ORIGINAL RESEARCH published: 08 May 2019 doi: 10.3389/fpsyg.2019.00678 Language as a Threat: Multimodal Evaluation and Interventions for Overwhelming Linguistic Anxiety in Severe Aphasia María José Torres-Prioris 1,2* , Diana López-Barroso 1,2* , José Paredes-Pacheco 3,4 , Núria Roé-Vellvé 3,5 , Marc S. Dawid-Milner 6 and Marcelo L. Berthier 1 1 Cognitive Neurology and Aphasia Unit (UNCA), Centro de Investigaciones Médico-Sanitarias, Instituto de Investigación Biomédica de Málaga – IBIMA, University of Málaga, Málaga, Spain, 2 Area of Psychobiology, Faculty of Psychology, University of Málaga, Málaga, Spain, 3 Molecular Imaging Unit, Centro de Investigaciones Médico-Sanitarias, General Foundation of the University of Malaga, Málaga, Spain, 4 Molecular Imaging and Medical Physics Group, Department of Psychiatry, Radiology and Public Health, University of Compostela, Santiago de Compostela, Spain, 5 Biomedical Research Networking Center in Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Barcelona, Spain, 6 Neurophysiology of Autonomic Nervous System Laboratory, Centro de Investigaciones Médico-Sanitarias, University of Málaga, Málaga, Spain Edited by: Agustin Ibanez, Linguistic anxiety (LA) is an abnormal stress response induced by situations that require Institute of Cognitive and Translational Neuroscience (INCyT), Argentina the use of verbal behavior, and it is accentuated during language testing in persons with Reviewed by: aphasia (PWA). The presence of LA in PWA may jeopardize the interpretation of cognitive Alfredo Ardila, evaluations, leading to biased conclusions about the severity of the language alteration Florida International University, United States and the effectiveness of the treatments. In the present study, we report the case of a Adolfo M. García, woman (Mrs. A) with severe chronic mixed transcortical aphasia due to left frontal and Laboratory of Experimental parietal hemorrhages that partially spared the perisylvian area. Mrs. A was treated with Psychology and Neuroscience, Argentina the dopamine agonist Rotigotine alone and combined with Intensive Language-Action *Correspondence: Therapy (ILAT). Complementary evaluations included autonomic reactivity during the María José Torres-Prioris performance of different language tasks, resting state functional magnetic resonance mjprioris@uma.es Diana López-Barroso imaging (rs-fMRI) and [18F]-fluorodeoxyglucose positron emission tomography (18F- dlopbarroso@uma.es FDG-PET). We found that formal language testing in a clinical setting triggered a dramatic increase of automatic echolalia, perseverations and frustration, making the Specialty section: This article was submitted to task completion difficult. The treatment improved aphasia, but gains were more robust Clinical and Health Psychology, when evaluation was performed by Mrs. A’s husband at home than by clinicians. a section of the journal Autonomic evaluation under Rotigotine revealed higher reactivity during tasks tapping Frontiers in Psychology an impaired function in comparison with a task evaluating a preserved function (verbal Received: 19 June 2018 Accepted: 11 March 2019 repetition). Baseline 18F-FDG-PET analysis showed decreased metabolic activity in left Published: 08 May 2019 limbic-paralimbic areas, whereas rs-fMRI revealed compensatory activity in the right Citation: hemisphere. We also analyzed the different factors (e.g., premorbid personality traits, Torres-Prioris MJ, López-Barroso D, task difficulty) that may have contributed to LA in Mrs. A during language testing. Our Paredes-Pacheco J, Roé-Vellvé N, findings emphasize the usefulness of implicating adequately trained laypersons in the Dawid-Milner MS and Berthier ML evaluation and treatment of PWA showing LA. Further studies using multidimensional (2019) Language as a Threat: Multimodal Evaluation evaluations are needed to disentangle the interplay between anxiety and abnormal and Interventions for Overwhelming language as well as the neural mechanisms underpinning LA in PWA. Linguistic Anxiety in Severe Aphasia. Front. Psychol. 10:678. Keywords: linguistic anxiety, stress, aphasia, language assessment, autonomic response, neuroimage, doi: 10.3389/fpsyg.2019.00678 Rotigotine Frontiers in Psychology | www.frontiersin.org 1 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia INTRODUCTION “linguistic anxiety” (LA) has been coined to describe the stress reactions induced by language testing and everyday situations Aphasia is a neuropsychological syndrome characterized that require the use of language in PWA (Cahana-Amitay et al., by an impairment of language, affecting speech production 2011). The inability of PWA to use language to communicate (spontaneous speech, naming and repetition) and auditory effectively may precipitate an excessive emotional reactions comprehension as well as the ability to read and write. Aphasic (i.e., anxiety state, frustration, irritability) during language syndromes are frequently classified upon performance on testing, the so-called “catastrophic reaction” (Goldstein, 1948; verbal repetition tasks (Albert et al., 1981). In accordance, Starkstein et al., 1993). Thus, the anticipation of committing aphasias are broadly classified in two groups: (1) aphasias verbal errors or the inability to perform a given task may be characterized by impaired verbal repetition, the so-called perceived as a stressful situation that triggers frustration and perisylvian aphasias (e.g., Broca’s and Wernickes aphasias); and anger (Goldstein, 1948). This means that in clinical settings, the (2) those with preserved verbal repetition, named transcortical existing language deficits might be reinforced by the anxiety aphasias. Aphasia mainly occurs as a consequence of brain reaction which may lead to inaccurate conclusions about the damage in the left hemisphere, usually due to stroke (Hillis, clinical profile and severity of aphasia and the effectiveness of 2007). Previous research has evidenced that stroke is commonly therapeutic interventions. Intriguingly, despite the relevance associated to pervasive emotional changes (e.g., depression, and potential clinical implications of LA, very few studies have anxiety, irritability, apathy, impaired awareness) (Hackett et al., addressed its impact on language performance in the aphasic 2014; Babulal et al., 2015; Robinson and Jorge, 2016; Døli et al., population (Laures-Gore et al., 2007; Cahana-Amitay et al., 2015; 2017). These psychological and behavioral alterations may be Laures-Gore and Buchanan, 2015). In addition, there is limited a consequence of focal and distant brain changes caused by evidence on the contributing factors (e.g., premorbid personality the lesion which might disrupt brain networks implicated in traits, lesion location) and the behavioral, psychological and the control of mood and behavior (House, 1987; Castillo et al., physiological characteristics of anxiety response in PWA. 1993; Beblo et al., 1999; Chemerinski and Robinson, 2000). But Relevantly, it is still unclear whether brain damaged persons emotional changes may also be linked to adjustment problems will have similar physiological response as healthy subjects related to the imposed life changes resulting from cognitive exposed to anxiogenic events, and how lesions affecting areas and motor deficits (Chemerinski and Robinson, 2000). Further, involved in “higher” modulation of the autonomic response may factors such as personality traits (Aben et al., 2002; Greenop alter this reaction. et al., 2009), coping strategies (Vuger-Kovačić et al., 2007), family The paucity of reports on this issue stimulates performing support (Morris et al., 1991) and premorbid mood disorders exploratory studies that help to design protocols with an (Burvill et al., 1995; Sembi et al., 1998; Leppävuori et al., 2003) integrative approach suitable even for persons with severe aphasia may also influence outcomes. All this suggests the need to adopt (e.g., global aphasia, mixed transcortical aphasia [MTCA]). a multidimensional approach to identify the multiple factors that Herein, we report the case of an elderly woman (Mrs. A) contribute to the emotional and psychological consequences of with severe chronic post-stroke MTCA who displayed an post-stroke aphasia. overwhelming LA during language evaluation when this was Among the above-mentioned, anxiety is one of the most performed by the therapist, but not when a trained naïve frequent emotional consequences of stroke (Fure et al., 2006). relative performed the evaluation and treatment. Therefore, The prevalence of anxiety in stroke victims was estimated to the aims of the present study were: (a) to describe the be between 20 and 24% (Burton et al., 2013). However, this LA profile of Mrs. A highlighting its clinical implications; figure may be misrepresented since screening tools for anxiety (b) to explore Mrs. A’s autonomic responses during testing disorders are not always suitable for persons with aphasia (PWA), of different language functions; and (c) to characterize the a reason that explains why this stroke population is frequently associated structural and functional brain correlates of LA using excluded from anxiety studies (Burton et al., 2013). Recently, multimodal neuroimaging. Morris et al. (2017) reported that 44% of PWA have significant Before providing details on the case of Mrs. A, we analyze the levels of anxiety, as revealed by questionnaires filled by the current state of LA in aphasia. patients’ relatives, being this prevalence higher than the one found in the general stroke population (20–24%) (Burton et al., 2013; Shehata et al., 2015). Another study, based in self-reports FRAMEWORK FOR STUDIES ON of a sample of stroke patients, found no significant differences LINGUISTIC ANXIETY in anxiety prevalence between an aphasic group and a non- aphasic group (Døli et al., 2017). However, the authors did find Persons with aphasia logically show greater perceived stress than higher anxiety level (i.e., more reported symptoms) was than healthy subjects during linguistic tasks (Laures-Gore et al., associated with impaired repetition and comprehension deficits 2007), and this stress response may interfere with their linguistic (Døli et al., 2017). performance. Multiple variables may modulate the subjective An important area of concern and enquiry among emotional response triggered when the PWA is confronted with neuropsychologists and speech pathologists working with situations that need to be solved using language. Laures-Gore and PWA is how anxiety levels affect linguistic performance and Buchanan (2015) proposed a multifactorial framework for the how this may interfere with clinical decision making. The term study of LA in aphasia positing that the stress response in a given Frontiers in Psychology | www.frontiersin.org 2 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia language-situation is influenced by different individual factors. In language performance, at least in subjects aware of their deficits, the present study, we focused on two individual factors that we increasing the likelihood of LA. However, since self-efficacy suggest have special relevance in the causality of LA in Mrs. A. is a subjective judgment, LA is also present in mild aphasias First, we consider the role of premorbid personality traits such as (Cahana-Amitay et al., 2015). excessive perfectionism and high self-criticism. These obsessional Another individual factor that may influence the expression of traits may interact with other individual factors (i.e., expectation LA in PWA is the location and extension of the lesion. Anxiety of improvement, awareness of failures) by self-imposing a high triggers a hyperactivation of the sympatho-adrenomedullary performance to comply with the demands of language testing, (SA) and the hypothalamic-pituitary-adrenocortical axes (HPA) and ultimately potentiating the LA response. Second, we consider (Ulrich-Lai and Herman, 2009), a response regulated by cortical the role that direct and remote (e.g., due to diaschisis) effects of and subcortical brain mechanisms. Overlapping circuits in the the lesion may play in both