Available online at www.sciencedirect.com

ysiology 68 (2008) 41–50www.elsevier.com/locate/ijpsycho

International Journal of Psychoph

Gender differences in lateralization of mismatch negativity indichotic listening tasks

Satoru Ikezawa a,d,⁎, Kazuyuki Nakagome a,d, Masaru Mimura a, Junko Shinoda a,Kenji Itoh b, Ikuo Homma c, Kunitoshi Kamijima a

a Department of Psychiatry, Showa University School of Medicine, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8666, Japanb Department of Speech and Cognitive Science, Graduate School of Medicine, University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan

c Second Department of Physiology, Showa University School of Medicine, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8666, Japand Division of Neuropsychiatry, Department of Multidisciplinary Internal Medicine, Tottori University, 36-1 Nishi-Machi, Yonago, Tottori 683-8504, Japan

Received 18 July 2007; received in revised form 12 December 2007; accepted 8 January 2008Available online 26 January 2008

Abstract

Objective: With the aim of investigating gender differences in the functional lateralization subserving preattentive processing of language stimuli,we compared auditory mismatch negativities (MMNs) using dichotic listening tasks.Methods: Forty-four healthy volunteers, including 23 males and 21 females, participated in the study. MMNs generated by pure-tone and phoneticstimuli were compared, to check for the existence of language-specific gender differences in lateralization. Both EEG amplitude and scalp currentdensity (SCD) data were analyzed.Results: With phonetic MMNs, EEG findings revealed significantly larger amplitude in females than males, especially in the right hemisphere,while SCD findings revealed left hemisphere dominance and contralateral dominance in males alone. With pure-tone MMNs, no significantgender differences in hemispheric lateralization appeared in either EEG or SCD findings.Conclusion: While males exhibited left-lateralized activation with phonetic MMNs, females exhibited more bilateral activity. Further, thecontralateral dominance of the SCD distribution associated with the ear receiving deviant stimuli in males indicated that ipsilateral input as well asinterhemispheric transfer across the corpus callosum to the ipsilateral side was more suppressed in males than in females.Significance: The findings of the present study suggest that functional lateralization subserving preattentive detection of phonetic change differsbetween the genders. These results underscore the significance of considering the gender differences in the study of MMN, especially whenphonetic stimulus is adopted. Moreover, they support the view of Voyer and Flight [Voyer, D., Flight, J., 2001. Gender differences in laterality on adichotic task: the influence of report strategies. Cortex 37, 345–362.] in that the gender difference in hemispheric lateralization of languagefunction is observed in a well-managed-attention condition, which fits the condition adopted in the MMN measurement; subjects are required tofocus attention to a distraction task and thereby ignore the phonetic stimuli that elicit MMN.© 2008 Elsevier B.V. All rights reserved.

Keywords: Event-related potential (ERP); Mismatch negativity (MMN); Scalp current density (SCD); Dichotic listening task (DLT); Gender; Phoneme

1. Introduction

No consensus regarding hemispheric lateralization of lan-guage functions and of gender differences in language later-

⁎ Corresponding author. Division of Neuropsychiatry, Department of Multi-disciplinary Internal Medicine, Tottori University, 36-1 Nishi-Machi, Yonago,Tottori 683-8504, Japan. Tel.: +81 859 38-6547; fax: +81 859 38 6549.

E-mail address: satoruikezawa@grape.med.tottori-u.ac.jp (S. Ikezawa).

0167-8760/$ - see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.ijpsycho.2008.01.006

alization has yet been achieved. Although it was long held thatleft-hemisphere dominance exists in all aspects of languageprocessing, it has now come to be widely accepted that the righthemisphere is also involved in certain aspects of languageprocessing. It has been found, for example, that prosody ispredominantly processed in the right hemisphere rather than theleft (Mitchell et al., 2003). Neurophysiological and psycholin-guistic studies have produced results making it plausible that theleft hemisphere processes semantically close associations

42 S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

rapidly, while the right hemisphere activates semantically re-mote associations over a longer period of time (Chiarello andRichards, 1992; Abdullaev and Posner, 1998). This suggeststhat the left hemisphere is superior in processing of lexical-semantic information, while the right hemisphere is active in theuse of contextual information to integrate items of informationdistributed over larger language units.

Further, initial studies suggested that the cerebral substratesubserving language functions more bilaterally in the female thanin the male brain (Dorion et al., 2000; Gur et al., 2000). However,more recent neuroimaging studies have yielded inconsistentfindings on gender differences in cortical language representation.Sommer et al. (2004) performed a meta-analysis of studies thatassessed language activity with functional imaging in healthymenand women, and concluded that the putative gender difference inlanguage lateralization may be absent at the population level.

Findings obtained using dichotic listening tasks (DLTs),which provide a noninvasive method for assessing cerebraldominance in language functions, are also inconsistent. DLTstudies compare performance in identifying two different syl-lables simultaneously presented, one to each ear. If the normalperceptual dominance of the right ear, the right ear advantage(REA), is reduced, decreased left cerebral dominance can bepresumed. However, reduced perceptual asymmetry in detectinglanguage stimuli cannot clearly be attributed to decreased cere-bral dominance, since it may reflect non-language-specific fac-tors, such as attention and/or inhibition.

The discrepancies in findings in this field of investigationmight, at least in part, be due to the different types of languageprocessing tasks employed in the studies concerned. Genderdifferences in language lateralization may vary depending on theprocessors involved, for example, whether phonological, syntac-tic or semantic, productive or receptive, conscious or preattentive;however, few studies have attempted to separate out these variousprocessing elements. When Frost et al. (1999) failed to find agender effect on brain activation during a language comprehen-sion task using a large number of subjects, the authors themselvesnoted that the language and control tasks employed differed inmany respects, and that the results were therefore likely to repre-sent the combined activation of several language-specific and-nonspecific component processors. As regards non-language-specific processors, Voyer and Flight (2001) examined genderdifferences inDLT performance using various reporting strategiesthat differed in the attention management levels required, andobtained findings of a gender difference in language lateralizationonly for well-managed-attention tasks.

