Voxel-based morphometric analysis on the volume of gray matter in bipolar I disorder

Psychiatry Research: Neuroimaging 191 (2011) 92–97

Contents lists available at ScienceDirect

Psychiatry Research: Neuroimaging j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p s yc h r e s n s

Voxel-based morphometric analysis on the volume of gray matter in bipolar I disorder Mingli Lia, Liqian Cuia, Wei Denga, Xiaohong Maa, Chaohua Huang a, Lijun Jianga, Yingcheng Wanga, David A. Collier d, Qiyong Gongb, Tao Lia,c,d,⁎ a

The Department of Psychiatry & Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Sichuan 610041, China MRI Center, West China Hospital, Sichuan University, Sichuan, China King's College London, Department of Psychiatry and Psychological Medicine, Institute of Psychiatry, London, UK d King's College London, MRC Social Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, London, UK b c

a r t i c l e

i n f o

Article history: Received 14 December 2009 Received in revised form 29 August 2010 Accepted 9 September 2010 Keywords: Bipolar I disorder Magnetic resonance imaging Brain structure

a b s t r a c t A number of previous studies have found that bipolar disorder is associated with abnormalities of brain structure. In this study we used optimized voxel-based morphometry (VBM) to compare gray matter volume between patients with bipolar I disorder and healthy controls. Twenty-four bipolar I patients (15 males and nine females) and 36 healthy controls (21 males and 15 females), who were well matched for age and gender, were scanned using structural magnetic resonance imaging. Gray matter volume was assessed and compared using optimized VBM, and the correlation between duration of illness/number of episodes and regional volumes was analyzed. There was no difference in whole-brain gray matter volume between the two groups. Optimized vVBM showed that subjects with bipolar I disorder had smaller volumes in the left inferior parietal lobule, right superior temporal gyrus, right middle frontal gyrus and left caudate. Only the volume of the right middle frontal gyrus was correlated with duration of illness and number of episodes in patients. These results suggest widespread gray matter defects in bipolar I disorder, which may play an important role in onset of the illness. © 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Bipolar disorder (BD) is a common psychiatric disorder characterized by the core feature of recurrence of hypomanic or manic and depressive episodes that seriously affect the quality of life and social functions of patients. Twin studies show that BD is highly heritable, but only a few risk variants have been isolated thus far (van der Schot et al., 2009) and its etiology is still unknown. In recent years, the rapid progress in imaging techniques, especially non-invasive magnetic resonance imaging (MRI), has provided new approaches for the investigation into the etiology of mental illness. Previously, the brain structures of the prefronto-subcortical and the anterior limbic networks were thought to play a critical role in bipolar disorder. It had been suggested that bipolar disorder might result from structural abnormalities in this circuit and subsequent impairment of emotional regulation (Strakowski et al., 2005). However, some studies also found that, in addition to the prefrontal lobe–limbic system, local structures including the parietal lobe and the temporal lobe were also involved (Adler et al., 2005; Frazier et al., ⁎ Corresponding author. The Department of Psychiatry & Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Sichuan 610041, China. Tel.: +86 28 85423561; fax: +86 28 85164019. E-mail address: [email protected] (T. Li). 0925-4927/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pscychresns.2010.09.006

2005; Nugent et al., 2006; Chen et al., 2007). During the last 2 years, a meta-analysis on bipolar disorder indicated that there were robust changes of brain structure in patients with bipolar disorders (Kempton et al., 2008), but the abnormal regions were inconsistent. Lower total intracranial and white matter volumes were revealed in first-episode bipolar patients (Vita et al., 2009). Arnone et al. indicated bipolar disorder was characterized by lower whole-brain and prefrontal lobe volume as well as higher globus pallidus and lateral ventricular volume (Arnone et al., 2009). A systematic review of voxel-based morphometry studies showed less gray matter was present in the anterior cingulate and bilateral insula in bipolar patients (Ellison-Wright and Bullmore, 2010). These different findings might stem from the heterogeneity of subjects and the methodological differences. Bipolar disorder was divided into four subtype categories: bipolar I disorder (BD I), bipolar II disorder (BD II), cyclothymic disorder, and bipolar disorder not otherwise specified according to Diagnostic and Statistical Manual of Mental Disorders IV edition (DSM-IV). Patients with bipolar disorder were often found to have other comorbidities, such as anxiety, substance use, and eating disorders (McElroy et al., 2001). Two neuroimaging studies that compared BD I and BD II found that the volume of the lateral ventricles and the ratio of the lateral ventricles to the cerebrum were significantly larger in BD I patients (Hauser et al., 2000), and the two clinical subtypes had different

