Targeting interhemispheric balance to modulate language processing: a transcranial direct current stimulation study in healthy volunteers

Размещено на http://www.allbest.ru/

Размещено на http://www.allbest.ru/

Targeting interhemispheric balance to modulate language processing: a tdcs study in healthy volunteers

Table of contents

• 1. Study the neural bases of language functions in the brain
• 1.1 Representation of language in the brain
• 1.2 The interhemispheric competition hypothesis
• 1.3 Transcranial direct current stimulation
• 2. Methods
• 2.1 Participants
• 2.2 Transcranial direct current stimulation
• 2.3 Procedure and tasks
• 2.4 Stimuli
• 3. Data analysis
• 4. Results
• 4.1 Safety and tolerability
• 4.2 Lexical decision
• 4.3 Sentence comprehension
• 4.4 Correlation of stimulation effect with handedness score
• 5. Discussion
• Literature

1. Study the neural bases of language functions in the brain

1.1 Representation of language in the brain

One of the main purposes of modern neurolinguistics is to study the neural bases of language functions in the brain and their individual variability. Since the second half of the 19^th century, there have emerged first studies concerning the functional organization of the brain and localization of language functions in the left hemisphere.

Many years before advanced neuroimaging methods were developed, scientists based their research on studying patients with aphasia. Patients that had been studied by neurologists had lesions caused by brain injury - thus, it was possible to determine post-mortem what brain regions were damaged and make conclusions about their functions, based on what deficits had been observed in a patient. The French neurologist Marc Dax observed that all patients with aphasia that he studied had lesion in the left hemisphere (the article had not been published) (Benton, 1984). Later, in 1861, a famous scientist Paul Broca studied several patients with brain damage and tested their performance on language tasks. After their death, Broca examined their brains and concluded that they all had damage in the same area - the frontal lobe of the left hemisphere, namely, left inferior frontal gyrus (later called Broca's area), and that the left hemisphere dominates in the language function (Traxler, 2012). In 1874, Karl Wernicke repeated Broca's observations and found another brain area in the left hemisphere that is crucial for speech comprehension - posterior section of the superior temporal gyrus (now known as Wernicke's area) (Ahlsйn, 2006).

With regard to particular areas in the left hemisphere involved in language processing, one of the most historically influential models defining the neural organization of language is the Wernicke-Lichtheim-Geschwind (WLG) model (Traxler, 2012). The WLG-model describes three cortical regions belonging to the left hemisphere that are involved in producing and understanding language. The Wernicke's area includes brain regions at the junction of the parietal and temporal cortices, the posterior section of the superior temporal lobe and the angular gyrus. It was originally thought to support language comprehension; nowadays its role is mainly specified as basic processing and analysis of auditory stimuli (Binder et al., 1997). The angular gyrus is also considered to be involved in language processing (in the analysis of visual input, and in semantic processing) (Seghier, 2013). The Broca's area is located in the left inferior frontal gyrus. A third brain structure, called the arcuate fasciculus, is a white-matter tract, or a bundle of axons, that connects the Broca's and Wernicke's areas. Finally, the Geschwind territory, which is connected to both Broca and Wernicke's areasm might take part in conceptual processing.

Although studying patients with aphasia with behavioral tests and establishing their lesion site post-mortem allowed to illuminate some aspects of language representation in the brain, these methods still had major limitations. First, for post-mortem studies doctors had to wait until the patient's death; also, post-mortem changes in the brain could prevent doctors from getting a precise estimate of the lesion site and thus the neural correlates of aphasia in patients. Later, these limitations were solved by neuroimaging methods that provided insights into the neural correlates of language recovery after stroke. Patients are tested with various linguistic tasks during fMRI (functional magnetic resonanse imaging); then the correlations between different experimental and control conditions and areas activated during tasks are found. Sometimes researchers compare activation patterns during successfully completed stimuli with patterns that resulted from errors. Using this modern, more advanced methods, the view of the neural underpinnings of language was refined and developed further, from a more simple to a better defined and clarified claims. Today it is established that the strict division of the brain into areas responsible for particular language functions is too simplistic and there is a strong evidence for numerous other brain areas to play an important role in language abilities (Brunner, Kornhuber, Seemьller, Suger, Wallesch, 1982; Dronkers, 1996; Cole, 1968). Moreover, there are important linguistic processes that are more supported by the right hemisphere: lexico-semantic processing of high-frequency and short words, processing the meaning of ambiguous words, pragmatic skills, and identifying and expressing emotional prosody (Cocquyt, De Ley, Santens, Van Borsel, De Letter, 2017).