In the present study, to identify and compare the processingelements involved in gender differences in language lateraliza-tion, we focused specifically on phonological preattentive pro-cessing using mismatch negativity (MMN), a component of theauditory event-related potential (ERP). MMN is widely used asan index of preattentive auditory processing, reflecting auto-matic context-dependent information processing and auditorysensory memory, and it has been suggested that it is generated inthe primary and secondary auditory cortices (Sabri et al., 2004;Liebenthal et al., 2003; Müller et al., 2002; Opitz et al., 1999,2002; Downar et al., 2001; Celsis et al., 1999; Jemel et al., 2002;

Levänen et al., 1996; Alho, 1995; Scherg et al., 1989; Hari et al.,1984; Javitt et al., 1994) and in frontal cortex (Sabri et al., 2004;Liebenthal et al., 2003; Müller et al., 2002; Opitz et al., 2002;Downar et al., 2001, 2002; Jemel et al., 2002; Yago et al., 2001;Rinne et al., 2000; Levänen et al., 1996; Giard et al., 1990).Näätänen and Michie (1979) suggested a functional dissociationbetween the two generators, with the temporal generator asso-ciated with establishment of memory traces and comparisonwith incoming stimulus attributes and the frontal generatorrelated to involuntary triggering of attention invoked by thedetected change. Although a few studies have provided somesupport for this view (Sabri et al., 2004, 2006; Müller et al.,2002; Shalgi and Deouell, 2007), direct evidence for the sug-gested division of function between the superior temporal andfrontal cortices is still lacking.

Previous studies (Näätänen et al., 1997; Alho et al., 1998;Rinne et al., 1999; Koyama et al., 2000) have demonstrated that,in normal subjects, the left auditory cortex is predominantlyinvolved in preattentive categorical perception of speech sounds,but not all studies have (Kasai et al., 2003; Mathiak et al., 2000).Mathiak et al. (2000), using whole-head MEG, demonstratedthat preattentive processing of CV syllables failed to exhibitMMF (mismatch field) lateralized effects in the DLT, but thatattention toward target events resulted in left-hemisphericdominance, suggesting bilateral preattentive processing andsubsequent attention-related left-hemispheric- dominant encod-ing of CV syllables at the level of the supratemporal plane.Although the authors suggested that subjects with right earadvantage in behavioral testing tended to exhibit greater func-tional dissociation of the two hemispheres during auditory pro-cessing than those without such advantage, they did not refer togender differences. The few studies that have examined genderdifferences in terms of hemispheric dominance in neural activitysubserving MMN have, thus far, failed to reveal any significantdifferences in laterality between male and female subjects(Aaltonen et al., 1994; Kasai et al., 2003). On the other hand,Hertrich et al. (2002), using spoken and synthetic CV syllables,found that gender difference in MMF laterality emerged onlywith use of synthetic syllables, and that therefore this findingmight reflect gender-specific difference in the processing ofartificial signals rather than speech-related hemispheric dom-inance. However, studies of the gender difference in MMNlaterality during phonological processing have been insufficient.

Since ERP and EEG data have limitations with regard to thequality of spatial information obtained, in order to improvespatial resolution, we calculated SCD and performed SCD map-ping. SCD was calculated by determining the Laplacian of thescalp potential surface generated by spherical spline interpola-tion (Nunez, 1981; Perrin et al., 1989). Since SCD computationis equivalent to applying a high-pass spatial filter to the potentialfields, the maxima and minima of their distribution are clearerthan those of the potential fields. Moreover, since the sensitivityof SCDmapping decreases with the depth of the generator in thebrain, theymay allow the splitting of the overlapped componentsarising from cortex exhibiting distinct sink/source patterns.Formally, this method is based on simplifications of Poisson'sequation relating the field potential at any point in the conductive

Fig. 1. Electrode position map. Diagram of electrode locations projected ontothe 2-D surface used for portraying ERP topography. The EEG was recordedfrom electrode numbers 1–30 and 33–60, with a total of 58 electrodes, asindicated. Circled numbers indicate the electrodes included in the standard 10/20sites. The averaged amplitude data for the electrode sites inside the solid linesrepresent the amplitude data for the left and right frontotemporal regions, whilethe averaged SCD data for the electrode sites inside the curved dashed linerepresent the SCD data for the left and right temporal regions.

43S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

medium to their underlying current generators, so that a localgenerator of extracellular negativity (current sink) correspondsto local neuronal depolarization, while a local generator ofextracellular positivity (current source) corresponds to hyperpo-larization. Although the anatomical and physiological organiza-tion of putative generator regions must be carefully evaluatedbefore attributing physiological correlates to SCD components,it can be reasonably assumed that SCD is useful in elucidatingthe cortical activities associated with the component specified.Several studies using this technique have demonstrated its ad-vantage in elucidating hemispheric lateralization of ERP compo-nents (Tenke et al., 1998; Grimm et al., 2006) and also inproviding support for a dual-generator model of MMN (Deouellet al., 1998; Giard et al., 1990; Shalgi and Deouell, 2007). Giardet al. (1990), using this technique, distinctly exhibited a temporalgenerator which is predominantly activated in the hemispherecontralateral to the ear of stimulation, and a frontal generatorinvolving mainly the right hemisphere.

The present study examined gender differences in hemi-spheric dominance in neural activity subserving preattentiveprocessing of speech sound, and used DLT to elicitMMN, whichis likely to be more sensitive in detecting interhemisphericdifferences in brain activity than the binaural stimulus deliveryused in conventional MMN tasks (O'Leary, 2002). In order todetermine whether the putative lateralization ofMMN is specificto the processing of language-related stimuli, pure-tone MMNswere also investigated. The hypothesis of the present study wasthat the left-hemispheric dominance for the temporal subcom-ponent of phonetic MMN is more evident in male subjects thanin female subjects, while, by contrast, pure-tone MMN does notdiffer between genders.