M. Li et al. / Psychiatry Research: Neuroimaging 191 (2011) 92–97

patterns of gray matter abnormalities (Ha et al., 2009), which indicated the two clinical subtypes might have different neurobiological abnormalities and characteristics. Methodologies employed in investigations on brain structure in vivo involve largely involve the delineation of regions of interest (ROI) and voxel-based morphometry (VBM) (Giuliani et al., 2005). In the ROI approach, investigators delineate ROIs on images and thereby calculate the volume accordingly. It is partially dependent on the hypothesis of the disease and the investigators' understanding on brain anatomical structures. In VBM, point-wise comparison of the whole brain between different subjects is performed by computer, which then calculates the density or volume of different structures. Such a method is applicable to the whole brain rather than limited specific regions. In addition, optimized VBM does a better job in resolving the problem of segmentation error in traditional VBM and produces more precise results by volume modulation (Good et al., 2001). The aim of the present study was to explore the structure of cerebral gray matter in patients with BD I without any comorbidities, using structural MRI and optimized VBM. The hypothesis for the study was that the difference of cerebral gray matter volume between patients and controls might not be confined to one cerebral region, but might also involve several local cerebral structures closely connected to each other. 2. Study subjects and methods 2.1. Subjects Patients with bipolar I disorder (BDI) group: Twenty-four patients hospitalized in the Mental Health Center, West China Hospital of Sichuan University from November 2006 to April 2007 were included in the study. On the basis of the Structured Clinical Interview for DSM-IV (SCID-IP), they were found to fulfill the diagnostic criteria of bipolar I disorder as described in the DSM-IV. Of these patients, 15 were males and nine were females. Only two of patients were first onset. All of them were on medication at the time of examination, such as mood stabilizers (lithium carbonate and sodium valproate) and clozapine (Table 1). But they had never been treated with electroconvulsive therapy. The patients were in a relatively stable condition when they were scanned (i.e. they could follow the doctor's instruction to keep still during the process of scanning). Patients with other mental diseases, nervous system diseases, severe physical diseases, personality disorders and abuse of alcohol and drugs were excluded. Healthy control (HC) group: A total of 36 healthy controls were collected by recruiting volunteers were recruited, 21 males and 15 females. They were screened with the SCID-IP by an experienced psychiatrist. Subjects with any psychiatric, mental or severe physical diseases were excluded.

93

Both groups were Han Chinese and were right-handed. This study was approved by the Institutional Review Broad (IRB) of West China Hospital, Sichuan University. Written informed consents were obtained from all participants. 2.2. Methods Imaging variables: All participants underwent MRI scans in the Department of Radiology at West China Hospital with a Signa 3.0-T scanner (GE Medical Systems, Milwaukee, WI) with an eight-channel phase array head coil. High-resolution T1 images were acquired by 3D spoiled gradient echo sequence (SPGR) from all participants. The sequence used in this protocol was as follow: TR = 8.5 ms; TE= 3.93 ms; flip angle = 12°; thickness of slice = 1 mm; single shot; field of view (FOV) = 24 cm × 24 cm; matrix = 256 × 256; size of voxel = 0.47 × 0.47 × 1 mm3; in total, 156 slices of axial images were collected from a brain. 2.2.1. Pre-processing of the data DICOM format data collected from magnetic resonance scanning were transferred into Analyze format data in MRIcro software (Rorden and Brett, 2000). Then, the non-parametric non-uniformity intensity normalization technique (N3) in the MINC software package (http:// wiki.bic.mni.mcgill.ca/index.php/MINC) was used to rectify the nonuniformity of high-magnetic field signal in acquired images. Thereafter, the Brain Extraction Tool (BET) in FSL software was used to remove nonbrain tissue and structure in the images (Smith, 2002), so as to produce native images for further processing. On the Matlab7.0 platform, Statistical Parametric Mapping (SPM) software and the optimized VBM method were used for further processing of the data. Construction of an individual gray matter template was performed using SPM2 software, the native images of 60 subjects underwent tissue segmentation in native space, and images of gray matter, white matter and cerebrospinal fluid were obtained, respectively; gray matter images obtained from each subject underwent non-linear affine transformation and registration to the gray matter template (gray.mnc) in standard ICBM (International Consortium for Brain Mapping) space; gray matter images of all subjects were normalized and acquired parameters were applied to native images; tissue segmentation was again performed on normalized images to obtain images of gray matter, white matter and cerebrospinal fluid; an individual gray matter template was therefore constructed by averaging three types of segmented images from each of the 60 subjects and then Gaussian-smoothing with a kernel of 8 mm FWHM (Full Width at Half Maximization). Native images were normalized and segmented by the integrated generation mode (unified segmentation; Ashburner and

Table 1 Demographic data.

Age (years) Education duration (years) Gender (male/female) Right-handed Onset age (years) Disease duration (months) Number of episodes Current status (manic/depression) Young Mania Rating Scale Hamilton Depression Scale Medicine use, n (%) Lithium carbonate Sodium valproate Clozapine Others

Bipolar I disorder patient group (n = 24)

Healthy control group (n = 36)

T/χ2 value

P value

28.42(6.64) 13.21(3.09) 15/9 24 22.38(6.64) 72.00(99.75)* 3.00(3.50)* 17/7 25.94(4.49) 20.00(4.32)

26.56(6.70) 13.06(2.32) 21/15 36

−1.069 −0.219 0.104

0.289 0.828 0.747

20(83.33) 13(54.17) 11(45.83) 8(33.33)

Note: Demographic data are shown as mean value (standard deviation) except where otherwise specified: * shown as median (inter-quartile range).

