- Email: [email protected]
Cross-sectional study of abnormal amygdala development in adolescents and young adults with bipolar disorder
Cross-Sectional Study of Abnormal Amygdala Development in Adolescents and Young Adults with Bipolar Disorder Barbara K. Chen, Roberto Sassi, David Axelson, John P. Hatch, Marsal Sanches, Mark Nicoletti, Paolo Brambilla, Matcheri S. Keshavan, Neal D. Ryan, Boris Birmaher, and Jair C. Soares Background: In vivo imaging studies in adult bipolar patients have suggested enlargement of the amygdala. It is not known whether this abnormality is already present early in the illness course or whether it develops later in life. We conducted a morphometric MRI study to examine the size of specific temporal lobe structures in adolescents and young adults with bipolar disorder and healthy control subjects, as well as their relationship with age, to examine possible neurodevelopmental abnormalities. Methods: Subjects included 16 DSM-IV bipolar patients (16 ⫾ 3 years) and 21 healthy controls (mean age ⫾ SD ⫽ 17 ⫾ 4 years). Measures of amygdala, hippocampus, temporal gray matter, temporal lobe, and intracranial volumes (ICV) were obtained. Results: There was a trend to smaller left amygdala volumes in patients (mean volumes ⫾ SD ⫽ 1.58 ⫾ .42 mL) versus control subjects (1.83 ⫾ .4 mL; F ⫽ 3.87, df ⫽ 1,32, p ⫽ .06). Bipolar patients did not show significant differences in right or left hippocampus, temporal lobe gray matter, temporal lobe, or right amygdala volumes (analysis of covariance, age, gender, and ICV as covariates, p ⬎ .05) compared with healthy control subjects. Furthermore, there was a direct correlation between left amygdala volumes and age (r ⫽ .50, p ⫽ .047) in patients, whereas in healthy controls there was an inverse correlation (r ⫽ ⫺.48, p ⫽ .03). Conclusions: The direct correlation between left amygdala volumes and age in bipolar patients, not present in healthy control subjects, may reflect abnormal developmental mechanisms in bipolar disorder. Key Words: Adolescence, affective disorders, development, mood disorders, MRI, neuroimaging
B
ipolar disorder, or manic-depressive illness, affects more than 2 million American adults (Spearing et al 2000). The exact prevalence among adolescents and young adults is more difficult to identify, because the symptoms may be similar to or occur concurrently with other childhood psychiatric disorders (National Institute of Mental Health 2000). As a major psychiatric disorder, bipolar disorder is implicated in functional impairment and represents an important risk factor for suicide (Oquendo and Mann 2001). Research into the pathogenesis of this disorder may help prevent misdiagnosis and improve currently available therapeutic strategies. In the past, bipolar disorder was classified as being an “endogenous” disorder for which there was no identified anatomic substrate (Alvarez 2001). During the last two decades, however, the development of modern neuroimaging techniques has enabled researchers to detect subtle structural and functional changes in the brains of patients suffering from mood disorders, which challenged the traditional concept that these disorders were “endogenous” or functional illnesses. Brain neuroanatomic circuits interconnecting the prefrontal cortex, medial temporal From the Department of Psychiatry (BKC, JPH, MS, MN, PB, JCS), and Radiology (JCS), The University of Texas Health Science Center at San Antonio, San Antonio, Texas; Department of Psychiatry (RS, DA, MSK, NDR, BB), University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Department of Psychiatry (RS), Institute of Psychiatry, University of Sao Paulo School of Medicine, and Department of Psychiatry (MS), Federal University of Sao Paulo, Sao Paulo, Brazil; IRCCS S. Giovanni di Dio (PB), Fatebenefratelli, Brescia, Italy; and South Texas Veterans Health Care System (JCS), Audie L. Murphy Division, San Antonio, Texas. Address reprint requests to Jair C. Soares, M.D., Division of Mood and Anxiety Disorders, Department of Psychiatry (MC 7792), University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900; E-mail: [email protected]. Received September 16, 2004; revised June 1, 2004; accepted June 15, 2004.
0006-3223/04/$30.00 doi:10.1016/j.biopsych.2004.06.024
lobe, and basal ganglia are involved in mood regulation. Morphometric studies suggest anatomic abnormalities in key structures in these neuroanatomic circuits in the pathophysiology of bipolar disorder (Beyer and Krishnan 2002; Soares 2003; Soares and Mann 1997). The amygdala and hippocampus play key roles in mood and behavior. Both are components of the limbic system, which influences emotional excitability, subjective emotional experiences, and sexual libido. Specifically, because the hippocampus functions to identify and bridge relationships in memory, a disruption in this structure would impair cognitive functions and behavior (Nagode and Pardo 2002). Direct emotional responses to information, especially to fear and anxiety, occur in the amygdala (Adolphs et al 1994; Gallagher and Chiba 1996; Morris 1996). Together, amygdala and hippocampus translate cognitive input in emotional processing (LeDoux 1996), and as a result, abnormalities in these brain structures may be implicated in the pathophysiology of bipolar disorder. Most studies showed enlargement of the amygdala in adult bipolar patients compared with healthy control subjects, without clear indications of size abnormalities in the hippocampus (Altshuler et al 2000; Brambilla et al 2003; Strakowski et al 1999). Because the majority of existing studies focused on adults, it is not known whether such enlargement would be present in younger patients suffering from bipolar disorder. Blumberg et al (2003) recently reported smaller amygdala size in both pediatric and adult bipolar disorder patients. If present in the young age group, amygdala abnormalities would likely reflect neurodevelopmental insults or changes that develop early in the course of the illness. Through the processes of pruning and myelination, the brain removes unnecessary gray matter while the white matter grows during late adolescence (Giedd et al 1999; Gogate et al 2001; Gogtay et al 2002). In this cross-sectional study, amygdala, hippocampus, and temporal lobe volumes of bipolar patients and healthy control subjects were determined with high-resolution magnetic resonance imaging (MRI). We compared these volumes to determine whether any size differences existed between these two groups, BIOL PSYCHIATRY 2004;56:399 – 405 © 2004 Society of Biological Psychiatry
400 BIOL PSYCHIATRY 2004;56:399 – 405
B.K. Chen et al
Table 1. Clinical and Demographic Information Bipolar Subject
Age (years)
Race
Gender
Handedness
Age at Onset (years)
Lithium
1 2 3
10 10 12
White White White
M M M
Left Right Left
7 4 6
⫹ ⫹
4 5 6 7 8 9
13 14 14 14 15 15
White White White White White White
F M F M F F
Right Right Right Right Right Right
10 12 11 10 14 14
10 11 12 13 14 15 16
17 17 18 18 19 21 21
White White White White White White Black
F M F F F M M
Left Right Right Right Right Right Right
13 12 15 8 17 19 15
⫹ ⫹
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
Valproate
Other Medications Olanzapine Thyroid hormone, methylphenidate
⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹
Clomipramine, dexedrine Fluoxetine, lorazepam, thyroid hormone Trazodone Thyroid hormone Citalopram
F, female; M, male.
with the hypothesis that bipolar patients would have enlargement of the amygdala. Furthermore, age effects on temporal lobe structures were compared in the two groups to examine possible abnormal neurodevelopment.
