doi:10.1093/brain/awaa001 BRAIN 2020: 143; 684–700 | 684 Brain structural covariance networks in obsessive-compulsive disorder: a graph analysis from the ENIGMA Consortium Je-Yeon Yun,1,2 Premika S.W. Boedhoe,3,4 Chris Vriend,3,4 Neda Jahanshad,5 Yoshinari Abe,6 Stephanie H. Ameis,7,8 Alan Anticevic,9 Paul D. Arnold,10,11 Marcelo C. Batistuzzo,12 Francesco Benedetti,13 Jan C. Beucke,14 Irene Bollettini,13 Anushree Bose,15 Silvia Brem,16 Anna Calvo,17 Yuqi Cheng,18 Kang Ik K. Cho,19 Valentina Ciullo,20 Sara Dallaspezia,13 Damiaan Denys,21,22 Jamie D. Feusner,23 Jean-Paul Fouche,24 Mònica Giménez,25,26 Patricia Gruner,9 Derrek P. Hibar,5 Marcelo Q. Hoexter,12 Hao Hu,27 Chaim Huyser,28,29 Keisuke Ikari,30 Norbert Kathmann,14 Christian Kaufmann,14 Kathrin Koch,31,32 Luisa Lazaro,33,34,35,36 Christine Lochner,37 Paulo Marques,38 Rachel Marsh,39,40 Ignacio Martı́nez-Zalacaı́n,41,42 David Mataix-Cols,43 José M. Menchón,36,41,42 Luciano Minuzzi,44 Pedro Morgado,38,45,46 Pedro Moreira,38,45,46 Takashi Nakamae,6 Tomohiro Nakao,47 Janardhanan C. Narayanaswamy,15 Erica L. Nurmi,23 Joseph O’Neill,23,48 John Piacentini,23,48 Fabrizio Piras,20 Federica Piras,20 Y.C. Janardhan Reddy,15 Joao R. Sato,49 H. Blair Simpson,39,50 Noam Soreni,51 Carles Soriano-Mas,36,41,52 Gianfranco Spalletta,20,53 Michael C. Stevens,54,55 Philip R. Szeszko,56,57 David F. Tolin,54,58 Ganesan Venkatasubramanian,15 Susanne Walitza,16 Zhen Wang,27,59 Guido A. van Wingen,21 Jian Xu,60 Xiufeng Xu,60 Qing Zhao,27 ENIGMA-OCD working group,* Paul M. Thompson,5 Dan J. Stein,24 Odile A. van den Heuvel3,4 and Jun Soo Kwon61,62  Appendix 1. Brain structural covariance networks reflect covariation in morphology of different brain areas and are thought to reflect common trajectories in brain development and maturation. Large-scale investigation of structural covariance networks in obsessive-compul- sive disorder (OCD) may provide clues to the pathophysiology of this neurodevelopmental disorder. Using T1-weighted MRI scans acquired from 1616 individuals with OCD and 1463 healthy controls across 37 datasets participating in the ENIGMA-OCD Working Group, we calculated intra-individual brain structural covariance networks (using the bilaterally-averaged values of 33 cortical surface areas, 33 cortical thickness values, and six subcortical volumes), in which edge weights were proportional to the similarity between two brain morphological features in terms of deviation from healthy controls (i.e. z-score transformed). Global networks were characterized using measures of network segregation (clustering and modularity), network integration (global effi- ciency), and their balance (small-worldness), and their community membership was assessed. Hub profiling of regional networks was undertaken using measures of betweenness, closeness, and eigenvector centrality. Individually calculated network measures were integrated across the 37 datasets using a meta-analytical approach. These network measures were summated across the net- work density range of K = 0.10–0.25 per participant, and were integrated across the 37 datasets using a meta-analytical approach. Received June 11, 2019. Revised November 24, 2019. Accepted November 26, 2019 ß The Author(s) (2020). Published by Oxford University Press on behalf of the Guarantors of Brain. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 685 Compared with healthy controls, at a global level, the structural covariance networks of OCD showed lower clustering (P 5 0.0001), lower modularity (P 5 0.0001), and lower small-worldness (P = 0.017). Detection of community membership emphasized lower network segregation in OCD compared to healthy controls. At the regional level, there were lower (rank-trans- formed) centrality values in OCD for volume of caudate nucleus and thalamus, and surface area of paracentral cortex, indicative of altered distribution of brain hubs. Centrality of cingulate and orbito-frontal as well as other brain areas was associated with OCD illness duration, suggesting greater involvement of these brain areas with illness chronicity. In summary, the findings of this study, the largest brain structural covariance study of OCD to date, point to a less segregated organization of structural covariance networks in OCD, and reorganization of brain hubs. The segregation findings suggest a possible signature of altered brain morph- ometry in OCD, while the hub findings point to OCD-related alterations in trajectories of brain development and maturation, par- ticularly in cingulate and orbitofrontal regions. 1 Seoul National University Hospital, Seoul, Republic of Korea 2 Yeongeon Student Support Center, Seoul National University College of Medicine, Seoul, Republic of Korea 3 Amsterdam UMC, Vrije Universiteit Amsterdam, Department of Psychiatry, Amsterdam Neuroscience, Amsterdam, The Netherlands 4 Amsterdam UMC, Vrije Universiteit Amsterdam, Department of Anatomy & Neurosciences, Amsterdam Neuroscience, Amsterdam, The Netherlands 5 Imaging Genetics Center, Mark and Mary Stevens Neuroimaging & Informatics Institute, Keck School of Medicine of the University of Southern California, Marina del Rey, CA, USA 6 Department of Psychiatry, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan 7 The Margaret and Wallace McCain Centre for Child, Youth & Family Mental Health, Campbell Family Mental Health Research Institute, The Centre for Addiction and Mental Health, Department of Psychiatry, Faculty of Medicine, University of Toronto, Toronto, Canada 8 Centre for Brain and Mental Health, The Hospital for Sick Children, Toronto, Canada 9 Department of Psychiatry, Yale University School of Medicine, New Haven, Connecticut, USA 10 Mathison Centre for Mental Health Research and Education, Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada 11 Department of Psychiatry, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada 12 Departamento e Instituto de Psiquiatria do Hospital das Clinicas, IPQ HCFMUSP, Faculdade de Medicina, Universidade de Sao Paulo, SP, Brazil 13 Psychiatry and Clinical Psychobiology, Division of Neuroscience, Scientific Institute Ospedale San Raffaele, Milano, Italy 14 Department of Psychology, Humboldt-Universität zu Berlin, Berlin, Germany 15 Obsessive-Compulsive Disorder (OCD) Clinic Department of Psychiatry National Institute of Mental Health and Neurosciences, Bangalore, India 16 Department of Child and Adolescent Psychiatry and Psychotherapy, Psychiatric Hospital, University of Zurich, Zurich, Switzerland 17 Magnetic Resonance Image Core Facility, IDIBAPS (Institut d’Investigacions Biomèdiques August Pi i Sunyer), Barcelona, Spain 18 Department of Psychiatry, First Affiliated Hospital of Kunming Medical University, Kunming, China 19 Institute of Human Behavioral Medicine, SNU-MRC, Seoul, Republic of Korea 20 Laboratory of Neuropsychiatry, Department of Clinical and Behavioral Neurology, IRCCS Santa Lucia Foundation, Rome, Italy 21 Amsterdam UMC, University of Amsterdam, Department of Psychiatry, Amsterdam Neuroscience, Amsterdam, The Netherlands 22 Netherlands Institute for Neuroscience, Royal Netherlands Academy of Arts and Sciences, Amsterdam, The Netherlands 23 Semel Institute for Neuroscience and Human Behavior, University of California, Los Angeles, CA, USA 24 SAMRC Unit on Risk and Resilience in Mental Disorders, Department of Psychiatry, University of Cape Town, Cape Town, South Africa 25 Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Carlos III Health Institute, Barcelona, Spain 26 Department of Psychiatry, Bellvitge University Hospital, Bellvitge Biomedical Research Institute-IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain 27 Shanghai Mental Health Center Shanghai Jiao Tong University School of Medicine, PR China 28 De Bascule, Academic Center for Child and Adolescent Psychiatry, Amsterdam, The Netherlands 29 Amsterdam UMC, Department of Child and Adolescent Psychiatry, Amsterdam, The Netherlands 30 Department of Neuropsychiatry, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi Higashi-ku, Fukuoka, Japan 31 Department of Neuroradiology, Klinikum rechts der Isar, Technische Universität München, Germany 32 TUM-Neuroimaging Center (TUM-NIC) of Klinikum rechts der Isar, Technische Universität München, Germany 33 Department of Child and Adolescent Psychiatry and Psychology, Institute of Neurosciences, Hospital Clı́nic Universitari, Barcelona, Spain 34 Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain 35 Department of Medicine, University of Barcelona, Barcelona, Spain 36 Centro de Investigación Biomèdica en Red de Salud Mental (CIBERSAM), Barcelona, Spain 37 SAMRC Unit on Anxiety and Stress Disorders, Department of Psychiatry, University of Stellenbosch, South Africa 686 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. 