Does Executive Function Training Impact on Communication? A Randomized Controlled tDCS Study on Post-Stroke Aphasia

New approaches in aphasia rehabilitation have recently identified the crucial role of executive functions (EFs) in language recovery, especially for people with severe aphasia (PWSA). Indeed, EFs include high-order cognitive abilities such as planning and problem solving, which enable humans to adapt to novel situations and are essential for everyday functional communication. In a randomized double-blind crossover design, twenty chronic Italian PWSA underwent ten days of transcranial direct current stimulation (tDCS) (20 min, 2 mA) over the right dorsolateral prefrontal cortex (DLPFC). Two conditions were considered, i.e., anodal and sham, while performing four types of cognitive training (alertness, selective attention, visuo-spatial working memory, and planning), all of which were related to executive functions. After anodal tDCS, a greater improvement in selective attention, visuospatial working memory and planning abilities was found compared to the sham condition; this improvement persisted one month after the intervention. Importantly, a significant improvement was also observed in functional communication, as measured through the Communication Activities of Daily Living Scale, in noun and verb naming, in auditory and written language comprehension tasks and in executive function abilities. This evidence emphasizes, for the first time, that tDCS over the right DLPFC combined with executive training enhances functional communication in severe aphasia.


Introduction
The traditional model of language organization, often referred to as the Broca-Wernicke-Lichtheim-Geschwind model [1,2], focuses almost exclusively on the involvement of the inferior frontal and posterior temporal regions for expressive and receptive language functions, respectively, and the connection between these sites, named the arcuate fasciculus. However, in recent years, behavioral and neuroimaging results have shown that the network subserving the language function is much more widely distributed across the brain than previously thought [3][4][5]. Indeed, most contemporary models propose a much more complex architecture encompassing regions which might also include bilateral cortical networks as well as subcortical circuits [6][7][8][9][10]. In line with this view, a growing body of evidence has led to the concept of "neuronal multifunctionality", in which these complex neuronal circuits subserve both linguistic and non-linguistic information, creating dynamic cognition-language interactions in the brain [11,12]. Based on this perspective, new approaches in aphasia rehabilitation have emphasized that non-linguistic functions may also subserve language recovery. Accordingly, several works have already shown that multiple cognitive domains, including attention [13][14][15][16][17], memory [18,19] and executive functions, may improve communication and language performance in aphasia [20][21][22][23]. Since in persons with severe aphasia (PWSA), language does not always adequately meet the communicative needs of the individual [24], to communicate successfully, it is necessary awareness measured through the Error Awareness Task [58]. Results showed that anodal tDCS over the right DLPFC exerted the greater effect increasing error awareness compared to the other conditions. In a sample of fifty-five healthy subjects, Heinze and colleagues [59] investigated the effects of bilateral tDCS over the DLPFC combined with an eye-tracking while performing a planning task by using the Tower of London test. Results showed a reduction in initial thinking time following left cathodal/right anodal DLPFC stimulation in parallel with a shorter duration of the last gaze before task's solution. Based on these findings, the authors concluded that anodal stimulation over the right DLPFC is associated with a reduction in the time spent in evaluation processes during planning tasks.
Although the DLPFC has never been considered as specifically related to language tasks, its role for implementing functional connectivity between the language network and other cognitive domains has been widely recognized (for a review, see [45]). Thus, several studies have stimulated the left DLPFC to improve error detection in sentences [60] or to reduce interference in picture naming tasks [61]. Very recently, Pestalozzi and collaborators [62] have investigated whether strengthening executive control through anodal tDCS over the left DLPFC would facilitate lexical retrieval in a group of fourteen PWA. Results showed an increase in verbal fluency and in naming speed of high frequency words after anodal condition with respect to sham.
As far as we know, to date, none of the reported studies have investigated whether tDCS over the DLPFC combined with executive function training improves functional communication in PWSA.
As previously stated, several lines of evidence have already highlighted the role of executive functions in maintaining an adequate level of communication skills in aphasic patients, particularly, in severe cases [20,25,26,34,35].
Thus, in the present study, we investigated whether anodal tDCS over the right DLPFC combined with different executive function training would enhance the ability to communicate in everyday life in twenty PWSA. The choice to stimulate the right DLPFC is pertinent with all the most recent literature indicating this region as crucial for the performance of executive tasks.

