- Email: [email protected]
Changes in Gray Matter Volume in Patients with Bipolar Disorder
Changes in Gray Matter Volume in Patients with Bipolar Disorder Caleb M. Adler, Ari D. Levine, Melissa P. DelBello, and Stephen M. Strakowski Background: Several lines of evidence suggest the presence of neurofunctional abnormalities in patients with bipolar disorder. These functional abnormalities may stem from structural pathology in these or connected brain regions. Previous studies have generally used a region of interest (ROI) approach to study morphologic changes in bipolar disorder with inconsistent findings among research groups, which may reflect differences in how ROIs are defined. Voxel based morphometry (VBM) allows a more exploratory analysis without the necessity for predefined anatomic boundaries. In this study we utilized VBM to compare gray matter volume between groups of bipolar and healthy subjects. Methods: Thirty-two patients with bipolar disorder and 27 healthy subjects participated in structural magnetic resonance imaging (MRI) scans. MRI images were segmented, normalized to a standard stereotactic space, and compared on a voxel-by-voxel basis using statistical parametric mapping. Results: Bipolar subjects showed increased gray matter in several regions including portions of anterior cingulate, ventral prefrontal cortex, fusiform gyrus and parts of the primary and supplementary motor cortex. Bipolar subjects showed decreased gray matter volume in superior parietal lobule. Conclusions: These data support suggestions that neurofunctional deficits are related to structural brain abnormalities in patients with bipolar disorder. The increased gray matter observed in several regions suggests that some affected areas may demonstrate volumetric expansion, at least in some patient populations. Key Words: Bipolar Disorder, voxel based morphometry, MRI, brain, prefrontal cortex, anterior cingulate
B
ipolar disorder is a major psychiatric illness that affects approximately 1.5% of the population and is a significant source of individual morbidity and societal cost (Goldberg et al 1995; Greenberg et al 1993; Regier et al 1990; Strakowski et al 1996, 1998). Although the pathophysiology of bipolar disorder remains poorly understood, findings from functional imaging studies suggest that bipolar disorder is associated with neuropathologic changes in specific brain regions, particularly within the frontal and temporal cortex, and subcortical structures, making up the anterior limbic network (Strakowski et al 2000). The results of morphological studies of these brain regions however, have been mixed and not entirely congruent with functional imaging findings (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2000, 2002). One potential cause of this disparity in findings may be the almost exclusive use of region-of-interest (ROI) based analyses in morphologic studies of patients with bipolar disorder. While useful in hypothesis-driven structural studies, this approach may be compromised by the lack of uniform methodology, and particularly by variations in structural boundaries. The prefrontal cortex, an area widely hypothesized to be involved in the pathophysiology of bipolar disorder, provides an excellent example of this methodological issue. Only a minority of structural studies have observed significant differences between bipolar patients and healthy subjects in prefrontal and frontal cortical
From the Center for Bipolar Disorders Research (CMA, AL, MPD, SMS) and the Center for Imaging Research (CMA, SMS), University of Cincinnati College of Medicine, Cincinnati, Ohio. Address reprint requests to Caleb M. Adler, M.D., University of Cincinnati College of Medicine, Department of Psychiatry, PO Box 670559, 231 Albert Sabin Way, Cincinnati, OH 45267; E-mail: cal.adler@psychiatry. uc.edu. Received December 6, 2004; revised February 15, 2005; accepted March 7, 2005.
0006-3223/05/$30.00 doi:10.1016/j.biopsych.2005.03.022
volume (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2000, 2002), despite differential activation observed across several functional imaging studies (Adler et al 2004; Blumberg et al 2003; Strakowski et al 2004, 2005, in press; Yurgelun-Todd et al 2000), and research suggesting correlations between prefrontal morphometry and cognitive performance in bipolar patients (Sax et al 1999), as well as inverse correlation between illness duration and prefrontal volume in some studies (Brambilla et al 2001; Lopez-Larson et al 2002). Among the possible causes of these discrepancies may be the poor match between functional and easily defined structural boundaries. Difficulties in mapping functional regions to clearly defined anatomic areas may play a role in the frequently ambiguous morphological findings in other portions of the anterior limbic network as well. Morphometric studies of the temporal cortex, and of some medial temporal structures, for instance, have been largely unrevealing, despite strong functional evidence for their involvement in the pathophysiology of bipolar disorder (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2005). Furthermore, ROI based analyses of even fairly easily defined brain regions or structures combine both affected and unaffected areas, potentially vitiating differences. Volumetric studies of midline structures such as the basal ganglia for example, have been inconclusive. While the majority of investigators have not observed changes in striatal volume, some investigators have noted enlargement of portions of the basal ganglia (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2000, 2002). One alternative to ROI-based approaches is voxel-based morphometry (VBM). In VBM, individual structural MRI scans are normalized to a standard template to allow voxel-by-voxel comparisons. Advantages of this approach include the ability to perform an exploratory analysis without the need to define structures a priori. In addition, VBM analysis allows structural investigation of functional regions, such as portions of the prefrontal cortex that may be difficult to define anatomically. Potential disadvantages however, include the need to control for the large number of comparisons made in a structural study of the whole brain, with the potential loss of statistical power which that entails, and the possibility of blurring precise structural BIOL PSYCHIATRY 2005;58:151–157 © 2005 Society of Biological Psychiatry
152 BIOL PSYCHIATRY 2005;58:151–157
C.M. Adler et al
boundaries implicit in normalizing diverse brains to a single template. Five studies published to date have reported VBM analyses of bipolar disorder. Wilke and colleagues utilized an optimized comparison technique to measure differences in gray matter volume between ten adolescents with bipolar disorder and a matched group of 52 healthy subjects. In contrast to most ROI based studies, Wilke and colleagues observed areas of decreased gray matter volume in the left anterior cingulate, as well as the orbitofrontal cortex and medial temporal structures. Increased volume was observed in portions of the basal ganglia (Wilke et al 2004). Lochhead and colleagues compared groups of 11 bipolar patients and 31 healthy subjects. Bipolar patients were found to have areas of decreased gray matter volume in separate portions of the anterior cingulate and temporal cortex. Areas of increased gray matter volume were observed in the insular/ frontoparietal cortex (Lochhead et al 2004). In contrast, McIntosh and colleagues observed decreased gray matter in the right inferior frontal gyrus, as well as the right insula in a group of 19 bipolar patients from families with a history of both bipolar disorder and schizophrenia. A larger group of 26 bipolar patients from bipolar families demonstrated decreased gray matter only in the thalamus and caudate (McIntosh et al 2004). In the largest study, Bruno and colleagues compared 39 bipolar patients with 35 healthy controls. They failed to identify any differences in gray matter volume between groups. Using a different technique, Lyoo and colleagues measured differences in gray matter density between 39 patients with bipolar disorder to 43 healthy subjects. They observed decreased density in the anterior cingulate and portions of the frontal cortex including the left medial and right inferior frontal gyrus, as well as the right precentral gyrus (Lyoo et al 2004). Lyoo and colleagues (2004) also noted that gray matter density in the left medial and right inferior frontal gyri inversely correlated with the number of manic episodes experienced by bipolar subjects. These studies suggest that differences in gray matter volume and density may be present that are not readily apparent in studies utilizing a more traditional ROI-based approach. Previous studies of gray matter volume however, have for the most part included only small numbers of bipolar patients. Moreover, Wilke and colleagues studied adolescent bipolar patients, a group that may be clinically distinct from adults (Geller et al 2004). In this study we utilized an optimized voxel-based analysis to compare gray matter volumes between a larger group of patients with bipolar disorder, including relatively early onset and first-episode patients, and a matched group of healthy
subjects. Based on previous morphometric studies, we hypothesized that we would find decreases in gray matter volume, particularly within functional regions making up the anterior limbic network, hypothesized to be involved in the neuropathophysiology of bipolar disorder. Further, we hypothesized that our cohort of early onset patients would demonstrate less change in gray matter volume than was observed in previous studies.
