
==== Front
101671285
44575
Biol Psychiatry Cogn Neurosci Neuroimaging
Biol Psychiatry Cogn Neurosci Neuroimaging
Biological psychiatry. Cognitive neuroscience and neuroimaging
2451-9022
2451-9030

36906445
10.1016/j.bpsc.2022.12.007
nihpa1881530
Article
Free-Water Diffusion Magnetic Resonance Imaging Differentiates Suicidal Ideation From Suicide Attempt in Treatment-Resistant Depression
Vandeloo Katie L.
Burhunduli Patricia
Bouix Sylvain
Owsia Kimia
Cho Kang Ik K.
Fang Zhuo
Van Geel Amanda
Pasternak Ofer
Blier Pierre
Phillips Jennifer L.
University of Ottawa Institute of Mental Health Research, Ottawa, Ontario, Canada (KLV, PBu, KO, ZF, AVG, PBl, JLP); Department of Cellular and Molecular Medicine, University of Ottawa, Ottawa, Ontario, Canada (KLV, PBu, PBl); Department of Psychiatry, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts (SB, KIKC, OP); Department of Neuroscience, Carleton University, Ottawa, Ontario, Canada (AVG, JLP); Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts (OP); Department of Psychiatry, University of Ottawa, Ottawa, Ontario, Canada (PBl, JLP); and the Department of Biochemistry, Microbiology and Immunology, University of Ottawa, Ottawa, Ontario, Canada (JLP).
Address correspondence to Jennifer L. Phillips, Ph.D., at Jennifer.Phillips@theroyal.ca.
19 9 2024
4 2023
20 12 2022
24 9 2024
8 4 471481
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
BACKGROUND:

Suicide attempt is highly prevalent in treatment-resistant depression (TRD); however, the neurobiological profile of suicidal ideation versus suicide attempt is unclear. Neuroimaging methods including diffusion magnetic resonance imaging–based free-water imaging may identify neural correlates underlying suicidal ideation and attempts in individuals with TRD.

METHODS:

Diffusion magnetic resonance imaging data were obtained from 64 male and female participants (mean age 44.5 ± 14.2 years), including 39 patients with TRD (n = 21 and lifetime history of suicidal ideation but no attempts [SI group]; n = 18 with lifetime history of suicide attempt [SA group]), and 25 age- and sex-matched healthy control participants. Depression and suicidal ideation severity were examined using clinician-rated and self-report measures. Whole-brain neuroimaging analysis was conducted using tract-based spatial statistics via FSL to identify differences in white matter microstructure in the SI versus SA groups and in patients versus control participants.

RESULTS:

Free-water imaging revealed elevated axial diffusivity and extracellular free water in fronto-thalamo-limbic white matter tracts of the SA group compared with the SI group. In a separate comparison, patients with TRD had widespread reductions in fractional anisotropy and axial diffusivity, as well as elevated radial diffusivity compared with control participants (thresholded p < .05, familywise error corrected).

CONCLUSIONS:

A unique neural signature consisting of elevated axial diffusivity and free water was identified in patients with TRD and suicide attempt history. Findings of reduced fractional anisotropy, axial diffusivity, and elevated radial diffusivity in patients versus control participants are consistent with previously published studies. Multimodal and prospective investigations are recommended to better understand biological correlates of suicide attempt in TRD.
==== Body
pmcTreatment-resistant depression (TRD) is conceptualized as significant unresolved symptoms following multiple attempted pharmacotherapy trials for major depressive disorder (MDD) (1). Patients with TRD have a higher burden of illness, poorer quality of life, worse occupational and social outcomes (2), and greater risk of mortality (3) than patients who are responsive to treatment. TRD may increase an individual’s likelihood of engaging in suicidal behaviors (2), and an estimated 30% of patients with TRD will attempt suicide in their lifetime (4).

Clinical prediction of suicide has not improved significantly over the last 50 years (5), and many cited risk factors for suicide are strongly predictive of ideation rather than attempts (6). A crucial avenue for suicide prevention is understanding the independent roles of suicidal thoughts, behaviors, and attempts (7) and the neurobiological progression from passive suicidal thoughts to highly lethal suicide attempts.

In vivo neuroimaging techniques may help to distinguish between suicidal ideation and attempts on a neurobiological level (8). Previous magnetic resonance imaging (MRI) studies have identified both regional and whole-brain differences in patients with and without suicide attempt history (9), such as reduced structural gray matter volume and abnormal fronto-thalamo-limbic functional connectivity (10,11). However, the directionality of structural and functional changes have been inconsistent. For example, structural volumes of the parietal lobe in MDD patients with a history of suicidal thoughts and/or attempts are reduced in some studies but elevated in others (12). Similarly, both increased and decreased functional activation of the dorsal prefrontal cortex has been reported in relation to suicidal thoughts and behaviors in MDD (8).

Diffusion MRI (dMRI) could allow for detection of more subtle morphological differences between suicidal ideation and suicide attempts, due to its ability to resolve subvoxel microstructural information (13). dMRI quantifies the magnetic resonance signal of the random dispersion of water molecules through white matter fiber tracts, which vary based on tissue architecture (14). Diffusion tensor imaging (DTI) analysis involves mathematical modeling of the direction and magnitude of diffusion at each voxel, from which microscopic tissue features can be inferred (15). The measurement of fractional anisotropy (FA) (the directional preference of water diffusion) is the most commonly reported, in addition to mean diffusivity (MD) (the overall mobility of water molecules in each voxel), axial diffusivity (AD) (the rate of diffusion parallel to the primary diffusion direction), and radial diffusivity (RD) (the rate of diffusion perpendicular to the primary diffusion direction) (16).

To date, few published DTI studies have examined white matter microstructure in MDD in relation to suicidal ideation (17–19) or suicide attempt (20–23). Collectively, these studies report abnormalities in white matter tracts including the inferior-frontal occipital fasciculus, uncinate fasciculus, insula, corona radiata, and anterior thalamic radiation. However, only two studies examined suicide attempt, one of which was uncorrected for multiple comparisons (21). While these investigations provide a foundation for further characterization of brain microstructure associated with suicide-related outcomes, no previous DTI studies have directly compared patients with suicidal ideation only to those with suicide attempt history.

DTI is most often conceptualized as a single diffusion tensor, where a 3 × 3 matrix depicts the strength and directionality of diffusion alongside microstructural white matter tracts (24). However, representation of diffusion using a single matrix is overly simplistic, as it may not fully characterize underlying neuropathological changes (14). For example, partial-volume effects occur when voxels are contaminated by signal from extracellular fluid in addition to brain tissue; DTI metrics are no longer tissue specific, as the diffusion signal will represent the weighted average of both white matter and free water (14). Postprocessing techniques such as free-water imaging can estimate the independent contribution of freewater and brain tissue in individual voxels. Signal contamination from extracellular fluid is removed, and the fractional volume of the free-water compartment can be approximated (25). Previous studies have suggested that free-water volume may provide insight into the pathology underlying isotropic changes (26). Although free-water dMRI has been used in recent studies with various neurological and psychiatric disorders, the application of free-water imaging in TRD and suicide is unique.

