Deformations of amygdala morphology in familial pediatric bipolar disorder

Amygdala structure and function in paediatric bipolar disorder and high-risk youth: A systematic review of magnetic resonance imaging findings

OBJECTIVE

Converging evidence from structural and functional magnetic resonance imaging (MRI) studies points to amygdala alteration as crucial in the development of paediatric bipolar disorder (pBP). The high number of recent studies prompted us to comprehensively evaluate findings. We aimed to systematically review structural and functional MRI studies investigating the amygdala in patients with pBP and in youth at high-risk (HR) for developing pBP.

METHODS

We searched PubMed from any time to 25 September 2020 using: 'amygdala AND (MRI OR magnetic resonance imaging) AND bipolar AND (pediatr* OR child OR children OR childhood OR adolescent OR adolescents OR adolescence OR young OR familial OR at-risk OR sibling* OR offspring OR high risk)'. In this review, we adhered to the PRISMA statement.

RESULTS

Amygdala hyperactivity to emotional stimuli is the most commonly reported finding in youth with pBP and HR compared to healthy peers (HC), whereas findings from structural MRI studies are inconsistent.

CONCLUSIONS

Hyperactivation of the amygdala might be an endophenotype of pBP.

Clinical and neurostructural characteristics among youth with familial and non-familial bipolar disorder: Family history and youth bipolar disorder

BACKGROUND

Bipolar disorder (BD) is highly heritable and often severe, particularly when illness onset occurs early in life. There is limited knowledge regarding the clinical and neurostructural correlates of family history of BD among youth with BD.

METHODS

Clinical characteristics were evaluated in 197 youth with BD, ages 13-20 years, including 87 with familial BD and 110 with non-familial BD. Structural neuroimaging was examined in a subsample of familial BD (n=39), non-familial BD (n=42), and healthy control (HC, n=58) youth. Region of interest (ROI) analyses of anterior cingulate cortex (ACC), inferior frontal gyrus (IFG), and amygdala were complemented by whole-brain vertex-wise analyses.

RESULTS

Youth with familial BD had more family history of other psychiatric disorders, less severe worst manic episode, and less treatment with lithium, selective serotonin reuptake inhibitor (SSRI) antidepressants, and any lifetime psychiatric medications. None of these findings survived after correction for multiple comparisons. There were no significant between-group differences in ROI analyses. In whole-brain analyses, significant differences in cortical thickness were as follows: familial and non-familial BD < HC in left precentral gyrus and right inferior parietal lobe; familial BD < HC in left superior frontal gyrus; non-familial BD < HC in right precentral gyrus.

LIMITATIONS

Relatives did not complete full diagnostic interviews.

CONCLUSIONS

There were relatively few differences in clinical and neurostructural correlates related to family history of BD in youth with BD. Current findings suggest that family history of BD is not a strong contributor to the clinical or neuroimaging phenotypes in youth with BD.

Preliminary study of structural magnetic resonance imaging phenotypes related to genetic variation in Interleukin-1β rs16944 in adolescents with Bipolar Disorder

BACKGROUND

Bipolar disorder (BD), among the most heritable psychiatric conditions, is associated with increased pro-inflammatory blood markers and pro-inflammatory gene expression in post-mortem brain. We therefore examined the effects of pro-inflammatory single nucleotide polymorphism interleukin-1β (IL-1β) rs16944 on brain structure in adolescents with BD and healthy control (HC) adolescents.

METHODS

T₁-weighted 3-T magnetic resonance imaging data were acquired for 38 adolescents with BD and 32 HC adolescents (14-20 years). Using FreeSurfer, a priori regions of interest analyses, examining hippocampus, amygdala, dorsolateral prefrontal cortex (DLPFC), and caudal anterior cingulate cortex, were complemented by exploratory whole-brain vertex-wise analyses. General linear models assessed the association between IL-1β rs16944 and the ROIs, controlling for sex, age, and intracranial volume.