the subjective emotional appraisal of limbic forebrain, the hypothalamus, and the brainstem are the the situation (i.e., language testing) and the related physiological main components of the physiological response to stress as response. Lastly, we also take into account other factors that they can generate a rapid reaction to re-establish and maintain may influence the emergence of LA and that are related to the the homeostatic status through direct connections with the context of evaluation and the difficulty of the tasks required to SA and HPA (Ulrich-Lai and Herman, 2009). The role of perform, in direct relation with the profile and severity of the limbic and paralimbic cortices such as amygdala, hippocampus aphasia. In the next section we review the mentioned factors (HC), medial prefrontal cortex (mPFC), anterior cingulate cortex to provide evidence on their contribution to the response of (ACC), and orbitofrontal cortex (OFC) is related to the subjective LA. Similarly, we also review the characteristics of the anxiety appraisal of immediate and potential stressors and the high- response both in healthy subjects and the existing evidence in order regulation of the physiological response via subcortical subjects with brain damage. connections (Ulrich-Lai and Herman, 2009; Shin and Liberzon, 2010; Fiddick, 2011). Individual Factors Influencing Anxiety Thus, the cortical-subcortical interplay seems to be crucial for Response the emotional regulation of anxiety. Anatomical and functional Premorbid personality traits, self-efficacy, aphasia severity, and differences together with changes in this network induced by brain areas affected by the lesion may be of special importance local or remote effects of a lesion may explain interindividual in the emergence of LA. In some cases, post-stroke anxiety differences in the level of perceived anxiety and the resultant may not only be the consequence of brain damage (Robinson physiological response (Critchley et al., 2003; Buchanan et al., and Starkstein, 1990), or be linked to the imposed life changes, 2009, 2010; Hahn et al., 2011; Makovac et al., 2016). Like but it may also reflect premorbid mood profiles or disorders. this, the disruption of the fronto-amygdalar network has been Burvill et al. (1995) found that 33.3% of men and 50% of proposed as a fingerprint of anxiety disorders. Decreased resting women with post-stroke anxiety disorders showed evidence of state functional connectivity (rs-fMRI) between the amygdala either depression or anxiety before the stroke onset (see also and the OFC has been identified in social anxiety disorders Sembi et al., 1998; Leppävuori et al., 2003 for similar data (Hahn et al., 2011). Additionally, autonomic response in brain on this issue, and Burton et al., 2013 for a review). Thus, it damaged patients may be reduced due to disruption in the seems reasonable to suggest that some premorbid individual above-mentioned network (Buchanan et al., 2010). characteristics, such as an obsessional personality style, might have an important role in triggering anxiety in PWA. Obsessional Task and Context Related Factors traits include perfectionism, disproportionate preoccupation for details, order and organization, and excessive devotion to Influencing the Anxiety Response work and productivity (American Psychiatric Association, 2013). Clinical contexts may induce an excessive stress response in many High comorbidity of obsessional traits, obsessive-compulsive subjects. In the medical field, this phenomenon is known as disorder and other anxiety disorders have been reported the “white coat effect” and it refers to the autonomic reactivity (Pena-Garijo et al., 2013)1 . induced by the mere presence of a physician (Pickering et al., Further, the interaction between low self-efficacy - defined 2002). Nevertheless, neuropsychological assessment may even as the perceived (in)ability to perform a given task (Bandura, be a more powerful stress-inducing situation since it also 1982) - and high levels of anxiety has been widely studied in involves direct testing. Anxiety elicited by a testing situation the psychology field (Bandura, 1988). Studies of foreign language has been associated with lower cognitive performance (e.g., in learning suggest that low self-efficacy negatively correlates with working memory tasks) (Coy et al., 2011), so that language the level of LA, and this relationship determines language evaluation by professionals in PWA can be viewed by as a achievements (Cheng et al., 1999; Chen and Lin, 2009). In potential stressor. In addition, a factor directly related to LA this vein, it seems obvious that the presence of a severe response is the difficulty of the task, which is also related aphasia may itself favor the perception of low self-efficacy in to the individual profile of aphasia. For instance, speech production tasks would imply more difficulty in a person 1 with non-fluent aphasia (e.g., Broca’s aphasia), than in cases Note that obsessive–compulsive disorder was considered an anxiety disorder until the last version of the Diagnostic and Statistical Manual of Mental Disorders of fluent aphasia with impaired and reduced awareness of (DSM-5) (American Psychiatric Association, 2013). language deficits. Frontiers in Psychology | www.frontiersin.org 3 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia Linguistic Anxiety Response: woman (24 years of formal education), previously working as a Physiological Correlates and Linguistic full professor of mathematics at a Spanish University. Mrs. A was referred to our unit for aphasia evaluation 27 months post-onset. Performance At that moment, she had a mild right hemiparesis, right visual Previous studies in healthy subjects suggest that anxiety response field defect and a severe aphasia. Mrs. A had a severe non-fluent is characterized by an increase in the respiration rate, sympathetic aphasia characterized by markedly reduced spontaneous speech activation [increased heart rate (HR), electrodermal activity and and almost nil auditory comprehension in the face of preserved blood pressure] (Kreibig, 2010), and cortisol levels (Laures-Gore language repetition (MTCA) (Berthier, 1999). Note that this et al., 2007). It remains to be determined, however, whether type of aphasia is equivalent to global aphasia, except for the the autonomic correlates of anxiety in healthy subjects and in preservation of repetition capacity. Automatic echolalia, verbal PWA are the same, and which specific factors may influence perseverations, and ready-made expressions were very frequent the physiological response to stressful situations in PWA. Few in Mrs. A. Further, she could sing overlearned songs without studies on this topic have been reported up to now. Cahana- prompting using excellent lyrics and melody. Amitay et al. (2015) reported increased HR during a speech Several episodes of LA were detected in Mrs. A throughout preparation task in comparison to a non-linguistic task in a the language evaluations. During language testing sessions she person with mild aphasia, but no prominent changes were would frequently get frustrated and when she was not able to observed in electrodermal activity (EDA). Further, a study of perform a linguistic task, she blocked out and reiteratively said a person with acquired apraxia of speech reported increased “I don’t know, I don’t know, I don’t know. . .”. After that, the systolic and diastolic blood pressure when speech therapy was evaluation had to be stopped, jumping to the next task since she administered by a clinician (stressful context) in comparison to refused to continue with that task. In other occasions, when she a conditions where the patient self-directed the same treatment was blocked out with an item that she did not understand or (Sakamoto et al., 1999), thus suggesting that persons with speech was not able to name, she would go back to the items, even after disorders are highly susceptible to stress induced by clinical several items had passed, and tried to get feedback (on the name settings. In a study of cortisol reactivity, Laures-Gore et al. (2007) or meaning) from the therapist. This was usually accompanied found that the linguistic task did not induce changes in cortisol by face expressions and body gestures indicating worry and level, although the participants reported higher perceived stress tension. We observed that during testing her verbal behavior than in a control non-linguistic task. Nonetheless, in a later study was replete with echolalic emissions, which markedly increased these authors did find cortisol reactivity induced by a speech when she was anxious and upset. However, Mrs. A’s husband production task in PWA (Laures-Gore et al., 2010). reported that echolalia was not excessively pronounced in other Importantly, LA in PWA has psychological and behavioral contexts (e.g., home, interactions with family members). The consequences that may negatively impact speech performance presence of severe LA and worsening of language performance interfering with recovery. For instance, in the abovementioned when Mrs. A was assessed in a clinical setting by a therapist study Cahana-Amitay et al. (2015) showed that autonomic jeopardized the interpretation of cognitive evaluations, leading to reactivity was accompanied by lower discourse productivity biased conclusions about the severity of the language alteration in a PWA. In healthy subjects, public speaking tasks induce and the effectiveness of the treatments. However, when language autonomic changes that are related to reduced speech fluency evaluation was performed by Mrs. A’s husband her performance (i.e., higher pause time) (Buchanan et al., 2014). More evidence improved significantly. on the detrimental effect of anxiety in language performance comes from studies that examined the impact of anxiety in Study Design the performance of a foreign/non-mastered language. Significant Mrs. A underwent 5 formal language evaluations in our negative relationship between second language performance and unit over a 1-year period (see Figure 2). Baseline evaluation anxiety is frequently reported in a variety of language domains (T0) was completed 27 months after stroke. Thereafter, she such as speaking (Phillips, 1992; Aida, 1994; Woodrow, 2006), began pharmacological treatment with Rotigotine (transdermal listening (Zhang, 2013), reading (Saito et al., 1999), and writing patches) alone, a drug that modulates the dopaminergic system (Cheng et al., 1999). (Gorgoraptis et al., 2012). Mrs. A initially received Rotigotine alone (without language therapy) for 4 months. Three language evaluations were performed during this period [month 28 (T1); MATERIALS AND METHODS month 29 (T2); and month 31 (T3)]. She also received treatment with oral Domperidone to prevent nausea and vomiting. Case Description Assessment at T2 consisted of two parts. In the first part, Mrs. Mrs. A was a 66-years-old right-handed woman who suffered A underwent a formal language evaluation conducted by one a stroke the day of her professional retirement, causing her therapist (the same therapist that assessed her at T0 and T1) in two simultaneous hemorrhagic lesions that affected frontal and a quiet room. In the second part, Mrs. A was invited to move to parietal areas in the left hemisphere, partially sparing the an adjacent unit where she was evaluated by a different therapist core perisylvian language areas (see Figure 1). Mrs. A was while her autonomic response was monitored. Three persons a monolingual native Spanish speaker with normal language were present at the time of this latest evaluation, one physician development during childhood. She was a highly educated who registered the autonomic activity and two therapists who Frontiers in Psychology | www.frontiersin.org 4 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia FIGURE 1 | T1-weighted MRI axial images depicting extensive brain damage in the left hemisphere. There are two old cavitated lesions, one involving the left dorsolateral and medial frontal cortex with partial compromise of the anterior perisylvian language area (inferior frontal gyrus, pars