Transient and executive function working memory in schizophrenia

https://doi.org/10.1016/j.schres.2006.04.013Get rights and content

Abstract

Transient working memory requires attention and temporary storage of information, whereas executive function working memory requires additional mental manipulation of that information. Working memory impairment is common in schizophrenia patients, but only some studies have found differential impairment in executive function working memory compared to transient working memory. We measured both types of working memory using the Digit Span forward (DF) and backward (DB) tasks in a large sample of schizophrenia patients (n = 267) and normal comparison subjects (n = 82); in the patients, we also examined associations between performance on the Digit Span tasks and Letter–Number Sequencing (LNS), a putative executive function working memory test. Compared to healthy subjects, the schizophrenia patients showed impairment in the medium effect size range on both DF (d =  0.55) and DB (d =  0.68). DB scores predicted LNS performance, whereas DF scores did not. Worse negative symptoms were associated with worse performance on DF, DB and LNS. These results do not reflect differential executive function working memory dysfunction in schizophrenia, but appear to support transient and executive function working memory as separable constructs.

Introduction

Working memory impairment is common in schizophrenia and has been proposed as a possible “core deficit” that contributes to multiple features of the disorder (Goldman-Rakic, 1994). Working memory is commonly defined as the capacity for “temporary storage and manipulation of the information necessary for such complex cognitive tasks as language comprehension, learning and reasoning” (Baddeley, 1992, p. 556, italics added). A recent re-conceptualization parses the storage and manipulation aspects of working memory by characterizing tasks as measuring either “transient online storage and retrieval” or “executive function working memory”, respectively (Perry et al., 2001). The term “working memory” has been used inconsistently in the schizophrenia literature and elsewhere, but these more precise definitions should be useful in clarifying the findings of past and future investigations.

The “transient” type of working memory is measured by “hold and repeat” tests in which the individual creates and maintains an internal representation of external stimuli. Examples of such tasks include repeating a series of digits, or holding a spatial location in mind and indicating the location following a brief delay. This latter type of task has been used by Goldman-Rakic (1994) and Park and Holzman (1992) to demonstrate transient working memory impairment in schizophrenia.

“Executive function” working memory is measured by tests that require transient working memory as well as additional processing or manipulation of the information held in mind. Common tests of executive function working memory include repeating digits in reverse order, pointing to a sequence of items in reverse order, and re-ordering a jumbled sequence of numbers and letters (Digit Span backward, Spatial Span backward and Letter–Number Sequencing from the Wechsler Adult Intelligence and Memory Scales (Wechsler, 1997a, Wechsler, 1997b)).

Digit span tests are among the oldest in the history of psychology, appearing in some of the first widely used intelligence tests (e.g., the Binet-Simon Scale in 1905 and Wechsler's Bellevue Intelligence Examination in 1939; see Ramsay and Reynolds, 1995 for review). The Digit Span test purportedly measures both types of working memory. For the Digit Span forward (DF) items, the examinee is required to simply repeat, in order, a series of numbers read aloud by the examiner at the rate of one digit per second. For the Digit Span backward (DB) items, the examinee must repeat the series of numbers in reverse order. The DF and DB tasks become more difficult as the number of digits to be repeated grows longer. It has long been recognized that DF and DB are tasks that measure somewhat different abilities. DF requires reception, attention, temporary storage and repetition of the stimuli, whereas DB requires these abilities as well as manipulation (reordering) of the stimuli. Thus, DF appears to be a prototypical transient working memory task, whereas DB seems to measure executive function working memory, because it demands manipulation of the items held in mind (Perry et al., 2001). (It should be recognized that, although DF does not involve re-ordering of items, there may be an “executive” component of successful performance on long digit span lengths, during which respondents often use a chunking strategy.) The difference in span between DF and DB (DIF) may also be a good measure of executive function working memory, because it controls for “hold and repeat” or transient working memory function assessed by DF. Presumably, greater DIF scores indicate more difficulty with the executive function component of working memory compared with transient working memory.

In a comprehensive review of the literature on digit span testing, Ramsay and Reynolds (1995) concluded that more investigations (n = 13) supported separate scaling for DF and DB than supported combined scaling (n = 4). Compelling evidence suggests that people often use visuospatial strategies to perform DB, but they rarely do so while performing DF (Ramsay and Reynolds, 1995). Brain imaging studies suggest that both tasks recruit left hemisphere brain regions, whereas DB appears to involve an additional right hemisphere substrate in the dorsolateral prefrontal cortex (Hoshi et al., 2000). Factor analytic studies, too, suggest that DF and DB often load on different factors (Ramsay and Reynolds, 1995).

Some investigations have found that patients with schizophrenia perform worse than healthy comparison subjects on both DF and DB (Conklin et al., 2000, Perry et al., 2001, Stefansson and Jonsdottir, 1996, Stratta et al., 1997); however, others have found that schizophrenia patients exhibit either no impairment on either test (Park and Holzman, 1992) or selective impairment in DB, but not DF (Stone et al., 1998). Many of these studies relied on small samples of 12–52 patients. A meta-analysis (Aleman, 1999) including 18 studies of DF and 7 studies of DB found no significant difference between the effect sizes comparing schizophrenia patients' performance to that of normal subjects (d = 0.71 for DF and d = 0.82 for DB). Evidence regarding the relationship between clinical symptoms, DF and DB performance have been mixed, with one study finding an association between severity of positive symptoms and poor performance on DF (Berman et al., 1997), and another study finding an association between severity of negative symptoms and poor performance on DB (Moritz et al., 2001).

