The clinical neuropsychology of schizophrenia
- Premorbid deficits
- Illness-related deficits
- Symptom- and syndrome-related deficits
- Executive function
- Reality monitoring
- Psychophysiological and conditioning models
- Hemisphere dysfunction
- Practical implications
Neuropsychology forms a bridge between the phenomenology and clinical features of schizophrenia, and the underlying pathology. The cognitive sciences have been increasingly adopted as a framework for discussing such features at the expense of a purely descriptive psychopathology. The advantage of this is that cognitive models strive to be mechanistic and explanatory although, like phenomenology, they sometimes succeed only in displacing one lot of arcane jargon for another. The neuro- prefix is a relatively new addition and simply reflects the often tacit acceptance of a biological basis for the changes observed, but in fact earlier discussions of for instance ‘disorders of attention in schizophrenia' which made no reference to the brain, would now be placed under the heading of neuropsychology. (1)
In his description of dementia praecox, Kraepelin emphasized the presence of clear-cut intellectual decline (see: A conceptual history of schizophrenia), whereas Bleuler (2) remarked:
It shows a complete misunderstanding of the peculiarities of schizophrenia if one believes that schizophrenic dementia can be proved or excluded by means of an ‘intelligence test'...The actual amount of knowledge remains preserved...but it is not always available or it is employed in the wrong way.
Paradoxically, both positions are correct. There are aspects of the neuropsychology of schizophrenia which are akin to a dementia (i.e. an apparent decline in function across several cognitive domains) in conjunction with other aspects most unlike a straightforward dementia, such as the variability of test performance across and within individuals, and lack of inexorable progression. However, attempts to distinguish schizophrenia from neurological disorders on the basis of neuropsychological tests have failed—the overlap in test scores is surprisingly large and classification rates on this basis, no better than chance. (3)
It was previously common for all-embracing psychological theories of schizophrenia to be advanced. This is less common nowadays, where empirical findings outweigh theoretical speculation. This article is not intended to be a comprehensive review of the work in this area but rather a brief overview of clinically relevant findings in the neuropsychology of schizophrenia. The approach taken here will be to divide the neuropsychology of schizophrenia into premorbid, illness-related, and symptom-related deficits, with discussion of whether these deficits are generalized or specific.
The majority of patients with schizophrenia have cognitive impairment. (4) This is not entirely due to ‘cognitive decline', but to some extent a lower than average IQ detectable premorbidly. Part of the evidence for this comes from two large population-based cohort studies in the United Kingdom (5,6) and a Swedish conscript cohort. (7) In these studies individuals were given intelligence tests during childhood or adolescence, at which time, none was manifesting signs of schizophrenia. After several years, some of the cohort developed schizophrenia and, as a group, when compared to the remainder, they showed a lower than average IQ (by the equivalent of about 5–10 points). Similarly, when school reports of children who later developed schizophrenia are compared with their classmates or siblings, or best of all, identical twins, the results show the same trend. (8) These premorbid differences on IQ tests are, by definition, generalized although language and planning deficits may be slightly more obvious.
Getting an accurate picture of premorbid deficits in routine clinical practice is difficult unless school reports over many years have been retained. However, a rough idea can be gained simply by establishing the level of education achieved (if clearly within the premorbid period) and contrasting this to family expectations and social norms. Deviation from an expected level or in comparison to siblings can provide additional comparisons. Tests of reading ability such as the National Adult Reading Test, (9) have been shown to be relatively immune from acquired cognitive decline so may serve as a proxy for premorbid ability.
When individuals with a diagnosis of schizophrenia are studied in contrast with ‘normal controls', the neuropsychological gulf between them is (again, on average) wide, much wider than that between those destined to develop the disorder but not identified as abnormal at the time of testing; in fact the discrepancy is of the order of 1.5 to 3.0 standard deviations (SD) below the normal control mean (10) (1 SD is equal to 15 IQ points). This applies to established cases, as well as first onset and drug-free patients (11) and there is surprisingly little resolution after the episode has abated. (12) Hence cognitive impairment is not an artefact of medication, chronicity, or ‘institutionalization'. Although truly longitudinal neuropsychological studies spanning premorbid and post-onset periods are elusive, the obvious assumption is that the onset of the illness brings with it a further substantial cognitive decline. Such studies carried out in the 1950s and 1960s in North America using conscript cohorts, before the days of reliable diagnosis, confirm an illness-related decline. (13)
Again, the question arises, are these deficits generalized or specific? Different studies provide different answers. All show deficits across the board including visual perception, recognition memory, motor skills, language comprehension and expression, episodic memory, and planning. However, some authors point out that some test scores seem to be disproportionately impaired and other relatively spared; this gives rise to the claim for specific neuropsychological deficits in schizophrenia. Visual perception, recognition, naming, and procedural or motor learning are relatively spared while executive functions (including tasks of set shifting, ignoring irrelevance, doing more than one thing simultaneously, forward planning) and learning and semantic memory tend to be the functions more obviously affected.