94

M. Li et al. / Psychiatry Research: Neuroimaging 191 (2011) 92–97

Friston, 2005) in SPM5 software. During this process, native images underwent non-linear transformation and registration to individual gray matter templates, and then gray matter images were obtained by segmentation using an iterative algorithm; re-obtained gray matter images underwent Jacobian modulation (volume modulation) and then whole-brain gray matter volume was calculated; after volume modulation, the gray matter images were Gaussian-smoothed with a kernel of 6 mm FWHM. Obtained images were used for comparison of gray matter volume. 2.2.2. Statistical analysis On the whole-brain level, whole-brain gray matter volume of two groups was compared by group t test in SPSS13.0 statistical software, and statistical significance was indicated when P b 0.05. On the regional VBM level, SPM5 software was used to design the matrix for comparing the BD I group and the HC group, with whole-brain gray matter volume, gender and age as covariates. Threshold masking, which was set at the absolute value of 0.1, and explicit masking, which was constructed from the 60 subjects by using “SPM Masking Toolbox” (Ridgway et al., 2008), were applied to ensure only voxels within the gray matter mask were analyzed. Based on multiple comparison correction using False Discovery Rate and a P value of b0.05 for each voxel, cerebral regions with more than 200 linked voxels (cluster of voxels) and a Pcorrected value of b0.05 (cluster level) were regarded as differential regions. In the patient group, using the ‘volume of interest’ (VOI) function (spm_regions.m) in SPM5, the eigenvariate gray matter volume values within a 10-mm radius sphere centered at the peak voxel in each differential region revealed by VBM were extracted and used for Spearman rank correlation analysis with the disease duration and number of episodes, and for Pearson correlation analysis, with the onset age in the patients by SPSS13.0 statistical software.

3.2. Whole-brain gray matter volume (WGV) in subjects of two groups The average WGV was 0.7088 (S.D. = 0.0763) L in BD I patients and 0.7265 (S.D. = 0.0602) L in controls, respectively. Group t test revealed a t value of 1.0022, degrees of freedom (d.f.) value of 58, and P value of 0.3204, which indicated no statistical difference. And the WGV was neither correlated to disease duration nor number of episodes. 3.3. Cerebral regions with statistically different gray matter volume between BD I patients and controls When comparing BD I patients to healthy controls, cerebral regions with less gray matter volume were mainly distributed in the left inferior parietal lobule (x = −51, y = −35, z = 43, T = 5.46, Pcorrected = 0.0348), right superior temporal gyrus (x = 55, y = −37, z = 5, T = 5.36, Pcorrected = 0.0420; x = 56, y = −50, z = 13, T = 5.32, Pcorrected = 0.0136), right middle frontal gyrus (x = 36, y = 47, z = 13, T = 5.07, Pcorrected = 0.0114) and left caudate nucleus (x = −8, y = 12, z = 9, T = 4.99, Pcorrected = 0.0015). No cerebral regions with higher gray matter volume were identified in BD I patients. See Table 2 and Fig. 1. 3.4. Rank correlation analysis between differential region volumes and disease duration in BD I patients The volume of the right middle frontal gyrus was positively correlated to disease duration in BD I patients, with a correlation coefficient of 0.5442 and a P value of 0.0060 (d.f. = 22). No correlation was identified between disease duration and volume of other regions including left inferior parietal lobule, right superior temporal gyrus and left caudate nucleus (Table 3). 3.5. Rank correlation analysis between differential region volumes and the number of episodes in BD I patients

3. Results 3.1. Demographic characteristics The demographic characteristics of the participants are shown in Table 1. The patients with BD I were aged between 16 and 39 (mean age = 28.42, S.D. = 6.64) years; onset age was 13 to 31 (mean age = 22.38, S.D. = 5.82) years; education duration ranged from 7 to 19 (mean duration = 13.21, S.D. = 3.09) years; disease duration ranged from 1 to 240 months; the number of episodes varied from one to seven. At the time of inclusion, 17 patients were in a manic phase and seven were in a depressive phase; mood state was assessed with the Young Mania Rating Scale and the Hamilton Depression Scale. Healthy controls aged from 16 to 40 (mean age = 26.56, S.D. = 6.70) years; education duration was 8 to 19 (mean duration = 13.06, S.D. = 2.32) years. A chisquare test was performed on genders of subjects in the two groups and revealed no statistical difference (χ2 = 0.104, P = 0.747). Age and education duration were tested by independent-sample t test, which revealed statistically insignificant differences for both factors.

