Supplementary Information

1. Exclusion of participants

Sample 1, Institute of Psychiatry, London, UK

Forty-eight male participants were recruited but two subjects were excluded due to claustrophobia during scanning, one subject due to current major depression and one subject due to motion artefacts.

Sample 2, CHU Montpellier, Montpellier, France

Eighty-nine male participants were recruited. Two suicide attempters were excluded due to claustrophobia and data from one patient controlwas incomplete.

Sample 3, CHU Montpellier, Montpellier, France

One hundred and five female participants were recruited. Three healthy controls were excluded due to excessive motion artefacts and panic attack. Twopatient controls were excluded due to current depression.Three suicide attempters were excluded: one due to past head trauma, another due to a current manic episodeand the third due to current depression.

2. Analyses

SPM8

For this VBM pipeline, structural data was analysed with SPM v.4667 without use of any other Matlab toolboxes or SPM plugins. SPM8 new segmentation function was used to segment participants’ T1 scans into gray matter, white matter, and cerebral spinal fluid component before applyingDiffeomorphic AnatomicalRegistration Through Exponentiated Lie Algebra (DARTEL) to create a study specific template. Individual subject data was subsequently normalized to MNI space while preserving tissue amount (modulation) such that ultimately modulated tissue volumes are compared.For the detailed stey-by-step settings we have used, see Ashburner 1. VBM data were smoothed using a 8mm full width at half maximum Gaussian kernel due to previous investigations suggesting increased sensitivity for smaller kernel size in larger sample size2.

References:

1.VBM tutorial. 2010, Accessed Date Accessed 2010 Accessed.

2.Shen S, Sterr A. Is DARTEL-based voxel-based morphometry affected by width of smoothing kernel and group size? A study using simulated atrophy. J Magn Reson Imaging 2013; 37(6): 1468-1475.

3.Ashburner J, Friston KJ. Voxel-based morphometry--the methods. NeuroImage 2000; 11(6 Pt 1): 805-821.

4.Ashburner J, Friston KJ. Unified segmentation. NeuroImage 2005; 26(3): 839-851.

FreeSurferSBM

Surface-based morphometry (SBM) analyses were conducted using the Freesurfer image analysis suite (version v 5.1.0) ( The technical details and references for these procedures are described in prior publications listed online ( Briefly, the processing includes motion correction5, removal of non-brain tissue using a hybrid watershed/surface deformation procedure6, automated Talairach transformation, segmentation of the subcortical white matter and deep gray matter volumetric structures7, 8intensity normalization9, tessellation of the gray matter/white matter boundary, automated topology correction10, 11, and surface deformation following intensity gradients to optimally place the gray/white and gray/cerebrospinal fluid borders at the location where the greatest shift in intensity defines the transition to the other tissue class12, 13.

Once the cortical models are complete, a number of deformable procedures can be performed for further data processing and analysis including surface inflation 14, registration to a spherical atlas which utilizes individual cortical folding patterns to match cortical geometry across subjects 15, parcellation of the cerebral cortex into units based on gyral and sulcal structure 16, 17, and creation of a variety of surface based data including maps of surface area, curvature and sulcal depth. This method uses both intensity and continuity information from the entire three dimensional MR volume in segmentation and deformation procedures to produce representations of cortical thickness, calculated as the closest distance from the gray/white boundary to the gray/CSF boundary at each vertex on the tessellated surface 13. The maps are created using spatial intensity gradients across tissue classes and are therefore not simply reliant on absolute signal intensity. These maps were used to measure surface area and cortical thickness for the relevant SBM analyses.

SBM data were smoothed using 20mm full width at half maximum surface Gaussian kernel to optimize sensitive toward small structural alterations18.

References:

5.Reuter M, Rosas HD, Fischl B. Highly Accurate Inverse Consistent Registration: A Robust Approach. NeuroImage 2010; 53(4): 1181-1196.