RESULTS

There was an IL-1β rs16944 polymorphism-by-diagnosis interaction effect on the DLPFC; T-carriers with BD had greater surface area compared to non-carriers with BD. Whereas, HC T-carriers had smaller DLPFC volume compared to HC non-carriers. In vertex-wise analyses, similar interactions were evident in a pars triangularis surface area cluster and a lateral occipital cortex volume cluster. Whole-brain analyses also yielded a main effect of IL-1β rs16944 polymorphism, whereby T-carriers had greater lateral occipital cortex surface area and volume.

CONCLUSIONS

The IL-1β rs16944 polymorphism is associated with neurostructural phenotypes in cognitive and visual regions that subserve functions, including facial recognition and response inhibition, which are known to be aberrant in BD. Future studies are warranted to evaluate whether the observed rs16944-related structural differences are relevant to neurocognitive function, functional neuroimaging phenotypes and IL-1β protein levels.

Correlation Between Decreased Amygdala Subnuclei Volumes and Impaired Cognitive Functions in Pediatric Bipolar Disorder

BACKGROUND

The amygdala has been proposed to be involved in the pathophysiology of pediatric and adult bipolar disorder (BD). The goal of this structural magnetic resonance imaging (sMRI) study was to investigate the morphometric characteristics of amygdala subnuclei in patients with pediatric bipolar disorder (PBD) compared to healthy controls (HCs). Simultaneously, we examined correlation between amygdala subnuclei volumes and cognitive dysfunction.

MATERIALS AND METHODS

We assessed 40 adolescent outpatients, diagnosed with manic or euthymic PBD according to the DSM-5 criteria for BD and 19 HCs. Cognitive functions were evaluated using a Stroop color-word test (SCWT), trail making test (TMT), visual reproduction immediate recall subtest (VR I), and digit span subtest (DST). Amygdala and its subnuclei structures were automated segmented using FreeSurfer software and the volumes of them were compared between groups and correlation with clinical and cognitive outcomes was conducted.

RESULTS

Manic patients exhibited significantly decreased volumes in the bilateral whole amygdala and its basal nucleus, cortico-amygdaloid transition (CAT), and accessory basal nucleus (ABN) compared with HCs. Euthymic patients had decreased volume in the bilateral ABN and left CAT. In addition, we found significant positive associations between VR I scores and the right whole amygdala and its bilateral basal, right lateral, and ABN volumes in the manic group.

CONCLUSION

These findings support previous reports of smaller amygdala volumes and cognitive dysfunctions in PBD, and further mapping abnormalities to specific amygdala subnuclei. Correlation between basolateral volume and VR I of PBD may expand our understanding of neural abnormalities that could be targeted by treatment.

Cortical Volume and Thickness Across Bipolar Disorder Subtypes in Adolescents: A Preliminary Study

OBJECTIVES

Neuroimaging studies of adults with bipolar disorder (BD) have identified several BD subtype distinctions, including greater deficits in prefrontal gray matter volumes in BD-I (bipolar I disorder) compared to BD-II (bipolar II disorder). We sought to investigate BD subtype differences in brain structure among adolescents and young adults.

METHODS

Forty-four youth with BD (14 BD-I, 16 BD-II, and 14 BD-not otherwise specified [NOS], mean age 17) underwent 3T-MRI and images were analyzed using FreeSurfer software. Cortical volume and thickness were analyzed for region of interest (ROI): ventrolateral prefrontal cortex, ventromedial prefrontal cortex, anterior cingulate cortex (ACC), subgenual cingulate cortex, and amygdala, controlling for age, sex, and total intracranial volume. ROIs were selected as found to be implicated in BD in prior studies. A whole brain vertex-wise exploratory analysis was also performed. Uncorrected results are presented.

RESULTS

There were group differences in ACC thickness (F = 3.88, p = 0.03, η² = 0.173 uncorrected), which was reduced in BD-II in comparison to BD-I (p = 0.027 uncorrected) and BD-NOS (p = 0.019 uncorrected). These results did not survive correction for multiple comparisons and no other group differences were observed. The exploratory vertex-wise analysis found a similar pattern of lower cortical thickness in BD-II in the left and right superior frontal gyrus and left caudal middle frontal gyrus.