Methods and Materials Participants Thirty-two patients with bipolar disorder, type I were recruited from ongoing outcome studies and twenty-seven healthy subjects were recruited in response to local advertisement or by word of mouth. Bipolar and healthy subjects did not statistically differ in age, education, sex, or handedness (Table 1). Subjects with bipolar disorder were free of concurrent axis I psychiatric or medical illness, including substance use disorders (with the exception of nicotine addiction). No healthy subjects had been diagnosed with any axis I psychiatric conditions, including substance dependence, current substance abuse, or major medical conditions (with the exception of nicotine addiction). No healthy controls reported a history of axis I psychiatric conditions in first-degree relatives. No healthy subjects were receiving psychotropic medication at the time of the study. Diagnosis in subjects with bipolar disorder, and the absence of axis I conditions in healthy controls were determined using the Structured Clinical Interview for DSM-IV (SCID) administered by a board certified psychiatrist or Ph.D. level psychologist (First et al 1997). Patients had an average age of onset of 22.5 years (SD ⫾ 7.7), and had been ill for an average of 8.7 years (SD ⫾ 9.2). The number of quantifiable past depressive episodes in subjects with bipolar disorder ranged from zero to eleven with an average of 2.9 (SD ⫾ 3.2) episodes. Eleven subjects reported that their number of depressive episodes had been too numerous to count, or were otherwise unable to accurately quantify the number. The number of past manic and mixed episodes ranged from one to six with an average of 2.0 (SD ⫾ 1.5) episodes. Nine subjects were unable to accurately quantify the number of manic and mixed episodes that they had experienced. Nine subjects were unmedicated at the time of the study. The remainder of the bipolar subjects was receiving standard pharmacotherapy including lithium, divalproex, anti-seizure medications including gabapentin and topiramate, atypical antipsychotics, antidepressants, and benzodiazepines, alone or in combination. While the major-
Table 1. Demographics of Bipolar and Healthy Subjects Characteristics Age, years Education, years Sex, n (%) women Handednessa, n (%) right Age of Onset, years Mood State, manic (%)/depressed (%) Number of Depressive Episodesb Number of Manic Episodesc Medication Status, n (%) current
Bipolar Subjects (n ⫽ 32)
Healthy Subjects (n ⫽ 27)
Group Difference (p-value)
31.2 (9.4) 13.6 (2.2) 13 (41) 31 (97) 22.5 (7.7) 5 (16)/2 (6) 2.9 (3.2) 2.0 (1.5) 23 (72)
30.5 (9.7) 14.0 (1.6) 15 (56) 26 (96) NA NA NA NA NA
.79 .33 .26 .90 NA NA NA NA NA
All values are mean (SD) unless otherwise noted. Handedness was determined using the Crovitz Handedness Inventory (Crovitz and Zener 1962). b n ⫽ 21. c n ⫽ 23. a
www.sobp.org/journal
BIOL PSYCHIATRY 2005;58:151–157 153
C.M. Adler et al ity of the bipolar patients were euthymic at the time of the scan, seven subjects were either manic or depressed (Table 1). All subjects demonstrated facility with English and a clear understanding of the procedures, risks and intent of the study. All subjects provided written informed consent, which was obtained after study procedures had been explained in detail, prior to their participating in the study. This study was approved by the University of Cincinnati College of Medicine Institutional Review Board. Imaging Scans were performed with a 3.0 Tesla Bruker Biospec scanner (Bruker Medizintechnik, Karisruhe, Germany). During the study, subjects reclined in a supine position on the bed of the scanner and a RF coil (Bruker NMR Instruments Inc., Fremont, California) was placed over the participant’s head. Earplugs and headphones were provided to block background noise. Following a 3-plane gradient echo scan for alignment and localization, a shim procedure was performed to generate a homogenous, constant magnetic field. One hundred-twenty contiguous 1 mm axial slices extending superiorly from the inferior aspect of the cerebellum to encompass most of the brain were selected from a sagittal localizer scan. A high-resolution T1-weighted 3-D brain scan was then obtained using a modified driven equilibrium Fourier transform (MDEFT) protocol (TI ⫽ 550 msec, TR ⫽ 16.5, TE ⫽ 4.3 msec, FOV ⫽ 25.6 ⫻ 19.2 ⫻ 14.4, 256 ⫻ 128 ⫻ 96, flip angle ⫽ 20°). Precautions were taken to minimize subject motion during the MRI study by instructing subjects to remain still and packing around their heads with foam padding. Data Pre-Processing All automated image processing was done using statistical parametrical mapping software (SPM99, Wellcome Department of Cognitive Neurology, University College London, United Kingdom) running in MATLAB (MathWorks, Natick, Massachusetts). Determination of the anterior commissure was made in each image by a single investigator (AL). Images were uniformly aligned with regard to head position, to provide optimal starting estimates for subsequent spatial normalization. Analysis followed the “optimized” voxel-based morphometry (VBM) processing strategy of Good and colleagues (Good et al 2001). Images were segmented in native space, and any remaining extra-neuronal tissue was removed by modulation with an extracted, individualized brain mask. These native-space gray matter partitions were then normalized to the corresponding SPM99 tissue-probability map. This procedure weights normalization parameters heavily towards gray matter and minimizes contributions from other tissue types. The normalization parameters from this transformation were then applied to the original images, yielding a normalized whole brain image. These images were then segmented and again modulated with a newly extracted, individual brain mask in order to remove nonbrain tissue. As part of the segmentation procedure, residual image inhomogeneities were removed by modeling smoothly varying intensity changes (Ashburner and Friston 2000). This procedure involves the estimation of an intensity nonuniformity field, which is then applied, to yield a corrected image. This procedure has been shown to markedly increase the reproducibility of SPM99segmentation results (Chard et al 2002). Spatial normalization was achieved using an initial 12-parameter affine transformation, followed by 12 nonlinear iterations using 7 ⫻ 8 ⫻ 7 discrete cosine transform basis functions
(Ashburner and Friston 1999, 2000). Images were written out in 1 ⫻ 1 ⫻ 1 mm resolution. To allow for the detection of true gray matter volume changes, images were modulated by the Jacobian determinant of the normalization matrix, resulting in images that take into account global and local volume changes during spatial normalization (Good et al 2001). Final images were smoothed using a Gaussian kernel with a full-width half-maximum (FWHM) of 12 mm to create a local weighted average of the surrounding pixels. This filter width, as per the matched filter theorem, determines the spatial scale at which changes are most sensitively detected, and also accounts for structural variability and possibly inexact spatial normalization. Image Analysis and Statistics Processed images from both datasets (healthy controls and bipolar patients) were analyzed within SPM99, employing the framework of the general linear model (Friston et al 1995). A model was designed in which age and gender were used as covariates of no interest; diagnosis was considered the parameter of interest. Two contrasts were calculated, testing for a positive or negative correlation of gray matter volume with the parameter of interest. Significance was set at a p-value of p ⫽ .001; with a minimum cluster size of 200 voxels (Wilke et al 2001, 2004). Points of maximum correlation were converted from Montreal Neurological Institute to Talairach coordinates using a nonlinear transformation (Brett 2002). Voxels of maximal correlation within the anterior cingulate and ventral prefrontal cortices were plotted across individual scans, using SPM99. A secondary analysis was performed on the bipolar data set using first episode (defined as having only one manic episode (n ⫽ 13), versus multiple episodes (n ⫽ 10) as the parameter of interest. Contrasts testing for positive or negative correlation of gray matter volume with the parameter of interest were calculated for regions in which significant differences were observed between bipolar and healthy subjects. Because this was a limited post-hoc analysis, significance was set at p ⫽ .01; with a minimum cluster size of 200 voxels (Wilke et al 2001, 2004). Points of maximum correlation were converted from MNI to Talairach coordinates using a nonlinear transformation (Brett 2002).
Results Patients with bipolar disorder showed increased gray matter volume in portions of the anterior cingulate, and of the ventral prefrontal cortex. Increased gray matter was also observed in the temporal cortex, including a portion of the left fusiform gyrus, and in areas involved with motor control, including parts of the supplemental motor cortex and the left precentral gyrus (Figure 1, Table 2). Between group separation was particularly consistent across individuals within each group in the anterior cingulate and ventral prefrontal cortex (Figure 2). Decreased gray matter volume in patients with bipolar disorder was observed only in a portion of the right superior parietal lobule (Figure 1, Table 2). The secondary analysis found increased gray matter volume in first- versus multi-episode patients only in portions of the ventral prefrontal cortex (Figure 3, Table 3).