38 Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Braga, Portugal 39 Columbia University Medical College, Columbia University, New York, NY, USA 40 The New York State Psychiatric Institute, New York, NY, USA 41 Department of Psychiatry, Bellvitge University Hospital, Bellvitge Biomedical Research Institute-IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain 42 Department of Clinical Sciences, University of Barcelona, Spain 43 Department of Clinical Neuroscience, Centre for Psychiatry Research, Karolinska Institutet, Stockholm, Sweden 44 McMaster University, Department of Psychiatry and Behavioural Neurosciences, Hamilton, Ontario, Canada 45 Clinical Academic Center–Braga, Braga, Portugal 46 ICVS-3Bs PT Government Associate Laboratory, Braga, Portugal 47 Department of Neuropsychiatry, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan 48 Division of Child and Adolescent Psychiatry, University of California, Los Angeles, CA, USA 49 Center for Mathematics, Computing and Cognition, Universidade Federal do ABC, Santo Andre, Brazil 50 Center for OCD and Related Disorders, New York State Psychiatric Institute, New York, NY, USA 51 Pediatric OCD Consultation Service, Anxiety Treatment and Research Center, St. Joseph’s HealthCare, Hamilton, Ontario, Canada 52 Department of Psychobiology and Methodology of Health Sciences, Universitat Autònoma de Barcelona, Spain 53 Beth K. and Stuart C. Yudofsky Division of Neuropsychiatry, Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, Texas, USA 54 Yale University School of Medicine, New Haven, Connecticut, USA 55 Clinical Neuroscience and Development Laboratory, Olin Neuropsychiatry Research Center, Hartford, Connecticut, USA 56 Icahn School of Medicine at Mount Sinai, New York, USA 57 James J. Peters VA Medical Center, Bronx, New York, USA 58 Institute of Living/Hartford Hospital, Hartford, Connecticut, USA 59 Shanghai Key Laboratory of Psychotic Disorders, PR China 60 Department of Psychiatry, The First Affiliated Hospital of Kunming Medical University, Kunming, Yunnan, PR China 61 Department of Psychiatry, Seoul National University College of Medicine, Seoul, Republic of Korea 62 Department of Brain and Cognitive Sciences, Seoul National University College of Natural Sciences, Seoul, Republic of Korea Correspondence to: Prof Jun Soo Kwon Department of Psychiatry, Seoul National University College of Medicine, 101, Daehak-ro, Jongno-gu, Seoul 03080, Republic of Korea E-mail: kwonjs@snu.ac.kr Keywords: brain structural covariance network; graph theory; obsessive-compulsive disorder; pharmacotherapy; illness duration Abbreviation: OCD = obsessive-compulsive disorder connectivity of the fronto-parietal and salience networks, as Introduction well as reduced inter-network connectivity between the sali- Three decades of neuroimaging research support the view ence, fronto-parietal and default-mode networks (Gürsel that structural brain abnormalities in obsessive-compulsive et al., 2018). disorder (OCD) do not merely involve alterations in discrete Brain structural covariance networks reflect intra-individ- brain regions, but rather are best characterized in terms of ual (Yun et al., 2016; Seidlitz et al., 2018a) or inter-individ- altered networks of brain structures (Boedhoe et al., 2017). ual (Alexander-Bloch et al., 2013; Kaczkurkin et al., 2019; More specifically, brain-based models of OCD have empha- Wannan et al., 2019) covariation in morphology of different sized the role of the cortico-striato-thalamo-cortical loops brain areas, which may in turn point to common trajectories and have also suggested the involvement of fronto-limbic, in brain development and maturation (Yun et al., 2015, fronto-parietal and cerebellar regions (Menzies et al., 2008; 2016; Hunt et al., 2016). Such networks may focus on a Milad and Rauch, 2012; de Wit et al., 2014; Piras et al., range of morphological features including regional brain vol- 2015; van den Heuvel et al., 2016; Boedhoe et al., 2017, ume (Spreng et al., 2019), cortical thickness (Solé-Casals 2018; Fouche et al., 2017). Most studies of brain networks et al., 2019), cortical surface area (Sharda et al., 2017), and in OCD have used resting state functional MRI (rs-fMRI) cortical white-grey contrast (Makowski et al., 2019), as well (Soriano-Mas and Harrison, 2017; Gürsel et al., 2018), with as the paired or conjoint patterns between different brain alterations evident in intra-network connections of fronto- regions (Seidlitz et al., 2018b; Hoagey et al., 2019) Brain limbic and fronto-striatal networks (Anticevic et al., 2014; structural covariance has been estimated using Pearson’s cor- Gottlich et al., 2014; Posner et al., 2014; Armstrong et al., relation coefficient (Seidlitz et al., 2018a; Solé-Casals et al., 2016; de Vries et al., 2017; Takagi et al., 2017). 2019; Wannan et al., 2019), partial least squares (Hoagey Furthermore, a meta-analysis of rs-fMRI studies comparing et al., 2019; Spreng et al., 2019), non-negative matrix factor- OCD to healthy controls found decreased intra-network ization (Kaczkurkin et al., 2019), and inverse exponential of Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 687 the difference between z-score transformed brain morpho- Networks were characterized at the global level using meas- logical values (Wee et al., 2013; Yun et al., 2015, 2016), ures of network segregation (clustering coefficient and among others. Structural covariance networks are more modularity), network integration (global efficiency), and similar to patterns of functional connectivity than the archi- their balance (small-worldness), as well as at the regional tecture of white matter connections, suggesting that areas level using betweenness, closeness, and eigenvector central- that co-vary in morphological characteristics also belong to ities (Lancichinetti and Fortunato, 2009; Rubinov and the same functional network (Zielinski et al., 2010; Soriano- Sporns, 2010; Cao et al., 2016; Palaniyappan et al., 2016; Mas et al., 2013). Such networks are thought to be shaped Vriend et al., 2018). For preservation of the network edge by genetic and environmental influences from early child- weights-related information in the derived graph metrics, the hood (Richmond et al., 2016) and may continue to be global and regional graph metrics were summed across the reshaped during the lifespan (Alexander-Bloch et al., 2013; network density range of K = 0.10–0.25 (Uehara et al., Aboud et al., 2019; Qi et al., 2019) by a range of trophic 2014). influences (Ferrer et al., 1995; Draganski et al., 2004; Previous neuroimaging studies of global network metrics Mechelli et al., 2005). have reported more (Zhang et al., 2011, 2014), less (Shin Inter-individual brain structural covariance networks et al., 2014; Armstrong et al., 2016; Jung et al., 2017; have been explored in a few studies of OCD and healthy Reess et al., 2018a), or similar levels (Reess et al., 2016) of controls. For example, Pujol et al. (2004) found a negative segregated organization of white matter-based structural association between relative volume reduction for OCD connectivity networks, resting state functional connectivity (compared to healthy controls) in the medial prefrontal- networks, or local gyrification index-based structural covari- insulo-opercular cortical regions and relative volume en- ance networks in individuals with OCD, compared to largement of ventral striatum, suggesting that abnormal healthy controls. These inconsistent findings raise the need brain morphology in OCD might be distributed in coordi- for larger-scale meta-analysis. Therefore, the current study nated fashion across diverse brain regions. In addition, a aimed to assess the level of global network segregation, as recent mega-analysis found higher covariance between determined by the global clustering coefficient, using volumes of left putamen and left frontal operculum, and the largest dataset of structural covariance networks in higher covariance between volumes of right amygdala and OCD to date. ventromedial prefrontal cortex in OCD compared to healthy controls (Subira et al., 2016). Further, local cor- tical gyrification (associated with cortical maturation)- based structural covariance network demonstrated lower Materials and methods covariance among mainly ventral brain regions in OCD compared to healthy controls (Reess et al., 2018b). Samples However, few studies have explored intra-individual brain This study included 37 datasets from 26 international re- structural covariance networks in OCD; consequently our search institutes participating in the OCD Working Group of understanding of the factors that influence changes in glo- the ENIGMA (Enhancing NeuroImaging and Genetics bal and regional network characteristics within individu- through Meta-Analysis) Consortium used in the meta-analytic als with OCD is limited. between-group comparisons of OCD and healthy controls in The ENIGMA-OCD Working Group has collaborated on terms of the subcortical volumes (Boedhoe et al., 2017), cor- tical surface area and cortical thickness (Boedhoe et al., developing a large database of structural brain imaging in 2018), in addition to the cortical and subcortical asymmetry OCD and healthy controls, providing a unique opportunity (Kong et al., 2019). Each dataset included demographic and to undertake such an exploration. Here we constructed neuroimaging data from OCD and healthy controls, as well intra-individual structural covariance networks from region as OCD clinical data (Table 1 and Supplementary material). of interest-based brain morphological features using 37 data- The diagnosis of psychiatric disorders including OCD and sets worldwide (n = 1616 for OCD; n = 1463 for healthy other comorbid disorders (if any) was made using a struc- controls), and investigated network topology using a graph tured or semi-structured interview; the Structured Clinical theory approach. The current study aimed to capture the Interview for DSM-IV [SCID-I (First et al., 2002); n = 23 intra-individual distribution of brain morphological changes datasets], the Mini-International Neuropsychiatric Interview (Wee et al., 2013; Yun et al., 2015, 2016) in OCD across [MINI (Sheehan et al., 1998); n = 6 datasets], the Anxiety 33 cortical surface areas, 33 cortical thickness values, and Disorder Interview Schedule [ADIS (Silverman et al., 2001; Grisham et al., 2004); n = 2 datasets], or the Schedule for six subcortical volumes (Kremen et al., 2013; Amlien et al., Affective Disorders and Schizophrenia for School-Aged 2016; Sussman et al., 2016; Vijayakumar et al., 2016; Children: Present and Lifetime Version [K-SADS-PL Krongold et al., 2017; Schmaal et al., 2017). Thus edge (Kaufman and Schweder, 2003); n = 7 datasets] (Table 1 and weights of the intra-individual structural covariance net- Supplementary material). Comorbid lifetime depressive dis- works were estimated in proportion to the similarity be- order was present in 256 individuals with OCD, and comor- tween two brain morphological features in terms of bid lifetime anxiety disorder was present in 267 (Table 1 and deviation from healthy controls (i.e. z-score transformed). Supplementary material). At the time of MRI acquisition, 721 Table 1 Demographic and clinical information 688 Study Study PI Study site MRI Total, n Age, mean (SD) Sex, male / female Comorbid Comorbid Y-BOCS Medicated Illness field lifetime lifetime total