The objective of this study was to characterize standard diffusion metrics (i.e., FA, AD, and RD); free-water–corrected FA (FAT), AD (ADT), and RD (RDT); and the free-water index (FW) using tract-based spatial statistics (TBSS) and a MATLAB-based free-water imaging script. In our primary analysis, patients with a lifetime history of suicidal ideation but no suicide attempt (SI group) were compared with those with lifetime history of suicide attempt (SA group). A second analysis compared the entire patient group to age- and sex-matched healthy control participants. Differences in microstructural white matter and free-water–corrected diffusion metrics were expected in the SA group relative to the SI group and in patients compared with control participants.

METHODS AND MATERIALS

Participants

Participants included male and female outpatients with TRD (age range, 18–65 years) recruited through referrals to the Mood Disorders Research Unit at the University of Ottawa Institute of Mental Health Research and the Mood and Anxiety Program at the Royal Ottawa Mental Health Centre in Ottawa, Canada. Age- (±2 years) and sex-matched healthy control participants were recruited through community advertisement. For participants with TRD, a primary diagnosis of MDD and lack of excluding comorbid psychiatric disorders was confirmed using the Structured Clinical Interview for the DSM-5 Research Version (27). Treatment resistance was defined as lack of response to two or more consecutive medications for depression with different mechanisms of action in the current major depressive episode (28). Inclusion criteria for patients were presence of a major depressive episode of at least 6 months in length, a score of ≥ 25 on the Montgomery–Åsberg Depression Rating Scale (29), and endorsement of lifetime presence of suicidal ideation according to the Columbia Suicide Severity Rating Scale (C-SSRS) (C-SSRS suicide ideation severity score ≥ 1) (30). A history of manic, hypomanic, or mixed episode(s), comorbid posttraumatic stress disorder, obsessive-compulsive disorder, substance or alcohol use disorder, eating disorder(s), and/or psychotic disorder(s) were exclusion criteria. Control participants had no Axis I diagnoses, confirmed through the Structured Clinical Interview for the DSM-5 Research Version, and no history of suicidal ideation or attempts. For all participants, positive urine toxicology screen, pregnancy, body mass index ≥ 35, history of head injury with a loss of consciousness, major medical or neurological illness, and contraindications to MRI scanning were exclusion criteria. The study protocol was approved by the Royal Ottawa Mental Health Centre Research Ethics Board. All participants provided written informed consent.

Clinical Data Collection and Analysis

Self-report and clinician-rated assessments of depression and suicidal ideation severity were completed using the Montgomery–Åsberg Depression Rating Scale, Patient Health Questionnaire 9-item (31), C-SSRS, and Beck Scale for Suicidal Ideation (32). Suicide attempt history was assessed using the C-SSRS, with suicide attempt defined as a self-injurious act enacted with at least some intent or wish to die. Demographic and medical information as well as family history of depression and suicide attempt were obtained through self-report. Handedness was determined using the Edinburgh Handedness Inventory (33). For naturally cycling females, to minimize the effects of endogenous ovarian hormone fluctuations on MRI data (34), the date of start of last menstrual cycle was obtained and used to schedule MRI scans during the estimated follicular phase (approximately cycle days 1–10). Analysis of clinical and demographic information was performed using IBM SPSS Statistics (Version 27; IBM Corp.). Independent samples t tests and χ2 tests were used to explore continuous and categorical demographic and clinical variables as appropriate. Results were considered significant at p < .05.

Imaging Data Acquisition

Magnetic resonance images were acquired on a 3T Siemens positron emission tomography-magnetic resonance system (Siemens Biograph mMR; Siemens Corp.) using a 32-channel head coil. dMRI data were obtained using a single shot 2-dimensional echo-planar imaging pulse sequence in the axial plane. At the expense of susceptibility-induced distortions (resulting in potential signal loss or signal pile up in the phase encoding direction), echo-planar imaging allows for reduced scanning time (35). Diffusion-weighted images were acquired in the anterior-to-posterior direction (64 volumes of b = 1000 s/mm2), and a single nondiffusion volume was acquired (b = 0 s/mm2). To correct for susceptibility-induced distortions, a posterior-to-anterior volume was also acquired (b = 0 s/mm2). Diffusion-weighted image parameters included repetition time = 10,900 ms, echo time = 105.0 ms, field of view = 256 mm2, slice thickness = 2.0 mm, voxel size = 2.0 × 2.0 × 2.0 mm3, bandwidth = 1776 Hz/Px, and echo-planar imaging factor = 128. Prior to inclusion of the ninth participant, the positron emission tomography–magnetic resonance system underwent a software upgrade from VB20 to VE11P; however, an exploratory TBSS comparison of matched participants before and after scanner upgrade revealed no significant differences. Therefore, data from participants collected prior to the scanner upgrade were retained.

Image Preprocessing and Analysis

Diffusion-weighted images were preprocessed according to a Python-based framework (pnlNipype) (36). Briefly, brain masks were created using a convolutional neural network–based segmentation tool (37,38). Topup as well as eddy-current and motion-induced distortion correction were performed using FSL (RRID:SCR_003070) (39,40), followed by the fit of a diffusion tensor at each voxel across the whole brain. Free-water imaging was applied through an in-house MATLAB-based script (25). Using this method, conventional diffusion tensors were computed via a least squares fit, after which diffusion maps (including FA, AD, and RD) were calculated from the tensors. Furthermore, by fitting the 2-compartment free-water imaging model to the diffusion images using a regularized nonlinear fit, FW maps and FW-corrected diffusion maps (including FAT, ADT, and RDT) were derived. MD was not examined as MD and FW have an approximate 1:1 relationship; therefore, elevations in MD are conceptualized as elevations in free water in this context. Furthermore, free-water–corrected MD was not included in this analysis as its contrast is approximately uniform and therefore not biologically meaningful (41).

Voxelwise statistics of diffusion data were accomplished using TBSS from FSL (42–44). The general linear model was set to perform 2 independent sample t tests using mean-centered age and sex as covariates, including SI versus SA groups, and patients versus control participants. Threshold-free cluster enhancement was used (45), with the number of randomized permutations set to 5000. The FSL cluster tool was used to define independent white matter clusters (46). Results were corrected for multiple comparisons through familywise error (45) and reported at a significance threshold of p < .05.

RESULTS

Sample Characteristics and Clinical Measures

Between July 2017 and December 2021, 84 participants provided consent and were screened. Sixty-eight participants underwent an MRI scan, and 64 participants had analyzable dMRI data (Figure S1). Four patients were excluded from the analyses: 1 stopped the scan prematurely due to anxiety, 2 were excluded due to incidental findings, and 1 was excluded due to motion artifacts. The final dMRI dataset comprised 39 patients with TRD (including 21 in the SI group, and 18 in the SA group), and 25 age- and sex-matched healthy control participants.

Table 1 characterizes the demographic and clinical information for participants. Comparing the SA and SI groups and patients with control participants, there were no significant between-group differences in age, sex, handedness, body mass index, or marital status (all p > .05). Patient and control groups differed in terms of the highest level of education obtained (p = .001) and smoking history (p = .038). For all clinical scales, patients had significantly higher scores than control participants (all p < .001). A larger proportion of patients had a family history of depression than did control participants (p = .001).

