Prefrontal activation during a working memory task differs between patients with unipolar and bipolar depression: A preliminary exploratory study

Author’s Accepted Manuscript Prefrontal activation during a working memory task differs between patients with unipolar and bipolar depression: a preliminary exploratory study Yue Zhu, Wenxiang Quan, Huali Wang, Yantao Ma, Jun Yan, Hua Zhang, Wentian Dong, Xin Yu www.elsevier.com/locate/jad

PII: DOI: Reference:

S0165-0327(17)30033-2 http://dx.doi.org/10.1016/j.jad.2017.07.031 JAD9097

To appear in: Journal of Affective Disorders Received date: 6 January 2017 Revised date: 12 July 2017 Accepted date: 17 July 2017 Cite this article as: Yue Zhu, Wenxiang Quan, Huali Wang, Yantao Ma, Jun Yan, Hua Zhang, Wentian Dong and Xin Yu, Prefrontal activation during a working memory task differs between patients with unipolar and bipolar depression: a preliminary exploratory study, Journal of Affective Disorders, http://dx.doi.org/10.1016/j.jad.2017.07.031 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Prefrontal activation during a working memory task differs between patients with unipolar and bipolar depression: a preliminary exploratory study Yue Zhu a,1, Wenxiang Quan b,1, Huali Wang a,c, Yantao Ma d, Jun Yan d, Hua Zhang e, Wentian Dong b*, Xin Yu a,c* a

Clinical Research Division, Peking University Sixth Hospital, Peking University Institute of Mental Health, Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China b Department of Translational Medicine, Peking University Sixth Hospital, Peking University Institute of Mental Health, Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China c National Clinical Research Center for Mental Disorders (Peking University Sixth Hospital), Beijing, China d Inpatient Unit, Peking University Sixth Hospital, Peking University Institute of Mental Health, Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China e Research Center of Clinical Epidemiology, Peking University Third Hospital, Beijing, China [email protected] [email protected] *

Corresponding authors. Peking University Sixth Hospital, No. 51 Hua Yuan Bei Road, Haidian District, Beijing 100191, P. R. CHINA. Tel.: +86-010-82801999; fax: +86-010-82801999.

Abstract Background: To identify bipolar disorder during the initial stages of a depressive episode has always been a great clinical challenge. Patterns of functional brain activity may underlie the differences in the neural mechanisms of bipolar depression (BD) and unipolar depression (UD). This study aimed to investigate the differences in neural

1

These authors contributed equally to this study. 1

activity between BD and UD patients during executive task.

Methods: We performed a 52-channel near-infrared spectroscopy (NIRS) scan in 39 patients with BD, 35 patients with UD, and 36 healthy controls (HCs). The relative concentration changes in oxygenated hemoglobin ([oxy-Hb]) and deoxygenated hemoglobin ([deoxy-Hb]) during a 1-back working memory task were measured for each channel. Clinical characteristics including current mood were evaluated within one week prior to NIRS examination.

Results: Compared to HCs, BD (CH34: Z=-2.354, P=0.019) and UD patients (CH18: Z=-2.358, P=0.018; CH30: Z=-2.174, P=0.030; CH34: Z=-1.990, P=0.047) showed reduced activation of [oxy-Hb] in the inferior prefrontal region. Compared to patients with UD, patients with BD showed less decreased [oxy-Hb] changes in the left frontopolar cortex (FPC) (CH18: Z=-2.366, P =0.018), left pars opercularis and pars triangularis (POPE/PTRI) regions (Broca's area) (CH30: Z=-2.333, P=0.020). No correlation existed between clinical characteristics and NIRS measurements.

Limitations: The effect of medication could not be excluded, and behavioral data was not systematically collected. 2

Conclusion: The results from this preliminary exploratory study suggest distinct prefrontal activation patterns underlie BD and UD, especially in the left frontopolar region and Broca’s area. The NIRS-based prefrontal activation measurement may serve as a potential marker to aid in differentiating bipolar from unipolar depression. Abbreviations: BD, Bipolar depression; UD, Unipolar depression; HCs, Healthy controls; NIRS, Near-infrared spectroscopy; [oxy-Hb], Oxygenated hemoglobin; [deoxy-Hb], Deoxygenated hemoglobin; CH, Channel; FPC, Frontopolar cortex; POPE/PTRI, Pars opercularis and Pars triangularis.

Keywords: Bipolar, Unipolar, Depression, Working memory (WM) , Near-infrared spectroscopy, (NIRS) , Prefrontal

3

1. Introduction The early and accurate diagnosis of mood disorders is critical for subsequent treatment and prognosis. However, the similar psychopathologies of depressive episodes pose a great challenge for the differentiation of bipolar from unipolar depression (or major depressive disorder) during the depressed phase (Judd et al., 2003, 2002; Zimmermann et al., 2009). Bipolar patients with depressive episodes are often misdiagnosed and initially treated as unipolar (Akiskal et al., 1995; Goldberg et al., 2001), which leads to the delay of appropriate treatment, poorer outcome, and higher health care costs (Dudek et al., 2013; Hirschfeld et al., 2003; Perlis, 2005). The key clinical question is how to identify bipolar disorder among depressed individuals, which will subsequently improve the outcome for patients with bipolar disorder. Several strategies including clinical, neurocognitive and neuroimaging methods have been applied to make a differential diagnosis during a depressive episode. While clinical assessments such as rating scales and questionnaires might identify unrecognized subthreshold hypomanic symptoms, they are unable to provide a distinction between bipolar and unipolar depression at the first depressive episode (de Almeida and Phillips, 2013). Previous neurocognitive work so far has demonstrated inconsistent findings. Although most studies (Borkowska and Rybakowski, 2001; Gruber et al., 2007; Maalouf et al., 2010) reported poorer cognitive function in bipolar depression (BD), others (Daniel et al., 2013; Godard et al., 2012; Scott et al., 2013) indicated an unspecific deficiency that did not allow differentiating between BD and recurrent unipolar depression (UD). Our research team observed negative results as 4

