13.1 Introduction

In 1977, Frans F Jöbsis pioneered a noninvasive method for measuring the hemodynamic oxygenation of biological tissue using near-infrared light (1). This method fostered a new era of near-infrared spectroscopy (NRIS) studies in the field of neuroscience. Over the last two decades, functional NIRS (fNIRS) has been applied to evaluate brain activation in humans in vivo and functional abnormalities in patients with psychiatric illnesses. Along with other functional neuroimaging modalities, such as functional MRI (fMRI), single-photon emission computed tomography (SPECT), and positron emission tomography (PET), studies using fNIRS to investigate mood disorders have been accumulating given the increasingly widespread use of NIRS in the study of psychiatric disorders. Novel and distinct imaging methods, such as fNIRS, will likely contribute to an increased understanding of brain pathophysiology in mood disorders. In this chapter, we discuss the principals of NIRS and its application in the study of mood disorders.

13.2 Principle of Near-Infrared Spectroscopy

Here, we summarize the principles of NIRS. Details of physiological, technical, and theoretical principles ((1–7)), and the main characteristics of NIRS ((8–10)) are described elsewhere.

Briefly, NIRS can measure the absorbance of light in certain tissues at several wavelengths in the 650–1,000 nm spectral range. It can also noninvasively and continuously quantify alterations in oxygenated (oxy-Hb) and deoxygenated (deoxy-Hb) hemoglobin. In the NIR spectral window (650–1,000 nm), which is called an “optical window,” human tissues are mainly transparent to light. Near-infrared light travels and scatters across tissues and is absorbed into them. The tissue oxymetry in NIRS was based on the modified Beer–Lambert law (1):

Attenuation(OD) =−log(I/I0) = εcLB+G,

where OD represents optical densities, I₀ represents the incident light intensity, I represents the detected light intensity, ɛ represents the absorption coefficient of the chromophore (mM⁻¹ cm⁻¹), c represents the concentration of chromophore (mM), L represents the physical distance between the points where light enters and leaves the tissue (cm), B represents a “path-length factor,” which takes into account the scattering of light into the tissue, and G represents a factor related to the measurement geometry and type of tissue. The absorption coefficients of oxy-Hb and deoxy-Hb have been measured in pure hemoglobin solutions.

For the measurement of brain tissue, NIR light diffusely penetrates the tissue layers of the head (skin, skull, and cerebrospinal fluid) beneath an optical probe. NIR light, attenuated in tissue, indicates the quantity of chromophore hemoglobin (the oxygen transport red blood cell protein) located in microcirculation vessels (<1 mm in diameter), such as capillary, arteriolar, and venular beds. The blood vessels >1 mm absorb the majority of NIR light. Then, a small amount of NIR light returns to the surface through the tissues of the head. Such an NIR light tract appears to form a “banana-shape” ((3)). Adequate depth of NIR light penetration (approximately one-half of the source-detector distance) can be achieved using a source-detector distance of around 30 mm. The detector optode detects NIR signals of hemodynamic changes in the cerebral cortex, as well as the skin ((11)). This signal depends on the distance between the source of NIR light and the detectors. If the distance of the source-detector is 20, 30, and 40 mm, the oxygenated change, resulting from cerebral activation, would be estimated to contribute 33.0, 54.8, and 68.5% of NIR signals detected, respectively ((12)). The units for oxy-Hb and deoxy-Hb signal changes should be expressed as μmolar*cm or mmolar*mm when the optical path-length of tissue is longer than the distance between the source and the detector, since the scattering effects of different tissue layers in the brain are unknown ((4)).

NIRS gauges changes in oxy-Hb and deoxy-Hb relevant to brain activity while excluding most of the effects of skin blood flow ((2)). An increase in oxy-Hb and corresponding decrease in deoxy-Hb, when linked with neural activity, is thought to indicate an increase in local arteriolar vasodilatation, local cerebral blood flow, and cerebral blood volume. This mechanism has been termed “neurovascular coupling.” Oxy-Hb is transported excessively to brain tissues where neuronal cells utilize the oxygen for their activity, resulting in an overabundance of cerebral blood oxygenation in the activated brain tissues ((13)). Oxy-Hb change, measured by NIRS, has been shown to be highly correlated with changes in regional cerebral blood flow, as determined by PET ((14)), and blood-oxygen-level-dependent (BOLD) signals, as determined by fMRI ((15–18)).