All brain areas interact with each other and are incorporated into networks of anatomically and functionally linked areas. Hickok and Poeppel (2004) proposed a model that suggests that dorsal and ventral streams are utilized in language processing. Namely, at earlier stages of speech perception, auditory areas in the posterior superior temporal gyrus are involved bilaterally. At further steps, the ventral stream (a pathway that projects dorso-posteriorly from temporal regions toward the parietal lobe and ultimately to frontal regions) takes part in mapping sound onto meaning and the dorsal stream (a pathway that projects ventro-laterally from temporoparietal toward inferior posterior temporal cortex (posterior middle temporal gyrus) is involved in mapping sound onto articulatory-based representations. Networks that take part in language processing are maintained by white-matter tracts. For example, Ivanova, Isaev, Dragoy, Akinina, Petrushevskiy, Fedina, and Dronkers (2016) studied language processing in aphasia using DTI (diffusion-tensor imaging). They concluded that fiber pathways of white matter are important in comprehension at the word and sentence level.

In the brain of the majority of individuals language is lateralized in the left hemisphere; thus, the work referenced above discusses the bases of language processing mainly in the left hemisphere. However, other studies demonstrate that the left hemisphere does not necessarily dominate and that there is some degree of variability in language lateralization. The main factor determining the lateralization of language functions is handedness. In the brain of right-handed people, language processing relies mainly on the left hemisphere, left-handed people often have bilateral or right-hemispheric lateralization (Knecht et al., 2000). Although handedness and language lateralization show partial pleotropy, there is a substantial amount of ontogenetic differences between them (Ocklenburg, Beste, Arning, Peterburs, and Gьntьrkьn, 2014). Nevertheless, there is indeed a correlation between handedness and language lateralization (Klar, 1999; Medland et al., 2009). Knecht et al. (2000) observed right-hemisphere language dominance in only 7.5% of neurologically healthy right-handed subjects. Szaflarski et al. (2002) found atypical (right or bilateral) language lateralization in 22% of neurologically healthy left-handed and ambidextrous subjects and in only 4-6% of normal right-handed subjects. This demonstrates that the vast majority of right-handers have left hemisphere lateralization.

1.2 The interhemispheric competition hypothesis

Aphasia is a language disorder resulting from aсquired brain damage. Most patients with aphasia lose a considerable amount of language functions as the result of a stroke or a head injury in a language-dominant hemisphere (which is the left hemisphere in a majority of right-handers, as discussed above) (Traxler, 2012). However, it is possible to recover at least a part of impaired language functions. Neurolinguists aim to investigate how to treat patients with aphasia to provide a maximum possible recovery, as well as how to make a precise prognosis of recovery in an individual patient. Various neuroimaging studies have given an insight in understanding of the neural correlates of language recovery after stroke.

In general, four reorganization patterns of the neural bases of language functions after stroke can be logically distinguished, namely:

1) Functional (beneficial for the recovery of language processing) activation of perilesional areas (areas in the left hemisphere);

2) Functional activation of homotopic areas in the right hemisphere;

3) Dysfunctional (maladaptive; i.e., hindering the recovery of language processing) activation of homotopic areas in the right hemisphere; that is what is referred to as `interhemispheric competition hypothesis' (Hamilton, Chrysikou, Coslett, 2011);

4) Dysfunctional (maladaptive) non-activation of the right hemisphere.

Thus, the role of the right hemisphere in the recovery of aphasia is still under debate. Some studies have reported a facilitatory effect of right hemisphere activity on aphasia recovery, while others have argued against this (Cocquyt, De Ley, Santens, Van Borsel, De Letter, 2017), proposing that intact areas of the right hemisphere interfere with language recovery and their overactivation is maladaptive. This hypothesis was called the interhemispheric competition hypothesis and has been entertained since the 19^th century by Wernicke, Henschen, Luria, Geschwind, etc. (Papanicolaou, Moore, Deutsch, Levin, Eisenberg, 1987; Papanicolaou, Moore, Levin, Eisenberg , 1988). The interhemispheric competition hypothesis claims that successful recovery from aphasia following left-hemisphere damage is mediated by activation of perilesional left- hemisphere areas; right-hemisphere activity is maladaptive and prevents the left hemisphere from restoring the functions that it is better suited to serve (Hamilton, Chrysikou, Coslett, 2011).

