Sage Journals: Discover world-class research

Abstract

Objective: To investigate structural abnormalities in bipolar disorder (BD) using optimized voxel-based morphometry (VBM) in closely matched patients and controls, and to examine the relationship of clinical features with regional gray matter (GM) volumes.

Methods: Twenty-four patients (six male) aged 19–59 years (mean=38.21 years, SD=11.04 years) with DSM-IV bipolar I disorder were compared with 25 control subjects, matched on age, sex, and years of education. VBM analyses were conducted on high-resolution T1-weighted brain magnetic resonance imaging to detect regional GM volume differences between groups, ensuring statistical correlation for age, sex and total intracranial volumes. Within the patient groups, regional GM changes were also investigated.

Results: Compared to controls, BD patients had increased GM volume in left parahippocampal gyrus and decreased GM volume in left middle temporal gyrus. Family history, psychotic symptoms and lithium status were associated with regional GM abnormalities in BD patients.

Conclusions: This study presents evidence of gray matter volume abnormalities in adults with bipolar I disorder. Regional variation in relation to clinical factors suggests a neurobiological basis for clinical heterogeneity and posits the possibility of trait deficits.

Keywords

bipolar disorder gray matter voxel-based morphometry

Bipolar disorder (BD) is one of the leading causes of disability worldwide. While many recent advances have occurred in understanding the disorder, its pathophysiology remains elusive. Recent advances in neuroimaging technologies have made it possible to investigate structural and functional abnormalities in the brain of bipolar patients in vivo. Structural abnormalities have been described in particular brain regions, for example prefrontal cortex, anterior cingulate cortex and amygdala, but the findings have not been consistent [1].

The inconsistency in the volumetric studies of brain structures is quite striking. For example, temporal lobe volumes of BD patients vary considerably across different studies [2–5]. This inconsistency also applied to amygdale volume [6, 7]. The discrepancy may in part be attributable to different methods and insufficient image quality used to study brain volumes in early studies. Regions of interest (ROI)-based structural magnetic resonance imaging (MRI) approaches have been widely used in morphometric studies, but measurement methods differ, and bias is unavoidable because of rater subjectivity in measuring ROI. Furthermore, the selection of ROI is based on assumptions that may not have empirical validity.

Voxel-based morphometry (VBM) is an assumption-free method [8], which applies voxel-wise comparisons throughout the brain to detect differences in gray matter (GM) concentration in different study groups, providing objective and operator-independent results. Optimized VBM analysis has been developed to allow comparisons of volumes [9] rather than concentrations of brain regions that are used in traditional VBM protocol, by incorporating a modulation step and improving the quality of segmentation with a customized template.

A number of studies of BD using VBM have been published [10–16]. Conflicting results suggest that although regional brain changes are present in bipolar patients, they can be affected by different populations and sample sizes of the study, as well as different clinical features of BD patients studied. Given the lack of consistent findings, we undertook this study to replicate some of the reported findings using optimized VBM in closely matched patients and controls. We broadly hypothesize that GM abnormalities will be present in the temporal lobes of BD patients. Further, we wished to examine the relationship between abnormalities in GM and the clinical features of mood disorder.

Method

Subjects

Twenty-four patients (six male) aged 19–59 years (mean = 38.21 years, SD = years) with DSM-IV bipolar I disorder were recruited from the Prince of Wales Hospital in Sydney and a specialist Bipolar Disorder Clinic. Diagnoses of the patients were made by research psychiatrists based on the Structured Clinical Interview for DSM-IV Patient Edition (SCID-P) [17], supplemented by case note review. Subjects were excluded if they had a history of ongoing substance misuse, neurological disease or closed head injury, or had an additional Axis-I or Axis-II DSM-IV diagnosis, or a medical disorder currently necessitating treatment. Patients were compared with volunteers recruited through local advertisement. Twenty-five control subjects were screened for personal or family history of neurological or psychiatric disorder by Structured Clinical Interview for DSM-III-R Nonpatient version (SCID-NP) [17], which has been used previously. They underwent the same clinical assessments and self-report questionnaires as the patients immediately prior to scanning. The Prince of Wales Hospital and University of New South Wales human research ethics committees approved the study and all participants provided written informed consent.

All participants were interviewed and clinical information regarding age, sex, years of education, medical history, duration of illness and medications was collected. Patients who had at least one first-degree relative with an affective disorder were defined as having a positive family history. Psychosis was defined if hallucinations and/or delusions had occurred during at least one affective episode for BD patients. At the time of MRI scanning, 12 patients were taking lithium (mean daily dose = 975.1±213.2 mg) with a mean plasma level of 0.77±0.1 mmol L⁻¹. Four of those patients taking lithium were also receiving sodium valproate. Eight patients were taking valproate, carbamazepine or a combination of the two, and the remaining four patients were unmedicated.