Comparing course of illness and clinical status, the SI and SA groups did not differ significantly on age of MDD onset, single versus recurrent major depressive episodes, length of current major depressive episode, prevalence of comorbid psychiatric diagnoses, or current disability leave status (all p > 05) (Table 1). At the time of imaging, the groups did not differ significantly in past-week depression severity (Montgomery–Åsberg Depression Rating Scale and Patient Health Questionnaire 9-item) or suicidal ideation severity on the C-SSRS. The SA group had significantly higher self-reported past-week suicidal ideation severity on the Beck Scale for Suicide Ideation than the SI group (p = .015). Medications taken at the time of the MRI scan were similar between patient groups (Table S1).

Compared with the SI group, a significantly higher proportion of patients in the SA group had lifetime and past-year history of inpatient psychiatric hospitalization (p < .001 and p = .002, respectively) (Table 1). The SA group had significantly higher lifetime suicidal ideation severity (p < .001). More patients in the SA group had a family history of suicide attempt than in the SI group (p = .005). For the SA group, 44% had an attempt within the past year, and the mean time since the most recent attempt was 7.9 (±11.9) years. Table S2 characterizes the frequency, timeframe, method, and lethality of previous suicide attempts for the SA group.

Neuroimaging

TBSS analyses comparing SI versus SA groups revealed significant differences in AD, ADT, and FW in fronto-thalamo-limbic white matter tracts as described below (familywise error–corrected p < .05) (Table 2; Figure S2). Compared with the SI group, the SA group had elevated AD in the bilateral inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, left uncinate fasciculus, and anterior thalamic radiation (Figure 1A). With free-water correction applied, the SA group had increased ADT in the left anterior limb of the internal capsule, anterior thalamic radiation, posterior limb of the internal capsule, and superior corona radiata, as well as the body of the corpus callosum and bilateral corticospinal tract (Figure 1B). Finally, the SA group had elevated FW in the left uncinate fasciculus and left sagittal striatum, including the inferior fronto-occipital fasciculus and inferior longitudinal fasciculus (Figure 1C). No significant differences were identified for any other diffusion metrics, including FA, RD, FAT, or RDT.

The TBSS analysis comparing patients and control participants did not reveal significant differences in FW; however, between-group differences in both standard and free-water–corrected diffusion metrics were widespread. Significant differences were found for FA, AD, FAT, ADT, and RDT (familywise error–corrected p < .05) (Table 3; Figure S3). Relative to control participants, patients had clusters of reduced FA in the middle cerebellar peduncle, bilateral corticospinal tract, right inferior fronto-occipital fasciculus, superior cerebellar peduncle, and left anterior thalamic radiation (Figure 2A). Furthermore, patients had reduced AD in the right corticospinal tract, uncinate fasciculus, inferior fronto-occipital fasciculus, superior longitudinal fasciculus, and forceps minor (Figure 2A). With free-water correction, patients had reductions in FAT in the bilateral inferior longitudinal fasciculus, inferior fronto-occipital fasciculus, left superior longitudinal fasciculus, and posterior thalamic radiation (Figure 2B). Reductions in ADT were also noted in patients in the right inferior longitudinal fasciculus, inferior fronto-occipital fasciculus, posterior limb of the internal capsule, anterior thalamic radiation, bilateral corticospinal tract, and middle cerebellar peduncle (Figure 2B). Elevations in RDT, but not RD, were identified in the bilateral anterior thalamic radiation, superior longitudinal fasciculus, posterior thalamic radiation, inferior longitudinal fasciculus, and inferior fronto-occipital fasciculus of patients relative to control participants (Figure 2B).

DISCUSSION

Herein, we report a unique neural correlate of suicide attempt in TRD, consisting of elevated AD (both standard and free-water corrected) and FW in fronto-thalamo-limbic white matter tracts. Furthermore, we confer support for previously published DTI studies reporting widespread microstructural alterations in patients with depression compared with healthy control participants, as reflected by findings of reduced FA and AD (both standard and free-water corrected) and elevated RD (free-water corrected). To date, these results provide the first indication of elevated extracellular free water specifically associated with a history of suicide attempt in depression.

From a clinical perspective, the SI and SA groups did not differ in terms of age of MDD onset, length and recurrence of major depressive episodes, severity of depressive symptoms, comorbid diagnoses, or current medications. Such findings highlight the homogeneity of the patient sample and underscore the difficulty in predicting suicide attempt risk among patients with TRD experiencing suicidal ideation. Indeed, past literature on the association between suicidal ideation and attempt is inconsistent. For example, in a meta-analysis of 81 studies, presence of suicidal ideation increased the risk of suicide death 4-fold; however, this finding was not significant in the context of mood disorder (47). Nevertheless, the patient groups differed on several suicide-related outcomes including self-reported past-week suicidal ideation severity (Beck Scale for Suicide Ideation scores), lifetime and past-year psychiatric inpatient hospitalization, and family history of suicide attempt. Psychiatric inpatient admission often occurs at times of crisis when individuals might pose a threat of harm to themselves (48). The acute period following inpatient psychiatric care is one of the highest-risk periods for subsequent suicide attempt and death, and individuals hospitalized for suicidal ideation and behaviors appear to be at the highest risk (49). Greater family history of suicide attempt in the SA group is also in line with previous reports (50,51). Personal history of suicide attempt and family history of suicidal behavior represent two of the most important nonmodifiable risk factors for suicide, suggesting a potential biological vulnerability (52).

Neuroimaging studies may help elucidate neurobiological substrates differentiating suicidal ideation from suicide attempt (8). In the present study, TBSS and free-water imaging revealed several relevant white matter tracts associated with suicidal ideation and suicide attempt in the context of TRD. In line with these findings, previous DTI investigations in depression have identified the importance of the uncinate fasciculus (53), a large white matter association tract connecting frontal and temporal cortices (54). Closely associated with the uncinate fasciculus are the inferior longitudinal fasciculus and inferior fronto-occipital fasciculus, associative temporal-limbic tracts joining the frontal, parietal, and temporal lobes (55). Together, all 3 fiber tracts have been hypothesized to contribute to a cortical-subcortical microstructural disconnection syndrome in MDD (56). Among patients with suicide attempt history in the present study, clusters of elevated free water were localized primarily to these 3 tracts, as were AD findings (without free-water correction). With conventional DTI analyses, altered uncinate fasciculus white matter microstructure has been specifically implicated in unipolar and bipolar depression with suicide attempt history (23,57,58). As elimination of noise caused by partial volume effects significantly impacted DTI analyses, future studies using DTI techniques should be aware of partial volume effects and address them if possible.

The largest clusters showing altered tissue microstructure in the SA group in our free-water–corrected analyses were localized to the corticospinal tract, body of the corpus callosum, and anterior limb of the internal capsule. Reduced FA has been reported in patients with mood disorders and suicide attempt history in the body of the corpus callosum (22,23) and anterior limb of the internal capsule (59). Tractography has also shown reduced fiber projections between the anterior limb of the internal capsule and the medial frontal cortex, orbitofrontal cortex, and thalamus in patients with MDD and suicide attempt history (20). This finding may suggest disrupted cognitive and emotional processing due to cortical-thalamic interruptions in patients with a history of suicide attempt (20). Past work has also reported widespread white matter microstructural differences in patients with depression and suicidal ideation (17). Consistent with our findings in patients versus control participants, reduced FA and AD in the anterior thalamic radiation may lead to cognitive and affective dysfunction in depression (60) and are correlated with reduced FA in the superior longitudinal fasciculus (61). The superior longitudinal fasciculus has a role in the perception of emotion through bidirectional connections with the frontal, parietal, and temporal cortices (62). Taken together, overlapping findings from our study and previous reports strengthen our understanding of the microstructural correlates of suicidal ideation and suicide attempt in TRD.