The hypothesis was tested in a number of neuroimaging studies that investigated whether an increase in activation of the right hemisphere after stroke was associated with better recovery. The results are mixed. In the studies discussed below the authors found that right-hemispheric activation might be beneficial. For example, Menke et. al. (2009) used fMRI to study treatment-related recovery from chronic aphasia. They tested eight patients (3 women; age: 34 to 67 years) with chronic aphasia and moderate to severe word finding difficulties (anomia). For the intensive anomia training, 50 concrete object names were individually selected for each patient. Patients received three hours of daily computer-assisted naming treatment over a period of two weeks. Prior to, immediately after, and eight months after training, patients overtly named trained and untrained objects during fMRI.

The analysis showed increased activity bilaterally in the hippocampal formation, the right precuneus and cingulate gyrus, and bilaterally in the fusiform gyri. The authors suggested that there may be other regions besides the `classical' language areas that are involved in short-term recovery, namely, brain regions responsible for memory encoding; these regions of the right hemisphere might take part in language recovery along with the intact or damaged areas of the left hemisphere.

Saur et al. (2006) conducted a study to investigate brain reorganization during language recovery with fMRI throughout all phases after stroke. They examined 14 patients with aphasia due to an infarction of the left middle cerebral artery territory and an age-matched control group with an auditory comprehension task. Participants were presented with sentences in a correct version and in version containing semantic violation. They were instructed to press a button when they hear an incorrect sentence. The data obtained suggests that recovery proceeds in three steps of brain reorganization: 1) in the acute phase (mean 1.8 days post stroke) a strongly reduced activation in intact left-hemisphere language areas; 2) in the subacute phase (mean 12.1 days post stroke) an upregulation with recruitment of homological language areas in the right hemisphere; 3) in the chronic phase (mean 321 days post stroke) a normalization of activation, possibly reflecting consolidation in the language system. The authors conclude that right-hemispheric activation might be essential in the subacute state, while in the chronic state activation of the left hemisphere is more beneficial for language recovery.

However, some studies show that right-hemispheric activation may interfere with the normal course of recovery. For example, Breier, Randle, Maher, and Papanicolaou (2010) used magnetoencephalography (MEG) to study two patients who underwent melodic intonation therapy (MIT). In the MEG paradigm, the patients were presented with a simple line drawing taken from the Action Naming Test (Obler Albert, 1979). They were told to silently name the action depicted on the pictures as quickly as possible. The stimuli consisted of 160 drawings. MEG was applied before and after each MIT period. MIT is a rehabilitation program that aims to recover language production in patients with nonfluent aphasia. It consists of repetition of linguistic phrases embedded in tonal patterns derived from possible speech prosody patterns (Helm-Estabrooks, Nicholas, Morgan, 1989). After the treatment, one of two patients showed improvement, while the other did not. The patient who did not show improvements after therapy exhibited right hemisphere activation after therapy. The paper suggests that activation in the right hemisphere may be detrimental for language recovery.

Szaflarski et al. (2013) conducted an fMRI study; they evaluated patients with post-stroke aphasia recovery, hypothesizing that although the clinical characteristics of the patients might be similar, the patients who experience better recovery may have different language activation patterns - namely, more left-hemispheric activation. 27 right-handed adults who suffered from LMCA (left middle cerebral artery) stroke at least one year prior to study enrollment took part in the study. According to their Token Test performance, participants were divided into two groups: LMCA-R (`recovered'), in which participants scored ? 41 (in the normal range), and LMCA-NR (`not recovered'), with participants scored ?40 (= aphasic patients). All subjects underwent language assessments before fMRI (the Boston Naming Test, Second edition (BNT), the Semantic Fluency Test (SFT) and Controlled Oral Word Association Test (COWAT), the Peabody Picture Vocabulary Test, Fourth edition (PPVT), the Complex Ideation subtest of the Boston Diagnostic Aphasia Examination (BDAE).

Independent samples t-tests (two-tailed) were performed to characterize language function differences between LMCA-R and LMCA-NR groups. The fMRI paradigm consisted of a semantic decision task as the experimental condition and a tone decision task as the control condition. The analysis revealed that patients from the LMCA-R group, who had normal or close to normal language functions, showed typical (left-hemispheric) fMRI activation, while in patients from LMCA-NR group shifts to the right hemispheric brain regions were observed. Thus, the authors suggest that the right-hemispheric compensatory reorganization is a less effective strategy of language function recovery.

For the summary of the neuroimaging studies on the role of the right hemisphere in language processing, see Table 1.

Despite the conflicting data from neuroimaging, a number of studies that employ brain stimulation techniques (tDCS and TMS) have supported the interhemispheric competition hypothesis. Below a number of studies that exploit a tDCS technique is reviewed.

