Elsevier

Psychiatry Research

Volume 110, Issue 3, 31 July 2002, Pages 259-271
Psychiatry Research

Spatial working memory span, delayed response and executive function in schizophrenia

https://doi.org/10.1016/S0165-1781(02)00105-1Get rights and content

Abstract

This study investigated the spatial working memory span (SWMS) as well as the immediate memory span of schizophrenic patients, and examined the contribution of each span to the patients’ executive function deficit in the visuospatial domain. SWMS measured the visuospatial working memory capacity that simultaneously processes and stores visuospatial information. Immediate memory span was measured with the spatial span (SS), a variant of Corsi's block-tapping test. A total of 16 patients diagnosed with schizophrenia and 16 normal control subjects participated in the study. SWMS, as well as the forward and backward SS, was significantly reduced in schizophrenia. The SWMS deficit observed in this study and previous findings of deficit in verbal working memory spans suggest that impairment in working memory capacity in schizophrenia is general, and not limited to the verbal domain. Executive function as assessed with the self-ordered pointing task (SOPT) was also impaired in the patients, which is consistent with clinical observations of self-monitoring impairment in schizophrenia. SWMS was able to account for the performance on the SOPT, but its contribution in the patients’ impairment did not reach statistical significance. Backward span deficit explained this executive function impairment. SWMS was effective in explaining schizophrenic patients’ impaired performance on the spatial delayed response, a prefrontal function task. Implications of the relations observed between the spans and the prefrontal function tasks are discussed.

Introduction

Working memory (WM) is an active memory system that simultaneously stores and processes information (Baddeley, 1986, Baddeley, 1995). The WM deficit hypothesis in schizophrenia has gained support from neuropsychological studies utilizing different WM paradigms (Park and Holzman, 1992, Fleming et al., 1995, Fleming et al., 1997, Gold et al., 1997, Salamé et al., 1998, Stone et al., 1998). The delayed-response (DR) paradigm derived from classical animal studies (Goldman and Rosvold, 1970) was first used by Park and Holzman (1992) to demonstrate the deficits in spatial WM and its neural substrate, prefrontal cortex, in schizophrenia, which have been repeatedly shown in other variations of the task (Keefe et al., 1995, Fleming et al., 1997, Goldman-Rakic and Selemon, 1997).

Deficits in executive function tests, such as the Wisconsin Card Sorting Test (WCST; Heaton, 1981), have also been interpreted as evidence of WM impairment in schizophrenia and its anatomical correlate, prefrontal dysfunction (Weinberger et al., 1986, Seidman et al., 1995). Since executive function tests require complex cognitive processes other than simple short-term storage, they have been considered useful in measuring WM. However, the multicomponent nature of the executive function tasks poses interpretive problems for impaired performance in schizophrenia. The WCST, for instance, involves cognitive processes other than WM that have been implicated in the pathological framework of schizophrenia, such as concept formation, selective attention, cognitive flexibility and the ability to use feedback to alter behavior. According to Gold et al. (1997), one solution to “…this interpretive complexity is to delineate the contribution of simpler, potentially more localizable, cognitive processes to…” executive function performance. Utilizing an auditory WM task, the Letter–Number span, Gold and his colleagues found a significant contribution of WM deficit in schizophrenics’ WCST performance. Their solution would prove useful in integrating results of basic cognitive impairment with data on higher cognitive dysfunction or symptomatology in schizophrenia.

More recent studies recognize the multi-dimensional nature of WM and utilize multiple WM measures to compare their effectiveness in explaining schizophrenic patients’ impairment (Stone et al., 1998, Perry et al., 2001). Stone et al. (1998) investigated the contribution of WM spans in executive function performance. Utilizing the listening and computation spans, Stone and her colleagues found that smaller listening and computation spans in schizophrenic patients accounted for their impaired performance in what they called “strategic long-term memory.” In contrast, immediate span, assessed by the forward Digit Span, did not account for the group difference in this memory. Strategic memory was measured with the free-recall, the temporal ordering and the self-ordered pointing procedures, which were assumed to involve executive function. The results demonstrated a significant contribution of the verbal WM spans to complex memory tasks that require executive function, and provided indirect evidence for prefrontal dysfunction in schizophrenia. WM spans, such as the reading (Daneman and Carpenter, 1980) and listening spans (Salthouse and Babcock, 1991), measure WM capacity by simultaneously drawing on the WM processing and storage resources in the verbal domain. These spans were developed by the information-processing approach to assess each individual's WM capacity (Just and Carpenter, 1992, Shah and Miyake, 1996), which differs from one individual to another.

Another study investigated the reading span and its contribution to impaired language comprehension in schizophrenia (Condray et al., 1996). The study found a significant association between the span and language comprehension in normal as well as schizophrenic participants, and concluded that language-comprehension deficits observed in schizophrenia could result from a general cognitive dysfunction that is associated with WM capacity. Verbal WM spans of schizophrenic patients, therefore, seem to contribute significantly to their deficits in executive function or complex cognition, at least in the verbal domain. No study, to the best of our knowledge, has yet examined schizophrenic patients’ WM span and its relation to executive function deficit in the visuospatial domain. In order to confirm that the WM span deficit is a general finding that is not limited to the verbal domain, this study examined the spatial working memory span (Shah and Miyake, 1996) and its contribution to executive function.

