Abstract
Craving is an important feature of models of addiction. However a lack of empirical support for a relationship between craving and relapse has triggered much controversy and debate within the field of addiction about the function and utility of craving. Aim and objectives: The present study aimed to test two competing models of craving: craving as an unconditional response to alcohol cessation (i.e. withdrawal craving) and craving as a conditioned response to alcohol cues (i.e. cue-elicited craving). The main objectives were to explore how each type of craving varied over time and to evaluate the relationship between craving and a behavioural indicator of intention to drink. Design: The study used a within-subjects repeated-measures design to obtain serial measures of withdrawal and cue-elicited craving. Repeated measures of craving in the absence of cues were obtained to represent withdrawal craving. Craving was also manipulated in a cue exposure experiment in which all participants were exposed to the sight and smell of an alcohol cue and a neutral cue to obtain repeated measures of cue-elicited craving. Setting: The research was based at the National Inpatient Alcohol Addiction Unit at Springfield Hospital in South London. Participants: 34 patients undergoing detoxification treatment for alcohol dependency volunteered to participate in the study. Measurements: Withdrawal craving was measured on Day 3, Day 4 and Day 12 of participant’s hospital admission. Cue-elicited craving, skin conductance and a behavioural measure of intention to drink were monitored during cue exposure trials conducted on Day 4 and Day 12. Findings: Withdrawal craving attenuated with the passage of time, whereas time by itself, did not impact on cue-elicited craving. Only cue-elicited craving was significantly related to a behavioural measure of intention to drink. Conclusions: Viewing craving as occurring under two conditions may provide a useful distinction between craving that has no predictive value particularly once the withdrawal phase has ended and cue-elicited craving with predictive value in relapse.