6.Segonne F, Dale AM, Busa E, Glessner M, Salat D, Hahn HK et al. A hybrid approach to the skull stripping problem in MRI. NeuroImage 2004; 22(3): 1060 - 1075.

7.Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 2002; 33: 341-355.

8.Fischl B, Salat DH, van der Kouwe AJW, Makris N, Ségonne F, Quinn BT et al.Sequence-independent segmentation of magnetic resonance images. NeuroImage 2004; 23(Supplement 1): S69 - S84.

9.Sled JG, Zijdenbos AP, Evans AC. A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE transactions on medical imaging 1998; 17(1): 87-97.

10.Fischl B, Liu A, Dale AM. Automated manifold surgery: constructing geometrically accurate and topologically correct models of the human cerebral cortex. IEEE Medical Imaging 2001; 20(1): 70-80.

11.Segonne F, Pacheco J, Fischl B. Geometrically accurate topology-correction of cortical surfaces using nonseparating loops. IEEE Trans Med Imaging 2007; 26: 518-529.

12.Dale A, Fischl B, Sereno MI. Cortical Surface-Based Analysis: I. Segmentation and Surface Reconstruction. NeuroImage 1999; 9(2): 179 - 194.

13.Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proceedings of the National Academy of Sciences of the United States of America 2000; 97(20): 11050-11055.

14.Fischl B, Sereno MI, Dale AM. Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system. NeuroImage 1999; 9(2): 195-207.

15.Fischl B, Sereno MI, Tootell RB, Dale AM. High-resolution intersubject averaging and a coordinate system for the cortical surface. Hum Brain Mapp 1999; 8(4): 272-284.

16.Desikan RS, Segonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage 2006; 31(3): 968-980.

17.Fischl B, van der Kouwe A, Destrieux C, Halgren E, Segonne F, Salat DH et al.Automatically parcellating the human cerebral cortex. Cerebral cortex 2004; 14(1): 11-22.

18.Lerch JP, Evans AC. Cortical thickness analysis examined through power analysis and a population simulation. Neuroimage 2005; 24(1): 163-173.

3. Definition of Regions-of-Interest

For all MRI analyses, we performed an a priori Region-of-Interest (ROI) approach. Since no single anatomical atlas can provide regional definitions in both volume and vertex space respectively used in VBM and SBM approaches, similar but not exactmatching ROIs were extracted based on the most commonly used atlases in eachsoftware.

SPM ROIs are anatomically defined with Brodmann’s Areas masks provided by the Wake Forest University PickAtlas software 3.0. ( Based on the atlas in PickAtlas, we defined four regions of interest, namely: orbitofrontal cortex (OFC) as BA47 and lateralpart of BA11;ventral lateral prefrontal cortex(VLPFC) as BA44 and 45; medial prefrontal cortex (MPFC) as medial part of BA11 and BA10, 24 and 32; dorsal prefrontal cortex (DPFC) as BA8, 9 and 46 (SI Figures 1 to 4).We expanded all ROIs by 1 pixel in all x-y dimensions in PickAtalas in 2D modeto ensure better coverage of gray matters.

Freesurfer requires different ROI definition because Brodmann areaatlas are typically defined in volume space. There is no known equivalency in vertex/surface space, which is the basis of SBM analysis. The most commonly used atlases in SBM are Desikan-Killiany Atlas for surface area and Destrieux Atlas for volume and thickness. There is no known one to one matching between the corresponding regions of Destrieux, Desikan or Brodmann area atlas.

