- Email: [email protected]
Reduced prefrontal activation during a verbal fluency task in Chinese-speaking patients with schizophrenia as measured by near-infrared spectroscopy
Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
Contents lists available at ScienceDirect
Progress in Neuro-Psychopharmacology & Biological Psychiatry
Reduced prefrontal activation during a verbal fluency task in Chinese-speaking patients with schizophrenia as measured by near-infrared spectroscopy Wenxiang Quan a,b,c,1, Tongning Wu d,1, Zhaohua Li e, Yuduo Wang e, Wentian Dong a,b,c,⁎, Bin Lv d,⁎⁎ a
Peking University Sixth Hospital, Beijing, China Peking University Institute of Mental Health, Beijing, China c Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China d China Academy of Telecommunication Research of Ministry of Industry and Information Technology, Beijing, China e School of Information and Communication Engineering, Beijing Information Science and Technology University, Beijing, China b
a r t i c l e
i n f o
Article history: Received 12 August 2014 Received in revised form 16 December 2014 Accepted 17 December 2014 Available online 24 December 2014 Keywords: Chinese version verbal fluency test Near-infrared spectroscopy (NIRS) Prefrontal cortex Schizophrenia
a b s t r a c t Near-infrared spectroscopy (NIRS) has been applied to examine the possible functional alternations during the performance of cognitive tasks in schizophrenia. With this technique, previous studies have observed that patients with schizophrenia are often associated with reduced brain activation in the prefrontal cortex during the verbal fluency task (VFT) of the English version or the Japanese version. However, it remains unclear whether there is a brain functional impairment in Chinese-speaking patients with schizophrenia. In this study, we designed a Chinese version of the VFT and performed a multichannel NIRS study in a large group of patients with schizophrenia and healthy controls. We investigated brain activation during the task period of the Chinese version of the VFT within a schizophrenia group and a healthy group, respectively, and compared the relative changes between the two groups. Our results confirmed that Chinese-speaking patients with schizophrenia had significantly lower brain activation in the prefrontal cortex and superior temporal cortex when compared with healthy controls. Such findings based on the NIRS data provided us reliable evidences about brain functional deficits in the Chinese-speaking patients with schizophrenia. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Schizophrenia is a psychiatric disorder which is characterized in general by the disruption in cognitive behavior and the incoordination of affective reaction (American Psychiatric Association, 2000). Besides these impairments in external performance, recent neuroimaging studies have demonstrated that patients with schizophrenia are often associated with the abnormalities of brain structure and function, especially in the prefrontal cortex (Lynall et al., 2010; Manoach et al., 2000; Narr et al., 2005; Palaniyappan et al., 2011; Steen et al., 2006; Abbreviations: BA, Brodmann's area; [deoxy-Hb], relative concentration of deoxygenated hemoglobin; FDR, false discovery rate; MRI, magnetic resonance imaging; NIRS, near-infrared spectroscopy; [oxy-Hb], relative concentration of oxygenated hemoglobin; [total-Hb], relative concentration of total hemoglobin; VFT, verbal fluency task ⁎ Correspondence to: W. Dong, Peking University Sixth Hospital. No. 51, Huayuanbei Road, Haidian District, Beijing 100191, China. Tel.: +86 10 82805966; fax: + 86 10 62026310. ⁎⁎ Correspondence to: B. Lv, China Academy of Telecommunication Research of Ministry of Industry and Information Technology, No. 52, Huayuanbei Road, Haidian District, Beijing 100191, China. Tel.: +86 10 62304633 2084. E-mail addresses: [email protected] (W. Dong), [email protected] (B. Lv). 1 These authors contributed equally to the work.
http://dx.doi.org/10.1016/j.pnpbp.2014.12.005 0278-5846/© 2014 Elsevier Inc. All rights reserved.
Weiss et al., 2004). With structural magnetic resonance imaging (MRI), converging evidences have revealed the structural changes of prefrontal cortex in schizophrenia including brain volume atrophy (Steen et al., 2006), cortical thickness thinning (Narr et al., 2005) and gyrification reduction (Palaniyappan et al., 2011). In addition, more and more studies have utilized functional neuroimaging technologies, e.g., functional MRI (fMRI), to investigate brain functional activation during the resting state (Lynall et al., 2010) as well as various cognitive tasks (Weiss et al., 2004). These results provided us imaging observation on the dysfunction of prefrontal cortex in patients with schizophrenia. Near-infrared spectroscopy (NIRS) is another promising and noninvasive functional neuroimaging technique which can measure the concentration changes of oxygenated hemoglobin ([oxy-Hb]) and deoxygenated hemoglobin ([deoxy-Hb]) near the brain surface (Ferrari and Quaresima, 2012). The NIRS signals ([oxy-Hb] and [deoxy-Hb]) are considered to reflect regional cerebral blood volumes and show strong correlations with fMRI signals (Sasai et al., 2012). Compared with fMRI, NIRS is more portable and less sensitive to motion artifacts. These advantages make it more suitable for the assessment of speech related tasks in schizophrenia (Ehlis et al., 2014; Ferrari and Quaresima, 2012). The verbal fluency task (VFT) is a common language
52
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
related neuropsychological task and patients with schizophrenia are associated with profound impaired verbal fluency (Dieler et al., 2012). Therefore, we can infer that schizophrenia has a risk of abnormal brain functional activation during the performance of a VFT. By means of NIRS, previous studies have indicated that schizophrenia patients had different changes in [oxy-Hb] than those of healthy controls during the VFT, although the brain regions showing abnormal activation are not always consistent (Koike et al., 2013). Watanabe et al. applied one channel NIRS system to investigate the cerebral hemoglobin response in 62 schizophrenia patients and 31 healthy controls (Watanabe and Kato, 2004). They described for the first time that patients with schizophrenia had reduced [oxy-Hb] in the left prefrontal cortex during the VFT. Till now, their sample size for schizophrenia patient is the largest in related NIRS studies (Koike et al., 2013). Then, Suto et al. described the spatial–temporal patterns of NIRS activation in the prefrontal and temporal cortices using two 24 multichannel NIRS systems (Suto et al., 2004). Their results on 13 schizophrenia patients and 16 healthy controls revealed that schizophrenia patients had lower NIRS activation in the bilateral prefrontal and temporal cortices at the start of the VFT task period. This dysfunction was further confirmed in a relatively large group (including 55 schizophrenia patients and 70 healthy subjects) (Takizawa et al., 2008). In addition, the delayed cortical activation was also detected particularly in the frontopolar region (Takizawa et al., 2008). By comparing the performance of phonological VFT and semantic VFT, the more pronounced cortical activation was observed during the phonological VFT compared to the semantic VFT in both schizophrenia patients and healthy controls, and their induced activations in schizophrenia patients were all significantly reduced than those in healthy controls (Ehlis et al., 2007). These findings demonstrated that it seems a reliable biomarker that patients with schizophrenia had significantly less frontal activation during performance in the VFT. However, most of these NIRS studies were conducted on Japanese-speaking (Suto et al., 2004; Takizawa et al., 2008; Watanabe and Kato, 2004) or English-speaking populations (Ehlis et al., 2007), and few studies have been performed in Chinesespeaking people, even for healthy subjects. A recent study demonstrated the effect of languages on the performance of the VFT using Japanese, Turkish, and English-speaking patients with schizophrenia (Sumiyoshi et al., 2014). Therefore, we need to investigate the response pattern of NIRS activition during the Chinese version VFT in healthy controls and its possible alternation in patients with schizophrenia. In the present study, we designed a Chinese version VFT and recruited a large group of native Chinese speakers to perform this cognitive task. The group included 140 patients with schizophrenia and 100 healthy controls. Then, a 52 multichannel NIRS system was used to examine the hemodynamic signals in the bilateral prefrontal and superior temporal cortices during the performance of the Chinese version VFT. Our aim was to evaluate the NIRS cerebral response pattern along with the Chinese version VFT and compare the brain activation between patients with schizophrenia and healthy controls. We hypothesized that the Chinese-speaking patients with schizophrenia had
an altered brain activation compared to healthy controls during the designed Chinese version VFT. 2. Methods 2.1. Subjects One hundred and sixty-five schizophrenia patients and 100 healthy comparison subjects took part in this study. 25 schizophrenia patients were excluded from further analysis because of incomplete collection/ obvious noise (19 patients) or handedness issue (6 left-handed patients). The final sample comprised 140 schizophrenia patients (male/female: 80/60, age: range 17–62 years, mean 33.81 ± 11.52 years) and 100 healthy subjects (male/female: 65/35, age: range 18–78 years, mean 34.43 ± 12.36 years), and they were all righthanded (Table 1). On education background, the schizophrenia group included 14 graduate degrees, 63 undergraduate degrees and 63 high school degrees, and the healthy group included 20 graduate degrees, 44 undergraduate degrees and 36 high school degrees. All participants were native speakers of the Chinese language and can read the Chinese language. Schizophrenia patients were recruited from Peking University Sixth Hospital. Each patient was diagnosed independently by two clinical psychiatrists (W.X. Quan and W.T. Dong) according to the Structured Clinical Interview for DSM-IV (American Psychiatric Association, 2000). Their age at onset was 24.25 ± 9.05 years, and the duration of illness was 9.56 ± 7.52 years. The chlorpromazine equivalent doses ranged from 56 mg to 1 575 mg daily (mean dose: 484.06 ± 240.04 mg/day). The healthy subjects were enrolled through the local community. Prior to attending the experiment, they were all interviewed to confirm that they had no history of neurological or psychiatric diseases. All schizophrenic and healthy subjects provided written informed consent after the experimental procedure had been fully explained. This study was approved by the ethics committee of Peking University Sixth Hospital. 2.2. Task description We designed a Chinese version of phonological VFT which was similar to a previous study (Chan and Chen, 2004). Participants were asked to sit in a comfortable chair with their eyes open and minimize head movements during the measurements. In order to reduce any possible distractions, the experiment needs to be performed under a quiet environment and participants should keep their mood stable for some time prior to the task. The task paradigm consisted of a 30-s pre-task baseline period, a 60-s task period and a 30-s post-task baseline period. There was a computer screen in front of the participant with a distance of 1 m. Before the pre-task baseline period of the VFT, we provided one Chinese character (‘门’, which indicates door) as an example to make each participant understand how to perform the experiment. During the pre-
Table 1 Demographic characteristics of study participants in each group.
