Differential alteration of automatic semantic processing in treated patients affected by bipolar mania and schizophrenia: An N400 study

Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

Contents lists available at SciVerse ScienceDirect

Progress in Neuro-Psychopharmacology & Biological Psychiatry journal homepage: www.elsevier.com/locate/pnp

Differential alteration of automatic semantic processing in treated patients affected by bipolar mania and schizophrenia: An N400 study☆ Vin Ryu a, Suk Kyoon An b, c, Ra Yeon Ha b, c, Jung Ae Kim b, c, Kyooseob Ha d, Hyun-Sang Cho b, c,⁎ a

Department of Psychiatry, College of Medicine, Konyang University, Daejeon, South Korea Department of Psychiatry, College of Medicine, Yonsei University, Seoul, South Korea Institute of Behavioral Science in Medicine, College of Medicine, Yonsei University, Seoul, South Korea d Department of Neuropsychiatry, Seoul National University Bundang Hospital, Gyeonggi-Do 463-707, South Korea b c

a r t i c l e

i n f o

Article history: Received 5 December 2011 Received in revised form 20 March 2012 Accepted 21 March 2012 Available online 29 March 2012 Keywords: Bipolar disorder Mania N400 Semantic priming Schizophrenia Word-matching task

a b s t r a c t Background: Various formal thought disorders are presented as symptoms by manic patients including pressure of speech, flight of ideas, and more complex speech with strong emotional components. N400 is the event-related potential, in which amplitude is suggested to be a general index of efforts to retrieve stored semantic context, which depends on the stored representation itself and the retrieval cue stimuli. The present study examines N400 components induced by a word-matching task in manic patients, and compare these responses to those induced by the task in schizophrenia and healthy controls. Methods: Twenty manic patients, twenty schizophrenic patients, and twenty healthy controls performed the word-matching task, in which they were presented with 120 (60 congruent and 60 incongruent) word pairs, they were instructed to discriminate whether each word pair was congruent or incongruent. During the task, we recorded the electroencephalogram. Results: Reaction time analysis revealed a main effect for priming, in which reaction times were longer in response to incongruent words than to congruent words in all three participant groups (F = 43.1, p b 0.001) with no group effects (F = 2.3, p = 0.11). N400 analysis showed the main effect for priming (F = 30.2, p b 0.001), for group (F = 5.0, p = 0.01), and the interaction of priming × group (F = 4.6, p = 0.02). Post-hoc analysis of this interaction revealed larger N400 amplitudes to congruent words in manic patients (F = 4.0, p = 0.02) and smaller N400 to incongruent words in schizophrenic patients than in other groups (F = 6.1, p = 0.004). No correlations were found between N400 and symptom severity within patient groups. Conclusions: These findings suggest that priming effects of contextually related word pairs are decreased in patients with bipolar mania, whereas priming N400 responses of contextually unrelated word pairs are increased in schizophrenia. This may be the neurophysiological evidence of abnormal automatic semantic processing in patients with bipolar mania, and it also reflects a qualitative difference in thought and speech disorders between bipolar manic and schizophrenia. © 2012 Elsevier Inc. All rights reserved.

1. Introduction Studies of thought pathology show similar levels or amount of formal thought disorder in mania and schizophrenia (Goodwin and Jamison, 2007). However, qualitative differences exist between the Abbreviations: SOA, stimulus-onset asynchrony; ERP, event-related potential; MINI, mini-international neuropsychiatric interview; IQ, intelligence quotients; K-WAIS, Korean Wechsler adult intelligence scales; BDI, Beck’s depressive inventory; YMRS, Young’s mania rating scales; BPRS, brief psychiatric rating scales; MADRS, Montgomery and Åsberg Depression Rating Scale; TMT, trail making task; ANOVA, analysis of variances. ☆ This study was supported by a grant (A101915) from the Korea Healthcare Technology R&D Project of the Ministry of Health & Welfare of the Republic of Korea. ⁎ Corresponding author at: Department of Psychiatry, College of Medicine, Yonsei University, 696-6 Tanbeol-dong, Gwangju-si, Gyeonggi-do, South Korea. Tel.: + 82 31 760 9403; fax: + 82 31 764 8662. E-mail address: [email protected] (H.-S. Cho). 0278-5846/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.pnpbp.2012.03.009

two patient groups. Manic patients tend to exhibit pressure of speech, flight of ideas, and complex speech with strong emotional components (Lott et al., 2002; Ragin and Oltmanns, 1987). Schizophrenic patients are more likely to exhibit loosened and disorganized associations, disordered thought contents, and peculiar words and phrases (Andreasen, 1979a, 1979b; Solovay et al., 1987). While there is some phenomenological understanding of thought pathology in manic patients, neural investigations into characteristics of manic thought or languages have been neglected. It has been proposed that loosened or bizarre associations observed in schizophrenia may result from abnormally heightened activation of semantic networks (Spitzer, 1997). The spread speed or extent of semantic network can be measured using semantic priming paradigms (Spitzer, 1997). The typical procedure of semantic priming is to have a subject read the prime and then read and respond to the target as fast as possible. The semantic priming effect is the tendency to report shorter reaction

