Visual–spatial working memory performance and temporal gray matter volume predict schizotypal personality disorder group membership

https://doi.org/10.1016/j.schres.2013.12.006Get rights and content

Abstract

Background

Prior work shows individuals with schizotypal personality disorder (SPD) evince temporal lobe volume abnormalities similar to schizophrenia but sparing of prefrontal cortex, which may mitigate psychosis and the severe neurocognitive impairments observed in schizophrenia. This study examined the extent to which frontal–temporal gray matter volume and neurocognitive performance predict: (1) SPD group membership in a demographically-balanced sample of 51 patients and 37 healthy controls; and (2) symptom severity in SPD.

Methods

Dimensional gray-matter volume (left frontal–temporal regions (Brodmann area (BA) 10, 21, 22)) and neurocognitive performance on key memory tasks (California Verbal Learning Test (CVLT), Dot Test, Paced Auditory Serial Addition Test (PASAT)), all salient to schizophrenia-spectrum disorders were examined in a multi-variable model.

Results

Middle temporal gyrus (BA21) volume and spatial–working memory (Dot Test) performance were significant predictors of SPD group membership likelihood, with poorer working-memory performance indicating increased probability of SPD membership. Combining across regional volumes or cognitive measures resulted in fair-to-good discrimination of group membership, but including neurocognitive and non-collinear regional volume measures together resulted in a receiver-operating-characteristic (ROC) curve with improved diagnostic discrimination. Larger BA10 volume in dorsolateral prefrontal cortex (DLPFC) significantly predicted less symptom severity in SPD.

Conclusions

These findings suggest that temporal lobe volume and spatial-working memory performance are promising biological/phenotype markers for likelihood of SPD classification, while greater DLPFC volume may serve as a protective factor.

Introduction

Schizotypal personality disorder (SPD) is considered to be the prototypic schizophrenia-spectrum disorder which shares common genetic, biological, and neurocognitive abnormalities with schizophrenia, but without the psychosis and severity of cognitive impairment observed in the latter (Kendler and Diehl, 1993, Cadenhead et al., 1999, Siever and Davis, 2004). The SPD diagnosis currently follows a categorical approach prescribed by the DSM-IV (American-Psychiatric-Association, 2000), in which an individual must demonstrate multi-contextual evidence of at least five out of nine criteria broadly encompassing domains of unusual behavior, speech, cognition, and limited social contact and interest. Limitations of the current diagnostic classification systems include: high degree of comorbidity, potentially artificial disorder boundaries, and a lack of diagnostic clarity, e.g., (Widiger and Samuel, 2005, Kupfer and Regier, 2011), which undermine its clinical utility. Additionally, current categorical classification systems do not incorporate knowledge gleaned from brain science (Insel and Cuthbert, 2009).

Krueger and Eaton (2010) recommend developing a comprehensive empirically derived system based on quantitative models of psychopathology. They focus on personality traits but argue that a valid diagnostic system to drive appropriate intervention strategies could be created using a combination of neurobehavioral indices and environmental elicitors. One aim of transdisciplinary diagnostic approaches is to address some of these limitations and enhance diagnostic classification, thereby aiding our understanding of etiological mechanisms and improving clinical utility.

Consistent with this approach, attempts to elucidate diagnostic clarity, understand disease pathology and course include, using empirical data to explore whether these variables could accurately classify diagnostic group membership and functional outcome. This strategy has proved fruitful by providing evidence indicating that neurocognitive performance (Pena et al., 2011), prodromal symptoms (McGorry et al., 2000) and activity in key brain regions as measured by fMRI scans (Whalley et al., 2006) correctly classify schizophrenia patients. Additionally, studies have shown that baseline neurocognitive performance predicts functioning five years after hospitalization for first psychotic episode (Gonzalez-Ortega et al., 2013) and psychotic symptoms predict future hospitalization (Werbeloff et al., 2012). Although there is consistent evidence that schizophrenia-spectrum disorder patients demonstrate neurocognitive performance deficits and a reduction of gray matter volume in key brain regions, no study has examined whether these measures can accurately classify patients with SPD. The extant literature primarily focuses on predicting schizophrenia group membership and differentiation of schizophrenia from other psychotic disorders. For example, Pena et al. (2011) reported that performance on the Wisconsin Card Sorting Test (WCST) correctly distinguished patients diagnosed with schizophrenia from those with bipolar disorder. Using this prior schizophrenia work as a novel model for SPD, the aim of this study was to examine the extent to which both gray matter volume in key frontal and temporal lobe regions, and neurocognitive performance effectively predict SPD group membership, and a dimensional measure of symptom severity in SPD.

