Retraining automatic approach tendencies in alcohol dependent patients

Aims: The study was designed to assess the short-term and long-term effects of a retraining of automatic approach tendencies in alcohol dependent patients. Furthermore, the specificity of the training effects was tested.

Design: Experimental study.

Participants: 225 alcohol-dependent patients in an inpatient treatment setting. The patients were assigned to four different training groups.

Measurements: We used three different versions of the adapted Approach Avoidance Task (AAT; Rinck & Becker, 2007) to train automatic alcohol and soda action tendencies in alcohol dependent patients. The training consisted of four training sessions, lasting 10 minutes each. Participants of the direct training group pulled or pushed a joystick in response to the contents of the presented pictures (push alcohol, pull soda). Participants of the indirect training group reacted to the pictures’ format, while alcohol always appeared in the pushing condition. Participants of the placebo training group also reacted to the pictures’ format, with 50% of the alcohol pictures appearing in the pushing condition and 50% appearing in the pulling condition. Finally, a waiting-list group received no training at all. Measuring short-term training effects, we tested generalization effects to new pictures in the AAT and to the Implicit Association Task (IAT). Long–term effects were measured using one-year follow-up relapse data. To test the specificity of training effects, we compared the training effects (short-term and long-term) of different subgroups of patients: Patients with high versus low initial approach bias, with high versus low short-term training effects, and with high versus low educational background.

Findings: The training effects of the direct and the indirect training group generalized to new pictures in the AAT and to words in the Implicit Association Test (IAT). These effects could not be found for the placebo training group or the group without training. Analyzing the one-year follow-up relapse data (which are not complete yet), we could find a (non-significant) trend towards a higher rate of successfully treated patients within the direct training group (70%) and the indirect training group (65%), compared to the placebo training group (60%) and the group without training (48%). The results of testing the specificity of training effects are not available yet.

Conclusion: Training automatic approach tendencies in alcohol dependent patients can lead to changes in alcohol approach actions and may be able to reduce the rate of relapse. To increase the positive effects on relapse prevention, we are now testing an intensified training version consisting of 12 training sessions.