V. Ryu et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

times with a target stimulus (“bread”) when it is preceded by a semantically related stimulus (“butter”) than with an unrelated stimulus (“tree”) (Meyer and Schvaneveldt, 1971). Semantic priming seems to occur through a network of semantically organized and interconnected nodes with the assumption that semantically closer nodes (words) will have more spreading activation together than distant nodes (Collins and Loftus, 1975). Increased priming is observed when the intervals between prime and target stimuli (the stimulus-onset asynchrony, SOA) is short (≤400 ms), especially in schizophrenia patients (Pomarol-Clotet et al., 2008). The short SOAs induce automatic spread of activation and reduce the influence of controlled processing (Neely, 1977). N400 is a negative-going waveform that peaks at approximately 400 ms following the onset of a target that is semantically primed or unprimed with the preceding context. The N400 amplitude is more negative (larger) in response to target words that are preceded by semantically incongruent contextual words (unprimed words). Primed targets elicit a minimal or absent N400 with relatively positive amplitude. The N400 effect may be defined as the attenuation of the N400 amplitude to congruent word relative to incongruent word (Kutas and Hillyard, 1980). The N400 amplitude is suggested to be a general index of effort to retrieve stored context related to a word, which depends on the stored representation of the context and the retrieval cue stimuli (Kutas and Federmeier, 2000). The semantic context can be provided by words, such as in word–word or picture–word-matching tasks (Bentin et al., 1985; Ford et al., 1996; Kiang et al., 2008; Mathalon et al., 2002), by sentences (Kutas and Federmeier, 2000; Sitnikova et al., 2002), and by discourse (Ditman et al., 2008). N400 responses of semantic priming have been studied, mainly in schizophrenia, with inconsistent results. Schizophrenic patients were found to have N400 amplitudes that were less negative (smaller) than normal in response to unprimed words (Mathalon et al., 2002, 2010), more negative (larger) (Kreher et al., 2008), or the same as normal controls (Condray et al., 2003; Kiang et al., 2008). In response to primed words in schizophrenic patients, they were found to have more negative N400 amplitude (Kiang et al., 2008; Mathalon et al., 2010) or the same (Condray et al., 2003; Mathalon et al., 2002) than normal controls. These findings may be interpreted as indicating abnormally broad spread of semantic activation or deficient use of the semantic context (Kiang et al., 2008; Mathalon et al., 2002; Mathalon et al., 2010). However, it remains undetermined whether these language-related N400 abnormalities are specific to schizophrenic thought disorder or whether they characterize schizophrenia as a whole (Kuperberg et al., 2010). N400 studies of mood disorders have been less common. Patients with major depression or dysthymia showed no abnormalities in semantic priming and N400 amplitudes using sentence context

Fig. 1. Schematic illustration of word–word-matching task.

195

(Deldin et al., 2006; Iakimova et al., 2009). These findings suggest intact semantic processes in depressive disorder. To our knowledge, there are no N400 event-related potential (ERP) studies that have investigated language-related semantic priming in manic patients who show characteristic and prominent thought or speech disorder. In the present study, manic patients were compared to schizophrenic and control subjects in N400 response using a lexical decision task. Since there are both qualitative differences and similarities of formal thought disorder between manic and schizophrenic patients, we focused specifically on comparisons of N400 response among manic, schizophrenic, and healthy subjects. We also explored the relationship of priming characteristics to manic symptoms. 2. Methods 2.1. Participants 20 bipolar manic patients and 20 schizophrenic patients were recruited among inpatients admitted to the Severance Mental Health Hospital of Yonsei University Health System. All manic and schizophrenic patients were diagnosed according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (American Psychiatric Association, 2000). Diagnoses of bipolar I disorder and schizophrenia were briefly assessed by the Mini-International Neuropsychiatric Interview (MINI) (Sheehan et al., 1998). Diagnostic workups were performed by two psychiatrists (V.R. and H.S.C.). Patients with schizoaffective disorder, severe personality disorder, recent substance abuse, or rapid cycling bipolar disorder, history of closed head injury, neurological diseases or any other current axis I disorders were excluded. At the time of the EEG recording, manic and schizophrenic patients were clinically stable enough to cooperate with the procedure. For control subjects, we posted a recruitment notice on a website and selected 20 healthy subjects, matching sex and age with patient groups. We also performed the diagnostic workups for healthy controls using the MINI. These healthy volunteers had no histories of bipolar disorder, schizophrenia or other psychiatric disease and did not show any mood or thought signs or symptoms during the interviews. All subjects were right handed as indicated by the Annett's handedness questionnaire (Annett, 1970). The intelligence quotients (IQ) were evaluated using a short form of the Korean Wechsler Adult Intelligence Scale (K-WAIS) comprising three subtests on: “information”, “digit span”, and “picture completion”. Subjective depressive symptoms were measured by Beck's Depressive Inventory (BDI) (Beck et al., 1961). Their mood status and psychopathology were also objectively measured using the YMRS (Young et al., 1978), the Brief Psychiatric Rating Scales (BPRS) (Overall and Gorham, 1962), and the Montgomery and Åsberg Depression Rating Scale (MADRS) (Asberg et al., 1978). This study was approved by the Institutional Review Board of Severance Mental Health Hospital, and it was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants with adequate understanding. The demographic characteristics of the two patient groups and the control group are presented in Table 1. There were no significant differences between groups with regard to age, sex, IQ, or education level. Education was slightly different between the two patient groups, but did not meet statistical significance (F = 2.78, p = 0.07). YMRS scores of manic patients were 25.2 ± 4.9. BPRS scores were 19.2 ± 4.9 in manic patients and 40.5 ± 7.0 in schizophrenic patients. Among manic patients, 8 patients were taking lithium and 12 patients were taking divalproex as mood stabilizers. Nineteen patients with mania were taking antipsychotics and all manic patients were on second generation antipsychotics. As for the antipsychotics, 9 patients took quetiapine (623.4 ± 216.5 mg of chlorpromazineequivalent doses) (American Psychiatric Association, 2006; Rijcken et al., 2003), 8 patients took risperidone (624.3 ± 271.2 mg), and

