Greeley et al., 1993).
The finer-grain analysis shows that cognitions represented by the subscale 'strong intentions and desires' also react to the alcohol cues and this replicates Schulze & Jones's (1999) earlier finding with this subscale. In their report they noted that two of the DAQ's four subscales comprised principally positive alcohol outcome expectancies: 'mild desires' comprised positive outcome expectancies that were positively reinforcing; 'negative reinforcement' comprised positive outcome expectancies that were negatively reinforcing. They also showed that the subscale 'mild desires' was alcohol cue-reactive (an increase in desire) but the subscale 'negative reinforcement' was not. This differential reactivity on positive expectancies supports Glautier & Tiffany's (1995) criticism that subjective alcohol cue-reactivity research requires a more fine-grain representation if theory is to be advanced--for a single item representation such as the DAQ total or a single numerical or analogue judgement (as described above) would not have captured this. However, although the differential reactivity was replicated in the current experiment, the direction of the difference was reversed. The lack of any significant alcohol cue response with the explicit positive expectancy construct (the AEQ total or the six AEQ subscales) in the current experiment does not help resolve this inconsistency and the extent to which positive alcohol expectancies might be alcohol cue-reactive needs further investigation. Just as alcohol cue-reactivity was not shown by the positive expectancy assessments (AEQ), it was also not shown by the negative expectancy assessments (NAEQ). One conclusion is that alcohol outcome expectancies are relatively stable, representing trait rather than state constructs (they are, after all, derived from memory structures that are the result of prolonged learning). However, Darkes & Goldman (1998) have shown that positive expectancies represented by the AEQ can, indeed, be changed in the relatively short term (hours, using an 'expectancy challenge' procedure) and Jones & McMahon (1998) have shown the same for negative expectancies represented by the NAEQ (days, using a 'motivational interviewing' procedure). This, coupled with the finding that the (implicit) positive expectancy components of the DAQ have demonstrated alcohol cue-reactivity in the current experiment and Schulze & Jones's (1999) earlier experiment, suggests that the state nature of expectancies within cue-reactivity paradigms warrants further investigation.
It has been one of the goals of contemporary alcohol researchers (basic and applied) to develop representations of alcohol motivations that integrate all relevant information into a 'final common path' (as, for example, described by Cox & Klinger, 1987). In the same vein (although within a much more limited framework), there has been a recognition of the need for a theoretical integration of learning manifest as cue-reactivity and learning manifest as outcome expectancy (e.g. Glautier & Remington, 1995). A better understanding of the variability in alcohol consumption might come from the collaboration of expectancy and cue-reactivity approaches--and within a single common framework provided by cognitive psychology, associative networks (Glautier & Spencer 1999).
- Schulze, Daniela; Jones, Barry T.; Desire for alcohol and outcome expectancies as measures of alcohol cue-reactivity; Addiction, Jul2000, Vol. 95 Issue 7
Reflection Exercise #4
The preceding section contained information
about helping clients understand cue reactivity and triggers. Write three case study examples
regarding how you might use the content of this section in your practice.
What are three cues or triggers for increased alcohol desire that clients on a controlled drinking program should be aware of?
Record the letter of the correct answer the Test