Experimental manipulation of attentional biases in heavy drinkers: do the effects generalise?

RATIONALE: In heavy drinkers, training attention towards alcohol cues increases alcohol craving, but it is not clear if effects of 'attentional training' generalise to novel stimuli and measurement procedures.
OBJECTIVES: We investigated possible generalisation of attentional training to novel alcohol cues and other methods of measuring cognitive bias.
MATERIALS AND METHODS: A modified visual probe task was used to train participants to direct their attention either towards ('attend alcohol' group) or away from ('avoid alcohol' group) alcohol cues; attentional bias was not manipulated in a control group (total N = 60). After attentional training, we measured cognitive bias (using visual probe, modified Stroop, flicker-induced change blindness and stimulus-response compatibility tasks), alcohol craving and alcohol consumption.
RESULTS: Attentional bias for alcohol cues increased in the 'attend alcohol' group, and this effect generalised to novel stimuli, but not to other cognitive bias tasks. In the 'avoid alcohol' group, attentional bias was reduced for the stimuli that were used during attentional training, but these effects did not generalise to different stimuli or cognitive bias tasks. Alcohol craving increased among participants in the 'attend alcohol' group, but only among participants who were aware of the experimental contingencies during attentional training. There were no group differences in alcohol consumption.
CONCLUSIONS: The effects of attentional training show limited generalisation to different alcohol cues and methods of measuring cognitive bias. Experimentally increased attentional bias seems to increase subjective craving, but only among participants who are aware of the experimental contingencies that were in place during attentional training.