The volume of the right middle frontal gyrus was also positively correlated to the number of episodes in BD I patients, with a correlation coefficient of 0.4779 and a P value of 0.0182 (d.f. = 22). No correlation was found between the number of episodes and the volume of other brain regions (Table 4). 3.6. Correlation analysis between differential region volumes and the age at onset in BD I patients There was no significant correlation between the volume of differential regions and onset age. 4. Discussion 4.1. Whole-brain gray matter volume A number of previous studies suggested no variance in wholebrain gray volume in patients with BD (Wilke et al., 2004; Chen et al.,

Table 2 Cerebral regions with variant gray matter volume in patients with bipolar I disorder. Regions

BD I b HC Left inferior parietal lobule Right superior temporal gyrus Right middle frontal gyrus Left caudate nucleus

Voxel

224 213 181 292 430

Talairach coordinates x

y

z

−51 55 56 36 −8

−35 −37 −50 47 12

43 5 13 13 9

T value

Pc value

Brodmann area

5.46 5.36 5.32 5.07 4.99

0.0348 0.0420 0.0136 0.0114 0.0015

40 22 22 10

Note: Pc indicates corrected P value on cluster level. Differential regions were defined as regions with an FDR of b0.05, more than 200 linked voxels and corrected P b 0.05 on a cluster level.

M. Li et al. / Psychiatry Research: Neuroimaging 191 (2011) 92–97

BD I
left inferior parietal lobule

right superior temporal gyrus

right middle frontal gyrus

95

right superior temporal gyrus

left caudate nucleus

Fig. 1. Brain regions with variant gray matter volume in patients with bipolar I disorder.

Our study found less gray matter volume in the left caudate nucleus, right middle frontal gyrus, right superior temporal gyrus and left inferior parietal lobule in BD I patients compared with controls. Our finding is consistent with a previous study, which showed patients with BD I had widespread decreases in gray matter in frontal, temporal, parietal and parahippocampal cortices while BD II only revealed less gray matter in ventromedial prefrontal regions and superior frontal gyrus (Ha et al., 2009). Some autopsy and magnetic resonance spectroscopic imaging studies have found

decreased N-acetyl aspartate concentration in the caudate nucleus (Port et al., 2008; Frye et al., 2007), gray matter of frontal lobe (Cecil et al., 2002; Molina et al., 2007; Yildiz-Yesiloglu and Ankerst, 2006) and superior temporal gyrus (Nudmamud et al., 2003) in BD as compared to that in healthy controls. N-acetyl aspartate is generally regarded as a marker for neurons and decreased concentrations may relate to loss or decreased activity of neurons. Therefore, less gray matter volume in these regions might be related to the loss or atrophy of neurons. However, there were fewer reports about decreased concentration of N-acetyl aspartate in the parietal lobe, which might relate to the fact that the role of the parietal lobe in the occurrence of BD was largely ignored in the past. Both the caudate nucleus and the middle frontal gyrus are part of the prefrontal–limbic system. The abnormal volume observed in these regions was consistent with numerous previous studies (Baumann et al., 1999; Beyer et al., 2004). Autopsy has found that volume of the caudate nucleus was relatively smaller in patients with emotional disorders (Baumann et al., 1999). In vivo studies also revealed reduced right caudate nucleus volume in older adult BD patients, and

Table 3 Results of rank correlation analysis on gray matter volume of differential regions and disease duration in bipolar I disorder patients.

Table 4 Results of rank correlation analysis on gray matter volume of differential regions and number of episodes in bipolar I disorder patients.

2007; Vita et al., 2009), or those with BD I or BD II (Ha et al., 2009). In agreement with these studies, we found no significantly higher or lower in GWV when comparing BD I patients to healthy controls. In addition, we did not find any relationships between WGV and disease duration or number of episodes. The absence of a correlation suggested that the occurrence and progression of BD might not involve changes in whole-brain gray matter volume. 4.2. Regional findings

Regions

Right middle frontal gyrus Left inferior parietal lobule Right superior temporal gyrus Left caudate nucleus

Talairach coordinates

Correlation coefficient

x

y

z

36 −51 55 56 −8

47 −35 −37 −50 12

13 43 5 13 9

0.5442 0.317 0.2300 0.3692 −0.0340

P value

0.0060 0.1314 0.2806 0.0758 0.8730

Regions

Right middle frontal gyrus Left inferior parietal lobule Right superior temporal gyrus Left caudate nucleus