Supplementary Figure 1:VLPFC ROI

Supplementary Figure 2: OFCROI

Supplementary Figure 3:MPFCROI

Supplementary Figure 4:DPFC ROI

For bilateral volume and thickness analyses with Destrieux atlas (shown in brackets), we considered OFC to be composed of orbital part of inferior frontal gyrus (G_front_inf-Orbital) and orbital gyrus (G_orbital); VLPFC ofpars triangularis (G_front_inf-Triangul) and opercularis (G_front_inf-Opercular) of inferior frontal gyrus;MPFC oftransverse frontopolar gyri and sulci(G_and_S_transv_frontopol), anterior and middle anterior part of cingulate gyrus and sulcus (G_and_S_cingul-Ant, G_and_S_cingul-Mid-Ant)and gyrus rectus(G_rectus); DPFC of middle and superior frontal gyrus.

For bilateral area analyses with the Desikan atlas, OFC was a combination of both lateral division of orbitalfrontal cortexand pars orbitalis of inferior frontal gyrus; VLPFC was composed of pars opercularis and pars triangularis of inferior frontal gyrus; MPFC of medial division of orbitofrontal cortex, frontal pole, rostraland caudaldivisions of anterior cingulate; DPFCof rostral middle frontal and superior frontal regions.

Volumeand areameasures are summed across these atlases’ defined sub-regions (i.e. giving total area, or volume of each ROI) whereas thicknessisaveraged across theses sub-regions (i.e. average thickness within each ROI).

1

Supplementary Table 1: Comparison of socio-demographic and clinical variables across the three participant groups in Sample 1 (top), Sample 2 (middle), andSample 3 (bottom)

