Medial temporal lobe abnormalities in pediatric unipolar depression

Transcription

1 Available online at Neuroscience Letters 427 (2007) Medial temporal lobe abnormalities in pediatric unipolar depression Sheila C. Caetano a,b,c, Manoela Fonseca a,b,d, John P. Hatch a,e, Rene L. Olvera f, Mark Nicoletti a, Kristina Hunter a, Beny Lafer c, Steven R. Pliszka f, Jair C. Soares a,b,g, a Division of Mood and Anxiety Disorders, Department of Psychiatry, The University of Texas Health Science Center at San Antonio, San Antonio, TX, USA b Psychiatry Service, South Texas Veterans Health Care System, Audie L. Murphy Division, San Antonio, TX, USA c Department of Psychiatry, University of São Paulo Medical School, São Paulo, Brazil d Psychiatry Research Unit, Federal University of Rio Grande do Sul, Medical School, Porto Alegre, Brazil e Department of Orthodontics, The University of Texas Health Science Center at San Antonio, San Antonio, TX, USA f Division of Child and Adolescent Psychiatry, Department of Psychiatry, The University of Texas Health Science Center at San Antonio, San Antonio, TX, USA g Center of Excellence for Research and Treatment of Bipolar Disorder (CERT-BD), Department of Psychiatry, UNC School of Medicine, Chapel Hill, NC, USA Received 13 April 2007; received in revised form 18 May 2007; accepted 6 June 2007 Abstract In vivo anatomical magnetic resonance imaging (MRI) studies in adults with major depressive disorder (MDD) have implicated neurocircuitries involved in mood regulation in the pathophysiology of mood disorders. Specifically, abnormalities in the medial temporal lobe structures have been reported. This study examined a sample of children and adolescents with major depressive disorder to investigate anatomical abnormalities in these key medial temporal brain regions. Nineteen children and adolescents with DSM-IV major depression (mean age ± S.D. = 13.0 ± 2.4 years; 10 unmedicated) and 24 healthy comparison subjects (mean age ± S.D. = 13.9 ± 2.9 years) were studied using a 1.5 T Philips MRI scanner. We measured hippocampus and amygdala gray matter volumes. MRI structural volumes were compared using analysis of covariance with age and total brain volumes as covariates. Pediatric depressed patients had significantly smaller left hippocampal gray matter volumes compared to healthy controls (1.89 ± 0.16 cm 3 versus 1.99 ± 0.18 cm 3, respectively; F = 5.0, d.f. = 1/39, p = 0.03; effect size: η 2 p = 0.11). Unmedicated depressed patients showed a trend towards smaller left hippocampal volumes compared to medicated patients and healthy subjects (F = 2.8, d.f. = 2/38, p = 0.07; effect size: η 2 p = 0.13). There were no statistically significant differences in mean volumes for left or right amygdala. Smaller left hippocampal volumes in children and adolescents with MDD are in agreement with findings from adult studies and suggest that such abnormalities are present early in the course of the illness. Amygdala volumes are not abnormal in this age group. Smaller hippocampal volumes may be related to an abnormal developmental process or HPA axis dysfunction Published by Elsevier Ireland Ltd. Keywords: Major depressive disorder; Children; Adolescents; Hippocampus; Amygdala; MRI Major depressive disorder (MDD) is a highly prevalent mental illness, occurring in 2.5% of children and 8.3% of adolescents in the USA [24]. Children and adolescents with MDD present with different clinical features compared to adults, such as high comorbidity with anxiety disorders [17]. Even though pediatric MDD has unique clinical presentation, few studies have investigated the pathophysiology of early-onset MDD or compared it with what is known about adults. Corresponding author at: CERT-BD, Department of Psychiatry, University of North Carolina School of Medicine, Neuroscience Hospital, CB#7160, Chapel Hill, NC , USA. Tel.: ; fax: addresses: jsoares@med.unc.edu, soares@uthscsa.edu (J.C. Soares). Magnetic resonance imaging (MRI) is a non-invasive tool that has been used to investigate the anatomy of the central nervous system and the pathophysiology of mood disorders. The most consistent structural MRI finding in adults with unipolar depression is hippocampal atrophy [10], while the amygdala seems to be preserved [7,9,37]. Smaller hippocampal volumes also have been demonstrated in MDD children and adolescents [26] but exceptions have been noted [34]. Smaller amygdala volumes in MDD children and adolescents compared to healthy subjects have also been reported [34]. Another study found larger amygdala/hippocampal volume ratios in pediatric MDD patients compared to healthy subjects, but no significant differences were found in either amygdala or hippocampal volumes examined /$ see front matter 2007 Published by Elsevier Ireland Ltd. doi: /j.neulet