Study Design
A randomized double-blinded cross-over design was conducted from January 2020 to June 2022 at the Behavioral Neurological Laboratory of the IRCCS Santa Lucia in Rome, Italy. Thirty chronic persons with post-stroke aphasia were examined through a detailed neuropsychological assessment. Ten were excluded for the following reasons: failure to meet the inclusion criteria, difficulty in transporting means and personal reasons. Thus, a final sample size of twenty patients was recruited (see Figure S1). G*Power 3.1 [63] was used to calculate the sample size with α = 0.05, a power = 90%, two measurements (anodal vs sham), and effect size f = 0.4. The analysis indicated that a total sample size of N ≥ 19 was necessary to detect a significant effect in our study.
All twenty patients received both interventions (AB → anodal-sham and BA → shamanodal). The order of conditions was randomized across subjects. Half of the participants (n = 10) started with condition A (anodal tDCS) followed by condition B (sham tDCS), while the other half began (n = 10) with condition B (sham tDCS) followed by condition A (anodal tDCS). The allocation sequences were generated by a technician of the laboratory. To avoid carryover effects, a washout period of four weeks was established between condition A and B (and vice versa). As this was a double-blinded study, both the examiner and the patient were blinded regarding the stimulation condition and the stimulator was turned on/off by a third person, who assigned participants to the AB or BA intervention.

Participants
Twenty left-brain-damaged participants (ten men and ten women, mean age: 61.04; SD 7.02) with severe chronic aphasia were included in the study. Inclusion criteria were Brain Sci. 2022, 12, 1265 4 of 17 native Italian proficiency, a single left ischemic stroke at least 6 months prior to the investigation, pre-morbid right handedness (based on the "Edinburgh Handedness Questionnaire"; [64]) and no acute or chronic neurological symptoms needing medication. Subjects over 75 years of age and those with seizures, implanted electronic devices (e.g., pacemaker) and previous brain damage were excluded. None of the participants were taking any kind of medication and none of them has received structured language therapy for at least 6 months before the time of inclusion in the study in order to prevent confounding therapy effects (see Table 1).  I  FT  0  0  0  0  0  0  0  0  0  0  5  7  F  58  13  2 y  H  FTI  5  5  0  0  20  10  10  0  0  0  5  8  F  72  8  1 y  I  FTP  0  0

Ethics Statement
The data analyzed in the current study were collected in accordance with the Declaration of Helsinki and the Institutional review board of the IRCCS Fondazione Santa Lucia, Rome, Italy. Before participation, all patients signed informed consent forms.
In the alertness training, the patient drove a motorcycle at varying speed along a winding road. The aim was to carefully observe the stretch of the road in front of him/her and to press the keyboard as quickly as possible when an obstacle appeared on the road (i.e., a level crossing closes, a tree falls unexpectedly in the driver's path), in order to brake promptly before it.
In the selective attention task, the patient was an explorer on a boat along a river surrounded by a forest. During the journey, several animals appeared, including hippos, giraffes, elephants. The patient was asked to press the keyboard only when he/she saw the hippos.
In the spatial working memory task, the patient watched colorful butterflies in a natural environment. The butterflies flew over a meadow or sandy area. From time to time, one butterfly landed, and another started its flight and so on. Depending on the level of difficulty, the patient was asked to remember the position of the last butterfly, the second-to-last butterfly, the third-to-last butterfly and so on.
In the planning training, the patient saw a map of the city with nine buildings (e.g., post office, café, insurance office, cultural center). On the right side of the map, a box appeared in which pending and completed errands were listed. The patient was asked to accomplish several tasks in each building, formulating an appropriate strategy to decide in which order running the errands. The task difficulty varied according to the number of errands to be completed and the time spent.
2.6. Procedure 2.6.1. Transcranial Direct Current Stimulation (tDCS) tDCS was applied using a battery driven Eldith (neuroConn GmbH) Programmable Direct Current Stimulator with a pair of surface-soaked sponge electrodes (5 × 7 cm). Anodal stimulation consisted of 20 min of 2 mA direct current with the anode placed over the right DLPFC (F4 of the extended International 10-20 system for EEG electrode placement) and the cathode (the reference electrode) above the contralateral frontopolar cortex (Fp1). For sham stimulation, the same electrode position was used. The current was ramped up to 2 mA and slowly decreased over 30 s to ensure the typical initial tingling sensation [73]. The order of conditions was randomized across subjects. Half of the participants started with the anodal condition and the remaining half with the sham condition. There were four weeks of intersession interval between the two experimental conditions. Thus, after four weeks, the order of condition was inverted. For each experimental condition (anodal vs. sham), the rehabilitative program consisted of 10 one-hour sessions over two weeks (Monday-Friday, weekends off, Monday-Friday). Although tDCS stimulation was delivered from the beginning of the cognitive training up to 20 min, the cognitive training lasted 1 h per day. At the end of each treatment condition (anodal vs. sham) and after four weeks (follow-up), the neuropsychological battery was readministered to all patients. During the training, none of the participants noticed differences in the intensity of sensation between the two stimulation conditions (anodal vs. sham), not being aware of what condition they were performing [74].