triangularis), whereas the other lesion involved the left inferior parietal lobe (angular gyrus) and medial occipital lobe. There are also marked asymmetric dilatation of the lateral ventricle especially of the left frontal and temporal horns. Note that great part of the left perisylvian language core is spared by the lesions and no damage was observed in the right hemisphere. Images are shown in neurological convention. L, left; R, right. FIGURE 2 | A schematic diagram of the pharmacological treatment and aphasia therapy shows baseline and different type points under treatment with Rotigotine alone and Rotigotine combined with Intensive Language-Action Therapy (ILAT). ∗ The autonomic study was performed in T2. In T4, the language assessment was divided in two parts. First, the therapist performed the first assessment, and 1 week after, the language evaluation was performed by Mrs. A’s husband. assessed and recorded Mrs. A’s verbal responses during three of post-stroke aphasia (see Berthier, 2005) considered that different language tasks. the prescription of this agent for an unapproved use (post- Thereafter, treatment with Rotigotine was combined with stroke aphasia) was appropriate for this particular case (Wittich aphasia therapy using the Spanish version of an Intensive et al., 2012). Moreover, since post-stroke MTCA like the one Language-Action therapy (ILAT) (Pulvermüller and Berthier, presented by Mrs. A is an exceptional syndrome (4 out of in 2008; Difrancesco et al., 2012; Berthier et al., 2014b). Mrs. 1.200 cases – 0.33%) (Bogousslavsky et al., 1988), the most A underwent a final evaluation 6 months after combined therapy viable option was to treat her in a clinical setting. Taking this (Rotigotine plus ILAT) [month 37 (T4)]. At this endpoint, we into consideration, ethics approval by an ethic committee was conducted two independent testing sessions. First, a language not required in accordance with applicable institutional and battery was administered by an experienced therapist (same national guidelines and regulations. Therefore, Rotigotine was therapist that have evaluated her in the previous time points) used according to the statement of ethical principles for medical and, 1 week later, the same battery was administered by Mrs. research involving human subjects of the Declaration of Helsinki A’s husband (see Language Assessment and Aphasia Therapy (section 37: Unproven Interventions in Clinical Practice). Mrs. Administration by a Relative). A and her husband were provided with the package leaflet of Rotigotine and they were also fully informed about the pharmacological characteristics, potential benefits and adverse Pharmacological Treatment events of the drug and neuroimaging (18-FDG-PET and MRI) Mrs. A was treated with the non-ergoline dopamine agonist acquisition. They were also fully informed about all the procedure Rotigotine (with high affinity for D1, D2, and D3 receptors) across the study (language evaluations, autonomic recording, to augment her markedly reduced speech production and other neuroimaging). Mrs A and her husband agreed to receive the language deficits (see Gill and Leff, 2014). Rotigotine was treatment. Written informed consent for research participation selected over more traditional ergoline dopaminergic agents, like in this study as well as for the publication of the case report bromocriptine, because it has better tolerance and adverse event was obtained from Mrs. A and from Mrs. A’s husband. During profile (cardiac valve fibrosis). She started with a dose of 2 mg/day both the titration phase (4 weeks) and whole drug treatment he for 1 week and then the dose was gradually tapered at a rate of was contacted to detect potential adverse event and evaluate the 2 mg per week up to 8 mg/day. adherence to the drug. At present, Rotigotine is approved for the treatment of Parkinson’s disease and restless legs syndrome (Benitez et al., 2014). However, given the favorable safety and efficacy profiles Language Therapy of Rotigotine found in stroke patients (Gorgoraptis et al., The ILAT is a language therapy based on three principles 2012; Zarola, 2018), the senior author (MLB), a board-certified that includes: (1) intensive practice, (2) use of language action neurologist with experience in the pharmacological treatment games to introduce a meaningful communication setting; and Frontiers in Psychology | www.frontiersin.org 5 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia (3) promotion of the use of verbal language, avoiding non-verbal developed after the stroke. The LOI also contains Resistance communication. It requires massive practice over a short period and Interference scales (Cooper, 1970) which evaluate symptom of time (3 h per day for 2 weeks) and guidance by modeling. severity. These two scales were no administered to Mrs. A’s Although it is normally applied in small groups as a therapeutic husband and son because they cannot rate the degree of game context (2 or 3 participants and a therapist), it can also interference in daily life caused by obsessions and compulsions be applied in a dyad (participant and therapist). Participants are as well as her effort to resist obsessive thoughts and compulsions. required to communicate verbally in order to ask for picture cards Moreover, based on the behavior of Mrs. A during language to each other (or to the therapist if the therapy is implemented testing, the therapist completed the Catastrophic Reaction Scale in dyad). For that, they might name or describe the depicted (CRS) (Starkstein et al., 1993). This is an 11-item scale based pictures, as well as comprehend requests made by others. The on symptoms and signs traditionally described in patients with ILAT comprises 1.100 picture cards of 6 different categories and CR including anxiety, fear, tearfulness, sadness, angry manner, frequencies (nouns, minimal pairs, colors, numbers, adjectives swearing, displacement, refusals, hopelessness, and boasting. It and actions) divided in two identical pairs of 550 cards (see was specially devised to rate the presence and severity of CR in methodology of administration in Difrancesco et al., 2012). In stroke patients. Scores range from 0 (no symptoms) to 3 (severe the present case, the therapy was administered in a dyad by a symptoms) and scores above 8 indicate the presence of CR. These naïve person, Mrs. A’s husband, who was previously trained in two scales were not tested during follow-up evaluations. the application of ILAT. Neuropsychological Testing Language Assessment and Aphasia Baseline evaluation testing (T0) included a detailed evaluation Therapy Administration by a Relative of language and other cognitive functions related to right As abovementioned, it was noted that Mrs. A became anxious and hemisphere functioning (see Table 1). The Aphasia Quotient frustrated during language testing. Therefore, therapists realized (AQ) of the Western Aphasia Battery (WAB) was used to rate that she was not a good candidate to be evaluated and treated by aphasia severity (Kertesz, 1982; Kertesz et al., 1990) across all a clinician and thence it was decided to train her husband in the evaluations. The WAB is a tool to assess linguistic functioning, evaluation of aphasia with the WAB (Kertesz, 1982; Kertesz et al., including the domains of spontaneous speech (fluency and 1990) and in the administration of the ILAT (Difrancesco et al., information content), naming, auditory-verbal comprehension, 2012). This last decision was made because there is evidence that and verbal repetition of words and phrases. For further language ILAT can be efficiently accomplished by trained laypersons with testing, selected subtests of the Psycholinguistic Assessments results comparable to that of experienced therapists (Meinzer of Language Processing in Aphasia (PALPA) (Kay et al., 1992; et al., 2007). Aphasia therapy was administered by Mrs. A’s Valle and Cuetos, 1995), targeting verbal repetition and reading husband after T3. He was a retired full professor of mathematics domains were used. Additional verbal repetition tasks with and he was deeply implicated with the treatment of Mrs. A. different linguistic stimuli were also used (see Table 1). Every time they visited our Unit for a follow-up, he brought In the following evaluations (T1, first part of T2, T3, and T4), notes with his every-day observations referred to the verbal and only the WAB was administered to Mrs. A (see Table 2). For communicative behavior showed by Mrs. A, highlighting day- the second part of T2, consisting in language assessment while to-day changes in her language and behavior. Before receiving autonomic response was acquired, three subtests of the WAB training with ILAT, he had spent about 1 h every day with Mrs. were used, two of them targeting auditory-verbal comprehension A training naming and writing skills since the stroke onset. (Yes/No Questions (YNQ), Auditory Word Recognition (AWR); Mrs. A’s husband was provided with the ILAT materials and one targeting Verbal Repetition (VR)] (see the Autonomic (manual of administration, set of cards, and registration sheets) Response Monitoring During Language Assessment). (Berthier et al., 2014b). He read the manual at home and during the following week, one of us explained the therapy and trained Neuropsychiatric Testing him through role-playing. Once he felt comfortable with the The Leyton Obsessional Inventory (LOI) (Cooper, 1970) was therapy, he started to apply it in a dyad to Mrs. A in a daily basis, used to identify whether Mrs. A had premorbid obsessive- 6 days/week (3 h/day). We also asked him to record fragments compulsive symptoms and traits (e.g., perfectionist, and self- of the therapy administration 2 time/week for the first 2 weeks criticism). The LOI is a self-rated scale that quantifies the range in order to provide him feedback of the execution, and after that of obsessional thoughts and compulsive behaviors. It contains telephonic interviews were regularly made with him to resolve 69 “yes/no” questions dealing with the subjective assessment any question related to the application of ILAT. He also was of obsessional symptoms (questions 1–46) and traits (questions trained in the use of the WAB. First, general information about 47–69) as well as the degree of resistance and interference the test was given, and then we pointed important aspects to with the patient’s life. Since Mrs. A had a severe aphasia, the consider in the application of each subtest. He was instructed LOI was completed by her husband and son independently to to allow her wife time to answer, without giving phonological ascertain agreement and disagreement in the type and number or semantic cues except for the object naming task, just when of responses. Both subjects were instructed to complete the enough time to answer was given. Also, he was informed to LOI by taken in consideration the premorbid symptoms and prevent her wife for lip-reading in the repetition task as well as traits of Mrs. A and not the potential changes in her behavior how to place the elements (pen, booklet, and comb) used in the Frontiers in Psychology | www.frontiersin.org 6 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia TABLE 1 | Neuropsychological baseline evaluation. Language assessment Score Normative data Western Aphasia Battery (WAB) Aphasia Quotient (AQ) (max = 100) 26 99.6 ± 0.3a,b Spontaneous Speech (max = 20) 0 20b Information Content (max = 10) 0 Fluency (max = 10) 0 Auditory Verbal Comprehension (max = 10) 2.1 10b Yes/No Question (max = 60) 33 Auditory Word Recognition (max = 60) 9 Sequential Commands (max = 80) 0 Repetition (max = 10) 9.2 9.9 ± 0.1b Naming and Word Finding (max = 10) 1.7 9.8 ± 0.1b Object Naming (max = 60) 15 Word Fluency (max = 20) 0 Sentences Completion (max = 10) 2 Responsive Speech (max = 10) 0 Sentences completion (max = 40) 9 39.3 ± 0.6c Verbal Repetition Tests PALPA 9. Repetition: Imageability × Frequency (max = 160) 135 Word repetition (max = 80) 80 78.77 ± 1.93b Non-word repetition (max = 80) 55 77.68 ± 3.35b Word triplets (high-frequency) Random combination (max = 20) 20 19 ± 0.8c Constrained