2 S.C. Caetano et al. / Neuroscience Letters 427 (2007) individually [27]. These studies point to both similarities and differences between adult and pediatric MDD patients related to medial temporal lobe structures. In this study, we evaluated the hippocampus and amygdala volumes of MDD children and adolescents and compared them to healthy subjects. We tested the hypothesis that, as in adults, the hippocampal volumes of MDD children and adolescents are smaller than those of healthy age matched subjects, while amygdala volumes are similar. This would indicate that hippocampal abnormalities are already present early in illness course in children who suffered from unipolar depression. The sample comprised 19 MDD pediatric outpatients and 24 age and sex matched healthy subjects. All participants were between 8 and 17 years old, were magnetic resonance compatible and did not have any serious non-psychiatric medical problems. All subjects and their parents or legal guardians were interviewed with the Schedule for Affective Disorders and Schizophrenia for school-age children-present and lifetime version (K-SADS-PL) interview [21]. We assessed the paternal socioeconomic status (SES) with the Hollingshead Socioeconomic status scale [18], Global Assessment of Functioning Scale (GAF), puberty status with the Pubertal Development Scale Petersen scale [32], the severity of depression with the Children s Depression Rating Scale, revised (CDRS) [33] and laterality with Oldfield s scale [31]. MDD children and adolescents had to meet DSM-IV diagnostic criteria for MDD, and they were excluded if they met lifetime diagnostic criteria for bipolar disorder, psychotic disorders, developmental disorders, substance abuse/dependence, eating disorders, tics, or mental retardation. Healthy subjects were excluded if they had history of or current psychiatric disorder, or history of any Axis I psychiatric disorder in first-degree relatives. The local IRB approved the study protocol. The parents or legal guardians of all subjects gave written informed consent, and all children gave their assent to participate after the study was explained. Participants underwent an MRI scan of the entire brain on a 1.5 T Philips Intera scanner at the South Texas Veterans Health Care System, Audie Murphy Division. We used a sagittal scout series to check each subject s head position and image quality and to find a midline sagittal image. We then obtained a T-1 weighted fast field echo sequence (3D T1-FFE) with repetition time (TR) = 25 ms, echo time (TE) = 5 ms, field of view (FOV) = 240 mm 220 mm, slice thickness = 1.0 mm, gap = 0, number of excitations (NEX) = 2, and matrix size = We analyzed MRI images using the Brains2 software (University of Iowa, Iowa) [1]. All ROIs were traced on the high resolution T1 weighted image, with subjects diagnoses and identities masked. The raters achieved reliability equal to or greater than 0.93 for all ROIs. The anatomical landmarks of each ROI are briefly described below. Hippocampus. Beginning with the slice where the thalamus connected to the superior colliculus, we traced the hippocampus until one slice before the mammilary bodies became visible. The superior border was the corona radiata, and the inferior border was the parahippocampal gyrus. The lateral border was the inferior horn of the lateral ventricle. Further details are provided in [6]. Amygdala. We traced from the slice where the mamillary bodies were first visualized until we could no longer distinguish the amygdala gray matter from the surrounding temporal lobe. The superior border was the temporal lobe white matter, and the inferior border was the parahippocampal gyrus [6]. Total brain. We traced the whole brain, excluding the cerebellum, dura-mater