2. Materials and methods

2.1. Subjects

A total of 23 male and 21 female Japanese subjects par-ticipated in the study. The mean ages of the male and femalesubjects were 28.5 [SD=8.5] and 27.6 [SD=8.7] years, re-spectively, and did not differ significantly. All subjects wereright-handed (Oldfield, 1971). The number of years of educationof males and females did not differ significantly (male: 15.3[SD=1.7]; female: 15.1 [SD=1.6]). None of the subjects hadreceived special education in music or was engaged in music-related work. None had a history of psychiatric or neurologicalillness, or suffered from hearing or visual impairment. After acomplete explanation of the study, written informed consent wasobtained from the subjects. The protocol of this study wasapproved by the Ethics Committee of Showa University.

2.2. Stimuli

The experiment was performed under two conditions. Thefirst condition was designed to generate MMNs in response to apitch change in pure-tone stimuli: standard stimuli werecomprised of a 1 kHz pure tone presented to both ears, whileright (left)-deviant stimuli were comprised of a 1 kHz tone to the

left (right) ear and 2 kHz tone to the right (left) ear. The secondcondition was designed to generate MMNs in response to aphoneme change: standard stimuli were comprised of theJapanese CV syllable da presented to both ears, while right(left)-deviant stimuli were comprised of da to the left (right) earand ba to the right (left) ear. These phoneme stimuli were spokenby a male native Japanese actor, digitized using the NeuroStimsystem (NeuroScan Inc., USA), and edited to a duration of250 ms and a loudness of 75 dBSPL. The frequency spectra ofthe syllables were as follows: /da/, F1=734, F2=1350,F3=2755, F4=3810, F5=4547 Hz; /ba/, F1=734, F2=1209,F3=2738, F4=3827, F5=4670Hz. The duration of the syllable-initial F2 transition was 35 ms for both ba and da. The pure-tonestimuli also lasted 250 ms, at a loudness of 75 dBSPL, witha rise/fall time of 10 ms. The order of presentation of the twoconditions was counterbalanced across the subjects and genders.

2.3. MMN task

Subjects were presented with auditory stimulus sequencesconsisting of randomly delivered standard stimuli (n=960;P=80%) and right and left deviant stimuli (n=120; P=10%each, total n=240; P=20%), except that each deviant stimuluswas preceded by at least one standard stimulus. Stimuli weredelivered dichotichally through headphones at intervals of490 ms. Subjects were instructed to perform a Flanker task as avisual negative priming task, (Neill and Valdes, 1992), and toignore the auditory stimuli.

Fig. 2. Pure-tone MMN waveform. Grand averaged pure-tone MMN by gender and stimulus category (left-ear-deviant, right-ear-deviant).

Table 1Summary of the statistical analyses on pure tone MMN

EEG SCD Latency

df F P df F P df F P

Gender 1,42 0.13 n.s. 1,42 0.72 n.s. 1,42 0.20 n.s.Ear 1,42 2.76 n.s. 1,42 1.89 n.s. 1,42 7.10 b0.05Hemisphere 1,42 1.27 n.s. 1,42 0.48 n.s.Gender⁎ear 1,42 0.32 n.s. 1,42 0.04 n.s. 1,42 2.94 b0.1Gender⁎

hemisphere1,42 0.01 n.s. 1,42 1.08 n.s.

Ear⁎hemisphere 1,42 0.14 n.s. 1,42 1.65 n.s.Gender⁎ear⁎

hemisphere1,42 0.33 n.s. 1,42 0.06 n.s.

44 S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

2.4. ERP recording

EEGs were recorded with a 58-electrode cap (Neuroscan,Inc.) including channels based on the International 10–20system. The tip of the nose was used as the reference for allelectrodes. One electrode was placed at the outer canthus andabove the left eye to monitor eye movements. The sampling ratewas 500 Hz/channel, and the analog filter band-pass was 0.3–60 Hz. The period of analysis was 512 ms, including 64-ms toestablish a prestimulus baseline. The baseline was correctedseparately for each channel in accordance with the meanamplitude of the EEG over the 64-ms prestimulus baselineperiod. Averaging and artifact rejection were performed offline.EEG epochs that contained peak-to-peak amplitudes exceeding75 μVat any site and electrooculograms exceeding 75 μV wereautomatically excluded from averaging. The average wave-forms were obtained separately for deviant and standard stimuli,and digitally filtered with a cutoff frequency of 30 Hz. Thenumber of recorded responses accepted exceeded 60 for allsubjects and did not differ significantly by gender.

2.5. ERP measurement

MMNs were measured using difference waveforms obtainedby subtracting the ERPs of standard stimuli from those ofdeviant stimuli. To determine the latency of the MMN, com-putation of potential field strength (global field power; GFP)was performed for the difference waveforms of all 58 electrodeswithin the area of the International 10–20 system for eachsubject and condition. The peak latency of the GFP within thelatency range of 100 to 200 ms was determined as the MMNlatency. Individual MMN amplitudes were determined as themean potential within an 80-ms period around the peak latency

of the averaged difference ERPs. The average amplitude data ofthe electrode sites in the left and right frontotemporal regionswere used for statistical analysis (Fig. 1).

The scalp potential was reconstructed by a spherical splineinterpolation. The SCD distributions were then obtained by com-puting the spatial derivatives of the spherical spline functions usedin the potential map interpolation [J=−σ((d2V/dx2)+(d2V/dy2)].The SCD distributions show the scalp areas where radial currentsemerge from the brain (sources; in red) and those where they enterthe brain (sinks; in blue). The 2-D locations of the electrodes weredetermined by projecting the 3-D locations of the electrodes ontothe 2-D surface. The electrode net was applied to the model headand the electrode positions were marked. The polar coordinates(from the vertex) of each electrode site were used to project thesite's location onto the 2-D surface. SCD data were alsocalculated as the mean level within the period of 80 ms aroundeach peak latency of the GFP. SCD data from the electrode sites

Fig. 3. Two-dimensional amplitude and SCDmappings of pure-tone MMN. Comparison of amplitude (left) and SCD (right) topographic mappings of pure-tone MMNin males and females. Each mapping corresponds to the MMN at a latency determined by applying GFP to each grand-averaged ERP. The SCD distribution suggeststhree MMN subcomponents indicated by the curved dashed lines. In the mappings, top depicts anterior, the left side the left hemiscalp, the right side the righthemiscalp, and the bottom the posterior.