Discussion In this study we observed increased gray matter volume in bipolar subjects in several regions hypothesized to be part of the anterior limbic network, and to play a role in emotional regulation, attention, and impulse control. These include portions of www.sobp.org/journal
154 BIOL PSYCHIATRY 2005;58:151–157
C.M. Adler et al Areas of decreased gray matter density were identified that substantially overlap regions of increased volume in the anterior cingulate and ventral prefrontal cortex (Lyoo et al 2004), suggesting concurrent changes in volume and density consistent with an apoptotic process. This model is also consistent with differences in findings between this study and that of previous investigations, which for the most part studied cohorts of older and more chronic subjects (Bruno et al 2004; Lochhead et al 2004; McIntosh et al 2004). Conversely, the subjects studied by Wilke and colleagues were adolescents whose prefrontal maturation is presumably incomplete, and who appear to constitute a clinically distinct group (Geller et al 2004). While limited in scope, the areas of greater gray matter volume we observed in first episode patients, in the ventral prefrontal cortex, at least partially support this hypothesis. Moreover, previous functional neuroimaging studies of young bipolar subjects have noted an absence of prefrontal differences (Blumberg et al 2003), leading to suggestions that ventral prefrontal abnormalities may emerge over the course of the illness (Blumberg et al 2004). Areas of increased gray matter volume were observed in several structures of the anterior limbic network, including Brodmann area 47 of the prefrontal cortex. Brodmann area 47 is adjacent and ventral to Brodmann area 10, an area that we have previously observed to activate significantly differently in patients with bipolar disorder compared with healthy subjects during cognitive tasks (Adler et al 2004; Strakowski et al 2004, in press). The ventral prefrontal cortex, including Brodmann areas 10/47, has been associated with elements of mood regulation that are hypothesized to be disrupted in bipolar disorder (Mega et al 1997; Ongür and Price 2000; Yamasaki et al 2002). Few morphometric studies however, have specifically examined this region in patients with bipolar disorder, and those areas of volumetric change that have been observed encompass large, functionally nonspecific portions of the ventral prefrontal cortex (Drevets et al 1997; Lopez-Larson et al 2002). Our findings suggest that morphological markers of neuropathology are present in only portions of the prefrontal cortex, and raise the possibility that functional changes elsewhere in the prefrontal cortex may be secondary to pathology in these ventral prefrontal regions.
Figure 1. Voxel-by-voxel comparison map of bipolar (n ⫽ 32) versus control (n ⫽ 27) subjects, overlaid on a T1-weighted anatomic image. Statistically significant differences in gray matter volume were defined p ⱕ .001, with a cluster size of 200.
the anterior cingulate, the ventral prefrontal cortex, and the fusiform gyrus. In addition, we observed increased gray matter volume in portions of the primary and secondary motor cortex. Decreased gray matter was observed only in a portion of the superior parietal lobule. Although somewhat speculative, differences in gray matter volume may represent anatomic markers of neuropathology. Decreased gray matter in the superior parietal lobule suggests neuronal apoptosis, or a loss of neuropil. In this study however, we identified several areas of increased gray matter volume in patients with bipolar disorder. We hypothesize that increased gray matter may be related to preapoptotic osmotic changes or hypertrophy, marking areas of early neuronal pathology. This suggestion is consistent with measures of gray matter density made by Lyoo and colleagues.
Table 2. Regional Coordinates of Areas of Differential Gray Matter Volume Brain Region Bipolar ⬎ Control Frontal Cortex L anterior cingulate R anterior cingulate L ventral prefrontal cortex (middle frontal gyrus) R ventral prefrontal cortex (middle frontal gyrus) L precentral gyrus Temporal Cortex L fusiform gyrus Parietal Cortex L supplemental motor cortex R supplemental motor cortex Control ⬎ Bipolar Parietal Cortex R superior parietal lobule L, left; R, right. a All regions p ⱕ .001. b Regions are contiguous.
www.sobp.org/journal
Talairach Coordinates
Brodmann Area
Cluster Sizea
787b 787b 370
⫺3 3 ⫺46
47 47 36
6 6 ⫺1
Area 42 Area 42 Area 47
40
40
⫺5
Area 47
1028
⫺49
4
6
Area 44
210
⫺53
⫺44
⫺5
Area 37
1087
⫺18 31
10 14
⫺53 51
Area 6 Area 6
225 608
43
⫺66
51
Area 7
466
C.M. Adler et al
Figure 2. Scan by scan fitted response of bipolar (n ⫽ 32) versus control (n ⫽ 27) subjects for anterior cingulate, left and right ventral prefrontal cortex. Between group differences were particularly consistent across individual subjects in these regions (fitted response: solid gray line; error: dashed line).
Despite cognitive and functional imaging based evidence of neurofunctional deficits in at least portions of the anterior cingulate, previous volumetric findings have been equivocal. With few exceptions, morphometry studies have failed to identify significant structural differences between bipolar and healthy subjects. Although Sassi and colleagues observed decreased left anterior cingulate volumes in a small cohort of untreated patients (Sassi et al 2004), and both Wilke and Lochhead observed decreased gray matter volumes in portions of the anterior cingulate (Lochhead et al 2004; Wilke et al 2004), all three studies enrolled only small numbers of bipolar patients. Moreover, the latter two, VBM studies identified changes in gray matter volume in spatially distinct portions of the functionally heterogeneous structure. Together with our findings, these studies suggest the presence of anatomic corollaries to previously observed functional changes in at least portions of the anterior cingulate, that are not reflected in ROI studies encompassing the entire structure. Others and we have previously observed neurofunctional changes in fusiform activation in patients with bipolar disorder, as well. The fusiform gyrus may play a role in the inhibition of emotional networks (Strakowski et al 2004; Yamasaki et al 2002) that is impaired in bipolar disorder. As in the prefrontal cortex and anterior cingulate, the changes observed here may be a structural corollary of functional impairment, and suggest that changes in functional activation of the fusiform gyrus during performance of cognitive tasks may represent a primary deficit, rather than exclusively representing a secondary response to deficits elsewhere, as has been previously suggested.
BIOL PSYCHIATRY 2005;58:151–157 155 Similarly, areas of increased gray matter in the primary and supplementary motor cortex may reflect primary deficits associated with bipolar disorder. Although motor dysfunction has not classically been described in patients with bipolar disorder, several studies have reported functional impairment of fine motor control and the presence of neurologic “soft signs” that may be related to neuropathology of the motor cortex (Nasrallah et al 1983; Negash et al 2004; Wilder-Willis et al 2001). We observed decreased gray matter volume in patients with bipolar disorder only in a small region in the right posterior parietal cortex, encompassing a portion of the superior parietal lobule. This is a region that has been suggested to be involved in sensory integration, which has been noted to be impaired in patients with bipolar disorder and corresponds with areas in which we have previously observed increased activation (Adler et al 2004; Negash et al 2004). Several important caveats need to be noted in discussing our findings. The majority of bipolar patients were receiving medication at the time of the MRI scan, and all patients had received either mood stabilizers or antipsychotic medications in the past. Several studies have suggested that these medications may have a neuroprotective effect and potentially impact neuroanatomic findings. A recent study examining morphometric changes in the anterior cingulate of patients with bipolar disorder found that those patients who were receiving medication did not show the same volumetric changes as those who were free of medication for at least one month prior to the scan (Sassi et al 2004). In addition, patients in different mood states were included in our analysis. While morphometric changes with mood have not been demonstrated in the past, current mood state could have an impact on our findings. In interpreting our comparison of first- versus multi-episode bipolar subjects, it is important to note that bipolar subjects who had experienced only a single episode of mania were defined as “first-episode.” Many of these subjects had experienced one or more episodes of depression as well. In addition, nicotine abuse
Figure 3. Voxel-by-voxel comparison map of first-episode bipolar (n ⫽ 13) versus multi-episode bipolar (n ⫽ 10) subjects, overlaid on a T1-weighted coronal anatomic image. Statistically significant differences in gray matter volume were defined p ⱕ .01, with a cluster size of 200.
www.sobp.org/journal
156 BIOL PSYCHIATRY 2005;58:151–157
C.M. Adler et al
Table 3. Regional Coordinates of Areas of Differential Gray Matter Volume Brain Region First-Episode ⬎ Multi-Episode Frontal Cortex L ventral prefrontal cortex (left inferior frontal gyrus) R ventral prefrontal cortex (right inferior frontal gyrus)
Talairach Coordinates
Brodmann Area
Cluster Size
⫺40
35
4
Area 46/47
1400
42
27
1
Area 46/47
576
L, left; R, right.
