Exploring Attentional Bias toward Alcohol Content: Insights from Eye-Movement Activity

Abstract

Introduction: Attentional bias (AB) is an implicit selective attention toward processing disorder-significant information while neglecting other environmental cues. Considerable empirical evidence highlights the clinical implication of AB in the onset and maintenance of substance use disorder. An innovative method to explore direct measures of AB relies on the eye-movement activity using technologies like eye-tracking (ET). Despite the growing interest regarding the clinical relevance of AB in the spectrum of alcohol consumption, more research is needed to fully determine the AB patterns and its transfer from experimental to clinical applications. The current study consisted of three consecutive experiments. The first experiment aimed to design an ad-hoc visual attention task (VAT) consisting of alcohol-related and neutral images using a nonclinical sample (n = 15). The objective of the second and third experiments was to analyze whether the effect of type of image (alcohol-related vs. neutral images) on AB toward alcohol content using the VAT developed in the first experiment was different for type of drinker (light vs. heavy drinker in the second experiment [n = 30], and occasional social drinkers versus alcohol use disorder (AUD) patients in the third experiment [n = 48]).

Methods: Areas of interest (AOIs) within each type of image (neutral and alcohol-related) were designed and raw ET-based data were subsequently extracted through specific software analyses. For experiment 1, attention maps were created and processed for each image. For experiments 2 and 3, data on ET variables were gathered and subsequently analyzed through a two-way ANOVA with the aim of examining the effects of the type of image and drinker on eye-movement activity.

Results: There was a statistically significant interaction effect between type of image and type of drinker (light vs. heavy drinker in experiment 2, F(1, 56) = 13.578, p < 0.001, partial η2 = 0.195, and occasional social drinker versus AUD patients in the experiment 3, F(1, 92) = 35.806, p < 0.001, partial η2 = 0.280) for "first fixation" with large effect sizes, but not for "number of fixations" and "dwell time." The simple main effect of type of image on mean "first fixation" score for AUD patients was not statistically significant.

Conclusion: The data derived from the experiments indicated the importance of AB in sub-clinical populations: heavy drinkers displayed an implicit preference for alcohol-related images compared to light drinkers. Nevertheless, AB fluctuations in patients with AUD compared to the control group were found. AUD patients displayed an early interest in alcohol images, followed by an avoidance attentional processing of alcohol-related images. The results are discussed in light of recent literature in the field.

Keywords: Alcohol use disorder; Attentional bias; Drinking patterns; Eye-tracking technology; Visual attention task.

Fig. 1.

Examples of images used in the VAT – first version.

Fig. 2.

Examples of attention maps of neutral images from the first version of the VAT. The top-left picture (pencils) reflects an attention map with high salience (centered attention or focal point), whereas the bottom-left picture indicates non-centered (dispersed) attention.

Fig. 3.

Schematic representation of the procedure.

Fig. 4.

Experiment 2: interaction effects of type of image and type of drinker on first fixation (plots of light vs. heavy drinkers).

Fig. 5.

Experiment 2: interaction effects of type of image and type of drinker on first fixation (plots of neutral vs. alcohol-related images).

Fig. 6.

Experiment 3: interaction effects of type of image and type of drinker on first fixation (plots of AUD patients vs. occasional social drinkers).

Fig. 7.

Experiment 3: interaction effects of type of image and type of drinker on first fixation (plots of neutral vs. alcohol-related images).