BACKGROUND: Only one laboratory study has examined the relation between stress and alcohol use in an ecologically valid drinking context. In that prior study, drinking was measured after the stressful situation. OBJECTIVE: To examine the effect of an anticipatory stressor, and trait social anxiety on "alcohol" consumption in a bar laboratory. METHODS:College students aged 18 and older (N = 127) in same-sex groups of two or three participants took part in a study that ostensibly examined alcohol's effect on language fluency. Using a between-subjects design, participants were randomly assigned to a stress or control condition. Participants in the stress condition anticipated giving a stressful speech for the fluency procedure, whereas those in the control group anticipated a nonstressful activity. Before the alleged fluency task, groups could order and consume mixed drinks ad lib in a bar laboratory. No beverages actually contained alcohol, but we used a validated procedure to ensure that participants included in these analyses were deceived. Primary analyses were performed with a hierarchical linear model (HLM) due to a substantial group/modeling effect. RESULTS: Counter to expectations, participants in the control group consumed more placebo alcohol than those in the stress condition. This main effect was moderated by past 3 months' drinks per week, such that the effect was attenuated (or reversed) among heavy drinkers. No main or interaction effects were observed for trait social anxiety. CONCLUSIONS: Some stressors (i.e., those invoking performance anxiety) may decrease consumption. People with higher levels of alcohol involvement might be especially motivated to drink for tension reduction purposes.
RCT Entities:
BACKGROUND: Only one laboratory study has examined the relation between stress and alcohol use in an ecologically valid drinking context. In that prior study, drinking was measured after the stressful situation. OBJECTIVE: To examine the effect of an anticipatory stressor, and trait social anxiety on "alcohol" consumption in a bar laboratory. METHODS: College students aged 18 and older (N = 127) in same-sex groups of two or three participants took part in a study that ostensibly examined alcohol's effect on language fluency. Using a between-subjects design, participants were randomly assigned to a stress or control condition. Participants in the stress condition anticipated giving a stressful speech for the fluency procedure, whereas those in the control group anticipated a nonstressful activity. Before the alleged fluency task, groups could order and consume mixed drinks ad lib in a bar laboratory. No beverages actually contained alcohol, but we used a validated procedure to ensure that participants included in these analyses were deceived. Primary analyses were performed with a hierarchical linear model (HLM) due to a substantial group/modeling effect. RESULTS: Counter to expectations, participants in the control group consumed more placebo alcohol than those in the stress condition. This main effect was moderated by past 3 months' drinks per week, such that the effect was attenuated (or reversed) among heavy drinkers. No main or interaction effects were observed for trait social anxiety. CONCLUSIONS: Some stressors (i.e., those invoking performance anxiety) may decrease consumption. People with higher levels of alcohol involvement might be especially motivated to drink for tension reduction purposes.
Entities:
Keywords:
Alcohol; anxiety; bar laboratory; stress