Methods and Materials Subjects Thirty-seven young bipolar patients and control subjects aged 10 through 21 (mean age ⫾ SD ⫽ 16 ⫾ 4 years) volunteered as subjects for this study. Table 1 shows the demographic information of the 16 DSM-IV bipolar disorder patients (mean age ⫾ SD ⫽ 16 ⫾ 3 years; 8 female subjects, 8 male subjects; 15 Caucasians, 1 African American; 12 bipolar I type, 3 bipolar II type, 1 bipolar not otherwise specified). Among the bipolar patients, 14 were euthymic and 2 were mildly depressed at the time of the study. All but one reported a familial history of mood disorders, defined as having a first-degree relative who ever received a diagnosis of unipolar or bipolar disorder. Fourteen patients were receiving treatment for bipolar disorder (six lithium only, four valproate only, and four lithium plus valproate). Eleven patients reported previous use of antipsychotic drugs. The mean duration of the disease among patients was 3.93 years (SD ⫽ 2.37). Comorbid mental disorders were found in seven patients and consisted of attention-deficit/hyperactivity disorder (five subjects), oppositional– defiant disorder (one subject), and conduct disorder (one subject). Twenty-one healthy control subjects without any history of psychiatric problems in first-degree relatives (mean age ⫾ SD ⫽ 17 ⫾ 4 years; 9 female subjects, 12 male subjects; 18 Caucasians, 3 African Americans) were studied. The psychiatric diagnoses were confirmed or ruled out through the Schedule for Affective Disorders and Schizophrenia for School-Age Children—Present and Lifetime Version (K-SADS; Kaufman et al 1997) for children up to 17 years old, or the Structured Clinical Interview for DSM-IV (SCID-IV; Spitzer et al 1998) for subjects who were 18 or older. The University of Pittsburgh Biomedical Institutional Review Board approved this study, and all subjects or their parents or legal representatives gave written informed consent after understanding all issues involved in study participation. www.elsevier.com/locate/biopsych
MRI Procedures The MRI images were obtained at a 1.5-T GE Signa Imaging System running the Signa 5.4.3 (General Electric Medical Systems, Milwaukee, Wisconsin) software at the University of Pittsburgh. All patients were placed in the MRI in a standardized position determined by sagittal scout series to verify head position and to decrease interscan variability. A plastic head holder was used to maintain placement of the subject’s head and to limit any head movement during the MRI session. Using three-dimensional gradient echo imaging by spoiled gradient recalled acquisition, 124 images covering the whole brain were acquired within the coronal plane with 1.5-mm-thick slices (repetition time ⫽ 25 msec, echo time ⫽ 5 msec, nutation angle ⫽ 40°, field of view ⫽ 24 cm, number of excitations ⫽ 1, matrix size ⫽ 256 ⫻ 192). T2 and protondensity images were obtained in the axial plane to screen for neuroradiologic abnormalities. Image Analysis All images were analyzed using Scion Image for Windows Beta-3b version on a PC workstation to obtain the volumes (in milliliters) of the amygdala, hippocampus, temporal lobe, and intracranial brain volume (ICV). Gray matter volumes were determined using a semiautomated segmentation algorithm (Keshavan et al 1994, 1995). Well-trained raters blinded to group assignments and subject identity performed all measurements, and interrater reliability was measured by intraclass correlation coefficients, determined by two independent raters tracing 10 scans. These values were as follows: right hippocampus (r ⫽ .92), left hippocampus (r ⫽ .93), right amygdala (r ⫽ .95) left amygdala (r ⫽ .91), total right temporal lobe (r ⫽ .96), total left temporal lobe (r ⫽ .99), right temporal lobe gray matter (r ⫽ .94), and left temporal lobe gray matter (r ⫽ .96). Hippocampus The right and left hippocampi were traced separately. The tracing of the hippocampus began where the superior colliculus completely joined the thalamus, as shown in Figure 1A. The hippocampus was visible inferiorly and laterally to the thalamus and directly laterally to the ambient cistern. Moving anteriorly,
BIOL PSYCHIATRY 2004;56:399 – 405 401
B.K. Chen et al
Table 3. Temporal Lobe Structures of Lithium-Free and Lithium-Treated Bipolar Patients Temporal Lobe Structure
No Lithium (n ⫽ 6) Mean ⫾ SD (mL)
Lithium (n ⫽ 10) Mean ⫾ SD (mL)
df
p
Right Hippocampus Left Hippocampus Right Amygdala Left Amygdala
3.07 ⫾ .51 3.07 ⫾ .55 2.00 ⫾ .59 1.59 ⫾ .58
3.07 ⫾ .29 3.16 ⫾ .31 2.21 ⫾ .22 1.58 ⫾ .33
14 14 14 14
.98 .68 .33 .96
Statistical testing was performed using the t test.
temporal lobe to the cerebrum, disappeared or where the amygdala gray matter became too diffuse to resolve from the surrounding temporal lobe gray matter (Figure 1D). As shown in Figure 1C, the more anteriorly located amygdala appeared in the same position as the hippocampus in more posterior slices. Its superior border extended into the white matter tract separating it from the globus pallidus. The inferior border was the parahippocampal gyrus. The areas of all slices were summed, and this sum was then multiplied by the slice thickness (.15 cm) to obtain volume (milliliters). Figure 1. Magnetic resonance imaging tracings of hippocampus and amygdala. The hippocampus and amygdala were traced using the following anatomical landmarks. (A) The most posterior or first slice of the hippocampus began where the superior colliculus joined the thalamus. (B) The most anterior or last slice of the hippocampus ended when the third ventricle splits from the cistern by the hypothalamus. (C) The most posterior slice of amygdala began where the mamillary bodies were separated. It appeared in the same position as the hippocampus in more posterior slices, but the superior border became the globus pallidus. (D) The amygdala tracing ended where the temporal bridge disappeared or the gray matter became too diffuse.
the ambient cistern appeared superior to the hippocampus. The tracing continued to the point where the third ventricle was split from the cistern by the hypothalamus as shown in Figure 1B. The amygdala was not included because it appeared above the hippocampus. Amygdala The right and left amygdalae were traced separately. Tracing began at the slice where the third ventricle and suprasellar cistern were clearly joined as the mamillary body (Figure 1C) and ended where the temporal bridge, the white matter that connects the
Temporal Lobe Temporal lobes were traced from the most posterior slice where the temporal lobe appeared, where the splenium of the corpus callosum was initially most pronounced. The superior border was the transverse fissure and the maximal limits of the gray and white matter of the temporal lobe were adopted as the inferior, medial, and lateral limits. The temporal stem, the white matter tract that links temporal lobe to the rest of the hemisphere, was excluded from the tracing. Intracranial Volume The ICV was obtained by tracing the border of the brain, which included total cerebral gray and white matter, cerebrospinal fluid, dura matter, sinuses, optic chiasm, pituitary, cerebellum, and brainstem. Every other slice was traced and multiplied by .3 cm, the thickness of two slices, because for measures of ICV we traced only every other slice. Statistical Analyses Statistical analysis was completed using Systat software, version 8.1 (SPSS, Chicago, Illinois) on a PC Windows workstation. Asymmetry in medial temporal lobe structures was evaluated
Table 2. Volumes of Temporal Lobe Structures Temporal Lobe Structure
Bipolar (n ⫽ 16) Mean ⫾ SD (mL)
Control (n ⫽ 21) Mean ⫾ SD (mL)
Right Hippocampus Left Hippocampus Right Amygdala Left Amygdala Right Temporal Lobe Gray Matter Left Temporal Lobe Gray Matter Right Temporal Lobe Left Temporal Lobe Intracranial Volume
3.07 ⫾ .37 3.12 ⫾ .40 2.13 ⫾ .39 1.58 ⫾ .42 62.23 ⫾ 7.45
3.06 ⫾ .27 3.15 ⫾ .33 2.14 ⫾ .57 1.83 ⫾ .42 58.75 ⫾ 6.81
61.12 ⫾ 8.71 84.94 ⫾ 11.18 77.26 ⫾ 12.77 1,518.56 ⫾ 141.14
F
df
p
.05 .41 .17 3.87 1.34
1/32 1/32 1/32 1/32 1/32
.82 .53 .69 .06 .26
60.36 ⫾ 7.57
.13
1/32
.72
79.43 ⫾ 10.32 76.31 ⫾ 9.62 1,479.81 ⫾ 162.51
3.13 .037 .58
1/32 1/32 1/35
.09 .85 .45
Statistical testing was performed using analysis of covariance with age, gender, and intracranial volume (ICV) as covariates; ICV was determined by analysis of variance. Unadjusted means and standard deviations are shown.
www.elsevier.com/locate/biopsych
402 BIOL PSYCHIATRY 2004;56:399 – 405
B.K. Chen et al
with repeated measures analyses of covariance (ANCOVA) with age, gender, and ICV as covariates. Left and right amygdala, hippocampus, temporal lobe gray matter, and temporal lobe total volumes of bipolar and control subjects were compared using ANCOVA with age, gender, and ICV as covariates. Analysis of variance (ANOVA) was performed to compare the ICV between bipolar patients and control subjects. Unadjusted volume means ⫾ SD are reported for all brain structures. In checking the ANCOVA assumption of equal regression slopes, it was determined that the interaction of the age covariate with diagnostic groups was statistically significant (p ⫽ .012) for the analysis of left amygdala. Therefore, this interaction term was retained in the analysis of the left amygdala. Pearson’s correlations were calculated to assess the relation of age to amygdala and hippocampus volumes. Z tests were performed to compare Pearson’s correlation and partial correlation coefficients adjusting for gender and ICV. A two-tailed p value less than .05 was considered for rejecting all null hypotheses.
Results The bipolar patients did not differ significantly from the healthy control subjects in age (t test, p ⫽ .256), gender (Fisher’s Exact Test, p ⫽ .746), race (Fisher’s Exact Test, p ⫽ .392), handedness (Fisher’s Exact Test, p ⫽ .296), height (t test, p ⫽ .330), weight (t test, p ⫽ .840), ICV (t test, p ⫽ .453), years of education (Mann-Whitney U test, p ⫽ .387), or educational achievement (Pearson chi-square, p ⫽ .218). No statistically significant differences were found between the bipolar and healthy groups for any of the structures measured (Table 2). There was a trend for the left amygdala to be smaller in the bipolar patients (F ⫽3.87, df⫽ 1,32, p ⫽ .06). No significant differences were found for any of the measured structures between lithium-treated bipolar patients and bipolar patients who were not on lithium (Table 3). Asymmetry was not found for either hippocampus (F ⫽ 1.27, df ⫽ 1,32, p ⫽ .269) or amygdala (F ⫽ .14, df ⫽ 1,32, p ⫽ .711). There also was no significant interaction between diagnostic group and side for either structure. Bipolar subjects with a comorbid diagnosis have significantly smaller left amygdala volumes (1.24 ⫾ .27 mL, n ⫽ 7) compared with bipolar subjects without a comorbid diagnosis (1.85 ⫾ .30 mL, n ⫽ 9; t ⫽ 4.21, p ⫽ .001). The left amygdala volume of bipolar patients without comorbidity was not statistically different from the values for the healthy control subjects (p ⬎ .05). The right amygdala and hippocampus volumes of bipolar subjects with comorbidity compared with those without comorbidity were not significantly different (p ⬎ .05). For healthy control subjects, there was a significant inverse relationship between age and left amygdala volumes (r ⫽ ⫺.48, df ⫽ 19, p ⫽ .03). In contrast, bipolar patients demonstrated a significant direct relationship between age and left amygdala volumes (r ⫽ .50, df ⫽ 14, p ⫽ .047), as seen in Figure 2. These correlations were significantly different (Z ⫽ 2.91, p ⫽ .004). Pearson’s correlation demonstrated no significant association between age and right hippocampus, left hippocampus, or right amygdala volumes in either group (bipolar: r ⫽ ⫺.15, p ⫽ .58; r ⫽ .03, p ⫽ .92; r ⫽ .39, p ⫽ .14; respectively; control: r ⫽ ⫺.02, p ⫽ .93; r ⫽ ⫺.26, p ⫽ .25; r ⫽ ⫺.31, p ⫽ .17). Partial correlations adjusting for gender and ICV of healthy control subjects (r ⫽ ⫺.36, df ⫽ 17, p ⫽ .134) and bipolar patients (r ⫽ .48, df ⫽ 12, p ⫽ .081) were not statistically significant. Nevertheless, the Z test comparing the partial correwww.elsevier.com/locate/biopsych
Figure 2. Age and left amygdala volumes in bipolar patients and healthy control subjects. Best fit curves showing left amygdala volumes versus age of bipolar and control subjects. (A) Bipolar left amygdala volumes show a direct relationship with age. (B) Normal subjects show an inverse relationship between left amygdala volume and age.
