Morphologic asymmetry of the human anterior cingulate cortex
Rene J. Huster,a,b,⁎Rene Westerhausen,aFrank Kreuder,a
Elisabeth Schweiger,aand Werner Wittlinga
aCenter for Neuropsychological Research, University of Trier, Germany
bInternational Center for Integrated Neuroscience, WIKO Greifswald, Germany
Received 4 August 2006; revised 6 September 2006; accepted 6 October 2006
Available online 11 December 2006
The anterior cingulate cortex (ACC) is thought to play a major role in
executive processes. Studies assessing neuroanatomical attributes of
this region report a high degree of morphological variability. Recent
theories consider the fissurization of the cortex to be a product of gross
mechanical processes related to cortical growth and local cytoarchi-
tectural characteristics. Hence, local sulcal patterning and gray matter
volume are supposed to be associated. ACC fissurization was
quantified in left- and right-handers of both sexes by recording the
presence and extension of the paracingulate sulcus (PCS). Differences
between groups regarding local gray matter volume were assessed by
means of optimized voxel-based morphometry (oVBM) including
additional modulation. Overall, the PCS occurred more often and was
more pronounced in the left as compared to the right anterior cingulate
region, although hemispheric differences were less pronounced in male
left- and female right-handers. These discrepancies between groups
seem to stem from variations of cingulate morphology in the left rather
than the right hemisphere. The pattern of relevant comparisons in the
oVBM analysis indicated a similar interaction. Therefore, evidence was
found for discrepancies between groups and hemispheres on the
© 2006 Elsevier Inc. All rights reserved.
The cingulate cortex is the most prominent structure of the
human medial wall. Especially its anterior portion has been subject
to intensive functional investigations. Recent research emphasizes
the role of the anterior cingulate cortex (ACC) in a variety of
cognitive processes, namely executive control (Fan et al., 2003),
conflict monitoring (Botvinick et al., 2004), response selection
(van Veen and Carter, 2005), error processing (Mathalon et al.,
2003) as well as reward-based decision making (Bush et al., 2002).
Up to now, no single unifying theory can explain the diversity
found in studies on ACC functioning. Importantly, an adequate
analysis of neuroimaging and electroencephalographic data may
also have been hindered by interindividual variability of local
The most apparent pattern of morphologic variability in the
ACC concerns the doubling of the cingulate gyrus (CG).
According to Vogt et al. (1995), the more dorsal or paracingulate
gyrus (PCG) contains, if present, portions of areas 24(V)c and 32(V).1
This latter cytoarchitectural field is believed to constitute a
cingulofrontal transition area. In case of a single cingulate gyrus,
area 32(V)is buried in the depth and the dorsal bank of the cingulate
sulcus (CS). From a functional perspective, fMRI studies often
show activations of areas 24(V)and 32(V)(e.g. van Veen and Carter,
2005; Laird et al., 2005). For these reasons, the PCG should be
included in the definition of an anterior cingulate region, as a
differentiation based on cytoarchitectural characteristics is not
possible in neuroimaging studies with human subjects. At present,
sulci of the anterior cingulate region seem to be the only, but
sufficient, guideline for the interpretation of functional and
structural MRI data.
Although extensively studied from a functional perspective,
research identifying structural characteristics of the ACC is sparse.
Paus and colleagues (1996b) quantitatively studied the morpholo-
gical variability of this region in a large sample of 247 subjects.
They found the paracingulate sulcus (PCS) to be present and also
more extended in the left as compared to the right hemisphere.
Beyond, whereas the PCS commonly was of intermediate size in
males, females more often showed an absent or stronger developed
PCS thereby demonstrating a higher degree of interindividual
macroanatomical variability. Extending these findings, Yücel et al.
(2001) additionally computed individual asymmetry metrics that
were also indicative of a leftward pattern in ACC folding. This
effect was more pronounced in male subjects while females tended
NeuroImage 34 (2007) 888–895
⁎Corresponding author. Center for Neuropsychological Research, Johan-
niterufer 15, D-54290 Trier, Germany. Fax: +49 651 201 3766.
E-mail address: firstname.lastname@example.org (R.J. Huster).
Available online on ScienceDirect (www.sciencedirect.com).
1When referring to areas 24(′)or 32(′), the relevant areas in perigenual
(24 or 32) and midcingulate (24′ or 32′) cortex are meant. This
nomenclature is given in accordance with the four compartment model of
Vogt and collaborators. Specific information can be found on this Web
1053-8119/$ - see front matter © 2006 Elsevier Inc. All rights reserved.
to be more symmetric. Taken together, both studies imply a
stronger cortical folding of the ACC in the left compared to the
right hemisphere, but they differ with respect to gender-associated
effects in regional fissurization.
