- Email: [email protected]
Bilateral temporal cortex transcranial direct current stimulation worsens male performance in a multisensory integration task
Neuroscience Letters 527 (2012) 105–109
Contents lists available at SciVerse ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Bilateral temporal cortex transcranial direct current stimulation worsens male performance in a multisensory integration task Olivia Morgan Lapenta a , Felipe Fregni b,c , Lindsay M. Oberman c , Paulo Sergio Boggio a,∗ a
Social and Cognitive Neuroscience Laboratory and Developmental Disorders Program, Center for Health and Biological Sciences, Mackenzie Presbyterian University, Sao Paulo, Brazil Laboratory of Neuromodulation, Department of Physical Medicine and Rehabilitation, Spaulding Rehabilitation Hospital and Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA c Berenson-Allen Center for Non-invasive Brain Stimulation, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA b
h i g h l i g h t s tDCS induces a polarity, gender and task specific effect in a multisensory integration task. Bilateral cathodal temporal cortex tDCS worsens task accuracy in males only. Bilateral anodal or cathodal temporal cortex tDCS results in decreased reaction time for both genders.
a r t i c l e
i n f o
Article history: Received 4 April 2012 Received in revised form 29 May 2012 Accepted 10 August 2012 Keywords: Multisensory integration tDCS Superior temporal sulcus Gender Extreme male brain theory
a b s t r a c t Somatosensory integration is a critical cognitive function for human social interaction. Though somatosensory integration has been highly explored in cognitive studies; only a few studies have explored focal modulation of cortical excitability using a speech perception paradigm. In the current study, we aimed to investigate the effects of tDCS applied over the temporal cortex of healthy subjects during a go-no-go task in which stimuli were shapes and non-words. Twenty-eight subjects were randomized to receive cathodal, anodal or sham tDCS bilaterally over the superior temporal cortex (the reference electrode was on deltoid) in a counterbalanced order. The effects on judgment of congruency between shapes and non-words in healthy volunteers were measured by a go-no-go task. Our findings show a significant modification of performance according to the polarity of stimulation, task and subject gender. We found that men performed worse on the no-go condition for congruent stimuli during cathodal tDCS. For reaction time, on the other hand, there was a similar effect for anodal and cathodal stimulation. There were significantly faster responses on incongruent trials during both anodal and cathodal tDCS. Along with previous literature showing gender differences in tasks associated with speech perception, the findings of this study provide additional evidence suggesting that men may have a more focal and restricted neural processing in this multisensory integration task. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Somatosensory integration is a critical component of cognitive processing. It is characterized by dual processing that integrates different sensory inputs such as visual and auditory stimuli in order to provide an adequate response. As these different signals can arise from common external events or objects, the integration between them can be useful [7]. It is a particularly important stage
∗ Corresponding author at: Social and Cognitive Neuroscience Laboratory and Developmental Disorders Program, Center for Health and Biological Sciences, Mackenzie Presbyterian University, Rua Piaui, 181, 10◦ andar, Sao Paulo 01241-001, SP, Brazil. Tel.: +55 11 2114 8001; fax: +55 11 2114 8563. E-mail address: [email protected] (P.S. Boggio). 0304-3940/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neulet.2012.08.076
of processing as most of our cognitive computations involve simultaneous processing of more than one type of sensory stimulation. Theoretical perspectives suggest that brain functions can present interactive distribution of functional systems, e.g. hearing a word is associated with the activation of articulatory motor areas. Thus, there are neural networks combining processing from different inputs. Additionally, inputs from a different modality can activate other areas through processes of association and integration [13]. The motor theory of speech perception suggests that the objects of perception of speech are the “phonetic gestures”, represented in the observer’s brain as a motor command of the signal, characterized by movements of the mouth, lips and tongue [6]. Specific neural networks have been described for different modalities of multisensory integration. For example, specific
106
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
regions of the cortex, including Broca’s area and superior temporal sulcus (STS) are likely involved in the integration of shapes and non-words. These regions may underlie the shared representation of the auditory perception of a sound and the motor representation of mouth and larynx movement needed to make the sound [11]. Neuropsychological studies have shown the relationship between motor action and speech production using cognitive tasks of association between images and written non-words where these images have been produced according to the phonetic gestures used to produce the sound of the corresponding non-word, e.g. sounds that need rounded movements of the mouth to be produced have rounded shapes to represent them [8,11]. We therefore hypothesize that a technique – transcranial direct current stimulation (tDCS) – that can alter cognitive processing might be useful to alter performance on a multisensory task. We chose a multisensory task integrating shapes and non-words as we were interested in exploring the effects of tDCS over temporal cortex in speech perception. We expect enhanced performance with anodal tDCS and a decreased performance for cathodal tDCS. Also, based on our preliminary findings, we expected to find a gender effect since language and some tasks associated with crossmodal processing differ in men and women.