N Age, year Handedness, right-/left-handed Gender, male/female Education background, graduate/undergraduate/high school degrees Age of onset, year Duration of illness, year Chlorpromazine equivalent dose, (mg/day) VFT performance a
Healthy controls
Patients with schizophrenia
100 34.43 ± 12.36 100/0 65/35 20/44/36 – – – 11.29 ± 4.50
140 33.81 ± 11.52 140/0 80/60 14/63/63 24.25 ± 9.05 9.56 ± 7.52 484.06 ± 240.04 9.05 ± 3.90
Chi-square test was used for testing group difference. Otherwise, t-test was used.
Group difference p-value 0.72 0.22a 0.07a
b0.05
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
task and post-task baseline periods, participants were instructed to fix their gaze at the center of the screen and repeat counting from one to five. During the 60-s task period, three Chinese characters (‘白’, ‘天’, and ‘大’, which indicate white, sky, and big, respectively) were presented on the screen in order. The number of characters and their sequences had no change for the different participants. Each character was presented for 2-s and then replaced by the marker ‘+’ which lasted for 18-s at the center of the screen. Therefore, there was a total of 20-s for each cue character. Participants were required to verbally generate as many phrases or four-character idioms as possible beginning with each given character. The reason why we selected these three Chinese characters is that they are more common in daily life and are easily used to generate phrases or four-character idioms by participants with different educational backgrounds. 2.3. NIRS measurement The optical signals were recorded using a 52 multichannel NIRS instrument (ETG-4000, Hitachi Medical Co., Japan) with two wavelengths of near-infrared light (695 nm and 830 nm). This system included 17 light emitters and 16 light detectors, and they were all arranged in a 3 × 11 array to form 52 measurement channels. The array was placed on the surface of the scalp with the lowest channels positioned along the Fp1–Fp2 line according to the international 10–20 system, which was the same as a previous study (Takizawa et al., 2008). Fig. 1 showed the position and distribution of 52 NIRS channels. The measurement region covered approximately the bilateral frontal and temporal cortices (including Brodmann's areas (BA) 9, 10, 44, 45, 46, and 47). We determined the brain region of each NIRS channel based on the mapping between Brodmann's areas and the 10–20 electrode positions (Okamoto et al., 2004). This arrangement allowed us to measure the relative concentration changes of [oxy-Hb] and [deoxy-Hb] in the frontal and superior temporal cortices throughout the experiment. Then the relative concentration of total hemoglobin ([total-Hb]) could be obtained as the sum of [oxy-Hb] and [deoxy-Hb]. The sampling frequency was 10 Hz. We adjusted the NIRS system to make all channels record the signals correctly for a period of time before the experiment began. 2.4. NIRS data analysis Prior to further processing, the raw data were filtered using a band pass filter with cutoff frequencies from 0.0016 Hz to 0.3 Hz (Chaudhary et al., 2011). This step was performed in order to remove the systemic noise and physiological artifacts such as respiratory and heart rhythms. Then, we performed the correlational analysis between demographic variables and brain activation (mean values of [oxy-Hb], [deoxy-Hb] and [total-Hb]) changes during the VFT in both groups. The demographic variables included age, gender and the VFT
53
performance which is defined as the number of produced Chinese phrases or four-character idioms during the VFT. Pearson's correlation was selected for most of the correlational analysis, and point-biserial correlation was used for the correlational analysis of gender due to the fact that gender is a categorical variable (Tate, 1954). For the schizophrenia group, we also calculated Pearson's correlation coefficients between brain activation and duration of illness and chlorpromazine equivalent dose to investigate their effects. In order to examine which channels were activated during the VFT, we firstly averaged the concentrations of NIRS signals ([oxy-Hb], [deoxy-Hb] and [total-Hb]) within two time periods (30-s pre-task baseline and 60-s task period), and then performed paired t tests between the mean values of these two periods at each channel (Chaudhary et al., 2011; Takizawa et al., 2008). The statistical results were corrected for multiple comparisons with a false discovery rate (FDR) method at p b 0.05 (Benjamini and Yekutieli, 2001; Singh and Dan, 2006). The analysis was performed in the schizophrenia group and the healthy group, respectively. Furthermore, we calculated the relative changes of NIRS signals by subtracting the mean values of the pre-task period from the mean values of the task period, and then performed a group comparison between the two groups using Student's two-sample t tests on each channel. FDR correction was also applied for multiple comparisons at p b 0.05 (Benjamini and Yekutieli, 2001; Singh and Dan, 2006). We aimed to assess the statistical differences in brain activation pattern between schizophrenia patients and healthy subjects. In addition, in order to evaluate whether the group differences are independent of sample size, we also calculated the effect size for the group differences of NIRS changes in each channel with Hedges' g method (Hedges, 1981). During the calculation of Hedges' g, we considered the healthy group as the treatment group and the schizophrenia group as the control group (Hedges, 1981; Hentschke and Stüttgen, 2011). 3. Results 3.1. Sample characteristics and behavioral data The sample characteristics of participants in each group are shown in Table 1. Patients with schizophrenia and healthy controls had no significant differences in age (t-test: t = − 0.35, df = 238, p = 0.72), gender (Chi-square test: χ2 = 1.51, df = 1, p = 0.22) and years of education (Chi-square test: χ2 = 5.28, df = 2, p = 0.07). For the cue character ‘白’ (white), participants generated the phrases or four-character idioms like: ‘白色’ (white color), ‘白人’ (white man), ‘白天’ (day), ‘白开水’ (plain water), ‘白吃白喝’ (freeload), et al. For the cue character ‘天’ (sky), the responses included: ‘天空’ (sky), ‘天生’ (inborn), ‘天地’ (heaven and earth), ‘天线’ (antenna), ‘天天向上’ (make progress every day), et al. And for the cue character ‘大’ (big), the responses included: ‘大人’ (adult), ‘大学’ (university), ‘大树’ (tall tree),
Fig. 1. The schematic diagram of near-infrared spectroscopy probe and channel settings. Left is the location of probe array (red band) according to the international 10–20 system. The probe array is centered in the NASION–INION line and its lowest boundary is positioned along the Fp1–Fp2 line. Right is the distribution of all 52 channels shown as yellow circles. Red and blue squares indicate the near-infrared light emitters and detectors, respectively. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
54
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
‘大家 (everybody), ‘大概’ (perhaps), et al. One Chinese character may have several different meanings. For example, besides the principal meaning of white (used in ‘白色’ and ‘白人’), the character ‘白’ has some other meanings like bright (used in ‘白天’), with nothing else (used in ‘白开水’), free (used in ‘白吃白喝’), et al. Therefore, we could determine that most of the generated phrases or four-character idioms are semantically related to the meaning of the given cue character in varying degrees. On the whole, schizophrenia patients produced significantly fewer phrases or four-character idioms during the VFT as compared with the healthy controls (9.05 ± 3.90 versus 11.29 ± 4.50; and t-test: t = − 4.11, df = 238, p b 0.05).
3.2. Correlational analysis We performed the correlational analysis between VFT performance and NIRS changes during the VFT in both groups. The range of correlation coefficient was from −0.38 to 0.34 in the healthy group and from − 0.14 to 0.29 in the schizophrenia group. All the correlation values were not significant after FDR correction (p b 0.05). Therefore, we determined that there was no effect of task performance on the brain activation changes during the VFT in our study. For the correlational analysis of age, most channels showed negative correlations between age and [oxy-Hb] (90%, 47/52 in the healthy group; 90%, 47/52 in the schizophrenia group)/[total-Hb] changes (83%, 43/52 in the healthy group; 77%, 40/52 in the schizophrenia group), and the positive correlations between age and [deoxy-Hb] changes (75%, 39/52 in the healthy group; 67%, 35/52 in the schizophrenia group). These results indicated that brain activations as measured by NIRS were weaker in the elderly participants compared with the young participants in both groups. Although there were several channels showing significant correlations (Ch. 31 and 42 for [deoxy-Hb] in the healthy group; Ch. 3, 14 and 19 for [oxy-Hb], Ch. 7 and 14 for [deoxy-Hb], and Ch. 3, 7 and 13–15 for [total-Hb] in the schizophrenia group), no significant correlations were found in any channels after FDR correction (p b 0.05). The point-biserial correlation of gender illustrated that most channels had positive correlations between gender and [oxy-Hb] (94%, 49/52 in the healthy group; 83%, 43/52 in the schizophrenia group)/[total-Hb] changes (98%, 51/52 in the healthy group; 77%, 40/52 in the schizophrenia group) changes, and negative correlations between gender and [deoxy-Hb] changes (71%, 37/52 in the healthy group; 63%, 33/52 in the schizophrenia group). Since we assigned male a value of 1 and female a value of 0 when calculating the pointbiserial correlation, the correlational results illustrated that the male participants had stronger brain activation than the female participants in both groups. And also, there were several channels showing significant correlation before FDR correction (Ch. 6–10, 15, 17, 19, 21, 26–27, 36–39 and 49–50 for [oxy-Hb], Ch. 1, 6, 9, 16, 19, 26–27 and 39 for [deoxy-Hb], and Ch. 8, 10, 15, 27, 37, 43 and 46 for the [totalHb] in the healthy group; Ch. 19 for [oxy-Hb], and Ch. 14 and 50 for the [total-Hb] in the schizophrenia group). However, no significant correlations between gender and brain activation were found in any channels after FDR correction (p b 0.05). In the schizophrenia group, we also performed the correlational analysis of brain activation with chlorpromazine equivalent dose and duration of illness. The Pearson correlation coefficients ranged from − 0.13 to 0.16 between chlorpromazine equivalent dose and brain activation changes, and ranged from − 0.18 to 0.15 for duration of illness. Our data showed that the brain activation changes during the VFT had no statistically significant correlations with chlorpromazine equivalent dose. Although significant correlations between duration of illness and brain activation changes were found in some channels (Ch. 29 and 49 for [oxy-Hb], Ch. 17, 37–39 and 47–49 for [deoxy-Hb]), none of them was significant after FDR correction (p b 0.05).