well, finding quite similar patterns of cognitive deficiency exist in BD and UD (Zhu et al., 2013). Brain imaging has also been used to help improve the understanding of the pathophysiological processes involved in mood disorders. Different patterns of brain functional activities have been found in bipolar versus unipolar individuals during resting-state (Liang et al., 2013) or task-based (Almeida et al., 2010; Cerullo et al., 2014; Chase et al., 2013; Diler et al., 2013) functional magnetic resonance imaging (fMRI) studies. However, poor accessibility and the high cost of fMRI testing limit its routine clinical use. Moreover, disadvantages such as restraint position, noise and the long duration of examination are especially intolerable for psychiatric patients. Functional near-infrared spectroscopy (fNIRS) is a recently developed, functional neuroimaging method allowing noninvasive detection of superficial brain activity by measuring the absorption of near-infrared light (Boas et al., 2004; Strangman et al., 2002). NIRS enables real-time monitoring of changes in the concentration of oxygenated hemoglobin ([oxy-Hb]) and deoxygenated hemoglobin ([deoxy-Hb]) in the micro-blood vessels of surface brain tissue，which are indicators of regional cortical activity (Hoshi et al., 2001; Obrig and Villringer, 2003; Strangman et al., 2002). Additionally, NIRS-Hb signals reflect the regional cerebral blood flow (rCBF) changes (Hock et al., 1997; Ohmae et al., 2006; Villringer et al., 1997) and are also comparable to the blood-oxygenation-level-dependent (BOLD) signals in gray matter (Sato et al., 2013). Because its performance is convenient and economical, NIRS could serve as a bedside technique and is well tolerated by psychiatric patients. 5

A growing number of studies have used NIRS with an executive function task to investigate brain functional activity in unipolar (Koseki et al., 2013; Noda et al., 2012; Pu et al., 2012b, 2011) or bipolar individuals (Kameyama et al., 2006; Matsuo et al., 2007; Nishimura et al., 2015a), but only a few (Matsubara et al., 2014; Ohtani et al., 2015; Schecklmann et al., 2011; Takei et al., 2014) have compared the difference in brain activation patterns between the two disorders. Working memory (WM) is one process of executive function that is impaired by both unipolar and bipolar depression (Cremaschi et al., 2013; Harvey et al., 2004; Rose and Ebmeier, 2006), and it refers to the system or systems that are assumed to be necessary to keep things in mind while performing complex tasks (Baddeley, 2010). A quantitative meta-analysis (Owen et al., 2005) suggested that the prefrontal cortex and frontopolar regions were robustly activated during n-back WM paradigm variants. To our knowledge, no NIRS study has applied the n-back task to investigate brain activity in unipolar and bipolar individuals. The only NIRS study (Schecklmann et al., 2011) that has been conducted showed unspecific deficits in BD and UD using a visual WM task, and indicated that this paradigm was probably too difficult. Our previous work (Zhu et al., 2013) also demonstrated similar WM performance with no significant difference in BD and UD. Thus, we performed the present study to examine whether there could be some differences in regional functional activity between BD and UD, while no difference exists in WM behavioral performance. We hypothesized that patients with BD exhibited significantly decreased prefrontal activation compared to healthy controls, and a less deactivation compared 6

to UD in response to the WM task. The aim of the present study was to test our hypothesis and examine whether prefrontal activation patterns can distinguish BD from UD. Furthermore, we will discuss the association between prefrontal activation and depression severity.

2. Materials and methods Written informed consent was obtained from all participants after providing comprehensive information about the study. The study procedures were approved by the Research Ethics Committee of Peking University Sixth Hospital (Beijing, China) and were in accordance with the Helsinki Declaration as revised 1989. Each individual in the study was compensated for their time and participation with a 50-yuan gift.

2.1. Subjects Thirty-nine patients with bipolar depression (BD) and thirty-five with recurrent unipolar depression (UD) were recruited from inpatient units of Peking University Sixth Hospital between October 2013 and May 2014. All research participants aged between 18-60 years and were right-handed. With the Mini International Neuropsychiatric Interview (M.I.N.I., Chinese version; Si et al., 2009), diagnosis of UD and BD with a current major depressive episode (296.5X, 296.89 or 296.3X) was made according to the Diagnostic and Statistical Manual of Mental Disorders, 4th. ed. (DSM-IV-TR) (APA, 2000). All subjects scored >=14 on the 17-item Hamilton 7

Depression Rating Scale (HAMD-17; Hamilton, 1967). Since bipolar depression is frequently misdiagnosed as unipolar depression during the initial stages of a depressive episode (Hirschfeld et al., 2003), in the present study, the illness duration of UD should be no less than five years to maintain the relative diagnostic stability (Morselli and Elgie, 2003). In addition, thirty-six healthy controls (HCs) participated, having no history of major psychiatric or neurological illness and no prescription of psychotropic medication. In order to decrease the heterogeneity of a depressive phase, those who manifest subthreshold bipolarity (score of Young Mania Rating Scale >5) were excluded (Angst et al., 2003; YMRS, Young et al., 1978; Zimmermann et al., 2009)

in order

to decrease the heterogeneity within a diagnostic group. The exclusion criteria also included: having current and previous substance dependence or abuse, having a history of electroconvulsive therapy within 6 months, having any physical or mental illness that may affect neurocognitive assessment or NIRS examination, and being unable to perform the cognitive task. To obtain detailed information on psychiatric symptoms, the participants were interviewed by two experienced psychiatrists (Y. Zhu and W. Quan) using MINI. Affective symptom severities were evaluated by well-trained psychiatrists using the Chinese version of the HAMD-17, YMRS and clinical global impression-bipolar version (CGI-BP; Spearing et al., 1997) within one week prior to each NIRS examination.

8

2.2. Activation task We used a 1-back version of the n-back working memory (WM) task to activate brain regions, which was aimed at enabling most patients to manage the task even with the cognitive impairment and decreased self-confidence that is associated with depression. The paradigm consisted of a 30 s pre-task baseline period, a 70 s task period and a 50 s post-task baseline period (Pu et al., 2012b, 2011). During the pre-task and post-task baseline condition, participants were instructed to fix their gaze on a cross at the center of a screen. The activation condition comprised 29 stimuli (14 targets, stimulus duration 0.5 s, stimulus interval 2 s), during which subjects viewed a random series of single-digit figures (0-9) and were instructed to press a button when the currently presented figure was identical to the preceding one (e.g., 1-3-3, but not 3-1-3 or 3-1-5-3). There was a computer screen in front of the participants at a distance of 1 m. Participants were instructed to sit in a relaxed position with eyes open and to minimize head movements during the entire experiment, while the NIRS technician made records of the “awake/eyes open” status and head movements. To reduce any possible distractions, the experiment was performed in a quiet environment and participants were asked to keep their mood stable for some time prior to the task.