There are noteworthy strengths and weaknesses in the use of NIRS to evaluate brain function. The strengths include (1) near-infrared light is noninvasive to the brain and body, which means it is able to be used safely and repeatedly; (2) NIRS has high time-resolution of Hb data on the order of 100 ms; (3) participants are able to undergo NIRS examinations without stabled or restricted body positioning (e.g., measurements can be taken in a comfortable position in a chair or at the bedside); (4) NIRS devices are relatively compact and portable compared to MRI and CT devices; and (5) NIRS does not require a sealed and purpose-built room. The weaknesses include (1) NIRS only measures relative changes in Hb concentration; (2) NIRS only assesses the function of the inner surface of brain (e.g., dorsolateral prefrontal area), but not of deep brain structures (e.g., cingulate, subcortical area, and hippocampus/amygdala); (3) NIRS uses a target task combined with control tasks to eliminate confounding factors that may impact absorption of hemoglobin (e.g., skin, muscle, skull absorption); and (4) NIRS has a low spatial resolution on the order of 10−30 mm. It is of note that the low spatial resolution indicates that NIRS evaluates brain function in a specific area (e.g., inferior prefrontal “area”), but not anatomically accurate brain cortical structures or regions (e.g., inferior prefrontal “cortex”). However, a probable brain map for use with NIRS has been created ((19)). Thus, within this chapter, we describe anatomical locations using the term “area,” although some studies cited here within use the term “cortex.”

13.3 Application to Mood Disorders

To our knowledge, the first fNIRS study for patients with mood disorders was reported by Okada et al. ((20)). They examined brain activation in the left and right frontal areas during a mirror-drawing task, to assess visuospatial function, and found that patients with major depressive disorder (MDD) showed lower brain activation during the task than healthy subjects. To date, there are around fifty NIRS studies investigating mood disorders. Twenty-four studies (around 50%) used a single verbal fluency task (VFT), with letter version, eight used a single other cognitive task, and eleven used a combination of tasks including VFT and other cognitive and physiological tasks. Twenty-seven studies (57%) were conducted in patients with MDD versus healthy subjects, eight (17%) were conducted in patients with bipolar disorder (BD) versus healthy subjects, and twelve (26%) compared patients with MDD and BD and patients with other mood disorders and psychiatric illnesses.

Many fNIRS studies using the VFT, with letter version, were conducted in Japan. VFT, with letter version requires subjects to produce words beginning with a certain letter of the alphabet, usually “F,” “A,” and “S.” This task has been preinstalled in certain NIRS devices since 2009 because the physiological examination of fNIRS in the frontal area, using this task, was approved by the Japanese health ministry as an “advanced medical technology” to assist psychiatric diagnoses in 2009. As such, fNIRS was covered by public health insurance in Japan, since 2014, as a supplementary laboratory test for the differential diagnosis of BD and schizophrenia from that of MDD. The Nature News reports such diagnostic assisting methods in fascination ((21)).

Very recently, a meta-analytic study of fNIRS in MDD provided some evidence that patients at remitted and depressed states showed increased oxy-Hb during cognitive activation of the prefrontal areas compared to healthy participants (22). However, this analysis did not reveal a significant difference in change of oxy-Hb during cognitive activation between remitted and depressed patients.