Schizophrenic patients lack self-awareness or self-monitoring abilities, which are important aspects of executive function (see Tranel et al., 1994, for a review). This clinical feature significantly contributes to the patients’ impairment in job and other role functioning that prevents them from leading active lives (Kraepelin, 1919, Levin, 1984, Frith and Dolan, 1996). Intact WM, according to Frith (1992), is crucial for self-monitoring, since it provides continuity in a person's past, present and future. Few studies, however, have investigated the role of WM in self-awareness or self-monitoring ability in schizophrenia.

The main purpose of this study was to investigate the spatial working memory span (SWMS) and its relation to self-monitoring ability in schizophrenia. Self-monitoring ability was examined with the abstract design condition of the self-ordered pointing task (Petrides and Milner, 1982). It is one of the few executive function tests that measure self-monitoring ability besides goal-directed processing. Although most executive function tests require some self-monitoring ability, the successful performance of other tests relies more heavily on utilizing outsource feedback or information. The crucial element of the self-ordered pointing task is keeping track of the testee's own behavior. We also examined the contribution of SWMS in spatial delayed response impairment, which has been consistently observed in schizophrenia. This study introduced the use of a multilevel model, a variant of regression analysis that can effectively deal with repeated measure or nested design data to examine the contribution of spans in the spatial delayed response impairment. Previous studies used ANCOVA to examine the contribution of WM spans in other cognitive tasks (Gold et al., 1997, Stone et al., 1998). Use of ANCOVA in this context, however, is problematic, since the diagnostic group is interdependent with the covariates, i.e. the WM spans (Miller and Chapman, 2001). Regression analysis is more effective in addressing the contribution of each span in prefrontal function tasks.

Four major predictions were possible concerning the WM measures. First, schizophrenic patients will show reduced SWMS due to their impaired WM. Second, as a result of their prefrontal dysfunction, they will demonstrate impairments in the executive function and the delayed response tasks. Third, SWMS deficit in schizophrenia may account for impairment in executive function, specifically self-monitoring ability, since temporal organization of behavior is crucial for monitoring one's own actions. Fourth, SWMS and delayed response performances would show strong association, both being WM measures that involve both storage and manipulation.

In addition, we measured immediate memory using forward and backward spatial spans (SS in WAIS-RNI; Kaplan et al., 1991). These spans were developed as non-verbal visual analogues to the digit spans to assess the capacity of immediate memory or attention span. The digit spans and the SS, however, have been both utilized as measures of WM. Backward spans (Carlesimo et al., 1994, Cherry et al., 1996, Brébion et al., 1998, Stone et al., 1998) have been regarded as measures of the central executive component of WM, whereas forward spans have been adopted for the slave systems, such as the articulatory loop and the visuospatial scratch pad (Smyth and Scholey, 1994, Cherry et al., 1996, Pantelis et al., 1997). Their relation to WM span and executive functions were investigated.

Digit span has consistently been found intact in schizophrenic patients whose other WM measures are impaired (Park and Holzman, 1992, Morice and Delahunty, 1996, Salamé et al., 1998), while SS results have been less consistent. Although Kolb and Whishaw (1983) found intact forward SS in schizophrenia, others found it impaired (Rizzo et al., 1996, Fleming et al., 1997, Pantelis et al., 1997, Salamé et al., 1998). Fleming et al. (1997) measured the scaled scores of SS, and found both forward and backward spans impaired in schizophrenia. The present study examined the relationship between SS and the SWMS, as well as their role in the delayed response and self-ordered pointing tasks, both to confirm the spans as measures of WM and to clarify relations among the spans. Reduced backward spatial span was predicted in schizophrenia; however, no specific prediction was made concerning the forward span due to the inconsistencies of previous results.

Section snippets

Participants

A total of 16 patients diagnosed with schizophrenia and 16 normal control subjects took part in this study; 11 inpatients and five outpatients in the psychiatric department of a public hospital in Seoul were tested. The patients met the following inclusionary criteria: (a) DSM-IV (American Psychiatric Association, 1994) criteria for schizophrenia, based on a review of medical records and clinical examinations by two experienced psychiatrists; (b) age between 20 and 50 years; (c) no mental

Spatial working memory span and spatial span

SWMS was significantly reduced in the schizophrenia group compared to the control group (t29=5.08, P=0.0001). The patients also had shorter Spatial Span (SS) than the control subjects, both forward (t29=3.96, P=0.0001) and backward (t29=2.05, P=0.05). The means and standard deviations for the two groups on these measures are presented in Table 2. Both SSs correlated significantly with the SWMS. Pearson correlation coefficients were 0.45 (P=0.012) for the forward span and 0.47 (P=0.008) for the

Discussion

This study investigated the working memory span as well as the immediate memory span of schizophrenic patients, and examined the contribution of each span in the patients’ executive function deficit in the visuospatial domain. Spatial working memory span (SWMS) (P<0.001), as well as forward (P<0.001) and backward (P=0.05) spatial spans (SS), were significantly reduced in schizophrenia. This and previous findings of verbal WM span deficits (Condray et al., 1996, Stone et al., 1998) together

Acknowledgements

The authors are grateful to Kyung-Hyung Cho, MD, Misun Park, MA, Sae-Hoon Chung, MD, and Hoyoung Kim, BA for their invaluable collaboration in this study. We also thank Cheongtag Kim, PhD, for his helpful statistical consultation.