2.3. NIRS data acquisition The relative concentration changes of [oxy-Hb] and [deoxy-Hb] during the WM task were measured using a 52-channel NIRS optical topography system (ETG-4000, 9

Hitachi Medical Co., Tokyo, Japan). The system uses two wavelengths of near-infrared light (695 and 830 nm) and calculates the amount of absorbed near-infrared light based on the modified Beer-Lambert law (Maki et al., 1995). The data sampling frequency was 10 Hz. The “3×11” measurement patches provided by Hitachi were used. Each patch consisted of 17 emitters and 16 detector probes that were alternately positioned at an inter-probe distance of 3 cm, resulting in a total of 52 measurement channels (Figure 1). The NIRS probes were placed on the participant’s frontotemporal region with the lowest probes positioned along the Fp1-Fp2 line (Matsubara et al., 2014), which enabled the measurement of hemoglobin (Hb) values from the bilateral prefrontal, frontopolar and the anterior part of the temporal cortical (aTC) surface regions based on the anatomical cranio-cerebral correction via the International 10/20 system (Okamoto et al., 2004). The spatial registration of NIRS channels to the Montreal Neurological Institute (MNI) space (Singh et al., 2005) was allowed by NIRS-SPM (available at http://bisp.kaist.ac.kr/NIRS-SPM) (YE et al., 2009) using 3D digitizer.

2.4. NIRS data processing The NIRS data processing from detrending to analysis was performed using NIRS-SPM (Jang et al., 2009; YE et al., 2009), a widely used software package based on Statistical Parametric Mapping (YE et al., 2009) and MATLAB platform (MATLAB 7.8 (R2009a), The Mathworks, Inc. Natick, MA, USA). NIRS signals during WM task were filtered using a low-pass filter from 0.008 to 10

0.2 Hz. Prior to further processing, Hb data were corrected for global drift by detrending using the wavelet-minimum description length (MDL) algorithm (Jang et al., 2009) and corrected for serial correlation, such as those produced by physiological noise sources, using the HRF precolouring method (Worsley and Friston, 1995) implemented within NIRS-SPM. NIRS channel timecourses were modeled with a general linear model (GLM) in NIRS-SPM. The changes in the [oxy-Hb] and [deoxy-Hb] concentrations were then calculated for each channel using the modified Beer-Lambert law (Maki et al., 1995). And the least-square estimation of beta (β) value representing the degree of regional activation during the task was obtained channel by channel for each individual subject.

2.5. Data analyses The statistical analyses were performed using the SPSS version 19 (IBM Corporation, New York, USA) and SAS 9.1.3 (North Carolina State University, USA). Continuous variables were subjected to normality tests using the single sample Kolmogorov-Smirnov test, where data in accordance with a normal distribution was expressed as the mean ± standard deviation, while those in accordance with a non-normal distribution were expressed as the median (Q25, Q75), and the dichotomous variables were described as the frequency (percentage). Clinical variables between BD and UD patients were analyzed using Student’s t test and Mann-Whitney test depending on whether they were normally distributed. 11

Dichotomous variables between the two groups were examined by a chi-square test. For the analysis of hemodynamic response data, we calculated the relative changes of Hb signals during the task for each channel by NIRS-SPM, and then performed comparisons of the [oxy-Hb] and [deoxy-Hb] changes for each channel among the three groups using the Kruskal-Wallis test due to abnormal distribution. Furthermore, for channels showing significant group differences, the Mann-Whitney test was used to assess the difference in brain activity between groups. Then, relative [Hb] changes were examined for correlations with demographic or clinical characteristics (age, education, duration of illness and HAMD scores) in both patient groups using the Spearman correlation coefficient and the partial correlation coefficient. Statistical significance was considered at p<0.05.

3. Results All three groups were comparable for gender (x2=3.123, P =0.210), age (F=0.722, P=0.488) and years of education (Z=1.396, P=0.498). Clinical variables including age at onset (Z=-0.049, P=0.961), illness duration (Z=-0.325, P=0.745), severity of depression (t=0.050, P=0.960) and severity of illness (Z=-0.698, P=0.485) showed no significant differences between unipolar depression (UD) and unipolar depression (BD) groups. As for medications, antidepressants were used more frequently in the UD group, and mood stabilizers were used more frequently in the BD group (Table 1). Only one patient with UD and two with BD were medication-free. None of the participants used tricyclic or tetracyclic antidepressants. 12

Compared to healthy controls (HCs), both BD and UD groups presented significantly reduced prefrontal activation in [oxy-Hb] during the working memory (WM) task. For patients with BD, the decrease in [oxy-Hb] change was observed only in the right inferior prefrontal gyrus (IPFG) (CH34: Z=-2.354, P=0.019). For patients with UD, decreases in [oxy-Hb] changes were demonstrated in the left frontopolar cortex (FPC; CH18: Z=-2.358; P=0.018), left pars opercularis and pars triangularis ((POPE/PTRI; CH30: Z=-2.174, P=0.030), and right IPFG regions (CH34: Z=-1.990, P=0.047). Table 2 illustrates channels with a significant difference in prefrontal activation among the three groups during the WM task. Compared to patients with UD, patients with BD showed less decreased [oxy-Hb] in the left FPC (CH18: Z=-2.366, P=0.018) and left POPE/PTRI regions (CH30: Z=-2.333, P=0.020) during the WM task (Figure 2). We also examined the changes in [deoxy-Hb] concentrations during the WM task. No significant differences in the [deoxy-Hb] changes were observed among the three groups in any channel. In addition, there were no significant group difference in the [oxy-Hb] or [deoxy-Hb] activation in any channel from the temporal region. For channels showing significant group differences in the [oxy-Hb] changes among the three groups (CH18, CH30, CH34), no correlation between activation and HAMD scores or other variables (age, education, and duration of illness) were observed in either BD or UD group. The results remained the same when age, education and duration of illness were adjusted for partial correlation analysis.

