Anxiety and depression are widespread psychosis which are believed to affect cerebral metabolism, especially in frontal and temporal cortex. The comorbidity patients of anxiety and depression (A&D) have more serious clinical symptoms. Functional near-infrared spectroscopy (fNIRS) is a noninvasive modality used to monitor human brain oxygenation, and it could be considered as a potential tool to detect psychosis which may lead to abnormal cerebral oxygen status when the brain is activated. However, how sensitive the cerebral oxygenation response to the cortex activation and whether these responses are consistent at different stages of A&D or different regions still remains unclear. In this study, a conventional physiological paradigm for cortex activation, i.e., verbal fluency task (VFT), and a relatively new paradigm, i.e., high-level cognition task (HCT), were compared to detect A&D through a longitudinal measurement of cerebral oxygen status by fNIRS. The A&D patients at the acute, consolidation and maintenance stages as well as the healthy subjects participated in the VFT and HCT paradigms, respectively. For the VTF paradigm, the subject was instructed to answer questions of phrase constructions within 60 s. For the HCT paradigm, the subject was instructed to categorize items, logical reasoning, and comprehensive judgment and write down the answers within 60 s. For most of the subjects, the oxy-Hb is found to increase remarkably, accompanied with a relatively small reduction in deoxy-Hb when subject to both paradigms. The statistical analyses show a relatively large variability within any group, leading to the significant difference that was only found between A&D at the acute stage and healthy subjects in the temporal lobe region (p < 0.001). Nevertheless, HCT would activate more oxygen increment when compared with the VFT, with a large integral value in oxy-Hb. On average, the oxy-Hb integral value of the A&D patients differs substantially at different stages when subject to HCT paradigm. Moreover, the prefrontal lobe and temporal lobe responses were more consistent to the HCT paradigm rather than the VFT paradigm. Under the VFT paradigm, however, no remarkable difference in integral value was found among the three stages, either at the prefrontal lobe or at the temporal lobe. This study indicated that HCT, which is intensively involved in brain function, would activate more oxygenation changes in the cerebral cortex. Additionally, with good performance at distinguishing different stages according to the oxy-Hb criterion, the HCT has the potential to evaluate the therapeutic effects for A&D patients.

Diagnosis and treatment of the patients with major depression (MD) or the combined anxiety and depression (A&D) depend on the questionnaire, sometimes accompanied by tasks such as verbal fluency task (VFT). Functional near infrared spectroscopy (fNIRS) is emerging as an auxiliary diagnostic tool to evaluate brain function, providing an objective criterion to judge psychoses. At present, the conclusions derived from VFT or rest (non-task) studies are controversial. The purpose of this study is to evaluate if task performs better than non-task in separating healthy people from psychiatric patients. In this study, healthy controls (HCs) as well as the patients with MD or A&D were recruited (n = 10 for each group) to participate in the non-task and VFT tasks, respectively, and the brain oxygenation was longitudinally evaluated by using fNIRS. An approach of spectral analysis is used to analyze cerebral hemoglobin parameters (i.e., Oxy and Deoxy), characterizing the physiological fluctuations in the non-task and task states with magnitude spectrum and average power. Moreover, the standard deviation of oxygenation responses during the non-task was compared with the peak amplitude during the task, with the aim to explore the sensitivity of the VFT task to brain activation. The results show that there is no significant difference (p > 0.05) among the three groups in average power during non-task. The VFT task greatly enhanced the magnitude spectrum, leading to significant difference (p < 0.05) in average power between any of two groups (HC, MD, and A&D). Moreover, 40% patients with A&D have an intermediate peak (around 0.05 Hz) in the magnitude spectrum when performing the VFT task, indicating its advantage in characterizing A&D. We defined a rate of the non-task standard variation to the task peak amplitude (namely, SD-to-peak rate) and found that this rate is larger than 20% in 90% of the MD subjects. By contrast, only 40% HC subjects have an SD-to-peak rate larger than 20%. These results indicate that the non-task may not be sufficient to separate MD or A&D from HC. The VFT task could enhance the characteristics of the magnitude spectrum, but its intensity needs to be elevated so as to properly explore brain functions related to psychoses.