196

V. Ryu et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

Table 1 Demographic and clinical variables among manic, schizophrenic patients, and healthy controls. Manic patients (N = 20)

Schizophrenic patients (N = 20)

Healthy controls (N = 20)

F, t or χ2

p

31.8 ± 8.5 11/9 13.4 ± 3.4 102.5 ± 12.5 25.2 ± 4.9 2.1 ± 0.9 18.2 ± 4.9 2.1 ± 1.1 41.7 ± 26.2 93.2 ± 32.5 9.7 ± 2.2

29.7 ± 7.4 10/10 14.6 ± 2.0 101.9 ± 10.9

28.6 ± 3.6 10/10 15.2 ± 1.4 105.5 ± 9.4

1.1 0.1 2.8 0.6

0.33 0.94 0.07 0.54

40.5 ± 7.0 2.6 ± 1.4 34.7 ± 14.4 76.7 ± 25.9 10.6 ± 2.2

2.1 ± 1.2 26.0 ± 8.2 54.2 ± 14.9 13.2 ± 2.8

12.1 1.2 3.9 11.8 11.1

b 0.001 0.32 0.03 b 0.001 b 0.001

Medication status Lithium, N (%) Valproic acid, N (%)

8 (40) 12 (60)

– –

– –

Antipsychotics Mean dosea Risperidone, N (%) Quetiapine, N (%) Olanzapine, N (%) Haloperidol, N (%)

621.2 ± 207.3 8(40) 9(45) 2(10) 0

653.1 ± 129.0 10(50) 8(40) 1(5) 1(5)

− 0.57

0.57

Age Sex (m/f) Education (year) Estimated IQ YMRS MADRS BPRS BDI TMT-a (s) TMT-b (s) Digit symbol (score)

Abbreviations: IQ (intelligence quotients); YMRS (Young Mania Rating Scale), MADRS (Montgomery Åsberg Depression Rating Scale, BPRS (Brief Psychiatric Rating Scale), BDI (Beck Depressive Inventory), TMT (Trail Making Task). a Chlorpromazine-equivalent dose.

2 patients took olanzapine (749.3 ± 599.3 mg). The average of chlorpromazine-equivalent dose was 621.2 ± 207.3 mg. Among schizophrenic patients, 10 participants were taking risperidone (673.2 ± 176.1 mg of chlorpromazine-equivalent doses) and 8 participants quetiapine (633.8 ± 196.6 mg). One patient was on olanzapine and another patient was on haloperidol. The mean chlorpromazine-equivalent dose was 653.1 ± 129.0 mg in schizophrenic patients (Table 1). 2.2. Task Participants performed a word-matching task, in which they made decisions about whether prime and target words in a word pair were related in meaning. 60 standardized, semantically congruent primetarget pairs and 60 semantically incongruent prime-target pairs were used (Kim and Lee, 2007) (Fig. 1). For example, two semantically congruent prime-target pairs were today–tomorrow ([o-nŭl]–[næil] in Korean), and picture–painter ([gŭ-rim]–[hwa-ga] in Korean), and two semantically incongruent prime-target pairs were today– shape ([o-nŭl]–[mo-sŭp] in Korean), or picture–someday ([gŭ-rim]– [gŭ-nal] in Korean). Each prime word was presented for 250 ms, followed by a 75-msec interval. Then a target stimulus was presented for 250 ms in a pseudo-randomized order with equal probabilities of being either semantically incongruent or not. Subjects were required to press counterbalanced and predetermined button according to the congruency of prime and target. 2.3. Electroencephalogram recording procedure Electroencephalographic activity was recorded from 64-channel AgCl lead cap according to the international 10/10 system with a 0.05 and 100 Hz band-pass filter and a sampling rate of 1000 Hz/ channel (SynAmpsII). The recordings were referenced to linked electrodes placed on the left and right mastoid processes. Eye blinks and