13

4. Discussion We measured brain functional activity during a 1-backWM task, and both UD and BD groups showed a reduced activation in the inferior prefrontal region. The results are consistent with previous NIRS studies, which unanimously reported hypofunctioning of the prefrontal region in patients with UD (Noda et al., 2012; Pu et al., 2012a, 2012b, 2011) and BD (Kameyama et al., 2006; Matsuo et al., 2007; Nishimura et al., 2015a, 2015b). Furthermore, in line with our hypothesis, we observed that the BD group exhibited less deactivation in the left prefrontal cortex compared with the UD group. Additionally, no clinical variables were significantly correlated with [oxy-Hb] changes in any channel during the task in either UD or BD group, suggesting activation was independent of the effect of these variables. Compared to UD, patients with BD exhibited less deactivation (less decreased [oxy-Hb]) mainly in the left inferior frontal gyrus (LIFG), comprising the left FPC and left POPE /PTRI regions. The FPC (Brodmann area (BA) 10), not only the most anterior part of the frontal cortex but also one of the least understood regions of the human brain, is unique in that it is only interconnected with other supramodal (high-level processing) cortical regions (Petrides and Pandya, 2007). Evidence from functional neuroimaging work suggests that FPC plays a high-level executive controlling function of WM, which enables optimal integration of multiple cognitive operations to maximize task performance (Fletcher and Henson, 2001; Ramnani and Owen, 2004). The FPC is also important in the pathophysiology of emotional dysregulation (Phillips et al., 2008), for which previous work has implied structural 14

(Hayashi et al., 2011) and functional (Takei et al., 2014) abnormalities in patients with mood disorders. Based on its integrative function, the FPC has been described as a neural “bridge” between emotion and cognition (Pochon et al., 2002). In the present study, the less activation in both BD and UD group may imply the inefficient engagement of FPC in mood disorder. The present finding that BD patients have a less deactivated frontopolar region than UD patients has not been shown elsewhere. Different from our results, the only study (Takei et al., 2014) that has focused on frontopolar activation in bipolar and unipolar patients during a face-to-face conversation task suggested an equally decreased activation in both groups, and the possible reason will be discussed. We speculate that, except for the distinct cognitive paradigms employed (Owen et al., 2005), another factor might arise from the heterogeneity of bipolar patients across mood states. A study (Nishimura et al., 2015a) which combined longitudinal and cross-sectional assessments suggested that prefrontal activation differs between hypomanic and depressive states in bipolar and was decreased after hypomanic symptoms resolved, though the effect of medication could not be excluded. We thus speculate that the mixture of mood states in bipolar subjects may have obscured the group differences in brain activation (Schecklmann et al., 2011; Takei et al., 2014). The present sample maintains homogeneity to a certain extent, due not only to the single depressive state of bipolar patients but also to the restriction on illness duration in the UD group to ensure the relative diagnostic stability. The influence of task design and clinical characteristics on NIRS activation should be further examined in 15

future studies. Another region detected in the present study is left POPE/PTRI, corresponding with the Brodmann's areas (BA) 44, 45 and 47 of the left hemisphere, which together comprise Broca's area, the crucial region of language in the left inferior frontal gyrus (LIFG). According to Baddeley’s (Baddeley, 2010) phonological loop model of WM, subjects in this study might enhance their initial memory trace of visual stimuli by verbal or subvocal rehearsal, which thus led to the activation of Broca's area. Although functional alterations in Broca’s area were occasionally reported by very few studies (Rasgon et al., 2008; Silverstone et al., 2005) in bipolar or unipolar, the rare information we obtained is far from a clear understanding of the regional function of Broca's area in mood disorders. Different methodologies also make it impossible to make comparisons between BD and UD. Further studies examining brain function of frontopolar and LIFG (especially of Broca’s area) in mood disorder are needed. We also explored the [deoxy-Hb] concentration changes during the WM task and demonstrated no among-group differences in any channel. The unparallel observations between [oxy-Hb] and [deoxy-Hb] changes may be attributable to their different nature in hemodynamic responses to neural activity and/or the difference in signal-to-noise ratio in the NIRS measurements. Under normal circumstances, the [oxy-Hb] signal reflects rCBF changes while the [deoxy-Hb] is generated through oxygen utilization in cerebral tissue. Since the [oxy-Hb] signals have previously been shown to reflect cognitive activation more directly than [deoxy-Hb] signals, as shown by a stronger correlation with BOLD signal from fMRI (Hoshi et al., 2001; 16

Strangman et al., 2002), our interpretations especially focused on the [oxy-Hb] changes. The current study has several limitations. The first limitation was medication. Most patients used psychotropic drugs when accepted NIRS measurement. A previous study (Kameyama et al., 2006) demonstrated that antidepressant dosages did not correlate with the [oxy-Hb] changes in both bipolar and unipolar groups. Similarly, a recent study (Nishimura et al., 2015a) confirmed no correlation between medication dose and NIRS signals for both cross-sectional and longitudinal assessments in bipolar patients. However, a controlled crossover trial (Kohmura et al., 2013) indicated that certain types of antidepressants may not be free of potentially confounding effect on NIRS signals, although no effect was observed on behavioral performance. These concerns are inconclusive by a lack of NIRS data directly addressing drug effects on brain activation in patients with mood disorder. Given this, the present study did not make specific restrictions on medication use. And the small number of drug-naive patients in this study does not allow definitive analysis. Future NIRS studies with large samples comparing medicated and unmedicated patients may help clarify the association between psychotropic drugs and brain functional activity. Second, behavioral data were not systematically collected here. Although some authors (Kohmura et al., 2013) used such data to explain NIRS results between groups, an fMRI study (Bertocci et al., 2012) suggested no significant group differences in n-back task performance (reaction time and accuracy) between medicated BD and UD patients. Our previous work (Zhu et al., 2013) and other researchers (Hill et al., 2009; 17

Sweeney et al., 2000; Xu et al., 2012) have also achieved parallel results in BD and UD groups. The present study is a preliminary exploratory fNIRS study, focusing mainly on the brain functional activity. For further study, we will systematically collect behavioral data and follow the relationship between regional activation and task performance. Future studies addressing these limitations are needed to confirm the present results further.