It has to be added that neither study assessed neuroanatomical
variations associated with subjects' handedness. Nevertheless, there
is ample evidence that this variable exerts influence on structural
neuroanatomy not only by simple main effects, but also by
complex interactions with other relevant factors. For example,
Amunts et al. (2000) measured the depth of the central sulcus as an
estimate of the hand motor area and found an inversion of cortical
asymmetry with left- compared to right-handed males. A similar
effect was not apparent in female subjects. Research on the planum
temporale also underscores the existence of complex interactions
between variables. In a study by Dos Santos Sequeira et al. (2006),
an overall asymmetry with larger planum temporale in the left
hemisphere emerged. However, the mean magnitude of this
asymmetry was not determined by handedness, gender or speech
lateralization alone but varied as a consequence of their specific
Because mechanical processes of cortical growth are believed
to drive cerebral fissurization (Armstrong et al., 1995), group
differences in gray matter volume (GMV) of the ACC should be
expected to mirror those of its regional folding. Interestingly,
studies on the cingulate GMV of schizophrenics and healthy
subjects are not always in accordance with this assumption (for a
discussion, see Kopelman et al., 2005). One important factor
contributing to discrepancies between studies is the definition used
to delineate the relevant region of interest. More specifically, the
inclusion or exclusion of the PCG might lead to varying volumetric
estimates. Manually tracing the intrasulcal gray matter of the
cingulate and the paracingulate sulcus, Paus and coworkers
(1996a) found a negative association between the respective
volumes. When both sulci were present, larger paracingulate came
along with smaller cingulate volumes. Unfortunately, the reported
metrics of total regional GMV did not differentiate between
individuals with single and double cingulate sulci. Nevertheless,
given the definitional irregularities, even inversions of anterior
cingulate volume asymmetries can be expected depending on the
chosen dorsal border of the ACC.
A recent study by Allen and colleagues (2003) took the
“outside” sulcus as boundary in cases of sulcal doubling.
Computing the cingulate GMV, they found a leftward asymmetry
in males but not in females. Although these results confirm what
can be expected from studies on anterior cingulate fissurization,
another major problem regarding the definition of the anterior
cingulate region has to be mentioned. The cingulate region in its
anterior to posterior extension at first glance seems to be a unitary
one from a gross morphological perspective. Nevertheless, recent
studies suggest a functional fragmentation of this cerebral
component. Focusing on anterior areas of the cingulate cortex,
functional as well as cytoarchitectural research implies the
existence of at least two major subdivisions subserving distinct
functions: a midcingulate cognitive and a perigenual affective
subdivision (Bush et al., 2000; Devinsky et al., 1995). It should be
noted that studies on the fissurization of the ACC (Paus et al.,
1996b; Yücel et al., 2001) approximated midcingulate portions of
the ACC (MCC), while those assessing regional GMV most often
referred to the whole anterior region. Actually, the above-cited
metrics of Allen and coworkers were based on the entire cingulate
region, therefore even including posterior parts.
Summarizing with respect to the cognitive or midcingulate
subdivision of the anterior cingulate cortex, findings suggest a
leftward asymmetry in cortical folding as operationalized via
assessment of PCS occurrence and expansion. Gender effects seem
to exist, but their exact nature still needs clarification. Whether
handedness is associated with differential patterns of gyrification in
the ACC has not yet been in the focus of scientific investigation.
Theories suggest a positive relationship between regional GMV
and the degree of fissurization. Therefore, group differences
regarding the amount of ACC gray matter should mirror those
results on its gyrification. Having yielded contradictory results
probably due to differences in the specific anterior cingulate
regions included, previously published studies are unable to
unambiguously clarify this topic.
To address the aforementioned issues, we assessed the main
characteristics of ACC folding in a large sample of left- and right-
handed subjects of both sexes utilizing structural magnetic
resonance imaging (MRI). In addition, group differences in
regional GMV were evaluated by optimized voxel-based morpho-
metry including additional modulation (oVBM). This method can
be regarded as a technically matured and automated procedure to
analyze structural MRI data (Ashburner and Friston, 2000;
Mechelli et al., 2005).
Materials and methods
A total of 146 female and male students of the University of
Trier underwent the MR scanning protocol described below. To
assess the effects of handedness on cingulate gyrification and
GMV, nominal left- and right-handed participants were recruited.
A modified version of the Edinburgh Handedness Inventory (EHI;
Oldfield, 1971; a detailed description of relevant modifications can
be found in Dos Santos Sequeira et al., 2006) was used to
psychometrically validate individual handedness. Non-correspon-
dence of psychometric and nominal hand preference resulted in
exclusion from analysis. Data sets showing artefacts were also
excluded from further processing. None of the subjects reported a
history of psychiatric or neurological disorders. Sulcal character-
istics were measured in 84 left (f=46, m=38) and 58 right handers
(f=32, m=26). The sample for oVBM consisted of 46 left-handed
(f=23, m=23) and 46 right-handed (f=23, m=23) subjects,
randomly chosen from the pool of all data sets available.
Fissurizational characteristics of these subgroups were assured to
match those of the source samples. An equalization of sample sizes
for oVBM was realized to circumvent problems associated with
unbalanced variance analytic designs, such as an increased
likelihood for the violation of its assumptions (e.g., multi-
collinearity or variance inhomogeneity). Participants' age ranged
from 19 to 38 years (mean=24±3.55 years) and did not differ
significantly between relevant subgroups. All participants gave
written informed consent prior to study participation.
All scans were performed on a 1.5-T Philips Gyroscan Intera
system (Philips, Netherlands). Movement-related artefacts were
minimized by stabilization of the participant's head with foam
cushions and an elastic forehead strap. Using Fast Field Echo
acquisition, 160 contiguous T1-weighted (TR=11.64 ms,
889R.J. Huster et al. / NeuroImage 34 (2007) 888–895
TE=3.3 ms) 1-mm thick slices were collected in the axial plane.
Given a congruent field of view (FOV) and in plane matrix
(256×256 mm2/256×256), resulting voxels were isotropic
Classification of ACC fissurization and its statistical analyses
In line with previous studies (Paus et al., 1996b; Vogt et al.,
1995; Yücel et al., 2001, 2002), we quantified individual ACC
morphology by registering the occurrence and extension of the
PCS. Anatomical images were analyzed using MRIcro (Rorden
and Brett, 2000). To guarantee rater's blindness to the subject and
hemisphere under investigation, all images were coded and half of
them were left/right flipped.
