Does participation in an alcohol administration study increase risk for excessive drinking?

It has long been thought that research protocols involving alcohol administration may exacerbate problem drinking in alcoholic subjects following their participation in such a study. However, recent studies suggest that involvement in an alcohol administration study does not, in fact, have a negative impact on subsequent drinking behavior. In the present study, 27 non-treatment-seeking alcohol-dependent subjects and 32 social drinkers participated in an alcohol administration study designed to investigate the effects of repeated doses of alcohol on craving, mood, and alcohol-seeking behavior. The volume of alcohol administered to the subjects was calculated in such a way that their blood alcohol concentration would reach a peak of 0.08 g/dl midway through testing. Before their release, alcohol-dependent subjects were given feedback regarding their level of alcohol consumption and provided with information about the potential harmful effects of their drinking behavior. Percentage of days abstinent (PDA), drinks per drinking day (D/DD), and percentage of heavy drinking days (PHDD, defined as >or=4 drinks per occasion for females and >or=5 drinks per occasion for males) were recorded for the 6 weeks preceding laboratory testing and for the 6 weeks following participation in the study. The alcohol-dependent subjects exhibited a significant increase of 24% in PDA during the poststudy period compared to the prestudy period. They also decreased their D/DD by 2.4 drinks per occasion, and decreased their PHDD by 21.6%. There were no differences in PDA or D/DD for the social drinkers between pre- and poststudy periods. There was, however, a small but significant increase of 3.5% in PHDD for the social drinkers following laboratory testing. These data suggest that participation in an alcohol administration study does not put alcoholic subjects at risk for increased alcohol consumption following study participation. In fact, participation in such studies may actually precipitate at least a temporary decrease in alcohol consumption, especially when paired with a brief intervention session. Thus, non-treatment-seeking alcoholics can be safely included in alcohol administration studies to provide results that are most relevant to the population of interest.