45S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

near the left and right superior temporal regions were averagedand subjected to statistical analysis (Fig. 1). Given our interest ingender- and stimulus-related differences in laterality, a lateralityindex for each subject was calculated for both amplitude and SCDdata [(− left+right) / (left2 +right2)1/2, with a positive value indi-cating left-lateralized pattern].

Fig. 4. Phoneme MMN waveform. Grand-averaged phoneme MMN by

2.6. Statistical analyses

Repeated-measures analyses of variance were performed onMMN amplitude and SCD data, latency, and laterality index.‘Gender’ was an interindividual factor, while ‘ear’ receivingdeviant stimulus (left, right) and ‘hemisphere’ (left, right) were

gender and stimulus category (left-ear-deviant, right-ear-deviant).

Fig. 5. Two-dimensional amplitude and SCD mappings of phoneme MMN. Comparison of amplitude (left) and SCD (right) topographic mappings in the phonemeMMN condition in males and females. Each mapping corresponds to the MMN at a latency determined by applying GFP to each grand-averaged ERP. The SCDdistribution suggests bilateral temporal MMN subcomponents indicated by curved dashed lines. In the mappings, top depicts anterior, the left side the left hemiscalp,the right side the right hemiscalp, and bottom the posterior.

46 S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

intraindividual factors. ‘Hemisphere’ was omitted from the intra-individual factors in the analysis of latency and laterality index.

3. Results

3.1. Pure-tone MMN

Grand-averaged waveforms of the difference ERPs (deviant-standard) exhibited similar magnitudes of bilaterally distributednegative components in the frontotemporal regions for male andfemale subjects in both left- and right-ear-deviant conditions(Fig. 2). Statistical analysis of EEG amplitude data revealed nosignificant effect of ‘gender’ or interaction between ‘gender’ andother factors (Table 1). SCD distribution indicated three separateMMN subcomponents, especially for the male subjects, one inthe mid frontal and the two others in bilateral temporal regions(Fig. 3). Visual inspection of the SCD mappings (sinks) sug-

Table 2Summary of the statistical analyses on phoneme MMN

EEG SC

df F P df

Gender 1,42 4.99 b0.05 1,Ear 1,42 2.16 n.s. 1,Hemisphere 1,42 4.73 b0.05 1,Gender⁎ear 1,42 0.75 n.s. 1,Gender⁎hemisphere 1,42 8.04 b0.01 1,Ear⁎hemisphere 1,42 3.77 b0.1 1,Gender⁎ear⁎hemisphere 1,42 0.91 n.s. 1,

gested left-hemisphere dominance in the temporal regions forthe right-ear-deviant condition, though statistical analysis didnot reveal a significant interaction between ‘ear’ and ‘hemi-sphere’, and there was no significant effect of ‘gender’ or inter-action between ‘gender’ and other factors (Table 1). Thus, nosignificant gender difference was observed for the lateralityindex for SCD findings.

A significant effect of ‘ear’was noted for latency, with MMNlatencies in the left-ear-deviant condition (146.7 [SD=21.6] ms)significantly shorter than those in the right-ear-deviant condi-tion (155.7 [SD=19.7] ms) (Table 1).

3.2. Phonetic MMN

Grand-averaged waveforms of the difference ERPs (deviant-standard) revealed a bilateral distribution of negative com-ponents in the frontotemporal regions, which was somewhat

D Latency

F P df F P

42 0.88 n.s. 1,42 1.56 n.s.42 4.97 b0.05 1,42 4.63 b0.0542 1.44 n.s.42 0.00 n.s. 1,42 0.43 n.s42 5.45 b0.0542 1.23 n.s.42 3.09 b0.1

47S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

skewed to the right hemisphere in the female subjects in boththe left- and right-ear-deviant conditions (Figs. 4 and 5). MMNamplitude appeared to be larger in female than in male subjects,especially in the right frontotemporal regions. Statisticalanalysis of the MMN amplitude data revealed significant ef-fects of ‘gender’ and ‘hemisphere’ as well as a significant inter-action between ‘gender’ and ‘hemisphere’ (Table 2). MMNamplitude was significantly greater in the right hemisphere thanthe left (Fig. 6). Furthermore, MMN amplitude was signifi-cantly greater in female than in male subjects, especially in theright hemisphere (F [1, 42]=8.34, P b 0.01) (Fig. 6). MMNamplitude patterns indicated significant right hemisphere domi-nance in female subjects (F [1, 20]=11.96, P b 0.01), thoughinterhemispheric amplitude difference was not significant inmale subjects.

SCD distribution indicated clear current sinks in the temporalregions, although the frontal MMN subcomponent was not asclear as in the pure-tone condition (Fig. 5). Interestingly, visualinspection of the SCD distribution indicated a clear differencebetween male and female subjects. SCD distribution in malesubjects exhibited left-hemisphere dominance, whereas in femalesubjects it was skewed to the right hemisphere. Gender differencewas more evident in the right-ear-deviant condition. Statisticalanalyses revealed a significant effect of ‘ear’ and a significantinteraction between ‘gender’ and ‘hemisphere’ (Table 2). SCDof MMN in the bilateral temporal regions was significantlymore active in the left-ear-deviant condition (−1.91 [SD=1.43]μV/m3) than in the right-ear-deviant condition (−1.53 [SD=1.48]μV/m3).Moreover, the current sinkwas stronger in female than inmale subjects in the right temporal region (F [1, 42]=6.05, P b0.05). The SCD distribution was significantly left-hemisphere-dominant in male subjects (F [1, 22]=8.16, P b 0.01), thoughthere was no significant interhemispheric difference in femalesubjects (Fig. 6).