2. Materials and methods We investigated the effects of anodal, cathodal and sham tDCS applied bilaterally over the STS in twenty eight volunteers (14 female and 14 male mean age of 23.2 ± 3.1). All subjects had a similar level of education. This sample is similar to the samples we have been testing in previous cognitive tDCS studies. Subjects were regarded as suitable to participate in this study if they fulfilled the following criteria: (1) age between 20 and 30 years; (2) no clinically significant or unstable medical, or neuropsychiatric disorder; (3) no use of central nervous systemeffective medication; (4) no history of brain surgery, tumor, or intracranial metal implantation. All study participants provided written, informed consent. This study was approved by the institutional ethics committee from Mackenzie Presbyterian University, Brazil. Before starting the experiment, all the participants had to complete successfully a pre-test consisting of visual presentation of pairs of images/non-words. Stimuli were provided from a previous study where neurotypical individuals corrected guessed the matching as expected, at a rate of 88% of trials [11]. Thus, before starting the experiment, all the participants had to complete successfully a pre-test consisting of visual presentation of pairs of images/nonwords. Subjects were included in this study if they assigned the correct correspondence between images and non-words with a minimum rate of 80%. All subjects received in different days with an interval between sessions of at least 48 h, sham, anodal and cathodal temporal cortex tDCS. In order to prevent learning effects, the tDCS sessions were randomized using the Latin Square method. As the number of subjects was not a multiple of 3, the final distribution was not fully counterbalanced. Two pairs of surface sponge electrodes (35 cm2 ) were soaked in saline solution and applied to the scalp at the desired areas of stimulation and to the right deltoid muscle as the reference electrode. We used this bi-temporal extracephalic montage before in other studies and obtained significant behavioral changes [5]. Rubber bandages were used to hold the electrodes in place for the duration of stimulation. For anodal stimulation anodal electrodes were placed over T3 and T4 (according to the 10–20 system for EEG electrode placement) and cathodal electrodes on the right arm; for cathodal stimulation the cathodal electrodes were
placed over T3 and T4 and anodal electrodes on the right arm; for sham stimulation we adopted the same position as for active stimulation, however, subjects only received stimulation in this condition for 20 s. Participants in the active conditions received a constant current of 1 mA for 14 min (5 min of tDCS only followed by 9 min of tDCS during the go-no-go task of congruency between shapes and non-words). The task is composed of 10 images, each having a corresponding written non-word. We created 10 stimuli which were 10 congruent and 10 incongruent pairs of images/non-word randomly presented. The shapes and written non-words were presented visually at the screen, side by side. The experiment was composed of 6 blocks of 40 trials. In half of blocks, subjects were instructed to press a button when the stimulus was congruent and in the other half when it was incongruent (Fig. 1). The blocks were presented alternately. At the first testing day, all subjects performed a training test composed by one go-congruent and one nogo-congruent blocks, each one containing 12 trials. Therefore most of the learning effect was achieved during this phase. Analyses were done with Statistica software (version 8.0, StatSoft Inc.). We performed repeated measures ANOVAs in which the dependent variables were number of correct responses on Go trials, number of correct responses on NoGo trials, and reaction time (for correct responses on Go trials) and the independent variables were condition of stimulation (anodal, cathodal or sham), gender (male or female), and congruency (congruent or incongruent stimuli). Despite previous tDCS go-nogo studies have performed analyses not separating go and nogo conditions, we performed separated analyses for both conditions as recent electrophysiological data suggest that the neural network associated with failure to press following go stimulus is different than that associated with pressing following nogo stimulus [12]. We also considered the interaction of gender vs. tDCS, congruency vs. tDCS, congruency vs. gender, and tDCS vs. gender vs. congruency. When appropriate, post hoc comparisons were carried out using Fisher’s LSD. Unless stated otherwise, all results are presented as means, confidence intervals, and standard errors. Statistical significance refers to a p value <0.05.
3. Results All subjects completed the entire experiment. All subjects tolerated the stimulation well and no side effects were reported. Also, bilateral stimulation was not associated with additional discomfort in any subject. Initially, we performed a repeated measures ANOVA in which the dependent variable was correct responses on Go trials (‘pressing a button’ responses). The ANOVA did not reveal significant effects for gender (F1,26 = 3.8; p = 0.1), tDCS (F2,52 = 0.6; p = 0.6), gender * tDCS (F2,52 = 0.2; p = 0.8), congruency * tDCS (F2,52 = 0.5; p = 0.6) or congruency * gender * tDCS (F2.52 = 0.9; p = 0.4). However, the ANOVA did reveal significant effects for congruency (F1,26 = 35.7; p = 0.000003, p 2 = 0.58) and for the interaction term congruency * gender (F1.26 = 7.5; p = 0.01, p 2 = 0.22. These significant effects, as it can be observed in Fig. 2, were due to a better performance of both genders on congruent trials as compared to incongruent ones. With regard to the interaction congruency * gender, Fischer LSD showed significant differences between congruent and incongruent correct responses for women (p = 0.000001) and for men (p = 0.04). Post hoc comparisons also revealed significant difference between women and men accuracy for incongruent trials (p = 0.03), but not for congruent trials (p = 0.7). As it can be seen in Fig. 2, women had poorer performance on incongruent trials as compared to men, but both presented similar performance for congruent trials.
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
107
Fig. 1. Diagram of the task. Example of go-congruent and go-incongruent blocks, respectively.
We then performed a repeated measures ANOVA in which the dependent variable was correct responses on NoGo trials (‘notpressing a button’ responses). The ANOVA did not reveal significant effects for Gender (F1,26 = 0.001; p = 1.0), tDCS (F2,52 = 0.3; p = 0.8), gender * tDCS (F2,52 = 0.1; p = 0.9), congruency (F1,26 = 0.1; p = 0.7), congruency * gender (F1.26 = 2.5; p = 0.12) or congruency * tDCS (F2,52 = 0.2; p = 0.8). However, the ANOVA did reveal a significant effect for the interaction congruency * gender * tDCS (F2.52 = 3.3; p = 0.047, p 2 = 0.11). With regard to this interaction, Fischer LSD showed significantly poorer performance of men during cathodal tDCS on congruent trials as compared to responses on NoGo congruent trials of men during anodal tDCS (p = 0.012), responses for NoGo incongruent trials of men during anodal tDCS (p = 0.035), responses for NoGo incongruent trials of men during cathodal tDCS (p = 0.012), responses for NoGo congruent trials of men during sham tDCS (p = 0.048) and responses for NoGo incongruent trials of men during sham tDCS (p = 0.012) (Fig. 3). There were no significant effects in the other comparisons (p > 0.05).
Fig. 2. Accuracy for congruent and incongruent conditions in Go trials. (*) Women presented poorer accuracy in the incongruent go trials when compared to congruent go trials (p = 0.000001) and when compared to men on incongruent go trials (p = 0.003). (**) Men presented poorer accuracy in incongruent go trials when compared to congruent go trials (p = 0.04). Both gender groups had similar performance in the congruent condition (p = 0.7).