and sinuses. We used version 12 of SPSS for Windows software (SPSS Inc., Chicago, IL) to perform the statistical analysis. We compared the MDD and healthy groups with respect to socio-demographic and some clinical variables using Fisher s exact test for categorical variables and Student s t-test or Mann Whitney U-test for variables measured on an interval or ordinal scale. We used analysis of covariance (ANCOVA) with age and total brain volumes (TBV) as covariates to test our primary hypothesis that MDD subjects have smaller hippocampal volumes than healthy comparison subjects. Left and right hippocampal volumes were analyzed separately. Our secondary hypothesis concerning the amygdala volumes was tested using similar ANCOVA models. In tertiary exploratory analysis, we also compared MDD patients on and off medication and healthy subjects using ANCOVA with age and TBV as covariates. We used Pearson s correlation coefficients to evaluate the association between age and ROI volumes, and Spearman s correlation coefficients for the association between ROI volumes and clinical variables (age of onset, length of illness, CDRS). We adopted a two-tailed significance level of p < 0.05 for all hypothesis testing. We report partial eta squared (η 2 p ) as an index of effect size. The demographic and clinical characteristics of the pediatric MDD patients and healthy subjects are displayed in Table 1. Patients and healthy subjects did not differ significantly regarding age (t = 1.1, d.f. = 41, p = 0.28), gender (Fisher s Exact test, p = 0.4), education (Mann Whitney Z = 1.2, p = 0.2), paternal SES (Mann Whitney Z = 0.3, p = 0.8) and pubertal status (Mann Whitney Z = 0.5, p = 0.6). MDD children and adolescents had significantly lower GAF scores compared to healthy subjects (t = 11.0, d.f. = 41, p = 0.001). The ROI volumes for MDD patients and healthy comparison subjects are displayed in Table 2. Our primary hypothesis received partial support as MDD patients had significantly smaller left hippocampal gray matter volumes compared to healthy subjects. Right hippocampus volumes were not significantly different. Furthermore, as predicted, there were no significant differences between patients and healthy subjects in amygdala volumes (p > 0.05) (Table 2). We included patients who were on psychotropic medication [n = 9; antidepressants: escitalopram (n = 2), sertraline (n = 2), bupropione (n = 1), paroxetine (n = 1); and/or stimulants: methylphenidate (n = 3), d-amphetamine (n = 2), and atomoxetine (n = 1)], and patients who were not taking psychotropic medication (n = 10; 6 were medication-naïve and 4 had not taken medication for 6 months or more prior to study entry).

3 144 S.C. Caetano et al. / Neuroscience Letters 427 (2007) Table 1 Demographic and clinical characteristics of MDD children and adolescents and healthy subjects MDD patients (n = 19) Healthy subjects (n = 24) Age (mean ± S.D.) (years) 13.0 ± ± 2.9 Range (years) Gender: number male (%) 13 (68.4%) 13 (54.2%) Race: number (%) Caucasian 10 (52.6%) 02 (8.3%) Hispanics 7 (36.8%) 18 (75.0%) African-Americans 1 (5.3%) 0 Asians 1 (5.3%) 04 (16.7%) Education: mean ± S.D. (years) 7.0 ± ± 3.0 SES: mean ± S.D (± 12.4) 43.6 (± 13.1) Puberty degree: number (%) Pre-pubertal 2 (10.5%) 2 (8.3%) Beginning 5 (26.3%) 7 (29.2%) Middle 5 (26.3%) 3 (12.5%) Advanced 6 (31.6%) 10 (41.7%) Post-pubertal 1 (5.3%) 2 (8.3%) Laterality: number right-handed (%) 18 (95%) 22 (92%) Mean GAF ± S.D ± ± 10.3 Mood disorders in first degree-relative: number (%) 17 (89.5%) 0 Mean age of onset ± S.D. (years) 10.3 ± 2.3 Mean duration of illness ± S.D. (months) 27.4 ± 12.2 Mean number of episodes ± S.D. 1.5 ± 1.4 Mean CDRS ± S.D. 44 ± 15.8 There was a non-significant trend towards smaller left hippocampal gray matter volumes in psychotropic medication-free patients (1.84 ± 0.15 cm 3 ; n = 10; mean age ± S.D. = 14.3 ± 1.7, 5 males) compared to patients on psychotropic medication (1.93 ± 0.16 cm 3 ; n = 9; mean age ± S.D. = 11.8 ± 2.3, 8 males) and to healthy subjects (1.99 ± 0.18 cm 3 )(F = 2.8, d.f. = 2/38, p = 0.07; ANCOVA with age and TBV as covariates; effect size: η 2 p = 0.13). Sixteen MDD patients had co-morbid disorders: attention deficit hyperactivity disorder (ADHD, n = 7), generalized anxiety (n = 7), social phobia (n = 2), panic disorder (n = 1) and oppositional defiant disorder (n = 1). Our MDD sample did not present psychotic features. As an exploratory analysis, we compared MDD patients with (n = 12; mean age ± S.D. = 12.6 ± 2.3 years, 9 males) and without comorbid anxiety disorders (n = 7; mean age ± S.D. = 13.8 ± 2.4 years, 4 males). There was a trend towards smaller right amygdala volumes in MDD patients with comorbid anxiety disorders compared with MDD without anxiety disorders (mean volume ± S.D. = 1.51 cm 3 ± 0.25 versus 1.66 cm 3 ± 0.19; F = 3.8, d.f. = 1/15, p = 0.07). MDD patients showed a significant inverse correlations between their current age and total brain volume (r = 0.50, p = 0.03). Healthy subjects did not show a significant correlation between age and any temporal lobe structures. Among MDD patients, CDRS was inversely correlated with total brain volume (rho = 0.59, p = 0.008) and also with right amygdala gray matter volume (rho = 0.55, p = 0.02). Age of onset was inversely correlated with total brain volume (rho = 0.47, p = 0.04). The finding of smaller left hippocampal volumes in children and adolescents with MDD compared to healthy subjects is in agreement with studies conducted on adults [10] and other studies of children and adolescents suffering from MDD [26]. MacMillan et al. [27] reported hippocampal volumes were about 3% smaller in MDD children and adolescents, but these findings did not reach statistical significance when age and intracranial Table 2 Mean gray matter volumes ± S.D. (cm 3 ) for each region of interest in pediatric MDD patients and healthy subjects Region of interest MDD patients (n = 19) Healthy subjects (n = 24) Statistical analysis (d.f. = 1/39) Effect size (η p ) F p Left hippocampus 1.89 ± ± Right hippocampus 1.97 ± ± Left amygdala 1.46 ± ± Right amygdala 1.56 ± ± Total brain ± ± a 0.01 a ANCOVA with age as covariate, d.f. = 1/40.

4 S.C. Caetano et al. / Neuroscience Letters 427 (2007) volume were controlled. Rosso et al. [34] in a predominantly female MDD child sample (85%) found smaller amygdala but not hippocampal volumes. In a meta-analysis, the magnitude of hippocampal atrophy in MDD adults ranged from 8 to 19% [10]. In a post-mortem study of adults with MDD, reduced hippocampal pyramidal neuron soma size and increased packing density of glia and pyramidal neurons were reported [40]. They suggested that these may be a result of reduced neuropil in the hippocampus of MDD patients. Indeed, smaller left hippocampal volumes seem to be present in adults, children and adolescents suffering from MDD. Hypercortisolemia induced neurotoxicity is believed to account for the hippocampal atrophy in MDD. Approximately, 30 50% of MDD adults present higher plasma cortisol levels than healthy subjects. Even when the cortisol levels are within the normal range, the combined dexamethasone/corticotropin releasing hormone test and the dexamethasone suppression tests are abnormal, showing an hypothalamic pituitary adrenal (HPA) axis dysfunction [41]. Cortisol abnormalities in teenagers with MDD have not been consistently demonstrated, with reports of normal [11] or higher cortisol levels [25]. Even though we did not have cortisol levels for our sample, this neuroendocrine theory may explain the 5% difference in left hippocampal volumes of MDD and healthy children and adolescents [28]. Hippocampal atrophy occurs during a depressive episode and might show some recovery during remission. In animal models, acute stress leads to reduction of the hippocampal dendrites of area CA3, and when the stress is eliminated the hippocampus recovers [20]. Our modest result may be explained in part by the presence of patients in remission (n = 6). It is unclear whether MDD adults show recovery from hippocampal atrophy during remission. Frodl et al. reported that MDD adults who do not achieve remission show bilateral hippocampal atrophy that persists through 1 year follow-up compared to patients in remission [14]. However, MDD adult patients in remission have been found to have smaller left hippocampal volumes compared to healthy subjects [7,36]. An inverse correlation between length of illness and hippocampal volumes has been found [3,38]. Noteworthy, Sheline et al. [36] demonstrated