CONCLUSIONS

This study found reduced cortical thickness for youth with BD-II, relative to BD-I, in regions associated with cognitive control. Further neurostructural differences between subtypes may emerge later during the course of illness.

Quantitative analysis of the amygdala, thalamus and hippocampus on magnetic resonance images in paediatric bipolar disorders and compared with the children of bipolar parents and healthy control

MR imaging studies in paediatric bipolar disorder have particularly focused on the amygdala and hippocampus, subcortical structures, and to a lesser extent on the thalamus. The purpose of this study was to perform structural analysis of the regions of interest (ROI) associated with mood regulation. In this study 18 children (between the ages of 12-18) were matched according to their age and sex and were divided into three groups. These were: a paediatric bipolar disorder group, risk group and a healthy control group. The structured diagnostic interviews were performed with children and their parents. T1 weighted MR images in the sagittal plane with a thickness of 1mm were taken from the subjects. Automatic structural brain analysis was performed, and the volume and volume fraction (VF) of the ROIs were obtained. Brain size in the patients with paediatric bipolar disorder (742.4 ± 110.1cm³) was significantly smaller than the healthy control group (880.7 ± 73.8cm³) (p≤0.05). MRI analysis between the paediatric bipolar disorder, risk group and healthy control group revealed no difference between them in terms of amygdala, thalamus or hippocampal volumes. In this study, there was no difference between the volumes of amygdala, thalamus or hippocampus.

Prospective longitudinal voxel-based morphometry study of major depressive disorder in young individuals at high familial risk

BACKGROUND

Previous neuroimaging studies indicate abnormalities in cortico-limbic circuitry in mood disorder. Here we employ prospective longitudinal voxel-based morphometry to examine the trajectory of these abnormalities during early stages of illness development.

METHOD

Unaffected individuals (16-25 years) at high and low familial risk of mood disorder underwent structural brain imaging on two occasions 2 years apart. Further clinical assessment was conducted 2 years after the second scan (time 3). Clinical outcome data at time 3 was used to categorize individuals: (i) healthy controls ('low risk', n = 48); (ii) high-risk individuals who remained well (HR well, n = 53); and (iii) high-risk individuals who developed a major depressive disorder (HR MDD, n = 30). Groups were compared using longitudinal voxel-based morphometry. We also examined whether progress to illness was associated with changes in other potential risk markers (personality traits, symptoms scores and baseline measures of childhood trauma), and whether any changes in brain structure could be indexed using these measures.

RESULTS

Significant decreases in right amygdala grey matter were found in HR MDD v. controls (p = 0.001) and v. HR well (p = 0.005). This structural change was not related to measures of childhood trauma, symptom severity or measures of sub-diagnostic anxiety, neuroticism or extraversion, although cross-sectionally these measures significantly differentiated the groups at baseline.

CONCLUSIONS

These longitudinal findings implicate structural amygdala changes in the neurobiology of mood disorder. They also provide a potential biomarker for risk stratification capturing additional information beyond clinically ascertained measures.

Subregional Shape Alterations in the Amygdala in Patients with Panic Disorder

BACKGROUND

The amygdala has been known to play a pivotal role in mediating fear-related responses including panic attacks. Given the functionally distinct role of the amygdalar subregions, morphometric measurements of the amygdala may point to the pathophysiological mechanisms underlying panic disorder. The current study aimed to determine the global and local morphometric alterations of the amygdala related to panic disorder.

METHODS

Volumetric and surface-based morphometric approach to high-resolution three-dimensional T1-weighted images was used to examine the structural variations of the amygdala, with respect to extent and location, in 23 patients with panic disorder and 31 matched healthy individuals.

RESULTS

There were no significant differences in bilateral amygdalar volumes between patients with panic disorder and healthy individuals despite a trend-level right amygdalar volume reduction related to panic disorder (right, β = -0.23, p = 0.09, Cohen's d = 0.51; left, β = -0.18, p = 0.19, Cohen's d = 0.45). Amygdalar subregions were localized into three groups including the superficial, centromedial, and laterobasal groups based on the cytoarchitectonically defined probability map. Surface-based morphometric analysis revealed shape alterations in the laterobasal and centromedial groups of the right amygdala in patients with panic disorder (false discovery rate corrected p < 0.05).