There is consistent evidence supporting reduced volume of temporal lobe regions in schizophrenia (reviewed by Shenton et al., 2001, Shepherd et al., 2012). While studies in SPD report temporal lobe reductions, they are less marked than those observed in schizophrenia (Takahashi et al., 2006, Hazlett et al., 2008). Specifically, SPD-related reduced gray matter volume has been found in the superior (Brodmann area (BA) 22) (Buchsbaum et al., 1997, Dickey et al., 1999), middle (BA21) (Hazlett et al., 2008), and inferior temporal gyrus (BA20) (Downhill et al., 2001). Dickey et al. (2002) found reduced gray volume matter in the left Heschl's gyrus of SPD patients; however, a subsequent study of female SPD patients failed to reveal differences in superior temporal gyrus (STG) volume (Dickey et al., 2003). We have proposed that reduced superior (BA22) and middle (BA21) temporal lobe gyrus volume may be an important endophenotype for schizophrenia-spectrum disorders (Siever and Davis, 2004, Goldstein et al., 2009).

While temporal lobe volume has been shown to be smaller in SPD, prefrontal cortex volume findings are more mixed. Some studies reported decreased prefrontal volume in SPD (Raine et al., 2002, Matsui et al., 2008), whereas our group (Hazlett et al., 2008) and others (Suzuki et al., 2005, Kuhn et al., 2012) have found increased dorsolateral prefrontal cortex (DLPFC) volume. SPD patients also show higher glucose metabolism in BA10 of the DLPFC compared with healthy controls (HCs) and schizophrenia patients during the California Verbal Learning Test (CVLT) adapted for PET (Buchsbaum et al., 2002a), and it has been proposed that this compensatory pattern may serve to protect these individuals from frank psychosis. In our earlier study, we concluded that increased gray matter volume in BA10 of the DLPFC and relative sparing of volume loss in the temporal cortex may be a protective factor in SPD, which reduces vulnerability to psychotic symptoms (Hazlett et al., 2008). Consistent with this concept, we recently reported that compared with HCs, SPD patients have significantly lower fractional anisotropy—a measure of white matter integrity—in left BA21 of the temporal lobe while showing sparing in prefrontal regions (Hazlett et al., 2011). Taken together, MRI findings support the model proposed by Siever and Davis (2004) that similar to schizophrenia, superior (BA22) and middle (BA21) temporal gyrus abnormalities are core to both SPD and schizophrenia, while frontal reserves (e.g., BA10 of DLPFC) protect the former from decompensating into florid psychosis.

Neurocognitive studies support the perspective that working memory impairment is a core, stable feature of schizophrenia, e.g., (Goldman-Rakic, 1991, Park and Holzman, 1992, Silver et al., 2003). Working memory involves the ability to manipulate and maintain information over a short period of time and has been associated with DLPFC function (Callicott et al., 2003, Reichenberg and Harvey, 2007) and frontal–temporal circuitry (Kang et al., 2011). Working-memory deficits in schizophrenia include verbal and non-verbal domains (Park and Holzman, 1992, Kareken et al., 1995, Park et al., 1996, Manoach et al., 1999, Reichenberg and Harvey, 2007). Similarly, individuals with SPD consistently exhibit problems with working memory (Voglmaier et al., 2000, Mitropoulou et al., 2005), including visual–spatial (Voglmaier et al., 1997, Farmer et al., 2000, Roitman et al., 2000) and verbal tasks (McClure et al., 2007), although deficits are typically observed with longer time delays, suggesting these deficits are more subtle than those observed in schizophrenia. Roitman et al. (2000) reported that SPD patients demonstrate visual–spatial impairments on the Dot Test at the delayed but not the more immediate conditions. Similarly, SPD-related deficits have been reported during the CVLT, e.g., (Bergman et al., 1998), particularly after a 20-minute delay (McClure et al., 2007). Mitropoulou et al. (2002) showed that SPD patients evince marked abnormalities on the PASAT, a test of auditory working memory, compared with HCs and individuals with non-schizophrenia spectrum personality disorders. Taken together, neurocognitive deficits in SPD are more subtle than those observed in schizophrenia but compared with HCs, SPD patients have shown deficits in auditory (PASAT) working memory, as well as visual working memory (Dot Test) and verbal learning (CVLT) during the more challenging delay conditions.

