Theory of mind profile and cerebellar alterations in remitted bipolar disorder 1 and 2: a comparison study

Abstract

The literature on social cognition abilities in bipolar disorder (BD) is controversial about the occurrence of theory of mind (ToM) alterations. In addition to other cerebral structures, such as the frontal and limbic areas, the processing of socially relevant stimuli has also been attributed to the cerebellum, which has been demonstrated to be involved in the above-mentioned disorder. Nevertheless, the cerebellar contribution to ToM deficits in bipolar patients needs to be elucidated further. To this aim, two tests assessing different components of ToM were used to evaluate the ability to appreciate affective and mental states of others in 17 individuals with a diagnosis of BD type 1 (BD1) and 13 with BD type 2 (BD2), both in the euthymic phase, compared to healthy matched controls. Cerebellar gray matter (GM) volumes were extracted and compared between BD1 and controls and BD2 and controls by using voxel-based morphometry. The results showed that BD1 patients were compromised in the cognitive and advanced components of ToM, while the BD2 ToM profile resulted in a more widespread compromise, also involving affective and automatic components. Both overlapping and differing areas of cerebellar GM reduction were found. The two groups of patients presented a pattern of GM reduction in cerebellar portions that are known to be involved in the affective and social domains, such as the vermis and Crus I and Crus II. Interestingly, in both BD1 and BD2, positive correlations were detected between lower ToM scores and decreased volumes in the cerebellum. Overall, BD2 patients showed a more compromised ToM profile and greater cerebellar impairment than BD1 patients. The different patterns of structural abnormalities may account for the different ToM performances evidenced, thus leading to divergent profiles between BD1 and BD2.

Keywords: emotion; gray matter; mentalizing; social cognition; voxel-based morphometry.

Figure 1

ToM profile of BD1 and BD2 patients. The results for each test are presented as the percentage of the total number of correct responses (accuracy); 0% indicates no correct answers, and 100% indicates totally correct. The mean and standard deviation of the accuracy are reported for BD1, HS-ToM, and BD2. *Statistical significance.

Figure 2

Between-group voxel-based comparison of cerebellar GM density. Cerebellar regions showing patterns of significantly reduced GM in BD1 (A) and BD2 (B) compared to MRI-HS are reported and superimposed on the Spatially Unbiased Infratentorial Template (SUIT; Diedrichsen et al., 2009) in coronal (y), sagittal (x), and axial (z) slices. The results are significant at p-values < 0.05 after FWE cluster-level correction. Images are shown in neurological convention. Regions of overlapping cerebellar GM loss (C) between BD1 (in red) and BD2 (in green) are reported in orange.

Figure 3

Data scatterplots. Significant correlations between decreased cerebellar GM volumes and the poor ToM scores in BD1 (A) and BD2 (B) patients are reported. (A) Positive correlations between the total score (r = 0.540; p = 0.025) and the cognitive component score (r = 0.511; p = 0.036) on the Faux Pas stories and GM volumes in the Vermis Crus II. (B) Positive correlations between the scores at the negative stimuli of the RMET and the reduced cerebellar GM volumes in the left crus I (r = 0.625; p = 0.022) and in the left lobule VI (r = 0.756; p = 0.003).