In order to expand on the previous findings of smaller studies, we examined the relationships among DF, DB and DIF scores in a large sample of patients with schizophrenia and normal comparison subjects (NCs). We hypothesized that schizophrenia patients would perform worse than NCs on DF, DB and DIF, but would demonstrate a differential impairment on the measures more likely to tap executive function working memory, i.e., DB and DIF. We also sought to examine the associations of the Digit Span tests with a more challenging measure of executive function working memory (WAIS-III Letter–Number Sequencing), which requires reordering of both numbers and letters; our hypothesis was that performances on LNS would be more strongly associated with those on DB and DIF than DF. Finally, we examined correlations between psychiatric symptom severity and the various measures of working memory performance.

Section snippets

Participants and procedures

The present report is based on a secondary analysis of an existing dataset of schizophrenia patients and NCs who were enrolled in the University of California, San Diego (UCSD) Advanced Center for Interventions and Services Research. Data from 267 outpatients with either schizophrenia (n = 212) or schizoaffective disorder (n = 55) and 82 NCs who had completed the Digit Span tests as part of several parent studies, were used in the current study. Of the schizophrenia patients, diagnostic subtypes

Results

Our first hypothesis, that schizophrenia patients would show differential impairment on DB and DIF, as compared to DF, was not supported. The patient group performed significantly worse than did NCs on both DF and DB. However, the groups did not differ significantly on DIF (see Table 1). Effect sizes (Cohen's d) for group differences in the T-scores were d =  0.55 for DF, d =  0.68 for DB and d =  0.10 for DIF. The proportion of patients impaired (i.e., T-score < 40) on DF was 29.8%, whereas the

Discussion

Many previous studies (Conklin et al., 2000, Perry et al., 2001, Stefansson and Jonsdottir, 1996, Stratta et al., 1997) and a meta-analysis (Aleman, 1999) have found that participants with schizophrenia are about equally impaired on both DF and DB. Our results, using a large sample and demographically corrected T-scores, are consistent with these findings. Schizophrenia patients' average performance was slightly over half of a standard deviation lower than normative expectation for both DF and

Acknowledgements

This study was supported, in part, by a grant from the National Institute of Mental Health (5P30MH066248), the National Alliance for Research on Schizophrenia and Depression, and the Department of Veterans Affairs.

References (23)

  • H.M. Conklin et al.

    Verbal working memory impairment in schizophrenia patients and their first-degree relatives: evidence from the digit span task

    Am. J. Psychiatry

    (2000)
  • Cited by (46)

    • Minimal effects of prolonged smoking abstinence or resumption on cognitive performance challenge the “self-medication” hypothesis in schizophrenia

      2018, Schizophrenia Research
      Citation Excerpt :

      Executive control and cognitive inhibition were evaluated with the Stroop Color-Word Test (SCWT) (Trenerry et al., 1989). Short-term working memory was measured using the Letter Number Sequencing tasks (LNS) and Digit Span (DS) (Kaplan, 1995; Twamley et al., 2006). Processing speed was measured using the WAIS-III Digit Symbol Task (DST) and Symbol Search (SS) tests (Wechsler, 1997).

    • Assessing working memory in children with ADHD: Minor administration and scoring changes may improve digit span backward's construct validity

      2018, Research in Developmental Disabilities
      Citation Excerpt :

      Moreover, the clinical literature has been criticized for using executive functioning tests that are not cognitively informed according to contemporary theoretical and modeling advancements (Snyder et al., 2015). The evidence base for these critiques is substantial, and includes data from large samples of healthy and clinically referred children, adolescents, and adults (Bowden, Petrauskas, Bardenhagen, Meade, & Simpson, 2013; Colom, Flores-Mendoza, Quiroga, & Privado, 2005; Egeland, 2015; Perry, Heaton, Potterat, & Braff, 2001; Tarle et al., 2017; Twamley, Palmer, Jeste, Taylor, & Heaton, 2006). The current study’s goal was to begin to address these criticisms by piloting relatively minor and straightforward test modifications involving scoring, task demands, and administration parameters to improve our ability to capture construct-relevant individual differences in working memory.

    • Impulsiveness and Inhibitory Mechanisms

      2016, Neuroimaging Personality, Social Cognition, and Character
    • A comparative study of the working memory multicomponent model in psychosis and healthy controls

      2015, Comprehensive Psychiatry
      Citation Excerpt :

      There are methodological issues that make difficult to explore working memory functioning in psychotic disorders. For instance, the tasks employed to assess working memory and executive functions are usually complex and need more than one executive function to be solved [18,19]. An interesting finding is the independence of the four executive functions studied, assessed with partial correlations.

    View all citing articles on Scopus
    View full text