There was a tendency toward significant interaction among‘gender’, ‘ear’, and ‘hemisphere’. Secondary analyses revealeda significant interaction between ‘ear’ and ‘hemisphere’ in male(F [1, 22]=6.66, P b 0.01) but not in female subjects. Thus, the

Fig. 6. Comparison of amplitude and SCD data of phoneme MMN. Comparisonof amplitude and SCD data of phoneme MMN for each hemisphere (left, right)in male and female subjects.

SCD distribution indicated contralateral dominance in malesubjects alone.

Although both amplitude and SCD findings implied genderdifferences in hemispheric lateralization, a significant effect of‘gender’ was obtained for the SCD lateralization index alone,with SCD distribution significantly skewed to the left hemi-sphere in the male subjects compared to the female subjects.The absence of a significant ‘gender’ effect on amplitude later-alization index suggests that the significant interaction between‘gender’ and ‘hemisphere’might be due in part to the significantdifference between male and female subjects in MMNamplitude in general.

A significant effect of ‘ear’ was observed for latency, withMMN latencies in the right-ear-deviant condition (141.1 [SD=14.5] ms) significantly shorter than in the left-ear-deviant con-dition (144.6 [SD=14.8] ms) (Table 2).

4. Discussion

The findings of the present study can be summarized asfollows: a) in the pure-tone condition, MMN appeared earlier inthe left-ear-deviant condition than in the right-ear-deviant con-dition, while the reverse occurred with phonetic stimuli; b) withphonetic MMN, the EEG exhibited significantly larger ampli-tude in female than in male subjects, especially in the righthemisphere, while SCD findings showed left hemisphere domi-nance and contralateral dominance in male subjects; c) nosignificant gender difference was observed in hemispheric lat-eralization of either EEG or SCD findings in relation to pure-tone MMN.

The difference in pitch frequencies between standard anddeviant stimuli in the pure-tone condition was somewhat largerthan those adopted in previous studies. Opitz et al. (2002)investigated the effects of degree of deviancy on frontal andtemporal lobe contribution to the process of generation ofMMN,and demonstrated that the larger the deviancy the stronger thetemporal lobe contribution and the weaker the frontal lobecontribution. Therefore, the relatively large deviancy used in thepresent study appeared consistent with our goal of examining thelaterality of the temporal MMN subcomponent. In the presentstudy, P3a was not as evident as in the study by Opitz et al.(2002), presumably due to the difference in visual task, whichrequired more attention in the present study. Although the devi-ancies in pure-tone and phoneme conditions cannot be directlycompared, the small P3a in both conditions (Figs. 2 and 4)suggested that attention was relatively well-controlled in eachcondition (Polich, 2007).

The above findings suggest that the gender differences inhemispheric lateralization were specific to the automatic pro-cessing of phonemes, for the following reasons. First, both pure-tone and phonetic MMN latencies differed depending on the earreceiving deviant stimuli. It is therefore implausible that MMNwas generated by the mismatch between the standard and de-viant stimuli composed of intermixed sounds obtained fromboth ears, as this would have resulted in a lack of differencedepending on the ear receiving deviant stimuli. It is reasonableto infer that the processes underlying MMN precede the process

48 S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

of integration of acoustic stimuli presented to both ears, andthat mismatch detection results, at least in part, from separateprocessing of the stimulus presented to each ear. It can thus beconcluded that mismatch of the pure tones was processed anddetected more promptly in the right hemisphere, which was thecontralateral side for the ear receiving deviant stimuli thatproduced shorter MMN latencies. It can similarly be concludedthat the process of mismatch detection for phonemes was fasterin the left than in the right hemisphere.

Although both male and female subjects exhibited a pro-cessing speed advantage in the speech-dominant hemisphere forphonemes, male subjects alone exhibited left-lateralized SCDdistribution in the phonetic MMN condition. This suggests thatphonemic mismatch detection was specifically processed in theleft hemisphere in male subjects, whereas in female subjectsbilateral processing occurred (Shaywitz et al., 1995; Baxter et al.,2003). Laterality indices were significantly different in males andfemales in the phonetic MMN condition, but not in the pure-toneMMN condition. These findings indicate that left-lateralizedcortical activity subservingMMN inmales and gender differencesin lateralization are specific to phonetic processing.

It should also be noted that male subjects alone exhibiteddominant SCD distributions contralateral to the ear receivingdeviant stimuli. The dominance of SCD distribution contral-ateral to the ear receiving deviant stimuli in the male subjectsindicates that ipsilateral input as well as interhemispherictransfer across the corpus callosum to the ipsilateral side wassuppressed, compared with female subjects.

It has been found that the size of the REA for perception oflanguage stimuli in DLTs is negatively correlated with the mid-sagittal cross-sectional area of the corpus callosum, suggestingthat the corpus callosum plays a role in interhemispheric transferof left-ear input to the left hemisphere. Although initial findingssuggested that women have a larger splenium area than men(DeLacoste-Utamsing and Holloway, 1982; de Lacoste et al.,1986), this is now contested. Two meta-analyses of postmortemand MRI studies found that the absolute size of the corpuscallosum is greater in males than in females, but that after ad-justing for whole brain size the gender effect disappears (BishopandWahlsten, 1997) or is even reversed (Driesen and Raz, 1995).

Recent studies using diffuse-tensor imaging (DTI) to ex-amine the microstructural organization of the corpus callosumhave found higher relative anisotropy values in males than infemales (Westerhausen et al., 2003, 2004), though contradictoryfindings have been obtained regarding this (Sullivan et al.,2001; Abe et al., 2002). The authors (Westerhausen et al., 2003,2004) speculate that the corpus callosum in men may containfewer but thicker myelinated fibers than that in women. Thesame group demonstrated a negative association between frac-tional anisotropy scores for the corpus callosum and perfor-mance in identifying language stimuli input to the left ear,suggesting that better interhemispheric transfer is associatedwith reduced density of axon and/or myelin material in thecorpus callosum (Westerhausen et al., 2006). This is consistentwith the findings for men and women in the present study.