3.3. Brain activation during the task period relative to the pre-task baseline The grand averaged waveforms of hemoglobin concentration changes ([oxy-Hb], [deoxy-Hb] and [total-Hb]) during the VFT in the healthy and schizophrenia groups are shown in Fig. 2. We then conducted the statistical tests for brain activation in both groups, respectively. Fig. 3 illustrates the group-level statistical results of paired t-tests between the mean values of the pre-task baseline and of the task period. We determined that, in most cases, the number of channels with statistical significance was more than that without statistical significance. In order to facilitate the observation, we used gray cycles to mark the channels which reached statistical significance during the task period relative to the pre-task baseline. In the healthy group, we observed that there were 50 channels with significant increases of [oxy-Hb] (Ch. 1–5, 7–16, and 18–52), 44 channels with significant decreases of [deoxy-Hb] (Ch. 1–3, 8–9, 11–14, 17–20, and 22–52) and 45 channels with significant increases of [total-Hb] (Ch. 1–5, 8–16, 18–35, 38–46, and 49–52). In the schizophrenia group, we determined that there were 39 channels with significant increases of [oxy-Hb] (Ch. 9, 11–12, 15, 17–20, and 22–52), 35 channels with significant decreases of [deoxy-Hb] (Ch. 9, 11–14, 19–27, 29–37, 39–46, and 49–52) and 16 channels with significant increases of [total-Hb] (Ch. 17, 28–29, 33–34, 38–40, 44–46, and 48–52). All results were corrected using the FDR method at p b 0.05. 3.4. Group comparison of brain activation In Fig. 2, we can easily find that the healthy group has more pronounced cortical activation than the schizophrenia group in most NIRS channels. Fig. 4 illustrates the results of group comparison detected by two-sample t tests. In Fig. 4A, we also used gray cycles to mark the channels which reached statistical significance. We found that schizophrenia patients had significantly lower increases of [oxy-Hb] than the healthy controls at 41 channels (Ch. 1–5, 9–16, 18–25, 27, 29–35, 39–46, and 49–52), lower decreases of [deoxy-Hb] at 24 channels (Ch. 12–13, 17, 22–24, 28–29, 32–35, 37–40, 43–46, and 48–51) and lower increases of [total-Hb] at 24 channels (Ch. 2–3, 11–13, 16, 19–21, 24–25, 30–31, 34–35, 39–42, 44–45, and 50–52). All results were corrected using the FDR method at p b 0.05. In addition, Fig. 4B shows some examples of the actual responses at the individual level with [oxy-Hb] in a representative channel (Ch. 24) from the healthy group and the schizophrenia group. We could observe that there are wide individual differences for the actual responses in both groups. In the healthy group, some actual responses are related to the different cue characters with obvious peaks and valleys (e.g., HC examples 1 and 2 in Fig. 4B), while others are not (e.g., HC examples 3 and 4 in Fig. 4B); In the schizophrenia group, some actual responses (e.g., SZ examples 2 and 3 in Fig. 4B) have more significant fluctuations over time than some others (e.g., SZ examples 1 and 4 in Fig. 4B). However, these fluctuations have no obvious differences among different cue characters. Fig. 5 shows the effect size of group differences in relative changes of brain activation between the healthy group and the schizophrenia group which were calculated based on three types of NIRS signals. 4. Discussion In this study, we collected the large sample of schizophrenia patients and healthy controls, and applied a multichannel NIRS to explore the possible changes of brain activation pattern in the bilateral prefrontal and superior temporal cortices during the performance of the Chinese version VFT. We investigated which NIRS channels were activated during the task period relative to the pre-task baseline within each group, and compared the relative changes of NIRS activation between the two groups. In terms of [oxy-Hb], [deoxy-Hb] and [total-Hb], we observed that the schizophrenia group had reduced brain activation in
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
55
Fig. 2. Grand average waveforms of hemoglobin concentration changes during the VFT in the healthy group (A) and the schizophrenia group (B). The X-axis represents the time (the unit is [sec]), and Y-axis represents hemoglobin concentration (the unit is [mMmm]). Different types of hemoglobin concentrations are presented with different colors (blue for [oxy-Hb], green for [deoxy-Hb] and red for [total-Hb]) in 52 channels. Abbreviations: HC: healthy controls; SZ: schizophrenia patients; R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
the prefrontal cortex and superior temporal cortex, in contrast to the healthy group. Our study demonstrated that, with the advantage of multichannel NIRS, functional impairment in Chinese-speaking patients with schizophrenia could be evaluated with the patterns of hemodynamic response during the designed Chinese version VFT. Assessment of verbal fluency is an important neuropsychological evaluation, and there are two main types of VFT: phonological VFT and semantic VFT (Dieler et al., 2012). Phonological VFT requires subjects to generate as many words as possible beginning with a specific letter, and semantic VFT is designed to produce as many items as possible which belong to a given category. Different types of VFT may relate to their own underlying cognitive conditions of language and executive functions, and result in different brain functional activities in several involved regions (Henry and Crawford, 2004). Compared with phonological VFT, semantic VFT requires the additional involvement of the semantic system (Henry and Crawford, 2004). In addition, one recent study has found the language-dependent effect on the performance of VFT (Sumiyoshi et al., 2014). With the NIRS technique,
previous studies have illustrated that phonological and semantic VFT may cause divergent brain activation patterns (Ehlis et al., 2007; Kubota et al., 2005; Marumo et al., 2014) and different language versions of VFT may also have different brain activation patterns especially for phonological VFT (Dan et al., 2013). Till now, most comparison studies were performed in the English-speaking (Ehlis et al., 2007; Heinzel et al., 2013; Tupak et al., 2012) or Japanese-speaking (Dan et al., 2013; Marumo et al., 2014; Suto et al., 2004; Takizawa et al., 2008; Watanabe and Kato, 2004) population, and few studies were conducted with the Chinese version VFT for native Chinese speakers. Chinese is an ideographic language which is quite different from some other languages (Chan and Chen, 2004). Thus, it's meaningful to develop a Chinese version VFT and investigate the pattern of brain activities during the performance of Chinese version VFT. In the current study, we designed a Chinese version of phonological VFT and applied a multichannel NIRS system to examine the changes of hemodynamic response that was associated with the execution of VFT in the prefrontal and superior temporal cortices. Compared to the pre-task baseline,
56
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
Fig. 3. The statistical results for brain activation during the task period relative to the pre-task baseline within the healthy group (left column) and the schizophrenia group (right column), respectively. The t-value maps are colored according to the color bar at the right. We use gray cycles to mark the channels which reach the statistical significance. The top row is calculated based on [oxy-Hb], the second row is based on [deoxy-Hb] and the bottom row is based on [total-Hb]. Abbreviations: HC: healthy controls; SZ: schizophrenia patients; R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
statistical results in the healthy group indicated that most NIRS channels had significant increases of [oxy-Hb] (about 50 channels), significant decreases of [deoxy-Hb] (about 44 channels) and also significant increases of [total-Hb] (about 45 channels) during the execution period of VFT over both hemispheres (Fig. 3). We determined that the number of activated NIRS channels is more than that in previous similar studies of the Japanese version VFT (Takizawa et al., 2008), and these channels covered the brain regions which are related to phonological association as well as semantic processing (e.g., BA 44 and 45). Such phenomenon can be explained by the fact that most of the generated phrases or four-character idioms are semantically related to the meaning of a given cue character in addition to beginning with the same character. Therefore, our study confirm that Chinese version of phonological VFT is an effective neuropsychological task to investigate the verbal fluency function for native Chinese speakers and NIRS is a suitable tool to
measure the hemoglobin concentration changes caused by cognitive tasks. Previous NIRS studies have applied the English version VFT or Japanese version VFT to investigate the functional impairment of verbal fluency in patients with schizophrenia (Azechi et al., 2010; Ehlis et al., 2007; Kubota et al., 2005; Suto et al., 2004; Takizawa et al., 2008). In the current study, we examined the changes of NIRS signals during the Chinese version VFT in the schizophrenia group. By comparing the brain activation during the task period relative to the pre-task baseline, we observed the significant changes of hemoglobin concentration in some NIRS channels. The number of activated channels in the schizophrenia group also reduced a lot while compared with healthy group (Fig. 3). In the schizophrenia group, the regions which became not activated were predominately located in the bilateral superior temporal and dorsolateral prefrontal cortices. When we compared the relative
Fig. 4. (A) The group comparison for relative changes of brain activation between the healthy group and the schizophrenia group. The t-value maps are colored according to the color bar at the right. We use gray cycles to mark the channels which reach statistical significance. Top row is calculated based on [oxy-Hb], second row is based on [deoxy-Hb] and bottom row is based on [total-Hb]; (B) Some examples of the actual responses with [oxy-Hb] in a representative channel (Ch. 24) from the healthy group and the schizophrenia group. We illustrate four examples in each group, respectively. Each time interval is separated by the black dotted lines. The X-axis represents the time (the unit is [sec]), and the Y-axis represents hemoglobin concentration (the unit is [mMmm]). Abbreviations: HC: healthy controls; SZ: schizophrenia patients; R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
57
Fig. 5. Effect size of group differences in the relative changes of NIRS activation between the healthy group and the schizophrenia group. The top row is calculated based on [oxy-Hb], second row is based on [deoxy-Hb] and bottom row is based on [total-Hb]. The effect size values are colored according to the color bars in the bottom of each subfigure, and channel numbers are labeled according to the distribution map in Fig. 1. Abbreviations: R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