There have been several attempts to classify those sorts of tasks more or less likely to be affected in schizophrenia, and these are summarized in Table 1. This tends to translate into difficult versus easy, which undermines the theoretical claims made for the distinctions.
|More affected||Less affected|
The Wisconsin Card Sorting Test is widely regarded as a specific test of set shifting and patients with schizophrenia do poorly on it as do those with frontal lobe lesions. However, the test requires the subject to shift set, to follow rules, and to keep a record of previous responses so that actions are not repeated. It is therefore not surprising that people with global deficits in IQ find the test difficult. (14) Heinrichs and Zakzanis (15) systematically reviewed studies comparing schizophrenic patients and controls on a variety of neuropsychological tests (including the Wisconsin Test). The result of a meta-analysis of some 204 published studies is that there is little evidence for specific deficits and much evidence for large overlap in performance between patients and controls. The greatest difference in performance was found in tests of verbal memory, bilateral motor skill, and performance (non-verbal) IQ although even these showed that around 25 per cent of patients were above the control median.
These studies of cognition have parallels in structural neuroimaging. Quantitative reviews of the studies on schizophrenia with CT (16) and magnetic resonance imaging show much evidence for generalized loss of tissue (large ventricles and smaller cortical thickness), (17) but more localized abnormalities are less consistently observed, with the possible exception of medial temporal lobe structures. (18)
Symptom- and syndrome-related deficits
Despite problems in showing that particular functions are specifically impaired in the disorder, many models have been put forward which have heuristic value. Instead of attempting to explain the whole of schizophrenia with all its diverse manifestation in terms of a single neuropsychological dysfunction, many such models take as their starting point isolated syndromes or symptoms and seek to account for these in cognitive terms. (19)
This is a loosely defined set of cognitive abilities which are thought to involve the integration of subsidiary abilities—such as memory and perception. The idea is that as well as requiring the basic ability to retain information, a complex organism in the social world has to be able to organize information, to know when to rely on memory, and to update its store of information. These are executive skills. Other executive functions include selecting appropriate responses and inhibiting inappropriate ones, generating plans, and solving problems. The maintenance of information on line and its manipulation to aid problem solving comes under the heading of ‘working memory' seen by most psychologists as an executive function. These functions are loosely allied with the ‘frontal lobes' although such brain-behaviour mapping is an oversimplification. With this broad definition it is easy to see how many of the behaviours and features of schizophrenia can be viewed as ‘executive function deficits'. The negative symptoms of lack of spontaneity, poor motivation, and socially inappropriate behaviour, in particular fit into this rubric. Difficulties in selecting relevant stimuli for further processing (selective attention or filtering) have been invoked to explain some positive symptoms like perceptual abnormalities (hallucinations) and, perhaps, ideas of reference. The Stroop paradigm, where the subject is asked to name the colour ink in which an incongruent colour word is printed, is a traditional test of ‘selective attention'.
The understanding of positive symptoms, particularly hallucinations, has benefited from the concepts of reality and source monitoring. These terms describe the everyday challenge of distinguishing between imagined events and memories for real events (reality monitoring) and in the case of actions or speech, between whether, for example, I said something or merely heard someone saying it to me (source monitoring). Building on the assumption that auditory hallucinations are the products of the subject's own mind (and not from aliens from outer space), reality monitoring provides a model whereby normal inner speech is mislabelled as not coming from the self. Why this occurs is another question. One suggestion is that if inner speech or other internal images are particularly vivid, the subject may confuse them for reality, although evidence for this is lacking. Another model advanced by Frith (20) is that our brains continually monitor or check our actions to see whether they tally with our intentions. If an action occurs (my arm reaches an object or I ask a question) but the brain had not registered a prior intention (owing to some failure of feedback), the action may then be regarded as alien. Again this provides a plausible explanation for auditory hallucinations and passivity phenomena. Frith and colleagues suggest that intentions arise from systems connecting the dorsolateral prefrontal cortex, the anterior cingulate, and the supplementary motor area. Others have argued that the hippocampus is a candidate for the ‘comparator' function between intentions and actions. (21)
Fundamental to psychosis is the attribution of the agency for one's own thoughts, actions, impulses, and so on to an ‘other'. However, social psychologists (22) have shown that this is not merely a defect but rather a bias towards certain types of information. For example, a tendency to attribute or blame others for bad things that happen in one's life, while accepting the credit for good things is an example of such a bias to which most people are prone. Another is paying more attention to events or ideas which seem to confirm one's prejudices. In the extreme these misattributions can amount to delusions. As well as misattributing events, some people are prone to misattribute intentions. Work from child development has proposed that normal children are adept ‘mind readers' in the sense of reading the intentions of others. People with autism frequently fail to infer intentions from social behaviour. (23) Frith (24) has proposed that patients with persecutory delusions do infer such intentions in others, but these are prone to error so that the intentions are invariably felt to be dishonourable.
