Elsevier

NeuroImage

Volume 36, Issue 1, 15 May 2007, Pages 188-193
NeuroImage

Brain activation mediates the association between structural abnormality and symptom severity in schizophrenia

https://doi.org/10.1016/j.neuroimage.2007.02.030Get rights and content

Thought disorder is a symptom of schizophrenia expressed as disorganized or incoherent speech. Severity of thought disorder correlates with decreased left superior temporal gyrus grey matter volume and cortical activation in posterior temporal regions during the performance of language tasks. The goal of this study was to determine whether language-related activation mediates the association between thought disorder and left superior temporal lobe grey matter volume. 12 patients with schizophrenia were assessed for thought disorder. FMRI images were acquired for each subject while they listened to English speech, along with a high resolution structural image. Thought disorder was used as a covariate in the functional analysis to identify brain regions within which activation correlated with symptom severity. Voxel based morphometry was used to calculate grey matter volume of the planum temporale. A mediation model wastested using a four-step multiple regression approach incorporating cortical volume, functional activation andsymptom severity. Thought disorder correlated with activation in a single cluster within the left posterior middle temporal gyrus during listening to speech. Grey matter volume within the planum temporale correlated significantly with severity of thought disorder and activation within the functional cluster. Regressing thought disorder on grey matter volume and BOLD response simultaneously led to a significant reduction in the correlation between grey matter volume and thought disorder. These results support the hypothesis that the association between decreased grey matter volume in the left planum temporale and severity of thought disorder is mediated by activation in the posterior temporal lobe during language processing.

Introduction

Schizophrenia is a psychiatric illness characterized by delusions, hallucinations, disordered thought and negative symptoms. Brain morphological and neuroimaging studies have reported subtle and wide-ranging irregularity in schizophrenia, frequently associated with specific symptom profiles. Among the few consistent anatomical findings reported in schizophrenia is abnormality of left posterior superior temporal cortex (Barta et al., 1997, Falkai et al., 1995, Menon et al., 1995, Rossi et al., 1992, Shenton et al., 1992), including the planum temporale (Barta et al., 1997, Falkai et al., 1995, Rossi et al., 1992). The planum temporale is the horizontal aspect of the posterior temporal lobe, it has a leftward asymmetry (Geschwind and Levitsky, 1968), overlapping with Wernicke's area, and is classically associated with language function. The planum temporale largely consists of auditory association cortex; lesion and imaging research indicates involvement in early auditory and phonological processing (Binder et al., 1997, Binder et al., 1996, Caplan et al., 1995, Petersen et al., 1988, Wise et al., 1991). In his review of the structural, functional and clinical literature regarding the planum temporale, Shapleske points out that the planum temporale likely engages in language tasks as a functional unit and is not inherently linguistic (Shapleske et al., 1999).

Reduced gray matter volume in left superior temporal cortex, including planum temporale, is associated with an increased severity of thought disorder in schizophrenia (Menon et al., 1995, Rajarethinam et al., 2000, Rossi et al., 1994, Shenton et al., 1992, Vita et al., 1995). The finding is robust, despite methodological variation in the measurement of cortical abnormality, anatomical definition of brain regions, and assessment of thought disorder. In a recent fMRI study we observed a correlation between left posterior temporal activation during listening to speech and severity of thought disorder in schizophrenic patients (Weinstein et al., 2006). Other investigators have described correlations between activation in temporal cortex and thought disorder during the performance of generative language tasks (Kircher et al., 2001a, Kircher et al., 2001b, Kircher et al., 2002, McGuire et al., 1998).

Although thought disorder has been linked with both temporal lobe function and gray matter volume separately, the possibility that these three variables may be interrelated has not been tested. The planum temporale operates at an earlier linguistic level than that of posterior superior and middle temporal gyri; therefore, the relationship between planum temporale structural abnormality and thought disorder may be arbitrated by the degree of activation in the posterior temporal lobe. Specifically, our hypothesis is that the association between decreased left planum temporale gray matter volume and greater severity of thought disorder is mediated by increased activation in left posterior temporal cortex during receptive language processing.

Mediation implies a very specific type of statistically testable relationship that extends beyond the concept of a partial correlation. True mediation suggests that an intermediate variable functions as a mechanism by which an independent variable exerts its influence on a dependent variable. Although this concept is widely recognized and tested in the social sciences, we are unaware of any investigation to date that has applied the methodology to neuroimaging data. In this study, we used a conventional mediation analysis procedure (Baron and Kenny, 1986) to directly investigate the proposal that temporal lobe functioning is the mechanism by which planum temporale structural abnormality influences severity of formal thought disorder.

