Anxious minds don’t always fixate on danger, according to new study

Anxiety does not appear to impair people’s ability to disengage attention from threatening stimuli when they are sufficiently motivated to do so, according to a study published in Cognition & Emotion.

Attentional biases toward threat are widely implicated in the development and maintenance of anxiety. Individuals with higher anxiety are thought to have difficulty disengaging attention from threatening stimuli, potentially fueling hypervigilance and a sense that danger is always present. Many cognitive models and therapies, including attentional bias modification, assume that this delayed disengagement is involuntary and resistant to motivational control.

Agnes Musikoyo and colleagues examined whether delayed disengagement from threat-related stimuli is truly an involuntary feature of anxiety or whether it might reflect voluntary attentional choices. Prior studies have often confounded initial orienting with disengagement and rarely considered the role of motivation in disengagement. The authors designed three pre-registered experiments using central fixation paradigms that isolate the disengagement process and manipulated participants’ motivation to disengage using punishment or reward.

Experiment 1 included 142 undergraduate participants who were assigned to low or high anxiety groups based on their DASS-21 scores (Depression Anxiety Stress Scales). After a Pavlovian conditioning phase where participants learned to associate one colored circle with an aversive noise (CS+) and another with no noise (CS–), they completed an eye-tracking task.

Participants had to saccade to a peripheral target while a CS+ or CS– stimulus appeared at fixation. Disengagement was penalized with a noise if too slow, incentivizing quick responses. Key measures included saccade latency from the central stimulus and self-reported anxiety, aversiveness ratings, and noise expectancy.

In Experiments 2 and 3, larger samples of participants (201 and 195 respectively) completed remote reaction-time tasks. Participants saw centrally presented images (angry/fearful/neutral faces in Experiment 2; angry/happy/neutral faces and snakes/spiders in Experiment 3) followed by a visual search task. Motivation was manipulated: half the participants received monetary incentives for quick responses, while the rest received only correctness feedback. Anxiety was measured using the DASS-21 (Experiment 2) and STAI-T (State Trait Anxiety Inventory; Experiment 3), and disengagement was inferred from slower response times when threat stimuli were shown.

In Experiment 1, participants were consistently slower to move their eyes away from the CS+ stimulus compared to the CS–, even though such delay increased their chance of being punished with the aversive noise. This suggests that motivationally salient threat signals can involuntarily capture attention. However, this effect was not influenced by participants’ anxiety level. High- and low-anxiety individuals showed comparable disengagement latencies. Even when analyzed using more sensitive trial-level models and including all participants, there was no evidence that anxiety predicted greater difficulty in disengaging from threat under high-motivation conditions.

In Experiment 2, the researchers found no evidence that emotional facial expressions, whether angry or fearful, slowed participants’ disengagement from the central stimulus, regardless of anxiety level or motivational context. While participants who were motivated by rewards responded faster overall, their disengagement speed did not differ depending on the emotional content of the face. Moreover, anxious individuals did not show any greater tendency to dwell on negative faces. The results were consistent across multiple analytic strategies, and again, no anxiety-related attentional bias emerged.

In Experiment 3, participants did exhibit slower responses when disengaging from images of snakes and spiders, suggesting these stimuli were indeed more attentionally captivating. However, this pattern did not interact with anxiety level or motivational condition, and no consistent effects emerged for emotional faces, even when participants had time to process them before the search task began.

Altogether, the findings across the three experiments suggest that attentional disengagement from threat is not inherently impaired in anxious individuals. Instead, the tendency to dwell on threat-related stimuli may be more situational and influenced by task demands and stimulus properties, rather than being a stable, anxiety-driven deficit.

Of note is that the threat stimuli used may not have been equally motivationally relevant across participants, particularly the use of emotional faces.

The research, “The role of motivation in delayed disengagement from threat in anxiety,” was authored by Agnes Musikoyo, Andrew E. Rayment, and Poppy Watson.