Neural synchrony and gray matter variation in human males and females: integration of 40 Hz gamma synchrony and MRI measures.

Abstract

Coherent cognition requires activity to be brought together across diverse brain networks. Synchronous, in-phase oscillations in the high-frequency (40 Hz) Gamma range are thought to be one mechanism underlying the functional integration of brain networks. While sex differences have been observed across a range of cognitive functions, their role in normal cortical synchronization has not been elucidated. We recorded Gamma phase synchrony in 500 male and 500 female subjects during an auditory oddball task, which taps discrimination of task-relevant signals. Results revealed a marked sex-linked dissociation in the spatio-temporal pattern of cortical synchronization. Females showed increased Gamma synchrony in the frontal brain, while males showed enhanced synchrony in the parieto-occipital region. These differences were not accounted for by sex differences in whole brain MRI volume. However, there were positive associations between Gamma synchrony and gray matter for females, while these relationships were negative for males. Sex differences in the profile of cortical synchronization may reflect distinct aspects of evolutionary advantage.

Full text

Page 1

Journal of Integrative Neuroscience, Vol. 4, No. 1 (2005) 77–93
c ? Imperial College Press
Research Report
NEURAL SYNCHRONY AND GRAY MATTER VARIATION IN
HUMAN MALES AND FEMALES: AN INTEGRATION OF 40HZ
GAMMA SYNCHRONY AND MRI MEASURES
LEANNE M. WILLIAMS∗
The Brain Dynamics Center, University of Sydney and
Westmead Hospital, NSW 2145, Australia
Discipline of Psychological Medicine, University of Sydney, NSW 2006, Australia
lea@psych.usyd.edu.au
STUART M. GRIEVE
The Brain Resource Company and the Brain Resource International Database
Ultimo, NSW 2007, Australia
THOMAS J. WHITFORD
The Brain Dynamics Center, University of Sydney and
Westmead Hospital, NSW 2145, Australia
Discipline of Psychological Medicine, University of Sydney, NSW 2006, Australia
C. RICHARD CLARK
Cognitive Neuroscience Laboratory, School of Psychology
Flinders University, Adelaide, SA 5042, Australia
RUBEN C. GUR
Section of Neuropsychiatry, Department of Psychiatry, University of Pennsylvania
Philadelphia, Pennsylvania, PA 19104, USA
ELKHONON GOLDBERG
Department of Neurology, School of Medicine, New York University
New York, NY 10016, USA
PIERRE FLOR-HENRY
Adult Psychiatry and Clinical Diagnostics and Research Center
Alberta Hospital Edmonton
Edmonton, AB T5J 2J7
∗Corresponding author.
77

Page 2

78Williams et al.
ANTHONY S. PEDUTO
The Brain Dynamics Center, University of Sydney and
Westmead Hospital, NSW 2145, Australia
MRI Unit, Department of Radiology, Westmead Hospital
Westmead, NSW 2145, Australia
EVIAN GORDON
The Brain Resource Company and the Brain Resource International Database
Ultimo, NSW 2007, Australia
Convenor, Integrative Neuroscience, Brain Dynamics Center, University of Sydney and
Westmead Hospital, NSW 2145, Australia
Discipline of Psychological Medicine, University of Sydney, NSW 2006, Australia
Received 1 February 2005
Revised 9 February 2005
Coherent cognition requires activity to be brought together across diverse brain networks.
Synchronous, in-phase oscillations in the high-frequency (40Hz) Gamma range are thought
to be one mechanism underlying the functional integration of brain networks. While sex
differences have been observed across a range of cognitive functions, their role in normal
cortical synchronization has not been elucidated. We recorded Gamma phase synchrony
in 500 male and 500 female subjects during an auditory oddball task, which taps discrim-
ination of task-relevant signals. Results revealed a marked sex-linked dissociation in the
spatio-temporal pattern of cortical synchronization. Females showed increased Gamma
synchrony in the frontal brain, while males showed enhanced synchrony in the parieto-
occipital region. These differences were not accounted for by sex differences in whole brain
MRI volume. However, there were positive associations between Gamma synchrony and
gray matter for females, while these relationships were negative for males. Sex differ-
ences in the profile of cortical synchronization may reflect distinct aspects of evolutionary
advantage.
Keywords: Gamma phase synchrony; MRI; frontal, occipital and parietal cortices; sex
differences; cognitive function; laterality.
1. Introduction
Drawing on a convergence of evidence across cellular and mesoscopic scales of func-
tion, it has been proposed that high-frequency (40Hz Gamma) oscillations that
are synchronous and in phase across spatially distributed neural populations, may
be one integrative mechanism by which the brain “binds” this activity [8–11].
In humans, synchronous 40Hz Gamma activity is derived from scalp-recorded
electroencephalographic data, time-locked to individual signals. The millisecond
timescale (40 cycles per second) of this measure is consistent with the presumed
speed of information processing, with Gamma synchronization thought to reflect
information transfer between brain regions [9, 10, 12].
There are at least two fundamental phases of Gamma synchrony observed in
the cortex in response to target detection tasks [13]. Early “Gamma 1” synchro-
nization, occurring around stimulus onset, is thought to reflect preparatory pro-
cessing (with constant ISI tasks) and the early perceptual integration of stimulus

