Reproducibility of fMRI activations during a story listening task in patients with schizophrenia

https://doi.org/10.1016/j.schres.2011.01.025Get rights and content

Abstract

A prerequisite to longitudinal fMRI studies in schizophrenia is the knowledge on fMRI signal reliability in schizophrenia patients. We assessed the reproducibility of activations elicited by two fMRI sessions, which were 21 months apart, of a story listening paradigm in 10 schizophrenia patients and 10 healthy subjects. In both groups, we observed a high degree of spatial overlap of activation maps as well as a good reproducibility of signal variations assessed on a voxel-wise basis in temporal areas underlying early stages of language processing. Task performance, assessed through a comprehension questionnaire, had no impact on the activation reproducibility.

Introduction

Longitudinal fMRI studies in schizophrenia patients may help identify functional biomarkers of the disease (Bertolino et al., 2004, Mendrek et al., 2004, Razafimandimby et al., 2007) as well as help monitor the brain's response to therapies (Davis et al., 2005). Good test–retest agreement is a prerequisite for a longitudinal study, so a crucial question is whether fMRI activations are reproducible in schizophrenia patients (Cho et al., 2005).

Manoach et al. (2001) used the Intraclass Correlation Coefficient to investigate the reproducibility of fMRI activations elicited by a working memory paradigm. They reported low reproducibility in schizophrenia patients as compared to healthy subjects in areas underpinning high-order cognitive processes and a high and similar reproducibility in primary motor areas. In a previous study, we assessed the reproducibility of fMRI activations elicited by a story listening paradigm that targeted high-order language integration areas. This study evidenced that the lower fMRI activation reproducibility in schizophrenia patients as compared to healthy subjects was explained by the lower cognitive performance of the patients (Maïza et al., 2010).

In the present study, we tested the hypothesis that the reproducibility of fMRI activations in auditory and language areas that are involved in early stages of speech analysis is high and is comparable between schizophrenia patients and healthy controls. Indeed, to date, low reproducibility in schizophrenia patients has been reported in areas involved in high-order cognitive processes raising the hypothesis that areas supporting more basic processes might be unaffected by this low reproducibility. Consistently, we also postulated that the reproducibility of activations in these areas is not influenced by cognitive performance.

Section snippets

Participants

The participants of the present study were selected among 21 pairs of schizophrenia patients and healthy controls who participated to a previous cross sectional study (Dollfus et al., 2005). Ten pairs also participated in a second session. Eleven of the patients were not included: 3 did not give their consent, 3 were admitted to hospital between the 2 sessions, 4 did not stay in contact with their practitioner and in one case, the data from the second session were not usable owing to an

Clinical and behavioral data

In patients, PANSS scores and antipsychotic medication doses were not significantly different between sessions and were highly correlated across sessions (Table 2).

The ANOVA on comprehension scores detected no Group-by-Session interaction (F[1,18] = 0.32; p = 0.58). Controls performed significantly better on the comprehension questionnaire than did the patients in both sessions (F[1,18] = 4.60; p = 0.046). The slight increase in scores between sessions was not significant (F[1,18] = 2.58; p = 0.12).

Discussion

This study assessed the reproducibility of fMRI activations elicited by a story listening task. We found that there was a similar degree of reproducibility both globally and locally in schizophrenia patients as in healthy controls. Task performance was lower in patients but had no influence on activation reproducibility.

Our results evidenced good reproducibility in both groups. This is in line with previous reports of a spatial overlap ranging from 49% to 64% for sensory tasks, especially for

Role of funding source

This work was funded by the French Health Ministry in a Programme Hospitalier de Recherche Clinique and by the French Research Ministry. The funding source had no role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Contributors

Drs. Dollfus and Tzourio-Mazoyer designed the study. Drs. Maïza, Dollfus, Razafimandimby and Tzourio-Mazoyer acquired the data, Drs. Maïza, Mazoyer and Hervé analyzed the data. Drs. Maïza and Hervé wrote the article, and Drs. Maïza, Mazoyer, Razafimandimby, Dollfus and Tzourio-Mazoyer reviewed it. All authors approved the article that was submitted for publication.

Conflicts of interest

None.

Acknowledgements

We thank G. Perchey, N. Delcroix, and F. Lamberton for expert technical assistance.

References (24)

  • S. Virtue et al.

    Neural activity of inferences during story comprehension

    Brain Res.

    (2006)
  • American Psychiatric Association

    DSM-IV: Diagnostic and Statistical Manual of Mental Disorders

    (1994)
  • Cited by (14)

    • Social cognition in schizophrenia: Validation of an ecological fMRI task

      2019, Psychiatry Research - Neuroimaging
      Citation Excerpt :

      Validating our hypothesis, the activations induced by the ecological task were reproducible. The mean spatial overlap of 63.51%, which is in line with previous reports of spatial overlap ranging from 31% to 94% for sensory tasks (Gonzalez-Castillo and Talavage, 2011; Maïza et al., 2011) and from 35% to 50% for language tasks (Gonzalez-Castillo and Talavage, 2011; Maïza et al., 2010; Maldjian et al., 2002) among healthy control subjects. However, such comparisons must be interpreted cautiously, given the differences in methods and tasks among different studies.

    • Impact of rTMS on functional connectivity within the language network in schizophrenia patients with auditory hallucinations

      2017, Schizophrenia Research
      Citation Excerpt :

      Participants listened to a factual story in French (4 blocks) alternating with periods of silence (5 blocks), for a duration of 5 min. This task is known to elicit activation in phonology and acoustical processing areas involved in AVHs and to induce good reproducible activations (Maïza et al., 2011). MRI was acquired on a Philips 3 Tesla scanner with T1-weighted and T2-weighted images.

    • Validity and reliability of four language mapping paradigms

      2017, NeuroImage: Clinical
      Citation Excerpt :

      The advantages of the Dice coefficient are that it is widely used, it can be calculated in any individual without reference to a group, it yields a single metric of overall activation similarity encompassing all brain regions under consideration, and it is intuitive and easy to interpret (Bennett and Miller, 2010). There is a substantial literature on test-retest reproducibility of language paradigms (Brannen et al., 2001; Maldjian et al., 2002; Rutten et al., 2002b; Fernández et al., 2003; Billingsley-Marshall et al., 2004; Harrington et al., 2006; Rau et al., 2007; Fesl et al., 2010; Gonzalez-Castillo and Talavage, 2011; Maïza et al., 2011; Gross and Binder, 2014). Many of these papers have reported Dice coefficients or related metrics, but to our knowledge, no previous studies have reported Dice coefficients except in specific regions of interest for any of the four paradigms we investigated in this study.

    • Comment on Milanovic et al.

      2012, Schizophrenia Research
    View all citing articles on Scopus
    View full text