Diffusion measurements were significantly influenced by free water, as evidenced by differing patterns of significantly altered white matter voxels between standard and free-water corrected–diffusion metrics. This replicates previous reports highlighting the ability of free-water correction to ameliorate erroneous diffusion signal and unmask results hidden by free-water contamination (63). Beyond this, FW was significantly elevated in the SA group compared with the SI group, which may provide evidence for the source of isotropic changes identified. For example, elevations in extracellular free water have been hypothesized to reflect neuroinflammation-associated edema (64,65). Multiple methodological approaches have provided supporting evidence for increased central and peripheral inflammation in relation to suicidal behavior (66). Increased microgliosis has been reported in the postmortem brain of individuals who died by suicide (67), a finding that appears specific to suicide rather than psychiatric diagnosis (68). Additionally, positron emission tomography studies have identified elevated microglial translocator protein (TSPO) binding in vivo in the anterior cingulate cortex, prefrontal cortex, and insula in patients with depression (69), with preliminary data suggesting increased TSPO binding specifically in those experiencing suicidal thoughts relative to patients without suicidal ideation (70). Finally, higher proinflammatory cytokine concentrations in blood and cerebrospinal fluid have been reported in individuals with depression and previous suicide attempt (71,72). Importantly, two previous studies conducted free-water imaging in nonresistant MDD (73,74), and neither identified FW changes in patients relative to matched control participants. Therefore, our results replicate these findings while suggesting that elevated free water may be specific to suicide attempt history in this cohort.

This study has several limitations that warrant consideration. First, the study was cross-sectional and suicide attempt history was retrospectively assessed. However, we thoroughly differentiated between suicidal ideation and suicide attempt history using the C-SSRS, which is critical to understanding the independent roles of these constructs (6,30). Nevertheless, the cross-sectional design does not allow for assessment of the potential predictive value of our results in identifying those who will progress from ideation to attempt, and we did not control for the severity of suicidal ideation at the time of the scan. Second, our sample size was modest. Despite this, significant results were obtained even with strict correction for multiple comparisons, which emphasizes the robustness of our results. Furthermore, the inclusion of clinically well-characterized patients and well-matched control participants strengthens the findings. For example, all previously published DTI investigations of suicide and depression focused on MDD rather than TRD (17,18,20,21); however, patients with TRD are at an elevated suicide risk (2). Third, patients were taking various medications at the time of imaging and potential medication effects on imaging correlates could not be analyzed. Fourth, our sample size did not allow for meaningful sex-based analyses. Regardless, all groups were well matched for sex, mean-centered sex was included as a covariate in imaging analyses, and menstrual cycle phase was considered during MRI scheduling (albeit without measurement of circulating hormone levels). Fifth, patient and control groups differed significantly in terms of the highest level of education obtained. The control sample may have been biased toward higher education as much of the recruitment was completed in a research hospital setting. Future studies may consider matching patients and control participants based on education. Sixth, history of psychiatric hospitalizations was proportionately higher in the SA group than in the SI group. Previous literature has reported an association between psychiatric hospitalization and diffusion metrics (75). The effects of psychiatric hospitalizations on dMRI were not investigated separately from suicide attempts due to their high correlation. With regard to neuroimaging methods, there are inherent limitations with TBSS, including the potential for rotational variance (76). TBSS does not account for the impact of edema and crossing fibers, which could impact the interpretation of FA (77). We addressed the former through free-water imaging; however, the impact of crossing fibers should be considered in future studies. More accurate estimation of the free-water measure could be achieved with multishell dMRI data, which was not available in this study. Finally, the addition of supplementary analyses such as tractography may further quantify fiber bundle characteristics (78).

In summary, suicide attempt is highly prevalent in the context of TRD; however, factors underlying the progression from suicidal ideation to suicide attempt are incompletely understood. In this study, standard and free-water–corrected diffusion metrics differed between SI and SA groups and between patients with TRD and suicidal ideation versus control participants. Using TBSS, altered white matter microstructure and elevated extracellular free water were identified in patients with TRD and suicide attempt history in fronto-thalamo-limbic tracts. As a group, patients with TRD and suicidal ideation had widespread alterations in white matter microstructure compared with healthy control participants across multiple diffusion metrics. Notably, the effect of free-water correction on diffusion metrics and the elevation of free water itself suggest a potential neurobiological mechanism leading to both anisotropic and isotropic white matter changes. To conclude, the findings reported herein strengthen the foundation for further characterization of neurobiological correlates of suicide-related outcomes in TRD.

Supplementary Material

2

ACKNOWLEDGMENTS AND DISCLOSURES

This work was supported by a grant from the University of Ottawa Medical Research Fund (to JLP), anonymous donor funding facilitated through the Ottawa Community Foundation and Royal Ottawa Foundation for Mental Health (to JLP), and imaging support from the University of Ottawa Institute of Mental Health Research (to JLP). KLV was supported by graduate scholarships from the Canadian Institutes of Health Research, the Ontario Ministry of Colleges and Universities, and the University of Ottawa.

The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

We thank Taylor Hatchard, Meagan Birmingham, Adel Farah, and Cyrus Maurice Sani for participant screening; Jeanne Talbot, Sandhaya Norris, and Charles Desfossés for medical support; and Natalia Jaworska for technical support. We acknowledge the contributions to data collection by former lab members Hassan Khan, Femi Carrington, Vanessa Zayed, Alyssa Stowe, and Dominique Vasudev.

A previous version of this article was published as a preprint on Research Square: https://doi.org/10.21203/rs.3.rs-1712962/v1.

Presented in part at the Annual Organization for Human Brain Mapping Meeting, June 19-23, 2022, Glasgow, Scotland; the 60th Annual Meeting of the American College of Neuropsychopharmacology, December 5-8, 2021, San Juan, Puerto Rico; and the 75th Annual Meeting of the Society of Biological Psychiatry, April 30-May 2, 2020, New York, New York.

PBI received grant funding and/or honoraria for lectures and/or participation in advisory boards for AbbVie, Bristol-Myers Squibb, Eisai, Elvium Life Sciences, Janssen, Lundbeck, Otsuka, Pierre Fabre Médicaments, Pfizer, Shire, and Takeda.

Figure 1. Elevated axial diffusivity (AD), free-water–corrected AD (ADT), and free-water index (FW) in patients with a history of suicide attempt compared with suicidal ideation: White matter tracts showing clusters of significantly elevated (A) AD, (B) ADT, and (C) FW in patients with a suicide attempt history compared with suicidal ideation history (familywise error–corrected p value < .05, corrected for multiple comparisons). Voxels with significant elevations in diffusion metrics are depicted in warm colors. Statistical images were projected onto a mean skeleton (green), and FSL’s TBSS-fill tool was used for visualization.