lations controlling for gender and ICV remained statistically significant (Z ⫽ 2.48, p ⫽ .007). Table 4 displays the Pearson’s correlations for male and female subjects separately. A significant direct relationship between age and left amygdala was observed in male bipolar patients (r ⫽ .73, df ⫽ 6, p ⫽ .04), whereas a significant inverse relationship was observed for male healthy control subjects (r ⫽ ⫺.70, df ⫽ 10, p ⫽ .01). Pearson’s correlations for female bipolar patients were nonsignificant (p ⬎ .05). Correlations between age and other temporal lobe structures were not significant in either gender. Figure 3 depicts the linear relationship of left amygdala volumes and age for male bipolar patients and healthy control subjects.
BIOL PSYCHIATRY 2004;56:399 – 405 403
B.K. Chen et al Table 4. Relationship Between Age and Volumes of Temporal Lobe Structures Male
Female
Bipolar (n ⫽ 8) Temporal Lobe Structure Right Hippocampus Left Hippocampus Right Amygdala Left Amygdala ICV
Control (n ⫽ 12)
Bipolar (n ⫽ 8)
Control (n ⫽ 9)
r
df
p
r
df
p
r
df
p
r
df
p
⫺.13 .07 .37 .73 .59
6 6 6 6 6
.77 .87 .37 .04 .12
.30 .15 ⫺.23 ⫺.70 ⫺.48
10 10 10 10 10
.35 .64 .47 .01 .11
⫺.07 .10 .48 .20 ⫺.30
6 6 6 6 6
.87 .81 .23 .64 .47
.11 ⫺.55 ⫺.26 .10 ⫺.23
7 7 7 7 7
.78 .12 .49 .79 .55
ICV, intracranial volume.
Discussion Our study found no significant differences in measures of temporal lobe, right amygdala, and hippocampus between young individuals with bipolar disorder and matched healthy control subjects. Nonetheless, there was a trend for smaller left amygdala volumes in patients, in contrast with several studies involving adult bipolar disorder individuals (Altshuler et al 1998, 2000; Brambilla et al 2003; Strakowski et al 1999) that found enlarged amygdala volumes. Conversely, Pearlson et al (1997) found smaller left amygdala volumes in adult bipolar patients compared with control subjects. Findings of smaller amygdala volumes in young patients with bipolar disorder suggest abnormalities in this brain structure that are already detectable early in the course of the illness. Such changes, in the context of adult studies that indicated enlarged amygdala, suggest that abnormal development of limbic structures may be involved in pathophysiology of bipolar disorder. The most pronounced abnormality was presented in bipolar patients with comorbidity, but this finding stemmed from a reduced sample size and is preliminary. It is unknown whether attention-deficit/hyperactivity disorder alone or together with bipolar disorder decreases amygdala volume. Studies with a larger sample size are necessary to further explore this question. Prior studies, consistent with our present findings, predominantly showed no evidence of anatomic abnormalities in the hippocampus of adult bipolar patients (Altshuler et al 1998, 2000; Brambilla et al 2003; Pearlson et al 1997; Strakowski et al 1999). Only one study (Swayze et al 1992) found a smaller hippocampal volume in adult bipolar individuals. Conversely, most studies of hippocampal volumes in unipolar patients found evidence of a size reduction (Beyer and Krishnan 2002; Bremner et al 2000; Mervaala et al 2000; Sheline et al 1999; Steffens et al 2000), whereas a few found no abnormalities (Altshuler et al 1998; Axelson et al 1993; Beyer et al 2002; Hauser et al 2000; Vakili et al 2000). Hippocampal volume reduction in unipolar patients is possibly caused by a hyperactivity within the hypothalamic– pituitary–adrenal axis and resulting hypercortisolemia. Our findings of normal temporal lobe volumes in young bipolar patients are consistent with most reports in adult bipolar patients (Altshuler et al 1998; Beyer et al 2002; Hauser et al 2000; Johnstone et al 1989). These findings differ from the studies by Hauser et al (1989) and Altshuler et al (1991), in which smaller temporal lobe volumes were found, and Harvey et al (1994), in which increased gray matter volumes in left temporal lobes were reported in adult bipolar individuals. Brain gray matter is progressively pruned during adolescence, especially between the ages of 14 and 17, which is the age span when the incidence of psychiatric disorders increases (Giedd 1999; Gogtay 2002). Giedd’s 1999 study involved healthy chil-
dren and adolescents and demonstrated pruning of gray matter in adolescence. Our findings of abnormal relationship between age and left amygdala volumes in bipolar patients suggest developmental abnormalities in this brain structure, which could reflect abnormal pruning, although the present cross-sectional study cannot definitively demonstrate this finding. Future longitudinal studies examining medial temporal lobe development in bipolar children, adolescents, and young adults will be necessary. A potential limitation of this study is that most patients were on medications at the time they were studied; however, our results suggest that medication effects are not potential confounders to our reported analyses. Drug-free bipolar patients did not differ from lithium-treated patients in measures of temporal lobe structures. Furthermore, our findings are in agreement with our prior report in adult bipolar patients, where amygdala and hippocampus volumes did not differ between lithium-treated and drug-free bipolar adults (Brambilla et al 2003). Although other studies did show that the treatment of bipolar disorder patients with lithium increased total brain gray matter (Moore et al 2000; Sassi et al 2002), any possible effects of lithium on amygdala and hippocampus size have not been characterized. Additionally, our study mainly included euthymic bipolar I patients. A study by Blumberg et al (2003) found no significant difference between euthymic, depressed, and manic bipolar patients, although their sample size for each state was small. Our study also indicates that left amygdala volumes among young male bipolar patients may increase as age increases, which is opposite of what is found in healthy male control subjects. In normal childhood development, the hippocampus and amygdala total volumes increase in both males and females, with the amygdala increasing significantly more in males and the hippocampus increasing significantly more in female children (Durston et al 2001; Giedd et al 1997); however, gray matter in the temporal lobe decreases in midadolescence and continues to decrease throughout adulthood in males and females (Bartzokis et al 2001; Giedd et al 1999). Our findings of increasing left amygdala size in males to the age of 21 years may indicate abnormal development of this brain structure in male bipolar patients, which could result in hypertrophy of this brain region in adult bipolar patients. Such findings, if confirmed in longitudinal studies, could be key steps underlying the pathophysiology of the emotional disregulation found in bipolar disorder. Patients with bilateral amygdala damage show placidness and hypoemotionality (Gallagher and Chiba 1996). Findings of enlargement of the left amygdala are consistent with findings of increased cerebral blood flow and glucose metabolism in this brain structure in adult bipolar patients (Drevets 2000). The abnormal relationship between age and left amygdala www.elsevier.com/locate/biopsych
404 BIOL PSYCHIATRY 2004;56:399 – 405
B.K. Chen et al In summary, no significant differences in volumes of hippocampus, amygdala, or temporal lobe were found between bipolar patients and matched healthy control subjects, even though there was a trend for smaller left amygdala volumes in patients. These findings suggest developmental abnormalities in amygdala in bipolar individuals. Even though left amygdala volumes may be smaller in younger bipolar patients because of possible abnormalities in pruning mechanisms during late adolescence, adult bipolar patients could end up with an enlargement of this brain structure, as suggested by the fact that the age effects on this structure in younger bipolar and healthy subjects are strikingly different. Nonetheless, because of the cross-sectional nature of our findings, future studies with larger patient samples and a longitudinal design are necessary.
This study was supported in part by National Institute of Mental Health Grants MH 01736, MH 55123, MH 30915 and MH 59929; the National Alliance for Research on Schizophrenia and Affective Disorders; the Office of the Dean, University of Texas Health Science Center at San Antonio Medical School (UTHSCSA), the Veterans Administration, and the Krus Endowed Chair in Psychiatry (UTHSCSA).
