Thanks for this collaborative effort, Claudia Ebrahimi, @endrass.bsky.social, @ldeserno.bsky.social, and all other coauthors!

While we cannot draw causal conclusions based on this study design, I found it fascinating that a relatively short task allows to predict real-life control over drinking.

9/9

…were less successful at drinking less than usual when they intended to do so during the following months. Put differently, they were worse at putting their drinking reduction intentions into practice.

8/9

People who behaved in a less goal-directed manner on the task and who had less pronounced hippocampal correlates of model-based RPE components…

7/9

In 67 participants with AUD who performed the two-step task during fMRI and were then tracked using smartphone-based ecological momentary assessments for up to one year, we found evidence for this idea.

6/9

Can we predict control over alcohol intake in the day-to-day life of individuals with AUD based on how goal-directedly they behave on the classic two-step task and how strongly they encode model-based prediction error components?

5/9

We thus moved away from looking at group differences, and instead asked:

4/9

…but there’s no consistent evidence for categorical differences in goal-directed behavior between individuals with and without AUD diagnosis.

3/9

This longitudinal project focused on alcohol use disorder (AUD). That’s because increased habitual and decreased goal-directed control have been suggested to underly addiction,…

2/9

Preprint will follow soon 📃