Finally, we performed a repeated measures ANOVA in which the dependent variable was response time on Go trials (‘pressing a button’ responses). Response time was only considered for correct responses. The ANOVA did not reveal significant effects for gender (F1,26 = 0.1; p = 0.7), tDCS (F2,52 = 0.3; p = 0.7), gender * tDCS (F2,52 = 1.0; p = 0.4), congruency * gender (F1.26 = 3.4; p = 0.08), or congruency * gender * tDCS (F2.52 = 1.8; p = 0.2). However, the ANOVA revealed a significant effect for congruency (F1,26 = 442.9; p < 0.000001, p 2 = 0.9) and the interaction of Congruency and tDCS (F2,52 = 4.6; p = 0.01, p 2 = 0.15). With regard to the congruency * tDCS interaction, Fischer LSD showed significant differences between response time for congruent trials vs. incongruent trials (p < 0.000001 for all comparisons). In addition, it was found that response time on incongruent trials were significantly reduced for anodal and cathodal tDCS as compared to sham tDCS (p = 0.003 and p = 0.004, respectively) (see Fig. 4).
Fig. 3. Accuracy on NoGo trials considering type of stimulation, congruency and gender. Men showed reduced performance in NoGo congruent trials when receiving cathodal tDCS when compared to all other conditions (NoGo congruent trials of men during anodal tDCS (p = 0.012), NoGo incongruent trials of men during anodal tDCS (p = 0.035), NoGo incongruent trials of men during cathodal tDCS (p = 0.012), NoGo congruent trials of men during sham tDCS (p = 0.048) and for NoGo incongruent trials of men during sham tDCS (p = 0.012).
108
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
Fig. 4. (**) Reaction time differed for congruent and incongruent trials (p = 0.000001). (*) Cathodal and anodal stimulation differed from sham stimulation during incongruent trials (p = 0.003 and p = 0.004, respectively).
4. Discussion Our findings show that tDCS is a useful tool to investigate multimodal integration as it induces a significant modification of performance according to the polarity of stimulation and subject gender. In this experiment the effects were gender-specific. We found that men performed worse on the no-go condition for congruent stimuli during cathodal tDCS. Accuracy on the no-go condition was indexed by the number of correct responses (notpressing a button in response to specific stimuli). We found that men pressed the button more times when faced with congruent stimuli during the block that they were asked to respond to incongruent ones. Therefore, they incorrectly indicated that the stimulus was congruent when it was incongruent when stimulated with cathodal tDCS. This suggests a performance worsening induced by cathodal tDCS on a specific multisensory integration task. We applied cathodal electrodes over both right and left superior temporal cortex, well-know regions of multisensory integration. Previous findings have already shown enhanced recruitment of this area during presentation of congruent stimuli as compared to incongruent one. Given these findings showing cortical recruitment during congruent stimuli, it is plausible to consider that the effect on congruent stimuli might be the result of the inhibitory effect of bilateral cathodal tDCS of this area in both hemispheres. Nevertheless, it is interesting that this effect was observed for men but not for women. A similar gender effect, but with a different tDCS montage, was observed earlier by our group. Boggio et al. [2] found that temporal cortex tDCS with an anodal electrode over T3 and cathodal over T4 lead to differential effects according to gender. In this study, authors found that men presented more errors on a facial expression go-no-go task as compared to sham. Both findings give evidence that the functionality of the temporal lobe and its neural network distribution are dependent on gender. The gender differential effect found here might be related to the type of task and the neural network underlying its processing. This finding is in accordance with the extreme male brain (EMB) theory that argues that ASD individuals have an exaggerated typically male brain, i.e. an extreme systemizing pattern related to more focal and restrict networks in males as opposed to larger and more distributed networks in females [1]. The findings of Oberman and Ramachandran [11] where ASD children were less likely to choose the correct non-word for the correspondent shape along with Boggio et al. [2] results of different tDCS effects according to gender are in line to the EMB theory. Therefore is possible that for multisensory
integration, male and female networks differ and given EBM theory in which males may have activation in more focal and restrict network, the inhibition of superior temporal cortex (an area that seems to play a role in the language deficits seen in children with ASD) by cathodal tDCS could induce an effect in males only. Furthermore, different activations of male and female brain areas have been shown before. Ruytjens et al. [14] using fMRI showed different activations between male and female during silent lipreading. Women activated key cortical areas known to be involved in lipreading in both hemispheres while men activated those areas only in the left hemisphere, except for the associative auditory temporal area. Furthermore, there was larger activation in premotor areas of females which might indicate that women relate visual speech image to motor speech production. This is in accordance with the motor theory of speech perception [6,10]. Besides premotor cortex, Ruytjens et al. [14] also found different activation related to gender in angular and inferior parietal areas (i.e., women showed larger and more intense activation when compared to men) indicating that women present a more diffuse processing while men present a more focal processing related to audio-visual stimuli related to language, also in line with EMB theory. Since the ASD children of Oberman, Ramachandran [11] study performed well on the Matrices subtest of the WAIS – that requires individuals to provide a missing piece to a pattern composed of shapes – a single modality stimuli, it is plausible to consider that the findings of impairment performance on the bouba-kiki task provides evidence of a relative dysfunction in cross-modal processing. Considering the STS and Broca’s area are regions of MNS also hypothesized to be involved in the Bouba-Kiki task and that those areas may play a role in the language deficits seen in children with ASD [11] along with the findings of Ruytjens [14] showing an activation pattern of premotor areas only in females, it is conceivable to conclude that in our task the activity of this area in females compensates for the disruption that occurs in male during cathodal tDCS. Interestingly, bilateral tDCS over the same area did not induce gender differential effects on a unimodal auditory task [5]. Thus, our gender-specific