Figure 3. Age and left amygdala volumes in male subjects. Best fit curves showing left amygdala volumes versus age of male bipolar patients and control subjects. (A) Left amygdala of male bipolar patients showing a direct relationship to age. (B) Left amygdala of healthy male subjects showing an inverse relationship to age.
volumes was restricted to male bipolar type I patients and was not present in female patients or control subjects. Our findings should be seen as preliminary, because sample sizes are relatively small. Nonetheless, these results suggest, as highlighted in previous studies (Jurjus et al 1993; Strakowski et al 1993), that male and female patients with mood disorders may differ in regard to neuroimaging findings and vulnerability to developmental factors implicated in the pathophysiology of mood disorders. For instance, Brown et al (1995) reported higher relative risk for affective psychosis, associated with prenatal factors such as exposure to famine, among male subjects. Furthermore, in Altshuler et al (2000), the findings of enlarged amygdala were restricted to a group of adult bipolar males. www.elsevier.com/locate/biopsych
Adolphs R, Tranel D, Damasio H, Damasio A (1994): Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala. Nature 372:669 – 672. Altshuler LL, Bartzokis G, Grieder T, Curran J, Jimenez T, Leight K, et al (2000): An MRI study of temporal lobe structures in men with bipolar disorder or schizophrenia. Biol Psychiatry 48:147–162. Altshuler LL, Bartzokis G, Grieder T, Curran J, Mintz J (1998): Amygdala enlargement in bipolar disorder and hippocampal reduction in schizophrenia: An MRI study demonstrating neuroanatomic specificity. Arch Gen Psychiatry 55:663– 664. Altshuler LL, Conrad A, Hauser P, Li X, Guze GN, Denicoff K, et al (1991): Temporal lobe volume reduction in bipolar disorder: A preliminary MRI report. Arch Gen Psychiatry 48:482– 483. Alvarez J (2001): Neuronal hypersynchronization, creativity and endogenous psychoses. Med Hypotheses 56:672– 685. Axelson DA, Doraiswamy P, McDonald WM, Boyko OB, Tupler LA, Patterson LJ, et al (1993): Hypercortisolemia and hippocampal changes in depression. Psychiatry Res 47:163–173. Bartzokis G, Beckson M, Lu PH, Nuechterlein KH, Edwards N, Mintz J (2001): Age-related changes in frontal and temporal lobe volumes in men: A magnetic resonance imaging study. Arch Gen Psychiatry 58:461– 465. Beyer JL, Krishnan KRR (2002): Volumetric brain imaging findings in mood disorders. Bipolar Disord 4:89 –104. Blumberg HP, Kaufman J, Martin A, Whiteman, R, Zhang JH, Gore JC, et al (2003): Amygdala and hippocampal volumes in adolescents and adults with bipolar disorder. Arch Gen Psychiatry 60:1201–1208. Brambilla P, Harenski K, Nicoletti M, Sassi RB, Mallinger AG, Frank E, et al (2003): MRI investigation of temporal lobe structures in bipolar patients. J Psychiatr Res 37:287–295. Bremner JD, Narayan M, Anderson ER, Staib LH, Miller HL, Charney DS (2000): Hippocampal volume reduction in major depression. Am J Psychiatry 157:115–117. Brown AS, Susser ES, Lin SP, Neugebauer R, Gorman JM (1995): Increased risk of affective disorders in males after second trimester prenatal exposure to the Dutch hunger winter of 1944-45. Br J Psychiatry 166:601– 606. Drevets WC (2000): Neuroimaging studies of mood disorders. Biol Psychiatry 48:813– 829. Durston S, Hulshoff Pol HE, Casey BJ, Giedd JN, Buitelaar JK, van Engeland H (2001): Anatomical MRI of the developing human brain: What have we learned? J Am Acad Child Adolesc Psychiatry 40:1012–1020. Gallagher M, Chiba AA (1996): The amygdala and emotion. Neurobiology 6:221–227. Giedd JN, Blumenthal J, Jeffries NO, Castellanos FX, Liu H, Zijdenbos A, et al (1999): Brain development during childhood and adolescence: A longitudinal MRI study. Nature Neurosci 2:861– 863.
B.K. Chen et al Giedd JN, Castellanos FX, Rajapakse JC, Vaituzis AC, Rapoport JL (1997): Sexual dimorphism of the developing human brain. Prog Neuropsychoparmacol Biol Psychiatry 21:1185–1201. Gogtay N, Giedd J, Janson K, Rapoport JL (2001): Brain imaging in normal and abnormal brain development: New perspectives for child psychiatry. Clin Neurosci Res 1:283–290. Gogtay N, Giedd J, Rapoport JL (2002): Brain development in healthy, hyperactive, and psychotic children. Arch Neurol 59:1244 –1248. Harvey I, Persaud R, Ron MA, Baker G, Murray RM (1994): Volumetric MRI measurements in bipolars compared with schizophrenics and healthy controls. Psychol Med 24:689 – 699. Hauser P, Altshuler LL, Berrettini W, Dauphinais ID, Gelernter J, Post RM (1989): Temporal lobe measurement in primary affective disorder by magnetic resonance imaging. J Neuropsychiatry Clin Neurosci 1:128 – 134. Hauser P, Matochik J, Altshuler LL, Denicoff KD, Conrad A, Li X, et al (2000): MRI-based measurements of temporal lobe and ventricular structures in patients with bipolar I and bipolar II disorders. J Affect Disord 60:25–32. Jurjus GJ, Nasrallah HA, Olson SC, Schwarzkopf SB (1993): Cavum septum pellucidum in schizophrenia, affective disorder and healthy controls: A magnetic resonance imaging study. Psychol Med 23:319 –322. Kaufman J, Birmaher B, Brent D, Ryan ND, Rao U (1997): Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL): Initial reliability and validity data. J Am Acad Child Adolesc Psychiatry 36:980 –988. Keshavan MS, Anderson S, Beckwith C, Nash K, Pettegrew JW, Krishnan KR (1995): A comparison of stereology and segmentation techniques for volumetric measurements of lateral ventricles in magnetic resonance imaging. Psychiatry Res 61:53– 60. Keshavan MS, Beckwith C, Bagwell W, Pettegrew W, Krishnan KR (1994): An objective method for edge detection in MRI morphometry. Eur Psychiatry 9:205–207. Johnstone EC, Owens DGC, Crow TJ, Frith CD, Alexandropolis K, Bydder G, Colter N (1989): Temporal lobe structure as determined by nuclear magnetic resonance in schizophrenia and bipolar affective disorder. J Neurol Neurosurg Psychiatry 53:736 –741. LeDoux J (1996): The Emotional Brain. New York: Simon & Schuster. Mervaala E, Fohr J, Kononen M, Valkonen-Korhonen M, Vainio P, Partanen K, et al (2000): Quantitative MRI of the hippocampus and amygdala in severe depression. Psychol Med 30:117–125. Moore GJ, Bebchuk JM, Wilds IB, Chen G, Manji HK (2000): Lithium-induced increase in human brain gray matter. Lancet 356:1241–1242. Morris JS, Frith CD, Perrett DI, Rowland D, Young AW, Calder AJ, Dolan RJ (1996): A differential neural response in the human amygdala to fearful and happy facial expressions. Nature 383:812– 815.
BIOL PSYCHIATRY 2004;56:399 – 405 405 Nagode JC, Pardo JV (2002): Human hippocampal activation during transitive inference. Neuroreport 13:939 –944. National Institute of Mental Health (2000): Child and adolescent bipolar disorder: An update from the National Institute of Mental Health. Available at: http://www.nimh.nih.gov/publicat/bipolarupdate.cfm. Accessed January 8, 2003. Oquendo MA, Mann JJ (2001): Identifying and managing suicide risk in bipolar patients. J Clin Psychiatry 62(supp 25):31–34. Pearlson GD, Buarta PE, Powers RE, Menon RR, Richards SS, Aylward EH, et al (1997): Ziskind-Somerfeld Research Award 1996. Medial and superior temporal gyral volumes and cerebral asymmetry in bipolar disorder. Biol Psychiatry 44:418 – 422. Sassi RB, Nicoletti M, Brambilla P, Mallinger AG, Frank E, Kupfer DJ, et al (2002): Increased gray matter volume in lithium-treated bipolar disorder patients. Neurosci Lett 329:243–245. Sheline YI, Sanghavi M, Mintun MA, Gado MH (1999): Depression duration but not age predicts hippocampal volume loss in medically healthy women with recurrent major depression. J Neurosci 19:5034 –5043. Soares JC (2003): Contributions from brain imaging to the elucidation of pathophysiology of bipolar disorder. Int J Neuropsychopharmacol 6:171– 180. Soares JC, Mann JJ (1997): The functional neuroanatomy of mood disorders. J Psychiatr Res 31:393– 432. Spearing M, Hendrix ML, Hyman SE, Rudorfer MV, Pearson JL, Wittenberg CK, et al (2000): Child and adolescent bipolar disorder: An update from the National Institute of Mental Health. Available at: http://www.nimh.nih. gov/publicat/bipolarupdate.cfm. Accessed January 8, 2003. Spitzer RL, Williams JBW, Gibbon M, First MG (1994), Structured Clinical Interview for DSM-III-R (SCID). Washington, DC: American Psychiatric Press. Strakowski SM., DelBello MP, Sax KW, Zimmerman ME, Shear PK, Hawkins JM, Larson ER (1999): Brain magnetic resonance imaging of structural abnormalities in bipolar disorder. Arch Gen Psychiatry 56:254 –260. Strakowski SM, Wilson DR, Tohen M, Woods BT, Douglass AW, Stoll AL (1993): Structural brain abnormalities in first-episode mania. Biol Psychiatry 33: 602– 609. Steffens DC, Byrum CE, McQuoid DR, Greenberg DL, Payne ME, Blitchington TF, et al (2000): Hippocampal volume in geriatric depression. Biol Psychiatry 48:301–309. Swayze VW2nd, Andreasen NC, Alliger RJ, Yuh WT, Ehrhardt JC (1992): Subcortical and temporal structures in affective disorder and schizophrenia: A magnetic resonance imaging study. Biol Psychiatry 31:221–240. Vakili K, Pillay SS, Lafer B, Fava M, Renshaw PF, Bonello-Cintron CM, et al (2000): Hippocampal volume in primary unipolar major depression: A magnetic resonance imaging study. Biol Psychiatry 47:1087–1090.
