 |
 |
|
|
|
The deficit in semantic context use has been regularly identified in numerous studies of language in schizophrenia patients, and its psychopathological relevance to the study of thought and communication disorders in schizophrenia has been stressed on many occasions. Thanks to their high level of temporal resolution, the event-related potentials (ERPs) simultaneously provide researchers with an accurate measure of on-line language processing and a direct manifestation of brain activity. The N400 component appears approximately 250 milliseconds to 500 milliseconds after the presentation of a potentially significant stimulus (1). Further, its amplitude varies systematically during semantic processing and can be reduced by various factors that increase an item’s predictability within the local context or by the presence of semantic links between the words. The N400 amplitude is extremely sensitive to the immediate context in which the word appears, whether this context consists of a word or a sentence (2). Generally, N400 reflects the integration of the context: integration is easier and N400 amplitude is lower when the word is congruent with the context (“N400 effect”). Likewise, it has been demonstrated that N400 is sensitive to the ease with which subjects can access the information stored in semantic memory, independently of the plausibility and the predictability of the item within the local context in relation to the semantic links between the words (3). N400 is habitually followed by a late positive component (LPC) which, like N400, has proved to be sensitive to semantic congruence (4,5). This effect, however, exhibits an inverse polarity: incongruent words evoke a more highly positive potential than do congruent words. Although the functional significance of LPC and its relation with N400 have not yet been clearly identified, the predominant opinion is that LPC reflects an additional stage in the processing of the stimulus, occurring after evaluation and before the choice of the response, that is related to the updating of working memory (6). In a somewhat similar vein, Van Petten and others consider that LPC reflects a more elaborative processing of the stimulus and its integration in memory (5). Finally, for Halgren, N400 and LPC reflect different stages in the same cognitive process (7). The influence of the semantic context on word processing can be studied using the lexical decision paradigm. In this task, subjects decide whether or not the target word exists in their language. The presence of a semantic or associative link between the target and the prime word that precedes it has been systematically found to reduce the time required to make this lexical decision. This phenomenon has been termed “semantic priming” (8). The 3 electrophysiologic studies that have used the lexical decision in subjects with schizophrenia report a normal N400 effect in these patients (9–11). These results are particularly unexpected because they contradict most behavioural studies (12), and because almost all N400 studies that have used semantic judgement tasks in a sentence context have found an anomaly of this ERP component (reduction or absence of the N400 effect and [or] an increase in amplitude in the congruent or related condition) (4,13–22). What the 3 lexical decision studies have in common is the high level of structuring of the semantic context, with the overall material containing over 30% related words. The absence of a deficit in schizophrenia patients could therefore be explained through the mobilization of context-processing strategies, owing to the highly structured verbal material used in these studies. At the same time, behavioural studies have clearly established that the semantic priming anomaly specifically concerns patients with thought-disordered schizophrenia (23). Several electrophysiologic studies have also revealed a correlation between formal thought disorders and N400 anomalies (20), or even between such disorders and N400 and LPC anomalies conjointly (24). This clinical variable, however, was not considered in the studies that used lexical decision. Finally, the studies that have investigated LPC in patients with schizophrenia have identified anomalies in this component, which have taken the form of a general reduction in amplitude and the congruence effect (4,16,20,22,24). To provide a more finely honed functional analysis of the processes that were mobilized in our experiments, we consider the data relating to LPC. Thus, the main aim of the current study was to compare the behavioural and electrophysiologic results obtained from a group of schizophrenia patients with formal thought disorders in a lexical decision task, using different levels of structuring of the context. In light of the data reported in the literature, we hypothesized that the reinforcement of the contextual constraints (high proportion of related words) should allow patients with schizophrenia to mobilize semantic context- processing strategies, which should be evidenced by the semantic priming effect and an N400 effect, comparable with those observed in a group of healthy control subjects. Conversely, in a weakly structured lexical decision task (low proportion of related words), the group of schizophrenia patients