movements were monitored by electrodes placed near the outer canthus and beneath the left eye. Recording procedures were performed in a dimly lit, quiet, and electrically shielded electroencephalography room. Subjects were seated in a comfortable reclining chair at an eye distance of 50 cm from the computer monitor (visual angle of 9° × 12°). Subjects were instructed to concentrate on the center of the monitor and to avoid eye-blinking as much as possible. The subjects’ performances were monitored by the closed-circuit camera, and subjects were not sleepy during the experiments. Analysis of electroencephalography was carried out on an off-line basis. Salient noises of the electroencephalography were removed by inspection. The electroencephalography was amplified by a 0.1–30 Hz band-pass filter. To control for eye movement artifacts, trials were adjusted by regression from electro-oculograms (Semlitsch et al., 1986). Artifacts were rejected if their amplitude exceeded ± 100 μV. A low-pass filter at 8.5 Hz was used to remove muscular movement, noise, and alpha-wave activity. Epochs were included in the analysis in cases of correct responses to target words. Average correct response rates were 92.5 ± 13.6% (incongruent target words) and 95.9 ± 4.0% (congruent target words). There was no significant difference among groups in task performance accuracy (F(2,59) = 0.26, p = 0.77 (incongruent target words); F(2,59) = 0.55, p = 0.58 (congruent target words)). Event-related potentials were averaged between pre-stimulus at 100 ms and post-stimulus at 900 ms. 2.4. Statistical analysis Reaction time was analyzed by repeated-measures analyses of variances (ANOVA) to assess effects of group (healthy control, schizophrenic patients, or bipolar patients), and congruency (congruent or incongruent). N400 amplitudes were analyzed by repeated-measures ANOVA to assess the effects of group, congruency, anterior–posterior (frontal, central, or parietal), and lateral (left, central, or right) sites of scalp electrode locations. Greenhouse–Geisser corrections for non-sphericity were applied. Pairwise comparisons for the priming effect within groups were performed with paired t-tests. For pairwise comparisons among the three groups, an adjusted p value for multiple comparisons (Least Significant Difference) was used. N400 amplitudes were collapsed across 9 electrodes. SPSS version 17.0 was used for statistical analyses. We calculated Pearson's correlational coefficient between N400 amplitudes at Pz electrodes and symptom severity. 3. Results 3.1. Reaction time A 2-way group × priming analysis of variance (ANOVA) revealed a main effect for prime, in which reaction times to incongruent words were longer than to congruent words (F(1,57) = 43.1, p b 0.001). There were no differences among three groups (F(2,57) = 2.3, p = 0.11). There were no interaction of group by priming (F(2,57) = 0.02, p = 0.98). Mean reaction times were shown in Fig. 2. 3.2. N400 components A 4-way group × priming × anterior–posterior site × lateral site ANOVA revealed a main effect for priming (F(1,57) = 30.2, p b 0.001), in which incongruent words elicited a more negative N400 amplitude than congruent words. This analysis also revealed a main effect of group (F(2,57) = 5.0, p = 0.01), in which N400 amplitudes in manic patients in response to congruent words were larger than those in schizophrenic patients. There was also a priming × group interaction (F(2,57) = 4.6, p = 0.02) (Table 2). We explored this interaction in 2 ways, as was performed in Mathalon et al.'s

V. Ryu et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

197

Fig. 2. Reaction time (ms) and N400 amplitude (μV) to congruous and incongruous words.

(2002) study. First, we assessed priming effects within each group separately. Second, we assessed group effects for congruent words and incongruent words separately. The effect of priming was significant for all groups (t = 3.1, p = 0.01 in manic patients; t = 1.8, p = 0.01 in schizophrenic patients; t = 4.5, p b 0.001 in controls). The group effect for congruent words was significant (F = 4.0, p = 0.02), with manic patients showing a larger (more negative) N400 amplitude compared to healthy controls and schizophrenic patients (p = 0.01 in manic vs. schizophrenic patients, p = 0.74 in schizophrenic patients vs. controls, and p = 0.03 in manic patients vs. controls). The group effect for incongruent words was also significant (F = 6.1, p = 0.004), with schizophrenic patients showing a smaller (less negative) N400 amplitude compared to healthy controls and manic patients (p = 0.002 in manic vs. schizophrenic patients, p = 0.01 in schizophrenic patients vs. controls, and p = 0.71 in manic patients vs. controls) (Fig. 2). Analysis of variance results of the subtraction of the primed N400 from the unprimed N400 (priming effect) revealed the main effect of group (F(2,59) = 3.6, p = 0.03). Post-hoc analysis also revealed that manic patients had a significantly lower priming effect than did control subjects (t = 0.81, p = 0.42 (schizophrenic vs. manic); t = 1.76, p = 0.09 (schizophrenic vs. controls); t = 2.36, p = 0.02 (manic vs. controls)). N400 latencies at Pz electrodes were not significantly different among manic (420.9 ± 41.8 ms), schizophrenic patients (416.0 ± 40.5 ms), and healthy controls (402.8 ± 51.4 ms) (F(2,59) = 0.9, p = 0.42). Grand averages of N400 are shown in Fig. 3. 3.3. Clinical correlations Within the manic group, the total YMRS scores and the scores of the YMRS items, related to manic thought disorder examined (speech (rate and amount), language–thought disorder, and content), were not correlated with N400 amplitudes in response to congruent Table 2 Results of prime (2) × anterior–posterior site (3) × lateral site (3) × group (3) ANOVA for N400 amplitudes. Effect

dfa

F

p

Group (G) Prime (P) Anterior–posterior site (A) Lateral site (L) P×G A×G L×G P×A P×L A×L P×A×G P×L×G A×L×G P×A×L P×A×L×G