www.elsevier.com/locate/biopsych
B
ipolar disorder, or manic-depressive illness, affects more than 2 million American adults (Spearing et al 2000). The exact prevalence among adolescents and young adults is more difficult to identify, because the symptoms may be similar to or occur concurrently with other childhood psychiatric disorders (National Institute of Mental Health 2000). As a major psychiatric disorder, bipolar disorder is implicated in functional impairment and represents an important risk factor for suicide (Oquendo and Mann 2001). Research into the pathogenesis of this disorder may help prevent misdiagnosis and improve currently available therapeutic strategies. In the past, bipolar disorder was classified as being an “endogenous” disorder for which there was no identified anatomic substrate (Alvarez 2001). During the last two decades, however, the development of modern neuroimaging techniques has enabled researchers to detect subtle structural and functional changes in the brains of patients suffering from mood disorders, which challenged the traditional concept that these disorders were “endogenous” or functional illnesses. Brain neuroanatomic circuits interconnecting the prefrontal cortex, medial temporal From the Department of Psychiatry (BKC, JPH, MS, MN, PB, JCS), and Radiology (JCS), The University of Texas Health Science Center at San Antonio, San Antonio, Texas; Department of Psychiatry (RS, DA, MSK, NDR, BB), University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Department of Psychiatry (RS), Institute of Psychiatry, University of Sao Paulo School of Medicine, and Department of Psychiatry (MS), Federal University of Sao Paulo, Sao Paulo, Brazil; IRCCS S. Giovanni di Dio (PB), Fatebenefratelli, Brescia, Italy; and South Texas Veterans Health Care System (JCS), Audie L. Murphy Division, San Antonio, Texas. Address reprint requests to Jair C. Soares, M.D., Division of Mood and Anxiety Disorders, Department of Psychiatry (MC 7792), University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900; E-mail: [email protected]. Received September 16, 2004; revised June 1, 2004; accepted June 15, 2004.
0006-3223/04/$30.00 doi:10.1016/j.biopsych.2004.06.024
lobe, and basal ganglia are involved in mood regulation. Morphometric studies suggest anatomic abnormalities in key structures in these neuroanatomic circuits in the pathophysiology of bipolar disorder (Beyer and Krishnan 2002; Soares 2003; Soares and Mann 1997). The amygdala and hippocampus play key roles in mood and behavior. Both are components of the limbic system, which influences emotional excitability, subjective emotional experiences, and sexual libido. Specifically, because the hippocampus functions to identify and bridge relationships in memory, a disruption in this structure would impair cognitive functions and behavior (Nagode and Pardo 2002). Direct emotional responses to information, especially to fear and anxiety, occur in the amygdala (Adolphs et al 1994; Gallagher and Chiba 1996; Morris 1996). Together, amygdala and hippocampus translate cognitive input in emotional processing (LeDoux 1996), and as a result, abnormalities in these brain structures may be implicated in the pathophysiology of bipolar disorder. Most studies showed enlargement of the amygdala in adult bipolar patients compared with healthy control subjects, without clear indications of size abnormalities in the hippocampus (Altshuler et al 2000; Brambilla et al 2003; Strakowski et al 1999). Because the majority of existing studies focused on adults, it is not known whether such enlargement would be present in younger patients suffering from bipolar disorder. Blumberg et al (2003) recently reported smaller amygdala size in both pediatric and adult bipolar disorder patients. If present in the young age group, amygdala abnormalities would likely reflect neurodevelopmental insults or changes that develop early in the course of the illness. Through the processes of pruning and myelination, the brain removes unnecessary gray matter while the white matter grows during late adolescence (Giedd et al 1999; Gogate et al 2001; Gogtay et al 2002). In this cross-sectional study, amygdala, hippocampus, and temporal lobe volumes of bipolar patients and healthy control subjects were determined with high-resolution magnetic resonance imaging (MRI). We compared these volumes to determine whether any size differences existed between these two groups, BIOL PSYCHIATRY 2004;56:399 – 405 © 2004 Society of Biological Psychiatry
400 BIOL PSYCHIATRY 2004;56:399 – 405
B.K. Chen et al
Table 1. Clinical and Demographic Information Bipolar Subject
Age (years)
Race
Gender
Handedness
Age at Onset (years)
Lithium
1 2 3
10 10 12
White White White
M M M
Left Right Left
7 4 6
⫹ ⫹
4 5 6 7 8 9
13 14 14 14 15 15
White White White White White White
F M F M F F
Right Right Right Right Right Right
10 12 11 10 14 14
10 11 12 13 14 15 16
17 17 18 18 19 21 21
White White White White White White Black
F M F F F M M
Left Right Right Right Right Right Right
13 12 15 8 17 19 15
⫹ ⫹
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
Valproate
Other Medications Olanzapine Thyroid hormone, methylphenidate
⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹
Clomipramine, dexedrine Fluoxetine, lorazepam, thyroid hormone Trazodone Thyroid hormone Citalopram
F, female; M, male.
with the hypothesis that bipolar patients would have enlargement of the amygdala. Furthermore, age effects on temporal lobe structures were compared in the two groups to examine possible abnormal neurodevelopment.
Methods and Materials Subjects Thirty-seven young bipolar patients and control subjects aged 10 through 21 (mean age ⫾ SD ⫽ 16 ⫾ 4 years) volunteered as subjects for this study. Table 1 shows the demographic information of the 16 DSM-IV bipolar disorder patients (mean age ⫾ SD ⫽ 16 ⫾ 3 years; 8 female subjects, 8 male subjects; 15 Caucasians, 1 African American; 12 bipolar I type, 3 bipolar II type, 1 bipolar not otherwise specified). Among the bipolar patients, 14 were euthymic and 2 were mildly depressed at the time of the study. All but one reported a familial history of mood disorders, defined as having a first-degree relative who ever received a diagnosis of unipolar or bipolar disorder. Fourteen patients were receiving treatment for bipolar disorder (six lithium only, four valproate only, and four lithium plus valproate). Eleven patients reported previous use of antipsychotic drugs. The mean duration of the disease among patients was 3.93 years (SD ⫽ 2.37). Comorbid mental disorders were found in seven patients and consisted of attention-deficit/hyperactivity disorder (five subjects), oppositional– defiant disorder (one subject), and conduct disorder (one subject). Twenty-one healthy control subjects without any history of psychiatric problems in first-degree relatives (mean age ⫾ SD ⫽ 17 ⫾ 4 years; 9 female subjects, 12 male subjects; 18 Caucasians, 3 African Americans) were studied. The psychiatric diagnoses were confirmed or ruled out through the Schedule for Affective Disorders and Schizophrenia for School-Age Children—Present and Lifetime Version (K-SADS; Kaufman et al 1997) for children up to 17 years old, or the Structured Clinical Interview for DSM-IV (SCID-IV; Spitzer et al 1998) for subjects who were 18 or older. The University of Pittsburgh Biomedical Institutional Review Board approved this study, and all subjects or their parents or legal representatives gave written informed consent after understanding all issues involved in study participation. www.elsevier.com/locate/biopsych
MRI Procedures The MRI images were obtained at a 1.5-T GE Signa Imaging System running the Signa 5.4.3 (General Electric Medical Systems, Milwaukee, Wisconsin) software at the University of Pittsburgh. All patients were placed in the MRI in a standardized position determined by sagittal scout series to verify head position and to decrease interscan variability. A plastic head holder was used to maintain placement of the subject’s head and to limit any head movement during the MRI session. Using three-dimensional gradient echo imaging by spoiled gradient recalled acquisition, 124 images covering the whole brain were acquired within the coronal plane with 1.5-mm-thick slices (repetition time ⫽ 25 msec, echo time ⫽ 5 msec, nutation angle ⫽ 40°, field of view ⫽ 24 cm, number of excitations ⫽ 1, matrix size ⫽ 256 ⫻ 192). T2 and protondensity images were obtained in the axial plane to screen for neuroradiologic abnormalities. Image Analysis All images were analyzed using Scion Image for Windows Beta-3b version on a PC workstation to obtain the volumes (in milliliters) of the amygdala, hippocampus, temporal lobe, and intracranial brain volume (ICV). Gray matter volumes were determined using a semiautomated segmentation algorithm (Keshavan et al 1994, 1995). Well-trained raters blinded to group assignments and subject identity performed all measurements, and interrater reliability was measured by intraclass correlation coefficients, determined by two independent raters tracing 10 scans. These values were as follows: right hippocampus (r ⫽ .92), left hippocampus (r ⫽ .93), right amygdala (r ⫽ .95) left amygdala (r ⫽ .91), total right temporal lobe (r ⫽ .96), total left temporal lobe (r ⫽ .99), right temporal lobe gray matter (r ⫽ .94), and left temporal lobe gray matter (r ⫽ .96). Hippocampus The right and left hippocampi were traced separately. The tracing of the hippocampus began where the superior colliculus completely joined the thalamus, as shown in Figure 1A. The hippocampus was visible inferiorly and laterally to the thalamus and directly laterally to the ambient cistern. Moving anteriorly,
BIOL PSYCHIATRY 2004;56:399 – 405 401
B.K. Chen et al
Table 3. Temporal Lobe Structures of Lithium-Free and Lithium-Treated Bipolar Patients Temporal Lobe Structure
No Lithium (n ⫽ 6) Mean ⫾ SD (mL)
Lithium (n ⫽ 10) Mean ⫾ SD (mL)
df
p
Right Hippocampus Left Hippocampus Right Amygdala Left Amygdala
3.07 ⫾ .51 3.07 ⫾ .55 2.00 ⫾ .59 1.59 ⫾ .58
3.07 ⫾ .29 3.16 ⫾ .31 2.21 ⫾ .22 1.58 ⫾ .33
14 14 14 14
.98 .68 .33 .96
Statistical testing was performed using the t test.