5. Conclusions In conclusion, this study is an early NIRS study investigating prefrontal activation during an n-back WM task in patients with BD and UD. The results from this preliminary exploratory study revealed distinct prefrontal activation patterns independent of depression severity underlying BD and UD during working memory processing. The specific activation regions that allow the measuring of prefrontal cortex functions were mainly located in the left inferior frontal gyrus (LIFG), which may aid the elucidation of the pathophysiology mechanisms of mood disorders. These findings raise the possibility that NIRS-based prefrontal activation may serve as a potential marker to aid in differentiating bipolar from unipolar depression at an early stage. Contributors Yue Zhu designed the study and drafted the manuscript. Wenxiang Quan helped with the drafting and Huali Wang revised the manuscript critically for important intellectual content. Yue Zhu, Wenxiang Quan, Huali Wang, Yantao Ma, Jun Yan and 18

Hua Zhang contributed significantly to data acquisition and interpretation. Yue Zhu and Hua Zhang undertook the statistical analysis. Wentian Dong and Xin Yu provided general advice for experiment and supervised completion of the manuscript. All authors contributed to and have approved the final manuscript.

Role of the funding source The research was supported in part by the Beijing Municipal Science and Technology Commission [Grant numbers 2011-4024-01, Z121107001012040]. The sponsors had no role in study design; in the collection, analysis and interpretation of data; in the preparation of the manuscript; or in the decision to submit the article for publication.

Acknowledgements The research was supported in part by the Beijing Municipal Science and Technology Commission [grant numbers 2011-4024-01 and Z121107001012040]. The authors appreciate the participants of the study. The authors also thank Jiuju Wang, PhD and Ms. Ju Tian, for assistance with the NIRS measurements, from Peking University Institute of Mental Health, Beijing, China; and Haijing Niu, PhD, for insights on methodology, from Beijing Normal University, Beijing, China. All acknowledged individuals have no conflicts of interest to declare.

Conflict of interest None. 19

References Akiskal, H.S., Maser, J.D., Zeller, P.J., Endicott, J., Coryell, W., Keller, M., Warshaw, M., Clayton, P., Goodwin, F., 1995. Switching from 'unipolar' to bipolar II. An 11-year prospective study of clinical and temperamental predictors in 559 patients. Arch Gen Psychiatry 52, 114-123. Almeida, J.R.C., Versace, A., Hassel, S., Kupfer, D.J., Phillips, M.L., 2010. Elevated Amygdala Activity to Sad Facial Expressions: A State Marker of Bipolar but Not Unipolar Depression. BIOL PSYCHIAT 67, 414-421. Angst, J., Gamma, A., Benazzi, F., Ajdacic, V., Eich, D., Rossler, W., 2003. Toward a re-definition of subthreshold bipolarity: epidemiology and proposed criteria for bipolar-II, minor bipolar disorders and hypomania. J Affect Disord 73, 133-146. APA, 2000. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association, Washington, DC. Baddeley, A., 2010. Working memory. CURR BIOL 20, R136-R140. Bertocci, M.A., Bebko, G.M., Mullin, B.C., Langenecker, S.A., Ladouceur, C.D., Almeida, J.R.C., Phillips, M.L., 2012. Abnormal anterior cingulate cortical activity during emotional n-back task performance distinguishes bipolar from unipolar depressed females. PSYCHOL MED 42, 1417-1428. Boas, D.A., Dale, A.M., Franceschini, M.A., 2004. Diffuse optical imaging of brain activation: approaches to optimizing image sensitivity, resolution, and accuracy. NEUROIMAGE 23 Suppl 1, S275-S288. Borkowska, A., Rybakowski, J.K., 2001. Neuropsychological frontal lobe tests indicate that bipolar depressed patients are more impaired than unipolar. BIPOLAR DISORD 3, 88-94. Cerullo, M.A., Eliassen, J.C., Smith, C.T., Fleck, D.E., Nelson, E.B., Strawn, J.R., Lamy, M., DelBello, M.P., Adler, C.M., Strakowski, S.M., 2014. Bipolar I disorder and major depressive disorder show similar brain activation during depression. BIPOLAR DISORD 16, 703-712. Chase, H.W., Nusslock, R., Almeida, J.R., Forbes, E.E., LaBarbara, E.J., Phillips, M.L., 2013. Dissociable patterns of abnormal frontal cortical activation during anticipation of an uncertain reward or loss in bipolar versus major depression. BIPOLAR DISORD 15, 839-854. Cremaschi, L., Penzo, B., Palazzo, M., Dobrea, C., Cristoffanini, M., Dell'Osso, B., Altamura, A.C., 2013. Assessing Working Memory via N-Back Task in Euthymic Bipolar I Disorder Patients: A Review of Functional Magnetic Resonance Imaging Studies. NEUROPSYCHOBIOLOGY 68, 63-70. Daniel, B.D., Montali, A., Gerra, M.L., Innamorati, M., Girardi, P., Pompili, M., Amore, M., 2013. Cognitive impairment and its associations with the path of illness in affective disorders: a comparison between patients with bipolar and unipolar depression in remission. J PSYCHIATR PRACT 19, 275-287. de Almeida, J.R., Phillips, M.L., 2013. Distinguishing between Unipolar Depression and Bipolar Depression: Current and Future Clinical and Neuroimaging Perspectives. Biol Psychiatry 73, 111-118. Diler, R.S., de Almeida, J.R.C., Ladouceur, C., Birmaher, B., Axelson, D., Phillips, M., 2013. Neural activity to intense positive versus negative stimuli can help differentiate bipolar disorder from unipolar major depressive disorder in depressed adolescents: A pilot fMRI study. Psychiatry Research: Neuroimaging 214, 277-284. Dudek, D., Siwek, M., Zielińska, D., Jaeschke, R., Rybakowski, J., 2013. Diagnostic conversions from 20