Page 3

Neural Synchrony in Males and Females 79
features for both task-relevant and task-irrelevant stimuli. By contrast, later (250–
500ms) “Gamma 2” synchronization arises in response to context evaluation and
integration of task-relevant stimuli, and is associated with reaction time to these
stimuli [13, 14]. In the oddball task, selective discrimination of task-relevant infor-
mation feasibly sets up a bias towards binding a particular set of association
networks [15].
Despite a growing focus on cognition as an active process, involving a dynamic
synchronization of diverse brain networks, sex differences in task-related brain func-
tion have not previously been examined using measures of functional connectivity
or synchrony. In this study we examined whether sex differences would be observed
in Gamma phase synchrony. Using magnetic resonance imaging (MRI), we also
explored the relationship between neuroanatomical differences in gray matter and
sex differences in neural synchrony at 40Hz.
It has been proposed that sex differences in information processing occur along
gradients of both lateralized and frontal-posterior cortical organization [5]. Neu-
roimaging and neuropsychological studies have reported greater hemispheric spe-
cialization in males compared to females, which may be largely independent of sex
differences in task performance [1]. For instance, males show preferential engage-
ment of the left hemisphere during verbal tasks (despite poorer performance), but
preferential right hemisphere involvement for visuo spatial tasks (with better per-
formance) [2–4]. By contrast, females tend to show bilateral engagement for these
tasks with more diffuse frontal engagement and enhanced performance during ver-
bal processing [2, 3]. It has been suggested that the greater functional specialization
present in males may not necessarily subserve greater efficiency. Rather, hemispheric
connectivity may be a more important factor [7]. In this regard, males have consid-
erably longer transmission times than females for transferring information from the
right hemisphere to the left cerebral hemisphere [6]. Greater hemispheric connectiv-
ity has been observed in females [7], consistent with their less lateralized functional
topography.
In this study, sex differences in integrative brain function were examined using
Gamma phase synchrony recorded from left/right hemispheres and frontal/posterior
cortical regions. To ensure robust findings, we employed a large sample (n =
1,000) of healthy individuals from the Brain Resource International Database
[16, 17]. Gamma phase synchrony was recorded in response to the oddball
task which elicits robust early and late phases of synchronization to target
discrimination and context evaluation [13]. The prediction tested in the study
was that during the discrimination of task-relevant stimuli, males and females
would be differentiated by cortical synchronization along lateralized and frontal-
posterior gradients. Our secondary, working hypothesis was that sex differences
in neural synchrony would be reflected in different patterns of correlations
between synchrony and regional (MRI generated) gray matter for males and
females.

Page 4

80Williams et al.
2. Methods
2.1. Subjects
Data from a total of 1,000 healthy subjects were acquired via the Brain Resource
International Database. Exclusion criteria included a history of mental illness, drug
or alcohol addiction, neurological disorder and brain injury, other serious medical
condition and left-handedness. The SPHERE questionnaire [18] was used to screen
out individuals with a potential Axis I disorder, and subjects were also excluded if
they reported a first-degree family member with an Axis I diagnosis. Subjects had a
mean age of 28.7 years. Male (n = 500; mean age = 28.6 years, SD = 19) and female
(n = 500; mean age = 28.8 years, SD = 19) subjects were matched (on a case by
case basis) for age at time of testing, such that they did not differ significantly on
this variable (F1,998= .05, p = .82). Subjects were asked to refrain from smoking
and caffeine for two hours prior to testing, and to refrain from alcohol for 12 hours
prior to testing.
Written informed consent form to participate was provided by all subjects, in
accordance with medical research council ethical guidelines.
2.2. Cognitive task
Gamma synchrony data were acquired during the completion of an auditory oddball
“target detection” task. Participants were seated in a sound and light attenuated
room, set at 24◦C. They were presented with a series of high and low tones, at
75dB for 50ms, with an ISI of 1s and rise and fall times of 5ms. Subjects were
instructed to button press with the index finger of each hand (to counterbalance
for possible motor effects) in response to “task-relevant” target tones (presented
at 1000Hz). They were asked not to respond to “task-irrelevant” or non-target
tones (presented at 500Hz). Speed and accuracy of response were equally stressed
in the task instructions (presented via standardised “.wav” files). The two tones
were presented in a quasi-random order, with the only constraint being that two
target tones could not appear consecutively.
2.3. Data acquisition and analysis: Behavioral data
For subjects’ responses to auditory oddball targets, we recorded accuracy and reac-
tion time. Sex differences in performance were assessed using between groups one
way analysis of variance (ANOVA).
2.4. Data acquisition and analysis: Gamma phase synchrony data
During the oddball task, EEG data were acquired from 26 recording sites: Fp1,
Fp2, F7, F3, Fz, F4, F8, FC3, FCz, FC4, T3, C3, Cz, C4, T4, CP3, CPz, CP4, T5,
P3, Pz, P4, T6, O1, Oz and O2 electrode and 4 EOG channels, orbicularis oculus
and masseter (Quikcap; NuAmps; 10–20 electrode international system). Data were