temporal lobe to the cerebrum, disappeared or where the amygdala gray matter became too diffuse to resolve from the surrounding temporal lobe gray matter (Figure 1D). As shown in Figure 1C, the more anteriorly located amygdala appeared in the same position as the hippocampus in more posterior slices. Its superior border extended into the white matter tract separating it from the globus pallidus. The inferior border was the parahippocampal gyrus. The areas of all slices were summed, and this sum was then multiplied by the slice thickness (.15 cm) to obtain volume (milliliters). Figure 1. Magnetic resonance imaging tracings of hippocampus and amygdala. The hippocampus and amygdala were traced using the following anatomical landmarks. (A) The most posterior or first slice of the hippocampus began where the superior colliculus joined the thalamus. (B) The most anterior or last slice of the hippocampus ended when the third ventricle splits from the cistern by the hypothalamus. (C) The most posterior slice of amygdala began where the mamillary bodies were separated. It appeared in the same position as the hippocampus in more posterior slices, but the superior border became the globus pallidus. (D) The amygdala tracing ended where the temporal bridge disappeared or the gray matter became too diffuse.
the ambient cistern appeared superior to the hippocampus. The tracing continued to the point where the third ventricle was split from the cistern by the hypothalamus as shown in Figure 1B. The amygdala was not included because it appeared above the hippocampus. Amygdala The right and left amygdalae were traced separately. Tracing began at the slice where the third ventricle and suprasellar cistern were clearly joined as the mamillary body (Figure 1C) and ended where the temporal bridge, the white matter that connects the
Temporal Lobe Temporal lobes were traced from the most posterior slice where the temporal lobe appeared, where the splenium of the corpus callosum was initially most pronounced. The superior border was the transverse fissure and the maximal limits of the gray and white matter of the temporal lobe were adopted as the inferior, medial, and lateral limits. The temporal stem, the white matter tract that links temporal lobe to the rest of the hemisphere, was excluded from the tracing. Intracranial Volume The ICV was obtained by tracing the border of the brain, which included total cerebral gray and white matter, cerebrospinal fluid, dura matter, sinuses, optic chiasm, pituitary, cerebellum, and brainstem. Every other slice was traced and multiplied by .3 cm, the thickness of two slices, because for measures of ICV we traced only every other slice. Statistical Analyses Statistical analysis was completed using Systat software, version 8.1 (SPSS, Chicago, Illinois) on a PC Windows workstation. Asymmetry in medial temporal lobe structures was evaluated
Table 2. Volumes of Temporal Lobe Structures Temporal Lobe Structure
Bipolar (n ⫽ 16) Mean ⫾ SD (mL)
Control (n ⫽ 21) Mean ⫾ SD (mL)
Right Hippocampus Left Hippocampus Right Amygdala Left Amygdala Right Temporal Lobe Gray Matter Left Temporal Lobe Gray Matter Right Temporal Lobe Left Temporal Lobe Intracranial Volume
3.07 ⫾ .37 3.12 ⫾ .40 2.13 ⫾ .39 1.58 ⫾ .42 62.23 ⫾ 7.45
3.06 ⫾ .27 3.15 ⫾ .33 2.14 ⫾ .57 1.83 ⫾ .42 58.75 ⫾ 6.81
61.12 ⫾ 8.71 84.94 ⫾ 11.18 77.26 ⫾ 12.77 1,518.56 ⫾ 141.14
F
df
p
.05 .41 .17 3.87 1.34
1/32 1/32 1/32 1/32 1/32
.82 .53 .69 .06 .26
60.36 ⫾ 7.57
.13
1/32
.72
79.43 ⫾ 10.32 76.31 ⫾ 9.62 1,479.81 ⫾ 162.51
3.13 .037 .58
1/32 1/32 1/35
.09 .85 .45
Statistical testing was performed using analysis of covariance with age, gender, and intracranial volume (ICV) as covariates; ICV was determined by analysis of variance. Unadjusted means and standard deviations are shown.
www.elsevier.com/locate/biopsych
402 BIOL PSYCHIATRY 2004;56:399 – 405
B.K. Chen et al
with repeated measures analyses of covariance (ANCOVA) with age, gender, and ICV as covariates. Left and right amygdala, hippocampus, temporal lobe gray matter, and temporal lobe total volumes of bipolar and control subjects were compared using ANCOVA with age, gender, and ICV as covariates. Analysis of variance (ANOVA) was performed to compare the ICV between bipolar patients and control subjects. Unadjusted volume means ⫾ SD are reported for all brain structures. In checking the ANCOVA assumption of equal regression slopes, it was determined that the interaction of the age covariate with diagnostic groups was statistically significant (p ⫽ .012) for the analysis of left amygdala. Therefore, this interaction term was retained in the analysis of the left amygdala. Pearson’s correlations were calculated to assess the relation of age to amygdala and hippocampus volumes. Z tests were performed to compare Pearson’s correlation and partial correlation coefficients adjusting for gender and ICV. A two-tailed p value less than .05 was considered for rejecting all null hypotheses.
Results The bipolar patients did not differ significantly from the healthy control subjects in age (t test, p ⫽ .256), gender (Fisher’s Exact Test, p ⫽ .746), race (Fisher’s Exact Test, p ⫽ .392), handedness (Fisher’s Exact Test, p ⫽ .296), height (t test, p ⫽ .330), weight (t test, p ⫽ .840), ICV (t test, p ⫽ .453), years of education (Mann-Whitney U test, p ⫽ .387), or educational achievement (Pearson chi-square, p ⫽ .218). No statistically significant differences were found between the bipolar and healthy groups for any of the structures measured (Table 2). There was a trend for the left amygdala to be smaller in the bipolar patients (F ⫽3.87, df⫽ 1,32, p ⫽ .06). No significant differences were found for any of the measured structures between lithium-treated bipolar patients and bipolar patients who were not on lithium (Table 3). Asymmetry was not found for either hippocampus (F ⫽ 1.27, df ⫽ 1,32, p ⫽ .269) or amygdala (F ⫽ .14, df ⫽ 1,32, p ⫽ .711). There also was no significant interaction between diagnostic group and side for either structure. Bipolar subjects with a comorbid diagnosis have significantly smaller left amygdala volumes (1.24 ⫾ .27 mL, n ⫽ 7) compared with bipolar subjects without a comorbid diagnosis (1.85 ⫾ .30 mL, n ⫽ 9; t ⫽ 4.21, p ⫽ .001). The left amygdala volume of bipolar patients without comorbidity was not statistically different from the values for the healthy control subjects (p ⬎ .05). The right amygdala and hippocampus volumes of bipolar subjects with comorbidity compared with those without comorbidity were not significantly different (p ⬎ .05). For healthy control subjects, there was a significant inverse relationship between age and left amygdala volumes (r ⫽ ⫺.48, df ⫽ 19, p ⫽ .03). In contrast, bipolar patients demonstrated a significant direct relationship between age and left amygdala volumes (r ⫽ .50, df ⫽ 14, p ⫽ .047), as seen in Figure 2. These correlations were significantly different (Z ⫽ 2.91, p ⫽ .004). Pearson’s correlation demonstrated no significant association between age and right hippocampus, left hippocampus, or right amygdala volumes in either group (bipolar: r ⫽ ⫺.15, p ⫽ .58; r ⫽ .03, p ⫽ .92; r ⫽ .39, p ⫽ .14; respectively; control: r ⫽ ⫺.02, p ⫽ .93; r ⫽ ⫺.26, p ⫽ .25; r ⫽ ⫺.31, p ⫽ .17). Partial correlations adjusting for gender and ICV of healthy control subjects (r ⫽ ⫺.36, df ⫽ 17, p ⫽ .134) and bipolar patients (r ⫽ .48, df ⫽ 12, p ⫽ .081) were not statistically significant. Nevertheless, the Z test comparing the partial correwww.elsevier.com/locate/biopsych
Figure 2. Age and left amygdala volumes in bipolar patients and healthy control subjects. Best fit curves showing left amygdala volumes versus age of bipolar and control subjects. (A) Bipolar left amygdala volumes show a direct relationship with age. (B) Normal subjects show an inverse relationship between left amygdala volume and age.
lations controlling for gender and ICV remained statistically significant (Z ⫽ 2.48, p ⫽ .007). Table 4 displays the Pearson’s correlations for male and female subjects separately. A significant direct relationship between age and left amygdala was observed in male bipolar patients (r ⫽ .73, df ⫽ 6, p ⫽ .04), whereas a significant inverse relationship was observed for male healthy control subjects (r ⫽ ⫺.70, df ⫽ 10, p ⫽ .01). Pearson’s correlations for female bipolar patients were nonsignificant (p ⬎ .05). Correlations between age and other temporal lobe structures were not significant in either gender. Figure 3 depicts the linear relationship of left amygdala volumes and age for male bipolar patients and healthy control subjects.