However, Hausmann et al. (2002) took an opposing view ongender differences in the functional lateralization of phonetic

MMN, suggesting that reduction in functional lateralization, asobserved in the female subjects in the present study, is causedby a decrease in transcallosal neuronal activation. Hausmannand colleagues (Hausmann, 2005; Hausmann et al., 2002, 2006)systematically demonstrated that high levels of sex hormones(estradiol, progesterone) in female subjects reduced functionallateralization by inhibiting transcallosal transfer. In view of thelack of significant interactions among gender, ear, and hemi-sphere in our SCD findings for pure-tone MMN, these inter-pretations should be considered tentative. Gender differences ininterhemispheric transfer across the corpus callosum warrantfurther study.

Contrary to expectation, the amplitude data for our femalesubjects indicated right-hemisphere dominance in the phoneticMMN condition. A highly speculative explanation of this is thatthe use of a male voice in the present study was more attention-capturing for female than for male subjects, enhancing involun-tary triggering of attention, the activation of which was reflectedin the right frontal MMN subcomponent (Giard et al., 1990).Right-hemisphere involvement in phonetic mismatch detectionprocesses in our female subjects might have resulted in theirlarger MMN amplitude, especially in the right hemisphere.There was no interhemispheric difference in SCD findings forthe temporal subcomponent of phonetic MMN, supporting thissuggestion.

A possible limitation of the present study is its lack of data onmenstrual cycle stage and gonadal steroid hormone levels in thefemale subjects. These have been found to affect functionallateralization, although the effects determined have not beenconsistent. Hausmann et al. (2002) suggested that high levels ofprogesterone in the midluteal phase reduce functional cerebrallateralization, whereas other studies have shown strongest lat-eralization during cycle phases defined by high levels of go-nadal steroid hormones (Altemus et al., 1989; Hampson, 1990;Alexander et al., 2002).

All in all, the findings of the present study suggest that thefunctional lateralization subserving preattentive detection ofphonetic change differs in certain respects between men andwomen. With phonetic MMN, males exhibited left-lateralizedactivation in the supratemporal region, while bilateral activitywas greater in females. Further, the contralateral dominance ofthe SCD distribution associated with the ear receiving deviantstimuli in males indicated that ipsilateral input as well as inter-hemispheric transfer across the corpus callosum to the ipsi-lateral side was more suppressed in males than in females.These findings owing much to the SCD analyses proved thismethod to be useful in identifying the temporal cortical activityunderlying auditory processing.

The results of the present study underscore the significanceof considering the gender differences in the study of MMN,especially when phonetic stimulus is adopted. Moreover, theysupport the view of Voyer and Flight (2001) in that the genderdifference in hemispheric lateralization of language function isobserved in a well-managed-attention condition, which fits thecondition adopted in the MMN measurement; subjects are re-quired to focus attention to a distraction task and thereby ignorethe phonetic stimuli that elicit MMN.

49S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

Acknowledgements

This work was supported in part by a Showa UniversityGrant-in-Aid for Innovative Collaborative Research Projectsand a Special Research Grant-in-Aid for Development ofCharacteristic Education from the Japanese Ministry ofEducation, Culture, Sports, Science, and Technology.

References

Aaltonen, O., Eerola, O., Lang, A.H., Uusipaikka, E., Tuomainen, J., 1994.Automatic discrimination of phonetically relevant and irrelevant vowelparameters as reflected by mismatch negativity. J. Acoust. Soc. Am. 96,1489–1493.

Abdullaev, Y.G., Posner, M.I., 1998. Event-related brain potential imaging ofsemantic encoding during processing single words. Neuroimage 7, 1–13.

Abe, O., Aoki, S., Hayashi, N., Yamada, H., Kunimatsu, A., Mori, H.,Yoshikawa, T., Okubo, T., Ohtomo, K., 2002. Normal aging in the centralnervous system: quantitativeMRdiffusion-tensor analysis. Neurobiol. Aging23, 433–441.

Alexander, G.M., Altemus,M., Peterson, B.S.,Wexler, B.E., 2002.Replication of apremenstrual decrease in right-ear advantage on language-related dichoticlistening tests of cerebral laterality. Neuropsychologia 40, 1293–1299.

Alho, K., 1995. Cerebral generators of mismatch negativity (MMN) and itsmagnetic counterpart (MMNm) elicited by sound changes. EarHear. 16, 38–51.

Alho, K., Connolly, J.F., Cheour, M., Lehtokoski, A., Huotilainen, M., Virtanen,J., Aulanko, R., Ilmoniemi, R.J., 1998. Hemispheric lateralization in pre-attentive processing of speech sounds. Neurosci. Lett. 258, 9–12.

Altemus, M., Wexler, B.E., Boulis, N., 1989. Changes in perceptual asymmetrywith the menstrual cycle. Neuropsychologia 27, 233–240.

Baxter, L.C., Saykin, A.J., Flashman, L.A., Johnson, S.C., Guerin, S.J.,Babcock, D.R., Wishart, H.A., 2003. Sex differences in semantic languageprocessing: a functional MRI study. Brain Lang. 84, 264–272.

Bishop, K.M., Wahlsten, D., 1997. Sex differences in the human corpus callosum:myth or reality? Neurosci. Biobehav. Rev. 21, 581–601.

Celsis, P., Boulanouar, K., Doyon, B., Ranjeva, J.P., Berry, I., Nespoulous, J.L.,Chollet, F., 1999. Differential fMRI responses in the left posterior superiortemporal gyrus and left supramarginal gyrus to habituation and changedetection in syllables and tones. Neuroimage 9, 135–144.

Chiarello, C., Richards, L., 1992. Another look at categorical priming in thecerebral hemispheres. Neuropsychologia 30, 381–392.

de Lacoste, M.C., Holloway, R.L., Woodward, D.J., 1986. Sex differences in thefetal human corpus callosum. Hum. Neurobiol. 5, 93–96.