Table 1. Summary of the neuroimaging studies on the role of the right hemisphere in language processing

Some studies favour the interhemispheric competition hypothesis. For example, Marangolo et al. (2013) conducted a tDCS study, where they used bihemispheric stimulation over the Broca's area and its right homologue. Eight left-brain-damaged participants with non-fluent aphasia were included in the study. Real stimulation consisted of 20 min of 2 mA direct current with the anode placed over the ipsilesional and the cathode over the contralesional inferior frontal gyrus (F5 and F7 of the extended International 10-20 system for EEG electrode placement). For sham stimulation, the same electrode positions were used.

Before the treatment, 126 stimuli (syllables, words and sentences) were auditorily presented. The authors identified which stimuli were incorrectly named or omitted and included them in two experimental lists. The experimenter presented one stimulus at a time, and for each stimulus, the treatment, which lasted 10 days, involved the use of four different steps: 1) the experimenter asked the patient to repeat the stumulus after him/her; 2) The clinician presented the stimulus with a pause between syllables, prolonged the vowel sound, exaggerated the articulatory gestures, and asked the patient to do the same; 3) the same step as 2); 4) The experimenter auditorily presented one syllable at a time, prolonged the vowel sound, exaggerated the articulatory gestures, and asked the patient to do the same. When the patient could reproduce the articulatory gestures shown by the clinician, he or she would be asked to repeat the whole stimulus without the experimenter's help and only if he or she succeeded in doing so again the response was considered correct. The analysis revealed that articulatory errors significantly decreased and all patients were faster in repeating the stimuli compared to the sham condition, confirming that bihemispheric stimulation might be a useful tool for the treatment of aphasia. However, the authors did not explore how bilateral tDCS language treatment may influence brain functional connectivity reorganization in patients.

Marangolo et al. (2016) aimed to cover this issue and performed a bilateral tDCS study in patients with aphasia. They tested nine participants with non-fluent aphasia following damage to the left hemisphere. All patients underwent two tDCS conditions: sham and real stimulation; 15 daily sessions in each. Real stimulation consisted of 20 min of 2 mA direct current with the anode placed over the ipsilesional and the cathode over the contralesional IFG (F5 and F7 in 10-20 system for EEG electrode placement); the same montage was used for sham stimulation. Patients were administered all the standardized language tests at the beginning and at the end of each treatment condition. Before the treatment, 160 stimuli (syllables and words) were audibly presented.

The authors identified which stimuli were incorrectly named or omitted and included them in the two experimental lists. The experimenter presented one stimulus at a time, and for each stimulus, the treatment involved the use of four different steps: 1) the experimenter asked the patient to repeat the stumulus after him/her; 2) The clinician presented the stimulus with a pause between syllables, prolonged the vowel sound, exaggerated the articulatory gestures, and asked the patient to do the same; 3) the same step as 2); 4) The experimenter auditorily presented one syllable at a time, prolonged the vowel sound, exaggerated the articulatory gestures, and asked the patient to do the same. When the patient could reproduce the articulatory gestures shown by the clinician, he or she would be asked to repeat the whole stimulus without the experimenter's help and only if he or she succeeded in doing so again the response was considered correct. Before and after each session each participant underwent fMRI. The results showed that real bilateral stimulation resulted in improvements not only in patients' performance for the treated stimuli, but also for untrained items. Moreover, fMRI scans showed that real stimulation yielded stronger functional connectivity increase in the left hemisphere.

Fiori at al. (2017) tested the interhemispheric competition hypothesis in healthy volunteers. They aimed to compare the performance of young participants with the elderly group. Fifteen healthy right-handed volunteers aged 20-40 years (young group) and 60-80 years (elderly group) participated in the study. In both groups, each participant was randomly assigned to one of the three groups: 1) anodal stimulation over the left temporal cortex, which included Wernicke?s area (CP5 of the extended International 10-20 system for EEG electrode placement) with cathode over the right orbito-frontal cortex; 2) bilateral stimulation, with anode placed over the left temporal cortex, which included Wernicke?s area (CP5 of the extended International 10-20 system for EEG electrode placement) and cathode over the controlateral right homologue area (CP4 of the extended International 10-20 system for EEG electrode placement); 3) sham stimulation - the anode was placed over the left emporal cortex, while the cathode was placed for half of the subjects as in the unihemispheric condition and for the remaining half as in the bilateral condition.