Figure 2. Reduced fractional anisotropy (FA), axial diffusivity (AD), free-water–corrected FA (FAT), free-water–corrected AD (ADT), and elevated free-water–corrected radial diffusivity (RDT) in patients with treatment-resistant depression and suicidal ideation compared with healthy control participants: White matter tracts showing areas of significantly reduced (A) FA, AD, (B) FAT, ADT, and elevated RDT in patients with treatment-resistant depression compared with healthy control participants (familywise error–corrected p value < .05, corrected for multiple comparisons). Voxels with significant elevations in diffusion metrics are depicted in warm colors, and voxels with significant reductions in diffusion metrics are depicted in cool colors. Statistical images were projected onto a mean skeleton (green), and FSL’s TBSS-fill tool was used for visualization.

Table 1. Demographic and Clinical Characteristics of Study Participants (N = 64)

Characteristic	SA Group, n = 18	SI Group, n = 21	SA vs. SI Group Comparison	Patients, n = 39	Control Participants, n = 25	Patient vs. Control Group Comparison	
Age, Years, Mean (SD)	46.5 (16.3)	44.6 (12.9)	t37 = 0.41, p = .68	45.5 (14.3)	43.0 (14.5)	t62 = 0.66, p = .52	
Sex, Female/Male, %	56%/44%	33%/67%	χ21 = 1.84, p = .18	44%/56%	40%/60%	χ21 = 0.08, p = .78	
Handedness, Right/Ambidextrous/Lefta, n	15/2/0	16/4/1	χ22 = 0.93, p = .63	31/6/1	21/3/1	χ22 = 0.25, p = .88	
Smoking Status: Nonsmoker/Former/Smoker, n	10/5/3	15/6/0	χ22 = 9.62, p = .008	25/11/3	23/2/0	χ22 = 6.67, p = .038	
Education: High School/College/Universityb, %	37%/19%/44%	29%/19%/52%	χ22 = 6.77, p = .034	32%/19%/49%	0%/0%/100%	χ22 = 18.51, p = .001	
Body Mass Index, Mean (SD)	27.6 (5.5)	27.5 (5.3)	t37 = 0.09, p = .93	27.5 (5.3)	25.1 (3.9)	t62 = 1.99, p = .051	
Marital Status: Not Married, n (%)	11 (61%)	7 (33%)	χ21 = 3.07, p = .08	18 (46%)	10 (40%)	χ21 = 0.23, p = .63	
Family History of Depressiona, n (%)	12 (71%)	13 (62%)	χ21 = 4.26, p = .04	25 (66%)	6 (24%)	χ21 = 10.54, p = .001	
Family History of Suicide Attempta, n (%)	5 (29%)	1 (5%)	χ21 = 7.90, p = .005	6 (16%)	1 (4%)	χ21 = 2.12, p = .15	
MADRS Total Score, Mean (SD)	35.9 (4.1)	33.1 (4.9)	t37 = 1.95, p = .059	34.4 (4.7)	0.3 (0.5)	t62 = 36.0, p < .001	
C-SSRS Suicidal Ideation Severity Past Week, Mean (SD)	2.4 (1.5)	1.8 (1.5)	t37 = 1.28, p = .21	2.1 (1.5)	0 (0)	t62 = 6.65, p < .001	
Beck Scale for Suicide Ideationc, Mean (SD)	16.5 (7.5)	10.0 (8.0)	t35 = 2.55, p = .015	13.0 (8.3)	0.1 (0.4)	t60 = 7.70, p < .001	
Patient Health Questionnaire 9b, Mean (SD)	20.7 (4.5)	20.8 (3.3)	t35 = 0.06, p = .95	20.7 (3.8)	0.6 (1.1)	t60 = 25.70, p <.001	
Beck Hopelessness Scalea, Mean (SD)	16.8 (3.9)	15.9 (3.6)	t36 = 0.70, p = .49	16.3 (3.7)	1.2 (1.5)	t61 = 19.0, p < .001	
Age at Onset of MDD, Years, Mean (SD)	27.4 (12.3)	30.9 (13.6)	t33 = 0.77, p = .45	–	–	–	
Major Depressive Episodes, Single/Recurrent, %	21%/79%	24%/76%	χ21 = 0.03, p = .87	–	–	–	
Length of Current Episode, Years, Mean (SD)	4.2 (5.1)	4.4 (3.7)	t32 = 0.09, p = .93	–	–	–	
Currently on Disability Leave, n (%)	7 (39%)	11 (52%)	χ21 = 0.71, p = .40	–	–	–	
History of Lifetime Inpatient Hospitalization, n (%)	15 (83%)	5 (24%)	χ21 = 34.29, p < .001	–	–	–	
History of Past-Year Inpatient Hospitalization, n (%)	8 (44%)	1 (5%)	χ21 = 9.30, p = .002	–	–	–	
Current Comorbid Diagnoses	
 Persistent depressive disorder, n (%)	16 (89%)	17 (81%)	χ21 = 1.01, p = .61	–	–	–	
 Panic disorder, n (%)	2 (11%)	0 (0%)	χ21 = 2.46, p = .12	–	–	–	
 Social anxiety disorder, n (%)	6 (33%)	4 (19%)	χ21 = 1.04, p = .31	–	–	–	
 Generalized anxiety disorder, n (%)	8 (44%)	11 (52%)	χ21 = 0.24, p = .62	–	–	–	
 Specific phobia, n (%)	1 (6%)	1 (5%)	χ21 = 0.01, p = .91	–	–	–	
 Attention-deficit/hyperactivity disorder, n (%)	1 (6%)	2 (10%)	χ21 = 0.22, p = .90	–	–	–	
C-SSRS, Columbia Suicide Rating Scale; MADRS, Montgomery–Åsberg Depression Rating Scale; MDD, major depressive disorder; SA, suicide attempt; SI, suicidal ideation.

a Data missing for 1 participant.

b Data missing for 2 participants.

c Data missing for 3 participants.

Table 2. Elevated AD, ADT, and FW in Patients With History of Suicide Attempt (n = 18) Compared with Suicidal Ideation (n = 21) Identified by TBSS

Diffusion Metric	Cluster	Number of Voxels	Maximum Intensity Voxel Coordinates	Hemisphere	Corresponding Tract	pFWE Value	Effect Size (Cohen’s d)a	95% CI	
AD	1	10,455	[−36, 7, −28]	L	UF/ILF	.030	1.48	0.76–2.19	
2	84	[28, −49, 19]	R	IFOF	.047	1.59	0.86–2.31	
3	74	[39, −43, −5]	R	ILF	.048	1.15	0.46–1.82	
4	44	[−38, 27, 10]	L	ATR	.049	1.08	0.39–1.76	
5	36	[−22, −78, −1]	L	IFOF	.050	0.95	0.28–1.61	
ADT	1	5655	[12, −28, −26]	R	CST	.020	1.37	0.66–2.06	
2	1277	[3, 1, 25]	–	Body of corpus callosum	.030	1.07	0.39–1.74	
3	586	[−16, 8, 2]	L	ALIC/ATR	.030	1.11	0.43–1.79	
4	188	[−25, 4, 20]	L	SCR	.040	1.16	0.47–1.83	
5	71	[−23, −15, 13]	L	PLIC/CST	.048	0.72	0.07–1.37	
FW	1	190	[−37, 6, −26]	L	UF	.040	1.45	0.74–2.16	
2	90	[−37, −14, −13]	L	ILF/IFOF	.047	1.59	0.86–2.31	
AD, axial diffusivity; ADT, free-water–corrected axial diffusivity; ALIC, anterior limb of the internal capsule; ATR, anterior thalamic radiation; CST, corticospinal tract; FW, free-water index; FWE, familywise error–corrected; IFOF, inferior fronto-occipital fasciculus; ILF, inferior longitudinal fasciculus; L, left; PLIC, posterior limb of the internal capsule; R, right; SCR, superior corona radiata; TBSS, tract-based spatial statistics; UF, uncinate fasciculus.

a Effect sizes were derived from the extracted mean value for the largest cluster in each tract.

