The reproducibility of grey matter volume differences in psychiatric disorders
Presented During:
Wednesday, June 26, 2024: 11:30 AM - 12:45 PM
COEX
Room:
Grand Ballroom 101-102
Poster No:
549
Submission Type:
Abstract Submission
Authors:
Trang Cao1, James Pang1, Ashlea Segal2, Sidhant Chopra2, Mehul Gajwani3, Alex Fornito1
Institutions:
1Monash University, Clayton, Victoria, 2Yale University, New Haven, CT, 3Monash University, Clayton, VIC
First Author:
Co-Author(s):
Introduction:
Voxel-based morphometry (VBM) [1] has been used extensively to study anatomical differences between people with psychiatric illness and healthy controls. However, the results of these studies have often been difficult to replicate [2]–[6], which may be driven in part by the well-known clinical heterogeneity within psychiatric disorders. Despite this heterogeneity, a fundamental assumption in psychiatric neuroimaging is that each disorder is associated with some core neural phenotype that should be replicable and consistent across different samples and study locations. If such a consistent phenotype cannot be identified, there may be questionable value in ongoing attempts to examine group differences in small, individual studies. Here, we investigated the degree to which five psychiatric disorders––Autism Spectrum Disorder (ASD), Bipolar Disorder (BD), Mood Disorder (MDD), Schizoaffective (SCA), and Schizophrenia (SCZ)––show consistent neuroanatomical phenotypes by examining correlations between disorder- and site-specific maps of grey matter volume alterations.
Methods:
We used 19 neuroimaging datasets of T1-weighted images acquired at 31 different sites across the five disorders. The final sample consisted of 1595 healthy controls (HC) and 1664 cases (ASD, N=265; BD, N=222; MDD, N=248; SCA, N=187; SCZ, N=742) taken from a larger pool of individuals after performing additional quality control checks. We only considered data from adult samples (aged 18-60 years) and sites with at least 20 individuals in each control and patient group.
We estimated case-control differences in grey matter volume (GMV) separately for each site using VBM, implemented via the Computational Anatomy Toolbox [7]. We then created a common mask from all sites and adjusted for site-specific effects using the data harmonization method, ComBat [8]. We ran general linear models, taking into account total intracranial volume (TIV), sex, and age as covariates, and estimated t-statistics within each brain mask to quantify the magnitude of voxel-level case-control differences in GMV. To examine the consistency of the spatial patterns of GMV differences, we then calculate Pearson correlation (r) of the resulting t-maps between each pair of sites separately for each disorder. To investigate whether site differences in sample characteristics influenced consistency estimates, we used Mantel tests to quantify the similarity between site-by-site correlation matrices of GMV differences and matrices corresponding to differences in site-specific properties, including the number of participants, age, sex, age of illness onset, illness duration, medication exposure, scanner manufacturer, and voxel resolution.
We estimated case-control differences in grey matter volume (GMV) separately for each site using VBM, implemented via the Computational Anatomy Toolbox [7]. We then created a common mask from all sites and adjusted for site-specific effects using the data harmonization method, ComBat [8]. We ran general linear models, taking into account total intracranial volume (TIV), sex, and age as covariates, and estimated t-statistics within each brain mask to quantify the magnitude of voxel-level case-control differences in GMV. To examine the consistency of the spatial patterns of GMV differences, we then calculate Pearson correlation (r) of the resulting t-maps between each pair of sites separately for each disorder. To investigate whether site differences in sample characteristics influenced consistency estimates, we used Mantel tests to quantify the similarity between site-by-site correlation matrices of GMV differences and matrices corresponding to differences in site-specific properties, including the number of participants, age, sex, age of illness onset, illness duration, medication exposure, scanner manufacturer, and voxel resolution.