Page 5

Neural Synchrony in Males and Females 81
recorded relative to the virtual ground, but referenced offline to linked mastoids.
Horizontal eye movements were recorded with electrodes placed 1.5cm lateral to
the outer canthus of each eye. Vertical eye movements were recorded with electrodes
placed 3mm above the middle of the left eyebrow and 1.5cm below the middle of the
left bottom eye-lid. Skin resistance was <5kOhms. A continuous acquisition system
was employed and the bipolar data were EOG corrected offline [19]. The sampling
rate was 500Hz. A low pass filter with attenuation of 40dB per decade above 100Hz
was employed prior to digitization.
The measure of the phase synchronization of Gamma activity was extracted using
a protocol based on Haig et al. [13]. For each single-trial waveform a 128-sample
Welch window was moved along sample by sample, starting with the centre of the
Welch window at 500ms prior to the stimulus (−500ms) and ending with the centre
of the Welch window at 750ms after the stimulus. At each sample position, the phase
of the Gamma frequency component was computed by means of FFT, yielding a
time series of Gamma phase from −500 to 750ms for each single-trial from each site.
Since the sampling rate was 500Hz and the window length was 128 samples, the
width of each frequency bin was 500/128 or 3.91Hz. Thus there was a bin centered
at 39.1Hz, which extended from 37.1Hz to 41.0Hz.
We calculated phase synchrony for each region of interest for responses to both
target and non-target stimuli, within the epoch of −500 to 750ms post-stimulus.
Phase synchrony was defined by the circular variance of phase (or extent of phase-
locking) across sites (such that smaller values reflected less variance, and thereby
greater synchronization). As for a correlation coefficient, circular variance is a nor-
malized measure that ranges from 0 to 1 and, since it is independent of response
amplitude, it has no associated units of measure. For ease of interpretation the cir-
cular variance measure was inverted, such the closer the score was to 1, the greater
the degree of synchrony.
This procedure produced a time series of Gamma phase synchrony globally,
reflecting the extent of phase locking within each region of interest as a function
of time (see Fig. 1 for an example Gamma phase synchrony time series waveform).
Five search regions of interest were based on cortical lobe divisions: frontal (Fp1,
Fp2, Fz, F3, F4, FC3, FCz, FC4, F7 and F8), centro-temporal (C3, Cz, C4, T3, T4,
T5, T6), and parieto-occipital (CP3, CP4, P3, Pz, P4, O1, Oz and O2) regions. We
also examined laterality in terms of left hemisphere (Fp1, F3, F7, FC3, C3, CP3,
T3, T5, P3 and O1) and right hemisphere (Fp2, F4, F8, FC4, C4, T4, CP4, P4, T6
and O2), and left and right sides of each lobe-based region: left frontal (Fp1, F3,
F7 and FC3), right frontal (Fp2, F4, F8 and FC4), left centro-temporal (T3, C3,
T5), right centro-temporal (C4, T4, T6), left parieto-occipital (CP3, P3 and O1)
and right parieto-occipital (CP4, P4 and O2).
The mean of each Gamma synchrony time series for each subject was then com-
puted using the following procedure. First, the time series was smoothed with a
15 point running average. For statistical analysis, we calculated average synchrony
within the time windows for early (−100 to 150ms) and late (200 to 450ms), relative