Talairach coordinates x

y

z

36 −51 55 56 −8

47 −35 −37 −50 12

13 43 5 13 9

Correlation coefficient

P value

0.4779 0.3476 0.2548 0.3115 0.0406

0.0182 0.0961 0.2296 0.1385 0.8506

96

M. Li et al. / Psychiatry Research: Neuroimaging 191 (2011) 92–97

this finding was not influenced by disease duration or onset age (Beyer et al., 2004). Nevertheless, the role of the caudate nucleus in bipolar disorder is still unclear. Berk et al. (2007) has reviewed a large number of studies, among which pharmacological models suggested hyperactivity of dopamine affected the occurrence of mania while hypoactivity of dopamine was related to depression; unstable transportation of dopamine was hypothesized to lead to mood instability (Berk et al., 2007). The neostriatum (caudate nucleus and putamen) and substantia nigra are connected by numerous back-andforth nerve connections; there is a large body of dopaminergic neuron receptors in the caudate nucleus, and a decreased volume of the caudate nucleus and atrophy of neurons may result in disorder of dopamine functions, and thereby be related to the occurrence of the disorder. However, we failed to reveal a correlation between reduced volume of the caudate nucleus and disease duration, suggesting that the abnormality might have occurred before onset of the disease. In patients with emotional disorders, the presence of an abnormal frontal lobe (Konarski et al., 2008) and prefrontal lobe volume reductions (Arnone et al., 2009) has been highly consistent among studies. Some studies found reduced volume of both gray and white matter in the frontal lobe (Haznedar et al., 2005), but more studies focused on gray matter and revealed a lower volume, especially in the prefrontal cortex and cortex of fronto-orbital gyrus, regardless of study methods (ROI or VBM) (Lopez-Larson et al., 2002; Najt et al., 2007; Lyoo et al., 2004; Wilke et al., 2004; Dickstein et al., 2005). Studies using event-related functional MRI also revealed decreased (Yurgelun-Todd et al., 2000; Malhi et al., 2007) or absent (Strakowski et al., 2005) activity of the prefrontal lobe and limbic system in bipolar disorder patients, even in stable phases of the disease. These studies provided evidence for the defects in the frontal lobe in bipolar disorder. Our study also found less volume of left inferior parietal lobule and right superior temporal gyrus, which had been rarely seen in previous studies using ROI. The possible reason for this was that previous studies using ROI had mostly focused on the frontal lobe and limbic system, and even in those studies which included the parietal and temporal lobes the focus was on the delineation of the entire structure. However, the abnormality of brain structures in BD might only involve changes in regional gray matter volume in these structures. The VBM method is more sensitive to changes in regional gray matter volume. In line with previous findings, which showed less volume of bilateral parietal lobes and left temporal lobe in adolescents with BD (Frazier et al., 2005), and less cortex volume in the bilateral inferior parietal lobule, right superior temporal gyrus and cingulate gyrus in treatment-naïve patients (Nugent et al., 2006), our finding supported the volume changes in the parietal and temporal lobes in BD patients to some degree, regardless of age of onset, clinical phase of the disease or the use of medication. The parietal lobe is generally considered related to attention. Attention deficits, such as distractibility of attention, are prominent clinical symptoms in patients with BD. A number of transcranial magnetic stimulation studies have also suggested that the parietal lobe plays a critical role in spatial attention and relocation of behavior (Rushworth and Taylor, 2006). The temporal lobe is connected to the limbic system and the prefrontal lobe by extensive nerve connections, and is considered involved in the process of emotional prosody (Mitchell et al., 2003). In addition, patients with schizophrenia and bipolar disorder may show some left-lateralisation of the normal right-lateralised temporal lobe response to emotional prosody (Mitchell et al., 2004). Whether low volume of the right superior temporal gyrus plays an important role in the etiology of bipolar disorder or not needs to be confirmed by further studies.