MRI acquisition

Imaging was conducted on a 1.5 T GE Signa scanner (GE Medical Systems, Milwaukee, WI, USA). A 2-D scout mid-sagittal cut for anterior commissure-posterior commisure (AC-PC) plane alignment was first acquired. 3-D structural MRI was acquired using a T1-weighted fast spoiled gradient echo (FSPGR) sequence (TR/TE = 12.2ms/5.3ms; flip angle = 25^o; matrix size = 256×256; field of view = 250×250 mm, yielding coronal slices 1.6 mm thick with no gaps and an in-plane spatial resolution of 0.977×0.977 mm pixel⁻¹.

Voxel-based morphometry

Optimized VBM analysis [9] was applied using SPM2 (Wellcome Department of Cognitive Neurology, Institute of Neurology, London, UK) on Matlab 6.5 (Math Works, Natick, MA, USA). All imaging analyses were performed on a Windows 2000 workstation. Before VBM analysis was conducted, all images were inspected to ensure their quality for further analyses. A customized template was created from all included images by normalizing, averaging and smoothing. The structural MRI images were spatially normalized and segmented to create a GM template. Raw structural images were segmented into GM, white matter (WM) and cerebrospinal fluid (CSF) partitions. The total GM volumes, WM volumes and CSF volumes were computed by adding the GM, WM and CSF segments of images in native space for each subject. GM partitions were then normalized into the GM template to estimate optimal normalization parameters, which were reapplied to the original raw images so as to generate whole brain images in stereotactic space. These optimally normalized images were segmented to create GM partitions, which were modulated by multiplying voxel values with the Jacobian determinants derived from spatial normalization, to obtain absolute volume of GM [9]. Finally, to satisfy the Gaussian distribution assumption for statistical analysis to test regional differences, images were smoothed with a 12 mm FWHM Gaussian kernel.

Statistical analysis

Total GM volumes of the BD and control groups were compared using ANCOVA, controlling for age, sex and total intracranial volume (TIV). Regional GM volume differences were detected by applying an ANCOVA to form statistical parametric maps. Age, sex and TIV were included in the model as covariates to avoid confounding effects of these variables. VBM analyses were applied further to examine regional GM abnormalities in relation to clinical features, for example family history of mood disorder, psychotic symptoms, lithium taken or not, episodes of mania and depression and duration of disease. For each clinical feature, ANCOVA models were used to investigate effects of discrete variables on GM, and multiple regression models were used to examine effects of continuous variables on GM. Statistical threshold was set initially at p < 0.001, uncorrected with extended threshold of 200 contiguous voxels [10, 16] and further at p < 0.05, family-wise corrected for multiple comparisons over the whole brain [18], which was chosen to control type I error rate in multiple comparisons conservatively. Anatomical localization of results such as cerebral lobes and Brodmann areas (BA) were determined by an automatic labelling method [19].

Results

Twenty-four BD patients (mean age = 38.21 years, SD = 11.04 years) and 25 control subjects (mean age = 38.44 years, SD = 11.05 years), including six and seven male patients, respectively, were recruited into the study. BD group and control group did not differ in age, sex or years of education (p < 0.05). All subjects were right-handed. Fourteen patients in the BD group (58.33%) had positive family history, and the same number presented with psychotic features. BD patients had an average illness duration of 14.17±10.26 years, with 20.92±18.56 episodes of mania or depression. At the time of writing, 12 of them (50%) were on lithium and 21 (87.5%) were on mood stabilizing psychotropic medication. Total brain measures of BD patients were not different from the control group before or after controlling for age, sex and education (Table 1). The comparison of total brain measures within subgroups in BD and controls is given in Table 2. There was no significant difference in total brain volumetric measures in relation to clinical features between BD group and controls.

Table 1.

Brain volumetric results comparing bipolar disorder patients and controls, adjusting for age, sex and education

Volumes (mL)	Bipolar (n = 24)	Control (n = 25)	F	p
Total intracranial volume	1643.98±341.97	1547.30 (146.12)	2.104	0.154
Total gray matter volume	741.69±115.33	714.98 (67.10)	1.408	0.242
Total white matter volume	473.35±135.39	447.47 (55.06)	1.028	0.316
Total cerebrospinal fluid volume	428.95±115.67	384.85 (42.89)	3.436	0.071

Table 2.