2,57 1,57 2,114 2,114 2,57 4,114 4,114 2,114 2,114 4,228 4,114 4,114 8,228 4,228 8,228

5.0 30.2 3.6 1.3 4.5 1.3 2.9 4.4 11.7 6.7 0.3 3.0 2.5 2.1 0.8

0.01 b 0.001 0.05 0.27 0.02 0.28 0.03 0.02 b 0.001 b 0.001 0.83 0.02 0.03 0.09 0.63

a

Degree of freedom.

(0.10–0.54 of p value) or incongruent (0.36–0.84 of p value) words at Pz electrodes. Within the schizophrenic group, none of the total BPRS score or BPRS positive symptoms examined (Conceptual disorganization, Hallucinatory behavior, and Unusual thought content) were correlated with N400 amplitude to congruent (0.38–0.81 of p value) or incongruent (0.33–0.85 of p value) words at Pz electrodes. The completion time of trail making test A was not significantly correlated with N400 amplitudes to congruent (r = − 0.14, p = 0.28) and incongruent (r = −0.09, p = 0.47) stimuli at Pz electrodes. The completion time of trail making test B was also not significantly correlated with N400 amplitudes to congruent (r = − 0.20, p = 0.13) and incongruent (r = −0.03, p = 0.82) stimuli at Pz electrodes. The RT priming effect (RT for incongruent words minus RT for congruent words) was also not significantly related to YMRS subscale scores (0.76–0.90 of p value) in manic patients or BPRS positive symptoms (0.71–0.84 of p value). 3.4. Medication effect We performed further analyses to determine possible drug effects on N400 amplitude. We did not find correlation between the medications (chlorpromazine-equivalent doses) and N400 amplitudes (p > 0.10; for example, r = 0.03, p = 0.85 to incongruent stimuli and r = −0.14, p = 0.38 to congruent stimuli at Pz electrode) in bipolar patients. Chlorpromazine-equivalent doses were not statistically different between the lithium group and the valproate group (673.0 ± 214.6 mg in the lithium group, 583.5 ± 203.5 mg in the valproate group, t = 0.93, p = 0.37). There was no significant difference in N400 amplitudes at Pz electrode between the lithium and the valproate groups with independent t test (t = −0.60, p = 0.56 to incongruent stimuli; t = 0.18, p = 0.86 to congruent stimuli). Then, the ANOVA revealed no significant difference in N400 amplitudes to incongruent stimuli at Pz electrode among the lithium, valproate and control groups (F(2,39) = 0.67, p = 0.52). And the ANOVA revealed a significant difference in N400 amplitudes to congruent stimuli at Pz electrode among the lithium, the valproate and control groups (F(2,39) = 3.25, p = 0.05). Post-hoc analysis revealed a significant difference only between the valproate and control groups (p = 0.03). 4. Discussion We investigated semantic priming effects in bipolar manic and schizophrenic patients by using a semantic priming task and examining N400 responses. Our results replicate earlier findings that smaller (i.e., less negative) N400 amplitudes occur in response to unprimed words, which was reported by Mathalon et al.'s (2002; Mathalon et al., 2010 studies), in schizophrenic subjects compared to normal controls. In contrast, manic patients showed larger (i.e., more negative) amplitudes for primed words compared to normal controls, but no differences in amplitudes for unprimed words in comparison to normal controls. Interestingly, a larger amplitude in response to primed words was reported in an N400 schizophrenia study (Kiang et al., 2008), and it is

198

V. Ryu et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

Fig. 3. Event-related potentials associated with congruent and incongruent word pairs recorded from manic, schizophrenic patients and healthy subjects.