information (max = 20) 20 19.4 ± 0.6c Word triplets (low-frequency) Random combination (max = 20) 18 17 ± 2.5c Constrained information (max = 20) 20 18.7 ± 1.2c Idiomatic clichés (max = 40) 35 39.45 ± 0.9c Novel sentences (max = 40) 36 38.75 ± 1.1c Reading PALPA 29. Imageability × Frequency (max = 80) 73 79.53 ± 0.55b PALPA 30. Grammatical class (max = 80) 61 79.53 ± 0.42b PALPA 31. Grammatical class and Imageability (max = 40) 33 39.91 ± 0.29b PALPA 32. Morphology (max = 30)d 25 29.61 ± 0.46b PALPA 33. Regularity (max = 60) 43 55.9 ± 5.84b PALPA 34. Non-words (max = 24) 6 23.22 ± 0.94b Right Hemisphere Testing Classification (Percentile rank) Raven’s Colored Progressive Matrices 26 Bellow averagee Visual Form Discrimination Testf 30 Superior (86+) Facial Recognition testf 48 High average (72–85) a Patients with an Aphasia Quotient score under 97.3 are considered to be aphasics. b Normative data was extracted from their reference manuals, i.e., WAB subtests from the revised version of the Western Aphasia Battery (Kertesz, 2007); and PALPA subtests from the Spanish version of PALPA (i.e., EPLA) (Valle and Cuetos, 1995) c Normative data was extracted from Berthier (2001). d Only 30 out of the 60 stimuli included in this subtest were used. e Normative score based on patient’s age and education level is 32.8 (±3.9) (Measso et al., 2009). f Test from Benton (1994). Scores below the normative data are highlighted in bold. Sequential Commands subtest. At the time he administered the autonomic responses were monitored while she was required to WAB he recorded the whole application. WAB correction was perform the three selected subtests of the WAB. The recorded done by two therapists independently. The final score was reach autonomic variables included: electrodermal activity (EDA), by agreement between the two therapists. electrocardiogram (ECG) and instantaneous heart rate (HR), mean arterial blood pressure (ABP), systolic blood pressure Autonomic Response Monitoring During (SBP), and diastolic blood pressure (DBP). ECG and HR were Language Assessment recorded in DII. The corresponding devices for monitoring In order to have a measure of the impact of language ABP and EDA were placed at the left arm. Autonomic assessment in the physiological arousal of the Mrs. A, different variables were acquired using the Biopac MP150 equipped with Frontiers in Psychology | www.frontiersin.org 7 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia TABLE 2 | Results from language testing across evaluations. BL Rotigotine Treatment T0 T1 T2 T3 Western Aphasia Battery (WAB) Aphasia Quotient (AQ) (max = 100) 26 32.6 33.5 34.2 Spontaneous Speech (max = 20) 0 2 3 3 Information Content (max = 10) 0 2 3 3 Fluency (max = 10) 0 0 0 0 Auditory Verbal Comprehension (max = 10) 2.1 3.1 2.75 2.9 Yes/No Question (max = 60) 33 51 42 42 Auditory Word Recognition (max = 60) 9 11 13 16 Sequential Commands (max = 80) 0 0 0 0 Repetition (max = 10) 9.2 10 9.8 9.4 Naming and Word Finding (max = 10) 1.7 1.2 1.2 1.8 Object Naming (max = 60) 15 10 8 12 Word Fluency (max = 20) 0 0 0 0 Sentences Completion (max = 10) 2 2 4 4 Responsive Speech (max = 10) 0 0 0 0 Rotigotine + Speech Therapy T4 Therapist Relative Evaluation Evaluation Western Aphasia Battery (WAB) Aphasia Quotient (AQ) (max = 100) 35.3 47.4 Spontaneous Speech (max = 20) 3 7 Information Content (max = 10) 3 4 Fluency (max = 10) 0 3 Auditory Verbal Comprehension (max = 10) 3.35 3.4 Yes/No Question (max = 60) 42 42 Auditory Word Recognition (max = 60) 25 26 Sequential Commands (max = 80) 0 0 Repetition (max = 10) 10 10 Naming and Word Finding (max = 10) 1.3 3.3 Object Naming (max = 60) 11 21 Word Fluency (max = 20) 0 2 Sentences Completion (max = 10) 2 6 Responsive Speech (max = 10) 0 4 BL, Baseline. an electrocardiographic amplifier SG100B and EDA response were administered in the following order: (1) YNQ, (2) AWR, amplifier GSR100B. Continuous beat-to-beat arterial blood and (3) VR. The administration of each test was separated by a pressure was registered with bmeyer’s Nexfin. EDA was used as 3-min rest period. an index of implicit emotional state that may emerge beyond The selection of the WAB subtests was based on the explicit intent. All variables were analyzed using AcqKnowledge difficulty that they represented for the patient by considering software version 5 (Biopac Systems, Inc.). Evaluation was carried her performance in previous evaluation as well as the profile out in a quiet room of the Lab of Autonomic Nervous System. and severity of aphasia. The aim was to compare the autonomic Mrs. A was seated at a comfortable armchair with her left arm response triggered by highly demanding linguistic tasks against a placed at the armrest, and was instructed to be relaxed, attending low demanding task and rest (baseline). As high-demanding tasks to what the therapist was going to ask her and to do not an auditory comprehension task that required verbal responses move the left arm during the evaluation. Before starting testing (YNQ), and an auditory comprehension task that required motor session, a 10-min rest period was introduced while listening to responses (AWR) were selected. As a low-demanding task, a calm music. Baseline autonomic measures were registered at this repetition task composed of word, numbers and phrases (VR) was period. After this baseline period, the three subtests of the WAB selected. Other production subtests of the WAB (naming, verbal Frontiers in Psychology | www.frontiersin.org 8 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia fluency) could not be used to evaluate autonomic responses order. Thus, for all the reported ANOVA analyses, the different because the patient performed at floor and could have led her levels of the factor “task” were entered in the following order: to disengage from the task, as previously suggested (Laures- B, YNQ, AWR, and VR. Keeping this order fixed allowed to the Gore et al., 2010). It was hypothesized that the autonomic interpretation of potential significant quadratic effects positing reactivity would be more evident in the two tasks in which that a given autonomic response increases from the baseline to Mrs. A performed worse (high language demand) than in the task the YNQ, then kept stable during the AWR, and then from in which she showed a nearly perfect execution (low language the AWR to the VR task it decreases again. When a significant demand). YNQ and AWR stimuli were presented every 30 s, effect was found, Bonferroni post hoc multiple comparison test while VR items were presented every 15 s. This was done because was applied to explore specific pair-wise differences between Mrs. A’s verbal repetition was fully preserved and, hence, she conditions. Statistical analysis was performed with IBM SPSS could automatically repeat any verbal stimulus (words, numbers statistic software version 23.0. and phrases) immediately (during the first 2 or 3 s) after the auditory presentation. On the other hand, since the YNQ and Neuroimaging Acquisition AWR entailed more difficulty, longer time was needed to answer. MRI studies were performed on a 3-T MRI scanner (Philips AcqKnowledge software version 5 (Biopac Systems, Inc.) was Intera, Amsterdam, Netherlands), Release 3.2.3.4, with a used to extract autonomic data. First, each channel, reflecting MASTER gradient system (nominal maximum gradient each of the autonomic variables, was visually inspected for strength = 30 mT/m, maximum slew rate = 150 mT/m/ms), artifacts. In very few cases artifact coincided with the time after equipped with a six-channel Philips SENSE head coil. Head stimulus presentation (time in which the patient was requested movements were minimized using head pads and a forehead to give an answer). In those cases, data for that item had to strap. High-resolution T-1 structural images of the whole brain be discarded. Second, autonomic data for each stimulus was were acquired with three-dimensional (3D) magnetization manually extracted. For all three tasks, the average and the prepared rapid acquisition gradient echo (3D MPRAGE) maximum and minimum levels of each measure was calculated sequence. Structural brain scans were acquired for the patient for each event (time interval between the end of a given stimulus and for 25 healthy controls matched by age (mean age: and the beginning of the next one). In the YNQ and AWR tasks, 58 ± 5 years; range: 48–67 years) (Crawford t-test: t = 1.42, values (average, maximum and minimum) were determined in p = 0.16, two-tailed) and years of education (mean years of the 0 – 30 s time interval after the presentation of the stimulus, education: 18.9 ± 3.9 years; range: 10–25 years) (Crawford t-test: whereas in the VR task the values were calculated in the 0 – 15 s t = 1.26, p = 0.22, two-tailed). The acquisition parameters for the time interval after the stimulus presentation. To get baseline control group were: echo time (TE) about 4.6 ms; repetition time measures that can be statistically compared to the tasks, the (TR) about 9.9 ms; acquisition matrix 240/200; field of view 240; 10 min baseline recording was fragmented in events of 30 s. turbo field echo (TFE) factor 200; flip angle 8◦ ; reconstruction Afterward, the change in each autonomic measure was calculated voxel size 1 × 0.94 × 0.94 mm; 190 contiguous slices; total by subtracting the minimum value from the maximum in the acquisition time of the sequences about 170 s. The acquisition time interval after the stimulus. This was done separately for parameters for Mrs. A were as follows: echo time (TE) = 4.5 ms; each stimulus. Values above or below 2 standard deviations repetition time (TR) = 9.8 ms; acquisition matrix 312/297; field from the mean were considered outliers and were removed of view 250; turbo field echo (TFE) factor 179; flip angle 8◦ ; from the analysis. reconstruction voxel size 0.80 × 0.78 × 0.78 mm; 220 contiguous slices; total acquisition time of the sequence was 172 s. Positron emission tomography (PET) data acquisitions were Statistical Analyses of the Autonomic performed on a Discovery ST PET/CT camera (General Electric, Response During Language Evaluation Milwaukee, WI) after an intravenous injection of about 3.3 To analyze the effect that the different language subtests had MBq/Kg for 25 healthy controls and 3.9 MBq/Kg for Mrs. A. PET on the mean and change (maximum – minimum) of EDA, HR, images were reconstructed using 3D FORE-IR algorithm with CT ABP, SBP and DBP, 10 repeated-measures ANOVAs with task attenuation correction (Matrix size, 128 × 128 × 47; voxel size, (4 levels: B, YNQ, AWR, and VR) as within-subject factor, were 2.34 × 2.34 × 3.27 mm). conducted. Given that the number of items varied across tasks Resting state functional MRI (rs-fMRI) was acquired while and in addition outliers and artifacts were removed, the total the patient was instructed to lie still with open eyes, without number of items included in each ANOVA analysis varied. The falling sleep. Two hundred volumes were collected with the final n included in each analysis is specified in the result section. following parameters: FFE/EPI sequence, TR = 3000 ms, The general hypothesis was that the high-demanding language TE = 35 ms, flip angle 82◦ , FOV = 220 mm, image matrix of tasks (YNQ and AWR) would trigger a stronger emotional 73/128 r, 50 slices with no gap and an acquisition voxel size of response than the low-demanding language task and this should 1.72 mm × 1.72 mm × 3.00 mm. be evidenced in the autonomic responses. Therefore, a trend analysis was performed to ascertain for quadratic non-linear PET Statistical Analysis relationship between the studied autonomic measures and the Spatial preprocessing and statistical analysis were performed with different conditions. Note that to perform a trend analysis, it Statistical Parametric Mapping (SPM12), running on MATLAB is of utmost importance to enter the variables in a meaningful R2016b (Mathworks Inc., Natick, MA, United States). All T-1 Frontiers in Psychology | www.frontiersin.org 9 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia structural images and 18F-FDG-PET images were manually data normalization (Brett et al., 2001). For this purpose, a aligned to anterior-posterior commissure (AC-PC) orientation. binary lesion mask was created by manually delineating the Reoriented PET images were then coregistered with T-1 images. damaged areas directly on the T-1 image using MRIcron software A lesion mask, drawn over the T-1 image of the patient in (Rorden and Brett, 2000). The normalization parameters resulted native space, was applied to the T-1 image. Then, the coregistered from the T-1 segmentation were used for the normalization of volumes and the lesion mask were spatially normalized onto the functional images to the standard MNI space. Finally, a the MNI template (McGill University, Montreal, QC, Canada) smoothing with an 8-mm FWHM Gaussian kernel was applied using the T-1 images to calculate the deformation fields. The to the functional images. size of the resulting voxels was 2 × 2 × 2 mm. The normalized Exploratory seed-driven functional connectivity correlational PET studies were smoothed with a FWHM 8-mm Gaussian analysis were performed with the CONN functional connectivity kernel. Histogram-based intensity normalization was performed Toolbox v174 (Whitfield-Gabrieli and Nieto-Castanon, 2012). using in-house software. In this procedure, the smoothed images After preprocessing, non-neuronal sources of noise were of each subject were voxel-wise divided by the mean of the removed using CompCor method implemented in CONN normalized and smoothed images of the healthy controls. toolbox. Specifically, the signal resulting from the voxels Histograms of these masked ratio images were generated, corresponding with the cerebrospinal fluid, the white matter excluding damaged areas and ventricles. Finally, each original and the brain lesion, as well as the motion parameters derived smoothed PET study was divided by the most prevalent value in from the realignment preprocessing step, were removed with its ratio image. regression. A band-pass filter of 0.008 Hz – 0.09 Hz was A region of interest (ROI) analysis was performed with applied to the residual BOLD time course to remove the the preprocessed PET studies. An average normalized activity effects of low frequency and high frequency noise outside concentration was calculated for each ROI and subject, excluding our range of interest. ROI-based analysis was carried out damaged areas from the ROIs of Mrs. A. Then a general linear by computing the temporal correlation between the BOLD model was applied for each ROI with sex and group (control signals from each voxel within the seed ROIs to each voxel or patient) as regressors. After estimating the coefficients with within the target ROIs. For that, bi-variate correlations were a robust fit, a t-test was applied for the group regressor, and performed between each pair of ROIs. ROIs were selected the z-score for each ROI was then estimated. An absolute in base on their association to either language processing or z-score ≥ 1.96 reflects a p-value ≤ 0.05; and a z-score ≥ 2.58 anxiety/stress response. Specifically, the selected language-related reflects a p-value ≤ 0.01, meaning that a given ROI shows ROIs were the IFG (pTr, pOp), insula, posterior STG (pSTG), significant differences in metabolism between the patient and posterior MTG (pMTG), SMG, and AG; whereas the selected the group of healthy controls. The ROIs included in the anxiety-related ROIs were ACC, OFC, mPFC, amygdala, HC analysis were the amygdala, cingulate cortex (CC), OFC, HC, and PHC. All the ROIs were used as seed and as target parahippocampal gyrus (PHC) and perisylvian areas (PSA) ROIs in both hemispheres. For the rs-fMRI, we do not have in both hemispheres. The PSA ROI comprised the rolandic a control sample, therefore only the single-level 1st level operculum, subareas of the inferior frontal gyrus (IFG) [pars results are presented in this study. Correlations ≥ 0.35 are triangularis (pTr), pars opercularis (pOp) and pars orbitalis considered meaningful. (pOr), precentral gyrus, postcentral gyrus, supramarginal gyrus (SMG), angular gyrus (AG), inferior parietal cortex, middle temporal gyrus (MTG), inferior temporal gyrus (ITG), superior RESULTS temporal gyrus (STG), and the insula in both hemispheres]. All the ROIs were obtained from the AAL atlas as provided Neuropsychological Findings by the Wake Forest University (WFU) Pickatlas toolbox2 Tables 1, 2 show Mrs. A’s performance across all evaluations. (Maldjian et al., 2003). Baseline evaluation scores reflect that Mrs. A had a severe language impairment affecting spontaneous speech, auditory Resting State Functional Connectivity comprehension and naming. Performance in the repetition Preprocessing and Analysis subtest of the WAB was relative preserved compared to other Resting state image preprocessing was performed with Statistical linguistic domains. However, a more in depth evaluation with Parametric Mapping software (SPM12)3 using a standard pipeline the PALPA and experimental subtests (Berthier et al., 2014a) for resting state analysis. Functional data was AC-PC orientated, showed that repetition of non-words and completion of open- and the functional images were realigned to the first volume ended sentences were markedly impaired, findings reflecting of the series. Coregistration between the mean functional and some involvement of the left PSA (Grossi et al., 1991). Reading the structural T-1 was performed. Cost function masking was comprehension was markedly impaired and oral reading showed used during segmentation of the T-1 image in different tissues the pattern of phonological dyslexia (mild impairment of in order to disregard the damaged tissue when calculating the word reading and marked impairment of non-word reading) normalization transformations to be applied to the functional (Crisp and Lambon Ralph, 2006). 2 http://www.fmri.wfubmc.edu/cms/software 3 4 http://www.fil.ion.ucl.ac.uk/spm/ http://www.nitrc.org/projects/conn Frontiers in Psychology | www.frontiersin.org 10 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia The language testing made by the therapist revealed TABLE 3 | Results of the ROI analysis for the PET studies. clinical improvement on the WAB-AQ score (6.6 pts in Left Right the WAB-AQ) under Rotigotine (8 mg/day) treatment (T1) Amygdala −4.59 −0.15 with minimal point improvement in successive evaluations Cingulate Cortex −6.29 −4.05 during continued treatment with the drug alone (T2: 0.9, Orbitofrontal Cortex −5.29 −0.22 T3: 0.7) and when ILAT was added (T4:1.1). Pharmacological Hippocampus −6.22 0.15 treatment promoted improvements in spontaneous speech and Parahippocampus −5.93 0.86 auditory verbal comprehension, while naming stayed relatively Perisylvian Area −6.36 0.65 stable throughout. Since performance in verbal repetition was almost at celling, no major changes were observed in this The z-scores for the comparison between Mrs. A and the control group are shown. Values over 1.96 or below −1.96 reflect p-values below 0.05; and values higher domain. Surprisingly, although Mrs. A’s relatives reported that than 2.58 or smaller than −2.58 reflect p-values below 0.01. she had improved substantially with both interventions, no further improvements were observed after combined therapy (Rotigotine-ILAT) when she was assessed by the therapist. η2 = 0.83] trend. Post hoc tests revealed that EDA change in B However, when language evaluation was performed by her was lower compared to the change in YNQ (p = 0.000), AWR husband (previously trained), significant improvement in the (p = 0.000) and VR (p = 0.014); and that EDA change in VR severity of aphasia (WAB-AQ) was seen. In fact, a difference was lower compared to YNQ (p = 0.013) and AWR (p = 0.001), of 12.1 points in the WAB-AQ was observed when comparing indicating stronger autonomic reactivity in these two last tasks. the evaluations of the therapist and the husband, having better There was a significant main effect of task in mean HR outcomes in the last one (p = 0.057, Fisher exact test, one tailed) response [F(3,39) = 17.85, p = 0.000, η2 = 0.59]. Both a significant (see Table 2). Mrs. A received around 400 h of ILAT in 140 days. linear [F(1,13) = 25.03, p = 0.000, η2 = 0.65] and a quadratic [F(1,11) = 33.26, p = 0.000, η2 = 0.71] trends were found. Post Neuropsychiatric Findings hoc tests showed that the mean HR response was lower during At baseline, a high agreement in LOI items evaluating excessive VR task compared to B (p = 0.007), YNQ (p = 0.000) and perfectionism, self-demanding, and methodical style was found AWR (p = 0.001). ANOVA analysis on HR change disclosed a when comparing the answers of Mrs. A’ husband and son main effect of task [F(3,39) = 7.86, p = 0.000, η2 = 0.37], and a (see Table 3). Some items in which they agreed answering significant quadratic effect [F(1,13) = 9.98, p = 0.008, η2 = 0.43]. “yes” were: “Are you the sort of person who has to pay Post hoc tests evidenced that the HR change during AWR was a great deal of attention to details?”, “Are you ever over- higher than during both YNQ (p = 0.031) and VR (p = 0.002). conscientious or very strict with yourself?”; “Are you very In reference to the mean ABP, the repeated-measured ANOVA systematic and methodical in your daily life?”; “Do you disclosed a main effect of task [F(3,39) = 19.33, p = 0.000, like to get things done exactly right, down to the smallest η2 = 0.59] and a linear trend [F(1,13) = 37, 37, p = 0.000, detail?”; “Do you think it is important to follow rules and η2 = 0.74]. No quadratic effect was found. Post hoc comparisons regulations exactly?”. Note that both relatives were instructed revealed that mean ABP in B was lower than in YNQ (p = 0.008), to answer by taken in consideration the premorbid symptoms AWR (p = 0.001) and VR (p = 0.000). For ABP change, a main and traits of Mrs. A. The CRS completed by the therapist effect of task [F(3,33) = 4.45, p = 0.01, η2 = 0.28] and a quadratic during observation of Mrs. A’s behaviour yielded a score of 8 trend [F(1,11) = 12. 83, p = 0.004, η2 = 0.53] were found. Post points, just approaching the abnormal scores (Starkstein et al., hoc comparisons disclosed that ABP change during VR was lower 1993). High scores were rated in the following items: “patient than during AWR (p = 0.019). appeared to be anxious”, “complained of feeling anxious or Repeated-measured ANOVA analysis of mean SBP revealed a afraid” (scoring 3 in both indicating severe symptomatology) main effect of task [F(3,39) = 43.93, p = 0.000, η2 = 0.77] and and “patient refused to do something” (scoring 2 indicating a linear [F(1,13) = 61.58, p = 0.000, η2 = 0.82] and quadratic moderate symptomatology). [F(1,13) = 5. 