Brain volumetric comparisons in relation to clinical features within bipolar disorder group, adjusting for age, sex and education

Volumes (mL)	Clinical features	Bipolar (n = 24)		F	p
		Yes	No
Total intracranial volume	Family history	1689.85 (441.00)	1579.77 (100.39)	0.186	0.671
	Psychosis	1581.37 (123.65)	1731.64 (511.99)	0.572	0.459
	Lithium taking	1576.70 (99.10)	1711.26 (474.15)	1.722	0.205
Total gray matter volume	Family History	764.32 (145.99)	710.00 (36.03)	0.482	0.496
	Psychosis	738.90 (76.11)	745.59 (159.99)	0.037	0.849
	Lithium taking	722.35 (51.78)	761.02 (155.94)	1.801	0.195
Total white matter volume	Family History	497.74 (172.93)	439.21 (37.78)	0.468	0.502
	Psychosis	451.53 (53.52)	503.89 (202.31)	0.516	0.481
	Lithium taking	456.13 (37.35)	493.57 (189.84)	1.411	0.250
Total cerebrospinal fluid volume	Family History	427.80 (148.82)	430.55 (46.84)	0.030	0.864
	Psychosis	390. 94 (57.81)	482.15 (154.83)	2.636	0.121
	Lithium taking	401.23 (52.67)	456.67 (153.37)	1.313	0.266

The results of VBM analyses are presented in Table 3, Table 4 and Figure 1. Compared to healthy subjects, patients with BD had greater left parahippocampal gyrus GM volume (x = − 17, y = − 19, z = − 22, p_uncorrected<0.001) and less left middle temporal gyrus GM volume (x = − 44, y = − 74, z = 13, p_corrected<0.05). No increase in GM volume was found in BD subjects as compared to controls.

Figure 1.

SPM (T) map depicted in a glass brain, showing (a) reduced gray matter volumes and (b) greater gray matter volumes in bipolar subjects compared with controls, after controlling effects of age, total intracranial volume and sex. Cluster extension set at 200 contiguous voxel, significance threshold of p < 0.001, uncorrected for multiple comparison.

Table 3.

VBM analyses of patients with different clinical features vs control subjects

Contrasts	Relation	Coordinates			Regions	BA	t	p_u	p_c
		x	y	z
Patients with family history vs controls	Increase	52	−2	37	Right precentral gyrus	6	3.79	0.000	0.799
	Decrease	−43	−74	14	Left middle temporal gyrus	39	4.64	0.000	0.175
Patients without family history vs controls	Increase	−17	−20	−19	Left parahippocampal gyrus	28	4.90	0.000	0.091
	Decrease	−44	−74	13	Left middle temporal gyrus	39	4.29	0.000	0.380
		40	−71	28	Right middle temporal gyrus	39	4.16	0.000	0.485
		43	−66	34	Right precuneus	39	3.92	0.000	0.693
		−42	−58	38	Left angular gyrus	39	3.62	0.000	0.901
Patients with psychosis symptoms vs controls	Increase	50	−1	37	Right precentral gyrus	6	4.76	0.000	0.132
		52	17	43	Right middle frontal gyrus	8	3.85	0.000	0.755
	Decrease	−44	−73	13	Left middle temporal gyrus	39	4.28	0.000	0.391
Patients without psychosis symptoms vs controls	Increase	NA
	Decrease	−45	−75	14	Left Middle Temporal Gyrus	39	5.10	0.000	0.055
		−48	−81	9	Left Middle Occipital Gyrus	19	4.10	0.000	0.119
		−35	−74	29	Left Superior Occipital Gyrus	19	4.83	0.000	0.514
		41	−60	29	Middle Temporal Gyrus	39	3.85	0.000	0.754
Patients taking lithium vs controls	Increase	NA
	Decrease	−43	−72	14	Left middle temporal gyrus	39	4.72	0.000	0.145
		−49	−83	11	Left middle occipital gyrus	19	3.55	0.000	0.933
Patients not taking lithium vs controls	Increase	−17	−19	−22	Left parahippocampal gyrus	28	4.32	0.000	0.351
	Decrease	−44	−74	13	Left middle temporal gyrus	39	4.37	0.000	0.318

BA, Brodmann areas; VBM, voxel-based morphometry; NA, no significant regions available.

Table 4.

