The effects of alcohol and alcohol expectancies on subjective reports and physiological reactivity: a meta-analysis.

Abstract The balanced placebo design (BPD) has been used to understand the etiology and maintenance of alcohol consumption. The utility of this design lies in its ability to examine both actual alcohol consumption and the expectation of alcohol consumption. A meta-analysis of the BPD literature was conducted in the context of cue-reactivity, which may be characterized as an experimental phenomenon observed in studies utilizing alcohol. Sixty-four studies were obtained in literature searches and coded for type of experimental setting and cues present during the actual beverage consumption. Lab setting was a moderator for both pharmacological (alcohol) and expectancy effects with the largest effects (in the same direction) noted in natural environment labs (i.e., an easy chair and casual environment). Contrary to predictions, the bar lab produced the smallest effects. Cues present during alcohol consumption served as a moderator of pharmacological effect, with the largest effect observed when alcohol was placed on the rim of the glass. Implications of these findings for cue-reactivity studies and the treatment of alcohol abuse are discussed.