DeLacoste-Utamsing, C., Holloway, R.L., 1982. Sexual dimorphism in thehuman corpus callosum. Science 216, 1431–1432.

Deouell, L.Y., Bentin, S., Giard, M.H., 1998. Mismatch negativity in dichoticlistening: evidence for interhemispheric differences and multiple generators.Psychophysiology 35, 355–365.

Dorion, A.A., Chantome,M., Hasboun, D., Zouaoui, A., Marsault, C., Capron, C.,Duyme, M., 2000. Hemispheric asymmetry and corpus callosum morpho-metry: a magnetic resonance imaging study. Neurosci. Res. 36, 9–13.

Downar, J., Crawley, A.P., Mikulis, D.J., Davis, K.D., 2001. The effect of taskrelevance on the cortical response to changes in visual and auditory stimuli:an event-related fMRI study. Neuroimage 14, 1256–1267.

Downar, J., Crawley, A.P., Mikulis, D.J., Davis, K.D., 2002. A cortical networksensitive to stimulus salience in a neutral behavioral context across multiplesensory modalities. J. Neurophysiol. 87, 615–620.

Driesen, N., Raz, N., 1995. The influence of sex, age, and handedness on corpuscallosum morphology: a meta-analysis. Psychobiology 23, 240–247.

Frost, J.A., Binder, J.R., Springer, J.A., Hammeke, T.A., Bellgowan, P.S., Rao,S.M., Cox, R.W., 1999. Language processing is strongly left lateralized inboth sexes. Evidence from functional MRI. Brain 122, 199–208.

Giard,M.H., Perrin, F., Pernier, J., Bouchet, P., 1990. Brain generators implicatedin the processing of auditory stimulus deviance: a topographic event-relatedpotential study. Psychophysiology 27, 627–640.

Grimm, S., Roeber, U., Trujillo-Barreto, N.J., Schroger, E., 2006. Mechanismsfor detecting auditory temporal and spectral deviations operate over similartime windows but are divided differently between the two hemispheres.Neuroimage 32, 275–282.

Gur, R.C., Alsop,D., Glahn, D., Petty, R., Swanson, C.L.,Maldjian, J.A., Turetsky,B.I., Detre, J.A., Gee, J., Gur, R.E., 2000. An fMRI study of sex differences inregional activation to a verbal and a spatial task. Brain Lang. 74, 157–170.

Hampson, E., 1990. Estrogen-related variations in human spatial and articulatory-motor skills. Psychoneuroendocrinology 15, 97–111.

Hari, R., Hamalainen,M., Ilmoniemi, R., Kaukoranta, E., Reinikainen,K., Salminen,J., Alho, K., Naatanen, R., Sams, M., 1984. Responses of the primary auditorycortex to pitch changes in a sequence of tone pips: neuromagnetic recordings inman. Neurosci. Lett. 50, 127–132.

Hausmann, M., 2005. Hemispheric asymmetry in spatial attention across themenstrual cycle. Neuropsychologia 43, 1559–1567.

Hausmann, M., Becker, C., Gather, U., Gunturkun, O., 2002. Functional cerebralasymmetries during the menstrual cycle: a cross-sectional and longitudinalanalysis. Neuropsychologia 40, 808–816.

Hausmann,M., Tegenthoff,M., Sanger, J., Janssen, F.,Gunturkun,O., Schwenkreis,P., 2006. Transcallosal inhibition across the menstrual cycle: a TMS study. Clin.Neurophysiol. 117, 26–32.

Hertrich, I., Mathiak, K., Lutzenberger, W., Ackermann, H., 2002. Hemisphericlateralization of the processing of consonant-vowel syllables (formanttransitions): effect of stimulus characteristics and attentional demands onevoked magnetic fields. Neuropsychologia 40, 1902–1917.

Javitt, D.C., Steinschneider, M., Schroeder, C.E., Vaughan Jr., H.G., Arezzo, J.C.,1994. Detection of stimulus deviance within primate primary auditory cortex:intracortical mechanisms of mismatch negativity (MMN) generation. BrainRes. 667, 192–200.

Jemel, B., Achenbach, C.,Muller, B.W., Ropcke, B., Oades, R.D., 2002.Mismatchnegativity results from bilateral asymmetric dipole sources in the frontal andtemporal lobes. Brain Topogr. 15, 13–27.

Kasai, K., Nakagome, K., Iwanami, A., Fukuda,M., Itoh, K., Koshida, I., Kato, N.,2003. No effect of gender on tonal and phonetic mismatch negativity in normaladults assessed by a high-resolution EEG recording. Brain Res. Cogn. BrainRes. 13, 305–312.

Koyama, S., Gunji, A., Yabe, H., Oiwa, S., Akahane-Yamada, R., Kakigi, R.,Näätänen, R., 2000.Hemispheric lateralization in an analysis of speech sounds:left hemispheric dominance replicated in Japanese subjects. Brain Res. Cogn.Brain Res. 10, 119–124.

Levänen, S., Ahonen, A., Hari, R., McEvoy, L., Sams, M., 1996. Deviant auditorystimuli activate human left and right auditory cortex differently. Cereb. Cortex98, 493–501.

Liebenthal, E., Ellingson, M.L., Spanaki, M.V., Prieto, T.E., 2003. SimultaneousERP and fMRI of the auditory cortex in a passive oddball paradigm. Neuro-image 19, 1395–1404.

Mathiak, K., Hertrich, I., Lutzenberger, W., Ackermann, H., 2000. Encoding oftemporal features (formant transients) during binaural and dichotic stimulusapplication: a whole-head magnetoencephalography study. Cogn. Brain Res.10, 125–131.

Mitchell, R.L., Elliott, R., Barry, M., Cruttenden, A., Woodruff, P.W., 2003. Theneural response to emotional prosody, as revealed by functional magneticresonance imaging. Neuropsychologia 41, 1410–1421.