notable that we observed a similar result in manic patients in the current study. For bipolar mania, this cannot be interpreted as indicating deficient use of context or poor memory of context as for schizophrenia (Carter et al., 1996). Rather, it may indicate that the benefits of priming are lost or reduced in the manic state. That is to say, it seems that manic patients treat related words as unrelated words. In other studies, bipolar patients in the manic state have shown impairments in target detection and elevated commission errors in attention or disinhibition tasks (Clark et al., 2001; Strakowski et al., 2009). These are likely to reflect the phenomena of high disinhibition and impulsivity in the manic state and the diminished ability to process external information flexibly and selectively. In animal studies, when incoming stimuli become familiar through repetition or priming, less firing or suppression of some neurons has been observed, suggesting selective inhibition (Miller et al., 1991, 1993). This suggests that repeated or familiar stimuli may be recognized as novel stimuli if selective inhibition does not work. Speculatively, manic patients who have “unprimed” responses to familiar stimuli, instead of normal “primed” responses, may be easily captivated by various environmental stimuli, then inattentive to them and thereby experience a loss of priming benefits. Semantic priming tasks have been said to measure automatic spreading activation, the association strength, and to reduce controlled processes (Hutchison, 2003; Silva-Pereyra et al., 1999). Our experiment used a relatively short SOA of 325 ms. At short intervals (about 300 ms), the priming mechanism is an automatic activation of the semantic network (Barch et al., 1999; Vinogradov et al., 1992). With longer SOAs, controlled and strategic processes are also involved in priming mechanisms in addition to automatic processes (Ober et al., 1997). Although a previous study reported no differences with long (750 ms)

versus short (300 ms) SOAs in schizophrenia, we used a short SOAs since it may reduce possible confusion and make easier to interpret (Kiang et al., 2008). N400 amplitude is known to change inversely with contextual integration strength that reflects holding data, association strength, and ease of accessing data from memory (Brown and Hagoort, 1993; Fischler et al., 1983; Kutas and Federmeier, 2000; Penke et al., 1997). Therefore, the larger N400 amplitude in response to primed stimuli observed in our manic subjects may reflect deficits in working memory for maintaining contextual integration. As described earlier, we found no abnormalities in semantic priming and N400 amplitude in depressed patients (Deldin et al., 2006; Iakimova et al., 2009), and this reflects intact semantic processes in depressive disorder. However, one emotional processing theory suggested that nodes in a semantic network of memory and emotion may represent emotional states, and asymmetric effects of information could result from spreading activation between these connecting nodes (Bower, 1981; Isen et al., 1987). As for emotional impacts on semantic processing, behavioral studies have demonstrated priming effects for emotionally positive and neutral stimuli and inhibition for sad stimuli (Rossell and Nobre, 2004) as well as the facilitative effects of mood induction on lexical decision (Hesse and Spies, 1996). These findings suggest that emotional states, especially manic emotional status with disinhibition and impulsivity, might have specific influences on semantic processing, as indexed by the N400 wave. In the current study, schizophrenic patients demonstrated significantly smaller N400 response to unprimed words compared to normal controls. This finding suggests that unprimed words were more primed than would be expected. In other words, schizophrenic patients seem to regard incongruent words as congruent words. This may reflect insensitive response to language incongruities or

V. Ryu et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

enhanced semantic effects of incongruent words and may be interpreted as an abnormally broad spread of semantic activation or hyperactivity of the semantic network (Mathalon et al., 2002, 2010; Spitzer, 1997). However, it should be noted that other studies using automatic semantic priming did not demonstrate smaller N400 amplitudes in schizophrenia (Condray et al., 2003; Kiang et al., 2008; Kreher et al., 2008). We did not find any correlations between manic or manic-related thought severity and N400 amplitudes in manic patients or between severity of thought disorder and N400 amplitude in schizophrenic patients. This result for schizophrenia is consistent with many other studies which have found no significant relationship between N400 amplitude and symptom severity (Bobes et al., 1996; Mathalon et al., 2010), especially with a short SOA of 300 ms (Kiang et al., 2008), although some studies have reported a relationship (Kreher et al., 2008; Spitzer et al., 1993). Finding no such association suggests that it is the trait-related nature that may be the reason for automatic semantic disturbances in schizophrenia (Mathalon et al., 2010). Similarly, we found no correlation of N400 response with severity of manic state in the current study, and this may mirror trait disturbances of bipolar disorder. However, YMRS score cannot provide correct and extensive measurements of thought disturbances in manic patients, and symptoms of manic patients may fluctuate according to the disease course. Thus, further studies are needed to examine over euthymic phases or longitudinal courses. Our study has some limitations. First, we did not control the psychotropic medications taken by the patients. It has been reported that event-related potentials are not affected by antipsychotic drugs (J. M. Ford et al., 1994). However, recent reports showed possible effects of antipsychotics on N400. For example, a significant priming effect of N400 was shown during haloperidol treatment (Condray et al., 2003) and also increased amplitude of N400 was observed after quetiapine treatment in schizophrenia (Zhang et al., 2009). In this study, we did not find any significant correlation between medications (chlorpromazine-equivalent doses) and N400 amplitudes in our patient groups (p > 0.10). We did not find any significant difference in N400 amplitudes at Pz electrode between the lithium and valproate patients within the manic group. However, our sample size is insufficient to allow us to completely rule out possibility of medication effects. Second, we evaluated IQ using the short form of the K-WAIS, but we did not administer comprehensive cognitive tasks in our assessment. Their administration in future studies would be useful for investigating the relationships of N400 responses with specific cognitive domains. Third, we used the first version of BDI in this study. However, BDI-II as the new version was developed to correspond to the diagnostic criteria in DSM-IV (Beck et al., 1996) and the use of new version may be appropriate. However, recent validation of the Korean version of BDI-II may make its use to be limited. In summary, manic patients showed a larger N400 response to primed words while schizophrenic patients showed a smaller N400 response to unprimed words. These results suggest that manic patients have inappropriately decreased priming effects of closely context-related words, while schizophrenic patients have inappropriately increased priming effects of unrelated words. This electrophysiological finding might reflect the qualitative difference in formal thought or speech disorder between bipolar mania and schizophrenia.