Table 3. Reduced FA, AD, FAT, ADT, and Elevated RDT in Patients With Treatment-Resistant Depression and Suicidal Ideation (n = 39) vs. Healthy Control Participants (n = 25) Identified by TBSS

Diffusion Metric	Cluster	Number of Voxels	Maximum Intensity Voxel Coordinates	Hemisphere	Corresponding Tract	pFWE Value	Effect Size
(Cohen’s d)a	95% CI	
FA	1	41,781	[33, −48, 12]	R	IFOF	.003	1.61	1.03 to 2.18	
2	414	[9, −49, −30]	R	SCP	.041	1.42	0.85 to 1.97	
3	121	[−4, −37, −33]	L	CST/ATR	.046	1.30	0.74 to 1.85	
4	27	[4, −22, −35]	R	CST	.049	0.33	−0.20 to 0.87	
5	24	[21, −41, −35]	–	MCP	.049	0.76	0.23 to 1.27	
AD	1	25,682	[5, −32, −41]	R	CST	<.001	2.00	1.38 to 2.61	
2	410	[24, 16, −10]	R	UF/IFOF	.004	1.26	0.71 to 1.81	
3	114	[45, −5, 23]	R	SLF	.004	1.47	0.91 to 2.04	
4	24	[14, 31, 1]	–	Forceps minor	.004	1.28	0.73 to 1.83	
FAT	1	24,300	[40, −35, −13]	R	ILF/IFOF	<.001	1.95	1.34 to 2.56	
2	20,845	[−36, −51, −1]	L	PTR/IFOF/ILF	<.001	1.78	1.18 to 2.37	
3	77	[−47, −55, 0]	L	SLF	.007	1.44	0.88 to 2.00	
ADT	1	7342	[41, −36, −11]	R	ILF/IFOF	<.001	2.32	1.67 to 2.96	
2	469	[26, −26, 18]	R	CST/PLIC	.004	1.26	0.71 to 1.80	
3	133	[28, −32, −2]	R	ATR	.004	1.21	0.66 to 1.76	
4	113	[−13, −34, −38]	L/−	CST/MCP	.004	0.92	0.39 to 1.44	
RDT	1	20,575	[−34, −55, 7]	L	PTR/ILF/IFOF/SLF/ATR	.001	−1.72	−2.3 to −1.13	
2	17,714	[37, −51, −2]	R	PTR/IFOF/ILF	.001	−1.83	−2.42 to −1.23	
3	661	[15, 57, −4]	R	ATR	.020	−1.70	−2.23 to −1.11	
4	114	[44, −4, 28]	R	SLF	.040	−1.19	−1.73 to −0.64	
5	95	[−9, −9, 8]	L	ATR	.045	−1.46	−2.02 to −0.89	
AD, axial diffusivity; ADT, free-water–corrected AD; ATR, anterior thalamic radiation; CST, corticospinal tract; FA, fractional anisotropy; FAT, free-water–corrected FA; FWE, familywise error–corrected; IFOF, inferior fronto-occipital fasciculus; ILF, inferior longitudinal fasciculus; L, left; MCP, middle cerebellar peduncle; PLIC, posterior limb of the internal capsule; PTR, posterior thalamic radiation; R, right; RDT, free-water–corrected radial diffusivity; SCP, superior cerebellar peduncle; SLF, superior longitudinal fasciculus; TBSS, tract-based spatial statistics; UF, uncinated fasciculus.

a Effect sizes were derived from the extracted mean value for the largest cluster in each tract.

All other authors report no biomedical financial interests or potential conflicts of interest.

Supplementary material cited in this article is available online at https://doi.org/10.1016/j.bpsc.2022.12.007.
==== Refs
REFERENCES