changes of NIRS signals between healthy group and the schizophrenia group, we determined that there were significant statistical differences around the bilateral dorsolateral prefrontal cortex (BA 9 and 46) and ventrolateral prefrontal cortex (BA 44, 45, and 47). Specifically, the schizophrenia patients had decreased brain activities in these brain regions (Fig. 4A). And with the results of the effect size, we also found the significant differences between the two groups to be independent of sample size (Fig. 5). Our results based on the large sample of native Chinese speakers were similar to the findings in some other language version of VFT studies (Azechi et al., 2010; Ehlis et al., 2007; Kubota et al., 2005; Suto et al., 2004; Takizawa et al., 2008). In addition, one previous study showed that schizophrenia also had reduced cortical activation in a wide range of the frontopolar region (BA 10) during the Japanese version VFT (Takizawa et al., 2008). However, we observed that there were only a few NIRS channels showing reduced activation in the frontopolar region. It is well known that performing VFT requires multiple high-level cognitive processes, including language, executive, and working memory (Henry and Crawford, 2004). Therefore, evaluation of the brain activation by using NIRS helps us to understand which functional deficits occur during the VFT in patients with schizophrenia. The dorsolateral prefrontal cortex (BA 9 and 46) has been implicated in sustaining attention and working memory (Owen, 1997). Left BA 44 and 45 are parts of the Broca's area which plays a critical role in phonological and syntactic processing (Owen, 1997). The mid-ventrolateral frontal cortex (BA 47) is involved in some executive function such as selection, comparison and judgment of stimuli held in short-term and long-term memory (Ramnani and Owen, 2004). Therefore, the evidences from our present study
demonstrated the functional impairment of phonemic verbal fluency in Chinese-speaking patients with schizophrenia. There are some limitations of the current study and must be improved in the future. First, although our Chinese version of phonological VFT can simulate a similar mechanism of phonological VFT in English or Japanese, the given Chinese characters may still contain the semantic hint since the character itself is associated with a specific meaning in addition to a fixed syllable (Chan and Chen, 2004). In our study, the Chinese version of phonological VFT activates the brain regions which are related to phonological association as well as semantic processing. Previous studies have demonstrated that phonological VFT has more pronounced brain activation than semantic VFT in English (Ehlis et al., 2007; Heinzel et al., 2013; Tupak et al., 2012) and Japanese (Kubota et al., 2005; Marumo et al., 2014). Compared with their findings, our NIRS results illustrated that the magnitudes of hemoglobin concentration changes in most channels are larger than the Japanese version of phonological VFT, not only in the healthy controls but also in the schizophrenia patients (Marumo et al., 2014). However, based on the current experiment, we cannot determine whether the differences of magnitudes are derived from task-specific activation or language-specific activation. Therefore, further investigation is required to design different Chinese versions of VFT and then compare their similarities and differences. Second, the main objective of this study is not to investigate the effect of psychotropic drugs on the NIRS activation during the VFT, therefore the duration of medicine treatment is not strictly designed and may be changed according to the patient's relevant symptoms. Previous study illustrated that the effect of psychotropic drug in schizophrenia was based on not only the chlorpromazine
58
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
equivalent dose but also the duration of treatment (Chou et al., 2014). Our investigation showed that there were no significant correlations between chlorpromazine equivalent dose and brain activation. Therefore, further study needs to be designed to evaluate the effect of treatment duration. 5. Conclusion In summary, our study demonstrated that the designed Chinese version VFT is an effective experimental paradigm. Compared with healthy controls, the multichannel NIRS results on the large sample confirmed that schizophrenia patients in the Chinese population had significant lower brain activation over the prefrontal cortex and superior temporal cortex during the Chinese version VFT. Such findings can increase our understanding of the brain functional deficits in Chinese-speaking patients with schizophrenia. Acknowledgments We would like to thank all the participants enrolled in this study. This study was financially supported in part by grants from the National Key Basic Research Project (2011CB503705) and National Natural Science Foundation of China (Grant Nos. 61201066 and 61371187). References American Psychiatric Association. Diagnostic and statistical manual of mental disorders. Fourth Edition. Washington, DC: American Psychiatric Press; 2000. Azechi M, Iwase M, Ikezawa K, Takahashi H, Canuet L, Kurimoto R, et al. Discriminant analysis in schizophrenia and healthy subjects using prefrontal activation during frontal lobe tasks: a near-infrared spectroscopy. Schizophr Res 2010;117:52–60. Benjamini Y, Yekutieli D. The control of the false discovery rate in multiple testing under dependency. Ann Stat 2001;29:1165–88. Chan RCK, Chen EYH. Development of a Chinese verbal fluency test for the Hong Kong psychiatric setting. J Psychiatr 2004;14:8–11. Chaudhary U, Hall M, DeCerce J, Rey G, Godavarty A. Frontal activation and connectivity using near-infrared spectroscopy: verbal fluency language study. Brain Res Bull 2011;84:197–205. Chou P, Koike S, Nishimura Y, Kawasaki S, Satomura Y, Kinoshita A, et al. Distinct effects of duration of untreated psychosis on brain cortical activities in different treatment phases of schizophrenia: a multi-channel near-infrared spectroscopy study. Prog Neuropsychopharmacol Biol Psychiatry 2014;49:63–9. Dan H, Dan I, Sano T, Kyutoku Y, Oguro K, Yokota H, et al. Language-specific cortical activation patterns for verbal fluency tasks in Japanese as assessed by multichannel functional near-infrared spectroscopy. Brain Lang 2013;126:208–16. Dieler AC, Tupak SV, Fallgatter AJ. Functional near-infrared spectroscopy for the assessment of speech related tasks. Brain Lang 2012;121:90–109. Ehlis AC, Herrmann MJ, Plichta MM, Fallgatter AJ. Cortical activation during two verbal fluency tasks in schizophrenic patients and healthy controls as assessed by multichannel near-infrared spectroscopy. Psychiatry Res 2007;156:1–13. Ehlis AC, Schneider S, Dresler T, Fallgatter AJ. Application of functional near-infrared spectroscopy in psychiatry. Neuroimage 2014;85:478–88. Ferrari M, Quaresima V. A brief review on the history of human functional near-infrared spectroscopy (fNIRS) development and fields of application. Neuroimage 2012;63: 921–35. Hedges LV. Distribution theory for Glass's estimator of effect size and related estimators. J Educ Behav Stat 1981;6:107–28.
Heinzel S, Metzger FG, Ehlis AC, Korell R, Alboji A, Haeussinger FB, et al. Aging-related cortical reorganization of verbal fluency processing: a functional near-infrared spectroscopy study. Neurobiol Aging 2013;34:439–50. Henry JD, Crawford JR. A meta-analytic review of verbal fluency performance following focal cortical lesions. Neuropsychology 2004;18:284–95. Hentschke H, Stüttgen MC. Computation of measures of effect size for neuroscience data sets. Eur J Neurosci 2011;34:1887–94. Koike S, Nishimura Y, Takizawa R, Yahata N, Kasai K. Near-infrared spectroscopy in schizophrenia: a possible biomarker for predicting clinical outcome and treatment response. Front Psychiatr 2013;4:145. Kubota Y, Toichi M, Shimizu M, Mason RA, Coconcea CM, Findling RL, et al. Prefrontal activation during verbal fluency tests in schizophrenia: a near-infrared spectroscopy (NIRS) study. Schizophr Res 2005;77:65–73. Lynall ME, Bassett DS, Kerwin R, McKenna PJ, Kitzbichler M, Muller U, et al. Functional connectivity and brain networks in schizophrenia. J Neurosci 2010;30:9477–87. Manoach DS, Gollub RL, Benson ES, Searl MM, Goff DC, Halpern E, et al. Schizophrenic subjects show aberrant fMRI activation of dorsolateral prefrontal cortex and basal ganglia during working memory performance. Biol Psychiatry 2000;48:99–109. Marumo K, Takizawa R, Kinou M, Kawasaki S, Kawakubo Y, Fukuda M, et al. Functional abnormalities in the left ventrolateral prefrontal cortex during a semantic fluency task, and their association with thought disorder in patients with schizophrenia. Neuroimage 2014;85:518–26. Narr KL, Bilder RM, Toga AW, Woods RP, Rex DE, Szeszko PR, et al. Mapping cortical thickness and gray matter concentration in first episode schizophrenia. Cereb Cortex 2005;15:708–19. Okamoto M, Dan H, Sakamoto K, Takeo K, Shimizu K, Kohno S, et al. Three-dimensional probabilistic anatomical cranio-cerebral correlation via the international 10–20 system oriented for transcranial functional brain mapping. Neuroimage 2004;21: 99–111. Owen AM. The functional organization of working memory processes within human lateral frontal cortex: the contribution of functional neuroimaging. Eur J Neurosci 1997;9:1329–39. Palaniyappan L, Mallikarjun P, Joseph V, White TP, Liddle PF. Folding of the prefrontal cortex in schizophrenia: regional differences in gyrification. Biol Psychiatry 2011;69: 974–9. Ramnani N, Owen AM. Anterior prefrontal cortex: insights into function from anatomy and neuroimaging. Nat Rev Neurosci 2004;5:184–94. Sasai S, Homae F, Watanabe H, Sasaki AT, Tanabe HC, Sadato N. A NIRS–fMRI study of resting state network. Neuroimage 2012;63:179–93. Singh AK, Dan I. Exploring the false discovery rate in multichannel NIRS. Neuroimage 2006;33:542–9. Steen RG, Mull C, Mcclure R, Hamer RM, Lieberman JA. Brain volume in first-episode schizophrenia: systematic review and meta-analysis of magnetic resonance imaging studies. Br J Psychiatry 2006;188:510–8. Sumiyoshi C, Ertugrul A, Yağcıoğlu AE, Roy A, Jayathilake K, Milby A, et al. Languagedependent performance on the letter fluency task in patients with schizophrenia. Schizophr Res 2014;152:421–9. Suto T, Fukuda M, Ito M, Uehara T, Mikuni M. Multichannel near-infrared spectroscopy in depression and schizophrenia: cognitive brain activation study. Biol Psychiatry 2004; 55:501–11. Takizawa R, Kasai K, Kawakubo Y, Marumo K, Kawasaki S, Yamasue H, et al. Reduced frontopolar activation during verbal fluency task in schizophrenia: a multi-channel near-infrared spectroscopy study. Schizophr Res 2008;99:250–62. Tate RF. Correlation between a discrete and a continuous variable. Point-biserial correlation. Ann Math Stat 1954;3:603–7. Tupak SV, Badewien M, Dresler T, Hahn T, Ernst LH, Herrmann MJ, et al. Differential prefrontal and frontotemporal oxygenation patterns during phonemic and semantic verbal fluency. Neuropsychologia 2012;50:1565–9. Watanabe A, Kato T. Cerebrovascular response to cognitive tasks in patients with schizophrenia measured by near-infrared spectroscopy. Schizophr Bull 2004;30: 435–44. Weiss EM, Hofer A, Golaszewski S, Siedentopf C, Brinkhoff C, Kremser C, et al. Brain activation patterns during a verbal fluency test — a functional MRI study in healthy volunteers and patients with schizophrenia. Schizophr Res 2004;70:287–91.
