Anxiety is generally characterized by cognitive impairment. In particular, the worry dimension of anxiety is closely related to impaired cognitive performance [1, 2]. Worry interferes with the transmission of signal from action-monitoring to post-error adjustments, which leads to unimproved behavioral performance [3]. Additionally, if worry refers to the cognitive dimension of anxiety, fear can be considered as emotional and psychological arousal. The relationship between fear and worry has been explored in numerous studies. Individuals in fear are more sensitive to emotional stimuli than those in worry, which shows the distinction between fear and worry in attention engagement [4]. Besides, worry, as the contrast avoidance, may attempt to inhibit the fear experience [5].

The interaction between fear and worry has a significant effect on the anxious experience. For example, when a snake appears unexpectedly, one may jump and yell automatically (fear reflex). Once fear is generalized, the fearful individual may avoid the related cue, make unrealistic appraisals and even have anticipation about similar stimuli. With the negative bias, the individual experience more fear of the threat. Meanwhile, the individual engages in worrisome thinking (worry), such as “What would have happened if I had stepped on it?”. Although the intention is to control the fear, this linguistic activity actually leads to greater negative emotions and physiological activation. In addition, the worrisome thoughts impede performance during cognitive tasks. In brief, worry has an effect on the fear experience and cognitive function (in anxiety groups). However, it remains unclear how worry is involved in the translation of fear into disrupted cognitive performance under the condition of explicit threat.

Previous research has described five processes of uncertainty and anticipation in the model of anxiety [6]. The core of this model are disrupted expectancies regarding uncertainty about potential threats, which manifests as excessive estimates of probability and cost, heightened vigilance, biased attention toward threat, and impaired safety learning. However, maladaptive responses to specific fear stimuli cannot be explained in this model. In addition, under the condition of an explicit threat, the fear reflex is often automatic and elicited before worry. Therefore, to understand how maladaptive responses are engaged and disruptive to cognitive performance under the condition of specific threat, we propose four processes to address the translation of fear into impaired cognitive outcomes through worry, including the fear reflex; anticipation and appraisal with negative bias; contrast avoidance of worry; and impaired cognitive performance (Fig. 1). During these processes, linguistic activity during worry exacerbates the negative experience and plays a critical role in cognitive impairment.

Fig. 1
figure 1

(Color online) The translation of the fear reflex into impaired cognitive function mediated by worry under the condition of specific threat. Once the fear reflex is evoked, memories of the relevant experience are replayed. Specifically, the AMY is mainly recruited by the fear affect. The areas of the rC, VS, OFC, and DMPFC are linked to cognition. The network of the insula and ACC is involved in interoceptive awareness. During anticipatory and threat appraisal, the sgACC, VMPFC, and ACC are mainly implicated. Meanwhile, to control the fear, worry, or the inner verbal stream, surges (the IFG, ITG, MPFC, Precu, insula, etc. are involved). Nevertheless, worry leads to disrupted cognitive performance. AMY, amygdala; dACC, dorsal anterior cingulate cortex; HIPP, hippocampus; lIFG, left inferior frontal gyrus; MPFC, medial prefrontal cortex; VMPFC, ventromedial prefrontal cortex; MTC, medial temporal cortex; OFC, orbitofrontal cortex; rC, rostral cingulate; rTPC, right temporo-parietal cortex; rACC, rostral anterior cingulate cortex; Precu, precuneous; rITG, right inferior temporal gyrus; sgACC, sub-genual anterior cingulate; VS, ventral striatum

Full size image

When a salient threat emerges unexpectedly, the brain responds automatically (fear reflex). The bottom-up sensory mechanisms, for example, the right inferior temporal gyrus and right temporoparietal cortex are implicated [4]. During this process, the amygdala is a key structure in fear perception and the dACC is associated with the subjective experience of negative affect. Meanwhile, the amygdala and hippocampus-based memory system are recruited for acquisition and retrieval of fear memory traces. The memories of previous experience linked to interception, affect, and cognition are also recalled instantly. In this case, the related brain regions are engaged. Subsequently, all information is integrated and transmitted to the anterior mid-cingulate cortex (aMCC) [6]. The integration in the aMCC provides instructions that influence subsequent actions.