or addiction was not an exclusion criterion. At least one study has linked nicotine use with decreased gray matter volume in the prefrontal cortex and anterior cingulate (Brody et al 2004). Other caveats include differences in gender distribution between groups and small differences in age. Although age did not significantly vary between groups, even limited differences could give rise to significant confounds. We therefore covaried our comparison to control for both sex- and age-related group differences. This study complements and extends previous voxel based morphometry studies of bipolar disorder and supports suggestions of underlying neuropathology in the anterior cingulate and ventral prefrontal cortex, elements of the anterior limbic network. Our observation of gray matter volume changes in portions of the fusiform gyrus, motor cortex and superior parietal lobule suggests that these regions may also play a primary role in the mood and cognitive symptoms of bipolar disorder. Furthermore, these data raise the possibility that changes in regional gray matter volume may reflect neuropathologic effects of affective symptomatology. If confirmed by further research, these findings support the importance of aggressive intervention to minimize affective symptomatology in bipolar patients. Additional studies examining the effects of illness course and duration, as well as medication status, on neuroanatomy in patients with bipolar disorder may further clarify these findings. Support was provided by The Stanley Medical Research Institute (CMA, SMS) and The Theodore and Vada Stanley Foundation Scholars (AL), the National Alliance for Research on Schizophrenia and Depression (CMA), and National Institute of Health grants MH58170 (SMS), MH63373 (MPD), MH64086 (CMA). We would like to gratefully acknowledge the assistance of Marko Wilke, M.D. Adler CM, Holland SK, Schmithorst V, Tuchfarber MJ, Strakowski SM (2004): Changes in neuronal activation in patients with bipolar disorder during performance of a working memory task. Bipolar Disorder 6:540-549. Ashburner J, Friston KJ (1999): Nonlinear spatial normalization using basis functions. Hum Brain Mapp 7:254 –266. Ashburner J, Friston KJ (2000): Voxel-based morphometry - The methods. Neuroimage 11:805– 821. Brambilla P, Harenski K, Nicoletti M, Mallinger AG, Frank E, Kupfer DJ, et al (2001): Differential effects of age on brain gray matter in bipolar patients and healthy individuals. Neuropsychobiology 43:242–7. Beyer JL, Krishnan KR (2002): Volumetric brain imaging findings in mood disorders. Bipolar Disord 4:89 –104. Blumberg HP, Kaufman J, Martin A, Charney DS, Krystal JH, Peterson BS (2004): Significance of adolescent neurodevelopment for the neural circuitry of bipolar disorder. Ann N Y Acad Sci 1021:376 – 83. Blumberg HP, Leung HC, Skudlarski P, Lacadie CM, Fredericks CA, Harris BC, et al (2003): A functional magnetic resonance imaging study of bipolar
www.sobp.org/journal
disorder: State- and trait-related dysfunction in ventral prefrontal cortices. Arch Gen Psychiatry 60:601–9. Blumberg HP, Martin A, Kaufman J, Leung HC, Skudlarski P, Lacadie C, et al (2003): Frontostriatal abnormalities in adolescents with bipolar disorder: Preliminary observations from functional MRI. Am J Psychiatry 160:1345–7. Brett M (2002): The MNI brain and the Talairach atlas, [Online], Available: http://www.mrc-cbu.cam.ac.uk/Imaging/Common/mnispace.shtml [22 October 2004]. Brody AL, Mandelkern MA, Jarvik ME, Lee GS, Smith EC, Huang JC, et al (2004): Differences between smokers and nonsmokers in regional gray matter volumes and densities. Biol Psychiatry 55:77– 84. Bruno SD, Barker GJ, Cercignani M, Symms M, Ron MA (2004): A study of bipolar disorder using magnetization transfer imaging and voxel-based morphometry. Brain 127:2433–2440. Chard DT, Parker GJM, Griffin CMB, Thompson AJ, Miller DH (2002): The reproducibility and sensitivity of brain tissue volume measurements derived from an SPM-based segmentation methodology. J Magn Reson Imaging 15:259 –267. Crovitz HF, Zener K (1962): A group test for assessing hand- and eye-dominance. Am J Psychology 75:271– 6. Drevets WC, Price JL, Simpson JR Jr, Todd RD, Reich T, Vannier M, et al (1997): Subgenual prefrontal cortex abnormalities in mood disorders. Nature 386:824 –7. First MB, Spitzer RL, Gibbon M, Williams JBW (1997): Structured Clinical Interview for DSM-IV Axis I Disorders – Patient Edition (SCID-I/P, Version 2.0, 4/97 revision). New York, New York: Biometrics Research Department, New York State Psychiatric Institute. Friston KJ, Holmes AP, Worsley KJ, Poline JP, Frith CD, Frackowiak RSJ (1995): Statistical parametric maps in functional imaging: A general linear approach. Hum Brain Mapp 2:189 –210. Geller B, Tillman R, Craney JL, Bolhofner K (2004): Four-year prospective outcome and natural history of mania in children with a prepubertal and early adolescent bipolar disorder phenotype. Arch Gen Psychiatry 61: 459 – 67. Goldberg JF, Harrow M, Grossman LS (1995): Course and outcome in bipolar affective disorder: A longitudinal follow-up study. Am J Psychiatry 152: 379 – 84. Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001): A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 14:21–36. Greenberg PE, Stiglin LE, Finkelstein SN, Berndt ER (1993): The economic burden of depression in 1990. J Clin Psychiatry 54:405–18. Lochhead RA, Parsey RV, Oquendo MA, Mann JJ (2004): Regional brain gray matter volume differences in patients with bipolar disorder as assessed by optimized voxel-based morphometry. Biol Psychiatry 55:1154 – 62. Lopez-Larson MP, DelBello MP, Zimmerman ME, Schwiers ML, Strakowski SM (2002): Regional prefrontal gray and white matter abnormalities in bipolar disorder. Biol Psychiatry 52:93–100. Lyoo IK, Kim MJ, Stoll AL, Demopulos CM, Parow AM, Dager SR, et al (2004): Frontal lobe gray matter density decreases in bipolar I disorder. Biol Psychiatry 55:648 –51. McDonald C, Zanelli J, Rabe-Hesketh S, Ellison-Wright I, Sham P, Kalidindi S, et al (2004): Meta-analysis of magnetic resonance imaging brain morphometry studies in bipolar disorder. Biol Psychiatry 56:411–7. McIntosh AM, Job DE, Moorhead WJ, Harrison LK, Forrester K, Lawrie SM, et al (2004): Voxel-based morphometry of patients with schizophrenia or bipolar disorder and their unaffected relatives. Biol Psychiatry 56:544 –52.
C.M. Adler et al Mega MS, Cummings JL, Salloway S, Malloy P (1997): The limbic system: An anatomic, phylogenetic, and clinical perspective. J Neuropsychiatr Clin Neurosci 9:315–330. Nasrallah HA, Tippin J, McCally WM (1983): Neurologic soft signs in manic patients. A comparison with schizophrenic and control groups. J Affect Disord 5:45–50. Negash A, Kebede D, Alem A, Melaku Z, Deyessa N, Shibire T, et al (2004): Neurological soft signs in bipolar I disorder. J Affect Disord 80:221–30. Ongür D, Price JL (2000): The organization of networks within the orbital and medial prefrontal cortex of rats, monkeys, and humans. Cereb Cortex 10:206 –219. Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, et al (1990): Comorbidity of mental disorders with alcohol and other drug abuse. Results from the Epidemiologic Catchment Area (ECA) Study. JAMA 264: 2511– 8. Sassi RB, Brambilla P, Hatch JP, Nicoletti MA, Mallinger AG, Frank E, et al (2004): Reduced left anterior cingulate volumes in untreated bipolar patients. Biol Psychiatry 56:467–75. Sax KW, Strakowski SM, Zimmerman ME, DelBello MP, Keck PE Jr, Hawkins JM (1999): Frontosubcortical neuroanatomy and the continuous performance test in mania. Am J Psychiatry 156:139 – 41. Strakowski SM, Adler CM, DelBello MP (2002): Volumetric MRI studies of mood disorders: Do they distinguish unipolar and bipolar disorder? Bipolar Disord 4:80 – 8. Strakowski SM, Adler CM, Holland SK, Mills N, DelBello MP (2004: A Preliminary fMRI Study of Sustained Attention in Euthymic, Unmedicated Bipolar Disorder. Neuropsychopharmacology 29:1734 – 40. Strakowski SM, Adler CM, Holland SK, Mills N, DelBello MP, Eliassen JC
BIOL PSYCHIATRY 2005;58:151–157 157 (in press): Abnormal fMRI brain activation in euthymic bipolar disorder during a counting Stroop task. Am J Psychiatry. Strakowski SM, DelBello MP, Adler CM (2005): The functional neuroanatomy of Bipolar Disorder: A review of neuroimaging findings. Mol Psychiatry 10:105-116. Strakowski SM, DelBello M, Adler CM, Cecil KM, Sax KW (2000): Neuroimaging in bipolar disorder. Bipolar Disord 2:148 – 64. Strakowski SM, Keck PE Jr, McElroy SL, West SA, Sax KW, Hawkins JM, et al (1998): Twelve-month outcome after a first hospitalization for affective psychosis. Arch Gen Psychiatry 55:49 –55. Strakowski SM, McElroy SL, Keck PE Jr, West SA (1996): Suicidality among patients with mixed and manic bipolar disorder. Am J Psychiatry 153: 674 – 6. Wilder-Willis KE, Sax KW, Rosenberg HL, Fleck DE, Shear PK, Strakowski SM (2001): Persistent attentional dysfunction in remitted bipolar disorder. Bipolar Disord 3:58 – 62. Wilke M, Kaufmann C, Grabner A, Putz B, Wetter TC, Auer DP (2001): Gray matter changes and correlates of disease severity in schizophrenia: A statistical parametric mapping study. Neuroimage 13:814 –24. Wilke M, Kowatch RA, DelBello MP, Mills NP, Holland SK (2004): Voxel-based morphometry in adolescents with bipolar disorder: First results. Psychiatry Res 131:57– 69. Yamasaki H, LaBar KS, McCarthy G (2002): Dissociable prefrontal brain systems for attention and emotion. Proc Natl Acad Sci U S A 99:11447– 11451. Yurgelun-Todd DA, Gruber SA, Kanayama G, Killgore WD, Baird AA, Young AD (2000): fMRI during affect discrimination in bipolar affective disorder. Bipolar Disord 2:237– 48.