Contents lists available at ScienceDirect
Progress in Neuro-Psychopharmacology & Biological Psychiatry
Reduced prefrontal activation during a verbal fluency task in Chinese-speaking patients with schizophrenia as measured by near-infrared spectroscopy Wenxiang Quan a,b,c,1, Tongning Wu d,1, Zhaohua Li e, Yuduo Wang e, Wentian Dong a,b,c,⁎, Bin Lv d,⁎⁎ a
Peking University Sixth Hospital, Beijing, China Peking University Institute of Mental Health, Beijing, China c Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China d China Academy of Telecommunication Research of Ministry of Industry and Information Technology, Beijing, China e School of Information and Communication Engineering, Beijing Information Science and Technology University, Beijing, China b
a r t i c l e
i n f o
Article history: Received 12 August 2014 Received in revised form 16 December 2014 Accepted 17 December 2014 Available online 24 December 2014 Keywords: Chinese version verbal fluency test Near-infrared spectroscopy (NIRS) Prefrontal cortex Schizophrenia
a b s t r a c t Near-infrared spectroscopy (NIRS) has been applied to examine the possible functional alternations during the performance of cognitive tasks in schizophrenia. With this technique, previous studies have observed that patients with schizophrenia are often associated with reduced brain activation in the prefrontal cortex during the verbal fluency task (VFT) of the English version or the Japanese version. However, it remains unclear whether there is a brain functional impairment in Chinese-speaking patients with schizophrenia. In this study, we designed a Chinese version of the VFT and performed a multichannel NIRS study in a large group of patients with schizophrenia and healthy controls. We investigated brain activation during the task period of the Chinese version of the VFT within a schizophrenia group and a healthy group, respectively, and compared the relative changes between the two groups. Our results confirmed that Chinese-speaking patients with schizophrenia had significantly lower brain activation in the prefrontal cortex and superior temporal cortex when compared with healthy controls. Such findings based on the NIRS data provided us reliable evidences about brain functional deficits in the Chinese-speaking patients with schizophrenia. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Schizophrenia is a psychiatric disorder which is characterized in general by the disruption in cognitive behavior and the incoordination of affective reaction (American Psychiatric Association, 2000). Besides these impairments in external performance, recent neuroimaging studies have demonstrated that patients with schizophrenia are often associated with the abnormalities of brain structure and function, especially in the prefrontal cortex (Lynall et al., 2010; Manoach et al., 2000; Narr et al., 2005; Palaniyappan et al., 2011; Steen et al., 2006; Abbreviations: BA, Brodmann's area; [deoxy-Hb], relative concentration of deoxygenated hemoglobin; FDR, false discovery rate; MRI, magnetic resonance imaging; NIRS, near-infrared spectroscopy; [oxy-Hb], relative concentration of oxygenated hemoglobin; [total-Hb], relative concentration of total hemoglobin; VFT, verbal fluency task ⁎ Correspondence to: W. Dong, Peking University Sixth Hospital. No. 51, Huayuanbei Road, Haidian District, Beijing 100191, China. Tel.: +86 10 82805966; fax: + 86 10 62026310. ⁎⁎ Correspondence to: B. Lv, China Academy of Telecommunication Research of Ministry of Industry and Information Technology, No. 52, Huayuanbei Road, Haidian District, Beijing 100191, China. Tel.: +86 10 62304633 2084. E-mail addresses: [email protected] (W. Dong), [email protected] (B. Lv). 1 These authors contributed equally to the work.
http://dx.doi.org/10.1016/j.pnpbp.2014.12.005 0278-5846/© 2014 Elsevier Inc. All rights reserved.
Weiss et al., 2004). With structural magnetic resonance imaging (MRI), converging evidences have revealed the structural changes of prefrontal cortex in schizophrenia including brain volume atrophy (Steen et al., 2006), cortical thickness thinning (Narr et al., 2005) and gyrification reduction (Palaniyappan et al., 2011). In addition, more and more studies have utilized functional neuroimaging technologies, e.g., functional MRI (fMRI), to investigate brain functional activation during the resting state (Lynall et al., 2010) as well as various cognitive tasks (Weiss et al., 2004). These results provided us imaging observation on the dysfunction of prefrontal cortex in patients with schizophrenia. Near-infrared spectroscopy (NIRS) is another promising and noninvasive functional neuroimaging technique which can measure the concentration changes of oxygenated hemoglobin ([oxy-Hb]) and deoxygenated hemoglobin ([deoxy-Hb]) near the brain surface (Ferrari and Quaresima, 2012). The NIRS signals ([oxy-Hb] and [deoxy-Hb]) are considered to reflect regional cerebral blood volumes and show strong correlations with fMRI signals (Sasai et al., 2012). Compared with fMRI, NIRS is more portable and less sensitive to motion artifacts. These advantages make it more suitable for the assessment of speech related tasks in schizophrenia (Ehlis et al., 2014; Ferrari and Quaresima, 2012). The verbal fluency task (VFT) is a common language
52
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
related neuropsychological task and patients with schizophrenia are associated with profound impaired verbal fluency (Dieler et al., 2012). Therefore, we can infer that schizophrenia has a risk of abnormal brain functional activation during the performance of a VFT. By means of NIRS, previous studies have indicated that schizophrenia patients had different changes in [oxy-Hb] than those of healthy controls during the VFT, although the brain regions showing abnormal activation are not always consistent (Koike et al., 2013). Watanabe et al. applied one channel NIRS system to investigate the cerebral hemoglobin response in 62 schizophrenia patients and 31 healthy controls (Watanabe and Kato, 2004). They described for the first time that patients with schizophrenia had reduced [oxy-Hb] in the left prefrontal cortex during the VFT. Till now, their sample size for schizophrenia patient is the largest in related NIRS studies (Koike et al., 2013). Then, Suto et al. described the spatial–temporal patterns of NIRS activation in the prefrontal and temporal cortices using two 24 multichannel NIRS systems (Suto et al., 2004). Their results on 13 schizophrenia patients and 16 healthy controls revealed that schizophrenia patients had lower NIRS activation in the bilateral prefrontal and temporal cortices at the start of the VFT task period. This dysfunction was further confirmed in a relatively large group (including 55 schizophrenia patients and 70 healthy subjects) (Takizawa et al., 2008). In addition, the delayed cortical activation was also detected particularly in the frontopolar region (Takizawa et al., 2008). By comparing the performance of phonological VFT and semantic VFT, the more pronounced cortical activation was observed during the phonological VFT compared to the semantic VFT in both schizophrenia patients and healthy controls, and their induced activations in schizophrenia patients were all significantly reduced than those in healthy controls (Ehlis et al., 2007). These findings demonstrated that it seems a reliable biomarker that patients with schizophrenia had significantly less frontal activation during performance in the VFT. However, most of these NIRS studies were conducted on Japanese-speaking (Suto et al., 2004; Takizawa et al., 2008; Watanabe and Kato, 2004) or English-speaking populations (Ehlis et al., 2007), and few studies have been performed in Chinesespeaking people, even for healthy subjects. A recent study demonstrated the effect of languages on the performance of the VFT using Japanese, Turkish, and English-speaking patients with schizophrenia (Sumiyoshi et al., 2014). Therefore, we need to investigate the response pattern of NIRS activition during the Chinese version VFT in healthy controls and its possible alternation in patients with schizophrenia. In the present study, we designed a Chinese version VFT and recruited a large group of native Chinese speakers to perform this cognitive task. The group included 140 patients with schizophrenia and 100 healthy controls. Then, a 52 multichannel NIRS system was used to examine the hemodynamic signals in the bilateral prefrontal and superior temporal cortices during the performance of the Chinese version VFT. Our aim was to evaluate the NIRS cerebral response pattern along with the Chinese version VFT and compare the brain activation between patients with schizophrenia and healthy controls. We hypothesized that the Chinese-speaking patients with schizophrenia had
an altered brain activation compared to healthy controls during the designed Chinese version VFT. 2. Methods 2.1. Subjects One hundred and sixty-five schizophrenia patients and 100 healthy comparison subjects took part in this study. 25 schizophrenia patients were excluded from further analysis because of incomplete collection/ obvious noise (19 patients) or handedness issue (6 left-handed patients). The final sample comprised 140 schizophrenia patients (male/female: 80/60, age: range 17–62 years, mean 33.81 ± 11.52 years) and 100 healthy subjects (male/female: 65/35, age: range 18–78 years, mean 34.43 ± 12.36 years), and they were all righthanded (Table 1). On education background, the schizophrenia group included 14 graduate degrees, 63 undergraduate degrees and 63 high school degrees, and the healthy group included 20 graduate degrees, 44 undergraduate degrees and 36 high school degrees. All participants were native speakers of the Chinese language and can read the Chinese language. Schizophrenia patients were recruited from Peking University Sixth Hospital. Each patient was diagnosed independently by two clinical psychiatrists (W.X. Quan and W.T. Dong) according to the Structured Clinical Interview for DSM-IV (American Psychiatric Association, 2000). Their age at onset was 24.25 ± 9.05 years, and the duration of illness was 9.56 ± 7.52 years. The chlorpromazine equivalent doses ranged from 56 mg to 1 575 mg daily (mean dose: 484.06 ± 240.04 mg/day). The healthy subjects were enrolled through the local community. Prior to attending the experiment, they were all interviewed to confirm that they had no history of neurological or psychiatric diseases. All schizophrenic and healthy subjects provided written informed consent after the experimental procedure had been fully explained. This study was approved by the ethics committee of Peking University Sixth Hospital. 2.2. Task description We designed a Chinese version of phonological VFT which was similar to a previous study (Chan and Chen, 2004). Participants were asked to sit in a comfortable chair with their eyes open and minimize head movements during the measurements. In order to reduce any possible distractions, the experiment needs to be performed under a quiet environment and participants should keep their mood stable for some time prior to the task. The task paradigm consisted of a 30-s pre-task baseline period, a 60-s task period and a 30-s post-task baseline period. There was a computer screen in front of the participant with a distance of 1 m. Before the pre-task baseline period of the VFT, we provided one Chinese character (‘门’, which indicates door) as an example to make each participant understand how to perform the experiment. During the pre-
Table 1 Demographic characteristics of study participants in each group.
