Inventory of Depression and Anxiety Symptoms (IDAS), Kiddie Schedule for Affective Disorders and Schizophrenia, Children’s Depression Rating Scale-Revised
Resting-state fMRI
Adolescents with MDD showed lower positive resting-state functional connectivity (RSFC) between the amygdala and hippocampus, parahippocampus, and brainstem (, corrected ). Patients also demonstrated greater (positive) amygdala-precuneus RSFC (, corrected ) in contrast to negative amygdala-precuneus RSFC in the adolescents serving as controls.
Lower coherence in the left executive control network (), anterior default mode network (), and salience network () was independently associated with greater lifetime severity of suicidal ideation.
MDD+low suicidality vs. MDD+high suicidality vs. healthy controls
Child Depression Rating Scale (CDRS), Moods and Feelings Questionnaire (MFQ)
Task-based fMRI (facial self-recognition task)
High-suicidality (HS) depressed subjects showed greater activity in the bilateral cuneus and occipital gyrus versus both low-suicidality (LS) depressed subjects and healthy controls (HC) (). HS and LS subjects showed greater activity in the right inferior frontal gyrus than HC subjects (). A group by self by emotion interaction with HS showed lower activity in the right hippocampus (), the left hippocampus and amygdala (), and the medial prefrontal cortex (mPFC) and Brodmann’s area (BA) 10 () in the happy self versus other face condition relative to the LS group, who had less activity in these regions compared to HC subjects. In a subsequent analysis that included depression severity as a covariate, a group by self-interaction showed that LS demonstrated higher midline cortical structure (MCS) activity than HS for the self versus other faces including in the left anterior lobe of the cerebellum and culmen (); the bilateral precuneus, posterior cingulate, and BA31 (); and the bilateral anterior cingulate, mPFC, BA32, and BA10 ().
Unipolar depression with various comorbid conditions (40% GAD, 16% ADHD, and 12% eating disorder NOS)
MDD (within group)
Inventory of Depression and Anxiety Symptoms (IDAS)
Resting-state fMRI
IDAS suicidality score was positively correlated with resting-state functional connectivity (RSFC) between the left precuneus and left pre- and postcentral gyri, superior and middle frontal gyri, and superior parietal lobe (-score range 3.10-3.56 for entire cluster). IDAS suicidality score was also positively correlated with RSFC between the right precuneus and the left pre- and postcentral gyri (-score range 3.22-3.62 for entire cluster), bilateral cerebellum (-score range 3.02-4.06 for entire cluster), and right precentral gyrus and middle, inferior, and superior frontal gyri (-score range 2.81-3.70 for entire cluster). IDAS suicidality score was negatively correlated with RSFC between the left posterior cingulate cortex and left cerebellum, lateral occipital cortex, and temporal-occipital fusiform gyrus (-score range 3.37-3.96 for entire cluster).
In this longitudinal study, the degree of reductions in clinical measures of subject lifetime suicidal ideation was correlated with the degree of increase in resting-state salience network (SN) coherence (,).
Schedule for Affective Disorders and Schizophrenia for School-Aged Children (K-SADS-PL)
Task-based fMRI (balloon analog risk task), reconstructed sMRI
Subjects showed relative increase in frontoparietal activation when choosing the win option as explosion probability increased in balloon analog risk task. Subjects with SI showed increased activity in the left precentral gyrus when choosing the win option and decreased activity when choosing the inflate option as explosion probability increased. Subjects without SI showed similar activity regardless of explosion probability. As explosion probability increased, subjects with SI showed greater orbitofrontal cortex activation to inflations, while subjects without SI had greater orbitofrontal cortex activation during explosions.
Axis I disorders not specified (20.4%), healthy subjects
Heterogeneous
Suicidal Ideation Questionnaire (SIQ)-Junior High Version; Implicit Association Task (IAT)-Death Version
sMRI
Reduced gray matter volume (GMV) of the bilateral putamen was significantly associated with higher IAT -scores (left: ; right: ). Reduced left caudate GMV was also significantly associated with higher IAT -scores ().