major depressive disorder into bipolar disorder in an outpatient setting: Results of a retrospective chart review. J AFFECT DISORDERS 144, 112-115. Fletcher, P.C., Henson, R.N., 2001. Frontal lobes and human memory: insights from functional neuroimaging. BRAIN 124, 849-881. Godard, J., Baruch, P., Grondin, S., Lafleur, M.F., 2012. Psychosocial and neurocognitive functioning in unipolar and bipolar depression: a 12-month prospective study. Psychiatry Res 196, 145-153. Goldberg, J.F., Harrow, M., Whiteside, J.E., 2001. Risk for bipolar illness in patients initially hospitalized for unipolar depression. Am J Psychiatry 158, 1265-1270. Gruber, S., Rathgeber, K., Braunig, P., Gauggel, S., 2007. Stability and course of neuropsychological deficits in manic and depressed bipolar patients compared to patients with Major Depression. J Affect Disord 104, 61-71. Hamilton, M., 1967. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 6, 278-296. Harvey, P.O., Le Bastard, G., Pochon, J.B., Levy, R., Allilaire, J.F., Dubois, B., Fossati, P., 2004. Executive functions and updating of the contents of working memory in unipolar depression. J PSYCHIATR RES 38, 567-576. Hayashi, Y., Nihonmatsu-Kikuchi, N., Yu, X., Ishimoto, K., Hisanaga, S.I., Tatebayashi, Y., 2011. A novel, rapid, quantitative cell-counting method reveals oligodendroglial reduction in the frontopolar cortex in major depressive disorder. Mol Psychiatry 16, 1155-1158. Hill, S.K., Reilly, J.L., Harris, M.S., Rosen, C., Marvin, R.W., Deleon, O., Sweeney, J.A., 2009. A comparison of neuropsychological dysfunction in first-episode psychosis patients with unipolar depression, bipolar disorder, and schizophrenia. SCHIZOPHR RES 113, 167-175. Hirschfeld, R.M., Calabrese, J.R., Weissman, M.M., Reed, M., Davies, M.A., Frye, M.A., Keck, P.J., Lewis, L., McElroy, S.L., McNulty, J.P., Wagner, K.D., 2003. Screening for bipolar disorder in the community. J Clin Psychiatry 64, 53-59. Hirschfeld, R.M., Lewis, L., Vornik, L.A., 2003. Perceptions and impact of bipolar disorder: how far have we really come? Results of the national depressive and manic-depressive association 2000 survey of individuals with bipolar disorder. J Clin Psychiatry 64, 161-174. Hock, C., Villringer, K., Muller-Spahn, F., Wenzel, R., Heekeren, H., Schuh-Hofer, S., Hofmann, M., Minoshima, S., Schwaiger, M., Dirnagl, U., Villringer, A., 1997. Decrease in parietal cerebral hemoglobin oxygenation during performance of a verbal fluency task in patients with Alzheimer's disease monitored by means of near-infrared spectroscopy (NIRS)--correlation with simultaneous rCBF-PET measurements. BRAIN RES 755, 293-303. Hoshi, Y., Kobayashi, N., Tamura, M., 2001. Interpretation of near-infrared spectroscopy signals: a study with a newly developed perfused rat brain model. J Appl Physiol (1985) 90, 1657-1662. Jang, K.E., Tak, S., Jung, J., Jang, J., Jeong, Y., Ye, J.C., 2009. Wavelet minimum description length detrending for near-infrared spectroscopy. J BIOMED OPT 14, 34004. Judd, L.L., Akiskal, H.S., Schettler, P.J., Coryell, W., Endicott, J., Maser, J.D., Solomon, D.A., Leon, A.C., Keller, M.B., 2003. A prospective investigation of the natural history of the long-term weekly symptomatic status of bipolar II disorder. Arch Gen Psychiatry 60, 261-269. Judd, L.L., Akiskal, H.S., Schettler, P.J., Endicott, J., Maser, J., Solomon, D.A., Leon, A.C., Rice, J.A., Keller, M.B., 2002. The long-term natural history of the weekly symptomatic status of bipolar I disorder. Arch Gen Psychiatry 59, 530-537. Kameyama, M., Fukuda, M., Yamagishi, Y., Sato, T., Uehara, T., Ito, M., Soto, T., Mikuni, M., 2006. 21

Frontal lobe function in bipolar disorder: A multichannel near-infrared spectroscopy study. NEUROIMAGE 29, 172-184. Komura, K., Iwamoto, K., Aleksey, B., Masada, K., Kawano, N., Katayama, H., Noda, Y., Noda, A., Hidaka, T., Ozaki, N., 2013. Effects of sedative antidepressants on prefrontal cortex activity during verbal

fluency

task

in

healthy

subjects:

a

near-infrared

spectroscopy

study.

PSYCHOPHARMACOLOGY 226, 75-81. Koneke, S., Noda, T., Yokoyama, S., Kunisato, Y., Ito, D., Suyama, H., Matsuda, T., Sugimura, Y., Ishihara, N., Shimizu, Y., Nakazawa, K., Yoshida, S., Arima, K., Suzuki, S., 2013. The relationship between positive and negative automatic thought and activity in the prefrontal and temporal cortices: A multi-channel near-infrared spectroscopy (NIRS) study. J AFFECT DISORDERS 151, 352-359. Liang, M.J., Zhou, Q., Yang, K.R., Yang, X.L., Fang, J., Chen, W.L., Huang, Z., 2013. Identify changes of brain regional homogeneity in bipolar disorder and unipolar depression using resting-state FMRI. PLOS ONE 8, e79999. Maalouf, F.T., Klein, C., Clark, L., Sahakian, B.J., Labarbara, E.J., Versace, A., Hassel, S., Almeida, J.R., Phillips, M.L., 2010. Impaired sustained attention and executive dysfunction: Bipolar disorder versus depression-specific markers of affective disorders. NEUROPSYCHOLOGIA. Maki, A., Yamashita, Y., Ito, Y., Watanabe, E., Mayanagi, Y., Koizumi, H., 1995. Spatial and temporal analysis of human motor activity using noninvasive NIR topography. MED PHYS 22, 1997-2005. Matsubara, T., Matsuo, K., Nakashima, M., Nakano, M., Harada, K., Watanuki, T., Egashira, K., Watanabe, Y., 2014. Prefrontal activation in response to emotional words in patients with bipolar disorder and major depressive disorder. NEUROIMAGE 85, 489-497. Matsuo, K., Kouno, T., Hatch, J.P., Seino, K., Ohtani, T., Kato, N., Kato, T., 2007. A near-infrared spectroscopy study of prefrontal cortex activation during a verbal fluency task and carbon dioxide inhalation in individuals with bipolar disorder. BIPOLAR DISORD 9, 876-883. Morselli, P.L., Elgie, R., 2003. GAMIAN-Europe/BEAM survey I--global analysis of a patient questionnaire circulated to 3450 members of 12 European advocacy groups operating in the field of mood disorders. BIPOLAR DISORD 5, 265-278. Nishimura, Y., Takahashi, K., Ohtani, T., Ikeda-Sugita, R., Kasai, K., Okazaki, Y., 2015a. Dorsolateral prefrontal hemodynamic responses during a verbal fluency task in hypomanic bipolar disorder. BIPOLAR DISORD 17, 172-183. Nishimura, Y., Takahashi, K., Ohtani, T., Ikeda-Sugita, R., Okada, N., Kasai, K., Okazaki, Y., 2015b. Social Function and Frontopolar Activation during a Cognitive Task in Patients with Bipolar Disorder. NEUROPSYCHOBIOLOGY 72, 81-90. Noda, T., Yoshida, S., Matsuda, T., Okamoto, N., Sakamoto, K., Koneke, S., Numachi, Y., Matsushima, E., Kunugi, H., Higuchi, T., 2012. Frontal and right temporal activations correlate negatively with depression severity during verbal fluency task: A multi-channel near-infrared spectroscopy study. J PSYCHIATR RES 46, 905-912. Obrig, H., Villringer, A., 2003. Beyond the visible--imaging the human brain with light. J Cereb Blood Flow Metab 23, 1-18. Ohmae, E., Ouchi, Y., Oda, M., Suzuki, T., Nobesawa, S., Kanno, T., Yoshikawa, E., Futatsubashi, M., Ueda, Y., Okada, H., Yamashita, Y., 2006. Cerebral hemodynamics evaluation by near-infrared time-resolved