VBM analyses of clinical features within BD patients

Contrasts	Relation	Coordinates			Regions	BA	t	p_u	p_c
		x	y	z
BD patients with family history vs without	Increase	32	53	−3	Right Middle Frontal Gyrus	10	4.07	0.000	0.56
		31	59	5	Right Middle Frontal Gyrus	10	3.68	0.000	0.869
		31	61	17	Right Superior Frontal Gyrus	10	3.55	0.000	0.934
		−27	38	25	Left Middle Frontal Gyrus	9	3.90	0.000	0.713
	Decrease	NA
BD patients with psychosis symptoms vs without	Increase	20	25	8	Right Caudate Body	–	4.95	0.000	0.081
		−21	23	45	Left Superior Frontal Gyrus	8	4.32	0.000	0.361
		49	−2	36	Right Precentral Gyrus	6	4.23	0.000	0.432
		53	−1	45	Right Precentral Gyrus	6	3.73	0.000	0.846
		48	3	50	Right Middle Frontal Gyrus	6	3.59	0.000	0.921
		8	−78	43	Right Precuneus	7	3.72	0.000	0.848
		15	−71	48	Right Precuneus	7	3.66	0.000	0.887
		43	16	33	Right Middle Frontal Gyrus	9	3.64	0.000	0.899
		0	33	8	Left Anterior Cingulate	24	3.59	0.000	0.922
	Decrease	NA
BD patients on lithium vs not on lithium	Increase	NA
	Decrease	17	25	−30	Right orbital gyrus	47	3.98	0.000	0.637
Duration of illness	Positive	NA
	Negative	−12	15	−3	Left caudate head	–	4.11	0.000	0.921
Episodes of mania and depression	Positive	NA
	Negative	54	−32	32	Right inferior parietal lobule	40	4.23	0.000	0.881

BA, Brodmann areas; BD, bipolar disorder; VBM, voxel-based morphometry; NA, no significant regions available.

In the VBM analyses of BD patients with different clinical features compared with healthy controls, family history, psychotic symptoms and being on lithium had various effects on regional GM abnormalities (Table 4). Reduced GM in left middle temporal gyrus was found in all comparisons between subgroups of patients and healthy controls. Analyses within the BD patients found different patterns of GM changes in relation to clinical characteristic variables where initial uncorrected threshold was used. However, differences shown in these regions were not significant after stringent threshold corrections for multiple comparisons over the whole brain (p_corrected<0.05) were applied. We therefore report only in Table 4 clusters thresholded by uncorrected p< 0.001, and only if a cluster consists of >200 voxels, to include clusters that had a low likelihood of being detected by chance. Similar approaches have previously been used in the published literature [10, 16].

Discussion

Global volumetric measures

The present study shows that there was no difference in total intracranial volume, total GM volume and total WM volume between BD patients and healthy controls. This is consistent with previous studies in which generalized changes of brain volumes were not shown by volumetric methods in subjects with affective disorder [20–22], with the exception of one report of ventricular enlargement in BD [23]. A noteworthy aspect of these associations is that the relationship between increased ventricular volumes and BD has been observed mostly in subjects with psychotic symptoms [24], which were not a homogenous characteristic of the present group, with only 58.33% in the present study having had psychotic symptoms. Increased CSF volumes in relation to psychotic features were not observed in subgroup analyses either, and in fact those with psychosis had a trend toward reduced CSF volume. A linear trend for a relationship between ventricular enlargement and number of previous episodes of illness [25, 26], as well as number of hospital stays has been reported [27], suggesting the role of severity of illness and its duration on ventricular size. However, this trend was not observed in the current study. It should be noted that the total volume of CSF space used in the present study, included volumes of ventricular space as well as subarachnoid space, which were examined separately in previous studies. Sulcal widening was not a significant change in BD patients in previous studies [13], which might partly explain the discrepancy between the present study and some others.

Regional abnormalities in BD patients

The main finding of the present study was the GM reduction in the left middle temporal gyrus (BA 39) in BD patients as compared with control subjects. This finding remained significant even after a conservative statistical threshold was applied. Furthermore, it was consistently significant in the VBM comparisons between subgroups of the BD and control groups. Regional abnormalities of GM volume in temporal lobes have been described in previous VBM studies of BD, although anatomical locations had varied. GM reductions were found in bilateral medial temporal lobe structures [10], left ventromedial temporal cortex [11], left amygdala [12] and left superior temporal gyrus [28] of BD patients as compared with controls. For some of these regions, for example the amygdala, there are conflicting findings from ROI methods showing either increased [6, 29, 30] or decreased volume [31, 32]. It seems that there is diversity of change in temporal structures in relation to BD, although different methods might also account for this. In the present study BD patients had reduced GM in the left middle temporal gyrus (BA 39) as well, a novel finding that has not been reported by other VBM studies. Of interest is the positron emission tomography report that this region was activated during both positive and negative emotional processing [33]. Also, healthy subjects, but not BD depressed patients, show activation on functional MRI (fMRI) in this region during positive and negative emotional changes when presented with designed captioned pictures [34]. Clearly, further study is necessary to investigate the role of this region in the development and progression of BD.

Interestingly, increased GM volume was detected in left parahippocampal gyri of BD subjects in the present study, although this result was not significant if a conservative statistical threshold was applied. Frazier et al. found that parahippocampal gyrus was enlarged bilaterally in paediatric BD patients [35], which is consistent with the present findings. Left parahippocampal gyrus has been reported to be decreased in size in panic disorder [36]. Parahippocampal gyri, with abundant interrelations with hippocampus and prefrontal cortex, may play an important role in the route of reciprocating information between these regions. A single-photon emission computed tomography study found an inverse correlation between resting regional blood flow in parahippocampal gyrus and severity of depressive mood [37], indicating that impaired function in these area might be related to mood disorder. Increased GM in parahippocampal gyrus could be a compensatory mechanism for the impaired function of structures that surround the area, for example the hippocampus. Its significance in BD, therefore, warrants further exploration.