OCD, duration strength, T HC OCD HC OCD HC OCD depression anxiety (OCD), n (%) (OCD), (OCD) n (%) (OCD) n (%) mean (SD) mean (SD) 1 Beucke Berlin, GER 1.5 54 57 32 (11) 33 (11) 23 / 31 31 / 26 11 (19) 6 (11) 20 (7) 23 (40) 16.2 (11) 2 Cheng Kunming, CHN 1.5 28 16 32 (8) 32 (12) 8 / 20 5 / 11 4 (25) 6 (38) 31 (7) 10 (63) 4.2 (5.2) 3 van den Heuvel Amsterdam, NLD 1.5 35 37 31 (8) 35 (9) 12 / 23 11 / 26 11 (30) 6 (16) 23 (6) 0 (0) 20 (11.8) 4 Hoexter San Paulo, BRA 1.5 9 38 28 (6) 31 (9) 5/4 17 / 21 20 (53) 24 (63) 28 (6) 8 (21) 17.4 (10.3) 5 Kwon Seoul, KOR_01 1.5 103 45 24 (4) 25 (5) 57 / 46 34 / 11 0 (0) 0 (0) 20 (6) 11 (24) 7.3 (5.2) 6 Kwon Seoul, KOR_02 1.5 45 34 25 (5) 29 (7) 29 / 16 19 / 15 1 (3) 0 (0) 24 (6) 0 (0) 9.9 (7.1) | BRAIN 2020: 143; 684–700 7 Mataix-Cols Stockholm, SWE 1.5 28 34 36 (11) 39 (11) 9 / 19 15 / 19 9 (26) 9 (26) 25 (8) 14 (41) 20.5 (14.9) 8 Menchon Barcelona, ESP 1.5 55 95 32 (10) 35 (9) 22 / 33 47 / 48 15 (16) 19 (20) 25 (6) 91 (96) 13.9 (9.9) 9 Morgado Braga, Portugal 1.5 51 58 28 (6) 27 (8) 19 / 32 27 / 31 – – 26 (6) 58 (100) – 10 Nakamae Kyoto, JPN 1.5 48 81 30 (8) 32 (9) 25 / 23 37 / 44 18 (22) 8 (10) 25 (6) 39 (48) 6.7 (6.8) 11 Reddy India 1.5 20 29 26 (6) 28 (7) 14 / 6 16 / 13 – – 25 (9) 0 (0) 5.7 (5.3) 12 Benedetti Milan, ITA 3 23 22 29 (11) 35 (11) 19 / 4 13 / 9 0 (0) 0 (0) 31 (6) 13 (59) 18.7 (12) 13 Cheng Kunming, CHN 3 72 40 26 (4) 33 (11) 20 / 52 21 / 19 13 (33) 37 (93) 28 (6) 25 (63) 5.4 (5.8) 14 Denys Amsterdam, NLD 3 15 14 38 (12) 34 (11) 6/9 1 / 13 4 (29) 1 (7) 27 (6) 9 (64) 14.9 (13.5) 15 van den Heuvel Amsterdam, NLD 3 30 32 39 (11) 39 (11) 12 / 18 16 / 16 17 (53) 13 (41) 21 (6) 0 (0) 25.9 (12.9) 16 Koch Munchen, GER 3 71 75 30 (9) 31 (10) 28 / 43 28 / 47 0 (0) 0 (0) 21 (6) 45 (60) 14.3 (10.6) 17 Kwon Seoul, KOR 3 89 90 26 (7) 27 (7) 54 / 35 56 / 34 2 (2) 1 (1) 27 (7) 2 (2) 7.7 (6.7) 18 Nakamae Kyoto, JPN 3 39 34 30 (7) 33 (10) 19 / 20 12 / 22 7 (21) 3 (9) 22 (7) 0 (0) 8.1 (6.1) 19 Nakao Fukuoka, JPN 3 31 66 39 (13) 37 (10) 11 / 20 30 / 36 22 (33) 0 (0) 23 (6) 59 (89) 12.2 (9.3) 20 Nurmi Los Angeles, USA 3 22 45 31 (12) 34 (11) 14 / 8 22 / 23 9 (20) 16 (36) 25 (4) 12 (27) 23 (10.8) 21 Reddy India 3 139 201 26 (5) 30 (7) 86 / 53 107 / 94 31 (15) 15 (7) 26 (6) 82 (41) 7.3 (5.4) 22 Simpson New York, USA 3 31 30 28 (8) 30 (8) 17 / 14 17 / 13 10 (33) 7 (23) 26 (4) 0 (0) 15.1 (8.7) 23 Spalletta Rome, ITA 3 95 71 38 (11) 36 (11) 54 / 41 45 / 26 8 (11) 8 (11) 23 (9) 65 (92) 16.6 (11.4) 24 Stein Cape Town, ZAF 3 25 21 31 (11) 31 (11) 10 / 15 11 / 10 0 (0) 0 (0) 23 (4) 9 (43) 17.9 (11.3) 25 Tolin Conneticut, USA 3 32 27 48 (12) 32 (12) 7 / 25 18 / 9 11 (41) 12 (44) 23 (5) 21 (78) – 26 Walitza Zurich, CHE 3 15 13 33 (9) 31 (7) 4 / 11 7/6 6 (46) 7 (54) 18 (10) 6 (46) 12.8 (10) 27 Wang Shanghai, CHN 3 35 47 26 (8) 30 (9) 18 / 17 23 / 24 0 (0) 0 (0) 25 (5) 0 (0) 6.5 (5.5) 28 Lazaro Barcelona, ESP 1.5 29 29 15 (2) 14 (2) 14 / 15 18 / 11 0 (0) 5 (17) 22 (6) 15 (52) 2.1 (1.8) 29 Arnold Ontario, CAN 3 11 34 12 (2) 13 (2) 6/5 20 / 14 7 (21) 10 (29) 21 (8) 21 (62) 4.2 (2.6) 30 Gruner Conneticut, USA 3 17 10 14 (2) 15 (2) 8/9 9/1 2 (20) 6 (60) 27 (5) 6 (60) – 31 Hoexter San Paulo, BRA 3 26 27 12 (2) 13 (2) 15 / 11 16 / 11 6 (22) 20 (74) 27 (5) 12 (44) 5.5 (2.4) 32 Huyser Amsterdam, NLD 3 20 20 14 (3) 14 (2) 8 / 12 6 / 14 7 (35) 9 (45) 26 (5) 0 (0) 3.1 (2.6) 33 Lazaro Barcelona, ESP 3 43 53 15 (2) 15 (2) 23 / 20 30 / 23 3 (6) 14 (26) 19 (7) 42 (79) 2.5 (2.1) 34 Nurmi Los Angeles, USA 3 36 53 13 (2) 13 (3) 18 / 18 29 / 24 1 (2) 2 (4) 24 (4) 7 (13) – 35 Reddy India 3 10 14 14 (3) 14 (2) 5/5 8/6 1 (7) 3 (21) 22 (7) 12 (86) 1.5 (0.9) 36 Soreni Ontario, CAN 3 20 18 11 (3) 13 (2) 10 / 10 7 / 11 0 (0) 0 (0) 23 (4) 0 (0) – 37 Walitza Zurich, CHE 3 11 6 16 (2) 16 (1) 6/5 5/1 0 (0) 0 (0) 18 (10) 4 (67) 5 (2.4) A more detailed version of this table is provided in the Supplementary material. A dash indicates data were not available. BRA = Brazil; CAN = Canada; CHE = Switzerland; CHN = China; ESP = Spain; GER = Germany; HC = healthy control; ITA = Italy; KOR_01/02 = South Korea site 1/2; NLD = the Netherlands; PI = principal investigator; SWE = Sweden; Y-BOCS = Yale–Brown Obsessive Compulsive Scale; ZAF = South Africa. J.-Y. Yun et al. Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 689 individuals with OCD were on psychotropic medication and Graph theory approach: single 881 were not. Age of illness onset of OCD was 18.8  9.1 years, and illness duration was 10.8  10.1 years (n = 1415). subject level Severity of obsessive-compulsive symptoms was assessed with Global network characteristics the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS; for Intra-individual structural covariance networks were thresh- patients aged 518) or Children’s Y-BOCS (CY-BOCS; for olded (using ‘threshold_proportional.m’ function in network patients aged 518); the mean score of 24.2  6.8 (n = 1581) density range of K = 0.05–0.30; with interval of 0.01) and indicated a moderate to severe range of symptoms in the binarized (using the ‘weight-conversion.m’ function; e.g. when study population. All local institutional review boards per- we applied a density threshold of K = 0.10, the edge weights in mitted the use of extracted numerical measures for meta- the network were sorted into numerical order and a cut-off was analysis. applied to retain only the strongest 10% of edges with edge weights converted to ‘1’ and edges weights for other remaining edges becomes ‘0’ (Fig 2, steps 2A and 3A). From these thresh- Image acquisition and processing olded and binarized networks, four global metrics were deter- mined: (i) global clustering (a tendency for brain regions to Structural T1-weighted brain MRI scans were acquired and segregate into locally interconnected triplets of neighbouring processed at each study site. For acquisition parameters of each nodes); (ii) global modularity (a measure of the segregation of site see Supplementary Table 1. All parcellations were per- the network into communities where nodes are more strongly formed with fully automated segmentation software FreeSurfer connected with each other than nodes outside the community version 5.3. (Fischl, 2012), following standardized ENIGMA because of similar morphological characteristics; this measure is protocols (http://enigma.usc.edu/protocols/imaging-protocols/). operationalized as the most frequently occurring value over 500 To ensure quality control, we visually inspected the segmenta- runs of estimation using ‘modularity_und.m’) (Newman, 2006; tions of 68 (34 left and 34 right) cortical grey matter regions Reichardt and Bornholdt, 2006); (iii) global efficiency (how well and seven subcortical regions based on the Desikan-Killiany on average each node is connected to all others based on the atlas (Desikan et al., 2006) and statistically evaluated the data minimum number of steps nodes are separated from each for outliers (Boedhoe et al., 2017, 2018). We excluded the vol- other); and (iv) small-worldness (a measure of balance between ume values of bilateral entorhinal cortices and the nucleus the degree of segregation versus integration in brain network) accumbens because of segmentation issues (as calculation of using the Brain Connectivity Toolbox (Rubinov and Sporns, intra-individual brain structural covariance networks requires 2010) in MATLAB R2017a (Weinberg et al., 2016; Das et al., every region of interest to be adequately measured in each par- 2018; Zaremba et al., 2018). ticipant; inclusion of regions of interest with relatively poorer Among the diverse network density levels of K = 0.05–0.30 quality segmentations would effectively decrease sample size). (with density interval of 0.01), only in the narrower network density levels of K = 0.10–0.25, three criteria of (Uehara et al., 2014) (i) network connectedness (4 80% of nodes remain con- Intra-individual cortical-subcortical nected to other nodes within the network); (ii) modular organ- ization (modularity 4 0.3); and (iii) small-world organization structural covariance networks (small-worldness 4 1) were satisfied for 495% of the intra-in- As illustrated at ‘step 1’ in Fig. 1, bilaterally-averaged values dividual structural covariance networks comprising each dataset (where brain regions were poorly segmented in one hemisphere, (n = 37). Therefore, these network density levels of K = 0.10– the value from the contralateral hemisphere was used as a 0.25 (density interval = 0.01) were selected for the between- proxy) of 33 cortical surface area regions of interest, 33 cortical group comparison of global network characteristics, community thickness regions of interest, and six subcortical volume regions membership detection, and hub profiling using the regional net- of interest, were corrected for age, sex, and individual brain size work characteristics (Fig. 2, step 3A). Estimation of the global (Vuoksimaa et al., 2016) per dataset (n = 37). The resulting network characteristics was done using Brain Connectivity residuals were then z-score transformed using mean and SD val- Toolbox (https://www.nitrc.org/projects/bct/) in MATLAB ues of each region of interest calculated from healthy controls R2017a. (to derive the degrees of brain morphological variations per re- gion of interest relative to the ‘average healthy controls’ values). Detection of community membership Finally, a measure of joint variation (which is not the same as In addition, we assessed community membership (Fortunato, the classical statistical definition of covariance) between the 72 2010) for each structural covariance (joint variation) network. morphometric features (33 cortical surface area values, 33 cor- For thresholded (K = 0.10–0.25) and binarized intra-individual tical thickness values, and six subcortical values) represented the structural covariance (joint variation) networks, detection of edge-weights (distributed between 0 and 1) of the network and communities [i.e. densely connected subgroups of nodes in a was calculated using the following formula (Yun et al., 2015, network (Power et al., 2013)] was conducted using the InfoMap 2016): algorithm (Rosvall and Bergstrom, 