09, p = 0.042, η2 = 0.28] trends. Post hoc tests showed that mean SBP was lower in B compared to YNQ (p = 0.000), Autonomic Response AWR (p = 0.000) and VR (p = 0.000); and that mean SBP was Figure 3 shows the results of the autonomic response analysis. higher in VR compared to YNQ (p = 0.026) and AWR (p = 0.003). Repeated-measures ANOVA analysis for mean EDA revealed a For SBP change, a main effect of task [F(3,33) = 3.11, p = 0.039, significant quadratic trend [F(1,12) = 5.51, p = 0.037, η2 = 0.31]. η2 = 0.22] and a quadratic effect [F(1,11) = 4.98, p = 0.047, No main effect of task was found in mean EDA (p = 0.21). Post η2 = 0.31] were found. Pos-hoc comparisons disclosed that SBP hoc tests using Bonferroni correction revealed that mean EDA change was lower during B than during AWR (p = 0.013). level in YNQ was higher than in B (p = 0.018), whereas no other Finally, referred to mean DBP, ANOVA analysis revealed a significant differences were found in pair-wise comparisons. main effect of task [F(3,39) = 10.29, p = 0.000, η2 = 0.44] and The same ANOVA analysis was performed with EDA change a linear effect [F(1,13) = 21.41, p = 0.000, η2 = 0.62]. Post hoc (maximum value – minimum value, calculated for each trial tests showed that mean DBP was lower in B than in AWR independently), revealing a main effect of task [F(3,33) = 26,67, (p = 0.004) and in VR (p = 0.006). For DBP, a main effect of p = 0.000, η2 = 0.7], and a significant linear [F(1,11) = 18.29, task [F(3, 36) = 15.99, p = 0.000, η2 = 0.57] and a quadratic effect p = 0.001, η2 = 0.62] and quadratic [F(1,11) = 56.43, p = 0.000, [F(1,12) = 35.33, p = 0.000, η2 = 0.74] were found. Post hoc tests Frontiers in Psychology | www.frontiersin.org 11 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia FIGURE 3 | Plots showing the patterns of the different autonomic responses in three different linguistic subtests of the Western Aphasia Battery (WAB): Yes/No Questions (YNQ); Auditory Word recognition (AWR); Verbal Repetition: VR. These tasks entailed high (YNQ and AWR) or low (VR) difficulty for the patient which resulted from the type and severity of the aphasia profile. The range of possible changes in mean activity for each task is represented in the orderly, whereas baseline (B) and the three language tasks are shown in abscissa. Note that AWR has, in addition, a motor component (patient had to point to the correct picture) that may further increase autonomic reactivity. For each task, mean and error bars are plotted. Error bars represent the standard error of the mean (SEM). Frontiers in Psychology | www.frontiersin.org 12 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia disclosed that DBP change was lower both during B (p = 0.000) LA has detrimental consequences on linguistic performance and VR (p = 0.000) compared to AWR. (Cahana-Amitay et al., 2015) and its presence may guide clinical decisions in the wrong direction. In line with previous Neuroimaging Results studies (Buchanan et al., 2014; Cahana-Amitay et al., 2015), the Positron Emission Tomography Results data obtained in the present work patently shows that the ROI analysis in PET revealed significant hypometabolic activity performance of Mrs. A in standardized aphasia batteries (e.g., in the left Amygdala, CC, OFC, HC and some components WAB) was worse when language testing was performed by of the PSA (p = 0.01). In the right hemisphere, significant a therapist (stressful situation) than when assessment was hypometabolic activity was seen only for the CC. No statistical performed by a relative at home. The first evaluation (T1), significant areas of hypermetabolism were found in any of the performed in a clinical setting after 1 month of Rotigotine ROIs; however, higher activation than the mean activation of the treatment (8 mg/day), revealed clinical improvement (>5 points control group was found in the right PSA (z-score = 0.65) and in the WAB-AQ) indicating that Mrs. A was a “responder” right parahippocampus (0.86) (see Table 3). to this intervention (Berthier et al., 2009). This suggests that the powerful effect of the initial dopaminergic stimulation Resting State Functional Connectivity Results with Rotigotine attenuated, at least partially, the negative The resulting functional connectivity matrix reflects the bivariate impact of LA in language testing. Nevertheless, only minimal correlation between each pair of ROIs (Figure 4A). This improvements in formal language evaluations performed by matrix is plotted as a heat-map (warmer color represents the therapist were found thereafter with Rotigotine alone at larger positive correlations whereas green color represents larger the same dose (T2 and T3) as well as during combined negative correlations) (see Figure 4B). Functional connectivity treatment with Rotigotine and ILAT (T4). It was noteworthy at single level reveals a set of regions showing a trend to be that when language evaluation (T4) was administered by Mrs. more functionally coupled. Only the patterns considered more A’s husband, the WAB-AQ yielded significantly higher scores, meaningful for the current study were highlighted here. As increasing the score by 12.1 points in comparison with the a general pattern, there was an increased connectivity among evaluation performed by the therapist. This suggests that LA the ROIs within the right hemisphere. Specifically, within the had a negative impact on Mrs. A’s linguistic performance in subset of language-related ROIs and within the subset of anxiety clinical settings, masking treatment-induced improvements after response-related ROIs. In addition, the right OFC showed a prolonged treatment with Rotigotine and very intensive use of stronger coupling with the bilateral Amygdala and the right ILAT (∼400 h in 140 days!). pMTG, the Amygdala and HC showed stronger functional We documented that several factors may have contributed synchronization with regions within the anxiety network in to Mrs. A‘s negative appraisal of language in testing contexts the right hemisphere and posterior temporo-parietal PSA. increasing the distortion of her language behavior (i.e., worsened Interestingly, the mPFC was highly coupled with the right performance and increased echolalia and perseverations). These language-related PSA (pSTG, pMTG) but it did not show high key factors are described in turn (see Figure 5). First, the correlations with the left PSA. identified premorbid obsessional personality style of Mrs. A (i.e., perfectionism and self-criticism of potential failures) most likely contributed to increment her anxious reactions during language DISCUSSION testing. Second, the severity of aphasia in Mrs. A, with almost inexistent spontaneous speech (Parr, 2007; Thomas and Lincoln, In the present study we reported the case of Mrs. A, a 2008) together with a preserved awareness of her linguistic deficit woman who had a severe chronic MTCA and overwhelming LA may have also favored the strong feeling of low self-efficacy in after suffering two simultaneous left hemisphere hemorrhages. language performance, thus heightening LA (Bandura, 1988). Clinical observations during language testing suggested that Lastly, the direct and remote effects of the lesion may also have an evaluation contexts induced anxiety and catastrophic reactions important role since key areas involved in the subjective appraisal in Mrs. A, which in turn intensified automatic echolalia and of stressors and in the regulation of the associated autonomic and perseverative responses (e.g., “No, no, no. . .”, “I can’t, I can’t, behavioral consequences were affected. I can’t”). In fact, both repetitive verbal behaviors were the The analysis of Mr A’s autonomic reactivity in response to most evident linguistic manifestations of LA. For the study different language tasks under Rotigotine treatment (T2) showed of Mrs. A, we used an integrative approach that included a greater EDA change and mean HR during linguistic tasks for examination of language and cognitive functions, premorbid which she had greater difficulty (i.e., YNQ and AWR) than in personality traits and autonomic response induced by linguistic tasks that evaluated a preserved linguistic domain (i.e., VR). tasks as well as the exploration of the neural basis of LA using Specifically, both EDA change and mean HR across YNQ and multimodal neuroimaging. AWR were greater compared to the same indexes measured Since there are not suitable verbal self-reports of LA in during VR. In fact, EDA change for VR was not different to EDA persons with severe aphasia, its detection becomes difficult in change observed during baseline recording. The greater change this population (Morris et al., 2017). Therefore, identifying in the AWR should be taken with care since it could be boosted linguistic and non-linguistic warning signs that may transpire by the motor component of this task in which Mrs. A had to discomfort and anxiety has an important clinical value. In fact, point to the picture named by the examiner. In contrast, the YNQ Frontiers in Psychology | www.frontiersin.org 13 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia FIGURE 4 | Resting state functional connectivity results. (A) Regions of interest (ROI) used for functional connectivity analyses. Blue balls represent the ROIs associated to the anxiety response circuits, whereas pink balls represent ROIs related to language processing. (B) Heat map representing the matrix of bi-variate correlations values between each pair of ROIs for both hemispheres. The scale color bar located at the right side of the figure indicates the degree of connectivity. The boxes representing a correlation greater than 0.35 or smaller than –0.35 show the edges highlighted in black. ∗∗ mPFC and ACC were bilateral ROIs. ACC, anterior cingulate cortex; OFC, orbitofrontal cortex; mPFC, medial prefrontal cortex; HC, hippocampus; PHC, parahippocampus; SMG, supramarginal gyrus; AG, angular gyrus; IFG pOp, inferior frontal gyrus pars opercularis; IFG pTr, inferior frontal gryus pars triangularis; HG, Heschl’s gyrus; pSTG, posterior superior temporal gyrus; pMTG, posterior middle temporal gyrus. Frontiers in Psychology | www.frontiersin.org 14 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia FIGURE 5 | Model proposed to describe the linguistic anxiety response in Mrs. A. The schema depicts different individual and task-related factors that may interact among them, resulting ultimately in the perception of language as a threat, triggering the stress response and negatively modulating language performance. OC, obsessive–compulsive; EDA, electrodermal activity; HR, heart rate; BP, blood pressure. does not have a motor component; therefore, greater reactivity and heart rate response to tasks manipulation observed in in this task is likely to be associated with the anxiety response Mrs. A might be a consequence of extensive direct damage evoked by the challenging language context. We did not find the as well as the remote structural, metabolic and functional expected pattern in either mean or blood pressure change (ABP, connectivity changes induced by the brain lesions (diaschisis) SBP and DBP) or HR change. Thus, the autonomic study of Mrs. (Carrera and Tononi, 2014). Abnormal cardiovascular response A revealed the expected pattern for some measures (i.e., EDA) but in patients with focal damage to the anterior CC confronting not for others (i.e., ABP, SBP, DBP). with cognitive stressful situations has been previously reported Significant changes in blood pressure and HR in subjects (Buchanan et al., 2010). exposed to stressful situations have been reported in brain Understanding the nature of LA response and its negative damaged and healthy subjects. For instance, a previous study of effects in Mrs. A’s language performance, requires considering anxiety in a person with acquired apraxia of speech reported more that the structural and functional metabolic changes, besides striking changes in ABP (SBP and DBP) and HR when therapy affecting some components of the left PSA, additionally involved was performed by a therapist (stressful context) than during a limbic and paralimbic areas rarely damaged in PWA (Kreisler self-administered therapy (Sakamoto et al., 1999). Experimental et al., 2000; Fridriksson et al., 2018) (see Table 3). Studies of studies inducing anxiety in healthy subjects showed greater the neural basis of anxiety in healthy population have provided autonomic reactivity than the observed in Mrs. A. For instance, knowledge on the functional complexity of the brain cortical- Gonzalez-Bono et al. (2002) reported an average HR increase subcortical networks that participate in emotional and cognitive of > 10 beats/minute in participants expressing low cognitive control. The ACC, OFC and mPFC have been pinpointed as anxiety but > 20 beats/minute in high anxiety subjects when relevant areas for (1) the appraisal and expression (behavioral giving a public speech. In any case, this leaves open the possibility and autonomic) of emotions; (2) control of inhibitory processes; that the autonomic response of subjects like Mrs. A confronting and (3) the regulation of cognitive processes (Etkin and Wager, with stressful situations may differ from the response found 2007; Stevens et al., 2011). Moreover, the mPFC regulates the in healthy subjects and in persons with brain lesions that did HR response to social threats in a complex manner since not affect the cortical-subcortical network regulating autonomic different regions have a completely different role, sometimes response. Here, it is possible that the blunted blood pressure modulating opposite effects (Wager et al., 2009). Further, these Frontiers in Psychology | www.frontiersin.org 15 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia areas are strongly interconnected to the amygdala (affected assessed in different contexts. The results from this study provide in Mrs. A) and exert a regulatory effect over it, modulating further evidence on the detrimental effects of LA in language use, emotional responses (Kim et al., 2011; Motzkin et al., 2015). confirming poorest language performance when the evaluation Importantly, structural, functional and metabolic findings of Mrs. was performed by a therapist (stressful context) compared to A showed involvement of several areas of this cortico-subcortical testing performed by a well-trained relative. In addition, we network. The PET analysis revealed decreased metabolic activity identified some factors (e.g., premorbid personality traits) that in the left amygdala, left OFC and bilateral CC. Further, resting may increase LA, the underlying neural mechanisms and its state functional connectivity analyses suggested a functional detrimental effect on language performance. Further longitudinal decoupling between the left amygdala and both the mPFC and studies with larger sample of participants are needed to advance the OFC, whereas the right amygdala seemed to be synchronized in the characterization of LA and to help to design more optimal with regions from the PSA, the mPFC and the OFC in the protocols for PWA. right hemisphere. The fact that preserved sectors of the left PSA were functionally disconnected from the mPFC suggests that the release of automatic echolalia as a manifestation of LA resulted AUTHOR CONTRIBUTIONS from the disinhibition of the left PSA (Berthier et al., 2018). More support for this possibility comes from the PET finding All authors listed have made a substantial, direct and intellectual showing decreased metabolic activity in the left OFC. Therefore, contribution to the work, and approved it for publication. All despite that the OFC is functionally connected to the pMTG, this authors were involved in conception and design, acquisition of may not exert the required control over the activity of the left data, or analysis and interpretation of data. MJT-P, DL-B, and MB PSA, contributing to the increment of automatic echolalia under were directly implicated in performing language, and cognitive stressful conditions. Dysfunctional activity in the network formed testing, as well as training the patient’s husband in the application by the amygdala and prefrontal areas could also play a role at this of speech therapy and language assessment. MJT-P, DL-B, stage. Since mPFC areas are involved in both cognitive control JP-P, NR-V, and MB analyzed and/or interpreted neuroimaging and the regulation of emotional responses, demanding linguistic data. MJT-P, DL-B, and MD-M participated in the acquisition, tasks may require higher cognitive control, modifying the activity statistical analysis, and/or interpretation of autonomic data. All of frontal areas, and at the same time affecting the emotional authors drafted the article and revised it critically for important response to these tasks through cortico-subcortical connections. intellectual content. Although the current evidence suggests that this network was dysfunctional in Mrs. A, further studies are needed to gain more insight on the brain correlates of LA. FUNDING MJT-P and JP-P have been funded by Ph.D. scholarships from the Spanish Ministry of Education, Culture and Sport under CONCLUSION the FPU program (MJT-P: FPU14/04021; JP-P: FPU16/05108). DL-B has been supported by a postdoctoral grant from the Anxiety is a frequent aftermath of stroke. In PWA, anxiety University of Malaga. induced by language testing and everyday situations that require the use of language is challenging since it affects linguistic performance, jeopardizing the reliability of evaluations outcome. ACKNOWLEDGMENTS Intriguingly, there are very few studies aiming to explore the clinical features and neural correlates of LA. The present work We are deeply grateful to Mrs. A and her family (husband provides a detailed multimodal characterization of a person and son) for their continuous engagement and support with severe aphasia and LA under different interventions and during the study. REFERENCES Babulal, G. M., Huskey, T. N., Roe, C. M., Goette, S. A., and Connor, L. T. (2015). Cognitive impairments and mood disruptions negatively impact instrumental Aben, I., Denollet, J., Lousberg, R., Verhey, F., Wojciechowski, F., and Honig, A. activities of daily living performance in the first three months after a first stroke. (2002). Personality and vulnerability to depression in stroke patients: a 1- Top. Stroke Rehabil. 22, 144–151. doi: 10.1179/1074935714Z.0000000012 year prospective follow-up study. Stroke 33, 2391–2395. doi: 10.1161/01.STR. Bandura, A. (1982). Self-efficacy mechanism in human agency. Am. Psychol. 37, 0000029826.41672.2E 122–147. doi: 10.1037/0003-066X.37.2.122 Aida, Y. (1994). Examination of Horwitz, Horwitz, and Cope’s construct of foreign Bandura, A. (1988). Self-efficacy conception of anxiety. Anxiety Res. 1, 77–98. language anxiety: the case of students of Japanese. Mod. Lang. J. 78, 155–168. doi: 10.1080/10615808808248222 doi: 10.1111/j.1540-4781.1994.tb02026.x Beblo, T., Wallesch, C. W., and Herrmann, M. (1999). The crucial role of Albert, M. L., Goodglass, H., Helm, N. A., Rubens, A. B., and Alexander, M. P. frontostriatal circuits for depressive disorders in the postacute stage after stroke. (1981). Clinical Aspects of Dysphasia. Berlin: Springer. doi: 10.1007/978-3- Neuropsychiatry Neuropsychol. Behav. Neurol. 12, 236–246. 7091-8605-3 Benitez, A., Edens, H., Fishman, J., Moran, K., and Asgharnejad, M. (2014). American Psychiatric Association (2013). Diagnostic and Statistical Manual Rotigotine transdermal system: developing continuous dopaminergic delivery of Mental Disorders (DSM-5 R ). Washington, DC: American Psychiatric to treat Parkinson’s disease and restless legs syndrome. Ann. N. Y. Acad. Sci. Association. doi: 10.1176/appi.books.9780890425596 1329, 45–66. doi: 10.1111/nyas.12508 Frontiers in Psychology | www.frontiersin.org 16 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia Benton, A. L. (1994). Contributions to Neuropsychological Assessment: A Clinical Cooper, J. (1970). The Leyton obsessional inventory. Psychol. Med. 1, 48–64. doi: Manual. New York, NY: Oxford University Press. 10.1017/S0033291700040010 Berthier, M. L. (1999). “Transcortical aphasias,” in Brain Damage, Behaviour and Coy, B., O’Brien, W. H., Tabaczynski, T., Northern, J., and Carels, R. Cognition Series, eds C. Code and G. Humphreys (Hove: Psychology Press). (2011). Associations between evaluation anxiety, cognitive interference and Berthier, M. L. (2001). Unexpected brain-language relationships in aphasia: performance on working memory tasks. Appl. Cogn. Psychol. 25, 823–832. evidence from transcortical sensory aphasia associated with frontal lobe lesions. doi: 10.1002/acp.1765 Aphasiology 15, 99–130. doi: 10.1080/02687040042000179 Crisp, J., and Lambon Ralph, M. A. (2006). Unlocking the nature of the Berthier, M. L. (2005). Poststroke aphasia: epidemiology, pathophysiology and phonological–deep dyslexia continuum: the keys to reading aloud are in treatment. Drugs Aging 22, 163–182. doi: 10.2165/00002512-200522020-00006 phonology and semantics. J. Cogn. Neurosci. 18, 348–362. doi: 10.1162/ Berthier, M. L., Dávila, G., Green-Heredia, C., Moreno Torres, I., Juárez y Ruiz 089892906775990543 de Mier, R., De-Torres, I., et al. (2014a). Massed sentence repetition training Critchley, H. D., Mathias, C. J., Josephs, O., O’Doherty, J., Zanini, S., Dewar, B., can augment and speed up recovery of speech production deficits in patients et al. (2003). Human cingulate cortex and autonomic control: converging with chronic conduction aphasia receiving donepezil treatment. Aphasiology 28, neuroimaging and clinical evidence. Brain 126, 2139–2152. doi: 10.1093/brain/ 188–218. doi: 10.1080/02687038.2013.861057 awg216 Berthier, M. L., Green-Heredia, C., Juárez Ruiz, de Mier, R., Lara, J. P., and Difrancesco, S., Pulvermüller, F., and Mohr, B. (2012). Intensive language- Pulvermüller, F. (2014b). Rehabilitación Grupal Intensiva de la Afasia. Madrid: action therapy (ILAT): the methods. Aphasiology 26, 1317–1351. doi: 10.1080/ TEA Ediciones. 02687038.2012.705815 Berthier, M. L., Dávila, G., and Torres-Prioris, M. J. (2018). “Echophenomena Døli, H., Helland, T., and Andersen Helland, W. (2017). Self-reported symptoms in aphasia: causal mechanisms and clues for intervention,” in Aphasia of anxiety and depression in chronic stroke patients with and without aphasia. Rehabilitation: Clinical Challenges, eds P. Coppens and J. Patterson (Burlington, Aphasiology 31, 1392–1409. doi: 10.1080/02687038.2017.1280595 MA: Jones & Bartlett Learning), 143–171. Etkin, A., and Wager, T. D. (2007). Functional neuroimaging of anxiety: a Berthier, M. L., Green, C., Lara, J. P., Higueras, C., Barbancho, M. A., Dávila, G., meta-analysis of emotional processing in PTSD, social anxiety disorder, and et al. (2009). Memantine and constraint-induced aphasia therapy in chronic specific phobia. Am. J. Psychiatry 164, 1476–1488. doi: 10.1176/appi.ajp.2007. poststroke aphasia. Ann. Neurol. 65, 577–585. doi: 10.1002/ana.21597 07030504 Bogousslavsky, J., Regli, F., and Assal, G. (1988). Acute transcortical mixed aphasia. Fiddick, L. (2011). There is more than the amygdala: potential threat assessment A carotid occlusion syndrome with pial and watershed infarcts. Brain 111(Pt 3), in the cingulate cortex. Neurosci. Biobehav. Rev. 35, 1007–1018. doi: 10.1016/J. 631–641. doi: 10.1093/brain/111.3.631 NEUBIOREV.2010.09.014 Brett, M., Leff, A. P., Rorden, C., and Ashburner, J. (2001). Spatial normalization Fridriksson, J., den Ouden, D.-B., Hillis, A. E., Hickok, G., Rorden, C., Basilakos, A., of brain images with focal lesions using cost function masking. Neuroimage 14, et al. (2018). Anatomy of aphasia revisited. Brain 141, 848–862. doi: 10.1093/ 486–500. doi: 10.1006/NIMG.2001.0845 brain/awx363 Buchanan, T. W., Driscoll, D., Mowrer, S. M., Sollers, J. J., Thayer, J. F., Fure, B., Wyller, T. B., Engedal, K., and Thommessen, B. (2006). Emotional Kirschbaum, C., et al. (2010). Medial prefrontal cortex damage affects symptoms in acute ischemic stroke. Int. J. Geriatr. Psychiatry 21, 382–387. physiological and psychological stress responses differently in men and women. doi: 10.1002/gps.1482 Psychoneuroendocrinology 35, 56–66. doi: 10.1016/j.psyneuen.2009.09.006 Gill, S. K., and Leff, A. P. (2014). Dopaminergic therapy in aphasia. Aphasiology 28, Buchanan, T. W., Laures-Gore, J. S., and Duff, M. C. (2014). Acute stress reduces 155–170. doi: 10.1080/02687038.2013.802286 speech fluency. Biol. Psychol. 97, 60–66. doi: 10.1016/J.BIOPSYCHO.2014. Goldstein, K. (1948). Language and Language Disturbances; Aphasic Symptom 02.005 Complexes and their Significance for Medicine and theory of Language. Oxford: Buchanan, T. W., Tranel, D., and Kirschbaum, C. (2009). Hippocampal damage Grune & Stratton. abolishes the cortisol response to psychosocial stress in humans. Horm. Behav. Gonzalez-Bono, E., Moya-Albiol, L., Salvador, A., Carrillo, E., Ricarte, J., and 56, 44–50. doi: 10.1016/J.YHBEH.2009.02.011 Gomez-Amor, J. (2002). Anticipatory autonomic response to a public speaking Burton, C. A. C., Murray, J., Holmes, J., Astin, F., Greenwood, D., and Knapp, P. task in women: the role of trait anxiety. Biol. Psychol. 60, 37–49. doi: 10.1016/ (2013). Frequency of anxiety after stroke: a systematic review and meta-analysis S0301-0511(02)00008-X of observational studies. Int. J. Stroke 8, 545–559. doi: 10.1111/j.1747-4949. Gorgoraptis, N., Mah, Y.-H., Machner, B., Singh-Curry, V., Malhotra, P., Hadji- 2012.00906.x Michael, M., et al. (2012). The effects of the dopamine agonist rotigotine on Burvill, P. W., Johnson, G. A., Jamrozik, K. D., Anderson, C. S., Stewart-Wynne, hemispatial neglect following stroke. Brain 135, 2478–2491. doi: 10.1093/brain/ E. G., and Chakera, T. M. (1995). Anxiety disorders after stroke: results from aws154 the Perth Community Stroke Study. Br. J. Psychiatry 166, 328–332. doi: 10.1192/ Greenop, K. R., Almeida, O. P., Hankey, G. J., van Bockxmeer, F., and bjp.166.3.328 Lautenschlager, N. T. (2009). Premorbid personality traits are associated with Cahana-Amitay, D., Albert, M. L., Pyun, S.