1. Limandri BJ (2018): Treatment-resistant depression: Identification and treatment strategies. J Psychosoc Nurs Ment Health Serv 56 :11–15.
2. Mrazek DA , Hornberger JC , Altar CA , Degtiar I (2014): A review of the clinical, economic, and societal burden of treatment-resistant depression: 1996–2013. Psychiatr Serv 65 :977–987.24789696
3. Brenner P , Reutfors J , Nijs M , Andersson TM (2021): Excess deaths in treatment-resistant depression. Ther Adv Psychopharmacol 11 :20451253211006508.
4. Bergfeld IO , Mantione M , Figee M , Schuurman PR , Lok A , Denys D (2018): Treatment-resistant depression and suicidality. J Affect Disord 235 :362–367.29665520
5. O’Connor RC , Kirtley OJ (2018): The integrated motivational-volitional model of suicidal behaviour. Philos Trans R Soc B Biol Sci 373 :20170268.
6. Klonsky ED , May AM , Saffer BY (2016): Suicide, suicide attempts, and suicidal ideation. Annu Rev Clin Psychol 12 :307–330.26772209
7. Klonsky ED , Dixon-Luinenburg T , May AM (2021): The critical distinction between suicidal ideation and suicide attempts. World Psychiatry 20 :439–441.34505359
8. Schmaal L , van Harmelen AL , Chatzi V , Lippard ETC , Toenders YJ , Averill LA , (2020): Imaging suicidal thoughts and behaviors: A comprehensive review of 2 decades of neuroimaging studies. Mol Psychiatry 25 :408–427.31787757
9. Bani-Fatemi A , Tasmim S , Graff-Guerrero A , Gerretsen P , Strauss J , Kolla N , (2018): Structural and functional alterations of the suicidal brain: An updated review of neuroimaging studies. Psychiatry Res Neuroimaging 278 :77–91.29929763
10. Campos AI , Thompson PM , Veltman DJ , Pozzi E , van Veltzen LS , Jahanshad N , (2021): Brain correlates of suicide attempt in 18, 925 participants across 18 international cohorts. Biol Psychiatry 90 :243–252.34172278
11. Kang SG , Na KS , Choi JW , Kim JH , Son YD , Lee YJ (2017): Resting-state functional connectivity of the amygdala in suicide attempters with major depressive disorder. Prog Neuropsychopharmacol Biol Psychiatry 77 :222–227.28445688
12. Domínguez-Baleón C , Gutiérrez-Mondragón LF , Campos-González AI , Rentería ME (2018): Neuroimaging studies of suicidal behavior and non-suicidal self-injury in psychiatric patients: A systematic review. Front Psychiatry 9 :500.30386264
13. Grieve SM , Maller JJ (2017): High-resolution diffusion imaging: Ready to become more than just a research tool in psychiatry? Mol Psychiatry 22 :1082–1084.27725663
14. Alexander DC , Dyrby TB , Nilsson M , Zhang H (2019): Imaging brain microstructure with diffusion MRI: Practicality and applications. NMR Biomed 32 :e3841.29193413
15. Chanraud S , Zahr N , Sullivan EV , Pfefferbaum A (2010): MR diffusion tensor imaging: A window into white matter integrity of the working brain. Neuropsychol Rev 20 :209–225.20422451
16. Podwalski P , Szczygieł K , Tyburski E , Sagan L , Misiak B , Samochowiec J (2021): Magnetic resonance diffusion tensor imaging in psychiatry: A narrative review of its potential role in diagnosis. Pharmacol Rep 73 :43–56.33125677
17. Taylor WD , Boyd B , McQuoid DR , Kudra K , Saleh A , MacFall JR (2015): Widespread white matter but focal gray matter alterations in depressed individuals with thoughts of death. Prog Neuropsychopharmacol Biol Psychiatry 62 :22–28.25963377
18. Myung W , Han CE , Fava M , Mischoulon D , Papakostas GI , Heo JY , (2016): Reduced frontal-subcortical white matter connectivity in association with suicidal ideation in major depressive disorder. Transl Psychiatry 6 :e835.27271861
19. Reis JV , Vieira R , Portugal-Nunes C , Coelho A , Magalhães R , Moreira P , (2022): Suicidal ideation is associated with reduced functional connectivity and white matter integrity in drug-naïve patients with major depression. Front Psychiatry 13 :838111.35386522
20. Jia Z , Wang Y , Huang X , Kuang W , Wu Q , Lui S , (2014): Impaired frontothalamic circuitry in suicidal patients with depression revealed by diffusion tensor imaging at 3.0 T. J Psychiatry Neurosci 39 :170–177.24119793
21. Olvet DM , Peruzzo D , Thapa-Chhetry B , Sublette ME , Sullivan GM , Oquendo MA , (2014): A diffusion tensor imaging study of suicide attempters. J Psychiatr Res 51 :60–67.24462041
22. Cyprien F , de Champfleur NM , Deverdun J , Olié E , Le Bars E , Bonafé A , (2016): Corpus callosum integrity is affected by mood disorders and also by the suicide attempt history: A diffusion tensor imaging study. J Affect Disord 206 :115–124.27472413
23. Wei S , Womer FY , Edmiston EK , Zhang R , Jiang X , Wu F , (2020): Structural alterations associated with suicide attempts in major depressive disorder and bipolar disorder: A diffusion tensor imaging study. Prog Neuropsychopharmacol Biol Psychiatry 98 :109827.31778758
24. O’Donnell LJ , Westin CF (2011): An introduction to diffusion tensor image analysis. Neurosurg Clin N Am 22 :185–196.21435570
25. Pasternak O , Sochen N , Gur Y , Intrator N , Assaf Y (2009): Free water elimination and mapping from diffusion MRI. Magn Reson Med 62 :717–730.19623619
26. Lyall AE , Pasternak O , Robinson DG , Newell D , Trampush JW , Gallego JA , (2018): Greater extracellular free-water in first-episode psychosis predicts better neurocognitive functioning. Mol Psychiatry 23 :701–707.28348381
27. First MB (2015): Structured Clinical Interview for the DSM-V research version (SCID). In: The Encyclopedia of Clinical Psychology. Arlington, VA: American Psychiatric Association.
28. Sackeim HA (2001): The definition and meaning of treatment-resistant depression. J Clin Psychiatry 62 (suppl 16 ):10–17.
29. Montgomery SA , Asberg M (1979): A new depression scale designed to be sensitive to change. Br J Psychiatry 134 :382–389.444788
30. Posner K , Brown GK , Stanley B , Brent DA , Yershova KV , Oquendo MA , (2011): The Columbia-suicide severity rating scale: Initial validity and internal consistency findings from three multisite studies with adolescents and adults. Am J Psychiatry 168 :1266–1277.22193671
31. Kroenke K , Spitzer RL , Williams JB (2001): The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med 16 :606–613.11556941
32. Beck AT , Kovacs M , Weissman A (1979): Assessment of suicidal intention: The Scale for Suicide Ideation. J Consult Clin Psychol 47 :343–352.469082
33. Oldfield RC (1971): The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia 9 :97–113.5146491
34. Dubol M , Epperson CN , Sacher J , Pletzer B , Derntl B , Lanzenberger R , (2021): Neuroimaging the menstrual cycle: A multimodal systematic review. Front Neuroendocrinol 60 :100878.33098847
35. Poustchi-Amin M , Mirowitz SA , Brown JJ , McKinstry RC , Li T (2001): Principles and applications of echo-planar imaging: A review for the general radiologist. RadioGraphics 21 :767–779.11353123
36. Billah T , Bouix S , Rathi Y (2019): NIFTI MRI processing pipeline Available at: https://github.com/pnlbwh/pnlNipype. Accessed August 26, 2020.
37. Dey R , Hong Y (2018): Compnet: Complementary segmentation network for brain MRI extraction. In: Frangi J , Schnabel C , Davatzikos C , Alberola-Lopez GG , editors. Medical Image Computing and Computer Assisted Intervention – MICCAI 2018. Cham: Springer, 213–276.
38. Guha Roy A , Conjeti S , Navab N , Wachinger C , Alzheimer’s Disease Neuroimaging Initiative (2019): QuickNAT: A fully convolutional network for quick and accurate segmentation of neuroanatomy. Neuroimage 186 :713–727.30502445
39. Andersson JLR , Sotiropoulos SN (2016): An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. Neuroimage 125 :1063–1078.26481672
40. Andersson JL , Skare S , Ashburner J (2003): How to correct susceptibility distortions in spin-echo echo-planar images: Application to diffusion tensor imaging. Neuroimage 20 :870–888.14568458
41. Golub M , Neto Henriques R , Gouveia Nunes R (2021): Free-water DTI estimates from single b-value data might seem plausible but must be interpreted with care. Magn Reson Med 85 :2537–2551.33270935
