Impact of a pre-test measurement of alcohol craving in cue-exposure studies: Relationship with social desirability and demand effects.

Craving is one of the most studied concepts in the field of addiction. It is often investigated with repeated-measure experimental designs using self-reported scales. However, the explicit nature of self-reported craving scales may make them vulnerable to social desirability and demand effects. The aim of the present study was to test whether a pre-experimental measurement of craving affects its post-experimental assessment after an alcohol video exposure and whether these changes relate to social desirability, demand effects, and alcohol consumption. Seventy-five healthy volunteers aged 18-30 years were randomly assigned to two experimental groups: a pre-post and a post-only craving assessment group. They were invited to watch an alcohol-related video. Social desirability, demand effects, engagement toward the video, and severity of alcohol consumption were assessed in all participants. The results showed a significant effect of alcohol consumption (p < .001, η2p = .09) on post-experimental craving. The main effect of the repeated measure of craving was also significant (p = .04, η2p = .001), together with the interaction between these two variables (p = .03, η2p = .06). The pre-experimental craving measurement increased its post-experimental levels, but only in heavy drinkers. However, no evidence was found that these changes were related to social desirability or demand effects. Additional exploratory analyses indicated that craving increase in high alcohol drinkers was mediated by a stronger engagement toward the alcohol-related video. In summary, the repeated measurement of craving with explicit scales has a significant impact on the results of alcohol craving studies and may not always be desirable, especially when the true purpose of the study needs to be hidden from the participants. However, the present results also suggest that the pre-post experimental design is advisable when the experimenters seek to maximize the relationship between the individual levels of alcohol consumption and alcohol cue-exposure craving changes.