www.sobp.org/journal
B
ipolar disorder is a major psychiatric illness that affects approximately 1.5% of the population and is a significant source of individual morbidity and societal cost (Goldberg et al 1995; Greenberg et al 1993; Regier et al 1990; Strakowski et al 1996, 1998). Although the pathophysiology of bipolar disorder remains poorly understood, findings from functional imaging studies suggest that bipolar disorder is associated with neuropathologic changes in specific brain regions, particularly within the frontal and temporal cortex, and subcortical structures, making up the anterior limbic network (Strakowski et al 2000). The results of morphological studies of these brain regions however, have been mixed and not entirely congruent with functional imaging findings (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2000, 2002). One potential cause of this disparity in findings may be the almost exclusive use of region-of-interest (ROI) based analyses in morphologic studies of patients with bipolar disorder. While useful in hypothesis-driven structural studies, this approach may be compromised by the lack of uniform methodology, and particularly by variations in structural boundaries. The prefrontal cortex, an area widely hypothesized to be involved in the pathophysiology of bipolar disorder, provides an excellent example of this methodological issue. Only a minority of structural studies have observed significant differences between bipolar patients and healthy subjects in prefrontal and frontal cortical
From the Center for Bipolar Disorders Research (CMA, AL, MPD, SMS) and the Center for Imaging Research (CMA, SMS), University of Cincinnati College of Medicine, Cincinnati, Ohio. Address reprint requests to Caleb M. Adler, M.D., University of Cincinnati College of Medicine, Department of Psychiatry, PO Box 670559, 231 Albert Sabin Way, Cincinnati, OH 45267; E-mail: cal.adler@psychiatry. uc.edu. Received December 6, 2004; revised February 15, 2005; accepted March 7, 2005.
0006-3223/05/$30.00 doi:10.1016/j.biopsych.2005.03.022
volume (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2000, 2002), despite differential activation observed across several functional imaging studies (Adler et al 2004; Blumberg et al 2003; Strakowski et al 2004, 2005, in press; Yurgelun-Todd et al 2000), and research suggesting correlations between prefrontal morphometry and cognitive performance in bipolar patients (Sax et al 1999), as well as inverse correlation between illness duration and prefrontal volume in some studies (Brambilla et al 2001; Lopez-Larson et al 2002). Among the possible causes of these discrepancies may be the poor match between functional and easily defined structural boundaries. Difficulties in mapping functional regions to clearly defined anatomic areas may play a role in the frequently ambiguous morphological findings in other portions of the anterior limbic network as well. Morphometric studies of the temporal cortex, and of some medial temporal structures, for instance, have been largely unrevealing, despite strong functional evidence for their involvement in the pathophysiology of bipolar disorder (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2005). Furthermore, ROI based analyses of even fairly easily defined brain regions or structures combine both affected and unaffected areas, potentially vitiating differences. Volumetric studies of midline structures such as the basal ganglia for example, have been inconclusive. While the majority of investigators have not observed changes in striatal volume, some investigators have noted enlargement of portions of the basal ganglia (Beyer and Krishnan 2002; McDonald et al 2004; Strakowski et al 2000, 2002). One alternative to ROI-based approaches is voxel-based morphometry (VBM). In VBM, individual structural MRI scans are normalized to a standard template to allow voxel-by-voxel comparisons. Advantages of this approach include the ability to perform an exploratory analysis without the need to define structures a priori. In addition, VBM analysis allows structural investigation of functional regions, such as portions of the prefrontal cortex that may be difficult to define anatomically. Potential disadvantages however, include the need to control for the large number of comparisons made in a structural study of the whole brain, with the potential loss of statistical power which that entails, and the possibility of blurring precise structural BIOL PSYCHIATRY 2005;58:151–157 © 2005 Society of Biological Psychiatry
152 BIOL PSYCHIATRY 2005;58:151–157
C.M. Adler et al
boundaries implicit in normalizing diverse brains to a single template. Five studies published to date have reported VBM analyses of bipolar disorder. Wilke and colleagues utilized an optimized comparison technique to measure differences in gray matter volume between ten adolescents with bipolar disorder and a matched group of 52 healthy subjects. In contrast to most ROI based studies, Wilke and colleagues observed areas of decreased gray matter volume in the left anterior cingulate, as well as the orbitofrontal cortex and medial temporal structures. Increased volume was observed in portions of the basal ganglia (Wilke et al 2004). Lochhead and colleagues compared groups of 11 bipolar patients and 31 healthy subjects. Bipolar patients were found to have areas of decreased gray matter volume in separate portions of the anterior cingulate and temporal cortex. Areas of increased gray matter volume were observed in the insular/ frontoparietal cortex (Lochhead et al 2004). In contrast, McIntosh and colleagues observed decreased gray matter in the right inferior frontal gyrus, as well as the right insula in a group of 19 bipolar patients from families with a history of both bipolar disorder and schizophrenia. A larger group of 26 bipolar patients from bipolar families demonstrated decreased gray matter only in the thalamus and caudate (McIntosh et al 2004). In the largest study, Bruno and colleagues compared 39 bipolar patients with 35 healthy controls. They failed to identify any differences in gray matter volume between groups. Using a different technique, Lyoo and colleagues measured differences in gray matter density between 39 patients with bipolar disorder to 43 healthy subjects. They observed decreased density in the anterior cingulate and portions of the frontal cortex including the left medial and right inferior frontal gyrus, as well as the right precentral gyrus (Lyoo et al 2004). Lyoo and colleagues (2004) also noted that gray matter density in the left medial and right inferior frontal gyri inversely correlated with the number of manic episodes experienced by bipolar subjects. These studies suggest that differences in gray matter volume and density may be present that are not readily apparent in studies utilizing a more traditional ROI-based approach. Previous studies of gray matter volume however, have for the most part included only small numbers of bipolar patients. Moreover, Wilke and colleagues studied adolescent bipolar patients, a group that may be clinically distinct from adults (Geller et al 2004). In this study we utilized an optimized voxel-based analysis to compare gray matter volumes between a larger group of patients with bipolar disorder, including relatively early onset and first-episode patients, and a matched group of healthy
subjects. Based on previous morphometric studies, we hypothesized that we would find decreases in gray matter volume, particularly within functional regions making up the anterior limbic network, hypothesized to be involved in the neuropathophysiology of bipolar disorder. Further, we hypothesized that our cohort of early onset patients would demonstrate less change in gray matter volume than was observed in previous studies.