Page 6

82Williams et al.
Fig. 1.
the figure is the graph of global phase synchrony as a function of time. Along the bottom are
topographic maps of phase from selected points in time along the waveform. Peaks in Gamma
phase synchrony are represented by higher inverted circular variance scores, and by homogenous
color in the corresponding color phase maps (below graph).
Gamma phase synchrony is illustrated in an example simulated data set. At the top of
to a pre-stimulus baseline average. Mixed-model analyses of variance were under-
taken for Gamma synchrony (to target stimuli), with sex as the between-subjects
factor and region (lobe-based regions or hemisphere) as a within subjects factor,
with repeated measures. Subsequent planned contrasts were undertaken to explore
significant effects involving gender. For comparison purposes, parallel MANOVAs
were undertaken to examine sex differences in Gamma 1 and 2 synchrony, in
response to non-target stimuli. We also repeated each analysis with age as a covari-
ate to ensure that sex differences were present over and above individual differences
in age.
2.5. Data acquisition and analysis: Structural MRI data
In addition to Gamma phase synchrony data, high resolution structural magnetic
resonance imaging (MRI) data were acquired for a subset of 151 subjects (79 males,
72 females) from the total pool employed in this study. There was an average
of two weeks between Gamma synchrony and MRI testing sessions. We used a

Page 7

Neural Synchrony in Males and Females83
T1 (MPRAGE) sequence, with images acquired in the saggital plane: 180 slices,
1mm cubic voxels, 256 × 256 matrix, TR 9.7, TE 4, TI 200 and flip angle 12.
Post-testing, a radiologist reviewed each structural MRI scan to confirm that
there was no neuropathology. Volumetric analysis of MRI data was performed using
voxel based morphometry (VBM) in SPM2 (http://www.fil.ion.ucl.ac.uk/spm).
Images were firstly normalized to a Brain Resource International Database-specific
T1-weighted template. This template was made using 255 subject images which
had been previously normalized to the MNI T1-weighted template. The protocol
for normalization and segmentation is described in detail elsewhere [20]. Segmented
images were modulated with the Jacobean determinants of the deformation field
generated in the spatial normalization, in order to preserve the original tissue vol-
umes [20, 21]. The images were then smoothed with a Gaussian kernel of 12mm
FWHM. Images were segmented into gray matter, white matter and cerebrospinal
fluid (CSF). Anatomical assignments were made with reference to the anatomical
parcellation (Automated Anatomical Labeling; AAL) masks performed by Tzourio-
Mazoyer et al. [22] in the MNI coordinate system. In addition to total gray matter,
we extracted regional gray matter volumes (calculated as absolute values and rela-
tive to the whole brain) for the left and right frontal, temporal, parietal and occipital
cortices to correspond to the regions used in Gamma synchrony analyses.
ANOVA was used to compare males and females on total gray matter volume and
on relative gray matter for the regions of interest. Subsequent exploratory Pearson
correlation analyses were undertaken to determine whether significant sex differences
in gray matter were related to sex differences in Gamma synchrony.
3. Results
3.1. Behavioral data
A total of 48 subjects (4.8% of the sample; 22 males and 25 females) had missing
Gamma data for one or more of the regions of interest, and were therefore excluded
from focal analyses.
Females were found to have significantly slower responses to oddball targets
(F1,950= 20.09, p < .0001), but there was no significant sex difference in number of
errors (F1,950= 3.74, p = .20).
3.2. Gamma phase synchrony data
The means (and standard deviations) for early and late Gamma synchrony for males
and females are presented in Table 1.
3.2.1. Early gamma
For early Gamma synchrony, there was an interaction of borderline significance
between gender and hemisphere (F1,950 = 3.56, p = .056). Within-group planned

Page 8

84Williams et al.
Table 1.
Means, standard deviations and range for early and late Gamma synchrony in response to target stimuli and p values for the planned
contrasts between males and females*.
Group
Females (F)
Males (M)
M vs F Contrasts (p Values)
Region
Early Gamma
Late Gamma
Early Gamma
Late Gamma
Early Gamma
Late Gamma
Left Hemisphere
.67 (.09)
.68 (.09)
.66 (.09)
.67 (.09)
.16
.12
Right Hemisphere
.68 (.09)
.69 (.09)
.65 (.08)
.66 (.09)
.016
.005
Frontal
.70 (.10)
.71 (.10)
.66 (.10)
.67 (.10)
.0001
.0001
Left Frontal
.74 (.10)
.74 (.11)
.70 (.11)
.71 (.11)
.0001
.0001
Right Frontal
.74 (.10)
.75 (.11)
.70 (.11)
.71 (.11)
.0001
.0001
Centro-Temporal
.68 (.09)
.69 (.09)
.68 (.08)
.70 (.09)
.54
.61
Left Centro-Temporal
.76 (.10)
.77 (.09)
.77 (.09)
.78 (.09)
.11
.30
Right Centro-Temporal
.77 (.10)
.79 (.10)
.78 (.09)
.79 (.09)
.12
.25
Parieto-Occipital
.74 (.09)
.76 (.09)
.76 (.08)
.77 (.08)
.035
.13
Left Parieto-Occipital
.78 (.08)
.80 (.08)
.82 (.07)
.83 (.07)
.0001
.0001
Right Parieto-Occipital
.79 (.08)
.81 (.07)
.82 (.07)
.83 (.06)
.027
.14
*Gamma indexed by circular variance (possible minimum value of 0 and maximum of 1, and range across cortical regions of .26 to .98).