Research has shown a direct relationship between structural brain abnormalities and cognitive functioning, as measured by neurocognitive tests (Goldberg et al., 1994, Antonova et al., 2004). In SPD, reduced frontal lobe size and dysfunction as measured by perseveration errors on the WCST, have been reported (Raine et al., 2002). Matsui and colleagues (Matsui et al., 2008) found that smaller left inferior frontal white matter volume was associated with better semantic clustering in SPD compared with healthy individuals and suggested the operation of a compensatory mechanism, i.e. brain areas other than the frontal lobe playing an important role in memory performance. Our lab (Goldstein et al., 2011) reported that in HCs, but not SPD patients, better spatial working memory performance was associated with larger ventrolateral prefrontal cortex gray matter volume. To date, no study of SPD has examined gray matter brain volume of key regions (frontal and temporal) and working-memory (both visual–spatial and verbal) function together in the same sample, which is an aim of this study.

Neuroimaging and neurocognitive findings in SPD suggest this disorder is characterized by reduced temporal lobe volume but frontal lobe sparing, as well as cognitive impairments, particularly on tasks involving working memory. We specifically selected cortical regions (BAs 22, 21, and 10) and variables from neurocognitive tests (PASAT, Dot Test and CVLT long-delay condition) that consistently reveal SPD differences from HC participants. This study tested the following hypotheses: (1) smaller left temporal, but not frontal lobe volume is associated with increased odds for SPD group membership; (2) poorer working memory and specifically, visual–spatial memory performance is associated with increased odds for SPD classification; and (3) frontal lobe sparing (i.e. larger frontal lobe volume) predicts less severity of SPD symptoms.

Section snippets

Participants

We studied two demographically-balanced groups of participants: 51 individuals with SPD and 37 HCs. Demographic and clinical details are presented in Table 1. A small subgroup of these participants was previously included in independent studies examining regional brain volume (Hazlett et al., 2008) and neurocognitive functioning (McClure et al., 2008) (n = 10 subjects, five in each study). Participants provided written informed consent in accordance with the Mt. Sinai and JJPVAMC Institutional

Results

The MANOVA indicated a main effect of Diagnostic group (F[1,86] = 13.0, p < 0.001) and all interactions with Diagnostic group were significant (Diagnostic group × Domain interaction, F[1,86] = 12.3, p < 0.001; Diagnostic group × Domain × Region/Test interaction, F[2,172] = 11.9, p < 0.0001). Post-hoc comparisons indicated the groups differed significantly for relative left BA21 volume and all three of the neurocognitive tests examined, while the groups did not differ for relative BA10 (DLPFC) or BA22 gray matter

Discussion

The main findings of this study are: (1) Frontal–temporal lobe gray matter volume and working-memory measures previously shown to be abnormal in schizophrenia-spectrum disorder illnesses provided good diagnostic classification of SPD; and (2) Consistent with our hypothesis that frontal lobe sparing in SPD may serve as a protective factor, larger DLPFC volume (BA10) is associated with less symptom severity in SPD patients.

Consistent with prior work indicating that visual–spatial memory

Limitations

A limitation of the study is that we did not include patients with schizophrenia. Subsequent studies should examine additional diagnostic indicators along the schizophrenia spectrum to ascertain, for example, the extent to which indicators of spatial-working memory and frontal–temporal gray matter volume overlap with or differentiate SPD patients from those with schizophrenia and other related disorders. This is of particular interest given that SPD will now be primarily coded as a

Conclusions

In conclusion, this study shows that combining a neurocognitive test that measures spatial-working memory (the Dot Test) together with a gray matter volume measure of the temporal lobe (BA21) provides good detection of the SPD phenotype. Study findings are consistent with a “frontal lobe sparing” theory in which individuals with SPD lack the significant frontal lobe impairment and, therefore, the overt psychotic symptoms seen in schizophrenia. The use of both biological and neurocognitive

Role of funding source

The funding sources had no role in the study design, collection, analysis, interpretation of data, writing of the manuscript, or in the decision to submit the paper for publication.

Contributors

Dr. Hazlett was responsible for overseeing all aspects of the study and the write-up of the manuscript. Under Dr. Hazlett's supervision, Ms. Lamade, a graduate student helped draft some sections of the manuscript. Drs. Moshier and Godbold provided statistical expertise and conducted the multivariate regression analyses. All other coauthors were involved in recruitment, diagnostic assessment, and/or testing of the participants, and provided feedback on the paper. All authors have approved the

Conflict of interest

None.