First, the relevant region of interest (ROI) had to be delineated.
In accordance with Yücel and colleagues, we defined the anterior
border of a sulcus as the turning point of its ventral to dorsal
course. These turning points were specified by lines running
perpendicular to the anterior/posterior commissural plane (AC–PC
plane). The posterior boundary of the ROI was given by another
line perpendicular to the AC–PC plane but crossing the anterior
Second, a protocol was generated for the classification of ACC
folding explicitness to guarantee reliability of these ratings. The
PCS was defined as a sulcus running dorsal and parallel to the CS.
Starting at a mid-sagittal slice the PCS had to be certified on at
least three more lateral slices in the respective hemisphere. This
was done to avoid confounding of hemispheres in the mid-sagittal
view and to distinguish the PCS from other, often more superficial,
sulci such as the intralimbic sulcus. If present, the extension of the
PCS was registered voxelwise yielding a measure in millimeter
resolution. Interruptions or gaps in the PCS course did not
contribute to the length measurements.
These data were then additionally transformed by grouping
individual measurements per hemisphere in steps of 15 mm. This
resulted in five categories of ACC morphology ranging from
absent (0) to prominent (4). This additional transformation to a
discrete representation allows for statistical inferences otherwise
not suitable due to a violation of relevant assumptions. The high
number of cases without a PCS leads to a non-normal distribution.
Non-parametric statistical approaches were thus used to examine
the effects of interest. Specifically, the McNemar test was used to
assess differences between hemispheres regarding the distribu-
tional similarity of PCS absence and presence as it compares paired
proportions. As a generalization of the former statistical method,
the Bowker test examines matched pairs of categorical data with
more than two levels and was applied to the discrete representa-
tions of PCS measurements as described above. A non-parametric
alternative to the paired t-test is the Wilcoxon signed-rank test,
which involves comparisons of differences between dependent
measurements on the interval level. It was employed to directly
appraise differences in PCS extent between hemispheres that not
necessarily have to be reflected in varying distributions of
categorical data. The Kruskal–Wallis H-test is often viewed as
the non-parametric equivalent of the parametric one-way analysis
of variance (ANOVA). Group comparisons were done for each
hemisphere separately with the continuous measurement as
dependent variable. These statistical analyses were conducted
using the software package SPSS 12. In addition to non-parametric
tests of significance, effect size estimates were computed using the
procedures described in Rosenthal (1994). Here, r2are reported.
This measure is an estimate for the explained variance and ranges
from zero to one.
To quantify individual asymmetries in ACC fissurization for
each subject, the relevant parameter for the right was subtracted
from that of the left hemisphere. Therefore, a positive value of this
metric indicates a leftward fissurization asymmetry.
Intra-rater reliability for the classification protocol was
evaluated by repeated measurements of the main rater (R.J.H.)
for 50 randomly chosen anatomical images. After sufficient
training in the assessment of ACC morphology, a second rater
repeated the classification on another 50 randomly selected data
sets for calculation of inter-rater reliability. According to Wirtz and
Caspar (2005), Spearman Rho coefficients were calculated.
Reliabilities for both hemispheres were computed and Fisher Z
transformed. These coefficients did not differ significantly. Mean
reliabilities were 0.89 and 0.92 for between- and within-rater
Optimized voxel-based morphometry
VBM analysis was done with SPM2 developed by the
Wellcome Institute (London, UK) running on MATLAB 6.1
(The MathWorks, Natrick, USA). Images were analyzed following
the “optimized” procedure as described in Mechelli et al. (2005).
Additional modulation was applied to identify differences in
regional gray matter volume rather than concentration by
multiplying the voxel densities with the Jacobian determinants
obtained during spatial normalization. The information about
absolute volumes is thereby preserved. This approach was chosen
based on the assumed associations between regional fissurization
and gray matter volume. Beyond, the volume of the anterior
cingulate cortex has been studied many times by means of manual
tracing methods but still no consensus regarding group or
hemispheric differences has been reached. An oVBM with
modulation is less biased by specific sulcal borders and might
therefore add information to this topic. Some processing steps
were adapted to the needs of this study: templates and a priori
images used were symmetrical and calculated from a subsample
of all images suited for this analysis (N=80; each of the four
subgroups contributing equally). These images were normalized
to match the MNI 305 template and flipped vertically in the mid-
sagittal plane. Then, flipped and unflipped images were averaged.
This symmetrical mean image was smoothed with an 8-mm
FWHM isotropic Gaussian kernel. Tissue-specific templates were
created after normalization and segmentation of the original
images, followed by processing as mentioned. Symmetric a priori
maps were made by flipping and averaging of maps provided with
After modulation, images underwent successive masking
processes to avoid hemispheric confounding in the area of interest.
First, hemispheres were masked out and separately smoothed with
a 12-mm FWHM Gaussian kernel. These images were masked
again to prevent smearings across the midline into the other
hemisphere. Then, both hemispheres were reintegrated into one
data set, from which the left and right ROIs were extracted.
Anterior and posterior boundaries were determined as described for
the fissurizational analyses. Its lower boundary was given by the
corpus callosum, whereas the dorsal border was defined being
890 R.J. Huster et al. / NeuroImage 34 (2007) 888–895
about 5 mm above the average position of the PCS. These regions
finally entered the statistical evaluation.
The statistical analysis was done within the framework of the
general linear model (GLM) in SPM2. Between-group analyses of
GMV were performed with the total volume of the hemispheres as
covariate. To derive estimates of bilateral hemispheric volume,
images were first normalized and segmented. Subsequently, white
and gray matter images were masked to retain the hemispheres
while excluding the cerebellum and brainstem. Inverted deforma-
tion fields were computed from the spatial normalization
parameters and applied to the masked and normalized images.
