Elsevier

Schizophrenia Research

Volume 180, February 2017, Pages 13-20
Schizophrenia Research

Grey matter volume patterns in thalamic nuclei are associated with familial risk for schizophrenia

https://doi.org/10.1016/j.schres.2016.07.005Get rights and content

Abstract

Previous evidence suggests reduced thalamic grey matter volume (GMV) in patients with schizophrenia (SCZ). However, it is not considered an intermediate phenotype for schizophrenia, possibly because previous studies did not assess the contribution of individual thalamic nuclei and employed univariate statistics. Here, we hypothesized that multivariate statistics would reveal an association of GMV in different thalamic nuclei with familial risk for schizophrenia. We also hypothesized that accounting for the heterogeneity of thalamic GMV in healthy controls would improve the detection of subjects at familial risk for the disorder.

We acquired MRI scans for 96 clinically stable SCZ, 55 non-affected siblings of patients with schizophrenia (SIB), and 249 HC. The thalamus was parceled into seven regions of interest (ROIs). After a canonical univariate analysis, we used GMV estimates of thalamic ROIs, together with total thalamic GMV and premorbid intelligence, as features in Random Forests to classify HC, SIB, and SCZ. Then, we computed a Misclassification Index for each individual and tested the improvement in SIB detection after excluding a subsample of HC misclassified as patients.

Random Forests discriminated SCZ from HC (accuracy = 81%) and SIB from HC (accuracy = 75%). Left anteromedial thalamic volumes were significantly associated with both multivariate classifications (p < 0.05). Excluding HC misclassified as SCZ improved greatly HC vs. SIB classification (Cohen's d = 1.39). These findings suggest that multivariate statistics identify a familial background associated with thalamic GMV reduction in SCZ. They also suggest the relevance of inter-individual variability of GMV patterns for the discrimination of individuals at familial risk for the disorder.

Introduction

The thalamus is comprised of numerous nuclei with sparse reciprocal connections, hence belonging to relatively independent circuits (Jones, 2007). The anatomical segregation of thalamic nuclei is reflected in their functional specialization. Accordingly, focal lesions yield different clinical correlates based on lesion localization (Pergola and Suchan, 2013). Consistent with the functional specialization of thalamic nuclei, post-mortem histology findings suggest that the volume of specific nuclei is reduced in patients with schizophrenia (SCZ) (reviewed by Alelu-Paz and Gimenez-Amaya, 2008, Pakkenberg et al., 2009). In particular, volume reductions in the mediodorsal thalamic nucleus (MD), pulvinar (Pul), anterior and midline thalamic nuclei (AT), and intralaminar nuclei (ILN) together explain the volume reduction of the entire thalamus (Byne et al., 2009). However, the small sample sizes limit the statistical power of post-mortem studies in SCZ.

In vivo studies with MRI using larger samples have shown significant grey matter volume (GMV) reduction in the thalamus of SCZ (Van Erp et al., 2015). However, a number of questions regarding the relationship between reduced thalamic volume and schizophrenia have not been addressed by the previous literature. First, it is not clear if reduced thalamic volume documented by imaging studies pertains to specific nuclei. In fact, the study of the thalamus as a whole does not take into account the functional and anatomical segregation of thalamic nuclei. On the other hand, examining GMV reduction in specific thalamic nuclei has led to controversial findings, possibly because assessing thalamic GMV requires specific imaging processing procedures (reviewed by Pergola et al., 2015).

Second, previous studies mostly employed univariate statistics. However, a substantial overlap is observed between patients and controls when evaluating single brain regions with univariate methods (Kambeitz et al., 2015). These procedures cannot test whether thalamic GMV estimates have a specific configuration across multiple nuclei that differentiates SCZ from healthy controls (HC). Multivariate statistics, instead, consider ensembles of brain regions at once to discriminate patients from controls. Even if individual components of complex brain patterns are not strongly associated with the illness, the multivariate pattern can be very sensitive in discriminating different populations (Kambeitz et al., 2015). For this reason, multivariate techniques are thought to yield greater sensitivity compared to univariate statistics (Zarogianni et al., 2013). Accordingly, these methods have been successful in identifying brain GMV patterns associated with diagnostic and prognostic variables in SCZ (Koutsouleris et al., 2012, Mourao-Miranda et al., 2012). However, to the best of our knowledge multivariate techniques have not been applied to study thalamic nuclei.