2007; Fortunato, 2010; [Intra-individual brain structural covariance (joint variation) be- Power et al., 2011; Kawamoto and Rosvall, 2015). First, a par- tween the ith (for i = 1 to 72) and j-th (for j = 1 to 72) regions of ticipant-level co-classification matrix (Dwyer et al., 2014) that interest in the k-th (for k = 1 to ‘total number of participants per represented the fraction of network density level, in which each dataset’) participant] = 1/exp{[(z-transformed value of i-th region of pair of nodes was clustered into the same community according interest in k-th participant) – (z-transformed value of j-th region of to the InfoMap algorithm (Rosvall and Bergstrom, 2007; interest in k-th participant)]2} (1) Kawamoto and Rosvall, 2015), was generated. Second, the 690 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. Figure 1 Schematic description of the study procedures: construction of intra-individual brain structural covariance net- works. HC = healthy controls; L = left; M = mean; R = right; ROI = region of interest; SD = standard deviation. InfoMap algorithm was applied to this co-classification matrix Meta-analysis of graph metrics to generate a participant-level consensus of community member- ship (Fornito et al., 2016). All procedures other than the Global network characteristics InfoMap-based community estimation were done using Meta-analysis of between-group differences in global network MATLAB R2017a software (https://kr.mathworks.com). characteristics across the whole dataset (n = 37; Fig. 2, step 3A) was performed using a random-effects meta-analytic model (Hedges and Vevea, 1998; Kambeitz et al., 2016) incorporating the bias-corrected standardized mean difference (SMD = Hub profiling and regional network Hedges’ g) between OCD and healthy controls for each of the characteristics four global network characteristics (summated over the network density range of K = 0.10–0.25) that satisfied network connect- Principal brain regions that could be essential indicators of brain edness, modular organization, and small-world organization; morphological changes within the network were assessed using see ‘Graph theory approach: single subject level’ section). hub profiling, which provided three local network measures: Summary effect sizes were calculated with restricted maximum- (i) betweenness centrality (the frequency of a node being located likelihood estimator (REML) (Raudenbush, 2009; Viechtbauer, in the shortest path for each pair of two other nodes in a net- 2010). Estimates for heterogeneity were assessed with the I2 work); (ii) closeness centrality (the ease with which one node value (Raudenbush, 2009). For all analyses, a significance level can reach all other nodes within a network); and (iii) eigen- of P 5 0.01 was used, i.e. P 5 0.05/5 number of global net- vector centrality (a self-referential measure of centrality that work characteristics (= 4) plus local network-related measure of reflects the presence of connectedness of one node to other the Dice coefficient (= 1; see section below) (Kambeitz et al., nodes with high eigenvector centrality) (Rubinov and Sporns, 2016). All statistical analyses were conducted using the R pack- 2010) (Fig 2, step 2B). As distribution of these local network age ‘metafor’ version 2.0.0 (Viechtbauer, 2010). measures does not follow normal distribution in a scale-free net- work, prior to the between-group comparison and meta-ana- lysis, these regional centrality metrics were rank-transformed Community membership using the ‘tiedrank.m’ function of MATLAB R2017a and were First, summation of network-transformed community profiles averaged in the network density range of K = 0.10–0.25 to be for each individual provided dataset-level co-classification matri- re-ranked at participant-level; participant-level hubs were ces (in which higher edge weights indicated that two nodes were selected as top-10 ranked nodes in two or three centralities. All clustered in the same community across a large proportion of of the procedures described above were conducted using the participants in dataset) for OCD and for healthy controls Brain Connectivity Toolbox (Rubinov and Sporns, 2010) and (Fornito et al., 2016). Second, consensus of community mem- MATLAB R2017a software (https://kr.mathworks.com). bership at dataset level (for OCD and healthy controls Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 691 Figure 2 Schematic description of the study procedures. (A) Calculation of graph theory metrics from the intra-individual brain struc- tural covariance networks at single-subject level and (B) meta-analytic integration of graph theory metrics for 37 datasets. HC = healthy con- trols; ROI = region of interest. separately) was estimated by applying the InfoMap algorithm to participants number per dataset, and summed to generate the the weighted and thresholded (at density level of K = 0.10) ver- meta-analytic co-classification matrices of OCD or healthy con- sion of the dataset-level co-classification matrices. Third, data- trols (n = 37). Finally, a weighted and thresholded (at density set-level consensus community profiles of OCD and healthy level of K = 0.10) version of these meta-analytic co-classification controls were binarized, multiplied by the square root of matrices underwent InfoMap-based community detection, to 692 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. determine the meta-analytic consensus community profile for Influence of comorbid lifetime brain structural covariance networks of OCD and healthy con- trols. All procedures other than the InfoMap-based community depressive or anxiety disorders in estimation were performed using MATLAB R2017a. patients with OCD Thirty-five (of 37) datasets provided information about comor- bid lifetime depressive and anxiety disorders in OCD individu- Hub profiling and regional network characteristics als; meta-analysis of global network characteristics and Dice coefficients was conducted to assess between-group differences In the current study, hub profiling was done to find the princi- in (i) OCD with and without comorbid lifetime depressive dis- pal brain regions that could be essential indicators of intra-indi- order (n = 10 datasets, in which n 4 10 for both OCD sub- vidual distribution of brain morphological changes (= deviation groups); and (ii) OCD with and without comorbid lifetime from healthy controls) based on the three local metrics of anxiety disorders (n = 7 datasets, in which n 4 10 for both betweenness, closeness, and eigenvector centralities (Fig. 2). OCD subgroups). Three rank-transformed centralities (betweenness, closeness, and eigenvector) were rank-transformed at the participant level, and were averaged in the network density range of K = 0.10–0.25. Influence of medication The top 10 ranked nodes (i.e. 10 nodes illustrated in Fig. 4 and Twenty-seven (of 37) datasets provided information about medi- Supplementary Fig. 1) for two or three centralities as calculated cation status (= presence or absence of psychotropic medication from the summation of participant-level centrality values within prescribed at the time of MRI data acquisition) of OCD individu- each dataset (n = 37) were classified as dataset-level hubs for als; meta-analytic integration for the between-group comparison OCD or healthy controls. Finally, meta-analytic hub scores for of regional network characteristics (= centralities) between medi- all network nodes (= 33 cortical surface area values + 33 cor- cated OCD versus unmedicated OCD was undertaken for these tical thickness values + six subcortical volumes) were calculated datasets. Furthermore, meta-analytic integration of between- by summing the values of [(presence (= 1) or absence (= 0) of group differences for global network metrics and Dice coefficients network nodes in the hub profile of each dataset)  (square were conducted using results retrieved from 12 datasets (in which root of participants number per dataset)] across the whole data- n 4 10 for both medicated and unmedicated subjects). set (n = 37) for OCD and healthy controls separately; top-10 ranked nodes for this meta-analytic hub score were defined as meta-analytic hubs for OCD or healthy controls, respectively. Influence of OCD illness duration Between-group comparison of rank-transformed centrality Fisher’s z-transformed correlation coefficients between the OCD values at the dataset level (n = 37) was performed using the illness duration and four global network metrics were calculated Wilcoxon rank sum test. Nodes that showed statistically signifi- per dataset (n = 32 datasets). Each of these correlation coefficients cant differences between OCD and healthy controls (P 5 0.05) per dataset and per global network characteristics were meta-ana- were recoded into MNI coordinates using brainGraph (https:// lytically integrated using the same pipeline as for the global net- cran.r-project.org/web/packages/brainGraph), and underwent work characteristics. Likewise, Spearman correlation coefficients coordinate-based meta-analysis, i.e. activation likelihood estima- between the OCD illness duration and rank-transformed (betwe- tion (ALE), using gingerALE version 2.3.6. (Eickhoff et al., enness, closeness, or eigenvector) centrality measures were also 2017). In this ALE-based meta-analysis, nodes that showed sig- calculated per dataset. Meta-analysis of the dataset-level nodes nificant effect sizes [cluster-level corrected threshold of P 5 0.05 that showed significant correlation with OCD illness duration (family-wise error, FWE); cluster-forming threshold at voxel (P 5 0.05) was performed using using gingerALE version 2.3.6 level of P 5 0.001] for between-group differences in two or [P 5 0.05 (cluster-level FWE)] (Eickhoff et al., 2017). three centralities were considered valid (Fig. 2, step 3B). Lastly, to explore the difference in hubs in terms of their topo- graphical location between OCD and healthy controls, we also Data availability calculated the Dice similarity coefficient (Dice, 1945), a measure De-identified data are available from the corresponding author of