N Age, year Handedness, right-/left-handed Gender, male/female Education background, graduate/undergraduate/high school degrees Age of onset, year Duration of illness, year Chlorpromazine equivalent dose, (mg/day) VFT performance a
Healthy controls
Patients with schizophrenia
100 34.43 ± 12.36 100/0 65/35 20/44/36 – – – 11.29 ± 4.50
140 33.81 ± 11.52 140/0 80/60 14/63/63 24.25 ± 9.05 9.56 ± 7.52 484.06 ± 240.04 9.05 ± 3.90
Chi-square test was used for testing group difference. Otherwise, t-test was used.
Group difference p-value 0.72 0.22a 0.07a
b0.05
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
task and post-task baseline periods, participants were instructed to fix their gaze at the center of the screen and repeat counting from one to five. During the 60-s task period, three Chinese characters (‘白’, ‘天’, and ‘大’, which indicate white, sky, and big, respectively) were presented on the screen in order. The number of characters and their sequences had no change for the different participants. Each character was presented for 2-s and then replaced by the marker ‘+’ which lasted for 18-s at the center of the screen. Therefore, there was a total of 20-s for each cue character. Participants were required to verbally generate as many phrases or four-character idioms as possible beginning with each given character. The reason why we selected these three Chinese characters is that they are more common in daily life and are easily used to generate phrases or four-character idioms by participants with different educational backgrounds. 2.3. NIRS measurement The optical signals were recorded using a 52 multichannel NIRS instrument (ETG-4000, Hitachi Medical Co., Japan) with two wavelengths of near-infrared light (695 nm and 830 nm). This system included 17 light emitters and 16 light detectors, and they were all arranged in a 3 × 11 array to form 52 measurement channels. The array was placed on the surface of the scalp with the lowest channels positioned along the Fp1–Fp2 line according to the international 10–20 system, which was the same as a previous study (Takizawa et al., 2008). Fig. 1 showed the position and distribution of 52 NIRS channels. The measurement region covered approximately the bilateral frontal and temporal cortices (including Brodmann's areas (BA) 9, 10, 44, 45, 46, and 47). We determined the brain region of each NIRS channel based on the mapping between Brodmann's areas and the 10–20 electrode positions (Okamoto et al., 2004). This arrangement allowed us to measure the relative concentration changes of [oxy-Hb] and [deoxy-Hb] in the frontal and superior temporal cortices throughout the experiment. Then the relative concentration of total hemoglobin ([total-Hb]) could be obtained as the sum of [oxy-Hb] and [deoxy-Hb]. The sampling frequency was 10 Hz. We adjusted the NIRS system to make all channels record the signals correctly for a period of time before the experiment began. 2.4. NIRS data analysis Prior to further processing, the raw data were filtered using a band pass filter with cutoff frequencies from 0.0016 Hz to 0.3 Hz (Chaudhary et al., 2011). This step was performed in order to remove the systemic noise and physiological artifacts such as respiratory and heart rhythms. Then, we performed the correlational analysis between demographic variables and brain activation (mean values of [oxy-Hb], [deoxy-Hb] and [total-Hb]) changes during the VFT in both groups. The demographic variables included age, gender and the VFT
53
performance which is defined as the number of produced Chinese phrases or four-character idioms during the VFT. Pearson's correlation was selected for most of the correlational analysis, and point-biserial correlation was used for the correlational analysis of gender due to the fact that gender is a categorical variable (Tate, 1954). For the schizophrenia group, we also calculated Pearson's correlation coefficients between brain activation and duration of illness and chlorpromazine equivalent dose to investigate their effects. In order to examine which channels were activated during the VFT, we firstly averaged the concentrations of NIRS signals ([oxy-Hb], [deoxy-Hb] and [total-Hb]) within two time periods (30-s pre-task baseline and 60-s task period), and then performed paired t tests between the mean values of these two periods at each channel (Chaudhary et al., 2011; Takizawa et al., 2008). The statistical results were corrected for multiple comparisons with a false discovery rate (FDR) method at p b 0.05 (Benjamini and Yekutieli, 2001; Singh and Dan, 2006). The analysis was performed in the schizophrenia group and the healthy group, respectively. Furthermore, we calculated the relative changes of NIRS signals by subtracting the mean values of the pre-task period from the mean values of the task period, and then performed a group comparison between the two groups using Student's two-sample t tests on each channel. FDR correction was also applied for multiple comparisons at p b 0.05 (Benjamini and Yekutieli, 2001; Singh and Dan, 2006). We aimed to assess the statistical differences in brain activation pattern between schizophrenia patients and healthy subjects. In addition, in order to evaluate whether the group differences are independent of sample size, we also calculated the effect size for the group differences of NIRS changes in each channel with Hedges' g method (Hedges, 1981). During the calculation of Hedges' g, we considered the healthy group as the treatment group and the schizophrenia group as the control group (Hedges, 1981; Hentschke and Stüttgen, 2011). 3. Results 3.1. Sample characteristics and behavioral data The sample characteristics of participants in each group are shown in Table 1. Patients with schizophrenia and healthy controls had no significant differences in age (t-test: t = − 0.35, df = 238, p = 0.72), gender (Chi-square test: χ2 = 1.51, df = 1, p = 0.22) and years of education (Chi-square test: χ2 = 5.28, df = 2, p = 0.07). For the cue character ‘白’ (white), participants generated the phrases or four-character idioms like: ‘白色’ (white color), ‘白人’ (white man), ‘白天’ (day), ‘白开水’ (plain water), ‘白吃白喝’ (freeload), et al. For the cue character ‘天’ (sky), the responses included: ‘天空’ (sky), ‘天生’ (inborn), ‘天地’ (heaven and earth), ‘天线’ (antenna), ‘天天向上’ (make progress every day), et al. And for the cue character ‘大’ (big), the responses included: ‘大人’ (adult), ‘大学’ (university), ‘大树’ (tall tree),
Fig. 1. The schematic diagram of near-infrared spectroscopy probe and channel settings. Left is the location of probe array (red band) according to the international 10–20 system. The probe array is centered in the NASION–INION line and its lowest boundary is positioned along the Fp1–Fp2 line. Right is the distribution of all 52 channels shown as yellow circles. Red and blue squares indicate the near-infrared light emitters and detectors, respectively. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
54
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
‘大家 (everybody), ‘大概’ (perhaps), et al. One Chinese character may have several different meanings. For example, besides the principal meaning of white (used in ‘白色’ and ‘白人’), the character ‘白’ has some other meanings like bright (used in ‘白天’), with nothing else (used in ‘白开水’), free (used in ‘白吃白喝’), et al. Therefore, we could determine that most of the generated phrases or four-character idioms are semantically related to the meaning of the given cue character in varying degrees. On the whole, schizophrenia patients produced significantly fewer phrases or four-character idioms during the VFT as compared with the healthy controls (9.05 ± 3.90 versus 11.29 ± 4.50; and t-test: t = − 4.11, df = 238, p b 0.05).
3.2. Correlational analysis We performed the correlational analysis between VFT performance and NIRS changes during the VFT in both groups. The range of correlation coefficient was from −0.38 to 0.34 in the healthy group and from − 0.14 to 0.29 in the schizophrenia group. All the correlation values were not significant after FDR correction (p b 0.05). Therefore, we determined that there was no effect of task performance on the brain activation changes during the VFT in our study. For the correlational analysis of age, most channels showed negative correlations between age and [oxy-Hb] (90%, 47/52 in the healthy group; 90%, 47/52 in the schizophrenia group)/[total-Hb] changes (83%, 43/52 in the healthy group; 77%, 40/52 in the schizophrenia group), and the positive correlations between age and [deoxy-Hb] changes (75%, 39/52 in the healthy group; 67%, 35/52 in the schizophrenia group). These results indicated that brain activations as measured by NIRS were weaker in the elderly participants compared with the young participants in both groups. Although there were several channels showing significant correlations (Ch. 31 and 42 for [deoxy-Hb] in the healthy group; Ch. 3, 14 and 19 for [oxy-Hb], Ch. 7 and 14 for [deoxy-Hb], and Ch. 3, 7 and 13–15 for [total-Hb] in the schizophrenia group), no significant correlations were found in any channels after FDR correction (p b 0.05). The point-biserial correlation of gender illustrated that most channels had positive correlations between gender and [oxy-Hb] (94%, 49/52 in the healthy group; 83%, 43/52 in the schizophrenia group)/[total-Hb] changes (98%, 51/52 in the healthy group; 77%, 40/52 in the schizophrenia group) changes, and negative correlations between gender and [deoxy-Hb] changes (71%, 37/52 in the healthy group; 63%, 33/52 in the schizophrenia group). Since we assigned male a value of 1 and female a value of 0 when calculating the pointbiserial correlation, the correlational results illustrated that the male participants had stronger brain activation than the female participants in both groups. And also, there were several channels showing significant correlation before FDR correction (Ch. 6–10, 15, 17, 19, 21, 26–27, 36–39 and 49–50 for [oxy-Hb], Ch. 1, 6, 9, 16, 19, 26–27 and 39 for [deoxy-Hb], and Ch. 8, 10, 15, 27, 37, 43 and 46 for the [totalHb] in the healthy group; Ch. 19 for [oxy-Hb], and Ch. 14 and 50 for the [total-Hb] in the schizophrenia group). However, no significant correlations between gender and brain activation were found in any channels after FDR correction (p b 0.05). In the schizophrenia group, we also performed the correlational analysis of brain activation with chlorpromazine equivalent dose and duration of illness. The Pearson correlation coefficients ranged from − 0.13 to 0.16 between chlorpromazine equivalent dose and brain activation changes, and ranged from − 0.18 to 0.15 for duration of illness. Our data showed that the brain activation changes during the VFT had no statistically significant correlations with chlorpromazine equivalent dose. Although significant correlations between duration of illness and brain activation changes were found in some channels (Ch. 29 and 49 for [oxy-Hb], Ch. 17, 37–39 and 47–49 for [deoxy-Hb]), none of them was significant after FDR correction (p b 0.05).