effect of cathodal tDCS together with findings from neuroimaging studies indicated that for some specific multisensory processing men recruited a more restricted neural network focused at the auditory temporal areas while women had a widespread network activation including areas such as the premotor cortex. A further interesting follow-up experiment would be to test the effects of tDCS over premotor areas. Also, differential gender effects of tDCS needs to be further investigated as other factors may explain these gender differences such as skull thickness between males and females. In addition, the observed effect is due to a 3-interaction effect with a small effect size. Therefore, further studies should explore the impact of tDCS on gender during multisensory tasks. We also found an interaction effect for gender and congruence during go trials. Both men and women were worse on incongruent go trials as compared to the congruent ones. In addition, women had worse performance on incongruent go trials as compared to men in the same condition. This effect was not expected given the different pattern of brain activation suggesting that females had more efficient connections between the visual to the motor aspects of speech production as compared to males [14] and previous findings showing better performance of women on verbal tasks [4]. However, the poorer accuracy on incongruent trials by women is not due to a failure to discriminate congruent and incongruent pairs as both men and women had to achieve more than 80% accuracy to participate on this experiment. This gender effect at baseline might be due to attentional deficits. We also found an interaction effect of tDCS and congruency showing that active tDCS (both anodal and cathodal) interferes with reaction time for incongruent pairs but not for congruent
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
ones. The lack of effect on congruent trials might be due to a ceiling effect on this condition. Nevertheless, there was a significant effect of tDCS on reaction time during incongruent trials; i.e., we found faster responses on incongruent trials during both anodal and cathodal tDCS (thus a lack of polarity effects) as compared to sham. Polarity-specific effects of tDCS are usually reported for several cognitive domains [5,15] however, some previous studies have also reported similar effects for cathodal and anodal tDCS. Boggio et al. [3] found significant decrease of alcohol craving for both anodal left/cathodal right and anodal right/cathodal left dorsolateral prefrontal cortex stimulation as compared to sham stimulation. The authors argue that both active stimulation conditions lead to interference in the pathological imbalance between right and left DLPFC activity that might be necessary for craving states. Also tDCS might have modulated other cortical and subcortical areas associated with processing of motivational behavior and the reward system. Medonc¸a et al. [9] found pain reduction (measured by visual analog scale) in fibromyalgia patients in both cathodal and anodal supraorbital tDCS. The authors argue that both excitatory and inhibitory tDCS might result at the same effect since multiple neural networks are simultaneously activated. Here similarly the effect of both polarities may be the result of an unspecific or widespread tDCS effect on reaction time for the incongruent trials. Given our initial findings suggesting an interaction between tDCS, gender and task on a multisensory task, it would be interesting to explore further other tasks of multisensory integration and also combine in the same study neurophysiological measurements with techniques of non-invasive brain stimulation to understand further gender differences on multisensory processing. Acknowledgments PSB is supported by a CNPq researcher grant (305718/ 2009-6). OML was supported by a Master grant (CAPESPROSUP–IES modality I).
109
References [1] S. Baron-Cohen, The extreme male brain theory of autism, Trends in Cognitive Sciences 6 (2002) 248–254. [2] P.S. Boggio, R.R. Rocha, M.T. Silva, F. Fregni, Differential modulatory effects of transcranial direct current stimulation on a facial expression go-no-go task in males and females, Neuroscience Letters 447 (2008) 101–105. [3] P.S. Boggio, N. Sultani, S. Fecteau, L. Merabet, T. Mecca, A. Pascual-Leone, A. Basaglia, F. Fregni, Prefrontal cortex modulation using transcranial DC stimulation reduces alcohol craving: a double-blind, sham-controlled study, Drug and Alcohol Dependence 92 (2008) 55–60. [4] R. Halari, T. Sharma, M. Hines, C. Andrew, A. Simmons, V. Kumari, Comparable fMRI activity with differential behavioural performance on mental rotation and overt verbal fluency tasks in healthy men and women, Experimental Brain Research 169 (2006) 1–14. [5] A. Ladeira, F. Fregni, C. Campanhã, C.A. Valasek, D. de Ridder, A.R. Brunoni, P.S. Boggio, Polarity-dependent transcranial direct current stimulation effects on central auditory processing, PLoS One 6 (2011) (art. no. e25399). [6] A.M. Liberman, I.G. Mattingly, The motor theory of speech perception revised, Cognition 21 (1985) 1–26. [7] E. Macaluso, J. Driver, Multisensory spatial interactions: a window onto functional integration in the human brain, Trends in Neurosciences 28 (2005) 264–271. [8] D. Maurer, T. Pathman, C.J. Mondloch, The shape of boubas: sound–shape correspondences in toddlers and adults, Developmental Science 9 (2006) 316–322. [9] M.E. Mendonc¸a, M.B. Santana, A.F. Baptista, A. Datta, M. Bikson, F. Fregni, C.P. Araújo, Transcranial DC Stimulation in fibromyalgia: optimized cortical target supported by high-resolution computational models, Journal of Pain 12 (2011) 610–617. [10] R. Mottonen, J. Jarvelainen, M. Sams, R. Hari, Viewing speech modulates activity in the left SI mouth cortex, Neuroimage 24 (2005) 731–737. [11] L.M. Oberman, V.S. Ramachandran, Preliminary evidence for deficits in multisensory integration in autism spectrum disorders: the mirror neuron hypothesis, Society for Neuroscience 3 (2008) 348–355. [12] R.G. O’Connell, P.M. Dockree, M.A. Bellgrove, A. Turin, S. Ward, J.J. Foxe, I.H. Robertson, Two types of action error: electrophysiological evidence for separable inhibitory and sustained attention neural mechanisms producing error on go/no-go tasks, Journal of Cognitive Neuroscience 21 (2009) 93–104. [13] F. Pulvermuller, Brain mechanisms linking language and action, Nature Reviews 6 (2005) 576–582. [14] L. Ruytjens, F. Albers, P. van Dijk, H. Wit, A. Willemsen, Neural responses to silent lipreading in normal hearing male and female subjects, European Journal of Neuroscience 24 (2006) 1835–1844. [15] C.J. Stagg, J. O’Shea, Z.T. Kincses, M. Woolrich, P.M. Matthews, H. Johansen-Berg, Modulation of movement-associated cortical activation by transcranial direct current stimulation, European Journal of Neuroscience 30 (2009) 1412–1423.