Müller, B.W., Jüptner, M., Jentzen, W., Müller, S.P., 2002. Cortical activation toauditory mismatch elicited by frequency deviant and complex novel sounds:a PET study. Neuroimage 17, 231–239.

Näätänen, R., Michie, P.T., 1979. Early selective-attention effects on the evokedpotential: a critical review and reinterpretation. Biol. Psychol. 8, 81–136.

Näätänen, R., Lehtokoski, A., Lennes, M., Cheour, M., Huotilainen, M., Iivonen,A., Vainio, M., Alku, P., Ilmoniemi, R.J., Luuk, A., Allik, J., Sinkkonen, J.,Alho, K., 1997. Language-specific phoneme representations revealed byelectric and magnetic brain responses. Nature 385, 432–434.

Neill, W.T., Valdes, L.A., 1992. The persistence of negative priming. Steady-state or decay? J. Exp. Psychol. Learn. Mem. Cogn. 18, 565–576.

Nunez, P.L., 1981. Electric fields of the brain: The neurophysics of EEG. OxfordUniversity Press, Oxford.

Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburghinventory. Neuropsychologia 9, 97–113.

50 S. Ikezawa et al. / International Journal of Psychophysiology 68 (2008) 41–50

O'Leary, D.S., 2002. Effects of attention on hemispheric asymmetry. In:Davidson, R.J., Hugdahl, K. (Eds.), The Asymmetrical Brain. MIT Press,Cambridge, pp. 477–509.

Opitz, B., Mecklinger, A., von Cramon, D.Y., Kruggel, F., 1999. Combiningelectrophysiological and hemodynamic measures of the auditory oddball.Psychophysiology 36, 142–147.

Opitz, B., Rinne, T., Mecklinger, A., von Cramon, D.Y., Schröger, E., 2002.Differential contribution of frontal and temporal cortices to auditory changedetection: fMRI and ERP results. Neuroimage 15, 167–174.

Perrin, F., Pernier, J., Bertrand, O., Echallier, J.F., 1989. Spherical splines forscalp current density mapping. Electroencephalogr. Clin. Neurophysiol. 72,184–187.

Polich, J., 2007. Updating P300: an integrative theory of P3a and P3b. Clin.Neurophysiol. 118, 2128–2148.

Rinne, T., Alho, K., Alku, P., Holi, M., Sinkkonen, J., Virtanen, J., Bertrand, O.,Naatanen, R., 1999. Analysis of speech sounds is left-hemispherepredominant at 100–150 ms after sound onset. NeuroReport 10, 1113–1117.

Rinne, T., Alho, K., Ilmoniemi, R.J., Virtanen, J., Näätänen, R., 2000. Separatetime behaviors of the temporal and frontal mismatch negativity source.Neuroimage 12, 14–19.

Sabri, M., Kareken, D.A., Dzemidzic, M., Lowe,M.J., Melara, R.D., 2004. Neuralcorrelates of automatic auditory sensory memory and change detection.Neuroimage 21, 69–74.

Sabri, M., Liebenthal, E., Waldron, E.J., Medler, D.A., Binder, J.R., 2006.Attentional modulation in the detection of irrelevant deviance: asimultaneous ERP/fMRI study. J. Cognitive Neurosci. 18, 689–700.

Scherg, M., Vajsar, J., Picton, T.W., 1989. A source analysis of the late humanauditory evoked potentials. J Cognitive Neurosci. 1, 336–355.

Shalgi, S., Deouell, L.Y., 2007. Direct evidence for differential roles of temporaland frontal components of auditory change detection. Neuropsychologia 45,1878–1888.

Shaywitz, B., Shaywitz, S.E., Pugh, K.R., Constable, R.T., Skudlarski, P.,Fulbright, R.K., Bronen, R.A., Fletcher, J.M., Shankweiler, D.P., Katz, L.,Gore, J.C., 1995. Sex differences in the functional organization of the brainfor language. Nature 373, 607–609.

Sommer, I.E., Aleman, A., Bouma, A., Kahn, R.S., 2004. Do women really havemore bilateral language representation thanmen?Ameta-analysis of functionalimaging studies. Brain 127, 1845–1852.

Sullivan, E.V., Adalsteinsson, E., Hedehus, M., Ju, C., Moseley, M., Lim, K.O.,Pfefferbaum, A., 2001. Equivalent disruption of regional white mattermicrostructure in ageing healthy men and women. NeuroReport 12, 99–104.

Tenke, C.E., Kayser, J., Fong, R., Leite, P., Towey, J.P., Bruder, G.E., 1998.Response- and stimulus-related ERP asymmetries in a tonal oddball task: aLaplacian analysis. Brain Topogr. 10, 201–210.

Voyer, D., Flight, J., 2001. Gender differences in laterality on a dichotic task: theinfluence of report strategies. Cortex 37, 345–362.

Westerhausen, R., Walter, C., Kreuder, F., Wittling, R.A., Schweiger, E., Wittling,W., 2003. The influence of handedness and gender on themicrostructure of thehuman corpus callosum: a diffusion-tensor magnetic resonance imaging study.Neurosci. Lett. 351, 99–102.

Westerhausen, R., Kreuder, F., Dos Santos Sequeira, S., Walter, C., Woerner, W.,Wittling, R.A., Schweiger, E., Wittling, W., 2004. Effects of handedness andgender onmacro- andmicrostructure of the corpus callosum and its subregions:a combined high-resolution and diffusion-tensor MRI study. Brain Res. Cogn.Brain Res. 21, 418–426.

Westerhausen, R., Woerner, W., Kreuder, F., Schweiger, E., Hugdahl, K., Wittling,W., 2006. The role of the corpus callosum in dichotic listening: a combinedmorphological and diffusion tensor imaging study. Neuropsychology 20,272–279.

Yago, E., Corral, M.J., Escera, C., 2001. Activation of brain mechanisms ofattention switching as a function of auditory frequency change. NeuroReport12, 4093–4097.