ventricles were found in multiple-episode BD versus first-episode patients (Strakowski et al., 2002); higher gray matter volume was found to appear in BD when the ratio of patients using lithium increased (Kempton et al., 2008); and compared with healthy controls, BD patients displayed a larger decline in hippocampal, fusiform, and cerebellar gray matter density at 4-year follow-up (Moorhead et al., 2007). In view of these factors, we performed the correlation analysis between the volume of differential brain regions and disease duration/ number of episodes/age at onset. We only found the volume of the right middle frontal gyrus to be positively associated with both disease duration and number of episodes. This correlation might reflect a compensatory or other change in brain structure with progression of illness. However, it should noted that, although it was difficult to confirm the exact length of treatment time for every patient since some of them were ill for many years and did not take medicine regularly, there was a tendency for longer duration of illness to be correlated with a relatively longer treatment time. In addition, in the present study, nearly 80% of patients received lithium treatment. Considering the previous study which identified an increased volume of gray matter in the dorso-lateral frontal lobe and cingulate gyrus in healthy adults who had received lithium compounds, while no increase in gray matter volume of other cerebral regions was seen (Monkul et al., 2007), we presumed the alternative explanation of the correlation in present study might be due to the use of lithium. Yuan et al. (2004) and Wada et al. (2005) reviewed a number of studies and presumed that lithium compounds might have nourished and protected neurons by preventing death of nerve neurons and promoting regeneration (Yuan et al., 2004; Wada et al., 2005). It was presumed that atrophy of neurons and low volume in the right middle frontal gyrus might have already occurred at the beginning of the disease, and that usage of lithium compounds in subsequent course of the disease might have promoted the restoration of neurons in the middle frontal gyrus. In the present study, we also found that among the five differential brain regions, four regions were not related to disease duration or number of episodes and none of the regions were correlated with age at onset. These findings suggest that progression of illness has relatively little impact on the differences in brain structure. The causal mechanism in bipolar disorder seems very likely to involve some neurodevelopmental gray matter volumetric differences (Sanches et al., 2008). 4.4. Limitations There are a number of limitations of the present study. Firstly, the patients included had taken medicine prior to scanning and it is difficult to ascertain the specific duration of drug treatment for each patient. Therefore, the effects of medication could be confounding factors in the analysis. Secondly, the heterogeneity of the sample have affected the results reported here to some extent, as the age at examination, duration of illness and number of episodes varied in a relatively large range. Thirdly, it is impossible to clarify the natural evolution of brain abnormalities in such a cross-sectional study. Longitudinal prospective studies are needed in future research. In summary, by using the optimized VBM method, we found that in addition to abnormal volume in structures of the prefrontal-limbic system, the inferior parietal lobule and the superior temporal gyrus were also regionally impaired in patients with BD I. These finding demonstrated that the occurrence of the disease might not merely involve the prefrontal-limbic system, the latter structures (inferior parietal lobule and superior temporal gyrus) might also have important roles in the occurrence of the disease and deserve further attention in future investigation.