Psychophysiological and conditioning models
Several techniques, such as measurement of skin conductance, startle responses, early cerebral evoked potentials (P50, a positive deflection in scalp-recorded electrical potentials occurring 50 ms after a stimulus), etc., have been applied to the study of schizophrenia. These have generally supported theories which suggest that patients are abnormally ‘aroused' by benign stimuli or, as in attentional theories, are unable to filter or ‘gate' sensory inputs, thereby leading to sensory ‘overload'. Or, within a learning and conditioning framework, experimental results suggest that patients are unable to distinguish between relevant and irrelevant stimuli. These differ from cognitive theories in that they postulate dysfunction early on in the information processing stream, or with ‘unconscious' processes (such as conditioning).
Neuropsychological theories of schizophrenia have arisen by analogy from the effects on cognition and behaviour of neurological lesions. For instance, the disturbances in language production observed in some schizophrenia patients (thought disorder) have been likened to aphasia and hence left hemisphere abnormalities while deficits in affect regulation and perception have been likened to right hemisphere lesions. Integration of information across the hemispheres seemed to be an attractive notion to account for the range of schizophrenic disturbances as well as apparent ‘splits' between thoughts and actions, and thoughts and emotions. (25,26) Such theories suffer from the difficulty described above in establishing disproportionate deficits in the task of interest against other less localizable deficits.
The course taken by Liddle (27) combines a syndromic with a neurological approach. He found that symptoms of schizophrenia aggregated into three broad clusters: psychomotor poverty (affecting speech and movement and blunting of affect), reality distortion (essentially positive symptoms, hallucinations, and delusions), and finally disorganization (including thought disorder and inappropriate affect). Using a battery of tests borrowed from the neurology clinic, he showed that psychomotor poverty was associated with poorer performance on abstract reasoning and long-term memory tests, disorganization or impairments of attention and learning, while the reality distortion symptoms correlated with impaired figure ground perception (traditionally temporoparietal tests). One interpretation of these findings is that the poverty and disorganization syndromes reflect different focal disturbances in frontal function (dorsolateral and ventral, respectively). Functional neuroimaging techniques are able to test these hypotheses in vivo.
Subsequent research has offered support for this general scheme, namely that ‘negative symptoms' tend to correlate with tests of executive (and memory) deficits, (28) while positive symptoms seem less ‘anchored' to neuropsychological impairments as commonly described but rather, productive abnormalities of language (thought disorder), biases in information appraisal (paranoia), and reality monitoring errors (auditory hallucinations).
Finally, the concept of insight into illness or unawareness of illness has become a topic for discussion. Again, arguing from the neurological analogy of anosognosia (the lack of awareness of neurological deficit, most commonly left hemiparesis) a right hemisphere dysfunction has been postulated or alternatively a problem with executive function (part of an ability for self-reflection). The evidence to date again points to generalized neuropsychological impairment being related to lack of insight although there is some evidence for executive deficits as well. (29)
Neuropsychological function has been related to various indices of outcome and predictors of rehabilitation success. In a thorough review of the literature, Green (30) concluded that the most consistent finding was that verbal memory was associated with all types of functional outcome so that deficits in this function could limit the level of outcome. Vigilance was related to social problem solving and skill acquisition, while card sorting predicted the quality of more general functioning in the community.
1. David, A.S. and Cutting, J.C. (1994). The neuropsychology of schizophrenia. Erlbaum, Hove.
2. Bleuler, E. (1911). Dementia praecox or the group of schizophrenias (trans. H. Zinkin). International Universities Press, New York, 1950.
3. Heaton, R.K., Baade, L.E., and Johnson, K.L. (1978). Neuropsychological test results associated with psychiatric disorders in adults. Psychological Bulletin, 22, 141–62.
4. McKenna, P.J. (1994). Schizophrenia and related syndromes. Oxford University Press.
5. Jones, P., Rodgers, B., Murray, R., and Marmot, M. (1994). Child developmental risk factors for adult schizophrenia in the British 1946 birth cohort. Lancet, 334, 1393–402.