This procedure resulted in a representation of the segmented
hemispheres in native space. After thresholding these images at a
customized level, the bilateral hemispheric volume was calculated
by addition of gray and white matter voxels.
Contrasts were defined to detect voxels indicative of group
differences in regional GMV. Hemispheric comparisons within
each group were made by left/right flipping and contrasting
unflipped and flipped images. From the resulting statistical
parameter maps, voxelwise measures for the size of the effect of
interest were calculated. Here, values of ω2will be reported. This
coefficient quantifies the degree of systematic or explained
variance. As r2, values of ω2span the range between zero and
one. According to the classification of Kirk (1996), even small
effects were deemed relevant. Effect size maps were thresholded at
0.01 and extension as well as mean ω2of relevant clusters (with
nv>350)2will be reported. Why was this approach chosen? Given
the exploratory nature of this study, an a priori determination of the
sample size was not possible. On the other hand, preprocessing of
the images as well as a correction for multiple comparisons may
inadequately lessen the power of the analysis. Moreover, even
small differences might be relevant when regarding neuroanato-
mical variations between groups. Evaluating the results in terms of
effect sizes is appropriate, as future studies may thus base their
experimental design or interpretation on this report.
Hemispheric ACC folding asymmetry
Overall, the PCS was more often found to be present than absent
in the left hemisphere (75.4% vs. 24.6%), while the respective
a significant McNemar statistic (χ2=17.83, p<0.001; r2=0.13).
Considering all categories (five levels ranging from absent to
prominent) and again comparing the left to the right hemisphere, the
Bowker test confirms a distributional asymmetry (MH=4.23,
p<0.001; r2=0.08). The assumption of a more pronounced PCS
in the left hemisphere (continuously measured) is verified by a
significant Wilcoxon test (z=−4.67, p<0.001; r2=0.15).
Assessing hemispheric differences in regional fissurization in
each of the four groups, a disordinal interaction of sex and
handedness became apparent. Thus, main effects will not be
reported. Female right-handers as well as male left-handers did not
show significant hemispheric differences with respect to the
occurrence of the PCS (McNemar test), nor did the latter
demonstrate an asymmetrical distribution or statistically relevant
discrepancies in PCS length between hemispheres (McNemar and
Wilcoxon tests, respectively). Relevant test statistics are summar-
ized in Table 1. Nonetheless, all subgroups displayed a leftward
fissurization bias (see Fig. 1). Frequencies and percentages
specifying the PCS development are presented in Table 2.
Group differences per hemisphere
Whether left or right ACC folding contributes to the observed
interaction of sex and handedness was assessed by means of
Kruskal–Wallis H-test. This analysis was conducted on the
continuous data and separately for each hemisphere, yielding
statistically relevant group differences in the left (χ2(3)=16.13,
p≤0.001; r2=0.11) but not in the right anterior cingulate region
(r2=0.02). To further clarify this effect, post hoc comparisons were
computed according to the procedure advised by Conover (1971,
1980). Here, right-handed as well as left-handed males signifi-
cantly differed from all other groups (comparisons based on critical
mean rank differences corresponding to the significance level of
p≤0.05), with highest and lowest mean ranks for these subsamples,
respectively (Table 3). Again, due to the pattern of contrasts
indicating a disordinal interaction, an effect of handedness was not
taken into further consideration. Distributional differences between
groups in the categorical data were assessed by means of
Kolmogorov–Smirnov Z tests, none of which remained significant
after the Bonferroni correction.
Optimized voxel-based morphometry
Hemispheric asymmetry per group
Contrasting unflipped and flipped modulated images for each
group, relevant clusters in the left ACC indicate left-bigger-than-
right and those in the right hemisphere right-bigger-than-left
volumetric differences. In all groups, clusters of substantial size
emerged in both the left and right cingulate regions. Clusters
indicating bigger volume in the right than the left cingulate showed
a wider extension in all groups. Notably, right-handed males
displayed the strongest effects and widest spatial extension
regarding the left-bigger-than-right cingulate comparison (see
Fig. 2). Differences in the left hemisphere were clearly located in
the sulci, while those in the right cingulate region were more
focused near the gyral crests. Table 4 summarizes these results.
Effects of sex and handedness on regional gray matter volume
Right- and left-handed male and female subjects were
compared while controlling for global neocortical volumes. A
substantial cluster indicated a gender difference with higher
2nvindicates the number of voxels above the chosen threshold.
Hemispheric differences in regional fissurization
PCSOccurrence Distribution Length
GroupMcNemar test Bowker testWilcoxon test
891R.J. Huster et al. / NeuroImage 34 (2007) 888–895
regional gray matter volume in females in the right dorsal ROI
(nv=2016, mean ω2=0.02, max ω2=0.06). Besides, right-handers
as compared to left-handers showed higher gray matter volume in
two clusters in the left hemisphere (nv=1576, mean ω2=0.02, max
ω2=0.04; nv=1827, mean ω2=0.02, max ω2=0.07). Assessing the
interaction of sex and handedness, right-handed males showed
higher GMV than the other groups in two regions of the left dorsal
(nv=498, mean ω2=0.02, max ω2=0.03; nv=950, mean ω2=0.03,
max ω2=0.06) and one cluster in the right perigenual area
(nv=961, mean ω2=0.03, max ω2=0.09, see also Fig. 3).
Overall, the PCS occurred more often and was more
pronounced (as indicated by distributional and length differences)
in the left compared to the right anterior cingulate region.
