Perseveration in schizophrenia: failure to generate a plan and relationship with the psychomotor poverty subsyndrome

Abstract

Although perseveration in the Wisconsin Card Sorting Test (WCST) has been studied extensively in schizophrenia, the underlying cognitive dysfunctions are not yet clear. In schizophrenia, perseveration has been found to relate to frontal and striatal abnormalities. Therefore, both a failure to generate a plan as seen in patients with frontal abnormalities, or a failure to execute a plan as observed in Parkinson patients, who suffer primarily from striatal abnormalities, could explain perseveration in schizophrenia. The aim of the present study was to distinguish between these two cognitive dysfunctions, which are described by Frith in his routes-to-action model. The main difference between these dysfunctions is the ability to use external guidance. In the present study, 39 schizophrenic patients and 36 healthy controls were assessed with the California Verbal Learning Test (CVLT) and the WCST, in which use of external guidance can be measured, and with the Positive and Negative Syndrome Scale (PANSS) to determine the relationship with symptomatology. The results showed that half of the schizophrenic patients showed perseveration, which could be explained by a failure to generate a plan and was related to the psychomotor poverty subsyndrome. No evidence was found for a failure to execute a plan. Type of antipsychotic medication used (atypical vs. typical) proved not relevant. The results are discussed in the light of evidence for involvement of the dorsolateral prefrontal cortex in perseveration in schizophrenia.

Introduction

Schizophrenic patients have been found to show perseveration on the Wisconsin Card Sorting Test (WCST) (Butler et al., 1992, Franke et al., 1992, Cuesta et al., 1995, Haut et al., 1996), a test developed to measure the ability to sort stimuli into abstract categories and shift cognitive set (Berg, 1948). Although the WCST has been used with schizophrenic patients extensively, the cognitive dysfunctions underlying perseveration in these patients are not yet clear.

Several neuropsychological studies have provided evidence for an important role for the fronto-striatal loops in shifting cognitive sets in patients with schizophrenia (Robbins, 1991, Elliott et al., 1995, Pantelis et al., 1997, Pantelis et al., 1999). In imaging studies, correlations have been found between WCST performance and both striatal (Rossi et al., 1996, Stratta et al., 1997) and frontal abnormalities (Berman et al., 1986, Weinberger and Berman, 1988, Paulman et al., 1990, Kahn and Davidson, 1995, Okubo et al., 1997). These abnormalities in striatum and frontal cortex of schizophrenic patients have been observed in patients using medication as well as in neuroleptic-naive patients (Wong et al., 1986, Weinberger and Berman, 1988, Andreasen et al., 1992, Buchsbaum et al., 1992, Siegel et al., 1993, Andreasen et al., 1997). Since there are close connections between the prefrontal cortex (PFC) and striatum within the fronto-striatal loops (Alexander et al., 1986), both PFC and striatal abnormalities may lead to the same behavioural deficits.

It has been shown that both patients with frontal lesions and patients with Parkinson's disease (PD), who suffer from degeneration of the nigro-striatal pathway, have problems in shifting cognitive set. However, subtle differences can be detected. It has been found that PD patients are able to benefit from cues to shift set (Fimm et al., 1994, Hsieh et al., 1995), even when these cues do not convey information about the correct solution to a problem (van-Spaendonck et al., 1995). Patients with lesions of the PFC have been found to benefit only from cues that are explicitly related to the content of the solution, like explanation of the possible categories (Delis et al., 1992, Dimitrov et al., 1999, Stuss et al., 2000).

Cues that do not convey information about the solution are, for instance, present in the acquisition phase of the WCST, during which the first category has to be found (van-Spaendonck et al., 1996). In the acquisition phase, problem solving is much more externally guided than in the phase after the first shift. Instructions and feedback can be unequivocally used in the acquisition, whereas after the unannounced shift, internally guided problem solving is required. The subject has to discover that the former solution is no longer correct and has to find alternatives without the help of further instructions or other external guidance (van-Spaendonck et al., 1996). Indeed, it has been frequently found that PD patients show normal performance when no shifting is required (Beatty et al., 1989, Richards et al., 1993, Hsieh et al., 1995, van-Spaendonck et al., 1995). External guidance is also present after the second shift, because then alternative solutions have been generated and the set-up of the test has become clearer. PD patients have been found to improve after the second shift (van-Spaendonck et al., 1995). In contrast with PD patients, frontal patients have problems in all phases of the WCST (Heaton et al., 1993). These differential performances presumably reflect dysfunctions at different information processing levels.

