The Diagnostic and Statistical Manual of Mental Disorders (DSM), which is a widely used classification system of mental health and disease, makes diagnostic deductions based on symptoms and behaviors reported by patients or observed by clinicians [1]. However, this diagnostic system is limited in that individuality is lost in evaluation, treatment, and prognosis because heterogeneously expressed symptoms are categorized as a part of a single disease entity. For instance, the diagnostic criteria for major depressive disorder in the DSM-5 includes nine heterogeneous symptoms: depressed mood, diminished interest, sleep disturbances, psychomotor agitation or retardation, fatigue, worthlessness, cognitive dysfunction, and suicidal ideation [1]. In addition, although the symptom presentation of patients might vary widely, the diagnosis of major depressive disorder depends on counting the number of symptoms, which might result in lack of consideration of individuality in the evaluation and treatment process.
Another limitation of classic diagnostic systems in psychiatry, unlike other fields of medicine, is a lack of incorporating neuroscientific technology developed in recent decades for diagnostic purposes [2]. Reflecting the concerns for these limitations, the DSM-5 adopted a dimensional approach regarding psychiatric disorders, viewing them as spectral entities rather than as strictly categorical, and new research results have been reported based on this revised classification system [1].
Besides the changes in diagnostic systems, the Research Domain Criteria (RDoC) project was proposed by the National Institute of Mental Health in 2010 as a framework to integrate information from various sources and domains, including symptoms, genetics, neuroscience, and physiology, as applied to understanding and conducting research regarding psychiatric disorders. RDoC is not a new classification system replacing the DSM, but a framework for the research of mental health and disease. RDoC includes six research areas (negative valence, positive valence, cognitive systems, systems for social processes, arousal/regulatory systems, and sensorimotor systems), with research being actively conducted worldwide [3]. For example, with respect to anxiety, there is an RDoC Anxiety and Depression project (RAD project), in which brain imaging has been used to establish and explain the association between brain-based constructs and anxiety symptoms [4].
Anxiety, one of the most common psychopathologies, is a complex physical and psychological emotional response. It is an unpleasant psychological state in which one might experience tension and irritability with the anticipation of a future threat [1]. It is often accompanied by various physiological responses of the autonomic nervous system, such as tachycardia, tremor, and dizziness as well as muscle tension and vigilance [1]. To assess the severity of anxiety symptoms in patients, the State-Trait Anxiety Inventory (STAI) is one of the most widely used self-reporting checklists for the assessment of anxiety symptoms [5]. The STAI consists of two subdomains: state anxiety evaluates the intensity of current feelings “at this moment,” and trait anxiety assesses the proneness to anxiety as a temperamental trait [5]. Although both share common features in that they assess the anxiety of an individual, they are reported to have some distinct neurobiological characteristics. For instance, a functional imaging study reported that while the anterior insula and basolateral amygdala constitute a network linked to both, state and trait anxiety are distinctively linked to dynamic functional and more static structural neural aspects, respectively [6]. Differences between state and trait anxiety are also observed in clinical tasks assessing the performance of executive function. [7] reported that higher trait anxiety predicted lower performance in a Stroop task evaluating executive function whereas higher state anxiety was associated with better performance.
The assessment of anxiety has relied primarily on self-report. However, in recent years, there have been increasing reports that electroencephalography (EEG) might be helpful in the assessment of the mental health disorders [8, 9]. In particular, quantitative electroencephalography (qEEG) is a neurological technique that analyzes the spectral band power of the delta (1–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13.5–30 Hz), and gamma (30–80 Hz) frequency bands. qEEG has several benefits over other neuroimaging techniques, in that it is non-invasive with no radiation exposure, it has excellent time resolution, and a low inspection cost [10, 11]. While qEEG can assist the physician in making a diagnosis, it can do more than simply detecting EEG abnormalities and helping form a diagnosis [11]. Specifically, qEEG can define subtypes of a disorder by identifying differential pathophysiological processes in patients who present with similar clinical symptoms [11, 12]. Its use in this way has been replicated in the assessment of various psychiatric disorders, such as providing probability estimates of the likelihood that a child may have a learning or attentional disorder [11, 12]. qEEG might also help in differentiating the categories of dementing disorders such as Alzheimer’s disease, vascular dementia, alcoholism, and delirium [11].
qEEG has also been applied in studies investigating disorders that are associated with anxiety, such as panic, posttraumatic stress (PTSD), and obsessive-compulsive disorders (OCD) [9, 13, 14]. For instance, studies have reported diminished alpha activity in anxiety disorder [15–17] and increased theta activity in OCD [18, 19]. However, compared to other psychiatric disorders, the literature on qEEG and anxiety disorder is small and difficult to translate to the clinical setting [11] due to the heterogeneity of symptoms among the disorders included in anxiety disorders [20].
Thus, the present study aimed to investigate the neural correlates of anxiety with qEEG focusing on the state and trait anxiety defined according to the RDoC framework existing across the differential diagnosis, rather than focusing on the diagnosis. In pursuing this purpose, we used the self-reporting of outpatients to examine anxiety symptoms and the various factors that influence it, such as socio-demographic factors and depressive symptoms. After controlling for these factors, we expected to find that anxiety symptoms were independently related to qEEG.