the degree of overlap between each participant-level hub pro- upon reasonable request. file versus the reference (= hub profile of healthy controls per dataset). For meta-analysis, the bias-corrected SMD (Hedges’ g) of Dice similarity coefficient (i) between the healthy controls and OCD (37 dataset) as well as (ii) between unmedicated OCD Results and medicated OCD (12 dataset in which 410 participants existed for all of the two subgroups) were calculated and Patients with OCD versus healthy entered into a random-effects meta-analytic model (Schmidt et al., 2009; Kambeitz et al., 2016). Summary effect sizes were controls calculated with REML (Raudenbush, 2009; Viechtbauer, 2010), Demographic and clinical characteristics and estimates for the amount of heterogeneity were assessed by A total of 37 datasets worldwide (n = 1616 for OCD; way of the I2 value (= the percentage of total variability across dataset that is due to heterogeneity than by chance) (Higgins n = 1463 for healthy controls) were included in this study. et al., 2003). For all analyses, a significance level of P 5 0.05 Demographic and clinical characteristics for each dataset are (two-tailed) was used (Kambeitz et al., 2016) and all statistical described in Table 1 and Supplementary material. Between- analyses were conducted using the R package ‘metafor’ version group (OCD versus healthy controls) statistical tests for age 2.0.0 (Viechtbauer, 2010). (using the independent t-test) and sex ratio (using the Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 693 Figure 3 Forest plots of the meta-analysis of global graph metrics comparying the OCD and healthy control groups. (A) Global clustering, (B) small-worldness, (C) modularity, (D) global efficiency, and (E) dice similarity coefficient. HC = healthy controls; ROI = region of interest. chi-squared test) did not show statistically significant differ- 0.01). Global efficiency and small-worldness did not differ ences between OCD and healthy controls (P 4 0.05) for 31 significantly between OCD and healthy controls (all P’s 4 (83.8%) and 34 datasets (91.9%), respectively. On the other 0.01). When the sample was divided into two groups (adults hand, years of education (information available for 27 data- and adolescents), and analyses run in each, these findings sets) were fewer in OCD compared to healthy controls continued to hold true (Table 2). Additional meta-analyses (P 5 0.05) in 10 (27.0%) datasets. using years of education as a moderator did not show any significant influence of this variable (all P’s 4 0.05) on ei- ther the global network metrics of global clustering Global network characteristics (Qm = 1.456, df = 2, P = 0.483), modularity (Qm = 0.819, Meta-analysis of global network characteristics for the intra- df = 2, P = 0.664), global efficiency (Qm = 0.673, df = 2, individual brain structural covariance networks (Table 2 P = 0.714), and small-worldness (Qm = 0.139, df = 2, and Fig. 3A–D) showed lowered global clustering and P = 0.933), or on the Dice similarity coefficient (Qm = 1.447, modularity in OCD compared to healthy controls (all P’s 5 df = 2, P = 0.485). 694 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. Figure 3 Continued Community membership healthy controls) included], module 2 [comprising cortical Community membership analysis detected that the healthy thickness of inferior parietal lobule-precuneus (module 2 in controls network had six modules (or subgroups within the healthy controls) in addition to cuneus-lingual-pericalcarine network), while the OCD network had three modules, indi- gyri (module 4 in healthy controls)], and module 3 (includes cative of less global network segregation. The six community a hub named hippocampal volume). modules of the healthy controls network (Fig. 4 and Supplementary Fig. 1) were module 1 [the principal (31 nodes); including six hubs of cortical surface area for medial Regional network characteristics orbitofrontal, caudal middle frontal, and parahippocampal Of the 10 hubs for the OCD network (Fig. 4 and cortices, as well as cortical thickness for posterior cingulate, Supplementary Fig. 1), only one node, i.e. cortical thickness pars triangularis, and insula], module 2 [cingulate-parietal- of postcentral cortex [member of the paracentral-temporal inferior frontal (13 nodes)], module 3 [subcortical (six module in healthy controls; fifth community (red square) in nodes); including two hubs named pallidal and caudate vol- Supplementary Fig. 1], was found among the 10 healthy umes], module 4 [frontal pole-occipital (six nodes); including controls hubs. Meta-analysis of Dice similarity coefficients cortical thickness for cuneus as hub], module 5 [paracentral- showed lower Dice similarity coefficient in OCD compared temporal (six nodes); including a hub of paracentral cortical to healthy controls (Table 2 and Fig. 3E), indicating that the thickness], and module 6 [insula-perisylvian (five nodes)]. As nodes classified as hubs differed between OCD and healthy smaller communities with less than four nodes (55% of controls. In terms of the centralities, compared to healthy total nodes) were excluded, six nodes comprising module 2 controls, rank-transformed centrality of caudate nucleus vol- for healthy controls [cortical surface area of caudal-rostral ume was lower in OCD (healthy controls hub; Fig. 5A and anterior cingulate and lateral orbitofrontal cortices, in add- Supplementary Fig. 3). ition to cortical thickness of paracentral, superior parietal, and supramarginal cortices] were not classified in these communities. Influence of comorbid lifetime In contrast, community membership of individualized depressive or anxiety disorders in structural covariance networks for OCD (Fig. 4 and Supplementary Fig. 1) showed just three modules: module 1 patients with OCD [in which eight OCD hubs for cortical surface area of super- No significant differences in global network characteristics ior temporal sulcus (module 1 in healthy controls), posterior or Dice similarity coefficients were found between OCD cingulate (module 2 in healthy controls), rostral middle with comorbid lifetime depression versus OCD without life- frontal-insular-superior temporal (module 6 in healthy con- time depression, nor between OCD with comorbid lifetime trols), and pericalcarine cortices, as well as cortical thickness anxiety disorder versus OCD without lifetime anxiety dis- of caudal anterior cingulate-frontal pole (module 1 in order (Table 2). Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 695 Table 2 Meta-analysis of global network characteristics and Dice similarity coefficients logSMD k z P-value 95% CI I2 (%) Q P OCD versus HC Global clustering coefficient (total) 0.77 37 –6.94 50.001 0.72 to 0.83 0.01 44.8 0.149 Adults (518 years) 0.79 27 –5.89 50.001 0.73 to 0.85 50.001 26.8 0.418 Adolescents (518 years) 0.66 10 –3.16 0.002 0.50 to 0.85 45.7 16.5 0.058 Modularity (total) 0.82 37 –5.21 50.001 0.77 to 0.89 0.01 43.1 0.194 Adults (518 years) 0.84 27 –4.28 50.001 0.78 to 0.91 0.01 22.3 0.670 Adolescents (518 years) 0.68 10 –2.63 0.009 0.51 to 0.91 54.0 19.1 0.025 Small-worldness (total) 0.92 37 –2.39 0.017 0.85 to 0.98 0.001 26.2 0.886 Adults (518 years) 0.93 27 –1.82 0.069 0.86 to 1.01 50.001 18.1 0.872 Adolescents (518 years) 0.84 10 –1.82 0.068 0.70 to 1.01 50.001 7.2 0.621 Global efficiency (total) 0.98 37 –0.54 0.586 0.91 to 1.05 0.02 38.5 0.358 Adults (518 years) 0.97 27 –0.68 0.494 0.89 to 1.06 10.7 32.6 0.174 Adolescents (518 years) 1.05 10 0.50 0.621 0.87 to 1.26 50.001 5.3 0.809 Dice similarity coefficient (total) 0.48 37 –14.36 50.001 0.43 to 0.53 39.35 58.3 0.011 Adults (518 years) 0.49 27 –11.97 50.001 0.44 to 0.55 45.7 49.6 0.004 Adolescents (518 years) 0.41 10 –9.18 50.001 0.34 to 0.50 50.001 2.9 0.969 OCD patients with versus without lifetime comorbid depressive disorder Global clustering coefficient 0.89 10 –1.13 0.257 0.73 to 1.09 11.6 10.1 0.344 Modularity 0.90 10 –0.91 0.365 0.72 to 1.13 26.5 11.9 0.217 Small-worldness 0.96 10 –0.44 0.659 0.80 to 1.15 50.001 6.6 0.678 Global efficiency 1.00 10 –0.04 0.966 0.83 to 1.20 50.001 5.8 0.764 Dice similarity coefficient 1.04 10 0.32 0.751 0.84 to 1.28 21.4 13.2 0.155 OCD patients with versus without lifetime comorbid anxiety disorder Global clustering coefficient 0.99 7 –0.09 0.929 0.79 to 1.25 50.001 5.1 0.531 Modularity 0.96 7 –0.39 0.695 0.76 to 1.20 50.001 1.8 0.934 Small-worldness 1.00 7 0.01 0.993 0.79 to 1.27 3.1 6.6 0.357 Global efficiency 1.03 7 0.24 0.814 0.82 to 1.29 50.001 6.2 0.403 Dice similarity coefficient 1.15 7 0.96 0.338 0.87 to 1.52 31.2 8.3 0.215 Medicated OCD versus unmedicated OCD Global clustering coefficient 0.95 12 –0.63 0.531 0.82 to 1.11 0.00 11.03 0.441 Modularity 0.94 12 –0.83 0.408 0.8 to 1.09 0.00 8.73 0.647 Small-worldness 0.99 12 –0.08 0.934 0.83 to 1.18 15.25 12.97 0.295 Global efficiency 0.85 12 –1.66 0.097 0.7 to 1.03 28.05 13.32 0.273 Dice similarity coefficient 1.06 12 0.72 0.474 0.91 to 1.24 2.16 6.90 0.807 Correlation coefficient k z P-value 95% CI I2 Q P Illness duration in OCD Global clustering coefficient –0.03 32 –0.85 0.393 –0.11 to 0.04 40.13 53.37 0.008 Modularity –0.05 32 –1.32 0.188 –0.12 to 0.02 34.70 48.57 0.023 Small-worldness –0.02 32 –0.67 0.584 –0.10 to 0.05 34.00 46.28 0.038 Global efficiency –0.02 32 –0.59 0.558 –0.07 to 0.04 0.00 20.64 0.921 CI = 95% confidence interval; I2 = total heterogeneity/total variability; k = number of studies included in given meta-analysis; log SMD = log-transformed standardized mean differ- ence; P = P-value of heterogeneity test; P-value = P-value of random effect model (REML); Q = heterogeneity score; z = z-score. Influence of medication for OCD OCD illness duration showed significant positive relation- ships with centrality (Fig. 5C and Supplementary Fig. 5) of No significant differences in global network characteristics or cortical thickness for caudal anterior cingulate (OCD hub), Dice similarity coefficients were found between medicated and cortical surface area for posterior cingulate (OCD hub), and unmedicated OCD (Table 2). The structural covariance net- cortical surface area of lateral orbitofrontal cortex (non- works of healthy controls, medicated OCD, and unmedicated hub). Furthermore, OCD illness duration showed significant OCD demonstrated five, three, and two modules (or subgroups negative correlations with centrality of the cortical surface within the network), respectively (Supplementary Fig. 2). area for parahippocampal cortex (healthy control hub), cor- tical thickness for the frontal pole, cortical surface area for superior temporal and pericalcarine cortices (OCD hubs), Influence of OCD illness duration cortical thickness for inferior parietal lobule, and cortical OCD illness duration did not show significant correlations surface areas for inferior temporal and cingulate isthmus with global network characteristics (Table 2). However, cortices (non-hubs). 