This has clearly been illustrated by Frith et al. (Frith and Done, 1988, Frith, 1992, Cahill and Frith, 1996, Frith and Dolan, 1996). In their model, two routes-to-action: the willed route and the stimulus-driven route are distinguished (Fig. 1). In the willed route a self-generated plan is turned into action or, worded differently, behaviour is internally guided. In the stimulus-driven route, the action is externally guided. A disturbance in the willed route-to-action can lead to perseveration but might also lead to stimulus-bound behaviour or lack of action, depending on the given task. This willed route could be dysfunctional because of a failure to generate a plan or a failure to execute a plan. Frith (1992) proposed a role for the PFC in the generation of a plan and a role for the striatum in the execution of a plan. This model makes clear why PD patients benefit from cues that give no information about the correct solution. They are able to generate plans, but are in need of an external trigger to execute these plans. Frontal patients only take advantage of cues that convey clear information about the solution, because these cues can compensate for the inability to generate a plan.

The goal of the present study was to investigate whether perseveration in schizophrenia can be explained by a failure to generate a plan or a failure to execute a plan. We used a battery of neuropsychological tests including the WCST, the California Verbal Learning Test (CVLT), and memory and intelligence tests to control for confounding factors. The potentially confounding factors emotional discomfort and hostility (Bell et al., 1994) were taken into account because they might affect motivation. Motivational problems are a major concern in schizophrenia research (Elliott and Sahakian, 1995).

The WCST was used to measure perseveration, and was analysed in phases, as was introduced by van-Spaendonck et al. (1995). To test whether a failure to generate a plan can explain perseveration, patients were selected who showed perseveration in all phases of the WCST. Without a plan, one cannot benefit from the external guidance in the test. To test whether a failure to execute a plan can explain perseveration, patients were selected who showed perseveration only in the shift 1 phase, because then no external guidance is present.

After selection of patients on WCST perseveration, the subgroups were tested on performance pattern on the CVLT to confirm the presence of a failure to generate or execute a plan. In the CVLT, the subject can use externally guided serial ordering of the words or internally generated semantic ordering. When a plan cannot be executed, semantic ordering will be decreased, because this requires execution of an internally generated plan. Serial ordering will be increased, because this strategy is externally guided. Without a plan, there will be only a decrease in semantic ordering, because there is no benefit from the subtle cue provided by the serial ordering of the words. It has been found that frontal patients (Freedman and Cermak, 1986) and PD patients (Buytenhuijs et al., 1994) tend not to use semantic strategies to encode the words, and that PD patients instead rely increasingly on serial clustering (Buytenhuijs et al., 1994).

We preselected patients on the use of classical and atypical antipsychotics to detect effects on cognitive performance. Moreover, we wished to distinguish between a failure to execute a plan due to striatal abnormalities related to the disease and a failure to execute a plan caused by the side effects of the classical antipsychotics. It has been found that classical antipsychotics affect set-shifting without cues (Berger et al., 1989), most likely due to their effects at the level of the striatum (Ellenbroek and Cools, 2000). For the atypical antipsychotics, most studies show either no effect on cognition or an improvement (Keefe et al., 1999). We considered failure to execute a plan a side effect of the classical antipsychotics only if it occurred in patients using classical antipsychotics and not in patients using atypical antipsychotics. Failure to execute a plan was considered to be due to the disease schizophrenia, if it occurred irrespective of the antipsychotic used.

Since heterogeneous neuropsychological test performance in schizophrenia is a consistent finding (Crow, 1985, Braff et al., 1991, Heinrichs and Awad, 1993, Elliott and Sahakian, 1995) and the relationship between perseveration and symptomatology is not yet clear, we tested the relationship between symptomatology and failure to generate or execute a plan. We used the three syndromes as distinguished by Liddle: psychomotor poverty, disorganisation, and reality distortion (Liddle, 1987a, Liddle, 1987b, Malla et al., 1993, Cuesta et al., 1995, Gureje et al., 1995).