that hippocampal volume was indirectly associated with length of time without treatment. In our sample, we did not find any association between hippocampal volumes and length of illness probably because the mean duration of MDD was less than 1 year. Pediatric MDD patients such as our sample with high genetic loading may be susceptible to hippocampal reduction. At least 30% of healthy first degree relatives of MDD patients have HPA axis dysregulation [22]. They suggested that patients with MDD and positive family history could have abnormal corticosteroid receptor function. Healthy children whose mothers have MDD showed higher cortisol levels in response to a psychosocial stressor. We should note that the majority of our MDD sample had a first degree relative with mood disorders, whereas, we selected healthy subjects without any family history of mood disorders. Genetics may determine hippocampal volume [15] and 60% of the individual variance of the cortisol level during the day [2]. An additional concern is that early onset MDD patients tend to have a high frequency of recurrence, which might limit the opportunities for hippocampal volume recovery. Other mechanisms proposed to explain the hippocampal reduction include stress related reduction in neurotrophic factors, decreased neurogenesis and glial loss resulting in increased vulnerability to glutamate neurotoxicity [35]. Antidepressants probably act directly on the glucocorticoid receptors, increasing their expression, improving their function and promoting their nuclear translocation [19]. Preclinical studies have demonstrated the neuroprotective effects of antidepressants by increasing brain derived neurotrophic factor (BDNF) [30]. Specifically, the selective serotonin reuptake inhibitors (SSRIs) were found to increase the dendritic tree [13]. After 1-year treatment with the SSRI paroxetine, patients with post-traumatic stress disorder patients had a 5% increase in hippocampal volumes [8]. These findings may explain our trend towards larger left hippocampal volumes in patients taking medication, mostly SSRIs, compared to patients off psychotropic medication. Our finding of preservation of amygdala volumes in pediatric MDD is in line with studies on adults [7,9,37] and other studies of children and adolescents with MDD [27]. However, Rosso et al. [34] found smaller amygdala volumes in MDD children and adolescents compared with healthy subjects. Amygdala reduction has been consistently reported in bipolar children and adolescents compared to healthy subjects [4]. Considering that up to 49% of all MDD patients will present a manic episode [12], amygdala abnormalities in this group may be related to future development of bipolar disorder. Reduction of amygdala volumes could also be attributed to anxiety as suggested by previous studies of children and adolescents with anxiety disorders [29] and by our finding of a trend towards smaller right amygdala volumes in MDD patients with anxiety compared to MDD without anxiety. On the other hand, the amygdala has excitatory connections to the hippocampus, and in post-mortem samples of MDD patients, the glial density and the glia/neuron ratio of the amygdala are reduced [5],so we cannot completely rule out amygdala volume abnormalities in MDD. Steingard et al. [39] reported smaller total brain volumes in MDD adolescents compared with healthy subjects. No changes in total brain volumes of adults with MDD were reported [7]. However, an inverse correlation between length of MDD and total brain volumes was reported [23]. In our sample, the total brain volumes of the MDD subjects were not different from those of the healthy comparisons. One interpretation of our findings is that these subjects were too early in the course of illness to show smaller total brain volumes. We did find an inverse correlation between age and total brain volume in MDD children and adolescents. We also found that total brain volumes were inversely associated with CDRS scores and age of onset, suggesting a possible link between loss of total brain volume and the clinical presentation of MDD. Adolescence is a vital time for cortical pruning [16] therefore total brain volume merits further research to determine whether brain development is disrupted by MDD.