696 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. Figure 4 Meta-analysis of community membership and hubs. (A) Healthy bontrols (HC); and (B) OCD. Spheres represent nodes [= bi- laterally-averaged values of 33 cortical surface areas (CSAs), 33 cortical thickness (CT), and six subcortical volumes (vol)] comprising the intra- individual structural covariance network. Larger spheres represent hubs, and differential colours were used to denote the spheres (or network nodes) segregated as different modules. Figure 5 Meta-analysis of regional network characteristics (= rank-transformed betweenness, closeness, and eigenvector cen- tralities). (A) Comparing OCD and healthy controls (HC); (B) comparing medicated OCD with unmedicated OCD; and (C) estimating the degrees of relationship with illness duration for OCD. CSA = cortical surface areas; CT = cortical thickness. Discussion duration. These novel findings are important; the first sug- gests a possible signature of altered brain morphometry in The current meta-analysis of 37 datasets from 26 sites world- OCD compared to healthy controls, and the second provides wide is the largest investigation of structural covariance net- evidence for OCD-related alterations in trajectories of brain works in OCD to date. Two main findings emerged. First, we development and maturation. observed lower clustering, modularity, and small-worldness Lower clustering, modularity and small-worldness, but nor- of OCD brain structural covariance networks compared with mal global efficiency, are indicative of lower global segrega- healthy controls, with community membership analysis con- tion, but spared global integration of OCD networks. In firming a less segregated organization of the global structural particular, lower modularity might be related to over-connect- covariance network of OCD patients. Second, hub profiling edness of certain nodes and diminished ability of the network demonstrated reduced centralities in subcortical volumes of to adapt flexibly (Guye et al., 2010). This finding is consistent caudate nucleus and thalamus as well as cortical surface area with previous observations of abnormal brain network segre- of paracentral cortex in OCD. Alterations in hub organiza- gation in functional networks in OCD (Zhang et al., 2011). tion were associated with both medication status and illness Small-worldness relates to an optimal network organization Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 697 that combines regional specialization and efficient global orbitofrontal regions. However, these hypotheses will re- (Watts and Strogatz, 1998; Latora and Marchiori, 2001; quire confirmation in longitudinal studies. Lefort-Besnard et al., 2018). Thus, despite intact global effi- This study has some limitations that deserve emphasis. ciency, decreased levels of small-worldness and modularity in First, the current study analysed datasets that were extracted OCD point to a disrupted hierarchical network architecture. from brain MRI data collected from 26 international re- Global network findings were not impacted by medication search institutions using diverse acquisition parameters status or illness duration. This contrasts with previous research, (Boedhoe et al., 2017, 2018), which may have introduced which although based on functional MRI data, suggested that systematic biases. Nevertheless, our meta-analytical ap- abnormal global network characteristics may depend on psy- proach took into account differential site effects. Second, al- chotropic treatment (Shin et al., 2014). Although it is theoretic- though all brain segmentation results underwent quality ally possible that the effects of psychotropic medication on check procedures prior to extraction of numerical values, we OCD brain morphology differ in the acute versus chronic stage were unable to implement motion correction of structural of pharmacotherapy so that there the net result over time is one images, and it is theoretically possible that estimates of of no change, there is little evidence to support this idea. In our group differences are inflated by uncorrected motion. view, a more plausible conclusion is that the lower global net- Nevertheless, there is no reason to suspect increased motion work segregation found here may represent a possible signature in either group. Third, in the calculation of intra-individual of altered brain morphometry in OCD. Further research is structural covariance networks, the current study applied the needed to confirm this. bilaterally-averaged values of 33 cortical surface area regions The study also found reduced centralities of caudate nu- of interest, 33 cortical thickness regions of interest, and six cleus and thalamic volumes in OCD compared to healthy subcortical volume regions of interest and therefore did not controls. This is in line with our previous multicentre mega- explore the homologous connectivity between the brain analysis, which showed increased thalamic volume in OCD regions. However, we would like to emphasize that patterns compared to healthy controls, even though only in the of brain cortical-subcortical morphological asymmetry in paediatric patients (Boedhoe et al., 2017). Likewise, caudate adult OCD are not significantly different from healthy con- nucleus and thalamus showed marked expansion in OCD trols (Kong et al., 2019). Fourth, the current study did not and in their unaffected siblings compared to healthy con- explore the possible effect of other clinical features such as trols, suggesting genetic contributions to altered brain the severity of depressive or anxiety symptoms, and IQ morphology (Shaw et al., 2015). Of note, meta-analytic inte- score, on the brain morphological features, because of the gration of task-related functional MRI studies demonstrated lack of sufficient information. Fifth, this was a cross-section- OCD-specific differences in functional activation of the al study and any conclusions regarding developmental trajec- caudate nucleus. Similarly, nodal efficiency of the caudate tories are necessarily tentative. nucleus was reduced in OCD in a white matter-based struc- Taken together, this study showed that the structural co- tural connectivity network (Zhong et al., 2014), in line with variance networks of individuals with OCD are less segre- a resting state functional connectivity profile that showed gated and show a reorganization of brain hubs, compared increased intra-subcortical modular connections for caudate to healthy controls. These findings support the hypothesis nucleus and thalamus in OCD (Vaghi et al., 2017). that OCD brain abnormalities are best described at the net- Our data emphasize that alterations in hubs in OCD are work level and involve alterations in the hierarchical struc- associated with illness duration. This is consistent with pre- ture of the brain. The segregation findings here are vious work suggesting brain-related changes during the de- important insofar as they suggest a possible signature of velopment of OCD (van den Heuvel et al., 2016). In altered brain morphometry OCD, while the hub findings are particular, we found that centralities of brain regions includ- useful in emphasizing the importance of OCD-related altera- ing the cortical thickness of caudal anterior cingulate as well tions in trajectories of brain development and maturation, as the cortical surface areas for posterior cingulate and lat- particularly in cingulate and orbitofrontal regions. eral orbitofrontal cortices, were associated with longer illness duration in OCD. As an interface between sensorimotor, limbic and executive networks, the caudal anterior cingulate plays a major role in attentional control (Margulies et al., 2007) and self-referential sensorimotor processing (Jung Funding et al., 2015; Mao et al., 2017), the posterior cingulate cortex This research was funded by Basic Science Research and connected default mode network supports internally- Program through the National Research Foundation directed cognition, participates in the control of arousal of Korea (NRF) funded by the Ministry of Education state, and interacts with other brain regions for attentional (NRF-2017R1D1A1B03028464) and the Basic Research modulation and conscious awareness (Leech and Sharp, Laboratory Program through the National Research 2014). Orbitofrontal regions have also previously been Foundation of Korea (NRF) (Grant no. emphasized in OCD. The hub findings reported here point 2018R1A4A1025891). to OCD-related alterations in trajectories of brain develop- ENIGMA is supported, in part, by an NIH grant (U54 ment and maturation, particularly in cingulate and EB020403) for big data analytics. 