Every year, one million individuals world over die from suicide . This is a serious global health problem, but suicide is difficult to predict and prevent in majority of the people at risk. Although the pathogenesis of suicide is multifactorial, the strongest predictor of risk for a future attempt is a history of suicidal behavior and presence of a psychiatric disorder, particularly a major depressive disorder (MDD) [2, 3]. Although most patients with MDD never attempt suicide, those who do frequently exhibit a diathesis for suicidal behavior , a notion formalized in the diathesis–stress model of suicidal behavior . According to this model, the proximal stressors leading to suicidal behavior are commonly psychiatric disorders together with acute psychosocial crises, whereas the components of the diathesis for suicidal behavior are impulsivity, aggression, and hopelessness.
In vivo brain imaging is a promising tool for identifying neurological correlates of the diathesis for suicidal behavior [5, 6]. A magnetic resonance imaging (MRI) study found that depressed patients at high suicide risk had a significantly thinner cortex in the left dorsolateral and ventrolateral prefrontal regions than those at a lower suicide risk . A recent MRI study indicated that suicide attempters (SAs) with a past history of mood disorders have reduced activity in the left ventrolateral prefrontal cortices compared with healthy controls (HCs) and with patient controls with a past history of mood disorders but not of suicidal behavior . Furthermore, functional MRI (fMRI) studies have shown that SAs can be distinguished from non-attempters (NAs) by specific frontal cortex activation patterns in response to angry and happy versus neutral faces  and by decreased activation in the medial prefrontal cortex while reading autobiographical accounts of recent suicide attempts compared with while reading neutral scripts . Single-photon emission computed tomography (SPECT) studies have found that SAs have significant perfusion deficits in the prefrontal cortex  and bilateral superior frontal regions  compared with the control subjects and in the left frontal region compared with controls and depressed NAs . In addition, a positron emission tomography (PET) study of patients with MDD revealed prefrontal hypofunction that was exacerbated by low serotonergic activity and was proportional to the lethality of past suicide attempts . One recent investigation found that compared with depressed individuals with only suicidal thoughts, depressed individuals without suicide thoughts and plans, and healthy controls, depressed individuals with suicide plans showed relative hypometabolism in the right middle frontal gyrus and right inferior parietal lobe . Thus, these neuroimaging studies have revealed altered brain structure and regional activity associated with vulnerability for suicide risk in psychiatric disorders [16, 17].
Multichannel near-infrared spectroscopy (NIRS) is a non-invasive optical technique that allows monitoring of hemodynamic changes related to cortical neural activity by measuring relative changes in oxygenated hemoglobin (oxy-Hb) and deoxygenated hemoglobin (deoxy-Hb). NIRS has several advantages over other neuroimaging techniques for psychiatric research . Near-infrared light is completely non-invasive, and therefore, there are no concerns regarding radiation exposure typically observed with other neuroimaging techniques, such as SPECT/PET. In addition, its easy applicability and high ecological validity make NIRS particularly suitable for psychiatric patients who may be afraid of tight enclosures (e.g., in MRI/PET scanners) or who exhibit motor restlessness that interferes with motion-sensitive imaging methods, such as MRI, SPECT, and PET. Furthermore, NIRS has a relatively high time resolution. Thus, it is a neuroimaging tool that can be applied clinically for assessing potential biomarkers in patients with psychiatric disorders. Many studies using NIRS have demonstrated that changes in mean oxy-Hb levels in frontotemporal regions induced by a verbal fluency task (VFT) are significantly lower in patients with MDD than in control subjects [19–23]. However, it is still unclear whether these NIRS signal changes are associated with a history of suicide attempts and a diathesis for suicidal behavior in MDD.
Here we hypothesized that patients with MDD and a history of suicidal behavior can be distinguished from patients with no such history and from non-psychiatric control subjects by measuring hemodynamic responses in frontotemporal regions using NIRS. We used phonemic verbal fluency as the cognitive task to stimulate changes in hemoglobin levels because it is easy to understand and execute for patients with psychiatric disorders who present with depressive symptoms, and poor performance in this task has been linked to suicidal behavior in patients with psychiatric disorders . In addition, we examined whether VFT-induced hemodynamic responses were associated with impulsivity, aggression, and hopelessness, which would support the validity of NIRS responses as a measure of vulnerability for suicide risk in MDD. Accordingly, the aim of this study was to explore frontotemporal hemodynamic responses in depressed patients with a history of suicide attempts using 52-channel NIRS.