Attentional bias training did not improve outcomes for alcohol or cannabis use disorder | Research

The takeaway

A multicenter RCT of 169 participants found that an internet-based attentional bias modification program added to standard treatment produced no significant improvements in substance use, craving, or relapse rates.

Addiction treatment researchers and clinicians considering cognitive training interventions.

No significant differences on any outcome at any time point

What the researchers found

No significant differences emerged between the ABM group and control groups on any measure: substance use, craving, relapse rates, attentional bias, depression, anxiety, or stress, either at post-test or at 6- and 12-month follow-ups. The null results held across all time points.

Why it matters

Attentional bias (automatic attention toward substance-related cues) has been theorized as a key mechanism in addiction persistence. This well-designed trial found no support for ABM as an add-on treatment, raising questions about whether targeting attention is the right approach.

The numbers in context

169 participants: 50% ABM+TAU, 25% placebo+TAU, 25% TAU only. No significant differences on any outcome at any time point. Measures included substance use, craving, relapse, attentional bias, depression, anxiety, and stress.

How the study worked

Multicenter RCT with 169 participants diagnosed with alcohol or cannabis use disorder. Randomly assigned to treatment as usual (TAU) plus attentional bias modification (ABM), TAU plus placebo training, or TAU only. The ABM used a home-delivered, multi-session internet program called the Bouncing Image Training Task. Outcomes assessed at baseline, post-test, 6 months, and 12 months.

What this study cannot tell us

Mixed treatment goals (some pursuing moderation, others abstinence) may have diluted effects. The specific ABM task may not have targeted the right attentional process (engagement vs. disengagement). Home-based delivery reduces compliance monitoring.

How to read the evidence

Well-designed multicenter RCT with long follow-up. Null results are clearly negative rather than inconclusive due to the comprehensive outcome assessment.

When this study was published

2021 multicenter RCT from the Netherlands.

The bigger picture

The failure of ABM to improve addiction outcomes adds to growing skepticism about attentional bias interventions outside the lab. The results may reflect that real-world addiction is driven by factors far more powerful than automatic attentional patterns, or that the specific intervention did not successfully modify bias.

Questions still open

Would ABM work if it targeted disengagement rather than engagement bias? Does the mixed treatment goal of the sample explain the null results? Is attentional bias even a modifiable treatment target in real-world addiction?

Common questions

What is attentional bias modification?

It is a computer-based training designed to reduce automatic attention toward substance-related cues (like images of alcohol or cannabis). The theory is that this automatic attention helps maintain addiction.

Why might the training have failed?

The authors suggest three possibilities: the training may not have successfully changed attentional bias, it may have targeted the wrong attentional process, or the mixed treatment goals of participants may have reduced effectiveness.

Read the original research

Effectiveness of attentional bias modification training as add-on to regular treatment in alcohol and cannabis use disorder: A multicenter randomized control trial.

PloS one, 16(6), e0252494

View on PubMed ↗Publisher’s paper ↗

Citation

Heitmann, Janika; van Hemel-Ruiter, Madelon E; Huisman, Mark; Ostafin, Brian D; Wiers, Reinout W; MacLeod, Colin; DeFuentes-Merillas, Laura; Fledderus, Martine; Markus, Wiebren; de Jong, Peter J. (2021). Effectiveness of attentional bias modification training as add-on to regular treatment in alcohol and cannabis use disorder: A multicenter randomized control trial.. PloS one, 16(6), e0252494. https://doi.org/10.1371/journal.pone.0252494