Craving modulates attentional bias towards alcohol in severe alcohol use disorder: An eye-tracking study

Competing models disagree on three theoretical questions regarding alcohol-related attentional bias (AB), a key process in severe alcohol use disorder (SAUD): (1) is AB more of a trait (fixed, associated with alcohol use severity) or state (fluid, associated with momentary craving states) characteristic of SAUD; (2) does AB purely reflect the over-activation of the reflexive/reward system or is it also influenced by the activity of the reflective/control system and (3) does AB rely upon early or later processing stages? We addressed these issues by investigating the time-course of AB and its modulation by subjective craving and cognitive load in SAUD.

A free-viewing eye-tracking task, presenting pictures of alcoholic and non-alcoholic beverages, combined with a concurrent cognitive task with three difficulty levels.

A laboratory setting in the detoxification units of three Belgian hospitals.

We included 30 patients with SAUD self-reporting craving at testing time, 30 patients with SAUD reporting a total absence of craving and 30 controls matched on sex and age. All participants from SAUD groups met the DSM-5 criteria for SAUD.

We assessed AB through early and late eye-tracking indices. We evaluated the modulation of AB by craving (comparison between patients with/without craving) and cognitive load (variation of AB with the difficulty level of the concurrent task).

Dwell time measure indicated that SAUD patients with craving allocated more attention towards alcohol-related stimuli than patients without craving (P < 0.001, d = 1.093), resulting in opposite approach/avoidance AB according to craving presence/absence. SAUD patients without craving showed a stronger avoidance AB than controls (P = 0.003, d = 0.806). AB did not vary according to cognitive load (P = 0.962, η² _p  = 0.004).

The direction of alcohol-related attentional bias (approach/avoidance) appears to be determined by patients' subjective craving at testing time and does not function as a stable trait of severe alcohol use disorder. Alcohol-related attentional bias appears to rely on later/controlled attentional stages but is not modulated by the saturation of the reflective/control system.