spectroscopy:

correlation

with

simultaneous

positron

emission

tomography

measurements. NEUROIMAGE 29, 697-705. Ohtani, T., Nishimura, Y., Takahashi, K., Ikeda-Sugita, R., Okada, N., Okazaki, Y., 2015. Association 22

between longitudinal changes in prefrontal hemodynamic responses and social adaptation in patients with bipolar disorder and major depressive disorder. J AFFECT DISORDERS 176, 78-86. Okamoto, M., Dan, H., Sakamoto, K., Takeo, K., Shimizu, K., Kohno, S., Oda, I., Isobe, S., Suzuki, T., Kohyama, K., Dan, I., 2004. Three-dimensional probabilistic anatomical cranio-cerebral correlation via the international 10-20 system oriented for transcranial functional brain mapping. NEUROIMAGE 21, 99-111. Owen, A.M., McMillan, K.M., Laird, A.R., Bullmore, E., 2005. N-back working memory paradigm: A meta-analysis of normative functional neuroimaging studies. HUM BRAIN MAPP 25, 46-59. Perlis, R.H., 2005. Misdiagnosis of bipolar disorder. AM J MANAG CARE 11, S271-S274. Petrides, M., Pandya, D.N., 2007. Efferent association pathways from the rostral prefrontal cortex in the macaque monkey. J NEUROSCI 27, 11573-11586. Phillips, M.L., Travis, M.J., Fagiolini, A., Kupfer, D.J., 2008. Medication effects in neuroimaging studies of bipolar disorder. Am J Psychiatry 165, 313-320. Pochon, J.B., Levy, R., Fossati, P., Lehericy, S., Poline, J.B., Pillon, B., Le Bihan, D., Dubois, B., 2002. The neural system that bridges reward and cognition in humans: an fMRI study. Proc Natl Acad Sci U S A 99, 5669-5674. Pu, S., Nakagome, K., Yamada, T., Yokoyama, K., Matsumura, H., Mitani, H., Adachi, A., Nagata, I., Kaneko, K., 2012a. The relationship between the prefrontal activation during a verbal fluency task and stress-coping style in major depressive disorder: A near-infrared spectroscopy study. J PSYCHIATR RES 46, 1427-1434. Pu, S., Yamada, T., Yokoyama, K., Matsumura, H., Kobayashi, H., Sasaki, N., Mitani, H., Adachi, A., Kaneko, K., Nakagome, K., 2011. A multi-channel near-infrared spectroscopy study of prefrontal cortex activation during working memory task in major depressive disorder. NEUROSCI RES 70, 91-97. Pu, S., Yamada, T., Yokoyama, K., Matsumura, H., Mitani, H., Adachi, A., Kaneko, K., Nakagome, K., 2012b. Reduced prefrontal cortex activation during the working memory task associated with poor social functioning in late-onset depression: Multi-channel near-infrared spectroscopy study. Psychiatry Research: Neuroimaging 203, 222-228. Ramnani, N., Owen, A.M., 2004. Anterior prefrontal cortex: insights into function from anatomy and neuroimaging. NAT REV NEUROSCI 5, 184-194. Rasgon, N.L., Kenna, H.A., Geist, C., Small, G., Silverman, D., 2008. Cerebral metabolic patterns in untreated postmenopausal women with major depressive disorder. Psychiatry Res 164, 77-80. Rose, E.J., Ebmeier, K.P., 2006. Pattern of impaired working memory during major depression. J AFFECT DISORDERS 90, 149-161. Sato, H., Yahata, N., Funane, T., Takizawa, R., Katura, T., Atsumori, H., Nishimura, Y., Kinoshita, A., Kiguchi, M., Koizumi, H., Fukuda, M., Kasai, K., 2013. A NIRS-fMRI investigation of prefrontal cortex activity during a working memory task. NEUROIMAGE 83, 158-173. Schecklmann, M., Dresler, T., Beck, S., Jay, J.T., Febres, R., Haeusler, J., Jarczok, T.A., Reif, A., Plichta, M.M., Ehlis, A., Fallgatter, A.J., 2011. Reduced prefrontal oxygenation during object and spatial visual working memory in unipolar and bipolar depression. Psychiatry Research: Neuroimaging 194, 378-384. Scott, E.M., Hermens, D.F., Naismith, S.L., Guastella, A.J., De Regt, T., White, D., Lagopoulos, J., Hickie, I.B., 2013. Distinguishing young people with emerging bipolar disorders from those with unipolar depression. J Affect Disord 144, 208-215. 23