-B., Westwood, A., Jenkins, T., post-stroke behavioral and psychological symptoms: a three-month follow-up Wolford, S., et al. (2011). Language as a stressor in aphasia. Aphasiology 25, study in Perth, Western Australia. Int. Psychogeriatrics 21, 1063–1071. doi: 593–614. doi: 10.1080/02687038.2010.541469 10.1017/S1041610209990457 Cahana-Amitay, D., Oveis, A. C., Sayers, J. T., Pineles, S. L., Spiro, A., and Albert, Grossi, D., Trojano, L., Chiacchio, L., Soricelli, A., Mansi, L., Postiglione, A., M. L. (2015). Biomarkers of “Linguistic Anxiety” in aphasia: a proof-of-concept et al. (1991). Mixed transcortical aphasia: clinical features and neuroanatomical case study. Clin. Linguist. Phon. 29, 401–413. doi: 10.3109/02699206.2015. correlates. Eur. Neurol. 31, 204–211. doi: 10.1159/000116679 1014572 Hackett, M. L., Köhler, S., O’Brien, J. T., and Mead, G. E. (2014). Neuropsychiatric Carrera, E., and Tononi, G. (2014). Diaschisis: past, present, future. Brain 137, outcomes of stroke. Lancet Neurol. 13, 525–534. doi: 10.1016/S1474-4422(14) 2408–2422. doi: 10.1093/brain/awu101 70016-X Castillo, C. S., Starkstein, S. E., Fedoroff, J. P., Price, T. R., and Robinson, R. G. Hahn, A., Stein, P., Windischberger, C., Weissenbacher, A., Spindelegger, C., (1993). Generalized anxiety disorder after stroke. J. Nerv. Ment. Dis. 181, Moser, E., et al. (2011). Reduced resting-state functional connectivity between 100–106. doi: 10.1097/00005053-199302000-00005 amygdala and orbitofrontal cortex in social anxiety disorder. Neuroimage 56, Chemerinski, E., and Robinson, R. G. (2000). The neuropsychiatry 881–889. doi: 10.1016/J.NEUROIMAGE.2011.02.064 of stroke. Psychosomatics 41, 5–14. doi: 10.1016/S0033-3182(00) Hillis, A. E. (2007). Aphasia: progress in the last quarter of a century. Neurology 69, 71168-6 200–213. doi: 10.1212/01.wnl.0000265600.69385.6f Chen, M. C., and Lin, H.-J. (2009). Self-efficacy, foreign language anxiety as House, A. (1987). Mood disorders after stroke: a review of the evidence. Int. J. predictors of academic performance among professional program students in Geriatr. Psychiatry 2, 211–221. doi: 10.1002/gps.930020403 a general English proficiency writing test. Percept. Mot. Skills 109, 420–430. Kay, J., Lesser, R., and Coltheart, M. (1992). Psycholinguistic Assessments of doi: 10.2466/pms.109.2.420-430 Language Processing in Aphasia. (PALPA). Hove: Lawrence Erlbaum Associated. Cheng, Y., Horwitz, E. K., and Schallert, D. L. (1999). Language anxiety: Kertesz, A. (1982). The Western Aphasia Battery. New York, NY: Grune and differentiating writing and speaking components. Lang. Learn. 49, 417–446. Stratton. doi: 10.1111/0023-8333.00095 Kertesz, A. (2007). Western Aphasia Battery-Revised. San Antonio, TX: PsychCorp. Frontiers in Psychology | www.frontiersin.org 17 May 2019 | Volume 10 | Article 678 Torres-Prioris et al. Linguistic Anxiety in Severe Aphasia Kertesz, A., Pascual-Leone, A., and Pascual-Leone García, A. (1990). Bateriìa de Robinson, R. G., and Starkstein, S. E. (1990). Current research in affective disorders Afasiìas “Western”: (the Western Aphasia Battery en Versioìn y Adaptacioìn following stroke. J. Neuropsychiatry Clin. Neurosci. 2, 1–14. doi: 10.1176/ Castellana). Manual del test. Valencia: NAU. jnp.2.1.1 Kim, M. J., Loucks, R. A., Palmer, A. L., Brown, A. C., Solomon, K. M., Marchante, Rorden, C., and Brett, M. (2000). Stereotaxic display of brain lesions. Behav. Neurol. A. N., et al. (2011). The structural and functional connectivity of the amygdala: 12, 191–200. doi: 10.1155/2000/421719 from normal emotion to pathological anxiety. Behav. Brain Res. 223, 403–410. Saito, Y., Garza, T. J., and Horwitz, E. K. (1999). Foreign language reading anxiety. doi: 10.1016/J.BBR.2011.04.025 Mod. Lang. J. 83, 202–218. doi: 10.1111/0026-7902.00016 Kreibig, S. D. (2010). Autonomic nervous system activity in emotion: a review. Biol. Sakamoto, H., Sakamaki, T., Tani, T., Sugai, Y., Nakamura, T., Ono, Z., et al. (1999). Psychol. 84, 394–421. doi: 10.1016/J.BIOPSYCHO.2010.03.010 “White coat effect” induced by therapist’s presence during speech therapy for Kreisler, A., Godefroy, O., Delmaire, C., Debachy, B., Leclercq, M., Pruvo, J. P., stroke rehabilitation: a single case study. Arch. Phys. Med. Rehabil. 80, 470–472. et al. (2000). The anatomy of aphasia revisited. Neurology 54, 1117–1123. doi: doi: 10.1016/S0003-9993(99)90289-7 10.1212/WNL.54.5.1117 Sembi, S., Tarrier, N., O’Neill, P., Burns, A., and Faragher, B. (1998). Does post- Laures-Gore, J., Heim, C. M., and Hsu, Y.-S. (2007). Assessing cortisol reactivity traumatic stress disorder occur after stroke: a preliminary study. Int. J. Geriatr. to a linguistic task as a marker of stress in individuals with left-hemisphere Psychiatry 13, 315–322. doi: 10.1002/(SICI)1099-1166(199805)13:5<315::AID- stroke and aphasia. J. Speech Lang. Hear. Res. 50, 493–507. doi: 10.1044/1092- GPS766>3.0.CO;2-P 4388(2007/034) Shehata, G. A., El Mistikawi, T., Risha, A. S. K., and Hassan, H. S. (2015). The Laures-Gore, J. S., and Buchanan, T. W. (2015). Aphasia and the effect of aphasia upon personality traits, depression and anxiety among stroke neuropsychobiology of stress. J. Clin. Exp. Neuropsychol. 37, 688–700. patients. J. Affect. Disord. 172, 312–314. doi: 10.1016/j.jad.2014.10.027 doi: 10.1080/13803395.2015.1042839 Shin, L. M., and Liberzon, I. (2010). The neurocircuitry of fear, stress and anxiety Laures-Gore, J. S., DuBay, M. F., Duff, M. C., and Buchanan, T. W. (2010). disorders. Neuropsychopharmacology 35, 169–191. doi: 10.1038/npp.2009.83 Identifying behavioral measures of stress in individuals with aphasia. Starkstein, S. E., Fedoroff, J. P., Price, T. R., Leiguarda, R., and Robinson, J. Speech Lang. Hear. Res. 53, 1394–1400. doi: 10.1044/1092-4388(2010/09- R. G. (1993). Catastrophic reaction after cerebrovascular lesions: frequency, 0236) correlates, and validation of a scale. J. Neuropsychiatry Clin. Neurosci. 5, 189– Leppävuori, A., Pohjasvaara, T., Vataja, R., Kaste, M., and Erkinjuntti, T. (2003). 194. doi: 10.1176/jnp.5.2.189 Generalized anxiety disorders three to four months after ischemic stroke. Stevens, F. L., Hurley, R. A., and Taber, K. H. (2011). Anterior cingulate cortex: Cerebrovasc. Dis. 16, 257–264. doi: 10.1159/000071125 unique role in cognition and emotion. J. Neuropsychiatry Clin. Neurosci. 23, Makovac, E., Meeten, F., Watson, D. R., Herman, A., Garfinkel, S. N., D 121–125. doi: 10.1176/jnp.23.2.jnp121 Critchley, H., et al. (2016). Alterations in amygdala-prefrontal functional Thomas, S. A., and Lincoln, N. B. (2008). Predictors of emotional distress after connectivity account for excessive worry and autonomic dysregulation in stroke. Stroke 39, 1240–1245. doi: 10.1161/STROKEAHA.107.498279 generalized anxiety disorder. Biol. Psychiatry 80, 786–795. doi: 10.1016/j. Ulrich-Lai, Y. M., and Herman, J. P. (2009). Neural regulation of endocrine and biopsych.2015.10.013 autonomic stress responses. Nat. Rev. Neurosci. 10, 397–409. doi: 10.1038/ Maldjian, J. A., Laurienti, P. J., Kraft, R. A., and Burdette, J. H. (2003). nrn2647 An automated method for neuroanatomic and cytoarchitectonic atlas-based Valle, F., and Cuetos, F. (1995). EPLA: Evaluación del Procesamiento Lingüístico en interrogation of fMRI data sets. Neuroimage 19, 1233–1239. doi: 10.1016/ la Afasia. Hove: Lawrence Erlbaum Associates. S1053-8119(03)00169-1 Vuger-Kovačić, D., Gregurek, R., Kovačić, D., Vuger, T., and Kalenić, B. Measso, G., Zappalà, G., Cavarzeran, F., Crook, T. H., Romani, L., Pirozzolo, (2007). Relation between anxiety, depression and locus of control of F. J., et al. (2009). Raven’s colored progressive matrices: a normative study patients with multiple sclerosis. Mult. Scler. J. 13, 1065–1067. doi: 10.1177/ of a random sample of healthy adults. Acta Neurol. Scand. 88, 70–74. doi: 1352458507077629 10.1111/j.1600-0404.1993.tb04190.x Wager, T. D., Waugh, C. E., Lindquist, M., Noll, D. C., Fredrickson, B. L., and Meinzer, M., Streiftau, S., and Rockstroh, B. (2007). Intensive language training Taylor, S. F. (2009). Brain mediators of cardiovascular responses to social threat: in the rehabilitation of chronic aphasia: efficient training by laypersons. J. Int. part I: reciprocal dorsal and ventral sub-regions of the medial prefrontal cortex Neuropsychol. Soc. 13, 846–853. doi: 10.1017/S1355617707071111 and heart-rate reactivity. Neuroimage 47, 821–835. doi: 10.1016/j.neuroimage. Morris, P. L. P., Robinson, R. G., Raphael, B., and Bishop, D. (1991). 2009.05.043 The relationship between the perception of social support and post-stroke Whitfield-Gabrieli, S., and Nieto-Castanon, A. (2012). Conn: a functional depression in hospitalized patients. Psychiatry 54, 306–316. doi: 10.1080/ connectivity toolbox for correlated and anticorrelated brain networks. Brain 00332747.1991.11024559 Connect. 2, 125–141. doi: 10.1089/brain.2012.0073 Morris, R., Eccles, A., Ryan, B., and Kneebone, I. I. (2017). Prevalence of anxiety Wittich, C. M., Burkle, C. M., and Lanier, W. L. (2012). Ten common questions in people with aphasia after stroke. Aphasiology 31, 1410–1415. doi: 10.1080/ (and their answers) about off-label drug use. Mayo Clin. Proc. 87, 982–990. 02687038.2017.1304633 doi: 10.1016/j.mayocp.2012.04.017 Motzkin, J. C., Philippi, C. L., Wolf, R. C., Baskaya, M. K., and Koenigs, M. (2015). Woodrow, L. (2006). Anxiety and speaking English as a second language. RELC J. Ventromedial prefrontal cortex is critical for the regulation of amygdala activity 37, 308–328. doi: 10.1177/0033688206071315 in humans. Biol. Psychiatry 77, 276–284. doi: 10.1016/J.BIOPSYCH.2014. Zarola, F. (2018). Vascular Parkinsonism sensitive to Rotigotine therapy is found 02.014 in aged patients: a clinical case description. Acta Biomed 89, 99–100. doi: 10. Parr, S. (2007). Living with severe aphasia: tracking social exclusion. Aphasiology 23750/abm.v89i1.6590 21, 98–123. doi: 10.1080/02687030600798337 Zhang, X. (2013). Foreign language listening anxiety and listening performance: Pena-Garijo, J., Edo Villamón, S., Meliá de Alba, A., and Ruipérez, M. Á. (2013). conceptualizations and causal relationships. System 41, 164–177. doi: 10.1016/J. Personality disorders in obsessive-compulsive disorder: a comparative study SYSTEM.2013.01.004 versus other anxiety disorders. Sci. World J. 2013, 1–7. doi: 10.1155/2013/ 856846 Conflict of Interest Statement: The authors declare that the research was Phillips, E. M. (1992). The effects of language anxiety on students’ oral test conducted in the absence of any commercial or financial relationships that could performance and attitudes. Mod. Lang. J. 76, 14–26. doi: 10.1111/j.1540-4781. be construed as a potential conflict of interest. 1992.tb02573.x Pickering, T. G., Gerin, W., and Schwartz, A. R. (2002). What is the white-coat Copyright © 2019 Torres-Prioris, López-Barroso, Paredes-Pacheco, Roé-Vellvé, effect and how should it be measured? Blood Press. Monit. 7, 293–300. doi: Dawid-Milner and Berthier. This is an open-access article distributed under the 10.1097/01.mbp.0000053044.16575.8b terms of the Creative Commons Attribution License (CC BY). The use, distribution Pulvermüller, F., and Berthier, M. L. (2008). Aphasia therapy on a neuroscience or reproduction in other forums is permitted, provided the original author(s) and basis. Aphasiology 22, 563–599. doi: 10.1080/02687030701612213 the copyright owner(s) are credited and that the original publication in this journal Robinson, R. G., and Jorge, R. E. (2016). Post-stroke depression: a review. Am. J. is cited, in accordance with accepted academic practice. No use, distribution or Psychiatry 173, 221–231. doi: 10.1176/appi.ajp.2015.15030363 reproduction is permitted which does not comply with these terms. Frontiers in Psychology | www.frontiersin.org 18 May 2019 | Volume 10 | Article 678