Composite score: Scale for Suicidal Ideation, Diagnostic Interview Schedule for Children Version IV (DISC-IV), Self-Injurious Thoughts and Behaviors Interview
Task-based fMRI (emotion regulation task)
Subjects with suicidal ideation (SI) demonstrated greater activation of the right dorsolateral prefrontal cortex (dlPFC) compared to subjects without SI () on trials in which they attempted to regulate their emotional responses to stimuli compared with trials in which they passively viewed negative emotional stimuli. During passive viewing of negative emotional stimuli, subjects with SI demonstrated less activation in the right dlPFC (), right temporoparietal junction (), right thalamus (), and left cerebellum/lateral occipital lobe () than age-matched controls and greater activation in the left temporal pole () than age-matched controls.
Schedule for Affective Disorders and Schizophrenia for School-Aged Children (K-SADS); SI Composite of the Mood and Feelings Questionnaire (MFQ-SI41); Children’s Depression Rating Scale-Revised (CDRS-R)
Task-based fMRI
Subjects with high levels of peer victimization or ecological momentary assessment-measured daily negative social experiences showed positive association between right anterior insula activation and suicidal ideation.
Schedule for Affective Disorders and Schizophrenia for School-Aged Children (K-SADS-PL)
Resting-state fMRI, sMRI
Subjects with caregiver-reported suicidal thoughts and behaviors had thinner superior temporal sulcus left banks when compared with the never-suicidal group (,,,). Subjects with a history of suicide attempts exhibited greater activation in the pallidum and decreased activation in the ventral diencephalon during inhibitory control tasks.
MDD+low SI vs. MDD+high SI vs. MDD+SA vs. healthy controls
Schedule for Affective Disorders and Schizophrenia for School-Aged Children (K-SADS-PL), Child Depression Rating Scale
Task-based fMRI (facial emotion processing task)
High-suicidality (HS) and suicide-attempting (SA) depressed subjects showed greater connectivity between the amygdala and the dorsolateral prefrontal cortex (dlPFC) (LS: , HS: , SA: ), dorsomedial prefrontal cortex (dmPFC) (LS: , HS: , SA: ), and precuneus (LS: , HS: , SA: ) than low-suicidality (LS) depressed subjects, who demonstrated a greater degree of amygdala-dlPFC connectivity than healthy controls (HC). Suicide-attempting depressed subjects showed greater connectivity between the left amygdala and the rostral anterior cingulate cortex (rACC) (HC<LS: ; HC<HS: : HC<SA, ) than all other groups. HS showed greater right amygdala-rACC connectivity than LS () and HC ().
Compared with nonsuicidal subjects and healthy controls, the suicide-attempting group exhibited increased zALFF (-score fractional amplitude of low-frequency fluctuation) in the right superior temporal gyrus (, 4.23, respectively), left middle temporal gyrus (, 4.46, respectively), and left middle occipital gyrus (, 3.09, respectively). Significantly decreased zALFF was observed in the suicidal group in the left superior frontal gyrus () and left middle frontal gyrus () compared to the nonsuicidal group.
History of suicide attempt(s) was significantly correlated with the presence of periventricular white matter hyperintensities (PVHs) (Fisher’s exact test: ), as well as right hemisphere PVHs (Fisher’s exact test: ). Logistic regression analyses controlling for age, sex, and several clinical risk factors supported this finding (; 95% CI: 1.6-21.2).
While healthy controls demonstrated a positive association between cortical thickness in the anterior left rostral middle frontal region (vmPFC) and BIS-11 motor impulsivity scores, participants with MDD and a history of at least one suicide attempt demonstrated a negative association between vmPFC cortical thickness and BIS-11 motor impulsivity scores (Cohen’s ,).
Columbia Classification Algorithm of Suicide Assessment (C-CASA); Kiddie Schedule for Affective Disorders and Schizophrenia; Structured Clinical Interview for DSM-IV Diagnosis (SCID)
sMRI
The following brain regions were found to have greater thickness in subjects with suicidal ideation than in those who had attempted suicide, and thickness was found to be negatively correlated with suicide attempt: right lateral orbitofrontal thickness (,;,), left fusiform thickness (,;,), left temporal pole volume (,;,), left lateral orbitofrontal thickness (,;,), left posterior cingulate thickness (,;,), right pars orbitalis thickness (,;,), right posterior cingulate thickness (,;,), and left medial orbitofrontal thickness (,;,).