should be characterized by the absence of semantic priming and by N400 anomalies. Methods and MaterialsSubjects The control group consisted of 24 healthy control subjects (including 15 women), who were matched on age, academic level, and vocabulary level with the patients who suffered from schizophrenia (28). All subjects spoke French as their native language and had normal or corrected-to-normal vision. The exclusion criteria for the 2 groups of subjects were as follows: age under 20 years or over 50 years, neurological antecedents, alcoholic intoxication, or regular or recent use of toxics. Stimuli The type of semantic link within the pairs consisted of an antonym relation or membership of the same semantic category, or constituted a part or whole relation. All words were common, and the word length varied between 3 and 7 letters (mean 5 letters per word). To eliminate any bias because of the way the lists were constructed, the material was tested on 10 healthy subjects who did not take part in the study. Two training lists, with 90 items each, were constructed on the basis of the same rules. Procedure
ERP Recording and Analysis Average waveforms were calculated separately for each subject and stimulus type with reference to the 200-millisecond prestimulus baseline. The data were analyzed by calculating, separately for each subject and recording site, the mean ERP amplitude for each stimulus type over two 200-millisecond epochs: 1) 250 milliseconds to 450 milliseconds after stimulus onset for N400 and 2) 450 milliseconds to 650 milliseconds after stimulus onset for LPC. These epochs were selected after visual inspection of the waveforms. The response times and error percentages were analyzed separately using analyses of variance (ANOVAs), with group and structuring as intergroup factors and relatedness as intragroup factor. The ERP data analyses focused on the midline recording site (Fz, Cz, and Pz). Multivariate analyses of variance (MANOVAs) were performed using group and structuring as intergroup factors and relatedness and electrode site as intragroup factors. The significant interactions were analyzed separately using post hoc Scheffé tests. ResultsBehavioural results Electrophysiologic Results Figure 1a ERPs for control subjects in the highly structured task
Figure 1b ERPs for schizophrenia patients in the highly structured task
The results indicate the presence of a main effect of semantic relation. Analyses of this effect reveal that, in the N400 epoch, the amplitude of the related words is more positive than that of the unrelated words (related word, 1.3 microvolts; unrelated words, 0.2 microvolts; F1,58 = 37.99; P < 0.0001). Figure 2a ERPs for control subjects in the weakly structured task
Figure 2b ERPs for schizophrenia patients in the weakly structured task
We also observed a main effect of electrode, explained by a difference in amplitude between the electrodes in the LPC window (F2,116 = 29.68; P < 0.0001) with maximum positive amplitudes at Pz (2.1 microvolts), compared with Cz (1.6 microvolts; P < 0.04), and at Cz, compared with Fz (0.4 microvolts; P < 0.0001).
The group ´ relatedness interaction was also significant. The analyses show that patients with schizophrenia exhibit an increase in N400 amplitude on related words, compared with control subjects (schizophrenia patients, 0.5 microvolts; control subjects, 2.0 microvolts; F1,58 = 6.15; P < 0.02) and a decrease in LPC amplitude for unrelated words (schizophrenia patients, 1.0 microvolts; control subjects, 1.9 microvolts; F1,58 = 6.49; P < 0.01). Finally, the analyses reveal the presence of a significant structuring ´ relatedness interaction. The N400 effect is significant in the 2 context-structuring conditions. In contrast, the amplitudes for related and unrelated words differ significantly in the LPC window in the highly structured task only (related words, 0.8 microvolts; unrelated words, 1.8 microvolts; F1,58 = 22.42; P < 0.0001). Owing to the absence of a main effect of structuring or of any interaction between this factor and the diagnostic group, we will not conduct separate analyses for each context- structuring condition. Difference Waveform Figure 3 Difference wave form for control subjects and schizophrenia patients
The results reveal a main effect of group (Rao R2,57 = 8.32; P < 0.0007), which indicates a reduction in the amplitude of the N400 effect (F1,58 = 6.55; P < 0.01) and the LPC effect (F1,58 = 7.05; P < 0.01) in patients who suffer from schizophrenia, compared with control subjects (schizophrenia patients, 0.7 microvolts and 0.2 microvolts, respectively; control subjects: 1.8 microvolts and 0.6 microvolts, respectively). We also observed an effect of electrode (Rao R4,55 = 10.11; P < 0.0001), which indicates that the N400 effect is at its maximum for the midline electrodes (–1.5 microvolts), compared with the left hemisphere (–1.1 microvolts) and the right hemisphere (–1.1 microvolts) electrodes. The anterior and posterior distribution of the effects did not differ. DiscussionResults for the Control Subjects Conversely, we observed an effect of this structuring on LPC: in the highly structured task, the unrelated words evoked more highly positive amplitudes than did related words, whereas in the weakly structured task, we