Section snippets

Subjects

Twelve patients with a DSM-IV diagnosis of schizophrenia were recruited from the Schizophrenia Day Program at Vancouver Hospital or through the primary care physician. Subjects were selected according to psychopathology to produce a sample with a range of thought disorder (from none detectable to severe). Diagnosis according to DSM-IV criteria was confirmed by one of the authors (EN). All were right-handed (Annett, 1967) native English speakers with no history of head injury, hearing loss,

Results

TLI scores ranged from 1.50 to 12.75 (M = 5.01, SD = 3.67). SSPI scores ranged from 0 to 19 (M = 9.50, SD = 5.70). A whole-brain regression analysis of the fMRI data, with TLI score entered as a covariate, was conducted to identify regions active during listening to English speech in which BOLD correlated with thought disorder (cluster-level corrected p < 0.05). One cluster of 54 voxels was significant at this criterion, spanning the left middle–posterior superior and middle temporal gyri (maximum at − 48

Discussion

The objective of this study was to investigate whether the association between decreased left planum temporale gray matter volume and severity of thought disorder is mediated by increased activation in left posterior temporal cortex during language processing. Until now, the relationships among these three variables had only been investigated with a pair-wise approach. Brain activation during a simple listening task correlated positively with thought disorder in this sample, and we replicated

References (38)

  • A. Rossi et al.

    Planum temporale in schizophrenia: a magnetic resonance study

    Schizophr. Res.

    (1992)
  • A. Rossi et al.

    Planum temporale asymmetry and thought disorder in schizophrenia

    Schizophr. Res.

    (1994)
  • J. Shapleske et al.

    The planum temporale: a systematic, quantitative review of it's structural, functional and clinical significance

    Brain Res. Rev.

    (1999)
  • A. Vita et al.

    Language and thought disorder in schizophrenia: brain morphological correlates

    Schizophr. Res.

    (1995)
  • S. Weinstein et al.

    Do you hear what I hear? Neural correlates of thought disorder during listening to speech in schizophrenia

    Schizophr. Res.

    (2006)
  • M. Annett

    The binomial distribution of right, mixed, and left handedness

    Q. J. Exp. Psychol.

    (1967)
  • R.M. Baron et al.

    The moderator–mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations

    J. Pers. Soc. Psychol.

    (1986)
  • P.E. Barta et al.

    Planum temporale asymmetry reversal in schizophrenia: replication and relationship to grey matter abnormalities

    Am. J. Psychiatry

    (1997)
  • J.R. Binder et al.

    Function of the left planum temporale in auditory and linguistic processing

    Brain

    (1996)
  • Cited by (25)

    • A systematic review of task-based functional neuroimaging studies investigating language, semantic and executive processes in thought disorder

      2018, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      One avenue to help elucidate the contributions of these cognitive processes to the manifestation of TD is through the use of neuroimaging. The most replicated structural correlates of TD (for reviews, see Cavelti et al., 2018; Sumner et al., 2018) encompass reduced volumes within the left posterior superior temporal gyrus, including the planum temporale (Anderson et al., 2002; Barta et al., 1997; Horn et al., 2009; Rajarethinam et al., 2000; Sans-Sansa et al., 2013; Shenton et al., 1992; Weinstein et al., 2007). However, other replicated findings implicate the left and/or right orbitofrontal cortex (Gur et al., 2004; Horn et al., 2010; Nakamura et al., 2008; Sans-Sansa et al., 2013), the left amygdala-hippocampus complex (Bogerts et al., 1993; Rajarethinam et al., 2001; Sallet et al., 2003; Spalletta et al., 2010), the bilateral nucleus accumbens (Ballmaier et al., 2004; Spalletta et al., 2010), the cerebellar vermis (Sandyk et al., 1991), and the corpus callosum (Arnedo et al., 2015; Kubicki et al., 2008).

    • Is formal thought disorder in schizophrenia related to structural and functional aberrations in the language network? A systematic review of neuroimaging findings

      2018, Schizophrenia Research
      Citation Excerpt :

      Both steps were done separately for sMRI, DTI, resting-state fMRI and task-based fMRI studies. The three studies reporting both structural and functional findings (Horn et al., 2009; Skudlarski et al., 2010; Weinstein et al., 2007) were considered twice for the summaries given in the tables and Fig. 3. If the study design included three groups (HC, schizophrenia patients with FTD; SZ-FTD, schizophrenia patients without FTD; SZ-nFTD), only the differences between the patient groups are presented in the tables, because differences between HC and SZ patients could be due to other factors than FTD such as general aspects of pathology or medication.

    View all citing articles on Scopus
    View full text