Participants took part in a single 30-minute session each. The experiment included three phases: training, verification and word retrieval. During the training session, participants were shown pictures paired with nonwords written on them. Subjects were instructed to pay attention to pictures and memorize them. At Phase 2, participants were asked to press the button for whether they see a pairing `picture+nonword' that they have already seen at Phase 1. tDCS was applied at 2 mA intensity for 20 min. At Phase 3, participants underwent tDCS and were asked to name aloud pictures using the corresponding previously matched written pseudoword. The results showed that, although no difference between tDCS conditions in young volunteers was revealed, the elderly group was more accurate and faster in the bihemispheric condition compared to the other two. The authors suggest there might be the following explanations to this: 1) the word recall task was particularly difficult only for elderly, but not for young participants; 2) word recall abilities may decline with aging; therefore tDCS might influence language abilities only in individuals with the loss of linguistic processing abilities.

Costa, Giglia, Brighina, Indovino, and Fierro (2015) conducted a case study in one patient. The patient was a 57-year-old woman who suffered from severe non-fluent aphasia due to a stroke following the right hemisphere. For the pilot experiment, thirty-two pictures (16 objects and 16 actions), extracted from BADA (the Battery for the Analysis of the Aphasic Deficit) and matched for frequency, were presented to the patient along with the first or the second letter of the word. Three single sessions of bilateral tDCS were conducted with a 1-week interval between them. Two types of bilateral stimulation were used: in the first session, the anode was placed on the left and the cathode was placed on the right Broca's area (“anode on the left” condition); then the montage was inverted in the other two sessions - first real (“anode on the right” condition) and second sham stimulation (sham condition). Between the two real stimulations, performances in the naming task were measured without stimulation.

The current intensity was 1 mA, applied for 20 min. The results of the pilot experiment showed that tDCS efficacy was greater when the cathode was positioned on the right (“anode on the left” condition). Thus, for the two experimental treatments the “anode on the left” condition was used. In Experiment 1, Broca's area and its right-hemispheric homologue were stimulated, and in Experiment 2, Brodmann areas 39/40 were stimulated, for 2 weeks. In both experiments, tDCS was applied at 1 mA for 20 min daily for 2 weeks. The patient performed the naming task at the following times: before tDCS, immediately after the end of tDCS treatment (T1), and 3 days after the end of the treatment, and if an effect was found, her performance was checked every 3 days, until it decreased to the baseline level. After the effect was lost, sham stimulation was performed. Experiment 2 was performed 4 months after Experiment 1. 4 weeks after Experiment 2, Experiment 3 was conducted. Tasks, tDCS parameters, and experimental timeline were the same as for Experiments 1 and 2, but in this case, the anode was placed on the right BA 39/40 and the cathode was placed on the left BA 39/40. The results showed that anodal stimulation of the left Brodmann areas 44/45 and simultaneous cathodal stimulation of the right BA 44/45 (Experiment 1) significantly increased the sum score. By contrast, the same montage on Brodmann areas 39/40 (Experiment 2) caused no significant improvement, but the reverse polarity, anodal tDCS on the right and cathodal tDCS on the left Brodmann areas 39/40 (Experiment 3), boosted participant's performance. Although the authors did not specifically compare bihemispheric tDCS to unihemispheric stimulation, they found that bihemishperic stimulation can boost language performance in aphasic patients.

Lee, Cheon, Yoon, Chang, and Kim (2013) investigated the effects of bihemispheric tDCS on the language function of patients with aphasia in comparison to unihemispheric tDCS in order to determine the most effective tDCS method for aphasia treatment. 11 patients took part in the study; 6 patients had non-fluent aphasia, 5 patients had fluent aphasia. tDCS was delivered at 2 mA for 30 minutes. Active stimulation electrodes were placed over the left inferior frontal gyrus in the unihemispheric montage or both the left and right left inferior frontal gyrus in the bihemispheric montage, while reference electrodes were applied to the buccinator muscle of the same side. Each participant took part in both bihemispheric and unihemispheric stimulation session on different days. To assess language abilities of the patients, a picture naming test and a verbal fluency test were completed by participants before and after each tDCS session.

The results showed that reaction times in the picture naming test were significantly shortened only after interhemispheric tDCS, suggesting that interhemispheric tDCS may be more effective than unihemispheric tDCS. Accuracy was significantly boosted after both bihemispheric and unihemispheric tDCS. Performance in the verbal fluency test was not significantly affected neither by interhemispheric nor by unihemispheric tDCS.

However, a number of studies failed to support the interhemispheric competition hypothesis. For example, Meinzer et al. (2014) did a study that aimed to uncover whether tDCS administered to the primary motor cortex can enhance language functions. Participants simultaneously underwent fMRI to let the authors reveal the neural mechanisms underlying tDCS effects. Eighteen healthy older adults participated in the study. A constant direct current at 1 mA was administered by an MRI-compatible stimulator for 30 minutes. For anodal stimulation, anode was placed over the left M1 and cathode over the right supraorbital region. For bihemispheric tDCS anode was placed over the left M1 and cathode over the right M1. During sham-tDCS the reference electrode was pseudo-randomly assigned to either the right supraorbital region or right M1 in half of the participants. During intrascanner tDCS (tDCS that was applied during fMRI) participants were shown six semantic categories and were instructed to produce an exemplar of the corresponding semantic category. The authors found that both active stimulation conditions resulted in a smaller number of errors than in sham condition, but no significant difference between bihemispheric and unihemispheric condition.