5 146 S.C. Caetano et al. / Neuroscience Letters 427 (2007) A limitation to this study is the high prevalence of comorbid Axis I DSM-IV diagnoses; however, our sample is representative of childhood MDD samples in this respect. Children and adolescents suffering from MDD have up to a 70% prevalence of comorbid Axis I DSM-IV psychiatric disorders, specifically anxiety disorders (33 65%) and conduct disorder (20 37%) [17]. High rates of anxiety disorders in grandchildren of two generations with MDD were demonstrated. In relatives of grandparents with MDD, an increased risk of anxiety disorders (relative risk = 5.2) and other psychiatric disorders (relative risk = 5.5) was observed in children (mean age = 12 years) whose parent had MDD, when compared to children whose parents did not have MDD [42]. These authors suggested that anxiety disorders could be the first presentation of psychopathology in children who will develop MDD. Studies comparing pediatric anxiety disorders patients versus MDD patients with and without comorbid anxiety disorders are needed to clarify the relationship between each disorder and temporal lobe pathology. Another limitation is the use of the mamillary body as an external reference to hippocampus posterior boundary and amygdala s anterior boundary because part of the amygdala could be included in the hippocampus measurement. However, our tracing method is similar to others [34]. In conclusion, MDD children and adolescents present smaller left hippocampal volumes compared to healthy children while amygdala volumes are not abnormal. Our preliminary results on the SSRI treatment may indicate that these volume abnormalities can be ameliorated by earlier treatment. Acknowledgments This work was partly supported by M01-RR-01346, K23- MH , MH , MH 69774, RR , the Krus Endowed Chair in Psychiatry (UTHSCSA), and CAPES Foundation (Brazil). This work was partly presented at the 60th Society of Biological Psychiatry meeting in Atlanta, USA, References [1] N.C. Andreasen, G. Cohen, G. Harris, T. Cizadlo, J. Parkkinen, K. Rezai, V.W. Swayze II, Image processing for the study of brain structure and function: problems and programs, J. Neuropsychiatr. Clin. Neurosci. 4 (1992) [2] M. Bartels, E.J. de Geus, C. Kirschbaum, F. Sluyter, D.I. Boomsma, Heritability of daytime cortisol levels in children, Behav. Genet. 33 (2003) [3] S. Bell-McGinty, M.A. Butters, C.C. Meltzer, P.J. Greer, C.F. Reynolds 3rd, J.T. Becker, Brain morphometric abnormalities in geriatric depression: long-term neurobiological effects of illness duration, Am. J. Psychiatr. 159 (2002) [4] H.P. Blumberg, J. Kaufman, A. Martin, R. Whiteman, J.H. Zhang, J.C. Gore, D.S. Charney, J.H. Krystal, B.S. Peterson, Amygdala and hippocampal volumes in adolescents and adults with bipolar disorder, Arch. Gen. Psychiatr. 60 (2003) [5] M.P. Bowley, W.C. Drevets, D. Ongur, J.L. Price, Low glial numbers in the amygdala in major depressive disorder, Biol. Psychiatr. 52 (2002) [6] P. Brambilla, K. Harenski, M. Nicoletti, R.B. Sassi, A.G. Mallinger, E. Frank, D.J. Kupfer, M.S. Keshavan, J.C. Soares, MRI investigation of temporal lobe structures in bipolar patients, J. Psychiatr. Res. 37 (2003) [7] J.D. Bremner, M. Narayan, E.R. Anderson, L.H. Staib, H.L. Miller, D.S. Charney, Hippocampal volume reduction in major depression, Am. J. Psychiatr. 