3.3. Brain activation during the task period relative to the pre-task baseline The grand averaged waveforms of hemoglobin concentration changes ([oxy-Hb], [deoxy-Hb] and [total-Hb]) during the VFT in the healthy and schizophrenia groups are shown in Fig. 2. We then conducted the statistical tests for brain activation in both groups, respectively. Fig. 3 illustrates the group-level statistical results of paired t-tests between the mean values of the pre-task baseline and of the task period. We determined that, in most cases, the number of channels with statistical significance was more than that without statistical significance. In order to facilitate the observation, we used gray cycles to mark the channels which reached statistical significance during the task period relative to the pre-task baseline. In the healthy group, we observed that there were 50 channels with significant increases of [oxy-Hb] (Ch. 1–5, 7–16, and 18–52), 44 channels with significant decreases of [deoxy-Hb] (Ch. 1–3, 8–9, 11–14, 17–20, and 22–52) and 45 channels with significant increases of [total-Hb] (Ch. 1–5, 8–16, 18–35, 38–46, and 49–52). In the schizophrenia group, we determined that there were 39 channels with significant increases of [oxy-Hb] (Ch. 9, 11–12, 15, 17–20, and 22–52), 35 channels with significant decreases of [deoxy-Hb] (Ch. 9, 11–14, 19–27, 29–37, 39–46, and 49–52) and 16 channels with significant increases of [total-Hb] (Ch. 17, 28–29, 33–34, 38–40, 44–46, and 48–52). All results were corrected using the FDR method at p b 0.05. 3.4. Group comparison of brain activation In Fig. 2, we can easily find that the healthy group has more pronounced cortical activation than the schizophrenia group in most NIRS channels. Fig. 4 illustrates the results of group comparison detected by two-sample t tests. In Fig. 4A, we also used gray cycles to mark the channels which reached statistical significance. We found that schizophrenia patients had significantly lower increases of [oxy-Hb] than the healthy controls at 41 channels (Ch. 1–5, 9–16, 18–25, 27, 29–35, 39–46, and 49–52), lower decreases of [deoxy-Hb] at 24 channels (Ch. 12–13, 17, 22–24, 28–29, 32–35, 37–40, 43–46, and 48–51) and lower increases of [total-Hb] at 24 channels (Ch. 2–3, 11–13, 16, 19–21, 24–25, 30–31, 34–35, 39–42, 44–45, and 50–52). All results were corrected using the FDR method at p b 0.05. In addition, Fig. 4B shows some examples of the actual responses at the individual level with [oxy-Hb] in a representative channel (Ch. 24) from the healthy group and the schizophrenia group. We could observe that there are wide individual differences for the actual responses in both groups. In the healthy group, some actual responses are related to the different cue characters with obvious peaks and valleys (e.g., HC examples 1 and 2 in Fig. 4B), while others are not (e.g., HC examples 3 and 4 in Fig. 4B); In the schizophrenia group, some actual responses (e.g., SZ examples 2 and 3 in Fig. 4B) have more significant fluctuations over time than some others (e.g., SZ examples 1 and 4 in Fig. 4B). However, these fluctuations have no obvious differences among different cue characters. Fig. 5 shows the effect size of group differences in relative changes of brain activation between the healthy group and the schizophrenia group which were calculated based on three types of NIRS signals. 4. Discussion In this study, we collected the large sample of schizophrenia patients and healthy controls, and applied a multichannel NIRS to explore the possible changes of brain activation pattern in the bilateral prefrontal and superior temporal cortices during the performance of the Chinese version VFT. We investigated which NIRS channels were activated during the task period relative to the pre-task baseline within each group, and compared the relative changes of NIRS activation between the two groups. In terms of [oxy-Hb], [deoxy-Hb] and [total-Hb], we observed that the schizophrenia group had reduced brain activation in
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
55
Fig. 2. Grand average waveforms of hemoglobin concentration changes during the VFT in the healthy group (A) and the schizophrenia group (B). The X-axis represents the time (the unit is [sec]), and Y-axis represents hemoglobin concentration (the unit is [mMmm]). Different types of hemoglobin concentrations are presented with different colors (blue for [oxy-Hb], green for [deoxy-Hb] and red for [total-Hb]) in 52 channels. Abbreviations: HC: healthy controls; SZ: schizophrenia patients; R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
the prefrontal cortex and superior temporal cortex, in contrast to the healthy group. Our study demonstrated that, with the advantage of multichannel NIRS, functional impairment in Chinese-speaking patients with schizophrenia could be evaluated with the patterns of hemodynamic response during the designed Chinese version VFT. Assessment of verbal fluency is an important neuropsychological evaluation, and there are two main types of VFT: phonological VFT and semantic VFT (Dieler et al., 2012). Phonological VFT requires subjects to generate as many words as possible beginning with a specific letter, and semantic VFT is designed to produce as many items as possible which belong to a given category. Different types of VFT may relate to their own underlying cognitive conditions of language and executive functions, and result in different brain functional activities in several involved regions (Henry and Crawford, 2004). Compared with phonological VFT, semantic VFT requires the additional involvement of the semantic system (Henry and Crawford, 2004). In addition, one recent study has found the language-dependent effect on the performance of VFT (Sumiyoshi et al., 2014). With the NIRS technique,
previous studies have illustrated that phonological and semantic VFT may cause divergent brain activation patterns (Ehlis et al., 2007; Kubota et al., 2005; Marumo et al., 2014) and different language versions of VFT may also have different brain activation patterns especially for phonological VFT (Dan et al., 2013). Till now, most comparison studies were performed in the English-speaking (Ehlis et al., 2007; Heinzel et al., 2013; Tupak et al., 2012) or Japanese-speaking (Dan et al., 2013; Marumo et al., 2014; Suto et al., 2004; Takizawa et al., 2008; Watanabe and Kato, 2004) population, and few studies were conducted with the Chinese version VFT for native Chinese speakers. Chinese is an ideographic language which is quite different from some other languages (Chan and Chen, 2004). Thus, it's meaningful to develop a Chinese version VFT and investigate the pattern of brain activities during the performance of Chinese version VFT. In the current study, we designed a Chinese version of phonological VFT and applied a multichannel NIRS system to examine the changes of hemodynamic response that was associated with the execution of VFT in the prefrontal and superior temporal cortices. Compared to the pre-task baseline,
56
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
Fig. 3. The statistical results for brain activation during the task period relative to the pre-task baseline within the healthy group (left column) and the schizophrenia group (right column), respectively. The t-value maps are colored according to the color bar at the right. We use gray cycles to mark the channels which reach the statistical significance. The top row is calculated based on [oxy-Hb], the second row is based on [deoxy-Hb] and the bottom row is based on [total-Hb]. Abbreviations: HC: healthy controls; SZ: schizophrenia patients; R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
statistical results in the healthy group indicated that most NIRS channels had significant increases of [oxy-Hb] (about 50 channels), significant decreases of [deoxy-Hb] (about 44 channels) and also significant increases of [total-Hb] (about 45 channels) during the execution period of VFT over both hemispheres (Fig. 3). We determined that the number of activated NIRS channels is more than that in previous similar studies of the Japanese version VFT (Takizawa et al., 2008), and these channels covered the brain regions which are related to phonological association as well as semantic processing (e.g., BA 44 and 45). Such phenomenon can be explained by the fact that most of the generated phrases or four-character idioms are semantically related to the meaning of a given cue character in addition to beginning with the same character. Therefore, our study confirm that Chinese version of phonological VFT is an effective neuropsychological task to investigate the verbal fluency function for native Chinese speakers and NIRS is a suitable tool to
measure the hemoglobin concentration changes caused by cognitive tasks. Previous NIRS studies have applied the English version VFT or Japanese version VFT to investigate the functional impairment of verbal fluency in patients with schizophrenia (Azechi et al., 2010; Ehlis et al., 2007; Kubota et al., 2005; Suto et al., 2004; Takizawa et al., 2008). In the current study, we examined the changes of NIRS signals during the Chinese version VFT in the schizophrenia group. By comparing the brain activation during the task period relative to the pre-task baseline, we observed the significant changes of hemoglobin concentration in some NIRS channels. The number of activated channels in the schizophrenia group also reduced a lot while compared with healthy group (Fig. 3). In the schizophrenia group, the regions which became not activated were predominately located in the bilateral superior temporal and dorsolateral prefrontal cortices. When we compared the relative
Fig. 4. (A) The group comparison for relative changes of brain activation between the healthy group and the schizophrenia group. The t-value maps are colored according to the color bar at the right. We use gray cycles to mark the channels which reach statistical significance. Top row is calculated based on [oxy-Hb], second row is based on [deoxy-Hb] and bottom row is based on [total-Hb]; (B) Some examples of the actual responses with [oxy-Hb] in a representative channel (Ch. 24) from the healthy group and the schizophrenia group. We illustrate four examples in each group, respectively. Each time interval is separated by the black dotted lines. The X-axis represents the time (the unit is [sec]), and the Y-axis represents hemoglobin concentration (the unit is [mMmm]). Abbreviations: HC: healthy controls; SZ: schizophrenia patients; R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
57
Fig. 5. Effect size of group differences in the relative changes of NIRS activation between the healthy group and the schizophrenia group. The top row is calculated based on [oxy-Hb], second row is based on [deoxy-Hb] and bottom row is based on [total-Hb]. The effect size values are colored according to the color bars in the bottom of each subfigure, and channel numbers are labeled according to the distribution map in Fig. 1. Abbreviations: R: right; L: left. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this article.)