CONCLUSIONS

The current findings suggest that subregion-specific shape alterations in the right amygdala may be involved in the development and maintenance of panic disorder, which may be attributed to the cause or effects of amygdalar hyperactivation.

Neurodevelopmental origins of bipolar disorder: iPSC models

Bipolar disorder (BP) is a chronic neuropsychiatric condition characterized by pathological fluctuations in mood from mania to depression. Adoption, twin and family studies have consistently identified a significant hereditary component to BP, yet there is no clear genetic event or consistent neuropathology. BP has been suggested to have a developmental origin, although this hypothesis has been difficult to test since there are no viable neurons or glial cells to analyze, and research has relied largely on postmortem brain, behavioral and imaging studies, or has examined proxy tissues including saliva, olfactory epithelium and blood cells. Neurodevelopmental factors, particularly pathways related to nervous system development, cell migration, extracellular matrix, H3K4 methylation, and calcium signaling have been identified in large gene expression and GWAS studies as altered in BP. Recent advances in stem cell biology, particularly the ability to reprogram adult somatic tissues to a pluripotent state, now make it possible to interrogate these pathways in viable cell models. A number of induced pluripotent stem cell (iPSC) lines from BP patient and healthy control (C) individuals have been derived in several laboratories, and their ability to form cortical neurons examined. Early studies suggest differences in activity, calcium signaling, blocks to neuronal differentiation, and changes in neuronal, and possibly glial, lineage specification. Initial observations suggest that differentiation of BP patient-derived neurons to dorsal telencephalic derivatives may be impaired, possibly due to alterations in WNT, Hedgehog or Nodal pathway signaling. These investigations strongly support a developmental contribution to BP and identify novel pathways, mechanisms and opportunities for improved treatments.

Volume and shape analysis of subcortical brain structures and ventricles in euthymic bipolar I disorder

Previous structural magnetic resonance imaging (S-MRI) studies of bipolar disorder have reported variable morphological changes in subcortical brain structures and ventricles. This study aimed to establish trait-related subcortical volumetric and shape abnormalities in a large, homogeneous sample of prospectively confirmed euthymic bipolar I disorder (BD-I) patients (n=60), compared with healthy volunteers (n=60). Participants were individually matched for age and gender. Volume and shape metrics were derived from manually segmented S-MR images for the hippocampus, amygdala, caudate nucleus, and lateral ventricles. Group differences were analysed, controlling for age, gender and intracranial volume. BD-I patients displayed significantly smaller left hippocampal volumes and significantly larger left lateral ventricle volumes compared with controls. Shape analysis revealed an area of contraction in the anterior head and medial border of the left hippocampus, as well as expansion in the right hippocampal tail medially, in patients compared with controls. There were no significant associations between volume or shape variation and lithium status or duration of use. A reduction in the head of the left hippocampus in BD-I patients is interesting, given this region's link to verbal memory. Shape analysis of lateral ventricular changes in patients indicated that these are not regionally specific.

Intrinsic Amygdala Functional Connectivity in Youth With Bipolar I Disorder

OBJECTIVE

Bipolar disorder (BD) commonly begins during adolescence and may continue into adulthood. Studies in adults with BD suggest that disruptions in amygdalar neural circuitry explain the pathophysiology underlying the disorder. Importantly, however, amygdala subregion networks have not yet been examined in youth close to mania onset. The goal of this study was to compare resting state functional connectivity patterns in amygdala subregions in youth with bipolar I disorder with patterns in healthy controls.

METHOD

Centromedial, laterobasal, and superficial amygdala subdivisions were assessed during rest and examined in relation to clinical measures of mania in youth (14-20 years old) with bipolar I disorder who experienced only a single episode of mania (BD; n = 20) and age-matched healthy comparison youth without any personal or family history of DSM-IV Axis I disorders (HC; n = 23).