Methods and Materials Participants Thirty-two patients with bipolar disorder, type I were recruited from ongoing outcome studies and twenty-seven healthy subjects were recruited in response to local advertisement or by word of mouth. Bipolar and healthy subjects did not statistically differ in age, education, sex, or handedness (Table 1). Subjects with bipolar disorder were free of concurrent axis I psychiatric or medical illness, including substance use disorders (with the exception of nicotine addiction). No healthy subjects had been diagnosed with any axis I psychiatric conditions, including substance dependence, current substance abuse, or major medical conditions (with the exception of nicotine addiction). No healthy controls reported a history of axis I psychiatric conditions in first-degree relatives. No healthy subjects were receiving psychotropic medication at the time of the study. Diagnosis in subjects with bipolar disorder, and the absence of axis I conditions in healthy controls were determined using the Structured Clinical Interview for DSM-IV (SCID) administered by a board certified psychiatrist or Ph.D. level psychologist (First et al 1997). Patients had an average age of onset of 22.5 years (SD ⫾ 7.7), and had been ill for an average of 8.7 years (SD ⫾ 9.2). The number of quantifiable past depressive episodes in subjects with bipolar disorder ranged from zero to eleven with an average of 2.9 (SD ⫾ 3.2) episodes. Eleven subjects reported that their number of depressive episodes had been too numerous to count, or were otherwise unable to accurately quantify the number. The number of past manic and mixed episodes ranged from one to six with an average of 2.0 (SD ⫾ 1.5) episodes. Nine subjects were unable to accurately quantify the number of manic and mixed episodes that they had experienced. Nine subjects were unmedicated at the time of the study. The remainder of the bipolar subjects was receiving standard pharmacotherapy including lithium, divalproex, anti-seizure medications including gabapentin and topiramate, atypical antipsychotics, antidepressants, and benzodiazepines, alone or in combination. While the major-
Table 1. Demographics of Bipolar and Healthy Subjects Characteristics Age, years Education, years Sex, n (%) women Handednessa, n (%) right Age of Onset, years Mood State, manic (%)/depressed (%) Number of Depressive Episodesb Number of Manic Episodesc Medication Status, n (%) current
Bipolar Subjects (n ⫽ 32)
Healthy Subjects (n ⫽ 27)
Group Difference (p-value)
31.2 (9.4) 13.6 (2.2) 13 (41) 31 (97) 22.5 (7.7) 5 (16)/2 (6) 2.9 (3.2) 2.0 (1.5) 23 (72)
30.5 (9.7) 14.0 (1.6) 15 (56) 26 (96) NA NA NA NA NA
.79 .33 .26 .90 NA NA NA NA NA
All values are mean (SD) unless otherwise noted. Handedness was determined using the Crovitz Handedness Inventory (Crovitz and Zener 1962). b n ⫽ 21. c n ⫽ 23. a
www.sobp.org/journal
BIOL PSYCHIATRY 2005;58:151–157 153
C.M. Adler et al ity of the bipolar patients were euthymic at the time of the scan, seven subjects were either manic or depressed (Table 1). All subjects demonstrated facility with English and a clear understanding of the procedures, risks and intent of the study. All subjects provided written informed consent, which was obtained after study procedures had been explained in detail, prior to their participating in the study. This study was approved by the University of Cincinnati College of Medicine Institutional Review Board. Imaging Scans were performed with a 3.0 Tesla Bruker Biospec scanner (Bruker Medizintechnik, Karisruhe, Germany). During the study, subjects reclined in a supine position on the bed of the scanner and a RF coil (Bruker NMR Instruments Inc., Fremont, California) was placed over the participant’s head. Earplugs and headphones were provided to block background noise. Following a 3-plane gradient echo scan for alignment and localization, a shim procedure was performed to generate a homogenous, constant magnetic field. One hundred-twenty contiguous 1 mm axial slices extending superiorly from the inferior aspect of the cerebellum to encompass most of the brain were selected from a sagittal localizer scan. A high-resolution T1-weighted 3-D brain scan was then obtained using a modified driven equilibrium Fourier transform (MDEFT) protocol (TI ⫽ 550 msec, TR ⫽ 16.5, TE ⫽ 4.3 msec, FOV ⫽ 25.6 ⫻ 19.2 ⫻ 14.4, 256 ⫻ 128 ⫻ 96, flip angle ⫽ 20°). Precautions were taken to minimize subject motion during the MRI study by instructing subjects to remain still and packing around their heads with foam padding. Data Pre-Processing All automated image processing was done using statistical parametrical mapping software (SPM99, Wellcome Department of Cognitive Neurology, University College London, United Kingdom) running in MATLAB (MathWorks, Natick, Massachusetts). Determination of the anterior commissure was made in each image by a single investigator (AL). Images were uniformly aligned with regard to head position, to provide optimal starting estimates for subsequent spatial normalization. Analysis followed the “optimized” voxel-based morphometry (VBM) processing strategy of Good and colleagues (Good et al 2001). Images were segmented in native space, and any remaining extra-neuronal tissue was removed by modulation with an extracted, individualized brain mask. These native-space gray matter partitions were then normalized to the corresponding SPM99 tissue-probability map. This procedure weights normalization parameters heavily towards gray matter and minimizes contributions from other tissue types. The normalization parameters from this transformation were then applied to the original images, yielding a normalized whole brain image. These images were then segmented and again modulated with a newly extracted, individual brain mask in order to remove nonbrain tissue. As part of the segmentation procedure, residual image inhomogeneities were removed by modeling smoothly varying intensity changes (Ashburner and Friston 2000). This procedure involves the estimation of an intensity nonuniformity field, which is then applied, to yield a corrected image. This procedure has been shown to markedly increase the reproducibility of SPM99segmentation results (Chard et al 2002). Spatial normalization was achieved using an initial 12-parameter affine transformation, followed by 12 nonlinear iterations using 7 ⫻ 8 ⫻ 7 discrete cosine transform basis functions
(Ashburner and Friston 1999, 2000). Images were written out in 1 ⫻ 1 ⫻ 1 mm resolution. To allow for the detection of true gray matter volume changes, images were modulated by the Jacobian determinant of the normalization matrix, resulting in images that take into account global and local volume changes during spatial normalization (Good et al 2001). Final images were smoothed using a Gaussian kernel with a full-width half-maximum (FWHM) of 12 mm to create a local weighted average of the surrounding pixels. This filter width, as per the matched filter theorem, determines the spatial scale at which changes are most sensitively detected, and also accounts for structural variability and possibly inexact spatial normalization. Image Analysis and Statistics Processed images from both datasets (healthy controls and bipolar patients) were analyzed within SPM99, employing the framework of the general linear model (Friston et al 1995). A model was designed in which age and gender were used as covariates of no interest; diagnosis was considered the parameter of interest. Two contrasts were calculated, testing for a positive or negative correlation of gray matter volume with the parameter of interest. Significance was set at a p-value of p ⫽ .001; with a minimum cluster size of 200 voxels (Wilke et al 2001, 2004). Points of maximum correlation were converted from Montreal Neurological Institute to Talairach coordinates using a nonlinear transformation (Brett 2002). Voxels of maximal correlation within the anterior cingulate and ventral prefrontal cortices were plotted across individual scans, using SPM99. A secondary analysis was performed on the bipolar data set using first episode (defined as having only one manic episode (n ⫽ 13), versus multiple episodes (n ⫽ 10) as the parameter of interest. Contrasts testing for positive or negative correlation of gray matter volume with the parameter of interest were calculated for regions in which significant differences were observed between bipolar and healthy subjects. Because this was a limited post-hoc analysis, significance was set at p ⫽ .01; with a minimum cluster size of 200 voxels (Wilke et al 2001, 2004). Points of maximum correlation were converted from MNI to Talairach coordinates using a nonlinear transformation (Brett 2002).
Results Patients with bipolar disorder showed increased gray matter volume in portions of the anterior cingulate, and of the ventral prefrontal cortex. Increased gray matter was also observed in the temporal cortex, including a portion of the left fusiform gyrus, and in areas involved with motor control, including parts of the supplemental motor cortex and the left precentral gyrus (Figure 1, Table 2). Between group separation was particularly consistent across individuals within each group in the anterior cingulate and ventral prefrontal cortex (Figure 2). Decreased gray matter volume in patients with bipolar disorder was observed only in a portion of the right superior parietal lobule (Figure 1, Table 2). The secondary analysis found increased gray matter volume in first- versus multi-episode patients only in portions of the ventral prefrontal cortex (Figure 3, Table 3).
Discussion In this study we observed increased gray matter volume in bipolar subjects in several regions hypothesized to be part of the anterior limbic network, and to play a role in emotional regulation, attention, and impulse control. These include portions of www.sobp.org/journal
154 BIOL PSYCHIATRY 2005;58:151–157
C.M. Adler et al Areas of decreased gray matter density were identified that substantially overlap regions of increased volume in the anterior cingulate and ventral prefrontal cortex (Lyoo et al 2004), suggesting concurrent changes in volume and density consistent with an apoptotic process. This model is also consistent with differences in findings between this study and that of previous investigations, which for the most part studied cohorts of older and more chronic subjects (Bruno et al 2004; Lochhead et al 2004; McIntosh et al 2004). Conversely, the subjects studied by Wilke and colleagues were adolescents whose prefrontal maturation is presumably incomplete, and who appear to constitute a clinically distinct group (Geller et al 2004). While limited in scope, the areas of greater gray matter volume we observed in first episode patients, in the ventral prefrontal cortex, at least partially support this hypothesis. Moreover, previous functional neuroimaging studies of young bipolar subjects have noted an absence of prefrontal differences (Blumberg et al 2003), leading to suggestions that ventral prefrontal abnormalities may emerge over the course of the illness (Blumberg et al 2004). Areas of increased gray matter volume were observed in several structures of the anterior limbic network, including Brodmann area 47 of the prefrontal cortex. Brodmann area 47 is adjacent and ventral to Brodmann area 10, an area that we have previously observed to activate significantly differently in patients with bipolar disorder compared with healthy subjects during cognitive tasks (Adler et al 2004; Strakowski et al 2004, in press). The ventral prefrontal cortex, including Brodmann areas 10/47, has been associated with elements of mood regulation that are hypothesized to be disrupted in bipolar disorder (Mega et al 1997; Ongür and Price 2000; Yamasaki et al 2002). Few morphometric studies however, have specifically examined this region in patients with bipolar disorder, and those areas of volumetric change that have been observed encompass large, functionally nonspecific portions of the ventral prefrontal cortex (Drevets et al 1997; Lopez-Larson et al 2002). Our findings suggest that morphological markers of neuropathology are present in only portions of the prefrontal cortex, and raise the possibility that functional changes elsewhere in the prefrontal cortex may be secondary to pathology in these ventral prefrontal regions.