Accordingly, all groups of interest (left- or right-handed males
and females) displayed a bias towards a leftward asymmetry
relative to symmetric or rightward folding patterns. Nonetheless,
significant hemispheric differences on all levels of analysis
occurred in male right- and female left-handers, while discrepan-
cies were diminished in male left- and female right-handers (but
notice the differentiation in the latter group with respect to
measures of PCS occurrence and extension). This interaction-like
pattern in the asymmetry metrics and hemispheric comparisons
within groups seems to stem from variations of cingulate
morphology between groups in the left rather than the right
An overall asymmetry in fissurization of the anterior cingulate
was reported by qualitative post-mortem (Ide et al., 1999; Vogt et
al., 1995) as well as quantitative MRI-based studies (Paus et al.,
1996b; Yücel et al., 2001). Our results are consistent with these
findings. Moreover, this leftward asymmetry bias was more
common compared to a symmetric or rightward folding pattern
in all four subgroups. Gender differences observed in right-handed
participants replicate the findings of Yücel et al. (2001), while
corresponding effects of sex in left-handers have not been
investigated before. Here, the observed inversion in left-handers
mirrors results of studies on other anatomical structures such as the
planum temporale (Dos Santos Sequeira et al., 2006) or the
precentral area (Amunts et al., 2000). Hemispheric differences in
the cingulate region due to variations in cortical folding of the left
but not the right ACC are also found in schizophrenics (Yücel et
al., 2002). Here, patients lacked the sulcal leftward asymmetry,
which was explained by reduced fissurization of the left anterior
Fig. 1. Distribution of individual folding patterns of PCS occurrence and
extent (asymmetry index: left PCS–right PCS). Those cases were collapsed
showing a leftward asymmetric, rightward asymmetric or a symmetric
fissurization. In all groups, the leftward asymmetric pattern is the most
common when compared to a symmetric or rightward folding pattern.
Mean ranks of PCS length per group and hemisphere
Fig. 2. Hemispheric comparison in right-handed males. Colored areas in the
left hemisphere depict highergray matter volume in the left cingulate region,
while those in right hemisphere indicate the reversed pattern. The effect size
is reported as ω2thresholded at 0.01.
PCS classification in right- and left-handed males and females
PCS Absent Prominent
Frequencies and percentages (in parentheses; rounded to the first decimal) of
PCS classifications in the left and right hemisphere; the categories range
from 0 (absent, PCS<15 mm) to 4 (prominent, PCS>60 mm) in steps of
892 R.J. Huster et al. / NeuroImage 34 (2007) 888–895
Work on the ontogeny of gyrification suggests an association
between gross morphological and variations on the microstructural
level. Studies of Zilles et al. (1996) and Armstrong and colleagues
(1995) corroborate a mechanical hypothesis of cortical folding.
Mechanical forces generated by differently sized strata within the
cortical mantle are believed to drive the degree of gyrification. In
contrast, Welker (1990) assumes stronger growth processes in one
compared to a neighboring cortical field as major determinant of
folding, while the placement, orientation and depth of sulci may
additionally be influenced by characteristics of underlying
fasciculi. The already cited work of Vogt et al. (1995) additionally
supports the notion of a correlation between cingulate macro- and
microstructural features. As the distribution of area 32(V)differs
between cases with and without PCS, a prominent PCS seems to be
indicative of a more extended area 32(V)and therefore is likely to
denote a bigger regional GMV in the ACC. Whatever the exact
nature of microstructural characteristics underlying regional
fissurization is, results point to substantial differences between
groups in the structural organization of the ACC.
ACC gray matter volume
By comparing flipped and unflipped images per group, we
found relevant clusters indicating a left bigger than right
volumetric relationship in our region of interest. Although this
effect was present in all subsamples, it was most pronounced
regarding spatial extent and variance explained in right-handed
males. Interestingly, those clusters indicating higher GMV in the
left hemisphere approximated regions in the cingulate and
paracingulate sulci. This emphasizes the existence of an associa-
tion between local fissurization and GMV and is in accordance
with theories on cortical fissurization. Nevertheless, in each group
a higher number of voxels indicated an increased gray matter
volume in the right relative to the left anterior cingulate region.
These regions were strongly oriented towards the medial surface.
This pattern of results raises the questions whether more subtle
cortical asymmetries might exist than those normally reported.
While ROI measurements based on sulcal or gyral borders
necessarily reveal large scale volumetric characteristics, oVBM
analyses may be more sensitive to variations on a lower
neuroanatomical level as detailed information regarding precise
boundaries of a region is not requisite.
Group comparisons regarding regional GMV also indicated an
interaction with higher left cingulate volumes in right-handed
males compared to the other subsamples. This result again is
consistent with those obtained from fissurizational analyses.
Moreover, main effects of handedness and gender emerged with
higher gray matter volumes in the left and right ROI respectively.
These comparisons, on the other hand, are not in full agreement
with the results obtained from the analyses on ACC fissurization.
The regional amount of gray matter probably results from an
interaction of a variety of different factors. These may play a role in
early developmental cellular events such as proliferation, migra-
tion, aggregation or programmed cell death (see Toro and Burnod,
2005). Furthermore, postnatal factors also exert influence on the
development of the cortical mantle (Armstrong et al., 1995).
Individual experiences, nutritional factors and the development of
axonal connections are major candidates, just to name a few. Still,
the exact processes and its interactions driving cortical fissurization
and gray matter development need further clarification. Reviewing
the literature in that field is beyond the scope of this article (but see
Toro and Burnod, 2005; Armstrong et al., 1995; Chi et al., 1977;
Van Essen, 1997). Given these complexities of maturational
processes, the assumed association between regional fissurization
and GMV is probably far from perfect. Moreover, it is reasonable
to assume an additional decline regarding the interrelations of
abovementioned factors during aging.