Si, T.M., Shu, L., Dang, W.M., Su, Y.A., Chen, J.X., Dong, W.T., Kong, Q.M., H, Z.W., 2009. Evaluation of the reliability and validity of Chinese version of the mini-international neuropsychiatric interview in patient with mental disorders. Chinese Mental Health Journal 23, 493-497. Silverstone, P.H., Bell, E.C., Willson, M.C., Dave, S., Wilman, A.H., 2005. Lithium alters brain activation in bipolar disorder in a task- and state-dependent manner: an fMRI study. Ann Gen Psychiatry 4, 14. Singh, A.K., Okamoto, M., Dan, H., Jurcak, V., Dan, I., 2005. Spatial registration of multichannel multi-subject fNIRS data to MNI space without MRI. NEUROIMAGE 27, 842-851. Spearing, M.K., Post, R.M., Leverich, G.S., Brandt, D., Nolen, W., 1997. Modification of the Clinical Global Impressions (CGI) Scale for use in bipolar illness (BP): the CGI-BP. Psychiatry Res 73, 159-171. Strangman, G., Boas, D.A., Sutton, J.P., 2002. Non-invasive neuroimaging using near-infrared light. Biol Psychiatry 52, 679-693. Sweeney, J.A., Kmiec, J.A., Kupfer, D.J., 2000. Neuropsychologic impairments in bipolar and unipolar mood disorders on the CANTAB neurocognitive battery. Biol Psychiatry 48, 674-684. Takei, Y., Suda, M., Aoyama, Y., Sakurai, N., Tagawa, M., Motegi, T., Yamaguchi, M., Narita, K., Fukuda, M., 2014. Near-infrared spectroscopic study of frontopolar activation during face-to-face conversation in major depressive disorder and bipolar disorder. J PSYCHIATR RES 57, 74-83. Villringer, K., Minoshima, S., Hock, C., Obrig, H., Ziegler, S., Dirnagl, U., Schwaiger, M., Villringer, A., 1997. Assessment of local brain activation. A simultaneous PET and near-infrared spectroscopy study. ADV EXP MED BIOL 413, 149-153. Worsley, K.J., Friston, K.J., 1995. Analysis of fMRI time-series revisited--again. NEUROIMAGE 2, 173-181. Xu, G., Lin, K., Rao, D., Dang, Y., Ouyang, H., Guo, Y., Ma, J., Chen, J., 2012. Neuropsychological performance in bipolar I, bipolar II and unipolar depression patients: a longitudinal, naturalistic study. J Affect Disord 136, 328-339. YE, J., TAK, S., JANG, K., JUNG, J., JANG, J., 2009. NIRS-SPM: Statistical parametric mapping for near-infrared spectroscopy. NEUROIMAGE 44, 428-447. Young, R.C., Biggs, J.T., Ziegler, V.E., Meyer, D.A., 1978. A rating scale for mania: reliability, validity and sensitivity. Br J Psychiatry 133, 429-435. Zhu, Y., Ma, Y.T., Shi, C., Zhang, L., Yu, X., 2013. A comparision of neurocognitive function among patients with bipolar depression, recurrent unipolar depression and schizophrenia. Chin J Psychiatry 46, 75-81. Zimmermann, P., Bruckl, T., Nocon, A., Pfister, H., Lieb, R., Wittchen, H.U., Holsboer, F., Angst, J., 2009. Heterogeneity of DSM-IV major depressive disorder as a consequence of subthreshold bipolarity. Arch Gen Psychiatry 66, 1341-1352.

24

Table 1 Demographic and clinical characteristics of the three groups

Age (years) Gender (male/female) Education (25,75; years) Duration of illness (months) Onset age of illness (years) CGI (3,4,5,6; %)

BD (n=39) 37.0±12.9 19/20

UD (n=35) 35.9±13.2 11/24

HCs (n=36) 33.6±10.3 18/18

F/ x2/ t /Z

P-value

0.722 3.123

0.488 0.210

15 (9,16)

15 (9,16)

15 (9.8,16.8)

1.396

0.498

76(28,156)

72(67,90)

—

-0.325

0.745

24(22,32)

27(18,41)

—

-0.049

0.961

7.7/35.9/46.1/10. 3 22.6±8.2

8.6/28.6/45.7/17. 1 22.5±13.2

—

-0.698

0.485

0.050

0.960

12.819 1.918 22.331 0.375

<0.001 0.166 <0.001 0.540

HAMD — Medication (n; %) Antidepressants 21(53.8) 32(91.4) — 23(59.0) 15(42.9) Antipsychotics — 22(61.5) 3(8.6) Mood stabilizers — 14(35.9) 14(40.0) Sedative hypnotics — BD = bipolar disorder group; UD = major depressive disorder group; HCs = healthy controls group. CGI: clinical global impression Scale; HAMD: Hamilton Depression Rating Scale, 17-item version.

25

Table 2 Comparison of brain activation during 1-back task among the three groups Channel M (Q25, Q75)

Brain regions

BD (n=39)

UD (n=35)

HCs (n=36)

x2*

P-value

Brain activation

0.0036 -0.0399 0.0185 7.400 0.025 HC>UD, BD>UD (-0.0352, 0.1239) (-0.2374,0.0458) (-0.0343,0.0803) left 0.0026 -0.0807 0.0178 CH30 6.746 0.034 HC>UD, BD>UD POPE/PTRI (-0.0605, 0.1732) (-0.3725,0.0471) (-0.0955,0.1348) -0.0245 -0.0104 0.0678 CH34 right IPFG 6.377 0.041 HC>UD, HC>BD (-0.0828, 0.0701) (-0.0936,0.1172) (-0.0204,0.1789) BD = bipolar disorder group; UD = major depressive disorder group; HCs = healthy controls group; FPC = frontopolar cortex; IPFG = inferior prefrontal gyrus; POPE/PTRI = pars opercularis and pars triangularis; CH=channel. * 2 x values by Kruskal-Wallis test of group comparison. CH18

left FPC

Highlights



Prefrontal activation differs between unipolar (UD) and bipolar depression (BD).



BD exhibited less deactivation in the left frontopolar and Broca's area than UD.



NIRS measurement may aid in differentiating BD from UD at an early stage.

26

Fig. 1. Location of probes and channel settings in 52-channel near-infrared spectroscopy. (Left): The location of the probe array (red band) according to the international 10-20 system. (Right): 2-D topographic map in a back view.

Fig. 2. Differences in [oxy-Hb] changes of prefrontal cortex during the working memory task among the three groups (BD, UD, HCs). Figure (A): CH18, the left frontopolar region; (B): CH30, the left POPE/PTRI region; and (C): CH34, the right IPFG region. Abbreviations: [oxy-Hb], the concentration of oxygenated hemoglobin; BD, bipolar depression; UD, recurrent unipolar depression; HCs, healthy controls; POPE/PTRI, pars opercularis and pars triangularis; and IPFG, inferior prefrontal gyrus.

27