Cognitive Treatments
The cognitive treatment was administered through the Cogniplus software (Schuhfried). During each one-hour session, all participants underwent four types of training: alertness, selective attention, visuo-spatial memory, and planning presented in randomized order.

Data Analysis
Before, after the treatment and at follow-up (FU), the patients' performance was evaluated by comparing the mean score obtained in the alertness, selective attention, visuospatial working memory, and planning training. Data were analyzed using STATISTICA10 software. The Shapiro-Wilk test was applied which revealed a normal distribution of the data. Four repeated measures ANOVAs were performed separately for the four types of training. For each analysis, two "within" factors were considered: CONDITION (anodal vs. sham) and TIME (baseline (T0) versus end of treatment (T10) versus follow up (FU)). The post-hoc Bonferroni test was conducted on the significant effects observed in the ANOVA. The values of p ≤ 0.05 were considered statistically significant. Before and after each treatment condition, the patients' responses to the different re-administration of the standardized language tests (EDL and BADA), CADL-2 test, visual search test, Smirni subtest and TOL test were also analyzed using χ2-test.

Discussion
In the present study, we investigated whether different types of executive function training combined with tDCS would enhance functional communication in twenty persons with chronic severe aphasia. After the training, an improvement in selective attention, spatial working memory and planning abilities was found in both stimulation conditions (anodal vs. sham) but it was greater in the anodal condition compared to sham. More importantly, this improvement persisted one month after the intervention. Thus, the executive training alone exerted its own effectiveness, but the recovery process was further improved after anodal tDCS. No differences were found between the two conditions in the alertness task. This last result argues against an explanation simply due to enhanced cognitive arousal which should have also influenced the alertness task. Interestingly, after anodal stimulation, a significant improvement was also observed in functional communication, in noun and verb naming, in auditory and written language comprehension tasks and in different executive functions tests.
As stated in the Introduction, in persons with chronic severe aphasia, language skills might result dramatically impaired even years after the onset of the disease. This situation often impacts also on the person's ability to rely on functional communication. That is, the ability to effectively communicate his/her own needs in social contexts making use of compensatory strategies which allow to bypass the person's verbal limitations [75][76][77]. Indeed, all individuals can express their communicative intentions not only through language, but also through extralinguistic means such as hand gestures, body movements or facial expressions which are intentionally expressed to convey a message. The impact of inadequate strategic competence prevents severe aphasic people to maintain successful social relations and to pursue life goals [78]. Consequently, in addition to the assessment of formal aspects of language (phonological, lexical, and grammatical domains), in persons with severe aphasia, the adjunct of a functional communication scale is particularly relevant to test their communicative abilities.
In the present study, together with standardized language tests, we administered the Communication Activities of Daily Living scale (CADL 2) [68] which, among the different communication abilities assessment tools, is considered a valid ecological battery for functional communication assessment [79]. The CADL 2 assesses a person's communication abilities in activities of daily life by asking him/her to simulate communication acts in hypothetical natural environments (e.g., going to the doctor, making a phone call, asking for directions). It has been widely employed in the assessment of everyday language abilities in persons with aphasia [80,81] and in the evaluation of intensive language rehabilitation programs [82]. Interestingly, our results clearly showed that, while, before the training, all patients obtained very low percentage scores in the CADL-2, after two weeks of treatment, all of them reached an average percentage score but only when the training was combined with anodal tDCS. Thus, although the training was effective, the greater improvement obtained in the anodic condition was reflected in a positive change in the CADL-2 only in this condition. Similar results, although not for all patients, were obtained in the tests of attention, spatial memory, and planning skills.
As already mentioned, in recent years, there has been increasing interest in understanding the role of executive processes (e.g., cognitive control, attention, working memory) in the recovery of post-stroke language deficits [62,83,84]. The cerebral regions involved in executive function processing have been shown to be recruited in language tasks in post-stroke aphasia and in healthy subjects [85]. Researchers have also found that executive regions play a major role in recovery from aphasia [86]. Indeed, to enact goal-directed behavior and to respond to novel and challenging tasks of everyday life, people should make use of different cognitive components, such as cognitive flexibility, working memory and attention which rely on executive functions processing. Executive functions have been linked to pragmatic abilities and to social behavior [87] as they are involved in planning, monitoring, and inhibiting the discourse and in social exchanges. Moreover, intact executive functions system seems to be crucial to adaptive, motivated, and effective communication [88,89]. Impaired executive functions processing has been described in aphasic stroke patients and it has been shown to have a negative impact on rehabilitation outcomes [15,16,20,[90][91][92], functional communication [25], and quality of life [93]. However, in most of these studies, the relationship between executive control and functional communication has been investigated only by assessing the aphasic performance through neuropsychological tests, including measures of functional communication, executive functional ability, and language impairment [25,26,94,95]. Based on these findings, it appears that higher levels of executive functioning are linked to better functional communication [21] and conversational skills [96] and greater cognitive flexibility has been significantly correlated to better strategy use in functional communication tasks [97]. Indeed, in the Olsson and colleagues [25,26] studies, most of the aphasic patients (79%) presented with executive functions deficits with the nonverbal participants more severely affected than the verbal group. Few recent studies have directly investigated if spared executive control is an important predictor of treatment gains. In the Simic et al. study [35], ten patients with mild to severe aphasia were treated three times a week for five weeks with a phonological naming therapy. Difference scores in naming accuracy of treated and untreated words served as the primary outcome measures. Results showed that individuals with better executive functions abilities showed better maintenance of treated words at four and eight-weeks follow-ups.
As far as we know, to date, only one study has emphasized the importance of strengthening executive control through prefrontal tDCS to facilitate lexical retrieval and verbal fluency in aphasia [62]. However, until now, no studies have investigated the impact of executive function training combined with tDCS on functional communication.
In our study, after the training, all patients improved in their functional communication skills and in several language and executive functions tasks, indicating an improvement in executive control enhanced functional communication. Interestingly, as also noted in previous works (see [37] for a review), these generalization effects on untreated tasks were present only after anodal stimulation. Indeed, the sham condition impacted only on the cognitive treatments as such. These results might be ascribed to the severity and chronicity of the functional communication profile observed, at baseline, in all patients. Indeed, since the treatment lasted only ten days, the hypothesis might be advanced that the executive function training alone was insufficient to improve functional communication. On the contrary, interestingly, the same number of training sessions combined with anodal tDCS over the same time period exerted beneficial effects on the treated functions and generalized to functional communication and to the language domain. Thus, similar to previous results (see [37] for a review), combining tDCS with cognitive training boosted cognitive recovery overcoming the difficulties caused by the severity of the deficit. These findings are, thus, very promising, as ten days of tDCS produced generalization effects over untreated functions (i.e., functional communication and language) which lasted up to one month which were not obtained in the absence of stimulation.
It is well known that executive functions are dependent on the prefrontal lobes which are strongly interconnected with other cortical areas and subcortical structures ( [98,99] among which the frontal areas. In our study, we choose to stimulate the right DLPFC due to its role in selective attention, spatial working memory and planning abilities. Indeed, most of the recent tDCS literature have found that anodal tDCS over the right DLPFC enhances these skills [53][54][55][56][57]. Moreover, since all of our patients had extensive damage to the left language areas, we reasoned that stimulation of the right DLPFC would also have positive effects on language tasks. Indeed, even if the exact underlying tDCS mechanisms in our study remain largely speculative, the hypothesis might be advanced that, due to the strong interconnections between the prefrontal and the frontal regions [85,86,98,99], anodal tDCS over the right DLPFC has enhanced activation on the right frontal cortex which, in turn, serves as a supportive area for the observed recovery. Indeed, to date, several neuroimaging studies have already shown functional connectivity changes on cortical activity within the lesioned left hemisphere [100][101][102] and in the contralateral right homologues [103][104][105] due to tDCS treatment.