Nonetheless, we found substantial volumetric differences in the
anterior cingulate gray matter with the observed patterns under-
scoring the existence of an association between fissurizational and
gray matter characteristics on the group level. These results also
indicate that hemispheric asymmetries with respect to GMV might
even be more specific than initially assumed. Methodological
differences between studies seem to play a major role contributing
to divergent results. Modern applications of MRI allow for the
analysis of a variety of cortical attributes such as surface area,
cortical thickness, GM volume, sulcal depth, regional gyrification
or curvature (see Rettmann et al., 2006). It would not be surprising
if not all of these measures showed the same pattern of asymmetry.
Even within the same methodological approach, differences
between studies are to be expected due to variations in the image
preprocessing. For example, the use of a stronger or weaker
smoothing kernel could bias the results towards lower or more
pronounced regional effects in the left cingulate region, respec-
tively. Obviously, more research is needed to solve at least some of
the issues just outlined (an interesting discussion of some
methodological challenges can be found in Luders et al., 2006).
One additional point in the volumetric analysis might be of
interest. Comparisons indicating the expected effects in the left
anterior cingulate also displayed a cluster of higher GMV in the
VBM: volumetric anterior cingulate asymmetry
Cluster size is measured in number of relevant voxels (1 mm cubic); also
reported are values of mean and maximal ω2for each cluster.
Fig. 3. Clusters of voxels indicating higher gray matter volume in right-
handed males as compared to the other groups. Images are thresholded at
893R.J. Huster et al. / NeuroImage 34 (2007) 888–895
right perigenual ACC. Future studies will have to replicate this
finding. It would be interesting to see whether a leftward
asymmetry in midcingulate areas goes hand in hand with a
bihemispheric effect or a reversed asymmetry in the right
perigenual region. This finding would further corroborate those
models favoring the distinction of a cognitive and an emotional
subdivision of the ACC (e.g., Bush et al., 2000). Delineations of
the ACC in the studies of Paus and Yücel approximated the
borders of the MCC. We chose the definition of the region of
interest in accordance with these studies (more specifically with the
one of Yücel et al., 2001). This was done to protect against
differential results caused by variations in relevant borders defined.
Reviewing the literature in this respect, functional as well as
cytoarchitectural surveys suggest that both anterior and posterior
borders should probably be moved caudally by about 10 mm to
more precisely approximate the MCC and distinguish it from the
perigenual ACC (Bush et al., 2000; Vogt et al., 2003). Nonetheless,
it is unlikely that these changes would affect the observed patterns
in fissurization and GMV to a high degree because the PCS as well
as cingulate clusters indicating volumetric differences between
groups were most often found in midcingulate portions.
Taken together, findings from different scientific groups and
fields of research underscore the existence of cytoarchitectural
variations as well as differences in the amount of local gray matter
volume associated with the sulcal patterning of the ACC. One
might therefore hypothesize about qualitative or quantitative
differences in cognitive and neuropsychological functioning
between subjects with and without a PCS. This seems to be
interwoven with the question of group differences due to sex or
handedness in processes tapped by the ACC. Presently, the exact
contributions of this region to cognition are under investigation and
need clarification. Given that the occurrence of a PCS might
indicate a relative expansion of area 32(V), Paus and colleagues
(1996b) speculated about the anterior cingulate asymmetry being a
phenomenon similar to that found in the planum temporale. An
association of morphological cingulate asymmetry and language
lateralization was therefore expected. Their assumption was based
on findings showing area 32(V)involvement in vocalizations of
primates. Further support came from data of functional activation
studies during verbal tasks (Paus et al., 1996b) showing a cluster of
activations that matched the paracingulate region.
However, findings published after those of Paus and colleagues
advice reconsideration. Functional imaging during performance of
flanker task variants reveals activations in area 32(V). Importantly,
two studies report such results although the tasks did not require
processing or responding in the verbal domain (Casey et al., 2000;
Durston et al., 2003). Studies directly assessing the effects of
processing domain or response modality on the pattern of
activations also lack evidence for such a differentiation in the
ACC. Using a spatial Stroop task with conditions tapping both
verbal and spatial processing as well as responding, Barch et al.
(2001) found the MCC to respond in an unspecific manner.
Hazeltine and colleagues (2003) drew the same conclusion based
on a different operationalization utilizing manual responses to
colors and letters as task-relevant stimuli in a flanker paradigm.
Even though modality-specific effects were found in prefrontal
areas, anterior cingulate regions responded similarly to conflict
with colors and letters. The most recent study addressing this
question (Mitchell, 2005) again found no effect of task-relevant
information in the anterior cingulate when comparing responses to
color, number and shape Stroop tasks.
Nevertheless, there is evidence generally supporting the
functional relevance of morphologic asymmetry in the ACC. For
example, schizophrenic patients and young men at high-risk of
developing a psychotic illness show a substantial reduction of
sulcal asymmetry in the ACC (Le Provost et al., 2003; Yücel et al.,
2002, 2003). While a direct association between these neuroana-
tomical abnormalities and positive or negative symptoms was not
found in the study of Le Provost et al. (2003), many PETand fMRI
studies have shown decreases of ACC activations in schizophrenia
(van Veen and Carter, 2002). Direct evidence is added by Fornito et
al. (2004): subjects with varying PCS asymmetry were compared
in tasks engaging executive cognitive processes to differing
degrees. Those subjects displaying a leftward asymmetric pattern
showed superior performance in both non-verbal and verbal
executive tasks. This difference between groups was not apparent
in tasks which more strongly engaged memory processes.