Page 9

Neural Synchrony in Males and Females85
contrasts revealed that females had higher right than left hemisphere synchrony
(F1,474= 8.06, p = .005), whereas males did not show a difference (F1,477= .064,
p = .8). Between-group contrasts showed that in females, right hemisphere syn-
chrony was also higher than that for males (F1,950= 5.86, p = .016), while there
were no sex difference for the left hemisphere (F1,950= 1.99, p = .16) (Table 1).
For the lobe-based divisions, males and females had a distinctive profile of
Gamma synchrony across frontal, centro-temporal and parieto-occipital regions
reflected in a significant interaction between sex and synchrony across these regions
(F2,949= 32.92, p < .0001). While synchrony was highest in the parieto-occipital
region relative to other regions for both males and females, males had a signifi-
cant and sharply linear increase in synchrony from the frontal to centro-temporal to
parieto-occipital regions (F2,477= 135.29, p < .0001), while synchrony for females
followed a significant (F2,474 = 18.13, p < .0001) J-shape pattern across these
regions. Between-group planned contrasts (Table 1) showed that these differences
were reflected in greater synchrony in the parieto-occipital region for males com-
pared to females (F1,950= 4.46, p = .035), and significantly greater synchrony in
the frontal region for females relative to males (F1,950 = 24.30, p < .0001), but
no difference within the centro-temporal region (F1,950 = .37, p = .54). Figure 2
illustrates this distribution of synchrony for males and females.
In regard to laterality, males showed no difference between left frontal and
right frontal early synchrony (F2,477 = .98, p = .23), nor between left and right
parieto-occipital synchrony (F2,477= .17, p = .68). Similarly, females also showed
no difference between left and right frontal (F2,474 = .24, p = .62), nor left and
right parieto-occipital (F2,474= 2.39, p = .13) synchrony. Between-group planned
contrasts (Table 1) confirmed that enhanced parieto-occipital synchrony in males
compared to females was apparent bilaterally (left, F1,950= 14.42, p < .0001; right,
F1,950= 4.88, p = .027), as was the comparatively enhanced frontal synchrony in
females (left, F1,950= 29.28, p < .0001; right, F1,950= 29.73, p < .0001).
3.2.2. Late gamma
The differential pattern of late Gamma synchronization in males and females largely
corresponded to that for early Gamma. As for early Gamma, there was a significant
gender by hemisphere interaction for late Gamma synchrony (F1,950 = 5.42, p =
.02). Again, females had relatively greater right than left hemisphere synchrony
(F1,475 = 6.94, p = .00), while males showed no difference (F1,477 = .28, p =
.60). Between-group contrasts showed right hemisphere synchrony in females was
enhanced compared to males (F1,950 = 8.07, p = .005), while there was again no
differences for the left hemisphere (F1,950= 2.48, p = .12) (Table 1).
There was also a significant interaction for gender and region for lobe-based
frontal, centro-temporal and parieto-occipital regions (F2,949= 32.19, p < .0001).
As for early Gamma, males showed a sharply linear increase across frontal and
temporal to parieto-occipital regions (F2,477 = 112.41, p < .0001), while females

Page 10

86 Williams et al.
Fig. 2.
phase synchrony in males (yellow arrows) and frontally in females (red arrows). Both early and
late Gamma phase was enhanced in females relative to males across the frontal cortex bilaterally.
By contrast, males showed a relative enhancement of Gamma phase synchrony across the occipito-
parietal cortices, which was apparent bilaterally for early Gamma and most pronounced in the left
hemisphere for late Gamma. While these sex differences were small, they were highly robust.
An illustration of the cortical lobe topography of relatively enhanced posterior Gamma
showed more of a J-shape pattern across these regions (F2,47= 35.50, p < .0001).
As for early frontal Gamma, females showed enhanced late frontal synchrony relative
to males (F1,950= 26.10, p < .0001). Late Gamma synchrony was similarly more
pronounced in the parieto-occipital region for males compared females, but only at
weak trend level (F1,950 = 3.06, p = .13) (Figure 2). There was a corresponding
lack of sex difference in late synchrony for the centro-temporal region (F1,950= .27,
p = .61) (Table 1).
When the laterality of these regions was considered, males showed no difference
between left frontal and right frontal early synchrony (F2,477= .88, p = .35), nor
between left and right parieto-occipital synchrony (F2,477= .35, p = .56). However,
while females also showed no difference between left and right late frontal synchrony
(F2,474= .02, p = .88), the reduction in parieto-occipital synchrony was less appar-
ent in the right compared to left hemisphere (F2,474= 5.03, p = .02). Consistent with