Hamilton, Chrysikou, and Coslett (2011) reviewed recent studies applying non-invasive brain stimulation (TMS, tDCS) in patients with chronic aphasia. Studies that show that inhibitory stimulation of the right hemisphere interferes with the course of recovery reject the interhemispheric competition hypothesis; however, there is also evidence in favor of beneficial effects of stimulation of the right hemisphere. The authors suggest that improvements may be associated with stimulation of the pars triangularis, as most of the previous studies that showed beneficial effect of right hemisphere targeted this region.

For the summary of the tDCS studies on language, see Table 2.

The interhemispheric competition hypothesis was tested not only for language, but for various other cognitive functions. Lindenberg, Nachtigall, Meinzer, Sieg, and Flцel (2013) assessed the neural correlates of bihemispheric and unihemispheric tDCS applied during fMRI scanning. 20 healthy older subjects participated in the study. Subjects took part in three MRI sessions with bihemispheric, anodal (unihemispheric) or sham tDCS, separated by at least 1 week to prevent carry-over effects. For bihemispheric stimulation, anode was placed over the left M1 and cathode over the right M1; for anodal stimulation, anode was placed over the left M1 and cathode (reference electrode) over the contralateral supraorbital region. In the sham condition, the electrode setup was pseudo-randomly assigned to participants (either as in bihemispheric condition or in anodal condition) and balanced across the group. During real stimulation, a current of 1 mA was constantly delivered for 30 min. Participants were presented with three different symbols in the center of the visual field and instructed to respond with button or to withhold a response. Analysis showed that task-related activity was stronger in bilateral M1 during bihemispheric tDCS compared with unihemispheric (anodal) tDCS. However, both active conditions resulted in increased activation compared to sham.

Table 2. Summary of tDCS studies on language

Lindenberg, Sieg, Meinzer, Nachtigall, and Flцel (2016) aimed to explore the mechanisms underlying bihemispheric and unihemispheric tDCS in young healthy adults. 24 healthy right-handed subjects were assessed during three identical fMRI sessions (bihemisperic, anodal, or sham tDCS), separated by at least one week. In all stimulation conditions, the anode was placed over left M1. In the bihemispheric condition, position C4 served as a reference for the cathode placement. In the anodal condition, the cathode was placed over the contralateral supraorbital region. In the sham condition, the set-up was randomly assigned to participants (either “bihemispheric” or “anodal”) and balanced across the group. For real stimulation, a current of 1 mA was constantly delivered for 30 min. In between fMRI sequences, subjects participated in an fMRI overt semantic word-retrieval task (data were not analysed in this paticular study). Participants were also presented with three different symbols in the center of the visual field and instructed to respond with button or to withhold a response. The results showed that unihemispheric (anodal) stimulation appeared to specifically exert its effects “locally” on primary and non-primary motor cortices targeted by the anode. Bihemispheric tDCS was characterized by more complex bihemispheric changes (mainly, bilateral motor cortex disinhibition).

Giglia et al. (2011) conducted a study aiming to compare bilateral tDCS to cathodal tDCS of the right hemisphere. The authors studied 11 healthy volunteers. They examined participants' performance on a computerized visuospatial task several times: before, during and after tDCS. Cathode was set on P6 in bihemispheric and cathodal conditions; anode was set over P5 for bihemispheric condition and over the contralateral orbita for cathodal condition. For real stimulation conditions, tDCS was delivered at 1 mA for 15 min. Sham stimulation was also administered to control unspecific effects of bihemispheric tDCS; with electrode positioning the same as in real bihemispheric session. Participants were presented with visual stimuli (black lines transected by a vertical bar) and were instructed to press the button to decide whether the left part of it is longer, shorter or equal to the right part. In a baseline condition (before stimulation), a common phenomenon known as “pseudoneglect” was observed. However, a significant rightward bias in symmetry judgments as compared with baseline and sham conditions was observed in both the stimulation approaches. After bihemispheric tDCS compared with cathodal stimulation, the effect was stronger and appeared earlier, but no longer-lasting after effects were found.

For the summary of tDCS studies on other cognitive functions, see Table 3.