6. Crow, T.J., Done, D.J., and Sacker, A. (1995). Childhood precursors of psychosis as clues to its evolutionary origins. European Archives of Psychiatry. Clinical Neuroscience, 245, 61–9.
7. David, A.S., Malmberg, A., Brandt, L., Allebeck, P., and Lewis, G. (1997). IQ and risk for schizophrenia: a population-based cohort study. Psychological Medicine, 27, 1311–23.
8. Aylward, E., Walker, E., and Bettes, B. (1984). Intelligence in schizophrenia: meta-analysis of the research. Schizophrenia Bulletin, 10, 430–59.
9. Nelson, H.E. (1982). The National Adult Reading Test (NART). NFER-Nelson, Windsor.
10. Cannon, T.D., Zorrilla, L.E., Shtasel, D., et al. (1994). Neuropsychological functioning in siblings discordant for schizophrenia and healthy volunteers. Archives of General Psychiatry, 51, 651–61.
11. Saykin, A.J., Shtasel, D.L., Gur, R.E., et al. (1994). Neuropsychological deficits in neuroleptic naive patients with first-episode schizophrenia. Archives of General Psychiatry, 51, 124–31.
12. Censits, D.M., Ragland, D., Gur, R.C., and Gur, R.E. (1997). Neuropsychological evidence supporting a neurodevelopmental model of schizophrenia: a longitudinal study. Schizophrenia Research, 24, 289–98.
13. Rogers, D.G.C. (1996). The cognitive disorder of psychiatric illness: a historical perspective. In Schizophrenia: a neuropsychological perspective (ed. C. Pantelis, H.E. Nelson, and T.R.E. Barnes), pp. 19–29. Wiley, Chichester.
14. Laws, K.R. (1999). A meta-analytic review of Wisconsin Card Sort studies in schizophrenia: a generalised deficit in disguise. Cognitive Neuropsychiatry, 4, 1–35.
15. Heinrichs, R.W. and Zakzanis, K.K. (1998). Neurocognitive deficit in schizophrenia: a quantitative review of the evidence. Neuropsychology, 12, 426–45.
16. Van Horn, J.D. and McManus, I.C. (1992). Ventricular enlargement in schizophrenia. A meta-analysis of studies of the ventricle : brain ratio (VBR). British Journal of Psychiatry, 160, 687–97.
17. Elkis, H., Friedman, L., Wise, A., and Meltzer, H.Y. (1996). Meta-analyses of studies of ventricular enlargement and cortical sulcal prominence in mood disorders. Comparisons with controls or patients with schizophrenia. Archives of General Psychiatry, 53, 1165–7.
18. Nelson, M.D., Saykin, A.J., Flashman, L.A., and Riordan, H.J. (1998). Hippocampal volume reduction in schizophrenia as assessed by magnetic resonance imaging: a meta-analytic study. Archives of General Psychiatry, 55, 433–40.
19. David, A.S. (1993). Cognitive neuropsychiatry? Psychological Medicine, 23, 1–5.
20. Frith, C.D. (1992). The cognitive neuropsychology of schizophrenia. Erlbaum, Hove.
21. Gray, J.A., Feldon, J., Rawlins, J.N.P., Hemsley, D.R., and Smith, A.D. (1991). The neuropsychology of schizophrenia. Behavioural and Brain Sciences, 14, 1–84.
22. Bentall, R.P. (1996). Brains, biases, deficits and disorders. British Journal of Psychiatry, 167, 149–58.
23. Baron-Cohen, S. (1997). Mindblindness: an essay on autism and theory of mind. MIT Press, Cambridge, MA.
24. Frith, C. (1994). Theory of mind in schizophrenia. In The neuropsychology of schizophrenia (ed. A.S. David and J.C. Cutting), pp. 147–61. Erlbaum, Hove.
25. Cutting, J.C. (1985). The psychology of schizophrenia. Churchill Livingstone, Edinburgh.
26. Cutting, J.C. (1990). The right cerebral hemisphere and psychiatric disorders. Oxford University Press.
27. Liddle, P.F. (1987). The symptoms of chronic schizophrenia: a re-examination of the positive–negative dichotomy. British Journal of Psychiatry, 151, 145–51.
28. Tamlyn, D., McKenna, P.J., Mortimer, A.M., Lund, C.E., Hammond, S., and Baddeley, A.D. (1992). Memory impairment in schizophrenia: its extent, affiliations and neuropsychological character. Psychological Medicine, 22, 101–15.
29. Amador, X.F. and David, A.S. (ed.) (1998). Insight and psychosis. Oxford University Press.
30. Green, M.F. (1996). What are the functional consequences of neurocognitive deficits in schizophrenia? American Journal of Psychiatry, 153, 321–30.