RESULTS

Relative to HC youth, youth with BD exhibited decreased connectivity between the laterobasal subdivision of the amygdala and the hippocampus and precentral gyrus, and increased connectivity between the laterobasal subdivision and the precuneus. Connectivity between the right laterobasal amygdala and right hippocampus was positively correlated with levels of anxiety in BD but not in HC youth, and connectivity between the right laterobasal amygdala and right precuneus was negatively correlated with insight about bipolar illness.

CONCLUSION

Youth with BD have abnormal amygdala resting state network connections to regions that are critical for emotional processing and self-awareness. Longitudinal studies are needed to determine whether these aberrant patterns in youth with BD can be altered with intervention and can influence the course of disorder.

Morphometry of the amygdala in schizophrenia and psychotic bipolar disorder

Volumetric studies suggest smaller amygdalae in subjects with schizophrenia (SZ) than with bipolar disorder (BP). We use morphometry to identify subregions of amygdala differentially affected in SZ and psychotic BP. Based on template centered population analysis, the shape of the amygdala in psychotic BP differs from SZ (pleft=0.044, pright=0.042). Using a high-field 7 T atlas, the bilateral basolateral, basomedial and centromedial subregions and the right lateral subregion were significantly atrophied in SZ compared to psychotic BP (p<0.02). These results suggest that change in shape of amygdala may represent a morphologic feature distinguishing SZ from psychotic BP.

Aberrant amygdala intrinsic functional connectivity distinguishes youths with bipolar disorder from those with severe mood dysregulation

It remains unclear the degree to which youths with episodic mania (bipolar disorder; BD) vs. those with chronic, severe irritability (severe mood dysregulation, SMD) should be placed in similar or distinct diagnostic groups. Addressing this clinically meaningful question requires greater understanding of the neural alterations underlying both disorders. We evaluated resting state functional magnetic resonance imaging data of 53 youths (14 BD, 20 healthy volunteers (HV), and 19 SMD, ages 9-18.5 years). Seed regions of interest were the bilateral basolateral, superficial and centromedial amygdala, defined using the Juelich probabilistic atlas. We found a significant between-group difference in functional connectivity between the left basolateral amygdala and the medial aspect of the left frontal pole plus the posterior cingulate/precuneus. This finding was driven by hyperconnectivity among BD vs. HV or SMD youths. As with earlier data, these findings suggest that the pathophysiology of BD and SMD may differ.

Early risk factors for adult bipolar disorder in adolescents with mood disorders: a 15-year follow-up of a community sample

BACKGROUND

We aimed to outline the early risk factors for adult bipolar disorder (BPD) in adolescents with mood disorders.

METHODS

Adolescents (16-17 years old) with mood disorders (n = 287; 90 participants with hypomania spectrum episodes and 197 with major depressive disorder [MDD]) were identified from a community sample. Fifteen years later (at 30-33 years of age), mood episodes were assessed (n = 194). The risk of developing BPD (n = 22), compared with MDD (n = 104) or no mood episodes in adulthood (n = 68), was estimated via logistic regression. Adolescent mood symptoms, non-mood disorders, and family characteristics were assessed as potential risk factors.

RESULTS

Among the adolescents with mood disorders, a family history of BPD was the strongest predictor of developing BPD compared with having no mood episodes in adulthood (OR = 5.94; 95% CI = 1.11-31.73), whereas disruptive disorders significantly increased the risk of developing BPD compared with developing MDD (OR = 2.94; CI = 1.06-8.12). The risk that adolescents with MDD would develop adult BPD, versus having no mood episodes in adulthood, was elevated among those with an early disruptive disorder (OR = 3.62; CI = 1.09-12.07) or multiple somatic symptoms (OR = 6.60; CI = 1.70-25.67). Only disruptive disorders significantly predicted adult BPD among adolescents with MDD versus continued MDD in adulthood (OR = 3.59; CI = 1.17-10.97). Only a few adolescents with hypomania spectrum episodes continued to have BPD as adults, and anxiety disorders appeared to increase this risk.

CONCLUSIONS

Although most of the identified potential risk factors are likely general predictors of continued mood disorders, disruptive disorders emerged as specific predictors of developing adult BPD among adolescents with MDD.