Conclusions
We are aware that our study has some limitations. The major ones are the small sample size and the lack of longitudinal follow-ups and of neuroimaging recordings. However, considering these limitations, we believe that our study highlights several important aspects to be considered when making treatment decisions for people with severe aphasia. First of all, it points to the possibility of training cognitive functions other than language. Indeed, from a connectionist perspective which considers the language system as part of a network largely distributed across the brain, this allows clinicians to plan different cognitive treatments which, in turn, facilitate aphasia recovery. It also emphasizes the need to assess functional communication skills, the recovery of which, even in the most severely affected patients, will allow the patient to socially interact in everyday life contexts. Finally, it confirms several previous reports which suggest that post-stroke aphasics in the chronic phase can still benefit from combining the treatment with tDCS.
In conclusion, although future studies are needed to deepen our understanding of the role of executive control on functional communication and the underlying neural mechanisms by which tDCS affects verbal performance, we believe that our results are promising since, for the first time, they suggest that executive function training can positively impact functional communication in severe chronic aphasia.
Author Contributions: Conceptualization, F.P. and P.M.; methodology, F.P., A.M. and A.C.; formal analysis, F.P.; data curation, F.P., A.M. and A.C.; writing-original draft preparation, F.P.; writingreview and editing, P.M. and C.C.; supervision, P.M. All authors have read and agreed to the published version of the manuscript. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to ethical and privacy restrictions.