Processing of alcohol-related health threat in at-risk drinkers: an online study of gender-related self-affirmation effects

Aims: Defensiveness in response to threatening health information related to excessive alcohol consumption prevents appropriate behaviour change. Alternatively, self-affirmation may improve cognitive-affective processing of threatening information, thus contributing to successful self-regulation. Methods: Effects of an online self-affirmation procedure were examined in at-risk university student drinkers. Participants were randomly assigned to a self-affirmation (writing about personally relevant values) or control task (writing about values relevant to another person) prior to presentation of alcohol-related threatening information. Assessment of prosocial feelings (e.g. ‘love’) after the task served as a manipulation check. Generic and personalized information regarding the link between alcohol use and cancer was presented, followed by assessment of perceived threat, message avoidance and derogation. Page dwell-times served as indirect indices of message engagement. Alcohol consumption and intention to drink less were assessed during the first online session and at 1-week and 1-month follow-up. Results: Although self-affirmation resulted in higher levels of prosocial feelings immediately after the task, there was no effect on behaviour in the self-affirmation group. Effects on intention were moderated by gender, such that men showed lower intention immediately after self-affirmation, but this increased at 1-week follow-up. Women's intention to reduce consumption in the self-affirmation group reduced over time. Trend-level effects on indices of derogation and message acceptance were in the predicted direction only in men. Conclusion: It is feasible to perform self-affirmation procedures in an online environment with at-risk drinkers. However, use of internet-based procedures with this population may give rise to (gender-dependent) effects that are substantially diluted compared with lab-based experiments