Results:
Statistical t-maps quantifying GMV reductions in ASD, BD, and MDD show low consistency, having correlation medians of 0.15 (-0.01 ≤ r ≤ 0.81 ), 0.05 (-0.11 ≤ r ≤ 0.31), and 0.06 (-0.18 ≤ r ≤ 0.19), respectively (Fig 1). Correlations for SCA and SCZ indicated higher consistency having medians of 0.2 (-0.08 ≤ r ≤ 0.38, -0.06 ≤ r ≤ 0.5, respectively). Mantel tests revealed no significant correlation between the site-by-site consistency of GMV differences and variations in clinical and demographic characteristics of the samples.
·Figure: For each disorder, the upper panel: the between-site correlation matrix of the t-maps; the lower panel: the kernel density estimation of all elements of the between-site correlation matrix
Conclusions:
Our results suggest the SCZ and SCA show some evidence for a consistent core neuroanatomical phenotype, whereas ASD, BD, and MDD show no such evidence. This lack of consistency may reflect an extreme clinical and biological heterogeneity in these disorders or a limited capacity for VBM to detect the underlying disease phenotype, highlighting the need for more careful analysis and interpretation of psychiatric neuroimaging findings .
Disorders of the Nervous System:
Neurodevelopmental/ Early Life (eg. ADHD, autism)
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Cortical Anatomy and Brain Mapping 2
Keywords:
Cortex
Psychiatric Disorders
STRUCTURAL MRI
1|2Indicates the priority used for review
Provide references using author date format
Ashburner, J., and K. J. Friston. 2000. ‘Voxel-Based Morphometry--the Methods’. NeuroImage 11 (6 Pt 1): 805–21. https://doi.org/10.1006/nimg.2000.0582.
Brugger, Stefan P., and Oliver D. Howes. 2017. ‘Heterogeneity and Homogeneity of Regional Brain Structure in Schizophrenia: A Meta-Analysis’. JAMA Psychiatry 74 (11): 1104–11. https://doi.org/10.1001/jamapsychiatry.2017.2663.
Cattarinussi, Giulia, Parnia Pouya, David Antonio Grimaldi, Mahta Zare Dini, Fabio Sambataro, Paolo Brambilla, and Giuseppe Delvecchio. 2024. ‘Cortical Alterations in Relatives of Patients with Bipolar Disorder: A Review of Magnetic Resonance Imaging Studies’. Journal of Affective Disorders 345 (January): 234–43. https://doi.org/10.1016/j.jad.2023.10.097.
Fortin, Jean-Philippe, Nicholas Cullen, Yvette I. Sheline, Warren D. Taylor, Irem Aselcioglu, Philip A. Cook, Phil Adams, et al. 2018. ‘Harmonization of Cortical Thickness Measurements across Scanners and Sites’. NeuroImage 167 (February): 104–20. https://doi.org/10.1016/j.neuroimage.2017.11.024.
Gaser, Christian, Robert Dahnke, Paul M. Thompson, Florian Kurth, Eileen Luders, and Alzheimer’s Disease Neuroimaging Initiative. 2022. ‘CAT – A Computational Anatomy Toolbox for the Analysis of Structural MRI Data’. bioRxiv. https://doi.org/10.1101/2022.06.11.495736.
Li, Xiaoyi, Kai Zhang, Xiao He, Jinyun Zhou, Chentao Jin, Lesang Shen, Yuanxue Gao, Mei Tian, and Hong Zhang. 2021. ‘Structural, Functional, and Molecular Imaging of Autism Spectrum Disorder’. Neuroscience Bulletin 37 (7): 1051–71. https://doi.org/10.1007/s12264-021-00673-0.
Zhuo, Chuanjun, Gongying Li, Xiaodong Lin, Deguo Jiang, Yong Xu, Hongjun Tian, Wenqiang Wang, and Xueqin Song. 2019. ‘The Rise and Fall of MRI Studies in Major Depressive Disorder’. Translational Psychiatry 9 (1): 1–14. https://doi.org/10.1038/s41398-019-0680-6.