Volumetric morphology in relation to clinical features of BD

As far as we know, this study is the first one that systematically investigated regional GM abnormality patterns in BD patients in relation to a variety of clinical features, using VBM. As suggested by previous studies [38, 39], the use of ‘corrected p’ under SPM for multiple comparisons is a very stringent criterion when applied to structural analyses. In the subgroup analyses, thereafter, we used a more liberal statistical threshold that has been used previously in BD study [16] to increase the sensitivity of detecting GM abnormality in this sample.

Family history

The present study found that patients with positive family history had an increase in right precentral gyrus compared to healthy subjects. This was not present in patients without family history compared to healthy subjects. There are other areas that had decreases in GM in family history-negative patients when compared with controls. Family history-positive patients had more GM bilaterally in the middle and right superior frontal gyri.

Previous studies about volumetric changes in familial BD have reported inconsistent results in subgenual prefrontal cortices [40, 41]. Family history was negatively correlated with right prefrontal cortex in female subjects but positively in male subjects in that study [40]. Increased sulcal and ventricular CSF volumes were also found to be characteristic of male familial BD patients [42]. However, the study did not provide effects of volumetric abnormalities independent of sex or effects of BD patients without familial history [42], which contributed to the ambiguity when explaining their findings.

No previous VBM study has compared regional GM changes in relation to family history in BD. The present study showed that patients with or without family history of BD may have different changes of brain structure. Particularly, the prefrontal region seems to be better preserved in patients with family history than those without. However, whether the changes are related to different genetic factors in these two groups or whether they reflect different illness effects is not known. There was no association between genetic liability and brain structure abnormalities found in a recent genetics study [43]. Another possibility is that effects of family history may lead to an illness with particular vulnerability of certain brain regions, which might be difficult to detect in the early stages.

Psychotic symptoms

It has been suggested that BD and schizophrenia shared some characteristics in clinical symptoms and pathogenesis[44]. BD patients with psychotic symptoms might be a particular subgroup that shows this overlap clinically and neuroanatomically. Early research using MRI showed similar volumetric changes such as decreased cortical GM and increased cortical sulci in schizophrenia as well as in BD, which differed quantitatively [45]. BD patients with psychosis and schizophrenia patients shared some brain morphological changes, for example increase in ventricular size [24], which were not found in non-psychotic BD patients. This result was not repeated in the present study.

We found that psychotic BD patients had more GM in the right precentral gyrus than non-psychotic patients or controls. Moreover, psychotic BD patients had more GM in the right precuneus, middle frontal gyrus, caudate body, left superior frontal gyrus, and left anterior cingulate than non-psychotic BD patients. Although there have been many volumetric studies about BD and schizophrenia, direct comparisons between psychotic and non-psychotic BD are rare. It is difficult to determine if psychotic features imply a brain abnormality that lies in between schizophrenia and BD. However, psychotic and non-psychotic BD present distinct patterns of GM abnormality, which need to be considered when looking into volumetric changes in BD.

Lithium

Neurotrophic effects of mood stabilizers such as lithium and valproate have been reported previously with evidence in brain volumes [46, 47] and molecular changes [48]. A quantitative MRI study detected an average 3% increase in total GM volume in eight BDI patients (history of manic and depressive edisodes) after 4 weeks of lithium treatment [47], but these changes were not replicated in another study [20]. In one VBM study, both medicated BD and control subjects had more GM in the posterior cingulate cortex than unmedicated BD subjects [28], supporting the neurotrophic effects of medication. However, medicated BD, if compared to healthy controls in the same study, had less GM in the lateral orbital cortex, which is not likely be explained by neurotrophic effects. Although this GM loss might come from other confounding factors, for example severity of illness, it could possibly be due to various effects from different medication. We particularly compared BD patients taking lithium with those not, and found regional GM loss in left middle occipital gyrus in patients taking lithium compared with control, and decreased GM in right orbital gyrus compared with patients not taking lithium. The results might suggest regional effects exerted by lithium unrelated to the neurotrophic effects. It should be noted that we did not have sufficient medication-naive subjects in our sample to investigate this effect conclusively. Further studies are needed to elucidate the effect when controlling for other factors in a relatively homogenous group of subjects.