698 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. Alexander-Bloch A, Raznahan A, Bullmore E, Giedd J. The conver- Competing interests gence of maturational change and structural covariance in human cortical networks. J Neurosci 2013; 33: 2889–99. P.D.A (Alberta Innovates Translational Health Chair in Amlien IK, Fjell AM, Tamnes CK, Grydeland H, Krogsrud SK, Child and Youth Mental Health). D.P.H. is now an employ- Chaplin TA, et al. Organizing principles of human cortical develop- ee of Genentech, Inc. working on projects unrelated to this ment–thickness and area from 4 to 30 years: insights from compara- publication. D.M-C. receives royalties for contributing tive primate neuroanatomy. Cereb Cortex 2016; 26: 257–67. Anticevic A, Hu S, Zhang S, Savic A, Billingslea E, Wasylink S, et al. articles to UpToDate, Wolters Kluwer Health and fees from Global resting-state functional magnetic resonance imaging analysis Elsevier for editorial tasks (all unrelated to the submitted identifies frontal cortex, striatal, and cerebellar dysconnectivity in work). H.B.S. (Biohaven research support for clinical trial; obsessive-compulsive disorder. Biol Psychiatry 2014; 75: 595–605. Royalties from UpToDate, Inc and Cambridge University Armstrong CC, Moody TD, Feusner JD, McCracken JT, Chang S, Press). N.S. (Lundbeck-IIT). P.M.T. has received a research Levitt JG, et al. Graph-theoretical analysis of resting-state fMRI in pediatric obsessive-compulsive disorder. J Affect Disord 2016; 193: grant from Biogen, Inc. unrelated to the topic of this paper. 175–84. No further conflict of interest is reported. Boedhoe PSW, Schmaal L, Abe Y, Alonso P, Ameis SH, Anticevic A, et al. Cortical abnormalities associated with pediatric and adult obses- sive-compulsive disorder: findings from the ENIGMA obsessive-com- pulsive disorder working group. Am J Psychiatry 2018; 175: 453–62. Boedhoe PS, Schmaal L, Abe Y, Ameis SH, Arnold PD, Batistuzzo Supplementary material MC, et al. Distinct subcortical volume alterations in pediatric and adult OCD: a worldwide meta- and mega-analysis. Am J Psychiatry Supplementary material is available at Brain online. 2017; 174: 60–9. Cao M, Huang H, Peng Y, Dong Q, He Y. Toward developmental connectomics of the human brain. Front Neuroanat 2016; 10: 25. Appendix 1 Das T, Borgwardt S, Hauke DJ, Harrisberger F, Lang UE, Riecher- Rossler A, et al. Disorganized gyrification network properties during the transition to psychosis. JAMA Psychiatry 2018; 75: 613–22. de Vries FE, de Wit SJ, van den Heuvel OA, Veltman DJ, Cath DC, ENIGMA-OCD working group van Balkom A, et al. Cognitive control networks in OCD: a resting- state connectivity study in unmedicated patients with obsessive-com- members pulsive disorder and their unaffected relatives. World J Biol Odile3 A. van den Heuvel, Dan J. Stein, Premika S.W. Psychiatry 2017; 20: 230–42. Boedhoe, Paul M. Thompson, Neda Jahanshad, Chris de Wit SJ, Alonso P, Schweren L, Mataix-Cols D, Lochner C, Menchon JM, et al. Multicenter voxel-based morphometry mega- Vriend, Yoshinari Abe, Stephanie H. Ameis, Alan Anticevic, analysis of structural brain scans in obsessive-compulsive disorder. Paul D. Arnold, Marcelo C. Batistuzzo, Francesco Am J Psychiatry 2014; 171: 340–9. Benedetti, Jan C. Beucke, Irene Bollettini, Anushree Bose, Desikan RS, Segonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, Silvia Brem, Anna Calvo, Yuqi Cheng, Kang Ik K. Cho, et al. An automated labeling system for subdividing the human cere- Valentina Ciullo, Sara Dallaspezia, Damiaan Denys, Jamie bral cortex on MRI scans into gyral based regions of interest. Neuroimage 2006; 31: 968–80. D. Feusner, Jean-Paul Fouche, Mònica Giménez, Patricia Dice LR. Measures of the amount of ecologic association between spe- Gruner, Derrek P. Hibar, Marcelo Q. Hoexter, Hao Hu, cies. Ecology 1945; 26: 297–302. Chaim Huyser, Keisuke Ikari, Norbert Kathmann, Christian Draganski B, Gaser C, Busch V, Schuierer G, Bogdahn U, May A. Kaufmann, Kathrin Koch, Jun Soo Kwon, Luisa Lazaro, Neuroplasticity: changes in grey matter induced by training. Nature Christine Lochner, Paulo Marques, Rachel Marsh, Ignacio 2004; 427: 311–2. Dwyer DB, Harrison BJ, Yucel M, Whittle S, Zalesky A, Pantelis C, Martı́nez-Zalacaı́n, David Mataix-Cols, José M. Menchón, et al. Large-scale brain network dynamics supporting adolescent Luciano Minuzzi, Pedro Morgado, Pedro Moreira, Takashi cognitive control. J Neurosci 2014; 34: 14096–107. Nakamae, Tomohiro Nakao, Janardhanan C. Eickhoff SB, Laird AR, Fox PM, Lancaster JL, Fox PT. Narayanaswamy, Erica L. Nurmi, Joseph O’Neill, John Implementation errors in the GingerALE Software: description and Piacentini, Fabrizio Piras, Federica Piras, Y.C. Janardhan recommendations. Hum Brain Mapp 2017; 38: 7–11. Ferrer I, Blanco R, Carulla M, Condom M, Alcantara S, Olive M, Reddy, Joao R. Sato, H. Blair Simpson, Noam Soreni, et al. Transforming growth factor-alpha immunoreactivity in the Carles Soriano-Mas, Gianfranco Spalletta, Michael C. developing and adult brain. Neuroscience 1995; 66: 189–99. Stevens, Philip R. Szeszko, David F. Tolin, Ganesan First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Venkatasubramanian, Susanne Walitza, Zhen Wang, Guido clinical interview for DSM-IV-TR axis I disorders (SCID-I). New A. van Wingen, Jian Xu, Xiufeng Xu, Je-Yeon Yun, Qing York, NY: New York State Psychiatric Institute, Biometrics Department; 2002. Zhao. Fischl B. FreeSurfer. Neuroimage 2012; 62: 774–81. Fornito A, Zalesky A, Bullmore ET. Fundamentals of brain network analysis. London: Elsevier; 2016. References Fortunato S. Community detection in graphs. Phys Rep 2010; 486: 75–174. Aboud KS, Huo Y, Kang H, Ealey A, Resnick SM, Landman BA, et al. Fouche JP, Du Plessis S, Hattingh C, Roos A, Lochner C, Soriano-Mas Structural covariance across the lifespan: brain development and C, et al. Cortical thickness in obsessive-compulsive disorder: multi- aging through the lens of inter-network relationships. Hum Brain site mega-analysis of 780 brain scans from six centres. Br J Mapp 2019; 40: 125–36. Psychiatry 2017; 210: 67–74. Brain structural covariance network of OCD BRAIN 2020: 143; 684–700 | 699 Gottlich M, Kramer UM, Kordon A, Hohagen F, Zurowski B. Lefort-Besnard J, Bassett DS, Smallwood J, Margulies DS, Derntl B, Decreased limbic and increased fronto-parietal connectivity in unme- Gruber O, et al. Different shades of default mode disturbance in dicated patients with obsessive-compulsive disorder. Hum Brain schizophrenia: subnodal covariance estimation in structure and func- Mapp 2014; 35: 5617–32. tion. Hum Brain Mapp 2018; 39: 644–61. Grisham JR, Brown TA, Campbell LA. The anxiety disorders interview Makowski C, Lewis JD, Lepage C, Malla AK, Joober R, Lepage M, schedule for DSM-IV (ADIS-IV). Comprehensive handbook of psy- et al. Structural associations of cortical contrast and thickness in first chological assessment, vol 2: personality assessment. Hoboken, NJ: episode psychosis. Cerebral Cortex 2019; 29: 5009–21. John Wiley & Sons Inc; 2004. p. 163–77. Mao CV, Araujo MF, Nishimaru H, Matsumoto J, Tran AH, Hori E, Gürsel DA, Avram M, Sorg C, Brandl F, Koch K. Frontoparietal areas et al. Pregenual anterior cingulate gyrus involvement in spontaneous link impairments of large-scale intrinsic brain networks with aberrant social interactions in primates-evidence from behavioral, pharmaco- fronto-striatal interactions in OCD: a meta-analysis of resting-state logical, neuropsychiatric, and neurophysiological findings. Front functional connectivity. Neurosci Biobehav Rev 2018; 87: 151–60. Neurosci 2017; 11: 34. Guye M, Bettus G, Bartolomei F, Cozzone PJ. Graph theoretical ana- Margulies DS, Kelly AM, Uddin LQ, Biswal BB, Castellanos FX, lysis of structural and functional connectivity MRI in normal and Milham MP. Mapping the functional connectivity of anterior cingu- pathological brain networks. MAGMA 2010; 23: 409–21. late cortex. Neuroimage 2007; 37: 579–88. Hedges LV, Vevea JL. Fixed- and random-effects models in meta-ana- Mechelli A, Friston KJ, Frackowiak RS, Price CJ. Structural covariance lysis. Psychol Methods 1998; 3: 486–504. in the human cortex. J Neurosci 2005; 25: 8303–10. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsist- Menzies L, Chamberlain SR, Laird AR, Thelen SM, Sahakian BJ, ency in meta-analyses. BMJ 2003; 327: 557–60. Bullmore ET. Integrating evidence from neuroimaging and neuro- Hoagey DA, Rieck JR, Rodrigue KM, Kennedy KM. Joint contribu- psychological studies of obsessive-compulsive disorder: the orbito- tions of cortical morphometry and white matter microstructure in fronto-striatal model revisited. Neurosci Biobehav Rev 2008; 32: healthy brain aging: a partial least squares correlation analysis. 525–49. Hum Brain Mapp 2019; 40: 5315–29. Milad MR, Rauch SL. Obsessive-compulsive disorder: beyond segre- Hunt BA, Tewarie PK, Mougin OE, Geades N, Jones DK, Singh KD, gated cortico-striatal pathways. Trends Cogn Sci 2012; 16: 43–51. et al. Relationships between cortical myeloarchitecture and electro- Newman ME. Modularity and community structure in networks. Proc physiological networks. Proc Natl Acad Sci USA 2016; 113: 13510–5. Natl Acad Sci USA 2006; 103: 8577–82. Jung WM, Lee IS, Wallraven C, Ryu YH, Park HJ, Chae Y. Cortical Palaniyappan L, Marques TR, Taylor H, Mondelli V, Reinders A, activation patterns of bodily attention triggered by acupuncture Bonaccorso S, et al. Globally efficient brain organization and treat- stimulation. Sci Rep 2015; 5: 12455. ment response in psychosis: a connectomic study of gyrification. Jung WH, Yucel M, Yun JY, Yoon YB, Cho KI, Parkes L, et al. Schizophr Bull 2016; 42: 1446–56. Altered functional network architecture in orbitofronto-striato-thal- Piras F, Piras F, Chiapponi C, Girardi P, Caltagirone C, Spalletta G. amic circuit of unmedicated patients with obsessive-compulsive dis- Widespread structural brain changes in OCD: a systematic review of order. Hum Brain Mapp 2017; 38: 109–19. voxel-based morphometry studies. Cortex 2015; 62: 89–108. Kaczkurkin AN, Park SS, Sotiras A, Moore TM, Calkins ME, Cieslak Posner J, Marsh R, Maia TV, Peterson BS, Gruber A, Simpson HB. M, et al. Evidence for dissociable linkage of dimensions of psycho- Reduced functional connectivity within the limbic cortico-striato- pathology to brain structure in youths. Am J Psychiatry 2019; 176: thalamo-cortical loop in unmedicated adults with obsessive-compul- 1000–9. sive disorder. Hum Brain Mapp 2014; 35: 2852–60. Kambeitz J, Kambeitz-Ilankovic L, Cabral C, Dwyer DB, Calhoun VD, Power JD, Cohen AL, Nelson SM, Wig GS, Barnes KA, Church JA, van den Heuvel MP, et al. Aberrant functional whole-brain network et al. Functional network organization of the human brain. Neuron architecture in patients with schizophrenia: a meta-analysis. 