Contents lists available at SciVerse ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Bilateral temporal cortex transcranial direct current stimulation worsens male performance in a multisensory integration task Olivia Morgan Lapenta a , Felipe Fregni b,c , Lindsay M. Oberman c , Paulo Sergio Boggio a,∗ a
Social and Cognitive Neuroscience Laboratory and Developmental Disorders Program, Center for Health and Biological Sciences, Mackenzie Presbyterian University, Sao Paulo, Brazil Laboratory of Neuromodulation, Department of Physical Medicine and Rehabilitation, Spaulding Rehabilitation Hospital and Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA c Berenson-Allen Center for Non-invasive Brain Stimulation, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA b
h i g h l i g h t s tDCS induces a polarity, gender and task specific effect in a multisensory integration task. Bilateral cathodal temporal cortex tDCS worsens task accuracy in males only. Bilateral anodal or cathodal temporal cortex tDCS results in decreased reaction time for both genders.
a r t i c l e
i n f o
Article history: Received 4 April 2012 Received in revised form 29 May 2012 Accepted 10 August 2012 Keywords: Multisensory integration tDCS Superior temporal sulcus Gender Extreme male brain theory
a b s t r a c t Somatosensory integration is a critical cognitive function for human social interaction. Though somatosensory integration has been highly explored in cognitive studies; only a few studies have explored focal modulation of cortical excitability using a speech perception paradigm. In the current study, we aimed to investigate the effects of tDCS applied over the temporal cortex of healthy subjects during a go-no-go task in which stimuli were shapes and non-words. Twenty-eight subjects were randomized to receive cathodal, anodal or sham tDCS bilaterally over the superior temporal cortex (the reference electrode was on deltoid) in a counterbalanced order. The effects on judgment of congruency between shapes and non-words in healthy volunteers were measured by a go-no-go task. Our findings show a significant modification of performance according to the polarity of stimulation, task and subject gender. We found that men performed worse on the no-go condition for congruent stimuli during cathodal tDCS. For reaction time, on the other hand, there was a similar effect for anodal and cathodal stimulation. There were significantly faster responses on incongruent trials during both anodal and cathodal tDCS. Along with previous literature showing gender differences in tasks associated with speech perception, the findings of this study provide additional evidence suggesting that men may have a more focal and restricted neural processing in this multisensory integration task. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Somatosensory integration is a critical component of cognitive processing. It is characterized by dual processing that integrates different sensory inputs such as visual and auditory stimuli in order to provide an adequate response. As these different signals can arise from common external events or objects, the integration between them can be useful [7]. It is a particularly important stage
∗ Corresponding author at: Social and Cognitive Neuroscience Laboratory and Developmental Disorders Program, Center for Health and Biological Sciences, Mackenzie Presbyterian University, Rua Piaui, 181, 10◦ andar, Sao Paulo 01241-001, SP, Brazil. Tel.: +55 11 2114 8001; fax: +55 11 2114 8563. E-mail address: [email protected] (P.S. Boggio). 0304-3940/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neulet.2012.08.076
of processing as most of our cognitive computations involve simultaneous processing of more than one type of sensory stimulation. Theoretical perspectives suggest that brain functions can present interactive distribution of functional systems, e.g. hearing a word is associated with the activation of articulatory motor areas. Thus, there are neural networks combining processing from different inputs. Additionally, inputs from a different modality can activate other areas through processes of association and integration [13]. The motor theory of speech perception suggests that the objects of perception of speech are the “phonetic gestures”, represented in the observer’s brain as a motor command of the signal, characterized by movements of the mouth, lips and tongue [6]. Specific neural networks have been described for different modalities of multisensory integration. For example, specific
106
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
regions of the cortex, including Broca’s area and superior temporal sulcus (STS) are likely involved in the integration of shapes and non-words. These regions may underlie the shared representation of the auditory perception of a sound and the motor representation of mouth and larynx movement needed to make the sound [11]. Neuropsychological studies have shown the relationship between motor action and speech production using cognitive tasks of association between images and written non-words where these images have been produced according to the phonetic gestures used to produce the sound of the corresponding non-word, e.g. sounds that need rounded movements of the mouth to be produced have rounded shapes to represent them [8,11]. We therefore hypothesize that a technique – transcranial direct current stimulation (tDCS) – that can alter cognitive processing might be useful to alter performance on a multisensory task. We chose a multisensory task integrating shapes and non-words as we were interested in exploring the effects of tDCS over temporal cortex in speech perception. We expect enhanced performance with anodal tDCS and a decreased performance for cathodal tDCS. Also, based on our preliminary findings, we expected to find a gender effect since language and some tasks associated with crossmodal processing differ in men and women.