Duration of illness and episodes of mania and depression

In the present subgroup analysis, duration of illness was negatively correlated with GM volume in the left caudate head. Numbers of episodes of mania and depression were negatively correlated with GM in the right inferior parietal lobule. Both findings suggested effects of illness and symptoms on the brain volumetric measures, and imply that discrepancy can potentially arise if illness severity is not taken into account. Previous VBM studies in BD addressed the issue, although one gave a negative result for illness duration [14] and the other reported reduced GM in relation to prior manic episodes in left middle frontal gyrus and right inferior frontal gyrus [15]. Adler et al. found that first-episode BD patients had increased GM in the ventral prefrontal cortex than patients with multiple episodes, suggesting that regional GM loss might be associated with affective episodes [16].

Previous ROI studies have produced discrepant results about brain abnormalities in relation to the course of illness. One study showed increased left temporal lobe in BD with duration of illness [49]. However, cortical thinning of the left middle frontal cortex was found to be related to the increase of duration of illness in BD [50]. The present study found that illness course was related to GM loss in the caudate, which was consistent with some previous results, that showed a larger caudate in male adult BD patients [51] but decreased in older BD patients [52]. Likewise, one VBM study in BD adolescents also found increased GM in caudate head, thereby supporting the hypothesis of caudate decline in the course of the illness [10]. Although the caudate was reportedly enlarged in several studies [26, 51], basal ganglia have been shown to be correlated with BD inconclusively in terms of volumetric changes of putamen, caudate or globus pallidus [4, 29]. Additionally, functional activities in these regions are observed to be associated with affective states of BD patients, indicating the role of basal ganglia in affective disorder.

Limitations

There are several limitations to the current study. First, the relative small sample size prevented us from further investigating the interactions of contributing factors such as duration of illness and type of medication. Second, the statistical power of the results is limited, given the number of factors being studied within a moderate sample size. We therefore chose a conservative statistical threshold (p_corrected<0.05). Although in subgroup analyses we used an uncorrected statistical threshold, we reported only clusters with >200 voxels. The findings in subgroup analyses should nevertheless be seen as exploratory. Third, as mentioned here, we need to be cautious when explaining observations of GM differences in bipolar subjects in a cross-sectional study, and the causative deductions are tentative. Although some affected regions are compatible with results from functional studies in fMRI studies, the cause–effect relationship remains uncertain. We also acknowledge that other factors, for example genetic polymorphism and environmental heterogeneity, may be involved in this association. A longitudinal study on a large sample would be crucial in elucidating the neuropathophysiological mechanisms of BD, as well as the influence of mood states, disease progression, and medication on the pattern of regional GM volumes.

Conclusion

The current study using the optimized VBM method presents evidence of structural abnormalities in GM in adult BD, with decreased GM in left middle temporal gyrus and increased GM in left parahippocampal gyrus. Regional variation in relation to different clinical factors was also seen, suggesting a neurobiological basis for symptom profiles in BD patients. From the present study it is not possible to state whether these abnormalities are present early in the course of the disease, or are secondary to disease duration and the effects of medication.

Footnotes

Acknowledgements

The authors wish to thank all the staff of the Neuropsychiatric Institute and the Black Dog Institute who participated in this study, and the patients and volunteers for their generous contributions.

References

1. McDonald

Zanelli

Rabe-Hesketh

. Meta-analysis of magnetic resonance imaging brain morphometry studies in bipolar disorder. Biol Psychiatry 2004; 56: 411–417.

2. Altshuler

Conrad

Hauser

. Reduction of temporal lobe volume in bipolar disorder: a preliminary report of magnetic resonance imaging. Arch Gen Psychiatry 1991; 48: 482–483.

3. Harvey

Persaud

Ron

Baker

Murray

. Volumetric MRI measurements in bipolars compared with schizophrenics and healthy controls. Psychol Med 1994; 24: 689–699.

4. Swayze

II Andreasen

Alliger

Yuh

Ehrhardt

. Subcortical and temporal structures in affective disorder and schizophrenia: a magnetic resonance imaging study. Biol Psychiatry 1992; 31: 221–240.

5. Roy

Zipursky

Saint-Cyr

Bury

Langevin

Seeman

. Temporal horn enlargement is present in schizophrenia and bipolar disorder. Biol Psychiatry 1998; 44: 418–422.

6. Altshuler

Bartzokis

Grieder

. An MRI study of temporal lobe structures in men with bipolar disorder or schizophrenia. Biol Psychiatry 2000; 48: 147–162.

7. DelBello

Zimmerman

Mills

Getz

Strakowski

. Magnetic resonance imaging analysis of amygdala and other subcortical brain regions in adolescents with bipolar disorder. Bipolar Disord 2004; 6: 43–52.

8. Ashburner

Friston

. Voxel-based morphometry: the methods. Neuroimage 2000; 11: 805–821.

9. Good

Johnsrude

Ashburner

Henson

Friston

Frackowiak

. A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 2001; 14: 21–36.

10.