BIOL PSYCHIATRY 2004;56:399 – 405 403
B.K. Chen et al Table 4. Relationship Between Age and Volumes of Temporal Lobe Structures Male
Female
Bipolar (n ⫽ 8) Temporal Lobe Structure Right Hippocampus Left Hippocampus Right Amygdala Left Amygdala ICV
Control (n ⫽ 12)
Bipolar (n ⫽ 8)
Control (n ⫽ 9)
r
df
p
r
df
p
r
df
p
r
df
p
⫺.13 .07 .37 .73 .59
6 6 6 6 6
.77 .87 .37 .04 .12
.30 .15 ⫺.23 ⫺.70 ⫺.48
10 10 10 10 10
.35 .64 .47 .01 .11
⫺.07 .10 .48 .20 ⫺.30
6 6 6 6 6
.87 .81 .23 .64 .47
.11 ⫺.55 ⫺.26 .10 ⫺.23
7 7 7 7 7
.78 .12 .49 .79 .55
ICV, intracranial volume.
Discussion Our study found no significant differences in measures of temporal lobe, right amygdala, and hippocampus between young individuals with bipolar disorder and matched healthy control subjects. Nonetheless, there was a trend for smaller left amygdala volumes in patients, in contrast with several studies involving adult bipolar disorder individuals (Altshuler et al 1998, 2000; Brambilla et al 2003; Strakowski et al 1999) that found enlarged amygdala volumes. Conversely, Pearlson et al (1997) found smaller left amygdala volumes in adult bipolar patients compared with control subjects. Findings of smaller amygdala volumes in young patients with bipolar disorder suggest abnormalities in this brain structure that are already detectable early in the course of the illness. Such changes, in the context of adult studies that indicated enlarged amygdala, suggest that abnormal development of limbic structures may be involved in pathophysiology of bipolar disorder. The most pronounced abnormality was presented in bipolar patients with comorbidity, but this finding stemmed from a reduced sample size and is preliminary. It is unknown whether attention-deficit/hyperactivity disorder alone or together with bipolar disorder decreases amygdala volume. Studies with a larger sample size are necessary to further explore this question. Prior studies, consistent with our present findings, predominantly showed no evidence of anatomic abnormalities in the hippocampus of adult bipolar patients (Altshuler et al 1998, 2000; Brambilla et al 2003; Pearlson et al 1997; Strakowski et al 1999). Only one study (Swayze et al 1992) found a smaller hippocampal volume in adult bipolar individuals. Conversely, most studies of hippocampal volumes in unipolar patients found evidence of a size reduction (Beyer and Krishnan 2002; Bremner et al 2000; Mervaala et al 2000; Sheline et al 1999; Steffens et al 2000), whereas a few found no abnormalities (Altshuler et al 1998; Axelson et al 1993; Beyer et al 2002; Hauser et al 2000; Vakili et al 2000). Hippocampal volume reduction in unipolar patients is possibly caused by a hyperactivity within the hypothalamic– pituitary–adrenal axis and resulting hypercortisolemia. Our findings of normal temporal lobe volumes in young bipolar patients are consistent with most reports in adult bipolar patients (Altshuler et al 1998; Beyer et al 2002; Hauser et al 2000; Johnstone et al 1989). These findings differ from the studies by Hauser et al (1989) and Altshuler et al (1991), in which smaller temporal lobe volumes were found, and Harvey et al (1994), in which increased gray matter volumes in left temporal lobes were reported in adult bipolar individuals. Brain gray matter is progressively pruned during adolescence, especially between the ages of 14 and 17, which is the age span when the incidence of psychiatric disorders increases (Giedd 1999; Gogtay 2002). Giedd’s 1999 study involved healthy chil-
dren and adolescents and demonstrated pruning of gray matter in adolescence. Our findings of abnormal relationship between age and left amygdala volumes in bipolar patients suggest developmental abnormalities in this brain structure, which could reflect abnormal pruning, although the present cross-sectional study cannot definitively demonstrate this finding. Future longitudinal studies examining medial temporal lobe development in bipolar children, adolescents, and young adults will be necessary. A potential limitation of this study is that most patients were on medications at the time they were studied; however, our results suggest that medication effects are not potential confounders to our reported analyses. Drug-free bipolar patients did not differ from lithium-treated patients in measures of temporal lobe structures. Furthermore, our findings are in agreement with our prior report in adult bipolar patients, where amygdala and hippocampus volumes did not differ between lithium-treated and drug-free bipolar adults (Brambilla et al 2003). Although other studies did show that the treatment of bipolar disorder patients with lithium increased total brain gray matter (Moore et al 2000; Sassi et al 2002), any possible effects of lithium on amygdala and hippocampus size have not been characterized. Additionally, our study mainly included euthymic bipolar I patients. A study by Blumberg et al (2003) found no significant difference between euthymic, depressed, and manic bipolar patients, although their sample size for each state was small. Our study also indicates that left amygdala volumes among young male bipolar patients may increase as age increases, which is opposite of what is found in healthy male control subjects. In normal childhood development, the hippocampus and amygdala total volumes increase in both males and females, with the amygdala increasing significantly more in males and the hippocampus increasing significantly more in female children (Durston et al 2001; Giedd et al 1997); however, gray matter in the temporal lobe decreases in midadolescence and continues to decrease throughout adulthood in males and females (Bartzokis et al 2001; Giedd et al 1999). Our findings of increasing left amygdala size in males to the age of 21 years may indicate abnormal development of this brain structure in male bipolar patients, which could result in hypertrophy of this brain region in adult bipolar patients. Such findings, if confirmed in longitudinal studies, could be key steps underlying the pathophysiology of the emotional disregulation found in bipolar disorder. Patients with bilateral amygdala damage show placidness and hypoemotionality (Gallagher and Chiba 1996). Findings of enlargement of the left amygdala are consistent with findings of increased cerebral blood flow and glucose metabolism in this brain structure in adult bipolar patients (Drevets 2000). The abnormal relationship between age and left amygdala www.elsevier.com/locate/biopsych
404 BIOL PSYCHIATRY 2004;56:399 – 405
B.K. Chen et al In summary, no significant differences in volumes of hippocampus, amygdala, or temporal lobe were found between bipolar patients and matched healthy control subjects, even though there was a trend for smaller left amygdala volumes in patients. These findings suggest developmental abnormalities in amygdala in bipolar individuals. Even though left amygdala volumes may be smaller in younger bipolar patients because of possible abnormalities in pruning mechanisms during late adolescence, adult bipolar patients could end up with an enlargement of this brain structure, as suggested by the fact that the age effects on this structure in younger bipolar and healthy subjects are strikingly different. Nonetheless, because of the cross-sectional nature of our findings, future studies with larger patient samples and a longitudinal design are necessary.
This study was supported in part by National Institute of Mental Health Grants MH 01736, MH 55123, MH 30915 and MH 59929; the National Alliance for Research on Schizophrenia and Affective Disorders; the Office of the Dean, University of Texas Health Science Center at San Antonio Medical School (UTHSCSA), the Veterans Administration, and the Krus Endowed Chair in Psychiatry (UTHSCSA).