2. Materials and methods We investigated the effects of anodal, cathodal and sham tDCS applied bilaterally over the STS in twenty eight volunteers (14 female and 14 male mean age of 23.2 ± 3.1). All subjects had a similar level of education. This sample is similar to the samples we have been testing in previous cognitive tDCS studies. Subjects were regarded as suitable to participate in this study if they fulfilled the following criteria: (1) age between 20 and 30 years; (2) no clinically significant or unstable medical, or neuropsychiatric disorder; (3) no use of central nervous systemeffective medication; (4) no history of brain surgery, tumor, or intracranial metal implantation. All study participants provided written, informed consent. This study was approved by the institutional ethics committee from Mackenzie Presbyterian University, Brazil. Before starting the experiment, all the participants had to complete successfully a pre-test consisting of visual presentation of pairs of images/non-words. Stimuli were provided from a previous study where neurotypical individuals corrected guessed the matching as expected, at a rate of 88% of trials [11]. Thus, before starting the experiment, all the participants had to complete successfully a pre-test consisting of visual presentation of pairs of images/nonwords. Subjects were included in this study if they assigned the correct correspondence between images and non-words with a minimum rate of 80%. All subjects received in different days with an interval between sessions of at least 48 h, sham, anodal and cathodal temporal cortex tDCS. In order to prevent learning effects, the tDCS sessions were randomized using the Latin Square method. As the number of subjects was not a multiple of 3, the final distribution was not fully counterbalanced. Two pairs of surface sponge electrodes (35 cm2 ) were soaked in saline solution and applied to the scalp at the desired areas of stimulation and to the right deltoid muscle as the reference electrode. We used this bi-temporal extracephalic montage before in other studies and obtained significant behavioral changes [5]. Rubber bandages were used to hold the electrodes in place for the duration of stimulation. For anodal stimulation anodal electrodes were placed over T3 and T4 (according to the 10–20 system for EEG electrode placement) and cathodal electrodes on the right arm; for cathodal stimulation the cathodal electrodes were
placed over T3 and T4 and anodal electrodes on the right arm; for sham stimulation we adopted the same position as for active stimulation, however, subjects only received stimulation in this condition for 20 s. Participants in the active conditions received a constant current of 1 mA for 14 min (5 min of tDCS only followed by 9 min of tDCS during the go-no-go task of congruency between shapes and non-words). The task is composed of 10 images, each having a corresponding written non-word. We created 10 stimuli which were 10 congruent and 10 incongruent pairs of images/non-word randomly presented. The shapes and written non-words were presented visually at the screen, side by side. The experiment was composed of 6 blocks of 40 trials. In half of blocks, subjects were instructed to press a button when the stimulus was congruent and in the other half when it was incongruent (Fig. 1). The blocks were presented alternately. At the first testing day, all subjects performed a training test composed by one go-congruent and one nogo-congruent blocks, each one containing 12 trials. Therefore most of the learning effect was achieved during this phase. Analyses were done with Statistica software (version 8.0, StatSoft Inc.). We performed repeated measures ANOVAs in which the dependent variables were number of correct responses on Go trials, number of correct responses on NoGo trials, and reaction time (for correct responses on Go trials) and the independent variables were condition of stimulation (anodal, cathodal or sham), gender (male or female), and congruency (congruent or incongruent stimuli). Despite previous tDCS go-nogo studies have performed analyses not separating go and nogo conditions, we performed separated analyses for both conditions as recent electrophysiological data suggest that the neural network associated with failure to press following go stimulus is different than that associated with pressing following nogo stimulus [12]. We also considered the interaction of gender vs. tDCS, congruency vs. tDCS, congruency vs. gender, and tDCS vs. gender vs. congruency. When appropriate, post hoc comparisons were carried out using Fisher’s LSD. Unless stated otherwise, all results are presented as means, confidence intervals, and standard errors. Statistical significance refers to a p value <0.05.
3. Results All subjects completed the entire experiment. All subjects tolerated the stimulation well and no side effects were reported. Also, bilateral stimulation was not associated with additional discomfort in any subject. Initially, we performed a repeated measures ANOVA in which the dependent variable was correct responses on Go trials (‘pressing a button’ responses). The ANOVA did not reveal significant effects for gender (F1,26 = 3.8; p = 0.1), tDCS (F2,52 = 0.6; p = 0.6), gender * tDCS (F2,52 = 0.2; p = 0.8), congruency * tDCS (F2,52 = 0.5; p = 0.6) or congruency * gender * tDCS (F2.52 = 0.9; p = 0.4). However, the ANOVA did reveal significant effects for congruency (F1,26 = 35.7; p = 0.000003, p 2 = 0.58) and for the interaction term congruency * gender (F1.26 = 7.5; p = 0.01, p 2 = 0.22. These significant effects, as it can be observed in Fig. 2, were due to a better performance of both genders on congruent trials as compared to incongruent ones. With regard to the interaction congruency * gender, Fischer LSD showed significant differences between congruent and incongruent correct responses for women (p = 0.000001) and for men (p = 0.04). Post hoc comparisons also revealed significant difference between women and men accuracy for incongruent trials (p = 0.03), but not for congruent trials (p = 0.7). As it can be seen in Fig. 2, women had poorer performance on incongruent trials as compared to men, but both presented similar performance for congruent trials.
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
107
Fig. 1. Diagram of the task. Example of go-congruent and go-incongruent blocks, respectively.
We then performed a repeated measures ANOVA in which the dependent variable was correct responses on NoGo trials (‘notpressing a button’ responses). The ANOVA did not reveal significant effects for Gender (F1,26 = 0.001; p = 1.0), tDCS (F2,52 = 0.3; p = 0.8), gender * tDCS (F2,52 = 0.1; p = 0.9), congruency (F1,26 = 0.1; p = 0.7), congruency * gender (F1.26 = 2.5; p = 0.12) or congruency * tDCS (F2,52 = 0.2; p = 0.8). However, the ANOVA did reveal a significant effect for the interaction congruency * gender * tDCS (F2.52 = 3.3; p = 0.047, p 2 = 0.11). With regard to this interaction, Fischer LSD showed significantly poorer performance of men during cathodal tDCS on congruent trials as compared to responses on NoGo congruent trials of men during anodal tDCS (p = 0.012), responses for NoGo incongruent trials of men during anodal tDCS (p = 0.035), responses for NoGo incongruent trials of men during cathodal tDCS (p = 0.012), responses for NoGo congruent trials of men during sham tDCS (p = 0.048) and responses for NoGo incongruent trials of men during sham tDCS (p = 0.012) (Fig. 3). There were no significant effects in the other comparisons (p > 0.05).
Fig. 2. Accuracy for congruent and incongruent conditions in Go trials. (*) Women presented poorer accuracy in the incongruent go trials when compared to congruent go trials (p = 0.000001) and when compared to men on incongruent go trials (p = 0.003). (**) Men presented poorer accuracy in incongruent go trials when compared to congruent go trials (p = 0.04). Both gender groups had similar performance in the congruent condition (p = 0.7).