changes of NIRS signals between healthy group and the schizophrenia group, we determined that there were significant statistical differences around the bilateral dorsolateral prefrontal cortex (BA 9 and 46) and ventrolateral prefrontal cortex (BA 44, 45, and 47). Specifically, the schizophrenia patients had decreased brain activities in these brain regions (Fig. 4A). And with the results of the effect size, we also found the significant differences between the two groups to be independent of sample size (Fig. 5). Our results based on the large sample of native Chinese speakers were similar to the findings in some other language version of VFT studies (Azechi et al., 2010; Ehlis et al., 2007; Kubota et al., 2005; Suto et al., 2004; Takizawa et al., 2008). In addition, one previous study showed that schizophrenia also had reduced cortical activation in a wide range of the frontopolar region (BA 10) during the Japanese version VFT (Takizawa et al., 2008). However, we observed that there were only a few NIRS channels showing reduced activation in the frontopolar region. It is well known that performing VFT requires multiple high-level cognitive processes, including language, executive, and working memory (Henry and Crawford, 2004). Therefore, evaluation of the brain activation by using NIRS helps us to understand which functional deficits occur during the VFT in patients with schizophrenia. The dorsolateral prefrontal cortex (BA 9 and 46) has been implicated in sustaining attention and working memory (Owen, 1997). Left BA 44 and 45 are parts of the Broca's area which plays a critical role in phonological and syntactic processing (Owen, 1997). The mid-ventrolateral frontal cortex (BA 47) is involved in some executive function such as selection, comparison and judgment of stimuli held in short-term and long-term memory (Ramnani and Owen, 2004). Therefore, the evidences from our present study
demonstrated the functional impairment of phonemic verbal fluency in Chinese-speaking patients with schizophrenia. There are some limitations of the current study and must be improved in the future. First, although our Chinese version of phonological VFT can simulate a similar mechanism of phonological VFT in English or Japanese, the given Chinese characters may still contain the semantic hint since the character itself is associated with a specific meaning in addition to a fixed syllable (Chan and Chen, 2004). In our study, the Chinese version of phonological VFT activates the brain regions which are related to phonological association as well as semantic processing. Previous studies have demonstrated that phonological VFT has more pronounced brain activation than semantic VFT in English (Ehlis et al., 2007; Heinzel et al., 2013; Tupak et al., 2012) and Japanese (Kubota et al., 2005; Marumo et al., 2014). Compared with their findings, our NIRS results illustrated that the magnitudes of hemoglobin concentration changes in most channels are larger than the Japanese version of phonological VFT, not only in the healthy controls but also in the schizophrenia patients (Marumo et al., 2014). However, based on the current experiment, we cannot determine whether the differences of magnitudes are derived from task-specific activation or language-specific activation. Therefore, further investigation is required to design different Chinese versions of VFT and then compare their similarities and differences. Second, the main objective of this study is not to investigate the effect of psychotropic drugs on the NIRS activation during the VFT, therefore the duration of medicine treatment is not strictly designed and may be changed according to the patient's relevant symptoms. Previous study illustrated that the effect of psychotropic drug in schizophrenia was based on not only the chlorpromazine
58
W. Quan et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 58 (2015) 51–58
equivalent dose but also the duration of treatment (Chou et al., 2014). Our investigation showed that there were no significant correlations between chlorpromazine equivalent dose and brain activation. Therefore, further study needs to be designed to evaluate the effect of treatment duration. 5. Conclusion In summary, our study demonstrated that the designed Chinese version VFT is an effective experimental paradigm. Compared with healthy controls, the multichannel NIRS results on the large sample confirmed that schizophrenia patients in the Chinese population had significant lower brain activation over the prefrontal cortex and superior temporal cortex during the Chinese version VFT. Such findings can increase our understanding of the brain functional deficits in Chinese-speaking patients with schizophrenia. Acknowledgments We would like to thank all the participants enrolled in this study. This study was financially supported in part by grants from the National Key Basic Research Project (2011CB503705) and National Natural Science Foundation of China (Grant Nos. 61201066 and 61371187). References American Psychiatric Association. Diagnostic and statistical manual of mental disorders. Fourth Edition. Washington, DC: American Psychiatric Press; 2000. Azechi M, Iwase M, Ikezawa K, Takahashi H, Canuet L, Kurimoto R, et al. Discriminant analysis in schizophrenia and healthy subjects using prefrontal activation during frontal lobe tasks: a near-infrared spectroscopy. Schizophr Res 2010;117:52–60. Benjamini Y, Yekutieli D. The control of the false discovery rate in multiple testing under dependency. Ann Stat 2001;29:1165–88. Chan RCK, Chen EYH. Development of a Chinese verbal fluency test for the Hong Kong psychiatric setting. J Psychiatr 2004;14:8–11. Chaudhary U, Hall M, DeCerce J, Rey G, Godavarty A. Frontal activation and connectivity using near-infrared spectroscopy: verbal fluency language study. Brain Res Bull 2011;84:197–205. Chou P, Koike S, Nishimura Y, Kawasaki S, Satomura Y, Kinoshita A, et al. Distinct effects of duration of untreated psychosis on brain cortical activities in different treatment phases of schizophrenia: a multi-channel near-infrared spectroscopy study. Prog Neuropsychopharmacol Biol Psychiatry 2014;49:63–9. Dan H, Dan I, Sano T, Kyutoku Y, Oguro K, Yokota H, et al. Language-specific cortical activation patterns for verbal fluency tasks in Japanese as assessed by multichannel functional near-infrared spectroscopy. Brain Lang 2013;126:208–16. Dieler AC, Tupak SV, Fallgatter AJ. Functional near-infrared spectroscopy for the assessment of speech related tasks. Brain Lang 2012;121:90–109. Ehlis AC, Herrmann MJ, Plichta MM, Fallgatter AJ. Cortical activation during two verbal fluency tasks in schizophrenic patients and healthy controls as assessed by multichannel near-infrared spectroscopy. Psychiatry Res 2007;156:1–13. Ehlis AC, Schneider S, Dresler T, Fallgatter AJ. Application of functional near-infrared spectroscopy in psychiatry. Neuroimage 2014;85:478–88. Ferrari M, Quaresima V. A brief review on the history of human functional near-infrared spectroscopy (fNIRS) development and fields of application. Neuroimage 2012;63: 921–35. Hedges LV. Distribution theory for Glass's estimator of effect size and related estimators. J Educ Behav Stat 1981;6:107–28.
Heinzel S, Metzger FG, Ehlis AC, Korell R, Alboji A, Haeussinger FB, et al. Aging-related cortical reorganization of verbal fluency processing: a functional near-infrared spectroscopy study. Neurobiol Aging 2013;34:439–50. Henry JD, Crawford JR. A meta-analytic review of verbal fluency performance following focal cortical lesions. Neuropsychology 2004;18:284–95. Hentschke H, Stüttgen MC. Computation of measures of effect size for neuroscience data sets. Eur J Neurosci 2011;34:1887–94. Koike S, Nishimura Y, Takizawa R, Yahata N, Kasai K. Near-infrared spectroscopy in schizophrenia: a possible biomarker for predicting clinical outcome and treatment response. Front Psychiatr 2013;4:145. Kubota Y, Toichi M, Shimizu M, Mason RA, Coconcea CM, Findling RL, et al. Prefrontal activation during verbal fluency tests in schizophrenia: a near-infrared spectroscopy (NIRS) study. Schizophr Res 2005;77:65–73. Lynall ME, Bassett DS, Kerwin R, McKenna PJ, Kitzbichler M, Muller U, et al. Functional connectivity and brain networks in schizophrenia. J Neurosci 2010;30:9477–87. Manoach DS, Gollub RL, Benson ES, Searl MM, Goff DC, Halpern E, et al. Schizophrenic subjects show aberrant fMRI activation of dorsolateral prefrontal cortex and basal ganglia during working memory performance. Biol Psychiatry 2000;48:99–109. Marumo K, Takizawa R, Kinou M, Kawasaki S, Kawakubo Y, Fukuda M, et al. Functional abnormalities in the left ventrolateral prefrontal cortex during a semantic fluency task, and their association with thought disorder in patients with schizophrenia. Neuroimage 2014;85:518–26. Narr KL, Bilder RM, Toga AW, Woods RP, Rex DE, Szeszko PR, et al. Mapping cortical thickness and gray matter concentration in first episode schizophrenia. Cereb Cortex 2005;15:708–19. Okamoto M, Dan H, Sakamoto K, Takeo K, Shimizu K, Kohno S, et al. Three-dimensional probabilistic anatomical cranio-cerebral correlation via the international 10–20 system oriented for transcranial functional brain mapping. Neuroimage 2004;21: 99–111. Owen AM. The functional organization of working memory processes within human lateral frontal cortex: the contribution of functional neuroimaging. Eur J Neurosci 1997;9:1329–39. Palaniyappan L, Mallikarjun P, Joseph V, White TP, Liddle PF. Folding of the prefrontal cortex in schizophrenia: regional differences in gyrification. Biol Psychiatry 2011;69: 974–9. Ramnani N, Owen AM. Anterior prefrontal cortex: insights into function from anatomy and neuroimaging. Nat Rev Neurosci 2004;5:184–94. Sasai S, Homae F, Watanabe H, Sasaki AT, Tanabe HC, Sadato N. A NIRS–fMRI study of resting state network. Neuroimage 2012;63:179–93. Singh AK, Dan I. Exploring the false discovery rate in multichannel NIRS. Neuroimage 2006;33:542–9. Steen RG, Mull C, Mcclure R, Hamer RM, Lieberman JA. Brain volume in first-episode schizophrenia: systematic review and meta-analysis of magnetic resonance imaging studies. Br J Psychiatry 2006;188:510–8. Sumiyoshi C, Ertugrul A, Yağcıoğlu AE, Roy A, Jayathilake K, Milby A, et al. Languagedependent performance on the letter fluency task in patients with schizophrenia. Schizophr Res 2014;152:421–9. Suto T, Fukuda M, Ito M, Uehara T, Mikuni M. Multichannel near-infrared spectroscopy in depression and schizophrenia: cognitive brain activation study. Biol Psychiatry 2004; 55:501–11. Takizawa R, Kasai K, Kawakubo Y, Marumo K, Kawasaki S, Yamasue H, et al. Reduced frontopolar activation during verbal fluency task in schizophrenia: a multi-channel near-infrared spectroscopy study. Schizophr Res 2008;99:250–62. Tate RF. Correlation between a discrete and a continuous variable. Point-biserial correlation. Ann Math Stat 1954;3:603–7. Tupak SV, Badewien M, Dresler T, Hahn T, Ernst LH, Herrmann MJ, et al. Differential prefrontal and frontotemporal oxygenation patterns during phonemic and semantic verbal fluency. Neuropsychologia 2012;50:1565–9. Watanabe A, Kato T. Cerebrovascular response to cognitive tasks in patients with schizophrenia measured by near-infrared spectroscopy. Schizophr Bull 2004;30: 435–44. Weiss EM, Hofer A, Golaszewski S, Siedentopf C, Brinkhoff C, Kremser C, et al. Brain activation patterns during a verbal fluency test — a functional MRI study in healthy volunteers and patients with schizophrenia. Schizophr Res 2004;70:287–91.