To sum up, there is a number of studies both supporting and contradicting the interhemispheric competition hypothesis. Previous studies demonstrate mixed results: a number of neuroimaging studies favours the interhemispheric competition hypothesis, while others show evidence against it. A large body of, but not all, non-invasive brain stimulation studies (TMS, tDCS) also supports the interhemispheric competition hypothesis (Hamilton, Chrysikou, Coslett, 2011).

Table 3. Summary of tDCS studies on other cognitive functions

However, one of the important limitations of the previous studies is that bilateral condition was not compared to both necessary control conditions (separate anodal and cathodal stimulation). Thus, it might be important to discover whether bilateral stimulation causes greater effects than unilateral tDCS. Moreover, among the previous studies only Giustolisi, Vergallito, Cecchetto, Varoli, and Lauro (2018) tested the effect of tDCS in a sentence-level task, which is crucial for understanding how language processing is modulated by tDCS.

It might also be of importance to add further evidence to methodological issues related to tDCS and to the effect of stimulation itself in healthy participants. Previous studies on tDCS in both healthy and clinical populations employ a rather small number of participants and therefore have less statistical power, which might be especially important in studies with healthy individuals. This study aims to test both the interhemispheric competition hypothesis and the main effect of stimulation on a large sample (n = 72).

1.3 Transcranial direct current stimulation

Previous studies mentioned above tested the interhemispheric competition hypothesis with the use of various neuroimaging and brain stimulation methods. The current study employs tDCS - a non-invasive brain stimulation method, which is based primarily on polarization of neurons. tDCS delivers constant low current through the electrodes and modulates cortical excitability. tDCS acts upon the resting membrane potential through the modulation of sodium and calcium-dependent channels and NMDA-receptor activity.

tDCS subdivides into two types: anodal tDCS (active electrode (anode) placed on a target area delivering positive charge) and cathodal tDCS (active electrode (cathode) delivering negative charge). Until recently, it has been widely believed that anodal DCS increases cortical excitability, whereas cathodal stimulation decreases it. However, further investigations showed that it is not always the case: Pirulli, Fertonani, and Miniussi (2014) demonstrated improved performance in a visual task following cathodal tDCS over the primary visual cortex.

tDCS is considered to be a promising method for clinical use and effective treatment for patients with aphasia, as well as for patients with other neurologic and psychiatric conditions because:

· tDCS is a safe method. In a clinical context tDCS is an promising form of neurostimulation in chronic stroke populations because it has not been reported to provoke seizures (Holland and Crinion, 2012);

· tDCS is easy to use even without a doctor or a supervisor;

· tDCS is relatively cheap;

· tDCS can be used in combination with other treatments, including medicine;

· tDCS is well suited for online application - namely, for doing language tasks during stimulation. tDCS can be used both in clinical research, where tDCS can be applied during behavioral tasks, and in clinical practice simultaneously with the methods of speech therapy;

· tDCS has higher tolerability and never provokes uncontrolled movements.

Results of tDCS are believed to be very sensitive to stimulation factors. That is why defining parameters of stimulation is crucial. These parameters include electrode size and positioning, intensity, duration of stimulation, number of sessions per day, and interval between sessions.

In tDCS, low direct current is delivered through the electrodes, which generate an electric dipole between each other. One of the electrodes is placed over the region of interest (stimulating electrode) and the other electrode, the reference electrode, is placed in another location in order to complete the electrical circuit. There are two ways of positioning electrodes: `unipolar' and `bipolar'. In a unipolar montage, the reference electrode is placed on another body part. Bipolar montage is the montage with both electrodes on the head. Importantly, the reference electrode can also contribute to the stimulation outcomes and be a possible confounder in conducting research; that is why electrode positioning should be chosen carefully.

The most common current intensity is 1-2 mA. Such current intensity remains below the safety limits and provides effective stimulation of the target areas. In fact, the current that reaches the area of interest is lower and depends on a range of factors. These include skin and skull resistance, resistance of intracranial structures (e.g., blood vessels, cerebrospinal fluid) and the resistance of brain tissue. Moreover, skull diseases, neuropsychiatric disorders, use of pharmacotherapy etc. influence the amount of current delivered. Finally, the baseline of cortical excitability is dependent on age, gender, and smoking. Such variability of factors may result in negative findings. Saline-soaked sponges (with electrodes put in them) are used to minimize skin resistance (Brunoni et.al., 2012).