Finally, we performed a repeated measures ANOVA in which the dependent variable was response time on Go trials (‘pressing a button’ responses). Response time was only considered for correct responses. The ANOVA did not reveal significant effects for gender (F1,26 = 0.1; p = 0.7), tDCS (F2,52 = 0.3; p = 0.7), gender * tDCS (F2,52 = 1.0; p = 0.4), congruency * gender (F1.26 = 3.4; p = 0.08), or congruency * gender * tDCS (F2.52 = 1.8; p = 0.2). However, the ANOVA revealed a significant effect for congruency (F1,26 = 442.9; p < 0.000001, p 2 = 0.9) and the interaction of Congruency and tDCS (F2,52 = 4.6; p = 0.01, p 2 = 0.15). With regard to the congruency * tDCS interaction, Fischer LSD showed significant differences between response time for congruent trials vs. incongruent trials (p < 0.000001 for all comparisons). In addition, it was found that response time on incongruent trials were significantly reduced for anodal and cathodal tDCS as compared to sham tDCS (p = 0.003 and p = 0.004, respectively) (see Fig. 4).
Fig. 3. Accuracy on NoGo trials considering type of stimulation, congruency and gender. Men showed reduced performance in NoGo congruent trials when receiving cathodal tDCS when compared to all other conditions (NoGo congruent trials of men during anodal tDCS (p = 0.012), NoGo incongruent trials of men during anodal tDCS (p = 0.035), NoGo incongruent trials of men during cathodal tDCS (p = 0.012), NoGo congruent trials of men during sham tDCS (p = 0.048) and for NoGo incongruent trials of men during sham tDCS (p = 0.012).
108
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
Fig. 4. (**) Reaction time differed for congruent and incongruent trials (p = 0.000001). (*) Cathodal and anodal stimulation differed from sham stimulation during incongruent trials (p = 0.003 and p = 0.004, respectively).
4. Discussion Our findings show that tDCS is a useful tool to investigate multimodal integration as it induces a significant modification of performance according to the polarity of stimulation and subject gender. In this experiment the effects were gender-specific. We found that men performed worse on the no-go condition for congruent stimuli during cathodal tDCS. Accuracy on the no-go condition was indexed by the number of correct responses (notpressing a button in response to specific stimuli). We found that men pressed the button more times when faced with congruent stimuli during the block that they were asked to respond to incongruent ones. Therefore, they incorrectly indicated that the stimulus was congruent when it was incongruent when stimulated with cathodal tDCS. This suggests a performance worsening induced by cathodal tDCS on a specific multisensory integration task. We applied cathodal electrodes over both right and left superior temporal cortex, well-know regions of multisensory integration. Previous findings have already shown enhanced recruitment of this area during presentation of congruent stimuli as compared to incongruent one. Given these findings showing cortical recruitment during congruent stimuli, it is plausible to consider that the effect on congruent stimuli might be the result of the inhibitory effect of bilateral cathodal tDCS of this area in both hemispheres. Nevertheless, it is interesting that this effect was observed for men but not for women. A similar gender effect, but with a different tDCS montage, was observed earlier by our group. Boggio et al. [2] found that temporal cortex tDCS with an anodal electrode over T3 and cathodal over T4 lead to differential effects according to gender. In this study, authors found that men presented more errors on a facial expression go-no-go task as compared to sham. Both findings give evidence that the functionality of the temporal lobe and its neural network distribution are dependent on gender. The gender differential effect found here might be related to the type of task and the neural network underlying its processing. This finding is in accordance with the extreme male brain (EMB) theory that argues that ASD individuals have an exaggerated typically male brain, i.e. an extreme systemizing pattern related to more focal and restrict networks in males as opposed to larger and more distributed networks in females [1]. The findings of Oberman and Ramachandran [11] where ASD children were less likely to choose the correct non-word for the correspondent shape along with Boggio et al. [2] results of different tDCS effects according to gender are in line to the EMB theory. Therefore is possible that for multisensory
integration, male and female networks differ and given EBM theory in which males may have activation in more focal and restrict network, the inhibition of superior temporal cortex (an area that seems to play a role in the language deficits seen in children with ASD) by cathodal tDCS could induce an effect in males only. Furthermore, different activations of male and female brain areas have been shown before. Ruytjens et al. [14] using fMRI showed different activations between male and female during silent lipreading. Women activated key cortical areas known to be involved in lipreading in both hemispheres while men activated those areas only in the left hemisphere, except for the associative auditory temporal area. Furthermore, there was larger activation in premotor areas of females which might indicate that women relate visual speech image to motor speech production. This is in accordance with the motor theory of speech perception [6,10]. Besides premotor cortex, Ruytjens et al. [14] also found different activation related to gender in angular and inferior parietal areas (i.e., women showed larger and more intense activation when compared to men) indicating that women present a more diffuse processing while men present a more focal processing related to audio-visual stimuli related to language, also in line with EMB theory. Since the ASD children of Oberman, Ramachandran [11] study performed well on the Matrices subtest of the WAIS – that requires individuals to provide a missing piece to a pattern composed of shapes – a single modality stimuli, it is plausible to consider that the findings of impairment performance on the bouba-kiki task provides evidence of a relative dysfunction in cross-modal processing. Considering the STS and Broca’s area are regions of MNS also hypothesized to be involved in the Bouba-Kiki task and that those areas may play a role in the language deficits seen in children with ASD [11] along with the findings of Ruytjens [14] showing an activation pattern of premotor areas only in females, it is conceivable to conclude that in our task the activity of this area in females compensates for the disruption that occurs in male during cathodal tDCS. Interestingly, bilateral tDCS over the same area did not induce gender differential effects on a unimodal auditory task [5]. Thus, our gender-specific