Third, it remains unclear whether thalamic GMV reduction is a trait associated with genetic risk for the disorder, or whether it is associated, for instance, with the clinical course of schizophrenia, i.e., state related. A common strategy to tackle this problem is investigating non-affected siblings of patients (SIB). SIB share on average 50% of genetic variation with patients and are free of disease-associated confounding variables. Therefore, studies with SIB are important to characterize intermediate phenotypes for schizophrenia, which are measures related to molecular genetics of the disease that are heritable, co-segregate with a psychiatric illness, yet are state-independent (Bertolino and Blasi, 2009, Gottesman and Gould, 2003). Notably, although thalamic GMV reduction is heritable (Den Braber et al., 2013) and is characteristic of SCZ (Van Erp et al., 2015), it is yet unclear whether it can be considered an intermediate phenotype for schizophrenia (Allen et al., 2009, Boos et al., 2007, Goldman et al., 2008, Honea et al., 2008). Recently, a meta-analysis including only six voxel-based morphometry (VBM) studies revealed reduced thalamic GMV in first-degree relatives of SCZ (Cooper et al., 2014). Thus, one reason for the lack of consensus on thalamic GMV reduction as an intermediate phenotype for schizophrenia may be that few VBM studies investigated specifically this brain region. Another reason may be that these studies do not take in account inter-individual variability in brain patterns of GMV, which is present also in healthy subjects and may be associated with genetic factors (Pergola et al., 2015). Indeed, genetic factors can be associated with phenotypes related to schizophrenia also in healthy populations, as indicated by several imaging genetics studies (Bertolino and Blasi, 2009).

Here, our aim was to investigate these topics. In particular, we examined GMV of individual thalamic nuclei in HC, SIB and SCZ and assessed with univariate statistics nuclei-specific GMV reductions within the thalamus. We did find specific GMV reductions, but consistent with prior evidence, we found no intermediate phenotype for schizophrenia. Then, we studied whether multivariate analyses provided greater sensitivity to detect thalamic phenotypes of familial risk for schizophrenia. We considered in the prediction also premorbid intelligence, which can be easily collected in a clinical setting and may be relevant to early-stage detection of schizophrenia (Kendler et al., 2016). Finally, we investigated whether taking into account inter-individual variability in thalamic patterns of GMV in healthy subjects may increase the detection power of greater familial risk for schizophrenia. With this aim, we removed HC with SCZ-like thalamic features from the HC sample; then, we assessed whether detection of subjects at greater familial risk for schizophrenia, i.e., SIB, improved. Here, we hypothesized that this procedure could enhance the sensitivity of risk detection analyses. Improved identification of at-risk subjects would also support the hypothesis that thalamic GMV is compromised in subjects at greater familial risk for schizophrenia.

Section snippets

Participants

We recruited 400 Caucasian subjects: 96 SCZ (DSM-IV-TR) selected among consecutive outpatients at the University Hospital of Bari; 55 SIB, 249 HC. Table 1 reports demographic information. Exclusion criteria for all groups were history of drug or alcohol abuse in the past year, non-psychiatric clinically relevant conditions, history of neurological diseases and head trauma with loss of consciousness. Absence of psychiatric illness in SIB and HC was established using the Structured Clinical

Results

The whole-brain VBM analysis yielded a significant main effect of diagnosis on the GMV at the level of the right thalamus (MNI = 8,− 12,8; cluster extent = 363; F2,394 = 17; FWE-corrected p < 0.05). Both HC and SIB had greater thalamic GMV compared to SCZ (Supplemental Fig. 1). Other brain regions surviving the statistical threshold are reported in Supplemental Table 1.

Discussion

The present study aimed to investigate the association between thalamic GMV patterns and risk for schizophrenia. First, we found with univariate analysis that specific thalamic nuclei differentiated HC from SCZ but not from SIB, most prominently the MD. This finding suggests that GMV reduction in the MD using this statistical approach is associated with the disease but not with familial risk for schizophrenia. Second, multivariate analyses supported the idea that GMV reduction in the MD are

Conclusions

In this study we parceled the thalamus in several ROIs and examined the association between GMV estimates in these ROI and diagnosis of schizophrenia with univariate and multivariate approaches. We found that multivariate approaches combined with thalamic parceling and premorbid intelligence explain a significant portion of the differences between diagnostic groups and carry information related with familial risk for schizophrenia. The findings highlight the suitability of multivariate

Conflict of interest

Giulio Pergola is the academic supervisor of a collaborative research project with Hoffman-La Roche, Ltd. All other authors report no potential conflicts of interest.