13.3.1 Major Depressive Disorder versus Healthy Subjects

13.3.1.1 VFT, Letter Version

To the best of our knowledge, the first study of NIRS using the VFT was done by Matsuo et al. ((23)), and a modified and simplified version was used by Fukuda and his colleagues ((24)). This modified version is easier to use and has been broadly distributed (Figure 13.1). Thus, the procedure of VFT, with letter version, is very similar across recent studies in Japan ((25–39)). The VFT, with letter version, activates the frontal, and in particular, dorsolateral prefrontal cortex ((40, 41)) and is thought to assess cognitive response generation, working memory, and cognitive speed in the neuropsychological field ((42)). Cross-sectional fNIRS studies using the VFT often compare clinical features in depressed or euthymic patients with MDD such as melancholic versus non-melancholic ((30, 33)), suicidal versus non-suicidal ((39)), suicidal ideation versus non-suicidal ideation ((34)), vascular depression versus nonvascular depression ((29)), menopausal depression versus MDD ((43)), selective serotonin reuptake inhibitor responders versus nonresponders ((36)), positive and negative autonomic thought ((26)), and discrepancy between self-measured and observer-measured depression severity ((32)). The VFT studies also demonstrated associations between brain activation during the task with clinical variables such as depression severity ((27)), sleep quality ((37)), obsessive symptoms ((44)), dose of psychotropic medication [38], and stress-coping style ((45)). Furthermore, there are five longitudinal fNIRS that have investigated response to biological treatments including antidepressants ((31, 35, 36)) and electroconvulsive therapy ((25)). For instance, Tomioka et al. examined oxy-Hb changes during the VFT in medication-naive patients with MDD in the pre- and post-phase of antidepressant administration ((35)). They found that patients with MDD demonstrated blunted brain activation in frontotemporal areas, during the task, in both phases compared to healthy subjects. Although depressive symptoms were improved by the treatment, MDD patients in the pre-phase did not demonstrate differential brain activation compared to the same patients in the post-phase, which suggested that low frontotemporal activation during the task may be a trait characteristic of MDD. However, the results of fNIRS studies that use the VFT should be interpreted cautiously because these studies investigated similar comparison of a subset of patients and similar correlations with clinical variables, used the same task paradigm, and measured very similar areas (frontotemporal or frontal areas) due to them often being conducted for Japanese public health insurance-related reasons. Significant changes in oxy-Hb in the frontal and/or temporal brain areas were a common finding across studies regardless of the clinical characteristic found to be associated with this activation, which could indicate that a clinical characteristic implicated in a subgroup in one study may confound the results of a different study. For instance, wide frontotemporal area activation was significantly different between MDD patients with suicide ideas and healthy subjects ((34)); this same area, to some extent, also showed activation differences between MDD patients with melancholic features and healthy subjects ((33)). Further, NIRS studies using the VFT are required to replicate previous findings and to validate and further evaluate the effect of, or interaction between, these potentially confounding factors.

Figure 13.1 Change in oxygenated hemoglobin as measured by fNIRS

(a) A scene from an experiment using an NIRS device (ETG-4000, Hitachi Medical Co., Japan) in our laboratory. (b) Position of the probes for the source and detector and fifty-two channels in the frontotemporal area. The letter “S” represents a source probe, “D” represents a detector probe, and the number between S and D represents one of the channels measured. (c) A typical time course of oxygenated hemoglobin (oxy-Hb) change during a verbal fluency task, with letter version. The number represents one of the channels measured. (d) Expanding of the oxy-Hb change in the channel numbers 17 and 18, where it is estimated to be anatomically set in the left middle frontal area. A large increase of oxy-Hb change in a forty-eight-year-old healthy subject (black line) and small increase in oxy-Hb change in a thirty-five-year-old patient with major depressive disorder during a depressed state (gray line)

13.3.1.2 Other Tasks

Five cross-sectional and two longitudinal studies were conducted in patients with MDD using other tasks such as mirror drawing ((20, 46)), n-back ((45, 47)) to assess working memory and executive function, rock–paper–scissors ((48)) to evaluate cognitive inhibition ((49)), and arithmetic task ((46)). The other tasks included were physiological tasks used to evaluate microvascular sclerosis; these task included hyperventilation and paper-bag breathing ((50)), and carbon dioxide inhalation ((51)).

Matsuo et al. used an in-house VFT, with letter version, and carbon dioxide inhalation in older MDD patients during the full remission ((51)). The results demonstrated poor activation during the VFT and blunted hemodynamic response to carbon dioxide inhalation in frontal areas in patients when compared to healthy subjects. The authors suggested that prefrontal microvascular dysregulation, as shown in fNIRS, is involved in the trait pathophysiology of frontal hypofunction in later-life depression because the patients were at remitted state. This study demonstrates the effectiveness of fNIRS when combined with different tasks at the bedside. The use of different tasks to stimulate brain function may not be as easily done in other imaging modalities such as fMRI.

Related posts:

Chapter 1 – Brain Imaging Methods in Mood Disorders Chapter 2 – Neuroanatomical Findings in Unipolar Depression and the Role of the Hippocampus Chapter 7 – Functional Connectome in Bipolar Disorder Chapter 8 – Magnetic Resonance Spectroscopy Investigations of Bioenergy and Mitochondrial Function in Mood Disorders Chapter 15 – Magnetoencephalography Studies in Mood Disorders Chapter 16 – An Overview of Machine Learning Applications in Mood Disorders

Stay updated, free articles. Join our Telegram channel

Join