effect of cathodal tDCS together with findings from neuroimaging studies indicated that for some specific multisensory processing men recruited a more restricted neural network focused at the auditory temporal areas while women had a widespread network activation including areas such as the premotor cortex. A further interesting follow-up experiment would be to test the effects of tDCS over premotor areas. Also, differential gender effects of tDCS needs to be further investigated as other factors may explain these gender differences such as skull thickness between males and females. In addition, the observed effect is due to a 3-interaction effect with a small effect size. Therefore, further studies should explore the impact of tDCS on gender during multisensory tasks. We also found an interaction effect for gender and congruence during go trials. Both men and women were worse on incongruent go trials as compared to the congruent ones. In addition, women had worse performance on incongruent go trials as compared to men in the same condition. This effect was not expected given the different pattern of brain activation suggesting that females had more efficient connections between the visual to the motor aspects of speech production as compared to males [14] and previous findings showing better performance of women on verbal tasks [4]. However, the poorer accuracy on incongruent trials by women is not due to a failure to discriminate congruent and incongruent pairs as both men and women had to achieve more than 80% accuracy to participate on this experiment. This gender effect at baseline might be due to attentional deficits. We also found an interaction effect of tDCS and congruency showing that active tDCS (both anodal and cathodal) interferes with reaction time for incongruent pairs but not for congruent
O.M. Lapenta et al. / Neuroscience Letters 527 (2012) 105–109
ones. The lack of effect on congruent trials might be due to a ceiling effect on this condition. Nevertheless, there was a significant effect of tDCS on reaction time during incongruent trials; i.e., we found faster responses on incongruent trials during both anodal and cathodal tDCS (thus a lack of polarity effects) as compared to sham. Polarity-specific effects of tDCS are usually reported for several cognitive domains [5,15] however, some previous studies have also reported similar effects for cathodal and anodal tDCS. Boggio et al. [3] found significant decrease of alcohol craving for both anodal left/cathodal right and anodal right/cathodal left dorsolateral prefrontal cortex stimulation as compared to sham stimulation. The authors argue that both active stimulation conditions lead to interference in the pathological imbalance between right and left DLPFC activity that might be necessary for craving states. Also tDCS might have modulated other cortical and subcortical areas associated with processing of motivational behavior and the reward system. Medonc¸a et al. [9] found pain reduction (measured by visual analog scale) in fibromyalgia patients in both cathodal and anodal supraorbital tDCS. The authors argue that both excitatory and inhibitory tDCS might result at the same effect since multiple neural networks are simultaneously activated. Here similarly the effect of both polarities may be the result of an unspecific or widespread tDCS effect on reaction time for the incongruent trials. Given our initial findings suggesting an interaction between tDCS, gender and task on a multisensory task, it would be interesting to explore further other tasks of multisensory integration and also combine in the same study neurophysiological measurements with techniques of non-invasive brain stimulation to understand further gender differences on multisensory processing. Acknowledgments PSB is supported by a CNPq researcher grant (305718/ 2009-6). OML was supported by a Master grant (CAPESPROSUP–IES modality I).
109
References [1] S. Baron-Cohen, The extreme male brain theory of autism, Trends in Cognitive Sciences 6 (2002) 248–254. [2] P.S. Boggio, R.R. Rocha, M.T. Silva, F. Fregni, Differential modulatory effects of transcranial direct current stimulation on a facial expression go-no-go task in males and females, Neuroscience Letters 447 (2008) 101–105. [3] P.S. Boggio, N. Sultani, S. Fecteau, L. Merabet, T. Mecca, A. Pascual-Leone, A. Basaglia, F. Fregni, Prefrontal cortex modulation using transcranial DC stimulation reduces alcohol craving: a double-blind, sham-controlled study, Drug and Alcohol Dependence 92 (2008) 55–60. [4] R. Halari, T. Sharma, M. Hines, C. Andrew, A. Simmons, V. Kumari, Comparable fMRI activity with differential behavioural performance on mental rotation and overt verbal fluency tasks in healthy men and women, Experimental Brain Research 169 (2006) 1–14. [5] A. Ladeira, F. Fregni, C. Campanhã, C.A. Valasek, D. de Ridder, A.R. Brunoni, P.S. Boggio, Polarity-dependent transcranial direct current stimulation effects on central auditory processing, PLoS One 6 (2011) (art. no. e25399). [6] A.M. Liberman, I.G. Mattingly, The motor theory of speech perception revised, Cognition 21 (1985) 1–26. [7] E. Macaluso, J. Driver, Multisensory spatial interactions: a window onto functional integration in the human brain, Trends in Neurosciences 28 (2005) 264–271. [8] D. Maurer, T. Pathman, C.J. Mondloch, The shape of boubas: sound–shape correspondences in toddlers and adults, Developmental Science 9 (2006) 316–322. [9] M.E. Mendonc¸a, M.B. Santana, A.F. Baptista, A. Datta, M. Bikson, F. Fregni, C.P. Araújo, Transcranial DC Stimulation in fibromyalgia: optimized cortical target supported by high-resolution computational models, Journal of Pain 12 (2011) 610–617. [10] R. Mottonen, J. Jarvelainen, M. Sams, R. Hari, Viewing speech modulates activity in the left SI mouth cortex, Neuroimage 24 (2005) 731–737. [11] L.M. Oberman, V.S. Ramachandran, Preliminary evidence for deficits in multisensory integration in autism spectrum disorders: the mirror neuron hypothesis, Society for Neuroscience 3 (2008) 348–355. [12] R.G. O’Connell, P.M. Dockree, M.A. Bellgrove, A. Turin, S. Ward, J.J. Foxe, I.H. Robertson, Two types of action error: electrophysiological evidence for separable inhibitory and sustained attention neural mechanisms producing error on go/no-go tasks, Journal of Cognitive Neuroscience 21 (2009) 93–104. [13] F. Pulvermuller, Brain mechanisms linking language and action, Nature Reviews 6 (2005) 576–582. [14] L. Ruytjens, F. Albers, P. van Dijk, H. Wit, A. Willemsen, Neural responses to silent lipreading in normal hearing male and female subjects, European Journal of Neuroscience 24 (2006) 1835–1844. [15] C.J. Stagg, J. O’Shea, Z.T. Kincses, M. Woolrich, P.M. Matthews, H. Johansen-Berg, Modulation of movement-associated cortical activation by transcranial direct current stimulation, European Journal of Neuroscience 30 (2009) 1412–1423.