References (55)

  • S Raudenbush et al.

    Hierarchical Linear and Non-linear Modeling (HLM) Version 5 for Windows

    (2001)
  • T.A Salthouse et al.

    Decomposing adult age differences in working memory

    Developmental Psychology

    (1991)
  • N.C Andreasen et al.

    Positive and negative symptoms

  • Diagnostic and Statistical Manual of Mental Disorders

    (1994)
  • A.D Baddeley

    Working Memory

    (1986)
  • A.D Baddeley

    Working memory or working attention?

  • J.A Banken

    Clinical utility of considering digits forward and backward as separate components of the Wechsler Adult Intelligence Scale-Revised

    Journal of Clinical Psychology

    (1985)
  • D.L Braff

    Impaired speed of information processing in non-medicated schizotypal patients

    Schizophrenia Bulletin

    (1981)
  • A.S Bryk et al.

    Application of hierarchical linear models to assessing change

    Psychological Bulletin

    (1987)
  • G.A Carlesimo et al.

    Verbal and spatial memory spans in Alzheimer's and multi-infarct dementia

    Acta Neurologica Scandinavica

    (1994)
  • B.J Cherry et al.

    Memory span procedures in Alzheimer's disease

    Neuropsychology

    (1996)
  • E Fey

    The performance of young schizophrenics and young normals on the Wisconsin Card Sorting Test

    Journal of Consulting Psychology

    (1951)
  • C.D Frith

    The Cognitive Neuropsychology of Schizophrenia

    (1992)
  • J.M Fuster

    The Prefrontal Cortex

    (1997)
  • J.M Fuster et al.

    Neuron activity related to short-term memory

    Science

    (1971)
  • J.M Gold et al.

    Auditory working memory and Wisconsin Card Sorting Test performance in schizophrenia

    Archives of General Psychiatry

    (1997)
  • P.S Goldman-Rakic et al.

    Functional and anatomical aspects of prefrontal pathology in schizophrenia

    Schizophrenia Bulletin

    (1997)
  • Cited by (50)

    • Acoustic startle and prepulse inhibition deficits in adult monkeys with neonatal lesions of the hippocampus, amygdala and orbital frontal cortex

      2023, Behavioural Brain Research
      Citation Excerpt :

      Furthermore, the specificity of the sensory-motor deficits following the three types of neonatal lesions was also interesting and may provide knowledge on the specific neural substrate of the sensory-motor deficits reported in a diversity of neuropsychiatric disorders, such as schizophrenia, autism spectrum disorders, Tourette’s syndrome, obsessive compulsive disorder, Huntington’s Disease, and post-traumatic stress disorder [see for review 28]. For example, as in rodents [6,7,9,33], the sensory-motor gating deficits observed after neonatal hippocampal lesions are not restricted to PPI changes but also involved impaired working memory and prefrontal malfunction [9,22–24,65,66,90, for a review see 91], which recapitulate core neuropathological changes and symptoms reported in schizophrenia [92–100]. Lastly, these data further support the necessity of utilizing nonhuman primates in the research of neural networks that are the subject of investigation for their role in mental illness, confirming prior research showing that some networks are conserved [21] while others vary between rodents and primates [20].

    • A comparative study of the working memory multicomponent model in psychosis and healthy controls

      2015, Comprehensive Psychiatry
      Citation Excerpt :

      No specific impairment was found in the patients in the remaining working memory components. In the spatial span task, representative of the visuospatial sketchpad component, other studies have reported deficits in patients with schizophrenia in similar tasks [17,53–55]. We also found a decline in performance in the backwards Spatial span task with respect to the forward task in both groups, which may be due to the involvement of the central executive functions in the performance of the backwards tasks.

    • Animal models of working memory: A review of tasks that might be used in screening drug treatments for the memory impairments found in schizophrenia

      2013, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      It is possible, however, that the latter result could be related to semantic tagging of odours encountered. Individuals with schizophrenia show reduced spans on a task which requires memory for the location of stimuli on a computer screen, and on a Corsi-type block tapping task (Chey et al., 2002). Diminished spans are also seen patients with schizophrenia on the visual memory span subtest of the Wechsler Adult Intelligence scale, a task in which one must remember a sequence of squares that the tester points to on a card (Pirkola et al., 2005).

    View all citing articles on Scopus
    View full text