10. Wilke

Kowatch

DelBello

Mills

Holland

. Voxel-based morphometry in adolescents with bipolar disorder: first results. Psychiatry Res 2004; 131: 57–69.

11.

11. Lochhead

Parsey

Oquendo

Mann

. Regional brain gray matter volume differences in patients with bipolar disorder as assessed by optimized voxel-based morphometry. Biol Psychiatry 2004; 55: 1154–1162.

12.

12. Dickstein

Milham

Nugent

. Frontotemporal alterations in pediatric bipolar disorder: results of a voxel-based morphometry study. Arch Gen Psychiatry 2005; 62: 734–741.

13.

13. Nasrallah

McCalley-Whitters

Jacoby

. Cortical atrophy in schizophrenia and mania: a comparative CT study. J Clin Psychiatry 1982; 43: 439–441.

14.

14. Bruno

Barker

Cercignani

Symms

Ron

. A study of bipolar disorder using magnetization transfer imaging and voxel-based morphometry. Brain 2004; 127: 2433–2440.

15.

15. Lyoo

Kim

Stoll

. Frontal lobe gray matter density decreases in bipolar I disorder. Biol Psychiatry 2004; 55: 648–651.

16.

16. Adler

Levine

DelBello

Strakowski

. Changes in gray matter volume in patients with bipolar disorder. Biol Psychiatry 2005; 58: 151–157.

17.

17. First

, Spitzer

, Gibbon

, Williams

JBW

. Structured Clinical Interview for DSM-IV Axis I Disorders, Patient Edition (SCID-P), Version 2.0. New York, NY: New York State Psychiatric Institute, Biometrics Research, 1995.

18.

18. Ananth

Popescu

Critchley

Good

Frackowiak

RSJ

Dolan

. Cortical and subcortical gray matter abnormalities in schizophrenia determined through structural magnetic resonance imaging with optimized volumetric voxel-based morphometry. Am J Psychiatry 2002; 159: 1497–1505.

19.

19. Lancaster

Woldorff

Parsons

. Automated Talairach atlas labels for functional brain mapping. Hum Brain Map 2000; 10: 120–131.

20.

20. Brambilla

Harenski

Nicoletti

. Differential effects of age on brain gray matter in bipolar patients and healthy individuals. Neuropsychobiology 2001; 43: 242–247.

21.

21. Dupont

Jernigan

Heindel

. Magnetic resonance imaging and mood disorders. Localization of white matter and other subcortical abnormalities. Arch Gen Psychiatry 1995; 52: 747–755.

22.

22. Zipursky

Seeman

Bury

Langevin

Wortzman

Katz

. Deficits in gray matter volume are present in schizophrenia but not bipolar disorder. Schizophr Res 1997; 26: 85–92.

23.

23. Norris

Krishnan

Ahearn

. Structural changes in the brain of patients with bipolar affective disorder by MRI: a review of the literature. Progr Neuropsychopharmacol Biol Psychiatry 1997; 21: 1323–1337.

24.

24. Strasser

Lilyestrom

Ashby

. Hippocampal and ventricular volumes in psychotic and nonpsychotic bipolar patients compared with schizophrenia patients and community control subjects: a pilot study. Biol Psychiatry 2005; 57: 633–639.

25.

25. Brambilla

Harenski

Nicoletti

. MRI study of posterior fossa structures and brain ventricles in bipolar patients. J Psychiatr Res 2001; 35: 313–322.

26.

26. Strakowski

DelBello

Zimmerman

. Ventricular and periventricular structural volumes in first- versus multiple-episode bipolar disorder. Am J Psychiatry 2002; 159: 1841–1847.

27.

27. Pearlson

Garbacz

Breakey

Ahn

DePaulo

. Lateral ventricular enlargement associated with persistent unemployment and negative symptoms in both schizophrenia and bipolar disorder. Psychiatry Res 1984; 12: 1–9.

28.

28. Nugent

Milham

Bain

. Cortical abnormalities in bipolar disorder investigated with MRI and voxel-based morphometry. Neuroimage 2006; 30: 485–497.

29.

29. Strakowski

DelBello

Sax

. Brain magnetic resonance imaging of structural abnormalities in bipolar disorder. Arch Gen Psychiatry 1999; 56: 254–260.

30.

30. Brambilla

Harenski

Nicoletti

. MRI investigation of temporal lobe structures in bipolar patients. J Psychiatr Res 2003; 37: 287–295.

31.

31. Pearlson

Barta

Powers

. Ziskind-Somerfeld Research Award 1996. Medial and superior temporal gyral volumes and cerebral asymmetry in schizophrenia versus bipolar disorder. Biol Psychiatry 1997; 41: 1–14.

32.

32. Blumberg

Kaufman

Martin

. Amygdala and hippocampal volumes in adolescents and adults with bipolar disorder. Arch Gen Psychiatry 2003; 60: 1201–1208.