2011; 72: 665–78. Schizophr Bull 2016; 42: S13–21. Power JD, Schlaggar BL, Lessov-Schlaggar CN, Petersen SE. Evidence Kaufman J, Schweder AE. The schedule for affective disorders and for hubs in human functional brain networks. Neuron 2013; 79: schizophrenia for school age children: present and lifetime version 798–813. (K-SADS-PL). In: M Hersen, DM Segal, M Hilsenroth, editors. The Pujol J, Soriano-Mas C, Alonso P, Cardoner N, Menchon JM, Deus J, comprehensive handbook of psychological assessment (CHOPA), et al. Mapping structural brain alterations in obsessive-compulsive volume 2: personality assessment. New York: John Wiley and Sons; disorder. Arch Gen Psychiatry 2004; 61: 720–30. 2003. Qi T, Schaadt G, Cafiero R, Brauer J, Skeide MA, Friederici AD. The Kawamoto T, Rosvall M. Estimating the resolution limit of the map emergence of long-range language network structural covariance equation in community detection. Phys Rev E Stat Nonlin Soft and language abilities. Neuroimage 2019; 191: 36–48. Matter Phys 2015; 91: 012809. Raudenbush SW. Analyzing effect sizes: random effects models. New Kong XZ, Boedhoe PSW, Abe Y, Alonso P, Ameis SH, Arnold PD, York: Russell Sage Foundation; 2009. et al. Mapping cortical and subcortical asymmetry in obsessive-com- Reess TJ, Rus OG, Gursel DA, Schmitz-Koep B, Wagner G, Berberich pulsive disorder: findings from the ENIGMA Consortium. Biol G, et al. Network-based decoupling of local gyrification in Psychiatry 2019; pii: S0006-3223(19)31292-2. obsessive-compulsive disorder. Hum Brain Mapp 2018a; 39: Kremen WS, Fennema-Notestine C, Eyler LT, Panizzon MS, Chen CH, 3216–26. Franz CE, et al. Genetics of brain structure: contributions from the Reess TJ, Rus OG, Gürsel DA, Schmitz-Koep B, Wagner G, Berberich Vietnam era twin study of aging. Am J Med Genet B Genet 2013; G, et al. Network-based decoupling of local gyrification in obses- 162: b751–61. sive-compulsive disorder. Hum Brain Mapp 2018b; 39: 3216–26. Krongold M, Cooper C, Bray S. Modular development of cortical gray Reess TJ, Rus OG, Schmidt R, de Reus MA, Zaudig M, Wagner G, matter across childhood and adolescence. Cereb Cortex 2017; 27: et al. Connectomics-based structural network alterations in obses- 1125–36. sive-compulsive disorder. Transl Psychiatry 2016; 6: e882. Lancichinetti A, Fortunato S. Community detection algorithms: a com- Reichardt J, Bornholdt S. Statistical mechanics of community detec- parative analysis. Phys Rev E 2009; 80: 056117. tion. Phys Rev E Stat Nonlin Soft Matter Phys 2006; 74 (Pt 2): Latora V, Marchiori M. Efficient behavior of small-world networks. 016110. Phys Rev Lett 2001; 87: 198701. Richmond S, Johnson KA, Seal ML, Allen NB, Whittle S. Development Leech R, Sharp DJ. The role of the posterior cingulate cortex in cogni- of brain networks and relevance of environmental and genetic factors: tion and disease. Brain 2014; 137 (Pt 1): 12–32. a systematic review. Neurosci Biobehav Rev 2016; 71: 215–39. 700 | BRAIN 2020: 143; 684–700 J.-Y. Yun et al. Rosvall M, Bergstrom CT. An information-theoretic framework for Takagi Y, Sakai Y, Lisi G, Yahata N, Abe Y, Nishida S, et al. A neural resolving community structure in complex networks. Proc Natl marker of obsessive-compulsive disorder from whole-brain function- Acad Sci USA 2007; 104: 7327–31. al connectivity. Sci Rep 2017; 7: 7538. Rubinov M, Sporns O. Complex network measures of brain connectiv- Uehara T, Yamasaki T, Okamoto T, Koike T, Kan S, Miyauchi S, ity: uses and interpretations. Neuroimage 2010; 52: 1059–69. et al. Efficiency of a “small-world” brain network depends on con- Schmaal L, Yucel M, Ellis R, Vijayakumar N, Simmons JG, Allen NB, sciousness level: a resting-state FMRI study. Cereb Cortex 2014; 24: et al. Brain structural signatures of adolescent depressive symptom 1529–39. trajectories: a longitudinal magnetic resonance imaging study. J Am Vaghi MM, Vertes PE, Kitzbichler MG, Apergis-Schoute AM, van der Acad Child Adolesc Psychiatry 2017; 56: 593–601.e9. Flier FE, Fineberg NA, et al. Specific frontostriatal circuits for Schmidt FL, Oh IS, Hayes TL. Fixed- versus random-effects models impaired cognitive flexibility and goal-directed planning in obses- in meta-analysis: model properties and an empirical comparison of sive-compulsive disorder: evidence from resting-state functional con- differences in results. Br J Math Stat Psychol 2009; 62 (Pt 1): nectivity. Biol Psychiatry 2017; 81: 708–17. 97–128. van den Heuvel OA, van Wingen G, Soriano-Mas C, Alonso P, Seidlitz J, Vasa F, Shinn M, Romero-Garcia R, Whitaker KJ, Vertes Chamberlain SR, Nakamae T, et al. Brain circuitry of compulsivity. PE, et al. Morphometric similarity networks detect microscale cor- Eur Neuropsychopharmacol 2016; 26: 810–27. tical organization and predict inter-individual cognitive variation. Viechtbauer W. Conducting meta-analyses in R with the metafor Neuron 2018a; 97: 231–47.e7. Package. J Stat Soft 2010; 36: 1–48. Seidlitz J, Váša F, Shinn M, Romero-Garcia R, Whitaker KJ, Vértes Vijayakumar N, Allen NB, Youssef G, Dennison M, Yucel M, PE, et al. Morphometric similarity networks detect microscale cor- Simmons JG, et al. Brain development during adolescence: a mixed- tical organization and predict inter-individual cognitive variation. longitudinal investigation of cortical thickness, surface area, and vol- Neuron 2018b; 97: 231–47.e7. ume. Hum Brain Mapp 2016; 37: 2027–38. Sharda M, Foster NEV, Tryfon A, Doyle-Thomas KAR, Ouimet T, Vriend C, van den Heuvel OA, Berendse HW, van der Werf YD, Anagnostou E, et al. Language ability predicts cortical structure and Douw L. Global and subnetwork changes of the structural connec- covariance in boys with autism spectrum disorder. Cereb Cortex tome in de novo Parkinson’s disease. Neuroscience 2018; 386: 2017; 27: 1849–62. 295–308. Shaw P, Sharp W, Sudre G, Wharton A, Greenstein D, Raznahan A, Vuoksimaa E, Panizzon MS, Chen CH, Fiecas M, Eyler LT, Fennema- et al. Subcortical and cortical morphological anomalies as an endo- Notestine C, et al. Is bigger always better? The importance of cor- phenotype in obsessive-compulsive disorder. Mol Psychiatry 2015; tical configuration with respect to cognitive ability. Neuroimage 20: 224–31. 2016; 129: 356–66. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller Wannan CMJ, Cropley VL, Chakravarty MM, Bousman C, Ganella EP, E, et al. The Mini-International Neuropsychiatric Interview Bruggemann JM, et al. Evidence for network-based cortical thickness (M.I.N.I.): the development and validation of a structured diagnostic reductions in schizophrenia. Am J Psychiatry 2019; 176: 552–63. psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry Watts DJ, Strogatz SH. Collective dynamics of ‘small-world’ networks. 1998; 59: 22–33; quiz 4-57. Nature 1998; 393: 440–2. Shin DJ, Jung WH, He Y, Wang J, Shim G, Byun MS, et al. The effects Wee CY, Yap PT, Shen D. Prediction of Alzheimer’s disease and mild of pharmacological treatment on functional brain connectome in ob- cognitive impairment using cortical morphological patterns. Hum sessive-compulsive disorder. Biol Psychiatry 2014; 75: 606–14. Brain Mapp 2013; 34: 3411–25. Silverman WK, Saavedra LM, Pina AA. Test-retest reliability of anx- Weinberg D, Lenroot R, Jacomb I, Allen K, Bruggemann J, Wells R, iety symptoms and diagnoses with the Anxiety Disorders Interview et al. Cognitive subtypes of schizophrenia characterized by differen- Schedule for DSM-IV: child and parent versions. J Am Acad Child tial brain volumetric reductions and cognitive decline. JAMA Adolesc Psychiatry 2001; 40: 937–44. Psychiatry 2016; 73: 1251–9. Solé-Casals J, Serra-Grabulosa JM, Romero-Garcia R, Vilaseca G, Yun JY, Jang JH, Kim SN, Jung WH, Kwon JS. Neural correlates of Adan A, Vilaro N, et al. Structural brain network of gifted children response to pharmacotherapy in obsessive-compulsive disorder: indi- has a more integrated and versatile topology. Brain Struct Funct vidualized cortical morphology-based structural covariance. Prog 2019; 224: 2373–83. Neuropsychopharmacol Biol Psychiatry 2015; 63: 126–33. Soriano-Mas C, Harrison BJ. Brain functional connectivity in Yun JY, Kim SN, Lee TY, Chon MW, Kwon JS. Individualized covari- Obsessive-Compulsive disorder. Oxford: Oxford University Press; ance profile of cortical morphology for auditory hallucinations in 2017. first-episode psychosis. Hum Brain Mapp 2016; 37: 1051–65. Soriano-Mas C, Harrison BJ, Pujol J, Lopez-Sola M, Hernandez-Ribas Zaremba D, Dohm K, Redlich R, Grotegerd D, Strojny R, Meinert S, R, Alonso P, et al. Structural covariance of the neostriatum with re- et al. Association of brain cortical changes with relapse in patients gional gray matter volumes. Brain Struct Funct 2013; 218: 697–709. with major depressive disorder. JAMA Psychiatry 2018; 75: 484–92. Spreng RN, DuPre E, Ji JL, Yang G, Diehl C, Murray JD, et al. Zhang T, Wang J, Yang Y, Wu Q, Li B, Chen L, et al. Abnormal Structural covariance reveals alterations in control and salience net- small-world architecture of top-down control networks in obsessive- work integrity in chronic schizophrenia. Cereb Cortex 2019; 29: compulsive disorder. J Psychiatry Neurosci 2011; 36: 23–31. 5269–84. Zhong Z, Zhao T, Luo J, Guo Z, Guo M, Li P, et al. Abnormal topo- Subira M, Cano M, de Wit SJ, Alonso P, Cardoner N, Hoexter MQ, logical organization in white matter structural networks revealed by et al. Structural covariance of neostriatal and limbic regions in diffusion tensor tractography in unmedicated patients with obses- patients with obsessive-compulsive disorder. J Psychiatry Neurosci sive-compulsive disorder. Prog Neuropsychopharmacol Biol 2016; 41: 115–23. Psychiatry 2014; 51: 39–50. Sussman D, Leung RC, Chakravarty MM, Lerch JP, Taylor MJ. The Zielinski BA, Gennatas ED, Zhou J, Seeley WW. Network-level struc- developing human brain: age-related changes in cortical, subcortical, tural covariance in the developing brain. Proc Natl Acad Sci USA and cerebellar anatomy. Brain Behav 2016; 6: e00457. 2010; 107: 18191–6.