42. Smith SM , Jenkinson M , Woolrich MW , Beckmann CF , Behrens TEJ , Johansen-Berg H , (2004): Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 23 (suppl 1 ):S208–S219.15501092
43. Smith SM , Jenkinson M , Johansen-Berg H , Rueckert D , Nichols TE , Mackay CE , (2006): Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage 31 :1487–1505.16624579
44. Billah T , Bouix S , Pasternak O (2019): Generalized tract based spatial statistics (TBSS) pipeline Available at: https://github.com/pnlbwh/tbss. Accessed September 15, 2020.
45. Smith SM , Nichols TE (2009): Threshold-free cluster enhancement: Addressing problems of smoothing, threshold dependence and localisation in cluster inference. Neuroimage 44 :83–98.18501637
46. Jenkinson M (2013): FSL cluster Available at: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Cluster. Accessed September 15, 2020.
47. Hubers AAM , Moaddine S , Peersmann SHM , Stijnen T , Van Duijn E , Van Der Mast RC , (2018): Suicidal ideation and subsequent completed suicide in both psychiatric and non-psychiatric populations: A meta-analysis. Epidemiol Psychiatr Sci 27 :186–198.27989254
48. Ziegenbein M , Anreis C , Brüggen B , Ohlmeier M , Kropp S (2006): Possible criteria for inpatient psychiatric admissions: Which patients are transferred from emergency services to inpatient psychiatric treatment? BMC Health Serv Res 6 :150.17121672
49. Jordan JT , McNiel DE (2018): Characteristics of a suicide attempt predict who makes another attempt after hospital discharge: A decision-tree investigation. Psychiatry Res 268 :317–322.30096659
50. Rihmer Z , Gonda X , Torzsa P , Kalabay L , Akiskal HS , Eory A (2013): Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J Affect Disord 149 :350–354.23477849
51. Lopez-Castroman J , Guillaume S , Olié E , Jaussent I , Baca-García E , Courtet P (2015): The additive effect on suicidality of family history of suicidal behavior and early traumatic experiences. Arch Suicide Res 19 :275–283.25259671
52. Turecki G , Brent DA (2016): Suicide and suicidal behaviour. Lancet 387 :1227–1239.26385066
53. Bhatia KD , Henderson LA , Hsu E , Yim M (2018): Reduced integrity of the uncinate fasciculus and cingulum in depression: A stem-by-stem analysis. J Affect Disord 235 :220–228.29656270
54. Ocklenburg S , Güntürkün O (2018): The Lateralized Brain: The Neuroscience and Evolution of Hemispheric Asymmetries. Cambridge, MA: Academic Press.
55. Wu Y , Sun D , Wang Y , Wang Y (2016): Subcomponents and connectivity of the inferior fronto-occipital fasciculus revealed by diffusion spectrum imaging fiber tracking. Front Neuroanat 10 :88.27721745
56. Liao Y , Huang X , Wu Q , Yang C , Kuang W , Du M , (2013): Is depression a disconnection syndrome? Meta-analysis of diffusion tensor imaging studies in patients with MDD. J Psychiatry Neurosci 38 :49–56.22691300
57. Johnston JAY , Wang F , Liu J , Blond BN , Wallace A , Liu J , (2017): Multimodal neuroimaging of frontolimbic structure and function associated with suicide attempts in adolescents and young adults with bipolar disorder. Am J Psychiatry 174 :667–675.28135845
58. Fan S , Lippard ETC , Sankar A , Wallace A , Johnston JAY , Wang F , (2019): Gray and white matter differences in adolescents and young adults with prior suicide attempts across bipolar and major depressive disorders. J Affect Disord 245 :1089–1097.30699851
59. Jia Z , Huang X , Wu Q , Zhang T , Lui S , Zhang J , (2010): High-field magnetic resonance imaging of suicidality in patients with major depressive disorder. Am J Psychiatry 167 :1381–1390.20843871
60. Lai CH , Wu YT (2014): Alterations in white matter micro-integrity of the superior longitudinal fasciculus and anterior thalamic radiation of young adult patients with depression. Psychol Med 44 :2825–2832.25065445
61. Jenkins LM , Barba A , Campbell M , Lamar M , Shankman SA , Leow AD , (2016): Shared white matter alterations across emotional disorders: A voxel-based meta-analysis of fractional anisotropy. NeuroImage Clin 12 :1022–1034.27995068
62. Nakajima R , Kinoshita M , Shinohara H , Nakada M (2020): The superior longitudinal fascicle: Reconsidering the fronto-parietal neural network based on anatomy and function. Brain Imaging Behav 14 :2817–2830.31468374
63. Berlot R , Metzler-Baddeley C , Jones DK , O’Sullivan MJ (2014): CSF contamination contributes to apparent microstructural alterations in mild cognitive impairment. Neuroimage 92 :27–35.24503415
64. Pasternak O , Kubicki M , Shenton ME (2016): In vivo imaging of neuroinflammation in schizophrenia. Schizophr Res 173 :200–212.26048294
65. Oestreich LKL , Lyall AE , Pasternak O , Kikinis Z , Newell DT , Savadjiev P , (2017): Characterizing white matter changes in chronic schizophrenia: A free-water imaging multi-site study. Schizophr Res 189 :153–161.28190639
66. Brundin L , Bryleva EY , Thirtamara Rajamani K (2017): Role of inflammation in suicide: From mechanisms to treatment. Neuropsychopharmacology 42 :271–283.27377015
67. Brisch R , Steiner J , Mawrin C , Krzyz_ anowska M , Jankowski Z , Gos T (2017): Microglia in the dorsal raphe nucleus plays a potential role in both suicide facilitation and prevention in affective disorders. Eur Arch Psychiatry Clin Neurosci 267 :403–415.28229240
68. Suzuki H , Ohgidani M , Kuwano N , Chrétien F , Lorin de la Grandmaison G , Onaya M , (2019): Suicide and microglia: Recent findings and future perspectives based on human studies. Front Cell Neurosci 13 :31.30814929
69. Setiawan E , Wilson AA , Mizrahi R , Rusjan PM , Miler L , Rajkowska G , (2015): Role of translocator protein density, a marker of neuroinflammation, in the brain during major depressive episodes. JAMA Psychiatry 72 :268–275.25629589
70. Holmes SE , Hinz R , Conen S , Gregory CJ , Matthews JC , Anton-Rodriguez JM , (2018): Elevated translocator protein in anterior cingulate in major depression and a role for inflammation in suicidal thinking: A positron emission tomography study. Biol Psychiatry 83 :61–69.28939116
71. Janelidze S , Mattei D , Å Westrin , Träskman-Bendz L , Brundin L (2011): Cytokine levels in the blood may distinguish suicide attempters from depressed patients. Brain Behav Immun 25 :335–339.20951793
72. Lindqvist D , Janelidze S , Erhardt S , Träskman-Bendz L , Engström G , Brundin L (2011): CSF biomarkers in suicide attempters - A principal component analysis. Acta Psychiatr Scand 124 :52–61.21198458
73. Bergamino M , Pasternak O , Farmer M , Shenton ME , Hamilton JP (2016): Applying a free-water correction to diffusion imaging data uncovers stress-related neural pathology in depression. NeuroImage Clin 10 :336–342.27006903
74. Bergamino M , Kuplicki R , Victor TA , Cha YH , Paulus MP (2017): Comparison of two different analysis approaches for DTI free-water corrected and uncorrected maps in the study of white matter microstructural integrity in individuals with depression. Hum Brain Mapp 38 :4690–4702.28631276
75. Meinert S , Leehr EJ , Grotegerd D , Repple J , Förster K , Winter NR , (2020): White matter fiber microstructure is associated with prior hospitalizations rather than acute symptomatology in major depressive disorder. Psychol Med 4 :1–9.
76. Bach M , Laun FB , Leemans A , Tax CMW , Biessels GJ , Stieltjes B , Maier-Hein KH (2014): Methodological considerations on tract-based spatial statistics (TBSS). Neuroimage 100 :358–369.24945661
77. Ye Z , Gary SE , Sun P , Mustafi SM , Glenn GR , Yeh FC , (2021): The impact of edema and fiber crossing on diffusion MRI metrics assessed in an ex vivo nerve phantom: Multi-tensor model vs. diffusion orientation distribution function. NMR Biomed 34 :e4414.33015890
78. Wang D , Luo Y , Mok VCT , Chu WCW , Shi L (2016): Tractography atlas-based spatial statistics: Statistical analysis of diffusion tensor image along fiber pathways. Neuroimage 125 :301–310.26481677