4.3. Clinical variables Acknowledgements Some abnormalities of brain structure were considered to be affected by various factors, such as the number of episodes, age at onset, use of lithium, and different stages of illness. For example, larger lateral

This work was partly funded by the National Natural Science Foundation of China (30530300, 30770777 and 30125014), the National

M. Li et al. / Psychiatry Research: Neuroimaging 191 (2011) 92–97

Basic Research Program of China (973 Program 2007CB512301), a NARSAD independent Investigator Award, and the Wellcome Trust (International Collaborative award). References Adler, C.M., Levine, A.D., DelBello, M.P., Strakowski, S.M., 2005. Changes in gray matter volume in patients with bipolar disorder. Biological Psychiatry 58, 151–157. Arnone, D., Cavanagh, J., Gerber, D., Lawrie, S.M., Ebmeier, K.P., McIntosh, A.M., 2009. Magnetic resonance imaging studies in bipolar disorder and schizophrenia: metaanalysis. The British Journal of Psychiatry 195, 194–201. Ashburner, J., Friston, K.J., 2005. Unified segmentation. Neuroimage 26, 839–851. Baumann, B., Danos, P., Krell, D., Diekmann, S., Leschinger, A., Stauch, R., Wurthmann, C., Bernstein, H.G., Bogerts, B., 1999. Reduced volume of limbic system affiliated basal ganglia in mood disorders: preliminary data from a postmortem study. Journal of Neuropsychiatry and Clinical Neurosciences 11, 71–78. Berk, M., Dodd, S., Kauer-Sant'anna, M., Malhi, G.S., Bourin, M., Kapczinski, F., Norman, T., 2007. Dopamine dysregulation syndrome: implications for a dopamine hypothesis of bipolar disorder. Acta Psychiatrica Scandinavica 116, 41–49. Beyer, J.L., Kuchibhatla, M., Payne, M., Moo-Young, M., Cassidy, F., MacFall, J., Krishnan, K.R.R., 2004. Caudate volume measurement in older adults with bipolar disorder. International Journal of Geriatric Psychiatry 19, 109–114. Cecil, K.M., DelBello, M.P., Morey, R., Strakowski, S.M., 2002. Frontal lobe differences in bipolar disorder as determined by proton MR spectroscopy. Bipolar Disorders 4, 357–365. Chen, X.H., Wen, W., Malhi, G.S., Ivanovski, B., Sachdev, P.S., 2007. Regional gray matter changes in bipolar disorder: a voxel-based morphometric study. Australian and New Zealand Journal of Psychiatry 41, 327–336. Dickstein, D.P., Milham, M.P., Nugent, A.C., Drevets, W.C., Charney, D.S., Pine, D.S., Leibenluft, E., 2005. Frontotemporal alterations in pediatric bipolar disorder — results of a voxel-based morphometry study. Archives of General Psychiatry 62, 734–741. Ellison-Wright, I., Bullmore, E., 2010. Anatomy of bipolar disorder and schizophrenia: a meta-analysis. Schizophrenia Research 117, 1–12. Frazier, J.A., Breeze, J.L., Makris, N., Giuliano, A.S., Herbert, M.R., Seidman, L., Biederman, J., Hodge, S.M., Dieterich, M.E., Gerstein, E.D., Kennedy, D.N., Rauch, S.L., Cohen, B.M., Caviness, V.S., 2005. Cortical gray matter differences identified by structural magnetic resonance imaging in pediatric bipolar disorder. Bipolar Disorders 7, 555–569. Frye, M.A., Thomas, M.A., Yue, K., Binesh, N., Davanzo, P., Ventura, J., O'Neill, J., Guze, B., Curran, J.G., Mintz, J., 2007. Reduced concentrations of N-acetylaspartate (NAA) and the NAA-creatine ratio in the basal ganglia in bipolar disorder: a study using 3-Tesla proton magnetic resonance spectroscopy. Psychiatry Research: Neuroimaging 154, 259–265. Giuliani, N.R., Calhoun, V.D., Pearlson, G.D., Francis, A., Buchanan, R.W., 2005. Voxelbased morphometry versus region of interest: a comparison of two methods for analyzing gray matter differences in schizophrenia. Schizophrenia Research 74, 135–147. Good, C.D., Johnsrude, I.S., Ashburner, J., Henson, R.N.A., Friston, K.J., Frackowiak, R.S.J., 2001. A voxel-based morphometric study of ageing in 465 normal adult human brain. Neuroimage 14, 21–36. Ha, T.H., Ha, K., Kim, J.H., Choi, J.E., 2009. Regional brain gray matter abnormalities in patients with bipolar II disorder: a comparison study with bipolar I patients and healthy controls. Neuroscience Letters 456, 44–48. Hauser, P., Matochik, J., Altshuler, L.L., Denicoff, K.D., Conrad, A., Li, X.M., Post, R.M., 2000. MRI-based measurements of temporal lobe and ventricular structures in patients with bipolar I and bipolar II disorders. Journal of Affective Disorders 60, 25–32. Haznedar, M.M., Roversi, F., Pallanti, S., Baldini-Rossi, N., Schnur, D.B., LiCalzi, E.M., Tang, C., Hof, P.R., Hollander, E., Buchsbaum, M.S., 2005. Fronto–thalamo–striatal gray and white matter volumes and anisotropy of their connections in bipolar spectrum illnesses. Biological Psychiatry 57, 733–742. Kempton, M.J., Geddes, J.R., Ettinger, U., Williams, S.C.R., Grasby, P.M., 2008. Metaanalysis, database, and meta-regression of 98 structural imaging studies in bipolar disorder. Archives of General Psychiatry 65, 1017–1032. Konarski, J.Z., Mcintyre, R.S., Kennedy, S.H., Rafi-Tari, S., Soczynska, J.K., Ketter, T.A., 2008. Volumetric neuroimaging investigations in mood disorders: bipolar disorder versus major depressive disorder. Bipolar Disorders 10, 1–37. Lopez-Larson, M.P., DelBello, M.P., Zimmerman, M.E., Schwiers, M.L., Strakowski, S.M., 2002. Regional prefrontal gray and white matter abnormalities in bipolar disorder. Biological Psychiatry 52, 93–100. Lyoo, I.K., Kim, M.J., Stoll, A.L., Demopulos, C.M., Parow, A.M., Dager, S.R., Friedman, S.D., Dunner, D.L., Renshaw, P.F., 2004. Frontal lobe gray matter density decreases in bipolar I disorder. Biological Psychiatry 55, 648–651. Malhi, G.S., Lagopoulos, J., Owen, A.M., Ivanovski, B., Shnier, R., Sachdev, P., 2007. Reduced activation to implicit affect induction in euthymic bipolar patients: an fMRI study. Journal of Affective Disorders 97, 109–122.