Unipolar depression (treatment-resistant—failure to respond to ≥8-week SSRI trial)
MDD+SA vs. MDD-SA vs. healthy controls
Kiddie Schedule for Affective Disorders and Schizophrenia (KSADS)
sMRI
Compared with nonsuicidal depressed subjects and healthy controls, the depressed suicide-attempting group exhibited a significant reduction in right superior temporal gyrus (rSTG) volume ().
Columbia Classification Algorithm of Suicide Assessment
Task-based fMRI (facial emotion processing task)
In response to 50% intensity angry faces, suicide attempters showed significantly greater activity than nonattempters in the right anterior cingulate gyrus (,) and left dorsolateral prefrontal cortex (,) attentional control circuitry, right primary sensory cortex (,), and right middle temporal gyrus (,) and significantly greater activity than healthy controls in the left primary sensory cortex (,), while nonattempters had significantly lower activity than healthy controls in the right anterior cingulate gyrus (,) and right ventromedial prefrontal cortex (,). To neutral faces during the angry emotion processing task, suicide attempters had significantly lower activity than nonattempters in the left fusiform gyrus (,). Suicide attempters also showed significantly lower activity than healthy controls to 100% intensity happy faces in the left primary sensory cortex (,) and to neutral faces in the happy run in the right anterior cingulate (,) and left medial frontal gyrus (,). Psychophysiological interaction analyses revealed significantly reduced anterior cingulate gyral-insula functional connectivity to 50% intensity angry faces in suicide attempters compared to nonattempters (right insula: ,; left insula: ,) or healthy controls (right insula: ,; left insula: ,).
Columbia Classification Algorithm of Suicide Assessment
Task-based fMRI (Go/NoGo response inhibition and motor control task)
In a 3-group by 2-condition (“GoNoGo” response inhibition block versus “Go” motor control block) block-design whole-brain analysis, nonsuicide attempters with unipolar depression showed greater activity than depressed suicide attempters in the right anterior cingulate gyrus (). Depressed nonattempters, but not depressed attempters, showed significantly greater activity than healthy controls in the left insula () in response to GoNoGo response inhibition blocks.
Subjects with unipolar depression and a history of suicide attempt showed significantly reduced cortical volume in the right superior temporal gyrus (rSTG) than healthy controls (-value 6.08, ).
Columbia Classification Algorithm of Suicide Assessment
Task-based fMRI (Iowa gambling task)
During low-risk decisions in the Iowa gambling task (IGT), depressed nonsuicide attempters, but not depressed attempters, showed significantly greater hippocampal activation than healthy controls (). Attempters showed significantly greater activation than healthy controls to low-risk decisions in the left caudate (). During high-risk decisions in the IGT, nonattempters showed significantly greater activation than attempters in the right thalamus ().
Compared with healthy controls, depressed subjects with a history of suicide attempt showed increased connectivity in the left cerebellum (), right middle temporal gyrus (), right middle occipital gyrus (), and left middle frontal gyrus (), as well as decreased connectivity in the right posterior cingulate cortex (PCC) (). Compared to the nonsuicidal depressed subjects, the suicidal patients showed increased connectivity in the left cerebellum () and the left lingual gyrus (), as well as decreased connectivity in the right precuneus ().
Kiddie Schedule for Affective Disorders and Schizophrenia; Structured Clinical Interview for DSM-IV Diagnosis (SCID); Hamilton Depression Rating Scale; Beck Scale for Suicidal Ideation
sMRI
Compared to both healthy controls and subjects with MDD without history of suicide attempts, subjects with MDD with history of suicide attempts had larger bilateral hippocampal fissure volume.
