Worrying has been suggested to prevent emotional and elaborative processing of fears. In cognitive-behavioral therapy (CBT), generalized anxiety disorder (GAD) patients are exposed to their fears during the method of directed threat imagery by inducing emotional reactivity. However, studies investigating neural correlates of directed threat imagery and emotional reactivity in GAD patients are lacking. The present functional magnetic resonance imaging (fMRI) study aimed at delineating neural correlates of directed threat imagery in GAD patients.

Nineteen GAD patients and 19 healthy controls (HC) were exposed to narrative scripts of either disorder-related or neutral content and were encouraged to imagine it as vividly as possible.

Rating results showed that GAD patients experienced disorder-related scripts as more anxiety inducing and arousing than HC. These results were also reflected in fMRI data: Disorder-related v. neutral scripts elicited elevated activity in the amygdala, dorsomedial prefrontal cortex, ventrolateral prefrontal cortex and the thalamus as well as reduced activity in the ventromedial prefrontal cortex/subgenual anterior cingulate cortex in GAD patients relative to HC.

The present study presents the first behavioral and neural evidence for emotional reactivity during directed threat imagery in GAD. The brain activity pattern suggests an involvement of a fear processing network as a neural correlate of initial exposure during directed imagery in CBT in GAD.

Panic disorder (PD) patients are constantly concerned about future panic attacks and exhibit general hypersensitivity to unpredictable threat. We aimed to reveal phasic and sustained brain responses and functional connectivity of the amygdala and the bed nucleus of the stria terminalis (BNST) during threat anticipation in PD.

Using functional magnetic resonance imaging (fMRI), we investigated 17 PD patients and 19 healthy controls (HC) during anticipation of temporally unpredictable aversive and neutral sounds. We used a phasic and sustained analysis model to disentangle temporally dissociable brain activations.

PD patients compared with HC showed phasic amygdala and sustained BNST responses during anticipation of aversive v. neutral stimuli. Furthermore, increased phasic activation was observed in anterior cingulate cortex (ACC), insula and prefrontal cortex (PFC). Insula and PFC also showed sustained activation. Functional connectivity analyses revealed partly distinct phasic and sustained networks.

We demonstrate a role for the BNST during unpredictable threat anticipation in PD and provide first evidence for dissociation between phasic amygdala and sustained BNST activation and their functional connectivity. In line with a hypersensitivity to uncertainty in PD, our results suggest time-dependent involvement of brain regions related to fear and anxiety.

There is an ongoing debate whether transdiagnostic neural mechanisms are shared by different anxiety-related disorders or whether different disorders show distinct neural correlates. To investigate this issue, studies controlling for design and stimuli across multiple anxiety-related disorders are needed.

The present functional magnetic resonance imaging study investigated neural correlates of visual disorder-related threat processing across unmedicated patients suffering from panic disorder (n = 20), social anxiety disorder (n = 20), dental phobia (n = 16) and post-traumatic stress disorder (n = 11) relative to healthy controls (HC; n = 67). Each patient group and the corresponding HC group saw a tailor-made picture set with 50 disorder-related and 50 neutral scenes.

Across all patients, increased activation to disorder-related v. neutral scenes was found in subregions of the bilateral amygdala. In addition, activation of the lateral amygdala to disorder-related v. neutral scenes correlated positively with subjective anxiety ratings of scenes across patients. Furthermore, whole-brain analysis revealed increased responses to disorder-related threat across the four disorders in middle, medial and superior frontal regions, (para-)limbic regions, such as the insula and thalamus, as well as in the brainstem and occipital lobe. We found no disorder-specific brain responses.

The results suggest that pathologically heightened lateral amygdala activation is linked to experienced anxiety across anxiety disorders and trauma- and stressor-related disorders. Furthermore, the transdiagnostically shared activation network points to a common neural basis of abnormal responses to disorder-related threat stimuli across the four investigated disorders.