Acknowledgment

Support for this work came from a VA MERIT grant (I01 CX000261) to E.A.H. and the Department of Veterans Affairs VISN3 Mental Illness Research, Education, and Clinical Center (MIRECC) at the James J. Peters VA Medical Center.

References (80)

  • T.E. Goldberg et al.

    Relations between neuropsychological performance and brain morphological and physiological measures in monozygotic twins discordant for schizophrenia

    Psychiatry Res.

    (1994)
  • K.E. Goldstein et al.

    Smaller superior temporal gyrus volume specificity in schizotypal personality disorder

    Schizophr. Res.

    (2009)
  • K.E. Goldstein et al.

    Dorso- and ventro-lateral prefrontal volume and spatial working memory in schizotypal personality disorder

    Behav. Brain Res.

    (2011)
  • I. Gonzalez-Ortega et al.

    Working memory as a predictor of negative symptoms and functional outcome in first episode psychosis

    Psychiatry Res.

    (2013)
  • E.A. Hazlett et al.

    Cortical gray and white matter volume in unmedicated schizotypal and schizophrenia patients

    Schizophr. Res.

    (2008)
  • E.A. Hazlett et al.

    Cingulate and temporal lobe fractional anisotropy in schizotypal personality disorder

    Neuroimage

    (2011)
  • E.A. Hazlett et al.

    Anterior limb of the internal capsule in schizotypal personality disorder: fiber-tract counting, volume, and anisotropy

    Schizophr. Res.

    (2012)
  • T.R. Insel et al.

    Endophenotypes: bridging genomic complexity and disorder heterogeneity

    Biol. Psychiatry

    (2009)
  • S.S. Kang et al.

    Disrupted functional connectivity for controlled visual processing as a basis for impaired spatial working memory in schizophrenia

    Neuropsychologia

    (2011)
  • R.S. Keefe et al.

    A pen-and-paper human analogue of a monkey prefrontal cortex activation task: spatial working memory in patients with schizophrenia

    Schizophr. Res.

    (1995)
  • R.S. Keefe et al.

    Performance of patients with schizophrenia on a pen and paper visuospatial working memory task with short delay

    Schizophr. Res.

    (1997)
  • S. Kuhn et al.

    Higher prefrontal cortical thickness in high schizotypal personality trait

    J. Psychiatr. Res.

    (2012)
  • D.S. Manoach et al.

    Schizophrenic subjects activate dorsolateral prefrontal cortex during a working memory task, as measured by fMRI

    Biol. Psychiatry

    (1999)
  • M. Matsui et al.

    The relationship between prefrontal brain volume and characteristics of memory strategy in schizophrenia spectrum disorders

    Prog. Neuropsychopharmacol. Biol. Psychiatry

    (2008)
  • M.M. McClure et al.

    Visual-spatial learning and memory in schizotypal personality disorder: continued evidence for the importance of working memory in the schizophrenia spectrum

    Arch. Clin. Neuropsychol.

    (2007)
  • P.D. McGorry et al.

    Can we improve the diagnostic efficiency and predictive power of prodromal symptoms for schizophrenia?

    Schizophr. Res.

    (2000)
  • S.A. Mitelman et al.

    Volume of the cingulate and outcome in schizophrenia

    Schizophr. Res.

    (2005)
  • V. Mitropoulou et al.

    Neuropsychological performance in schizotypal personality disorder: evidence regarding diagnostic specificity

    Biol. Psychiatry

    (2002)
  • S. Park et al.

    Working memory and the syndromes of schizotypal personality

    Schizophr. Res.

    (1997)
  • J. Pena et al.

    Executive functioning correctly classified diagnoses in patients with first-episode psychosis: evidence from a 2-year longitudinal study

    Schizophr. Res.

    (2011)
  • C. Ranganath

    Working memory for visual objects: complementary roles of inferior temporal, medial temporal, and prefrontal cortex

    Neuroscience

    (2006)
  • M.E. Shenton et al.

    A review of MRI findings in schizophrenia

    Schizophr. Res.

    (2001)
  • A.M. Shepherd et al.

    Systematic meta-review and quality assessment of the structural brain alterations in schizophrenia

    Neurosci. Biobehav. Rev.

    (2012)
  • S.D. Slotnick et al.