Mixed: unipolar depression (52%) and other psychiatric diagnosis not specified (48%)
Heterogeneous +SA vs. heterogeneous-SA
History of prior attempt by clinical interview
sMRI
Deep white matter hyperintensities (DWMH) in the right posterior parietal lobe were observed significantly more frequently in the suicidal group compared to the nonsuicidal group, across all diagnostic groups ( test, ). Logistic regression analysis showed an odds ratio of 8.6 (95% ) for positive history of suicide attempt with the presence of parietal lobe DWMHs.
Mixed: unipolar depression (31.4%), bipolar disorder (22.9%), psychotic disorder NOS (15%), affective disorder NOS (8.5%), conduct disorder/ADHD (11.8%), other psychiatric disorder NOS (9.8%)
Heterogeneous+SA vs. heterogeneous-SA
Modified Pfeffer Rating Scale
sMRI
Subjects with unipolar depression and white matter hyperintensities (WMH) were found to have a significantly higher prevalence of prior suicide attempts compared to subjects without WMH across all diagnostic groups (72% versus 37%; Fisher’s exact test, ). Further logistic regression analysis showed that subjects with unipolar depression who also had WMH were 18.6 times (95% ) more likely to have had a past suicide attempt when compared with subjects with other psychiatric disorders, or with unipolar depression but no evidence of DWMH.
Bipolar disorder (55%) and unipolar depression (45%)
Bipolar disorder+SA vs. bipolar disorder–SA vs. MDD+SA vs. MDD-SA
Columbia Suicide History Form
sMRI and DTI
Across and within the bipolar disorder and unipolar depression groups, subjects with a history of suicide attempt showed significantly reduced gray matter volume in the left ventral prefrontal cortex (Brodmann’s areas 11 and 47) (), as well as reduced fractional anisotropy in the left frontotemporal white matter (including the uncinate fasciculus) (), compared to subjects without a history of suicide attempt. Numerical effect sizes were not reported.
(healthy controls); (comorbid borderline personality disorder and MDD)
Male (23%) and female (77%)
Subjects meeting criteria for both borderline personality disorder and MDD (50%); healthy controls (50%)
MDD+bipolar disorder vs. healthy controls
Lifetime Self-Destructiveness Scale (LSDS)
sMRI
Among the adolescent subjects with comorbid borderline personality disorder and MDD, the number of prior suicide attempts was associated with smaller overall volume of BA24 of the anterior cingulate gyrus (,) (averaged across gray and white matter), but not with the volume for any of the other anterior cingulate BAs assessed (BA25, BA23, BA29, and BA31). Among this group, a greater number of suicide attempts were also associated with greater white matter volume (,) but not gray matter volume in BA23 of the anterior cingulate.
Bipolar youth with a history of suicide attempt showed reduced left lateral orbitofrontal cortex (OFC) volumes compared to healthy controls (-value 5.94, ), but there was no significant difference between bipolar suicide attempters and bipolar nonattempters. Controls and bipolar nonattempters had significantly greater left (-value 4.44, p =0.017) and right (-value 13.11, ) lateral OFC cortical thickness than bipolar attempters.
sMRI, DTI, and task-based fMRI (facial emotion processing task)
Compared to the group without history of suicide attempt, the suicide-attempting group showed reductions in gray matter volume in the right orbitofrontal cortex (), right hippocampus (), and bilateral cerebellum (); white matter fractional anisotropy in the uncinate fasciculus (), left uncinate/ventral prefrontal region (), and right cerebellum (); and amygdala functional connectivity to left ventral prefrontal (happy face processing task (), neutral face processing task (), and fearful face processing task ()) and right rostral prefrontal cortex (neutral face processing task ()).
Abbreviations: ADHD = attention-deficit/hyperactivity disorder; BPD = bipolar disorder; DTI = diffusion tensor imaging; fMRI = functional magnetic resonance imaging; GAD = generalized anxiety disorder; HC = healthy controls; MDD = major depressive disorder; PTSD = posttraumatic stress disorder; RSFC = resting-state functional connectivity; SA = suicide attempt; SI = suicidal ideation; sMRI = structural magnetic resonance imaging.