Figure 1. Voxel-by-voxel comparison map of bipolar (n ⫽ 32) versus control (n ⫽ 27) subjects, overlaid on a T1-weighted anatomic image. Statistically significant differences in gray matter volume were defined p ⱕ .001, with a cluster size of 200.
the anterior cingulate, the ventral prefrontal cortex, and the fusiform gyrus. In addition, we observed increased gray matter volume in portions of the primary and secondary motor cortex. Decreased gray matter was observed only in a portion of the superior parietal lobule. Although somewhat speculative, differences in gray matter volume may represent anatomic markers of neuropathology. Decreased gray matter in the superior parietal lobule suggests neuronal apoptosis, or a loss of neuropil. In this study however, we identified several areas of increased gray matter volume in patients with bipolar disorder. We hypothesize that increased gray matter may be related to preapoptotic osmotic changes or hypertrophy, marking areas of early neuronal pathology. This suggestion is consistent with measures of gray matter density made by Lyoo and colleagues.
Table 2. Regional Coordinates of Areas of Differential Gray Matter Volume Brain Region Bipolar ⬎ Control Frontal Cortex L anterior cingulate R anterior cingulate L ventral prefrontal cortex (middle frontal gyrus) R ventral prefrontal cortex (middle frontal gyrus) L precentral gyrus Temporal Cortex L fusiform gyrus Parietal Cortex L supplemental motor cortex R supplemental motor cortex Control ⬎ Bipolar Parietal Cortex R superior parietal lobule L, left; R, right. a All regions p ⱕ .001. b Regions are contiguous.
www.sobp.org/journal
Talairach Coordinates
Brodmann Area
Cluster Sizea
787b 787b 370
⫺3 3 ⫺46
47 47 36
6 6 ⫺1
Area 42 Area 42 Area 47
40
40
⫺5
Area 47
1028
⫺49
4
6
Area 44
210
⫺53
⫺44
⫺5
Area 37
1087
⫺18 31
10 14
⫺53 51
Area 6 Area 6
225 608
43
⫺66
51
Area 7
466
C.M. Adler et al
Figure 2. Scan by scan fitted response of bipolar (n ⫽ 32) versus control (n ⫽ 27) subjects for anterior cingulate, left and right ventral prefrontal cortex. Between group differences were particularly consistent across individual subjects in these regions (fitted response: solid gray line; error: dashed line).
Despite cognitive and functional imaging based evidence of neurofunctional deficits in at least portions of the anterior cingulate, previous volumetric findings have been equivocal. With few exceptions, morphometry studies have failed to identify significant structural differences between bipolar and healthy subjects. Although Sassi and colleagues observed decreased left anterior cingulate volumes in a small cohort of untreated patients (Sassi et al 2004), and both Wilke and Lochhead observed decreased gray matter volumes in portions of the anterior cingulate (Lochhead et al 2004; Wilke et al 2004), all three studies enrolled only small numbers of bipolar patients. Moreover, the latter two, VBM studies identified changes in gray matter volume in spatially distinct portions of the functionally heterogeneous structure. Together with our findings, these studies suggest the presence of anatomic corollaries to previously observed functional changes in at least portions of the anterior cingulate, that are not reflected in ROI studies encompassing the entire structure. Others and we have previously observed neurofunctional changes in fusiform activation in patients with bipolar disorder, as well. The fusiform gyrus may play a role in the inhibition of emotional networks (Strakowski et al 2004; Yamasaki et al 2002) that is impaired in bipolar disorder. As in the prefrontal cortex and anterior cingulate, the changes observed here may be a structural corollary of functional impairment, and suggest that changes in functional activation of the fusiform gyrus during performance of cognitive tasks may represent a primary deficit, rather than exclusively representing a secondary response to deficits elsewhere, as has been previously suggested.
BIOL PSYCHIATRY 2005;58:151–157 155 Similarly, areas of increased gray matter in the primary and supplementary motor cortex may reflect primary deficits associated with bipolar disorder. Although motor dysfunction has not classically been described in patients with bipolar disorder, several studies have reported functional impairment of fine motor control and the presence of neurologic “soft signs” that may be related to neuropathology of the motor cortex (Nasrallah et al 1983; Negash et al 2004; Wilder-Willis et al 2001). We observed decreased gray matter volume in patients with bipolar disorder only in a small region in the right posterior parietal cortex, encompassing a portion of the superior parietal lobule. This is a region that has been suggested to be involved in sensory integration, which has been noted to be impaired in patients with bipolar disorder and corresponds with areas in which we have previously observed increased activation (Adler et al 2004; Negash et al 2004). Several important caveats need to be noted in discussing our findings. The majority of bipolar patients were receiving medication at the time of the MRI scan, and all patients had received either mood stabilizers or antipsychotic medications in the past. Several studies have suggested that these medications may have a neuroprotective effect and potentially impact neuroanatomic findings. A recent study examining morphometric changes in the anterior cingulate of patients with bipolar disorder found that those patients who were receiving medication did not show the same volumetric changes as those who were free of medication for at least one month prior to the scan (Sassi et al 2004). In addition, patients in different mood states were included in our analysis. While morphometric changes with mood have not been demonstrated in the past, current mood state could have an impact on our findings. In interpreting our comparison of first- versus multi-episode bipolar subjects, it is important to note that bipolar subjects who had experienced only a single episode of mania were defined as “first-episode.” Many of these subjects had experienced one or more episodes of depression as well. In addition, nicotine abuse
Figure 3. Voxel-by-voxel comparison map of first-episode bipolar (n ⫽ 13) versus multi-episode bipolar (n ⫽ 10) subjects, overlaid on a T1-weighted coronal anatomic image. Statistically significant differences in gray matter volume were defined p ⱕ .01, with a cluster size of 200.
www.sobp.org/journal
156 BIOL PSYCHIATRY 2005;58:151–157
C.M. Adler et al
Table 3. Regional Coordinates of Areas of Differential Gray Matter Volume Brain Region First-Episode ⬎ Multi-Episode Frontal Cortex L ventral prefrontal cortex (left inferior frontal gyrus) R ventral prefrontal cortex (right inferior frontal gyrus)
Talairach Coordinates
Brodmann Area
Cluster Size
⫺40
35
4
Area 46/47
1400
42
27
1
Area 46/47
576
L, left; R, right.