Page 11

Neural Synchrony in Males and Females87
these within-group planned contrasts, the enhancement in late frontal synchrony in
females compared to males was apparent bilaterally, following the same pattern
as early frontal synchrony (left, F1,950 = 26.47, p < .0001; right, F1,950 = 33.87,
p < .0001) (Table 1). Males showed a greater left-sided late synchrony in the parieto-
occipital region than females (F1,950= 12.37, p < .0001), which corresponded to that
for early synchrony, but nor the right side of this region (F1,950= 2.14, p = .14) due
to the parieto-occipital laterality effect in females (Table 1).
3.3. MRI data
Consistent with previous MRI studies, females had significantly less total gray mat-
ter volume than males (F1,149= 27.58, p < .0001). For the absolute values, males
also had greater gray matter volume than females for frontal, temporal and parieto-
occipital lobes, both left and right-sided (p < .0001). However, differences were less
striking when relative values (regional gray matter relative to the whole brain) were
considered. Gray matter was greater for males compared to females in the right
(F1,149 = 8.92, p = .003) and left (F1,149 = 9.00, p = .003) parietal cortex, and
in the left occipital cortex at trend level (F1,149= 3.41, p = .07). In the temporal
cortex, males had a trend towards greater right-sided gray matter (F1,149= 3.66,
p = .06), and females a trend towards greater left-sided gray matter (F1,149= 3.40,
p = .07). There were no sex differences in the frontal cortex.
3.4. MRI-Gamma synchrony relationships
While relationships between MRI and Gamma synchrony data were generally low
to moderate, males and females showed a distinct pattern of relationships between
synchrony and gray matter.
There was no relationship between Gamma phase synchrony and global gray
matter for either males or females. For females, right occipital gray matter was
related positively to both right (r = .30, p = .028) and left (r = .28, p = .04)
early frontal synchrony, and with early right parieto-occipital synchrony (r = .27,
p = .049). Right parietal gray matter was associated positively with right frontal
early synchrony (r = .33, p = .016). For late synchrony, significant correlations also
involved only right parietal and occipital gray matter for females. Right parietal gray
matter was positively associated with right frontal (r = .34, p = .014), while right
occipital gray matter was positively associated with left frontal (r = .28, p = .045),
right frontal (r = .29, p = .03) and right parieto-occipital (r = .31, p = .023) late
synchrony.
Males showed moderate and negative associations between left frontal early
Gamma synchrony and right frontal (r = −.26, p = .04), right occipital (r = −.36,
p = .004), right parietal (r = −.39, p = .002), left parietal (r = −.38, p = .003) and
right temporal (r = −.28, p = .03) relative gray matter. Right frontal early syn-
chrony showed a similar pattern of negative associations with right frontal (r = −.30,

Page 12

88Williams et al.
p = .02), right occipital (r = −.33, p = .01), right parietal (r = −.25, p = .05), left
parietal (r = −.34, p = .008) and right temporal (r = −.33, p = .01) relative gray
matter. In contrast to this global pattern of associations, both left parieto-occipital
(r = −.35, p = .007) and right parieto-occipital (r = −.27, p = .035) early synchrony
were negatively associated with right parietal gray matter only.
A similar pattern of negative associations was revealed for late Gamma phase
synchrony for males. Left frontal late synchrony showed moderate negative associ-
ations with right frontal (r = −.25, p = .06, trend level), right parietal (r = −.40,
p = .002), left parietal (r = −.38, p = .003), right occipital (r = −.37, p = .003)
and right temporal (r = −.28, p = .029) relative gray matter. Right frontal late
synchrony also showed moderate negative associations with right frontal (r = −.32,
p = .012), right parietal (r = −.25, p = .05), left parietal (r = −.36, p = .005)
and right occipital (r = −.34, p = .008) gray matter. Both left (r = −34, p = .008)
and right (r = −.34, p = .008) parieto-occipital late synchrony were negatively asso-
ciated with right parietal gray matter, as was the case for early synchrony. When
age was included as a covariate, the relationships between synchrony and regional
gray matter remained stable or became more clearly significant (with coefficients
increasing to .5).
4. Discussion
In this study, we used a measure of cortical synchronization (40Hz Gamma phase
synchrony) to examine sex differences in functional brain integration to task-relevant
stimuli. In a large sample of normative subjects, females and males were distin-
guished by the topography of synchronization. In terms of cerebral hemisphere,
females showed greater right hemisphere synchrony than males, while no differ-
ences were apparent within the left hemisphere. Females also showed greater frontal
synchronization, while males were defined by a relative enhancement of occipito-
parietal synchrony. The greater frontal synchrony in females was apparent bilaterally
for both early (within 150ms post-stimulus) and later (200–450 ms post-stimulus)
periods of synchronization in response to target stimuli. For males, greater occipito-
parietal synchrony was also apparent bilaterally for the early phase, but only in
the left hemisphere for the later phase. By contrast, there were no sex differences
in Gamma phase synchrony in the temporal cortex. While the mean differences
between females and males in frontal and occipital-parietal synchrony were gener-
ally small (Table 1), they were highly robust, pointing to the existence of consistent
yet subtle sex-related variation in neural synchronization. These differences may
only be revealed with a large standardized sample, such as the one employed in
this study.
The differentiation of males and females in terms of Gamma synchronization
along the frontal versus parieto-occipital gradient suggests the use of preferential
strategies for signal discrimination, tapped here by target stimuli. Discrimination of
salient stimuli draws on the integration of frontal with posterior attention networks,