The overgeneralization of fear is the core characteristic of anxiety disorders, which enhances the unpredictability and uncontrollability and makes it difficult to prepare properly for future events. Anxious individuals tend to make anticipation and appraisal about the uncertain threat with negative bias, which will draw individuals’ attention and induce hypervigilance to the threat [6]. In this process, the dynamic activation of the ACC depends on the level of threat during anticipation [7]. With the negative bias, the individuals with post-traumatic stress disorder (PTSD) prefer looking for potential dangers constantly in the external environment. Researches have indicated that the activation of amygdala is also related to the bias towards threat and the reduced activation of VMPFC and sgACC is associated with threat appraisal [8]. The individual cannot discriminate between threat and safety accurately, which makes the fearful experience more unbearable.

When the subjects cannot withstand discomfort, efforts are made to relieve or avoid the discomfort. In behavior, the individual may try to leave or escape. In terms of cognition, worry surges as verbal stream. The worriers show more left-hemisphere activation in the inferior temporal gyrus [9]. Besides, the enhanced neural activation to negative words in Broca’s area was also observed during worry. Verbal activity in worrying is directed at inhibiting emotional processing of both imagined and actual feared stimuli. However, rather than acting to suppress, worry extends negative affect and physiological activation [5]. In addition, the individual who is susceptible to worry is hard to ignore the fear cue or memory. Moreover, worry-related cognition, including negative self-referential thoughts and self-focused attention, is elevated by the level of worry [2]. All of those will aggravate the negative experience in worriers.

As the linguistic activity, worry is closely related to the impaired cognitive performance [1]. Research has shown that worry can explain the mechanism of error commission by recruiting stronger dorsolateral prefrontal cortex (DLPFC)-precuneus and DLPFC-posterior cingulate connectivity [2]. In addition, higher levels of worry are linked to reduced rACC activity and increased dACC activity. With levels of worry rising, the increased engagement of dACC is associated with the greater Stroop interference [10]. The patterns activated suggest that worry impairs the conflict resolution and proactive control processes. Moreover, worry is related to the error committed (the error-related negativity (ERN) by the right hand during the Eriksen flankers task [1]. The enlarged ERN reflects compensatory post-error adjustment in response to worry-induced attentional distraction.

The evidence reviewed here describes the translation of fear into disrupted cognitive performance mediated by worry. Some researches manifest that fear and worry are engaged consecutively. First, fear is not only a response to immediate and real danger, but also reflexes to future danger. During worry, individuals attempt to anticipate and prepare for possible negative outcomes in the future. Second, fear, as the emotional and physical arousal, is elicited automatically and transiently. Whereas, worry, in the form of inner language, is more perseverative and lasting. Furthermore, when the individual cannot endure the fear-experience, worry-related thoughts engaged attempt to inhibit the negative emotion [5]. Based on the characters of fear and worry, it is clear that worry occurs after fear in an attempt to control the fearful experience.

Worry, as the maladaptive response, also worsens the anxious experience. Research has proposed that worry functions as an adaptive avoidance strategy in healthy controls but not in pathological worriers. Once worry becomes a pathological response, it will prolong negative physical and emotional arousal [5]. Besides, worriers show increased sympathetic nervous system activity and decreased parasympathetic nervous system. Meanwhile, worry, as a negative intrusion, develops into verbal streams about threats, which may cause further problems in worriers. Among these, the relationship between worry and impaired cognitive performance has been a focus of many studies [2, 3]. One theory involves limited cognitive resources. It proposes that resources for maintaining attentional control can be depleted by worry. In this way, enhanced ERN amplitude shows compensatory recruitment of cognitive control through reactive control in worriers [1]. Another theory holds that increased worry is linked to elevated levels of spontaneous self-related concerns. The task unrelated thoughts are uncontrollable and disruptive, which leads to impoverished attention-control mechanism [2]. Moreover, further research has revealed that the worry–ERN relationship involves verbal processes in the left hemisphere [1]. In addition, verbal-form worry had more adverse effects on attentional control than that in the form of imagery. As noted earlier, Snyder and Hutchison et al. found in language-production task participants higher in worry have larger selection costs [11]. Therefore, we believe that worry, as the linguistic activity, may interfere with cognitive control. Language processing may act as a mediator between worry and cognitive deficiency. This is an area in need of further research.