References American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association; 2000. American Psychiatric Association. American Psychiatric Association practice guidelines for the treatment of psychiatric disorders. Compendium 2006. American Psychiatric Pub; 2006. Andreasen NC. Thought, language, and communication disorders: I. Clinical assessment, definition of terms, and evaluation of their reliability. Arch Gen Psychiatry 1979a;36:1315.

199

Andreasen NC. Thought, language, and communication disorders: II. Diagnostic significance. Arch Gen Psychiatry 1979b;36:1325. Annett M. A classification of hand preference by association analysis. Br J Psychol 1970;61:303–21. Asberg M, Montgomery SA, Perris C, Schalling D, Sedvall G. A comprehensive psychopathological rating scale. Acta Psychiatr Scand Suppl 1978:5-27. Barch DM, Carter CS, Perlstein W, Baird J, Cohen JD, Schooler N. Increased Stroop facilitation effects in schizophrenia are not due to increased automatic spreading activation. Schizophr Res 1999;39:51–64. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71. Beck AT, Steer RA, Brown GK. Manual for Beck depression inventory II (BDI-II). San Antonio, TX: Psychology Corporation; 1996. Bentin S, McCarthy G, Wood CC. Event-related potentials, lexical decision and semantic priming. Electroen Clin Neuro 1985;60:343–55. Bobes MA, Lei ZX, Ibanez S, Yi H, Valdes-Sosa M. Semantic matching of pictures in schizophrenia: a cross-cultural ERP study. Biol Psychiatry 1996;40:189–202. Bower GH. Mood and memory. Am Psychol 1981;36:129–48. Brown C, Hagoort P. The processing nature of the N400: evidence from masked priming. J Cognitive Neurosc 1993;5:34–44. Carter C, Robertson L, Nordahl T, Chaderjian M, Kraft L, O'Shora-Celaya L. Spatial working memory deficits and their relationship to negative symptoms in unmedicated schizophrenia patients. Biol Psychiatry 1996;40:930–2. Clark L, Iversen SD, Goodwin GM. A neuropsychological investigation of prefrontal cortex involvement in acute mania. Am J Psychiatry 2001;158:1605–11. Collins A, Loftus E. A spreading activation theory of semantic processing. Psychol Rev 1975;82:407–28. Condray R, Siegle GJ, Cohen JD, van Kammen DP, Steinhauer SR. Automatic activation of the semantic network in schizophrenia: evidence from event-related brain potentials. Biol Psychiatry 2003;54:1134–48. Deldin P, Keller J, Casas BR, Best J, Gergen J, Miller GA. Normal N400 in mood disorders. Biol Psychol 2006;71:74–9. Ditman T, Holcomb PJ, Kuperberg GR. Time travel through language: temporal shifts rapidly decrease information accessibility during reading. Psychon Bull Rev 2008;15:750–6. Fischler I, Bloom PA, Childers DG, Roucos SE, Perry Jr NW. Brain potentials related to stages of sentence verification. Psychophysiology 1983;20:400–9. Ford JM, White PM, Csernansky JG, Faustman WO, Roth WT, Pfefferbaum A. ERPs in schizophrenia: effects of antipsychotic medication. Biol Psychiatry 1994;36:153–70. Ford JM, Woodward SH, Sullivan EV, Isaacks BG, Tinklenberg JR, Yesavage JA, et al. N400 evidence of abnormal responses to speech in Alzheimer's disease. Electroen Clin Neuro 1996;99:235–46. Goodwin FK, Jamison KR. Manic-Depressive Illness: Bipolar Disorders and Recurrent Depression. 2nd Edition. New York: Oxford University Press; 2007. Hesse FW, Spies K. Effects of negative mood on performance: reduced capacity or changed processing strategy? Eur J Soc Psychol 1996;26:163–8. Hutchison KA. Is semantic priming due to association strength or feature overlap? A microanalytic review. Psychon Bull Rev 2003;10:785–813. Iakimova G, Passerieux C, Foynard M, Fiori N, Besche C, Laurent J, et al. Behavioral measures and event-related potentials reveal different aspects of sentence processing and comprehension in patients with major depression. J Affect Disord 2009;113: 188–94. Isen AM, Daubman KA, Nowicki GP. Positive affect facilitates creative problem solving. J Pers Soc Psychol 1987;52:1122–31. Kiang M, Kutas M, Light GA, Braff DL. An event-related brain potential study of direct and indirect semantic priming in schizophrenia. Am J Psychiatry 2008;165:74–81. Kim SK, Lee HW. The semantic priming effects of young and older adults in Korean word recognition. Korean J Exp Psychol 2007;19:279–97. Kreher DA, Holcomb PJ, Goff D, Kuperberg GR. Neural evidence for faster and further automatic spreading activation in schizophrenic thought disorder. Schizophr Bull 2008;34:473–82. Kuperberg GR, Kreher DA, Ditman T. What can event-related potentials tell us about language, and perhaps even thought, in schizophrenia? Int J Psychophysiol 2010;75:66–76. Kutas M, Federmeier KD. Electrophysiology reveals semantic memory use in language comprehension. Trends Cogn Sci 2000;4:463–70. Kutas M, Hillyard SA. Reading senseless sentences: brain potentials reflect semantic incongruity. Science 1980;207:203–5. Lott PR, Guggenbuhl S, Schneeberger A, Pulver AE, Stassen HH. Linguistic analysis of the speech output of schizophrenic, bipolar, and depressive patients. Psychopathology 2002;35:220–7. Mathalon DH, Faustman WO, Ford JM. N400 and automatic semantic processing abnormalities in patients with schizophrenia. Arch Gen Psychiatry 2002;59:641–8. Mathalon DH, Roach BJ, Ford JM. Automatic semantic priming abnormalities in schizophrenia. Int J Psychophysiol 2010;75:157–66. Meyer DE, Schvaneveldt RW. Facilitation in recognizing pairs of words: evidence of a dependence between retrieval operations. J Exp Psychol 1971;90:227–34. Miller EK, Li L, Desimone R. A neural mechanism for working and recognition memory in inferior temporal cortex. Science 1991;254:1377. Miller EK, Li L, Desimone R. Activity of neurons in anterior inferior temporal cortex during a short-term memory task. J Neurosci 1993;13:1460. Neely JH. Semantic priming and retrieval from lexical memory: roles of inhibitionless spreading activation and limited-capacity attention. J Exp Psychol Gen 1977;106: 226. Ober BA, Vinogradov S, Shenaut GK. Automatic versus controlled semantic priming in schizophrenia. Neuropsychology 1997;11:506–13.