Page 13

Neural Synchrony in Males and Females 89
as well as the coordination of left and right hemispheres [24]. In attention models, it
is proposed that the posterior parieto-occipital networks subserve vigilance in regard
to integrating sensory input necessary for shifts of attention and the amplification
of relevant signals and suppression of noise [24]. The frontal (anterior) system, on
the other hand underlies the coordination of attentional activities, the integration of
attention with working memory and with initiation of actions [24]. This distinction
might be summarized as the detection versus contextual processing of salient stimuli.
In our study, similarly high mean levels of synchronization were observed for early
perceptual integration and detection of target stimuli (within 150ms of stimulus
onset), and for subsequent context evaluation of these stimuli (200–450ms post-
stimulus) (Table 1), consistent with our previous studies of Gamma synchrony in
case:control studies [12, 23, 21, 22, 33].
The possibility that enhanced parieto-occipital synchrony in males reflects a
gender-specific bias towards the integration of sensory input necessary for effective
stimulus detection is consistent with animal evidence that parietal Gamma activity is
enhanced when cats are tracking their prey (mice) [25]. Indeed, one might speculate
that males have evolved traits selected for their adaptive value in vigilance and
orienting, related previously to the hunting role [26].
By contrast, the relatively heightened frontal synchrony in females suggests a bias
towards enhanced connectivity of frontal brain networks for coordination of attention
and context evaluation of salient stimuli. The greater efficiency of inter-hemispheric
transfer in females [7] might account for this bilaterally increased level of frontal
synchronization. Given the role of the frontal cortex in coordinating more complex
information processing, including language, relatively enhanced frontal integration
in females might underlie their better performance on verbal tasks [2, 3].
MRI data revealed the expected differences in neuroanatomy for males and
females [27], with males showing relatively greater gray matter globally, and across
each cortical lobe. Notably, when gray matter was considered relative to whole brain
volume, differences were observed only in the parietal lobe, with males showing rel-
atively greater parietal gray matter than females. This difference is in the reverse
direction to that observed previously in a sample of 84 subjects [27], suggesting
that it might be a sample-specific neuroanatomical variation. When we examined
the possibility that neuroanatomical variations could contribute to differences in
synchrony, we found a distinct pattern of MRI-synchrony relationships in males and
females.
For males, Gamma phase synchronization across the frontal region (both early
and late) was negatively associated with parietal and occipital gray matter. Parieto-
occipital synchrony also showed a negative association with right parietal gray mat-
ter in males. Females, by contrast, showed positive relationships between frontal
synchrony and parieto-occipital gray matter. While these synchrony-gray matter
relationships were only moderate in size (around .25 to .39), our finding that these
associations clearly demarcated males and females in both topography and direc-
tion points to their validity. It is nonetheless recognized that, given the large number