observed a persistence of the N400 effect in the LPC window, with the related words evoking greater positive amplitudes. This result indicates that the 2 lexical decision tasks did not mobilize the same cognitive processes. The highly structured task seems to have been processed at a deeper, more elaborative level than the weakly structured task. Further, the words that were unrelated to the context were the object of additional integrative processing. Results for Patients with Schizophrenia Contrary to our hypotheses, we observed N400 anomalies not only in the weakly structured task but also in the highly structured one. Thus, it appears that, in a group of patients with formal thought-disordered schizophrenia, the reinforcement of the contextual constraint, as operationalized in our study, does not permit the mobilization of contextual processes. Likewise, it should be noted that an examination of ERPs reveals that the N400 anomalies are lower in the weakly structured task, probably because of the more superficial level at which this task is processed. The results obtained in our highly structured task differ from those reported by Koyama and others (10,11), even though the proportions of related words are comparable. There are certain notable differences, however, between these studies and our own—first, concerning the clinical characteristics of the patients and second, concerning the nature of the semantic relation and the SOA duration. As we mentioned in our introduction, the semantic priming and N400 anomalies do not seem to affect all schizophrenia patients but only the subgroup characterized by the presence of formal thought disorders. The divergence between our results and those obtained by Koyama and others could be owing to the clinical heterogeneity of the patients, itself underpinned by a cognitive heterogeneity. Should we conclude that the deficit in contextualization processes in schizophrenia patients with formal thought disorders is stable? Should we seek to explain those studies based on behavioural variables that have reported comparable semantic priming effects in schizophrenia patients with formal thought disorders and in control subjects (31,32)? The functional analysis of the experimental tasks makes it possible to advance a complementary hypothesis: perhaps, our highly structured task did not solicit the same type of context processes as those mobilized by studies in which a context effect was reported. In fact, 3 types of mechanism have been proposed to explain the semantic priming effect (33). These are the automatic spreading of activation through the semantic network, a prelexical mechanism for the generation of expectations, and a postlexical mechanism for the verification of semantic compatibility. According to the classic opposition of automatic and controlled processes, the first of these mechanisms should be automatic and be independent of the experimental conditions, whereas the other 2 should be controlled and closely linked to the characteristics of the material and the task. In their verbal material, Koyama and others used only antonymic relations. The pregnancy of the antonymic relations—coupled with the precise anticipations that they permit, the high proportion of related words, and the long SOA (greater than 1000 milliseconds)—likely favoured the mobilization of the prelexical mechanism for the generation of expectations (34). This observation also applies to the studies conducted by Henik and others (31,32) who employed all the conditions that are conducive to the expectation generation mechanism. Conversely, our verbal material used various semantic relations. Our experimental conditions, which employed a less pregnant semantic relation and a shorter SOA (450 milli- seconds), did not call on the expectation-generation mechanism but instead on the postlexical-integration mechanism, known to operate independently of SOA and to have effects that increase as a function of the proportion of related words (34). The persistence of N400 anomalies under these conditions may indicate a stable deficit in the postlexical integrative strategies in schizophrenia patients with formal thought disorders. The value of distinguishing between prelexical predictive strategies and postlexical integrative strategies in a study of contextualization deficits in schizophrenia patients is also reflected in the general language comprehension model recently proposed by Kutas and Federmeier (35). According to this model, the processing of context is thought to preactivate the perceptual and semantic characteristics of the most coherent item or of the most highly predictable item within the given context. Activation then spreads through the semantic network to related concepts, and when the presented item corresponds to the predictions, it is recognized more easily. When this proactive strategy proves defective, the integrative strategy is used. The subject needs to allocate additional attentional resources to compare the characteristics of the target with those of the context (that is, to search for an association with the context words or semantic compatibility) and to integrate the word into the representation of the context stored in working memory. It seems permissible, then, to think that it is