    Prefrontal cortex hemispheric specialization for categorical and coordinate visual spatial memory

    Neuropsychologia

    (2006)
  • M. Suzuki et al.

    Prefrontal abnormalities in patients with simple schizophrenia: structural and functional brain-imaging studies in five cases

    Psychiatry Res.

    (2005)
  • T. Takahashi et al.

    Morphologic alterations of the parcellated superior temporal gyrus in schizophrenia spectrum

    Schizophr. Res.

    (2006)
  • M.M. Voglmaier et al.

    Neuropsychological dysfunction in schizotypal personality disorder: a profile analysis

    Biol. Psychiatry

    (1997)
  • H.C. Whalley et al.

    Functional imaging as a predictor of schizophrenia

    Biol. Psychiatry

    (2006)
  • American-Psychiatric-Association

    DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders

    (2000)
  • A.J. Bergman et al.

    Verbal learning and memory in schizotypal personality disorder

    Schizophr. Bull.

    (1998)
  • Cited by (17)

    • Frontotemporal thalamic connectivity in schizophrenia and schizotypal personality disorder

      2022, Psychiatry Research - Neuroimaging
      Citation Excerpt :

      Less superior temporal gyrus gray matter volume was reported in SPD compared to healthy controls (HCs) (Dickey et al., 1999) with subsequent work (Dickey et al., 2002) indicating that Heschl's gyrus (but not planum temporale) gray matter volume was 21% lower in SPD compared to controls making it a potential vulnerability factor for schizophrenia-spectrum disorders. In contrast, greater frontal “reserve” in SPD compared to schizophrenia (Hazlett et al., 2014; Chan et al., 2018, 2008; Lener et al., 2015), which has been linked to less symptom severity (Hazlett et al., 2014), may protect individuals with SPD from severe cognitive deterioration and social deficits (Siever and Davis, 2004; Hazlett et al., 2014). One of the frontal regions most widely implicated in the neurobiology of schizophrenia is the rostral middle frontal cortex, which includes part of Brodmann's area 46, and corresponds approximately to the dorsolateral prefrontal cortex (Rajkowska and Goldman-Rakic, 1995).

    • Working-memory impairment in schizophrenia and schizotypal personality disorder

      2021, Cognitive and Behavioral Dysfunction in Schizophrenia
    • Masking impairments in schizophrenia and schizotypal personality disorder

      2021, Cognitive and Behavioral Dysfunction in Schizophrenia
    • Frontal and temporal cortical volume, white matter tract integrity, and hemispheric asymmetry in schizotypal personality disorder

      2018, Schizophrenia Research
      Citation Excerpt :

      In SPD, one of the most robust finding is reduced gray matter volume in the superior temporal gyrus (STG; Asami et al., 2013; Dickey et al., 1999; Downhill et al., 2001; Goldstein et al., 2009; Kawasaki et al., 2004; Takahashi et al., 2006a, 2010) and its component parts including Heschl's gyrus (Dickey et al., 2002) and planum temporale (Kawasaki et al., 2004; Takahashi et al., 2006a). The middle temporal gyrus (MTG) and inferior temporal gyrus (ITG) are also involved in auditory and language processing (Hickok and Poeppel, 2004), and studies in SPD have shown volume reduction in the MTG (Asami et al., 2013; Hazlett et al., 2008, 2014; Koo et al., 2006) and ITG (Asami et al., 2013; Downhill et al., 2001). Many of the findings reported are predominantly in the left hemisphere (see Fervaha and Remington, 2013; Hazlett et al., 2012) where language processing is localized.

    • Predicting individualized clinical measures by a generalized prediction framework and multimodal fusion of MRI data

      2017, NeuroImage
      Citation Excerpt :

      Note that the IPL reveals strong involvement in semantic processing, social cognition or theory-of-mind in several meta-analysis reviews (Binder et al., 2009; Mars et al., 2011; Seghier, 2013). Hazlett, et al. recently reported that larger BA10 volumes in the DLPFC predict less symptom severity (Hazlett et al., 2014), and Kawada et al. found executive dysfunction scores were correlated with volume reduction in DLPFC in schizophrenia (Kawada et al., 2009). The role of the cerebellum in schizophrenia has also been highlighted by Andreasen's hypothesis of “cognitive dysmetria”, which suggests a general dyscoordination of sensorimotor and mental processes (Cerasa et al., 2012; Moulis et al., 2014).

    View all citing articles on Scopus
    View full text