Page 14

90Williams et al.
of correlations undertaken, and their exploratory nature, it would be important to
replicate the associations in another group of subjects.
Broadly, the pattern of negative associations between frontal Gamma synchrony
for males, but positive frontal synchrony-gray matter associations in females, sug-
gests that variations in gray matter may contribute in particular to variations in
synchrony of the frontal cortex. Our finding that parieto-occipital gray matter was
associated with frontal synchrony (negatively for males, positively for females) sug-
gests that gray matter may contribute indirectly to the topography of synchrony,
via the reciprocal region of the cortex. This proposal is consistent with evidence that
the formation of Gamma oscillations in interneuronal interactions involves reciprocal
inhibitory connections [28]. Similarly, both modelling and lesion data suggest that
the generation of Gamma is typically more contralateral than ipsilateral [28, 29;
review see 8], reflecting reciprocal transcallosal inhibition, rather than inhibition
between neighboring neurons [28]. We note of course that our MRI analysis was
constrained to cortical regions, and the likelihood that interactions between the
subcortical thalamus and cortex contribute to Gamma synchrony [8] requires fur-
ther investigation.
It is also possible that sex differences in Gamma synchrony are underpinned by
neurochemical differences, given the proposed neuropharmacological mechanisms of
synchrony. Given the role of gamma-aminobutyric acid (GABA)-ergic interneurons
in modulating short-range Gamma synchronization in particular [8], variation in
GABA-ergic activity may produce sex-related differences. Indeed, animal studies
suggest that hormone-mediated changes in the GABAergic system contribute to the
development of sexually dimorphic characteristics in the adult brain [30]. Similarly,
animal evidence suggests that sex differences in brain function may also reflect
hormone-related changes in the N-methyl-D-aspartate (NMDA) system, which has
been implicated in the mechanisms of long-range Gamma synchrony [8].
This study underscores the potential of Gamma phase synchrony as a measure
of individual differences in integrative function in the normative population. The
findings suggest that differences in cortical synchronization may contribute to sex
differences in the ability to discriminate task-relevant signals. Future studies might
further explore this area in regard to sex differences in verbal and visuo-spatial
functions, observed previously to differ among males and females. Elucidating sex
differences in the healthy human brain may also help us to understand the sex-
differentiated symptomatology in clinical conditions. Many brain-related disorders,
such as schizophrenia, post-traumatic stress (PTSD) and attention deficit hyperac-
tivity (ADHD) disorder, show sex differences in patterns of prevalence and severity.
Such clinical disorders are also marked by a deficit in the ability to discriminate and
appropriately respond to task-relevant stimuli. Our initial study of schizophrenia
suggests that sex differences in Gamma synchrony provide a pertinent window onto
the biological basis of these conditions and their progression [32]. Future studies will
undertake a further integration of Gamma phase synchrony measures with genetic,

Page 15

Neural Synchrony in Males and Females 91
neurochemical, fMRI and numerical simulation data to shed light on the mecha-
nisms, which predispose individuals to breakdowns in integrative brain function.
Acknowledgements
Scientists
(www.BrainResource.com).
andsubjectsfrom theBrainResourceInternationalDatabase
References
[1] Ashburner J, Friston KJ, Voxel-based morphometry — The methods, Neuroimage
11:805–821, 2000.
[2] Bouyer JJ, Montaron MF, Rougeul–Buser A, Buser A, Thalamo-cortical rhythmic sys-
tem accompanying high vigilance levels in the cat, in Pfurtscheller G, Buser P, Lopes
da Silva FH, Petsche H (eds.), Rhythmic EEG Activities and Cortical Functioning,
Elsevier, Amsterdam, 1980, pp. 63–77.
[3] Crick F, Koch C, A framework for consciousness, Nat Neurosci 6:119–126, 2003.
[4] Davis AM, Grattan DR, Selmanoff M, McCarthy MM, Sex differences in glutamic acid
decarboxylase mRNA in neonatal rat brain: Implications for sexual differentiation,
Horm Behav 30(4):538–52, 1996.
[5] Davitzikos C, Sresnick S, Sex differences in anatomic measures of interhemispheric
connectivity: Correlation in cognition in women but not men, Cereb Cortex 8:635–640,
1998.
[6] de Jongh A, de Munck JC, Baayen JC, Jonkman EJ, Heethaar RM, van Dijk BW, The
localization of spontaneous brain activity: first results in patients with cerebral tumors,
Clin Neurophysiology 112(2):378–85, 2001.
[7] Fernandez-Duque D, Posner MI, Brain imaging of attentional networks in normal and
pathological states, J Clin Exper Neuropsyc 23:74–93, 2001.
[8] Goldberg E, Cognitive bias, functional cortical geometry, and the frontal lobes: Later-
ality, sex and handedness, J Cognitive Neurosci 6:276–296, 1994.
[9] Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS, A voxel-
based morphometric study of ageing in 465 normal adult human brains, Neuroimage
14:21–36, 2001.
[10] Gordon E, Integrative neuroscience in psychiatry: the role of a standardized database,
Australas Psychiat 11(2):156–163, 2003.
[11] Gordon E, Integrative neuroscience and psychiatry, Neuropsycopharmacol 28(1):S2–S8,
2003.
[12] Gordon E, Williams LM, Haig AR, Wright J, Meares RA, Symptom profile and
“Gamma” processing in schizophrenia, Cognitive Neuropsychiatry 6:7–19, 2001.
[13] Gratton G, Coles MG, Donchin E, A new method for off-line removal of ocular artifact,
Electroen Clin Neurophysiol 55:468–484, 1983.
[14] Gur RC, Alsop D, Glahn D, Petty R, Swanson CL, Maldjian JA, Turetsky BI, Detre JA,
Gee J, Gur RE, An fMRI study of sex differences in regional activation to a verbal and
a spatial task, Brain Lang 74:157–170, 2000.