precisely this retroactive recourse to the contents of working memory and the additional cognitive effort it involves that is reflected in the LPC amplitude increase following N400. A literature review on schizophrenia patients, in light of this dual-process model, suggests that, whenever it is possible to use the predictive strategy, patients with schizophrenia exhibit a priming effect that is equivalent to, or sometimes greater than, that observed in control subjects. This is the case, for example, in lexical decision tasks that use a highly structured context or in lexical decision tasks containing ambiguous words whose nondominant meaning is strongly induced by the context. Similarly, when the experimental conditions do not necessitate a recourse to the postlexical integrative strategies (that is, passive reading tasks), schizophrenia patients do not exhibit N400 anomalies (4,24). Notably, in our study, the LPC anomaly in the schizophrenia patients was specific to the unrelated condition. (The same result was also obtained by Strandbourg and others [22].) Taken together, the N400 anomalies in the related condition only and the LPC in the unrelated condition only could indicate that subjects with schizophrenia, unlike the control subjects, are unable to use a semantic relation to facilitate word processing (that is, an increase in N400 amplitude). Further, these subjects do not make the additional effort that is required to integrate unrelated words (that is, a reduction in LPC amplitude). We can also observe that the ERP anomalies seen in our study in individuals with schizophrenia do not appear to relate to neuroleptic treatment. We witnessed no correlation between the daily dosage (measured in terms of chlorpromazine equivalent) on the one hand and the N400 or LPC amplitude on the other. This result is in line with those reported in several different studies (14,18–20,24). Moreover, the only study published to date that compares the N400 pattern in a group of schizophrenia patients receiving treatment (haloperidol) with that of a group of untreated schizophrenia patients and a healthy control group on a lexical decision task has revealed the presence of a significant N400 effect in the control group and in the treated patient group, but not in the untreated patient group (9). These results indicate that conventional neuroleptics actually tend to improve the effect of semantic context on N400 amplitude in schizophrenia patients. ConclusionThe absence of any improvement in the processing of context in a condition that facilitated the mobilization of postlexical integrative strategies suggests that the deficit in these processes is stable in schizophrenia patients. This deficit may relate to working-memory anomalies in schizophrenia (20,21,31). These results, taken together with the absence of N400 anomalies observed under conditions that strongly solicit the predictive strategies, lead us to believe that schizophrenia patients process language on the basis of the proactive anticipations induced by the context. They experience difficulties, however, when they have to revise these expectations via the retroactive recourse to the integration of meaning. The weakening of the associations, the loss of goal, and the associations based on assonance or semantic contiguity (that is, the disorganization of language and thought in schizophrenia) could be the clinical expression of this cognitive disorder. The hypothesis that the predictive strategies are preserved while only the integrative strategies are impaired remains to be tested in later research. References1. Kutas M, Hillyard SA. Event-related brain potentials to semantically inappropriate and surprisingly large words. Biol Psychol 1980;11:99–116. 2. Bentin S, McCarthy G, Wood CC. Event-related potentials, lexical decision and semantic priming. Electroencephalogr Clin Neurophysiol 1985;60:343–55. 3. Federmeier KD, Kutas M. A rose by any other name: long term memory structure and sentence processing. J Mem Lang 1999;41:469–95. 4. Mitchell PF, Andrews S, Fox AM, Catts SV, Ward PB, McConaghy N. Active and passive attention in schizophrenia: an ERP study of information processing in a linguistic task. Biol Psychol 1991;32:101–24. 5. Van Petten C, Kutas M, Kluender R, Mitchiner M, Melsaac H. Fractionating the word repetition effect with event-related potentials. J Cogn Neurosci 1991;3:131–50. 6. Garcia-Larrea L, Cézanne-Bert G. P3, positive slow wave and working memory load: a study on the functional correlates of slow wave activity. Electroencephalogr Clin Neurophysiol 1998;108:260–73. 7. Halgren E. Insights from evoked potentials into the neuropsychological mechanisms of reading. In: Scheibel AB, Wechsler AF, editors. Neurobiology of higher cognitive function. New York: Guilford Press; 1990. p 103–49. 8. Meyer DE, Schvanenveldt RW. Facilitation in recognizing pairs of words: evidence of a dependence between retrieval operations. J Exp Psychol 1971;90:227–34. 9. Condray R, Steinhauer SR, Cohen JD, Van Kammen DP, Kasparek A. Modulation of language processing in schizophrenia: effects of context and haloperidol on the event related-potential. Biol Psychiatry 1999;45:1336–55. 