To summarize, prior findings of gender differences regarding
anterior cingulate fissurization could be replicated. As was found
when assessing neuroanatomical attributes of other cortical regions,
the fissurizational pattern observed in right-handed subjects was
reversed in left-handers. Group differences in regional gray matter
volume similar to those just described were detected in the left
cingulate sulci. Nevertheless, overall the effects demonstrate a
higher global GMV in the right hemisphere, as was also frequently
found for the frontal lobes. Whether more subtle asymmetries do in
fact exist needs further clarification. Current findings suggest a
functional relevance of such morphologic variations, but their exact
impact on cognitive functioning still has to be investigated.
This work was supported by the Alfried Krupp von Bohlen und
Halbach-Stiftung (International Center for Integrated Neu-
roscience, Alfried Krupp Wissenschaftskolleg).
Allen, J.S., Damasio, H., Grabowski, T.J., Bruss, J., Zhang, W., 2003.
Sexual dimorphism and asymmetries in the gray-white composition of
the human cerebrum. NeuroImage 18, 880–894.
Amunts, K., Jancke, L., Mohlberg, H., Steinmetz, H., Zilles, K., 2000.
Interhemispheric asymmetry of the human motor cortex related to
handedness and gender. Neuropsychologia 38, 304–312.
Armstrong, E., Schleicher, A., Omran, H., Curtis, M., Zilles, K., 1995. The
ontogeny of human gyrification. Cereb. Cortex 5, 56–63.
Ashburner, J., Friston, K.J., 2000. Voxel-based morphometry—The
methods. NeuroImage 11, 805–821.
Barch, D.M., Braver, T.S., Akbudak, E., Conturo, T., Ollinger, J., Snyder,
A., 2001. Anterior cingulate cortex and response conflict: effects
of response modality and processing domain. Cereb. Cortex 11,
Botvinick, M.M., Cohen, J.D., Carter, C.S., 2004. Conflict monitoring and
anterior cingulate cortex: an update. Trends Cogn. Sci. 8, 539–546.
Bush, G., Luu, P., Posner, M.I., 2000. Cognitive and emotional influences in
anterior cingulate cortex. Trends Cogn. Sci. 4, 215–222.
Bush, G., Vogt, B.A., Holmes, J., Dale, A.M., Greve, D., Jenike, M.A.,
Rosen, B.R., 2002. Dorsal anterior cingulate cortex: a role in reward-
based decision making. Proc. Natl. Acad. Sci. U. S. A. 99, 523–528.
Casey, B.J., Thomas, K.M., Welsh, T.F., Badgaiyan, R.D., Eccard, C.H.,
894 R.J. Huster et al. / NeuroImage 34 (2007) 888–895
Jennings, J.R., Crone, E.A., 2000. Dissociation of response conflict, Download full-text
attentional selection, and expectancy with functional magnetic reso-
nance imaging. Proc. Natl. Acad. Sci. U. S. A. 97, 8728–8733.
Chi, J.G., Dooling, E.C., Gilles, F.H., 1977. Gyral development of the
human brain. Ann. Neurol. 1, 86–93.
Conover, W.J., 1971, 1980. Practical nonparametric statistics. New York:
Devinsky, O., Morrell, M.J., Vogt, B.A., 1995. Contributions of anterior
cingulate cortex to behaviour. Brain 118 (Pt. 1), 279–306.
Dos Santos Sequeira, S., Woerner, W., Walter, C., Kreuder, F., Lueken, U.,
Westerhausen, R., Wittling, R.A., Schweiger, E., Wittling, W., 2006.
Handedness, dichotic-listening ear advantage, and gender effects on
planum temporale asymmetry—A volumetric investigation using
structural magnetic resonance imaging. Neuropsychologia 44,
Durston, S., Davidson, M.C., Thomas, K.M., Worden, M.S., Tottenham, N.,
Martinez, A., Watts, R., Ulug, A.M., Casey, B.J., 2003. Parametric
manipulation of conflict and response competition using rapid mixed-
trial event-related fMRI. NeuroImage 20, 2135–2141.
Fan, J., Flombaum, J.I., McCandliss, B.D., Thomas, K.M., Posner, M.I.,
2003. Cognitive and brain consequences of conflict. NeuroImage 18,
Fornito, A., Yücel, M., Wood, S., Stuart, G.W., Buchanan, J.A., Proffitt, T.,
erson, V., Velakoulis, D., Pantelis, C., 2004. Individual differences in
anterior cingulate/paracingulate morphology are related to executive
functions in healthy males. Cereb. Cortex 14, 424–431.
Hazeltine, E., Bunge, S.A., Scanlon, M.D., Gabrieli, J.D., 2003. Material-
dependent and material-independent selection processes in the frontal
and parietal lobes: an event-related fMRI investigation of response
competition. Neuropsychologia 41, 1208–1217.
Ide, A., Dolezal, C., Fernandez, M., Labbe, E., Mandujano, R., Montes, S.,
Segura, P., Verschae, G., Yarmuch, P., Aboitiz, F., 1999. Hemispheric
differences in variability of fissural patterns in parasylvian and cingulate
regions of human brains. J. Comp. Neurol. 410, 235–242.
Kirk, R.E., 1996. Practical significance: a concept whose time has come.
Educ. Psychol. Meas. 56 (5), 746–759.
Kopelman, A., Andreasen, N.C., Nopoulos, P., 2005. Morphology of the
anterior cingulate gyrus in patients with schizophrenia: relationship to
typical neuroleptic exposure. Am. J. Psychiatry 162, 1872–1878.