The cognitive effects induced by tDCS are largely dependent on the current intensity. Several studies demonstrated effects such as enhanced verbal fluency improvement at 2 mA versus lower improvement at 1 mA (Iyer et al., 2005) and working memory improvement at 2 mA versus no improvement at 1 mA (Boggio et al., 2006) (Brunoni et.al., 2012). It is believed that increasing the current density increases the depth of the electrical field penetration. However, Batsikadze et al. (2013) showed that an enhancement of tDCS intensity does not necessarily results in increase of its efficacy, but might also change the direction of effects. Thus, relationship between stimulation effects and current density is not straightforward.

The duration of stimulation is substantial for defining relevant parameters of stimulation, especially for clinical purposes, where the aim is to get long-lasting effects. Short applications (up to a few minutes) induce changes of excitability during stimulation, but have no after-effects. Some studies show that repeated sessions of tDCS may have cumulative effects associated with greater magnitude and duration of behavioral effects. For example, tDCS sessions applied for 5 consecutive days (daily sessions) resulted in improvement lasting for up to two weeks, while such effects were not observed for weekly sessions (once a week). (Brunoni et al., 2012). Interestingly, Monte-Silva, Kuo, Liebetanz, Paulus, and Nitsche (2010) revealed that when the second stimulation was applied during the after-effects of the first, an enhancement of tDCS-induced effect was obtained. According to (Monti et al., 2013), current density of 1-2 mA for 20 min using electrode sizes of 35 cm² in repeated daily sessions (3- 5 days) might be the optimal parameters (Zelenkova, 2016 (unpublished manuscript)).

However, it still remains unclear whether tDCS can modulate language processing in healthy individuals. Previous studies testing the interhemispheric competition hypothesis in healthy population show that interhemispheric stimulation might enhance language processing (Fiori et al. 2017), while others do not support the hypothesis. For example, Westwood, Olson, Miall, Nappo, and Romani (2017) conducted four separate experiments with healthy individuals to examine the effects of anodal tDCS of frontal and temporal lobe on reading and naming tasks. The authors did not find any difference between sham and real stimulation of any of the target areas. The authors hypothesize, among other possible explanation, that the brain of healthy individuals is already in almost its maximum capacity and therefore their performance can not be altered by tDCS. The lack of negative findings may also be due to regression-to-the-mean in low-powered studies and/or publication bias (Westwood, Romani, 2017). Thus, this study aims to contribute to this issue and test the interhemispheric competition hypothesis on a large sample of healthy participants (n=72).

language brain transcranial aphasia

2. Methods

2.1 Participants

The participants were 72 young volunteers (49 females; mean age 22.94, SD 3.75, range 18-32 years), all self-reportedly right-handed (mean score 65.7, SD 19.0, range 22.7-100.0), monolingual native speakers of Russian, with normal or corrected-to-normal vision and no reported history of neurological, psychiatric, or speech-language disorders. Participants completed a tDCS safety questionnaire before the study to rule out any contraindications and to gather demographic information. Participants were blind to their experimental assignment and to the experimental design.

They also completed a handedness questionnaire to determine their handedness score. The handedness questionnaire was the Edinburgh Handedness Inventory (Oldfield, 1971), which is a measurement scale to assess a person's dominant hand in everyday activities: using a fork or a knife, kicking a ball, striking a match etc. (11 activities). Participants were asked to assess their preference on a scale ranging from 1 to 5 (1 - “I use only the left hand/leg to do this”, 5 - “I use only the right hand/leg to do this”). All participants were paid for the experiment if they took part in both sessions. The study protocol conformed to the Declaration of Helsinki and was approved by the local University Research Ethics committee.

2.2 Transcranial direct current stimulation

tDCS was delivered at 1.5 mA for 20 minutes using a battery-driven Starstim® stimulator (Neuroelectrics), via round 25 cm² rubber-sponge electrodes, soaked in saline and positioned in the supplied neoprene headcap. In a between-group design, participants were randomly assigned to one of the three stimulation conditions (n=24 per condition) - anodal, cathodal and bilateral condition - using a sealed-envelope approach. Participants in the left anodal condition received anodal stimulation over left posterior inferior frontal gyrus (IFG), corresponding to Broca's area (F7), with the reference electrode (cathode) at Pz. Participants in the right cathodal condition received cathodal stimulation over right IFG (F8), while the reference electrode (anode) was placed at Pz. Participants in the bilateral condition received a combination of anodal stimulation over left IFG and cathodal stimulation over right IFG (anode at F7, cathode at F8) (for the electrode placement, see Figure 1). Every participant underwent real and sham stimulation on different days; session order was counterbalanced across participants. For sham stimulation, current intensity was also ramped up to 1.5 mA but then ramped down in 50 seconds. The interval between sessions was 6.33 days on average (range 1-26 days) and did not differ between groups (F(2, 69) = 1.05, p = 0.35).