33.

33. Aalto

Naatanen

Wallius

. Neuroanatomical substrata of amusement and sadness: a PET activation study using film stimuli. Neuroreport 2002; 13: 67–73.

34.

34. Malhi

Lagopoulos

Ward

. Cognitive generation of affect in bipolar depression: an fMRI study. Eur J Neurosci 2004; 19: 741–754.

35.

35. Frazier

Breeze

Makris

. Cortical gray matter differences identified by structural magnetic resonance imaging in pediatric bipolar disorder. Bipolar Disord 2005; 7: 555–569.

36.

36. Massana

Serra-Grabulosa

Salgado-Pineda

. Parahippocampal gray matter density in panic disorder: a voxel-based morphometric study. Am J Psychiatry 2003; 160: 566–568.

37.

37. Perico

CA-M

Skaf

Yamada

. Relationship between regional cerebral blood flow and separate symptom clusters of major depression: a single photon emission computed tomography study using statistical parametric mapping. Neurosci Lett 2005; 384: 265–270.

38.

38. Sowell

Thompson

Holmes

Batth

Jernigan

Toga

. Localizing age-related changes in brain structure between childhood and adolescence using statistical parametric mapping. Neuroimage 1999; 9: 587–597.

39.

39. Wilke

Kaufmann

Grabner

Putz

Wetter

Auer

. Gray matter-changes and correlates of disease severity in schizophrenia: a statistical parametric mapping study. Neuroimage 2001; 13: 814–824.

40.

40. Sharma

Menon

Carr

Densmore

Mazmanian

Williamson

. An MRI study of subgenual prefrontal cortex in patients with familial and non-familial bipolar I disorder. J Affect Disord 2003; 77: 167–171.

41.

41. Brambilla

Nicoletti

Harenski

. Anatomical MRI study of subgenual prefrontal cortex in bipolar and unipolar subjects. Neuropsychopharmacology 2002; 27: 792–799.

42.

42. Davis

Kwon

Cardenas

Deicken

. Decreased cortical gray and cerebral white matter in male patients with familial bipolar I disorder. J Affect Disord 2004; 82: 475–485.

43.

43. McIntosh

Job

Moorhead

. Genetic liability to schizophrenia or bipolar disorder and its relationship to brain structure. Am J Med Genet B Neuropsychiatr Genet 2006; 141: 76–83.

44.

44. Coryell

Leon

Turvey

Akiskal

Mueller

Endicott

. The significance of psychotic features in manic episodes: a report from the NIMH collaborative study. J Affect Disord 2001; 67: 79–88.

45.

45. Lim

Rosenbloom

Faustman

Sullivan

Pfefferbaum

. Cortical gray matter deficit in patients with bipolar disorder. Schizophr Res 1999; 40: 219–227.

46.

46. Sassi

Nicoletti

Brambilla

. Increased gray matter volume in lithium-treated bipolar disorder patients. Neurosci Lett 2002; 329: 243–245.

47.

47. Moore

Bebchuk

Wilds

Chen

Manji

. Lithium-induced increase in human brain grey matter.[erratum appears in Lancet 2000 356(9247):2104 Note: Menji HK [corrected to Manji HK]. Lancet 2000; 356: 1241–1242.

48.

48. Manji

Moore

Chen

. Clinical and preclinical evidence for the neurotrophic effects of mood stabilizers: implications for the pathophysiology and treatment of manic-depressive illness. Biol Psychiatry 2000; 48: 740–754.

49.

49. Ali

Denicoff

Altshuler

. Relationship between prior course of illness and neuroanatomic structures in bipolar disorder: a preliminary study. Neuropsychiatry Neuropsychol Behav Neurol 2001; 14: 227–232.

50.

50. Lyoo

Sung

Dager

. Regional cerebral cortical thinning in bipolar disorder. Bipolar Disord 2006; 8: 65–74.

51.

51. Aylward

Roberts-Twillie

Barta

. Basal ganglia volumes and white matter hyperintensities in patients with bipolar disorder. Am J Psychiatry 1994; 151: 687–693.

52.

52. Beyer

Kuchibhatla

Payne

. Caudate volume measurement in older adults with bipolar disorder. Int J Geriatr Psychiatry 2004; 19: 109–114.

Regional Gray Matter Changes in Bipolar Disorder: A Voxel-Based Morphometric Study

Abstract

Keywords

Method

Subjects

MRI acquisition

Voxel-based morphometry

Statistical analysis

Results

Discussion

Global volumetric measures

Regional abnormalities in BD patients

Volumetric morphology in relation to clinical features of BD

Family history

Psychotic symptoms

Lithium

Duration of illness and episodes of mania and depression

Limitations

Conclusion

Footnotes

Acknowledgements

References