157 (2000) [8] J.D. Bremner, E. Vermetten, Neuroanatomical changes associated with pharmacotherapy in posttraumatic stress disorder, Ann. NY Acad. Sci (2004) [9] S.C. Caetano, J.P. Hatch, P. Brambilla, R.B. Sassi, M. Nicoletti, A.G. Mallinger, E. Frank, D.J. Kupfer, M.S. Keshavan, J.C. Soares, Anatomical MRI study of hippocampus and amygdala in patients with current and remitted major depression, Psychiatr. Res. 132 (2004) [10] S. Campbell, M. Marriott, C. Nahmias, G.M. MacQueen, Lower hippocampal volume in patients suffering from depression: a meta-analysis, Am. J. Psychiatr. 161 (2004) [11] R. Dahl, J. Puig-Antich, N. Ryan, B. Nelson, H. Novacenko, J. Twomey, D. Williamson, R. Goetz, P.J. Ambrosini, Cortisol secretion in adolescents with major depressive disorder, Acta Psychiatr. Scand. 80 (1989) [12] M.P. DelBello, G.A. Carlson, M. Tohen, E.J. Bromet, M. Schwiers, S.M. Strakowski, Rates and predictors of developing a manic or hypomanic episode 1 to 2 years following a first hospitalization for major depression with psychotic features, J. Child Adolesc. Psychopharmacol. 13 (2003) [13] R.S. Duman, G.R. Heninger, E.J. Nestler, A molecular and cellular theory of depression, Arch. Gen. Psychiatr. 54 (1997) [14] T. Frodl, E.M. Meisenzahl, T. Zetzsche, T. Hohne, S. Banac, C. Schorr, M. Jager, G. Leinsinger, R. Bottlender, M. Reiser, H.J. Moller, Hippocampal and amygdala changes in patients with major depressive disorder and healthy controls during a 1-year follow-up, J. Clin. Psychiatr. 65 (2004) [15] T. Frodl, E.M. Meisenzahl, P. Zill, T. Baghai, D. Rujescu, G. Leinsinger, R. Bottlender, C. Schule, P. Zwanzger, R.R. Engel, R. Rupprecht, B. Bondy, M. Reiser, H.J. Moller, Reduced hippocampal volumes associated with the long variant of the serotonin transporter polymorphism in major depression, Arch. Gen. Psychiatr. 61 (2004) [16] N. Gogtay, J.N. Giedd, L. Lusk, K.M. Hayashi, D. Greenstein, A.C. Vaituzis, T.F. Nugent III, D.H. Herman, L.S. Clasen, A.W. Toga, J.L. Rapoport, P.M. Thompson, Dynamic mapping of human cortical development during childhood through early adulthood, Proc. Natl. Acad. Sci. U.S.A. 101 (2004) [17] I.M. Goodyer, J. Herbert, S.M. Secher, J. Pearson, Short-term outcome of major depression: I. Comorbidity and severity at presentation as predictors of persistent disorder, J. Am. Acad. Child Adolesc. Psychiatr. 36 (1997) [18] A. Hollingshead, Two-Factor Index of Social Position, Department of Sociology, Yale University, New Haven, CT, [19] F. Holsboer, The corticosteroid receptor hypothesis of depression, Neuropsychopharmacology 23 (2000) [20] C. Hoschl, T. Hajek, Hippocampal damage mediated by corticosteroids a neuropsychiatric research challenge, Eur. Arch. Psychiatr. Clin. Neurosci. 251 (Suppl. 2) (2001) II81 II88. [21] J. Kaufman, B. Birmaher, D. Brent, U. Rao, N. Ryan, Diagnostic Interview Kiddie-Sads-Present and Lifetime Version (K-SADS-PL), [22] J.C. Krieg, C.J. Lauer, W. Schreiber, S. Modell, F. Holsboer, Neuroendocrine, polysomnographic and psychometric observations in healthy subjects at high familial risk for affective disorders: the current state of the Munich vulnerability study, J. Affect Disord. 62 (2001) [23] I.K. Lampe, H.E. Hulshoff Pol, J. Janssen, H.G. Schnack, R.S. Kahn, T.J. Heeren, Association of depression duration with reduction of global cerebral gray matter volume in female patients with recurrent major depressive disorder, Am. J. Psychiatr. 160 (2003) [24] P.M. Lewinsohn, G.N. Clarke, J.R. Seeley, P. Rohde, Major depression in community adolescents: age at onset, episode duration, and time to recurrence, J. Am. Acad. Child Adolesc. Psychiatr. 33 (1994) [25] J.L. Luby, A. Heffelfinger, C. Mrakotsky, K. Brown, M. Hessler, E. Spitznagel, Alterations in stress cortisol reactivity in depressed preschoolers