200

V. Ryu et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 38 (2012) 194–200

Overall JE, Gorham DR. The brief psychiatric rating scale. Psychological reports; 1962. Penke M, Weyerts H, Gross M, Zander E, Munte TF, Clahsen H. How the brain processes complex words: an event-related potential study of German verb inflections. Cognitive Brain Res 1997;6:37–52. Pomarol-Clotet E, Oh TM, Laws KR, McKenna PJ. Semantic priming in schizophrenia: systematic review and meta-analysis. Br J Psychiatry 2008;192:92–7. Ragin AB, Oltmanns TF. Communicability and thought disorder in schizophrenics and other diagnostic groups. A follow-up study. Br J Psychiatry 1987;150:494–500. Rijcken CA, Monster T, Brouwers JR, de Jong-van den Berg LT. Chlorpromazine equivalents versus defined daily doses: how to compare antipsychotic drug doses? J Clin Psychopharmacol 2003;23:657–9. Rossell SL, Nobre AC. Semantic priming of different affective categories. Emotion 2004;4:354. Semlitsch HV, Anderer P, Schuster P, Presslich O. A solution for reliable and valid reduction of ocular artifacts, applied to the P300 ERP. Psychophysiology 1986;23: 695–703. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. The MiniInternational Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 1998;59(Suppl. 20):22–33. quiz 4–57. Silva-Pereyra J, Harmony T, Villanueva G, Fernandez T, Rodriguez M, Galan L, et al. N400 and lexical decisions: automatic or controlled processing? Clin Neurophysiol 1999;110:813–24.

Sitnikova T, Salisbury DF, Kuperberg G, Holcomb PJ. Electrophysiological insights into language processing in schizophrenia. Psychophysiology 2002;39:851–60. Solovay MR, Shenton ME, Holzman PS. Comparative studies of thought disorders: I. Mania and schizophrenia. Arch Gen Psychiatry 1987;44:13. Spitzer M. A cognitive neuroscience view of schizophrenic thought disorder. Schizophr Bull 1997;23:29–50. Spitzer M, Braun U, Hermle L, Maier S. Associative semantic network dysfunction in thought-disordered schizophrenic patients: direct evidence from indirect semantic priming. Biol Psychiatry 1993;34:864–77. Strakowski SM, Fleck DE, DelBello MP, Adler CM, Shear PK, McElroy SL, et al. Characterizing impulsivity in mania. Bipolar Disord 2009;11:41–51. Vinogradov S, Ober BA, Shenaut GK. Semantic priming of word pronunciation and lexical decision in schizophrenia. Schizophr Res 1992;8:171–81. Young RC, Biggs JT, Ziegler VE, Meyer DA. A rating scale for mania: reliability, validity and sensitivity. Br J Psychiatry 1978;133:429–35. Zhang Y, Lehmann M, Shobeiry A, Höfer D, Johannes S, Emrich H, et al. Effects of quetiapine on cognitive functions in schizophrenic patients: a preliminary singletrial ERP analysis. Pharmacopsychiatry 2009;42:129–34.