Retraining automatic cogntieve processes in problematic alcohol use.

The aim of the current study is to investigate the effectiveness of three online CBM paradigms among problem drinkers aged 18-64 years: attentional bias retraining, alcohol/no-go training (targeting memory bias), and approach bias retraining. The main goal of the study is to test the effects of CBM on alcohol use at 3 months after the intervention, with changes in the number of alcoholic drinks consumed in the past two weeks as the primary outcome measure.

It is expected that, for each of the three CBM paradigms, participants in the intervention conditions will show a greater decrease in weekly alcohol use than participants in the control conditions. Generalization of each specific CBM paradigm to other biases will be explored, as will the effects of exposure to the various combinations of CBM paradigms. It is expected that each CBM paradigm will decrease or reverse the specific bias it trains and that these changes will mediate the effects on alcohol use. Furthermore, it is expected that participants with strong automatic biases and/or low working memory capacity and inhibitory control will benefit more from CBM than participants with weaker biases and/or stronger executive functions.

Baseline assessment, post-intervention assessment (about 6 weeks after baseline), follow-ups (3,6 and 12 months after the post-intervention assessment).

Each of the twelve CBM sessions consists of three tasks (of about 15 minutes each): attentional bias retraining, alcohol/no-go training, and approach bias retraining. The three CBM paradigms were designed to be as similar as possible, to enable clear comparisons of the training effects.



1. Attentional bias retraining:

Attentional bias is assessed and trained using the visual probe task. In this task, a picture of an alcoholic and a picture of a non-alcohol beverage are presented next to each other on the screen for 500 ms. When the pictures disappear, a small arrow pointing up or down appears at the location of one of the pictures. Participants are instructed to respond to the direction of the arrow, by pressing the corresponding arrow key on the keyboard. Attentional bias for alcohol (i.e. faster detection of the direction of the probe when it appears at the location of the alcohol picture, where attention was already focused, than when it appears at the location of the soda picture) has been related to alcohol use and craving.
In the assessment block, the arrow replaces the picture of the alcoholic beverage (alcohol trials) and the picture of the non-alcoholic beverage (non-alcohol trials) equally often. Attentional bias for alcohol is computed by subtracting RTs on alcohol trials from those on non-alcohol trials. In the CBM block, participants in the experimental condition will be trained to direct their attention away from alcoholic beverages towards non-alcoholic beverages, by exposing them only to non-alcohol trials, while participants in the placebo condition receive 50% alcohol and 50% non-alcohol trials (as in the assessment block).



2. Alcohol/no-go training:

Alcohol related memory bias (positive associations with alcohol) is measured and trained using the alcohol/no-go task. In this task, a picture of an alcoholic or non-alcoholic beverage is presented for 1500 ms, together with a go (i.e. the letter ‘p’) or no-go cue (‘f’) which is displayed randomly in one of four corners of the picture. Participants are instructed to respond to the presented letter, i.e. by pressing the spacebar as quickly as possible when they see the letter ‘p’, and do nothing (wait until picture disappears) when they see the letter ‘f’. The contingency between the letter and the response (p = press and f = inhibit, versus f = press and p = inhibit) is counterbalanced across participants. Previous studies on alcohol/no-go training have not investigated whether heavy drinkers show an ‘alcohol/go bias’, i.e. faster reaction times for alcohol/go trials compared to soda/go trials. In the current study, an assessment block is included (similar to the assessment block in the attentional and approach bias retraining) in order to explore whether problem drinkers demonstrate an ‘alcohol/go bias’ at baseline, and whether this bias is influenced by alcohol/no-go training.

In the assessment block, the pictures of alcoholic and non-alcoholic beverages are presented equally often with a go and with a non-go cue. Alcohol/go bias is computed by subtracting RT’s on alcohol/go trials from those on non-alcohol/go trials. In addition, the number of errors (incorrect go responses) can be compared between alcohol and non-alcohol no-go trials. In the CBM block, participants in the experimental condition will be trained to inhibit their response to alcohol, by exposing them only to alcohol/no-go trials and non-alcohol/go trials, while for participants in the placebo condition there is no contingency between the content of the picture and the required response (as in the assessment block).



3. Approach bias retraining:

Approach bias is trained with the modified Approach-Avoidance Task (AAT). In this paradigm, a picture of an alcoholic or non-alcoholic beverage is presented, which is tilted 3 degrees to the left or right. Participants are instructed to respond to format of the picture, e.g. by pushing all pictures tilted to the right away from them, and pulling all pictures tilted to the left towards them. The contingency between the format of the picture and the response (left = push and right = pull, versus left = pull and right = push) is counterbalanced across participants. Picture size gradually increases when the pull-key is pressed, while it decreases when the push-key is pressed. Heavy drinkers have been found to show an approach bias for alcohol, i.e. faster response to alcohol/pull trials than to alcohol/push trials.

In the assessment block, the pictures of alcoholic and non-alcoholic beverages are presented equally often in push and in pull format. Approach bias for alcohol is computed by comparing RTs for push, pull, alcohol and non-alcohol trials ((alcohol/push - alcohol/pull) - (non-alcohol/push - non-alcohol/pull)). In the CBM block, participants in the experimental condition will be trained to avoid alcohol, by exposing them only to alcohol/push and non-alcohol/pull trials, while for participants in the placebo condition there is no contingency between the content of the picture and the required response (as in the assessment block).