or addiction was not an exclusion criterion. At least one study has linked nicotine use with decreased gray matter volume in the prefrontal cortex and anterior cingulate (Brody et al 2004). Other caveats include differences in gender distribution between groups and small differences in age. Although age did not significantly vary between groups, even limited differences could give rise to significant confounds. We therefore covaried our comparison to control for both sex- and age-related group differences. This study complements and extends previous voxel based morphometry studies of bipolar disorder and supports suggestions of underlying neuropathology in the anterior cingulate and ventral prefrontal cortex, elements of the anterior limbic network. Our observation of gray matter volume changes in portions of the fusiform gyrus, motor cortex and superior parietal lobule suggests that these regions may also play a primary role in the mood and cognitive symptoms of bipolar disorder. Furthermore, these data raise the possibility that changes in regional gray matter volume may reflect neuropathologic effects of affective symptomatology. If confirmed by further research, these findings support the importance of aggressive intervention to minimize affective symptomatology in bipolar patients. Additional studies examining the effects of illness course and duration, as well as medication status, on neuroanatomy in patients with bipolar disorder may further clarify these findings. Support was provided by The Stanley Medical Research Institute (CMA, SMS) and The Theodore and Vada Stanley Foundation Scholars (AL), the National Alliance for Research on Schizophrenia and Depression (CMA), and National Institute of Health grants MH58170 (SMS), MH63373 (MPD), MH64086 (CMA). We would like to gratefully acknowledge the assistance of Marko Wilke, M.D. Adler CM, Holland SK, Schmithorst V, Tuchfarber MJ, Strakowski SM (2004): Changes in neuronal activation in patients with bipolar disorder during performance of a working memory task. Bipolar Disorder 6:540-549. Ashburner J, Friston KJ (1999): Nonlinear spatial normalization using basis functions. Hum Brain Mapp 7:254 –266. Ashburner J, Friston KJ (2000): Voxel-based morphometry - The methods. Neuroimage 11:805– 821. Brambilla P, Harenski K, Nicoletti M, Mallinger AG, Frank E, Kupfer DJ, et al (2001): Differential effects of age on brain gray matter in bipolar patients and healthy individuals. Neuropsychobiology 43:242–7. Beyer JL, Krishnan KR (2002): Volumetric brain imaging findings in mood disorders. Bipolar Disord 4:89 –104. Blumberg HP, Kaufman J, Martin A, Charney DS, Krystal JH, Peterson BS (2004): Significance of adolescent neurodevelopment for the neural circuitry of bipolar disorder. Ann N Y Acad Sci 1021:376 – 83. Blumberg HP, Leung HC, Skudlarski P, Lacadie CM, Fredericks CA, Harris BC, et al (2003): A functional magnetic resonance imaging study of bipolar
www.sobp.org/journal
disorder: State- and trait-related dysfunction in ventral prefrontal cortices. Arch Gen Psychiatry 60:601–9. Blumberg HP, Martin A, Kaufman J, Leung HC, Skudlarski P, Lacadie C, et al (2003): Frontostriatal abnormalities in adolescents with bipolar disorder: Preliminary observations from functional MRI. Am J Psychiatry 160:1345–7. Brett M (2002): The MNI brain and the Talairach atlas, [Online], Available: http://www.mrc-cbu.cam.ac.uk/Imaging/Common/mnispace.shtml [22 October 2004]. Brody AL, Mandelkern MA, Jarvik ME, Lee GS, Smith EC, Huang JC, et al (2004): Differences between smokers and nonsmokers in regional gray matter volumes and densities. Biol Psychiatry 55:77– 84. Bruno SD, Barker GJ, Cercignani M, Symms M, Ron MA (2004): A study of bipolar disorder using magnetization transfer imaging and voxel-based morphometry. Brain 127:2433–2440. Chard DT, Parker GJM, Griffin CMB, Thompson AJ, Miller DH (2002): The reproducibility and sensitivity of brain tissue volume measurements derived from an SPM-based segmentation methodology. J Magn Reson Imaging 15:259 –267. Crovitz HF, Zener K (1962): A group test for assessing hand- and eye-dominance. Am J Psychology 75:271– 6. Drevets WC, Price JL, Simpson JR Jr, Todd RD, Reich T, Vannier M, et al (1997): Subgenual prefrontal cortex abnormalities in mood disorders. Nature 386:824 –7. First MB, Spitzer RL, Gibbon M, Williams JBW (1997): Structured Clinical Interview for DSM-IV Axis I Disorders – Patient Edition (SCID-I/P, Version 2.0, 4/97 revision). New York, New York: Biometrics Research Department, New York State Psychiatric Institute. Friston KJ, Holmes AP, Worsley KJ, Poline JP, Frith CD, Frackowiak RSJ (1995): Statistical parametric maps in functional imaging: A general linear approach. Hum Brain Mapp 2:189 –210. Geller B, Tillman R, Craney JL, Bolhofner K (2004): Four-year prospective outcome and natural history of mania in children with a prepubertal and early adolescent bipolar disorder phenotype. Arch Gen Psychiatry 61: 459 – 67. Goldberg JF, Harrow M, Grossman LS (1995): Course and outcome in bipolar affective disorder: A longitudinal follow-up study. Am J Psychiatry 152: 379 – 84. Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001): A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 14:21–36. Greenberg PE, Stiglin LE, Finkelstein SN, Berndt ER (1993): The economic burden of depression in 1990. J Clin Psychiatry 54:405–18. Lochhead RA, Parsey RV, Oquendo MA, Mann JJ (2004): Regional brain gray matter volume differences in patients with bipolar disorder as assessed by optimized voxel-based morphometry. Biol Psychiatry 55:1154 – 62. Lopez-Larson MP, DelBello MP, Zimmerman ME, Schwiers ML, Strakowski SM (2002): Regional prefrontal gray and white matter abnormalities in bipolar disorder. Biol Psychiatry 52:93–100. Lyoo IK, Kim MJ, Stoll AL, Demopulos CM, Parow AM, Dager SR, et al (2004): Frontal lobe gray matter density decreases in bipolar I disorder. Biol Psychiatry 55:648 –51. McDonald C, Zanelli J, Rabe-Hesketh S, Ellison-Wright I, Sham P, Kalidindi S, et al (2004): Meta-analysis of magnetic resonance imaging brain morphometry studies in bipolar disorder. Biol Psychiatry 56:411–7. McIntosh AM, Job DE, Moorhead WJ, Harrison LK, Forrester K, Lawrie SM, et al (2004): Voxel-based morphometry of patients with schizophrenia or bipolar disorder and their unaffected relatives. Biol Psychiatry 56:544 –52.
C.M. Adler et al Mega MS, Cummings JL, Salloway S, Malloy P (1997): The limbic system: An anatomic, phylogenetic, and clinical perspective. J Neuropsychiatr Clin Neurosci 9:315–330. Nasrallah HA, Tippin J, McCally WM (1983): Neurologic soft signs in manic patients. A comparison with schizophrenic and control groups. J Affect Disord 5:45–50. Negash A, Kebede D, Alem A, Melaku Z, Deyessa N, Shibire T, et al (2004): Neurological soft signs in bipolar I disorder. J Affect Disord 80:221–30. Ongür D, Price JL (2000): The organization of networks within the orbital and medial prefrontal cortex of rats, monkeys, and humans. Cereb Cortex 10:206 –219. Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, et al (1990): Comorbidity of mental disorders with alcohol and other drug abuse. Results from the Epidemiologic Catchment Area (ECA) Study. JAMA 264: 2511– 8. Sassi RB, Brambilla P, Hatch JP, Nicoletti MA, Mallinger AG, Frank E, et al (2004): Reduced left anterior cingulate volumes in untreated bipolar patients. Biol Psychiatry 56:467–75. Sax KW, Strakowski SM, Zimmerman ME, DelBello MP, Keck PE Jr, Hawkins JM (1999): Frontosubcortical neuroanatomy and the continuous performance test in mania. Am J Psychiatry 156:139 – 41. Strakowski SM, Adler CM, DelBello MP (2002): Volumetric MRI studies of mood disorders: Do they distinguish unipolar and bipolar disorder? Bipolar Disord 4:80 – 8. Strakowski SM, Adler CM, Holland SK, Mills N, DelBello MP (2004: A Preliminary fMRI Study of Sustained Attention in Euthymic, Unmedicated Bipolar Disorder. Neuropsychopharmacology 29:1734 – 40. Strakowski SM, Adler CM, Holland SK, Mills N, DelBello MP, Eliassen JC
BIOL PSYCHIATRY 2005;58:151–157 157 (in press): Abnormal fMRI brain activation in euthymic bipolar disorder during a counting Stroop task. Am J Psychiatry. Strakowski SM, DelBello MP, Adler CM (2005): The functional neuroanatomy of Bipolar Disorder: A review of neuroimaging findings. Mol Psychiatry 10:105-116. Strakowski SM, DelBello M, Adler CM, Cecil KM, Sax KW (2000): Neuroimaging in bipolar disorder. Bipolar Disord 2:148 – 64. Strakowski SM, Keck PE Jr, McElroy SL, West SA, Sax KW, Hawkins JM, et al (1998): Twelve-month outcome after a first hospitalization for affective psychosis. Arch Gen Psychiatry 55:49 –55. Strakowski SM, McElroy SL, Keck PE Jr, West SA (1996): Suicidality among patients with mixed and manic bipolar disorder. Am J Psychiatry 153: 674 – 6. Wilder-Willis KE, Sax KW, Rosenberg HL, Fleck DE, Shear PK, Strakowski SM (2001): Persistent attentional dysfunction in remitted bipolar disorder. Bipolar Disord 3:58 – 62. Wilke M, Kaufmann C, Grabner A, Putz B, Wetter TC, Auer DP (2001): Gray matter changes and correlates of disease severity in schizophrenia: A statistical parametric mapping study. Neuroimage 13:814 –24. Wilke M, Kowatch RA, DelBello MP, Mills NP, Holland SK (2004): Voxel-based morphometry in adolescents with bipolar disorder: First results. Psychiatry Res 131:57– 69. Yamasaki H, LaBar KS, McCarthy G (2002): Dissociable prefrontal brain systems for attention and emotion. Proc Natl Acad Sci U S A 99:11447– 11451. Yurgelun-Todd DA, Gruber SA, Kanayama G, Killgore WD, Baird AA, Young AD (2000): fMRI during affect discrimination in bipolar affective disorder. Bipolar Disord 2:237– 48.
www.sobp.org/journal