Healthy
Controls
(n=18) / Patient
Controls
(n=14) / Suicide
Attempters
(n=12) / Omnibus
F / χ² / t / p / Post-Hoc
Male Gender, N (%) / 18 / (100) / 14 / (100) / 12 / (100) / - / - / -
Age, mean (SD) / 33.6 / (10.9) / 43.9 / (10.5) / 41.5 / (11.1) / 4.0 / 0.03 / HC < PC
Years of education, N (%) / 17.3 / (1.8) / 14.6 / (2.7) / 14.3 / (2.6) / 8.1 / 0.001 / HC > PC, SA
NART, mean (SD) / 0.79 / (0.12) / 0.81 / (0.10) / 0.70 / (0.12) / 3.4 / 0.04 / PC > SA
HDRS , mean (SD) / 1.1 / (1.5) / 2.9 / (2.1) / 2.2 / (2.2) / 3.7 / 0.03 / HC < PC
BDI , mean (SD) / 2.4 / (4.5) / 8.1 / (5.3) / 7.1 / (6.5) / 5.3 / 0.009 / HC < PC
Age at first mood episode, mean (SD) / - / - / 29.6 / (10.7) / 26.1 / (14.7) / 0.6 / 0.5 / -
Number of depressive episodes, mean (SD) / - / - / 3.8 / (3.3) / 2.7 / (2.3) / 8.4 / 0.3 / -
Bipolar disorder, N (%) / - / - / - / - / - / - / - / - / -
Number of hypo/manic Episode, mean (SD) / - / - / - / - / - / - / - / - / -
Anxiety disorders / - / - / 7 / (50.0) / 5 / (41.7) / 0.18 / 0.7 / -
OCD, current, N (%) / - / - / 0 / (0.0) / 0 / (0.0) / - / - / -
Alcohol/substance abuse, past, N (%) / - / - / 0 / (0.0) / 0 / (0.0) / - / - / -
BIS10, mean (SD) / 60 / (9.4) / 62 / (9.3) / 67 / (10.8) / 1.6 / 0.2 / -
Psychotropic medication, N (%) / - / - / 2 / (14.3) / 3 / (25.0) / 0.5 / 0.5 / -
Antidepressant, N (%) / - / - / 2 / (14.3) / 2 / (16.7) / 0.03 / 0.9 / -
Lithium, N (%) / - / - / 0 / (0.0) / 1 / (8.3) / 1.2 / 0.3 / -
Antipsychotics, N (%) / - / - / 0 / (0.0) / 0 / (0.0) / - / - / -
Anticonvulsivants, N (%) / - / - / 0 / (0.0) / 0 / (0.0) / - / - / -
Anxiolytics and hypnotics, N (%) / - / - / 0 / (0.0) / 1 / (8.3) / 1.2 / 0.3 / -
Age of first attempt, mean [min-max] / - / - / - / - / 28.0 / [13-52] / - / - / - / -
Number of suicidal attempts, mean [min-max] / - / - / - / - / 2.4 / [1-6] / - / - / - / -
Suicide intent scale, total score, most severe act, mean [min-max] / - / - / - / - / 18.3 / [13-26] / - / - / - / -
Risk rescue rating scale, total score, most severe act, mean [min-max] / - / - / - / - / 41.9 / [26-57] / - / - / - / -
Healthy Controls
(n=44) / Patient Controls
(n=29) / Suicide Attempters
(n=16) / Omnibus⁺
F / χ² / t / p / Post-Hoc
Male Gender, N (%) / 44 / (100.0) / 29 / (9.9) / 16 / (100) / - / - / -
Age, mean (SD) / 40.2 / (6.3) / 42.1 / (9.9) / 40.4 / (12) / 0.5 / 0.6 / -
Years of education, N (%) / 14.9 / (2.1) / 13.6 / (3.0) / 13.5 / (2.4) / 3.5 / 0.03 / -
NART, mean (SD) / 0.74 / (0.14) / 0.75 / (0.12) / 0.73 / (0.11) / 0.07 / 0.9 / -
HDRS , mean (SD) / 0.55 / (0.87) / 3.1 / (2.2) / 2.3 / (2.2) / 22 / <0.001 / HC < PC, SA
BDI , mean (SD) / 0.68 / (1.6) / 5.0 / (5.1) / 5.4 / (4.7) / 16 / <0.001 / HC < PC, SA
Age at first mood episode, mean (SD) / - / - / 25.6 / (9.2) / 30.0 / (13) / 1.3 / 0.2 / -
Number of depressive episodes, mean (SD) / - / - / 7.1 / (9.1) / 6.4 / (9.3) / 3.3 / 0.9 / -
Bipolar disorder, N (%) / - / - / 16 / (55) / 9 / (56) / 0.0 / 0.9 / -
Number of hypo/manic Episode, mean (SD) / - / - / 6.1 / (9.0) / 7.2 / (11) / 7.7 / 0.4 / -
Anxiety disorders / - / - / 8 / (28) / 5 / (31) / 0.07 / 0.8 / -
OCD, current, N (%) / - / - / 1 / (3.4) / 0 / (0.0) / 0.6 / 0.4 / -
Alcohol/substance abuse, past, N (%) / - / - / 20 / (69) / 5 / (31) / 5.9 / 0.02 / PC > SA