Figure 3. Age and left amygdala volumes in male subjects. Best fit curves showing left amygdala volumes versus age of male bipolar patients and control subjects. (A) Left amygdala of male bipolar patients showing a direct relationship to age. (B) Left amygdala of healthy male subjects showing an inverse relationship to age.
volumes was restricted to male bipolar type I patients and was not present in female patients or control subjects. Our findings should be seen as preliminary, because sample sizes are relatively small. Nonetheless, these results suggest, as highlighted in previous studies (Jurjus et al 1993; Strakowski et al 1993), that male and female patients with mood disorders may differ in regard to neuroimaging findings and vulnerability to developmental factors implicated in the pathophysiology of mood disorders. For instance, Brown et al (1995) reported higher relative risk for affective psychosis, associated with prenatal factors such as exposure to famine, among male subjects. Furthermore, in Altshuler et al (2000), the findings of enlarged amygdala were restricted to a group of adult bipolar males. www.elsevier.com/locate/biopsych
Adolphs R, Tranel D, Damasio H, Damasio A (1994): Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala. Nature 372:669 – 672. Altshuler LL, Bartzokis G, Grieder T, Curran J, Jimenez T, Leight K, et al (2000): An MRI study of temporal lobe structures in men with bipolar disorder or schizophrenia. Biol Psychiatry 48:147–162. Altshuler LL, Bartzokis G, Grieder T, Curran J, Mintz J (1998): Amygdala enlargement in bipolar disorder and hippocampal reduction in schizophrenia: An MRI study demonstrating neuroanatomic specificity. Arch Gen Psychiatry 55:663– 664. Altshuler LL, Conrad A, Hauser P, Li X, Guze GN, Denicoff K, et al (1991): Temporal lobe volume reduction in bipolar disorder: A preliminary MRI report. Arch Gen Psychiatry 48:482– 483. Alvarez J (2001): Neuronal hypersynchronization, creativity and endogenous psychoses. Med Hypotheses 56:672– 685. Axelson DA, Doraiswamy P, McDonald WM, Boyko OB, Tupler LA, Patterson LJ, et al (1993): Hypercortisolemia and hippocampal changes in depression. Psychiatry Res 47:163–173. Bartzokis G, Beckson M, Lu PH, Nuechterlein KH, Edwards N, Mintz J (2001): Age-related changes in frontal and temporal lobe volumes in men: A magnetic resonance imaging study. Arch Gen Psychiatry 58:461– 465. Beyer JL, Krishnan KRR (2002): Volumetric brain imaging findings in mood disorders. Bipolar Disord 4:89 –104. Blumberg HP, Kaufman J, Martin A, Whiteman, R, Zhang JH, Gore JC, et al (2003): Amygdala and hippocampal volumes in adolescents and adults with bipolar disorder. Arch Gen Psychiatry 60:1201–1208. Brambilla P, Harenski K, Nicoletti M, Sassi RB, Mallinger AG, Frank E, et al (2003): MRI investigation of temporal lobe structures in bipolar patients. J Psychiatr Res 37:287–295. Bremner JD, Narayan M, Anderson ER, Staib LH, Miller HL, Charney DS (2000): Hippocampal volume reduction in major depression. Am J Psychiatry 157:115–117. Brown AS, Susser ES, Lin SP, Neugebauer R, Gorman JM (1995): Increased risk of affective disorders in males after second trimester prenatal exposure to the Dutch hunger winter of 1944-45. Br J Psychiatry 166:601– 606. Drevets WC (2000): Neuroimaging studies of mood disorders. Biol Psychiatry 48:813– 829. Durston S, Hulshoff Pol HE, Casey BJ, Giedd JN, Buitelaar JK, van Engeland H (2001): Anatomical MRI of the developing human brain: What have we learned? J Am Acad Child Adolesc Psychiatry 40:1012–1020. Gallagher M, Chiba AA (1996): The amygdala and emotion. Neurobiology 6:221–227. Giedd JN, Blumenthal J, Jeffries NO, Castellanos FX, Liu H, Zijdenbos A, et al (1999): Brain development during childhood and adolescence: A longitudinal MRI study. Nature Neurosci 2:861– 863.
B.K. Chen et al Giedd JN, Castellanos FX, Rajapakse JC, Vaituzis AC, Rapoport JL (1997): Sexual dimorphism of the developing human brain. Prog Neuropsychoparmacol Biol Psychiatry 21:1185–1201. Gogtay N, Giedd J, Janson K, Rapoport JL (2001): Brain imaging in normal and abnormal brain development: New perspectives for child psychiatry. Clin Neurosci Res 1:283–290. Gogtay N, Giedd J, Rapoport JL (2002): Brain development in healthy, hyperactive, and psychotic children. Arch Neurol 59:1244 –1248. Harvey I, Persaud R, Ron MA, Baker G, Murray RM (1994): Volumetric MRI measurements in bipolars compared with schizophrenics and healthy controls. Psychol Med 24:689 – 699. Hauser P, Altshuler LL, Berrettini W, Dauphinais ID, Gelernter J, Post RM (1989): Temporal lobe measurement in primary affective disorder by magnetic resonance imaging. J Neuropsychiatry Clin Neurosci 1:128 – 134. Hauser P, Matochik J, Altshuler LL, Denicoff KD, Conrad A, Li X, et al (2000): MRI-based measurements of temporal lobe and ventricular structures in patients with bipolar I and bipolar II disorders. J Affect Disord 60:25–32. Jurjus GJ, Nasrallah HA, Olson SC, Schwarzkopf SB (1993): Cavum septum pellucidum in schizophrenia, affective disorder and healthy controls: A magnetic resonance imaging study. Psychol Med 23:319 –322. Kaufman J, Birmaher B, Brent D, Ryan ND, Rao U (1997): Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL): Initial reliability and validity data. J Am Acad Child Adolesc Psychiatry 36:980 –988. Keshavan MS, Anderson S, Beckwith C, Nash K, Pettegrew JW, Krishnan KR (1995): A comparison of stereology and segmentation techniques for volumetric measurements of lateral ventricles in magnetic resonance imaging. Psychiatry Res 61:53– 60. Keshavan MS, Beckwith C, Bagwell W, Pettegrew W, Krishnan KR (1994): An objective method for edge detection in MRI morphometry. Eur Psychiatry 9:205–207. Johnstone EC, Owens DGC, Crow TJ, Frith CD, Alexandropolis K, Bydder G, Colter N (1989): Temporal lobe structure as determined by nuclear magnetic resonance in schizophrenia and bipolar affective disorder. J Neurol Neurosurg Psychiatry 53:736 –741. LeDoux J (1996): The Emotional Brain. New York: Simon & Schuster. Mervaala E, Fohr J, Kononen M, Valkonen-Korhonen M, Vainio P, Partanen K, et al (2000): Quantitative MRI of the hippocampus and amygdala in severe depression. Psychol Med 30:117–125. Moore GJ, Bebchuk JM, Wilds IB, Chen G, Manji HK (2000): Lithium-induced increase in human brain gray matter. Lancet 356:1241–1242. Morris JS, Frith CD, Perrett DI, Rowland D, Young AW, Calder AJ, Dolan RJ (1996): A differential neural response in the human amygdala to fearful and happy facial expressions. Nature 383:812– 815.
BIOL PSYCHIATRY 2004;56:399 – 405 405 Nagode JC, Pardo JV (2002): Human hippocampal activation during transitive inference. Neuroreport 13:939 –944. National Institute of Mental Health (2000): Child and adolescent bipolar disorder: An update from the National Institute of Mental Health. Available at: http://www.nimh.nih.gov/publicat/bipolarupdate.cfm. Accessed January 8, 2003. Oquendo MA, Mann JJ (2001): Identifying and managing suicide risk in bipolar patients. J Clin Psychiatry 62(supp 25):31–34. Pearlson GD, Buarta PE, Powers RE, Menon RR, Richards SS, Aylward EH, et al (1997): Ziskind-Somerfeld Research Award 1996. Medial and superior temporal gyral volumes and cerebral asymmetry in bipolar disorder. Biol Psychiatry 44:418 – 422. Sassi RB, Nicoletti M, Brambilla P, Mallinger AG, Frank E, Kupfer DJ, et al (2002): Increased gray matter volume in lithium-treated bipolar disorder patients. Neurosci Lett 329:243–245. Sheline YI, Sanghavi M, Mintun MA, Gado MH (1999): Depression duration but not age predicts hippocampal volume loss in medically healthy women with recurrent major depression. J Neurosci 19:5034 –5043. Soares JC (2003): Contributions from brain imaging to the elucidation of pathophysiology of bipolar disorder. Int J Neuropsychopharmacol 6:171– 180. Soares JC, Mann JJ (1997): The functional neuroanatomy of mood disorders. J Psychiatr Res 31:393– 432. Spearing M, Hendrix ML, Hyman SE, Rudorfer MV, Pearson JL, Wittenberg CK, et al (2000): Child and adolescent bipolar disorder: An update from the National Institute of Mental Health. Available at: http://www.nimh.nih. gov/publicat/bipolarupdate.cfm. Accessed January 8, 2003. Spitzer RL, Williams JBW, Gibbon M, First MG (1994), Structured Clinical Interview for DSM-III-R (SCID). Washington, DC: American Psychiatric Press. Strakowski SM., DelBello MP, Sax KW, Zimmerman ME, Shear PK, Hawkins JM, Larson ER (1999): Brain magnetic resonance imaging of structural abnormalities in bipolar disorder. Arch Gen Psychiatry 56:254 –260. Strakowski SM, Wilson DR, Tohen M, Woods BT, Douglass AW, Stoll AL (1993): Structural brain abnormalities in first-episode mania. Biol Psychiatry 33: 602– 609. Steffens DC, Byrum CE, McQuoid DR, Greenberg DL, Payne ME, Blitchington TF, et al (2000): Hippocampal volume in geriatric depression. Biol Psychiatry 48:301–309. Swayze VW2nd, Andreasen NC, Alliger RJ, Yuh WT, Ehrhardt JC (1992): Subcortical and temporal structures in affective disorder and schizophrenia: A magnetic resonance imaging study. Biol Psychiatry 31:221–240. Vakili K, Pillay SS, Lafer B, Fava M, Renshaw PF, Bonello-Cintron CM, et al (2000): Hippocampal volume in primary unipolar major depression: A magnetic resonance imaging study. Biol Psychiatry 47:1087–1090.
www.elsevier.com/locate/biopsych