6 S.C. Caetano et al. / Neuroscience Letters 427 (2007) relative to psychiatric and no-disorder comparison groups, Arch. Gen. Psychiatr. 60 (2003) [26] F.P. MacMaster, V. Kusumakar, Hippocampal volume in early onset depression, BMC Med. 2 (2004) 2. [27] S. MacMillan, P.R. Szeszko, G.J. Moore, R. Madden, E. Lorch, J. Ivey, S.P. Banerjee, D.R. Rosenberg, Increased amygdala:hippocampal volume ratios associated with severity of anxiety in pediatric major depression, J. Child Adolesc. Psychopharmacol. 13 (2003) [28] G.M. MacQueen, S. Campbell, B.S. McEwen, K. Macdonald, S. Amano, R.T. Joffe, C. Nahmias, L.T. Young, Course of illness, hippocampal function, and hippocampal volume in major depression, Proc. Natl. Acad. Sci. U.S.A. 100 (2003) [29] M.P. Milham, A.C. Nugent, W.C. Drevets, D.P. Dickstein, E. Leibenluft, M. Ernst, D. Charney, D.S. Pine, Selective reduction in amygdala volume in pediatric anxiety disorders: a voxel-based morphometry investigation, Biol. Psychiatr. 57 (2005) [30] M. Nibuya, E.J. Nestler, R.S. Duman, Chronic antidepressant administration increases the expression of camp response element binding protein (CREB) in rat hippocampus, J. Neurosci. 16 (1996) [31] R. Oldfield, The assessment and analysis of handedness: the Edinburgh inventory, Neuropsychologia 9 (1971) [32] A.C. Pertersen, L. Crockett, M. Richards, A. Boxer, A self-report measure of pubertal status: reliability validity, and initial norms, J. Youth Adolesc. 17 (1988) [33] E.O. Poznanski, S.C. Cook, B.J. Carroll, A depression rating scale for children, Pediatrics 64 (1979) [34] I.M. Rosso, C.M. Cintron, R.J. Steingard, P.F. Renshaw, A.D. Young, D.A. Yurgelun-Todd, Amygdala and hippocampus volumes in pediatric major depression, Biol. Psychiatr. 57 (2005) [35] Y.I. Sheline, Neuroimaging studies of mood disorder effects on the brain, Biol. Psychiatr. 54 (2003) [36] Y.I. Sheline, M.H. Gado, H.C. Kraemer, Untreated depression and hippocampal volume loss, Am. J. Psychiatr. 160 (2003) [37] Y.I. Sheline, M.H. Gado, J.L. Price, Amygdala core nuclei volumes are decreased in recurrent major depression, Neuroreport 9 (1998) [38] Y.I. Sheline, M. Sanghavi, M.A. Mintun, M.H. Gado, Depression duration but not age predicts hippocampal volume loss in medically healthy women with recurrent major depression, J. Neurosci. 19 (1999) [39] R.J. Steingard, P.F. Renshaw, J. Hennen, M. Lenox, C.B. Cintron, A.D. Young, D.F. Connor, T.H. Au, D.A. Yurgelun-Todd, Smaller frontal lobe white matter volumes in depressed adolescents, Biol. Psychiatr. 52 (2002) [40] C.A. Stockmeier, G.J. Mahajan, L.C. Konick, J.C. Overholser, G.J. Jurjus, H.Y. Meltzer, H.B. Uylings, L. Friedman, G. Rajkowska, Cellular changes in the postmortem hippocampus in major depression, Biol. Psychiatr. 56 (2004) [41] S. Watson, P. Gallagher, D. Del-Estal, A. Hearn, I.N. Ferrier, A.H. Young, Hypothalamic pituitary adrenal axis function in patients with chronic depression, Psychol. Med. 32 (2002) [42] M.M. Weissman, P. Wickramaratne, Y. Nomura, V. Warner, H. Verdeli, D.J. Pilowsky, C. Grillon, G. Bruder, Families at high and low risk for depression: a 3-generation study, Arch. Gen. Psychiatr. 62 (2005)