10. Koyama S, Nageishi Y, Shimokochi M, Hokama H, Miyazato Y, Miyatani M, and others. The N400 component of event-related potentials in schizophrenic patients: a preliminary study. Electroencephalogr Clin Neurophysiol 1991;78:124–32. 11. Koyama S, Nageishi Y, Shimokochi M, Hokama H, Miyazato Y, Miyatani M, and others. ERPs in schizophrenic patients during word recognition task and reaction times. Electroencephalogr Clin Neurophysiol 1994;92:546–54. 12. Besche-Richard C, Passerieux C. Que nous apprennent les études de reconnaissance de mots chez les schizophrènes? Ann Psychol 2001;101:521–43. 13. Adams J, Faux SF, Nestor PG, Shenton M, Marcy B, Smith S, and others. ERP abnormalities during semantic processing in schizophrenia. Schizophr Res 1993;10:247–57. 14. Grillon C, Ameli R, Glazer WM. N400 and semantic categorisation in schizophrenia. Biol Psychiatry 1991;29:467–80. 15. Matsuoka H, Matsumoto K, Yamazaki H, Sakai H, Miwa S, Yoshida S, and others. Lack of repetition priming effect on visual event-related potentials in schizophrenia. Biol Psychiatry 1999;46:137–40. 16. Nestor PG, Matthew OK, O’Donnell BF, Smith L, Niznikiewicz M, Shenton M, and others. Aberrant semantic activation in schizophrenia: a neurophysiological study. Am J Psychiatry 1997;154:640–6. 17. Niznikiewicz MA, O’Donnell BF, Nestor PG, Smith L, Law S, Karapelou M, and others. ERP assessment of visual and auditory language processing in schizophrenia. J Abnorm Psychol 1997;106:85–94. 18. Ohta K, Uchiyama M, Matsushima E, Toru M. An event-related potential study in schizophrenia using Japanese sentences. Schizophr Res 1999;40:159–70. 19. Olichney JM, Iragui VJ, Kutas M, Nowacki R, Jeste DV. N400 abnormalities in late life schizophrenia and related psychoses. Biol Psychiatry 1997;42:13–23. 20. Salisbury DF, O’Donnell BF, McCarley RW, Nestor PG, Shenton ME. Event-related potentials elicited during a context-free homograph task in normal versus schizophrenic subjects. Psychophysiology 2000;37:456–63. 21. Salisbury DF, Shenton ME, Nestor PG, McCarley RW. Semantic bias, homograph comprehension and event-related potentials in schizophrenia. Clin Neurophysiol 2002;1:383–95. 22. Strandburg RJ, Marsh JT, Brown WS, Asarnow RF, Guthrie D, Harper R, and others. Event-related potential correlates of linguistic information processing in schizophrenics. Biol Psychiatry 1997;42:596–608. 23. Passerieux C, Segui J, Besche C, Chevalier JF, Widlöcher D, Hardy-Baylé MC. Heterogeneity in cognitive functioning of schizophrenic patients evaluated by a lexical decision task. Psychol Med 1997;27:1295–302. 24. Andrews S, Shelley AM, Ward PB, Fox A, Catts SV, McConaghy N. Event-related potentials indices of semantic processing in schizophrenia. Biol Psychiatry 1993;34:443–58. 25. Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull 1987;13:261–74. 26. Andreasen NC. Thought, language and communication disorders. I: Clinical assessment, definition of terms and evaluation of their reliability. Arch Gen Psychiatry 1979;36:1315–21. 27. Hardy-Baylé MC, Hardy P, Dantchev N. Stratégies et moyens thérapeutiques en psychiatrie. Paris: Doin; 1993. 28. Binois R, Pichot P. Test de vocabulaire. Paris: Editions du Centre de Psychologie Appliquée; 1959. 29. Silva-Pereyra J, Harmony T, Villanueva G, Fernandez T, Rodriguez M, Galan L, and others. N400 and lexical decisions: automatic or controlled processing? Clin Neurophysiol 1999;110:813–24. 30. McCarley RW, Niznikiewicz MA, Salisbury DF, Nestor PG, O’Donnell BF, Hirayasu Y. Cognitive dysfunction in schizophrenia: unifying basic research and clinical aspects. Eur Arch Psychiatry Clin Neurosci 1999;249(Suppl 4):IV/69–IV/82. 31. Henik A, Nissimov E, Priel B, Umansky R. Effects of cognitive load on semantic priming in patients with schizophrenia. J Abnorm Psychol 1995;104:576–84. 32. Henik A, Priel B, Umansky R. Attention and automaticity in semantic processing of schizophrenic patients. Neuropsychiatry Neuropsychol Behav Neurol 1992;5:161–9. 33. Neely JH, Keefe DE. Semantic context effects on visual word processing: a hybrid prospective retrospective processing theory. In: Bower JH, editor. The psychology of learning and motivation: advances in research and theory. New York: Academic Press; 1989. p 207–48. 34. Neely JH. Semantic priming effects in visual word recognition: a selective review of current findings and theories. In: Besner D, Humphreys G, editors. Basic processes in reading: visual word recognition. Hillsdale (NJ): Erlbaum; 1991. p 264–336. 35. Kutas M, Federmeier KD. Electrophysiology reveals semantic memory use in language comprehension. Trends Cogn Sci 2000;4:463–70. Author(s)Manuscript received August 2002, revised, and accepted December 2002. 1Assistant Professor, Université Paris VIII, Equipe de Recherche en Psychologie Clinique, Paris, France. 2Hospital Practitioner, Service de Psychiatrie, Centre hospitalier de Versailles, Versailles, France. 3Assistant Professor, Université Paris VIII, Equipe de Recherche en Psychologie Clinique, Paris, France. 4Professor and Hospital Practitioner, Service de Psychiatrie, Centre hospitalier de Versailles, Versailles, France. Address for correspondence: Dr M Kostova, Equipe de Recherche en Psychologie Clinique, Université Paris VIII, 2 rue de la Liberté, 93526 Saint-Denis Cedex 02, France e-mail: milena.kostova@univ-paris8.fr
1 | 2
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