Laird, A., McMillan, K., Lancaster, J., Kochunov, P., Turkeltaub, P., Pardo,
J., Fox, P., 2005. A comparison of label-based review and ALE meta-
analysis in the Stroop task. Hum. Brain Mapp. 25, 6–21.
Le Provost, J.B., Bartres-Faz, D., Paillere-Martinot, M.L., Artiges, E.,
Pappata, S., Recasens, C., Perez-Gomez, M., Bernardo, M., Baeza, I.,
Bayle, F., Martinot, J.L., 2003. Paracingulate sulcus morphology in men
with early-onset schizophrenia. Br. J. Psychiatry 182, 228–232.
Luders, E., Narr, K.L., Thompson, P.M., Rex, D.M., Jancke, L., Toga, A.W.,
2006. Hemispheric asymmetries in cortical thickness. Cereb. Cortex. 16,
Mathalon, D.H., Whitfield, S.L., Ford, J.M., 2003. Anatomy of an error:
ERP and fMRI. Biol. Psychol. 64, 119–141.
Mechelli, A., Price, C., Ashburner, J., 2005. Voxel-based morphometry of
the human brain: methods and applications. Curr. Med. Imaging Rev. 1,
Mitchell, R.L., 2005. The BOLD response during Stroop task-like inhibition
paradigms: effects of task difficulty and task-relevant modality. Brain
Cogn. 59, 23–37.
Oldfield, R.C., 1971. The assessment and analysis of handedness: the
Edinburgh inventory. Neuropsychologia 9, 97–113.
Paus, T., Otaky, N., Caramanos, Z., MacDonald, D., Zijdenbos, A.,
D'Avirro, D., Gutmans, D., Holmes, C., Tomaiuolo, F., Evans, A.C.,
1996a. In vivo morphometry of the intrasulcal gray matter in the human
cingulate, paracingulate, and superior-rostral sulci: hemispheric asym-
metries, gender differences and probability maps. J. Comp. Neurol. 376,
Paus, T., Tomaiuolo, F., Otaky, N., MacDonald, D., Petrides, M., Atlas, J.,
Morris, R., Evans, A.C., 1996b. Human cingulate and paracingulate
sulci: pattern, variability, asymmetry, and probabilistic map. Cereb.
Cortex 6, 207–214.
Rettmann, M.E., Kraut, M.A., Prince, J.L., Resnick, S.M., 2006. Cross-
sectional and longitudinal analyses of anatomical sulcal changes
associated with aging. Cereb. Cortex 16, 1584–1594.
Rorden, C., Brett, M., 2000. Stereotaxic display of brain lesions. Behav.
Neurol. 12, 191–200.
Rosenthal, R., 1994. Parametric measures of effect size. In: Cooper, H.,
Hedges, L.V. (Eds.), The Handbook of Research Synthesis. Russell Sage
Foundation, New York, pp. 231–244.
Toro, R., Burnod, Y., 2005. A morphogenetic model for the development of
cortical convolutions. Cereb. Cortex 15, 1900–1913.
Van Essen, D.C., 1997. A tension-based theory of morphogenesis and
compact wiring in the central nervous system. Nature 385, 313–318.
van Veen, V., Carter, C.S., 2002. The anterior cingulate as a conflict monitor:
fMRI and ERP studies. Physiol. Behav. 77, 477–482.
van Veen, V., Carter, C.S., 2005. Separating semantic conflict and response
conflict in the Stroop task: a functional MRI study. NeuroImage 27,
Vogt, B.A., Nimchinsky, E.A., Vogt, L.J., Hof, P.R., 1995. Human cingulate
cortex: surface features, flat maps, and cytoarchitecture. J. Comp.
Neurol. 359, 490–506.
Vogt, B.A., Berger, G.R., Derbyshire, S.W., 2003. Structural and functional
dichotomy of human midcingulate cortex. Eur. J. Neurosci. 18,
Welker, W.I., 1990.The significance of foliationand fissurationof cerebellar
cortex. The cerebellar folium as a fundamental unit of sensorimotor
integration. Arch. Ital. Biol. 128, 87–109.
Wirtz, M., Caspar, F., 2005. Beurteilerübereinstimmung und Beurteilerre-
liabilität. Hogrefe, Göttingen.
Yücel, M., Stuart, G.W., Maruff, P., Velakoulis, D., Crowe, S.F., Savage, G.,
Pantelis, C., 2001. Hemispheric and gender-related differences in the
gross morphology of the anterior cingulate/paracingulate cortex in
normal volunteers: an MRI morphometric study. Cereb. Cortex 11,
Yücel, M., Stuart, G.W., Maruff, P., Wood, S.J., Savage, G.R., Smith, D.J.,
Crowe, S.F., Copolov, D.L., Velakoulis, D., Pantelis, C., 2002.
Paracingulate morphologic differences in males with established
schizophrenia: a magnetic resonance imaging morphometric study.
Biol. Psychiatry 52, 15–23.
Yücel, M., Wood, S.J., Phillips, L.J., Stuart, G.W., Smith, D.J., Yung, A.,
Velakoulis, D., McGorry, P.D., Pantelis, C., 2003. Morphology of the
anterior cingulate cortex in young men at ultra-high risk of developing a
psychotic illness. Br. J. Psychiatry 182, 518–524.
Zilles, K., et al., 1996. Structural asymmetries in the human forebrain and
the forebrain of non-human primates and rats. Neurosci. Biobehav. Rev.
895R.J. Huster et al. / NeuroImage 34 (2007) 888–895