BIS10, mean (SD) / 68 / (8.5) / 73 / (12) / 75 / (11) / 4.0 / 0.02 / HC < SA
Psychotropic medication, N (%) / - / - / 17 / (59) / 10 / (63) / 0.06 / 0.8 / -
Antidepressant, N (%) / - / - / 7 / (24) / 5 / (31) / 0.3 / 0.6 / -
Lithium, N (%) / - / - / 6 / (21) / 4 / (25) / 0.1 / 0.7 / -
Antipsychotics, N (%) / - / - / 4 / (14) / 6 / (38) / 3.4 / 0.07 / -
Anticonvulsivants, N (%) / - / - / 6 / (21) / 3 / (19) / 0.02 / 0.9 / -
Anxiolytics and hypnotics, N (%) / - / - / 5 / (17) / 4 / (25) / 0.4 / 0.5 / -
Age of first attempt, mean [min-max] / - / - / - / - / 33.7 / [14-59] / - / - / -
Number of suicidal attempts, mean [min-max] / - / - / - / - / 1.5 / [1-4] / - / - / -
Suicide intent scale, total score, most severe act, mean [min-max] / - / - / - / - / 13.5 / [2-25] / - / - / -
Risk rescue rating scale, total score, most severe act, mean [min-max] / - / - / - / - / 38.6 / [25-50] / - / - / -
Healthy Controls
(n=20) / Patient Controls
(n=39) / Suicide Attempters
(n=39) / Omnibus⁺
F / χ² / t / p / Post-Hoc
Male Gender, N (%) / 0.0 / (0.0) / 0.0 / (0.0) / 0.0 / (0.0) / - / - / -
Age, mean (SD) / 36.1 / (6.9) / 35.9 / (7.9) / 38.0 / (9.9) / 0.7 / 0.5 / -
Years of education, N (%) / 15.3 / (1.7) / 14.5 / (2.0) / 13.9 / (1.9) / 3.4 / 0.04 / HC > SA
NART, mean (SD) / 0.65 / (0.12) / 0.67 / (0.10) / 0.67 / (0.11) / 0.2 / 0.8 / -
HDRS , mean (SD) / 1.6 / (1.8) / 4.2 / (2.2) / 3.8 / (2.2) / 11 / 0.001 / HC < PC, SA
BDI , mean (SD) / 0.7 / (1.3) / 5.0 / (5.2) / 4.4 / (3.7) / 7.7 / 0.001 / HC < PC, SA
Age at first mood episode, mean (SD) / - / - / 24.5 / (8.1) / 23.2 / (8.2) / 0.7 / 0.5 / -
Number of depressive episodes, mean (SD) / - / - / 3.9 / (8.9) / 6.1 / (8.6) / 27 / 0.03 / PC < SA
Bipolar disorder, N (%) / - / - / 14 / (36) / 21 / (54) / 2.5 / 0.1 / -
Number of hypo/manic Episode, mean (SD) / - / - / 2.2 / (8.1) / 3.6 / (9.5) / 10 / 0.5 / -
Anxiety disorders / - / - / 13 / (33) / 17 / (44) / 0.9 / 0.4 / -
OCD, current, N (%) / - / - / 2 / (5.1) / 0 / (0.0) / 2.1 / 0.2 / -
Alcohol/substance abuse, past, N (%) / - / - / 6 / (15) / 11 / (28) / 1.9 / 0.2 / -
BIS10, mean (SD) / 44 / (12) / 48 / (13) / 49 / (13) / 1.2 / 0.3 / -
Psychotropic medication, N (%) / - / - / 28 / (72) / 33 / (85) / 1.9 / 0.2 / -
Antidepressant, N (%) / - / - / 16 / (41) / 17 / (44) / 0.05 / 0.8 / -
Lithium, N (%) / - / - / 8 / (21) / 8 / (21) / 0.0 / 1.0 / -
Antipsychotics, N (%) / - / - / 2 / (5.1) / 11 / (28) / 7.5 / 0.006 / PC < SA
Anticonvulsivants, N (%) / - / - / 8 / (21) / 11 / (28) / 0.6 / 0.4 / -
Anxiolytics and hypnotics, N (%) / - / - / 9 / (23) / 15 / (39) / 2.2 / 0.1 / -
Age of first attempt, mean [min-max] / - / - / - / - / 25.1 / [11-43] / - / - / -
Number of suicidal attempts, mean [min-max] / - / - / - / - / 3.2 / [1-10] / - / - / -
Suicide intent scale, total score, most severe act, mean [min-max] / - / - / - / - / 13 / [5-26] / - / - / -
Risk rescue rating scale, total score, most severe act, mean [min-max] / - / - / - / - / 39 / [28-57] / - / - / -

Footnotes: BIS10: Barratt Impulsivity Scale Version 10; NART: National Adult Reading Test; HDRS: Hamilton Depression Rating Scale; BDI: Beck Depression Inventory; HC: Healthy Controls; PC: Patient Controls; SA: Suicide Attempters. N.S.: not significant.

1