Contributors

GP, PDC, AB, and GB designed the study. ST, PT, MM, MAN, IA, GC, TP, AR, ADG, AB, and GB collected the data. GP, ST, PDC, MM, NA, MAN, IA, GC, TP, ADG, AB, and GB performed the analyses. GP, PDC, PT, MM, NA, GC, ADG, AB, and GB interpreted the results. GP, PDC, AB, and GB wrote the manuscript. All authors critically revised the manuscript, proofread it, and prepared it for submission.

Role of the funding source

The funding sources had no role in the design of the study, in data collection and analysis.

Acknowledgments

This research has been partly funded by the “Capitale Umano ad Alta Qualificazione” grant awarded by Fondazione Con Il Sud (2011-PDR-06) to AB. This project has received funding from the European Union Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 602450. This paper reflects only the author's views and the European Union is not liable for any use that may be made of the information contained therein.

We thank Andrea Tateo for his

References (56)

  • A. Jakary et al.

    N-acetylaspartate reductions in the mediodorsal and anterior thalamus in men with schizophrenia verified by tissue volume corrected proton MRSI

    Schizophr. Res.

    (2005)
  • A. Krauth et al.

    A mean three-dimensional atlas of the human thalamus: generation from multiple histological data

    NeuroImage

    (2010)
  • J. Lisman

    Excitation, inhibition, local oscillations, or large-scale loops: what causes the symptoms of schizophrenia?

    Curr. Opin. Neurobiol.

    (2012)
  • J.E. Lisman et al.

    A thalamo-hippocampal-ventral tegmental area loop may produce the positive feedback that underlies the psychotic break in schizophrenia

    Biol. Psychiatry

    (2010)
  • S. Ocklenburg et al.

    Laterality and mental disorders in the postgenomic age—a closer look at schizophrenia and language lateralization

    Neurosci. Biobehav. Rev.

    (2015)
  • R.C. Oldfield

    The assessment and analysis of handedness: the Edinburgh inventory

    Neuropsychologia

    (1971)
  • G. Pergola et al.

    Recall deficits in stroke patients with thalamic lesions covary with damage to the parvocellular mediodorsal nucleus of the thalamus

    Neuropsychologia

    (2012)
  • G. Pergola et al.

    The role of the thalamic nuclei in recognition memory accompanied by recall during encoding and retrieval: an fMRI study

    NeuroImage

    (2013)
  • G. Pergola et al.

    The role of the thalamus in schizophrenia from a neuroimaging perspective

    Neurosci. Biobehav. Rev.

    (2015)
  • A. Qiu et al.

    Neuroanatomical asymmetry patterns in individuals with schizophrenia and their non-psychotic siblings

    NeuroImage

    (2009)
  • K.A. Young et al.

    Reduced number of mediodorsal and anterior thalamic neurons in schizophrenia

    Biol. Psychiatry

    (2000)
  • E. Zarogianni et al.

    Towards the identification of imaging biomarkers in schizophrenia, using multivariate pattern classification at a single-subject level

    NeuroImage Clin.

    (2013)
  • J.P. Aggleton et al.

    Episodic memory, amnesia, and the hippocampal-anterior thalamic axis

    Behav. Brain Sci.

    (1999)
  • R. Alelu-Paz et al.

    The mediodorsal thalamic nucleus and schizophrenia

    J. Psychiatry Neurosci.

    (2008)
  • H. Ananth et al.

    Cortical and subcortical gray matter abnormalities in schizophrenia determined through structural magnetic resonance imaging with optimized volumetric voxel-based morphometry

    Am. J. Psychiatry

    (2002)
  • H. Barbas et al.

    Frontal-thalamic circuits associated with language

    Brain Lang.

    (2013)
  • H.B. Boos et al.

    Brain volumes in relatives of patients with schizophrenia: a meta-analysis

    Arch. Gen. Psychiatry

    (2007)
  • L. Breiman

    Random forests

    Mach. Learn.

    (2001)
  • Cited by (35)

    • Joint structural-functional magnetic resonance imaging features are associated with diagnosis and real-world functioning in patients with schizophrenia

      2022, Schizophrenia Research
      Citation Excerpt :

      At the spatial level, our results show that structural–functional covariation between frontal, parietal, temporal, and thalamic regions varies between SCZ and HV. All these regions have been found as compromised in patients both at the structural and functional level (Antonucci et al., 2019; Antonucci et al., 2016; Di Carlo et al., 2020; John, 2009; Khadka et al., 2013; Pergola et al., 2015; Pergola et al., 2017; van den Heuvel et al., 2010). These findings are consistent with the hypothesis of altered development in schizophrenia.

    View all citing articles on Scopus
    View full text