97

McElroy, S.L., Altshuler, L.L., Suppes, T., Keck, P.E., Frye, M.A., Denicoff, K.D., Nolen, W.A., Kupka, R.W., Leverich, G.S., Rochussen, J.R., Rush, A.J., Post, R.M., 2001. Axis I psychiatric comorbidity and its relationship to historical illness variables in 288 patients with bipolar disorder. American Journal of Psychiatry 158, 420–426. Mitchell, R.L.C., Elliott, R., Barry, M., Cruttenden, A., Woodruff, P.W.R., 2003. The neural response to emotional prosody, as revealed by functional magnetic resonance imaging. Neuropsychologia 41, 1410–1421. Mitchell, R.L.C., Elliott, R., Barry, M., Cruttenden, A., Woodruff, P.W.R., 2004. Neural response to emotional prosody in schizophrenia and in bipolar affective disorder. British Journal of Psychiatry 184, 223–230. Molina, V., Sanchez, J., Sanz, J., Reig, S., Benito, C., Leal, I., Sarrarnea, F., Rebolledo, R., Palomo, T., Desco, M., 2007. Dorsolateral prefrontal N-acetyl-aspartate concentration in male patients with chronic schizophrenia and with chronic bipolar disorder. European Psychiatry 22, 505–512. Monkul, E.S., Matsuo, K., Nicoletti, M.A., Dierschke, N., Hatch, J.P., Dalwani, M., Brambilla, P., Caetano, S., Sassi, R.B., Mallinger, A.G., Soares, J.C., 2007. Prefrontal gray matter increases in healthy individuals after lithium treatment: a voxel-based morphometry study. Neuroscience Letters 429, 7–11. Moorhead, T.W.J., McKirdy, J., Sussmann, J.E.D., Hall, J., Lawrie, S.M., Johnstone, E.C., McIntosh, A.M., 2007. Progressive gray matter loss in patients with bipolar disorder. Biological Psychiatry 62, 894–900. Najt, P., Nicoletti, M., Chen, H.H., Hatch, J.P., Caetano, S.C., Sassie, R.B., Axelson, D., Rambilla, P., Keshavan, M.S., Ryan, N.D., Birmaher, B., Soares, J.C., 2007. Anatomical measurements of the orbitofrontal cortex in child and adolescent patients with bipolar disorder. Neuroscience Letters 413, 183–186. Nudmamud, S., Reynolds, L.M., Reynolds, G.P., 2003. N-acetylaspartate and Nacetylaspartylglutamate deficits in superior temporal cortex in schizophrenia and bipolar disorder: a postmortem study. Biological Psychiatry 53, 1138–1148. Nugent, A.C., Milham, M.P., Bain, E.E., Mah, L., Cannon, D.M., Marrett, S., Zarate, C.A., Pine, D.S., Price, J.L., Drevets, W.C., 2006. Cortical abnormalities in bipolar disorder investigated with MRI and voxel-based morphometry. Neuroimage 30, 485–497. Port, J.D., Unal, S.S., Mrazek, D.A., Marcus, S.M., 2008. Metabolic alterations in medication-free patients with bipolar disorder: a 3 T CSF-corrected magnetic resonance spectroscopic imaging study. Psychiatry Research: Neuroimaging 162, 113–121. Ridgway, G., Omar, R., Ourselin, S., Hill, D., Warren, J., Fox, N., 2008. Issues with threshold masking in voxel-based morphometry of atrophied brains. NeuroImage 44, 99–111. Rorden, C., Brett, M., 2000. Stereotaxic display of brain lesions. Behavioura. Neurology 12, 191–200. Rushworth, M.F.S., Taylor, P.C.J., 2006. TMS in the parietal cortex: updating representations for attention and action. Neuropsychologia 44, 2700–2716. Sanches, M., Keshavan, M.S., Brambilla, P., Soares, J.C., 2008. Neurodevelopmental basis of bipolar disorder: a critical appraisal. Progress in Neuro-Psychopharmacology & Biological Psychiatry 32, 1617–1627. Smith, S.M., 2002. Fast robust automated brain extraction. Human Brain Mapping 17, 143–155. Strakowski, S.M., DelBello, M.P., Zimmerman, M.E., Getz, G.E., Mills, N.P., Ret, J., Shear, P., Adler, C.M., 2002. Ventricular and periventricular structural volumes in first- versus multiple-episode bipolar disorder. American Journal of Psychiatry 159, 1841–1847. Strakowski, S.M., DelBello, M.P., Adler, C.M., 2005. The functional neuroanatomy of bipolar disorder: a review of neuroimaging findings. Molecular Psychiatry 10, 105–116. van der Schot, A.C., Vonk, R., Brans, R.G.H., van Haren, N.E.M., Koolschijn, P.C.M.P., Nuboer, V., Schnack, H.G., van Baal, G.C.M., Boomsma, D.I., Nolen, W.A., Pol, H.E.H., Kahn, R.S., 2009. Influence of genes and environment on brain volumes in twin pairs concordant and discordant for bipolar disorder. Archives of General Psychiatry 66, 142–151. Vita, A., De Peri, L., Sacchetti, E., 2009. Gray matter, white matter, brain, and intracranial volumes in first-episode bipolar disorder: a meta-analysis of magnetic resonance imaging studies. Bipolar Disorders 11, 807–814. Wada, A., Yokoo, H., Yanagita, T., Kobayashi, H., 2005. Lithium: potential therapeutics against acute brain injuries and chronic neurodegenerative diseases. Journal of Pharmacological Sciences 99, 307–321. Wilke, M., Kowatch, R.A., DelBello, M.P., Mills, N.P., Holland, S.K., 2004. Voxel-based morphometry in adolescents with bipolar disorder: first results. Psychiatry Research: Neuroimaging 131, 57–69. Yildiz-Yesiloglu, A., Ankerst, D.P., 2006. Neurochemical alterations of the brain in bipolar disorder and t heir implications for pathophysiology: a systematic review of the in vivo proton magnetic resonance spectroscopy findings. Progress in NeuroPsychopharmacology & Biological Psychiatry 30, 969–995. Yuan, P.X., Gould, T.D., Gray, N.A., Bachmann, R.F., Schloesser, R.J., Lan, M.J.K., Du, J., Moore, G.J., Manji, H.K., 2004. Neurotrophic signaling cascades are major long-term targets for lithium: clinical implications. Clinical Neuroscience Research 4, 137–153. Yurgelun-Todd, D.A., Gruber, S.A., Kanayama, G., Killgore, W.D.S., Baird, A.A., Young, A.D., 2000. fMRI during affect discrimination in bipolar affective disorder. Bipolar Disorders 2, 237–248.