Smaller Cerebellar Vermis But Not Hemisphere Volumes in Patients With Chronic Schizophrenia

Fifty-nine patients (38 men and 21 women) with chronic schizophrenia according to DSM-IV criteria and 57 healthy volunteers (35 men and 22 women), all Caucasians, were included in this study (Table 1). All subjects lived within the area of northern Stockholm. They were recruited at the Karolinska Hospital Department of Clinical Neuroscience, where they were investigated between August 1999 and April 2002. At the time of the investigation, 37 of the male and 19 of the female patients were receiving antipsychotic medication. Dose of current neuroleptic treatment was converted into haloperidol units (13). All subjects were found to be healthy according to physical examination and blood and urinary tests. All subjects underwent a psychiatric interview (Structured Clinical Interview for DSM-III-R—Non-Patient Version) (14) to confirm schizophrenia in the patients and a lack of mental disorder in the comparison subjects. The interview was conducted by an experienced psychiatrist well trained in the administration of this instrument. Exclusion criteria for the healthy volunteers were current or past treatment for a psychiatric disorder or any history of alcohol or drug addiction, head trauma with loss of consciousness for more than 5 minutes, or severe somatic disorder.

The subjects were examined with a 1.5-Tesla GE Signa (Milwaukee) system at the MR Research Center, Karolinska Hospital, Stockholm. T₁-weighted images, using a spoiled gradient/recall acquisition in the steady state sequence, were acquired with the following parameters: 1.5-mm coronal slices, no gap, flip angle=35°, TR=24 msec, TE=6.0 msec, number of excitations=2, field of view=24 cm, acquisition matrix=256×192. T₂-weighted images were acquired with the following parameters: 2.0-mm coronal slices, no gap, TE=84 msec, TR=6000 msec, number of excitations=2, field of view=24 cm, acquisition matrix=256×192. From visual inspection, all scans were judged to be excellent without any obvious motion artifact. We rated the scan according to motion artifact and did not find a difference between patients and comparison subjects. All scans were found to lack gross clinical pathology as evaluated by a neuroradiologist. The quantitative analysis was performed blind to subject diagnosis and gender until the final statistical analysis.

MRI data analysis was performed with BRAINS software (15, 16). Detailed information about the segmentation procedure and evaluation of validity has been described by Harris et al. (17). Reproducibility and reliability of the segmentation procedure have also been ascertained previously (17, 18).

The detailed delineation of the cerebellar structures was described in the previous report (12). The cerebellar vermis was manually parcellated into anterior vermis (lobules I–V), posterior superior vermis (lobules VI–VII), and posterior inferior vermis (lobules VIII–X) in sagittal planes on the segmented images (Figure 1). The lateral extent of the vermis was determined in the following way. The vermian subregions were delineated from the sagittal slice in which the vermian subregions appeared and until the slice where the anterior vermis, posterior superior vermis, or posterior inferior vermis disappeared. The outer contour of the total vermis on a coronal projection is shown in Figure 2.

Cerebellar hemispheres and cerebellar tonsils were traced manually on all sagittal slices. The volume estimate of the cerebellar hemispheres included the cerebellar hemispheres and the tonsils but not the vermis.

Ten scans were randomly selected and evaluated blindly for inter- and intraoperator reliability. Intraclass correlation coefficients (19) were used to establish reliability to measure volumes of the cerebellar hemispheres, anterior vermis, posterior superior vermis, and posterior inferior vermis.

Absolute volume of cerebellar vermis was the volume of anterior vermis, posterior superior vermis, and posterior inferior vermis added together. Absolute volumes of cerebellar hemispheres, anterior vermis, posterior superior vermis, and posterior inferior vermis were independent factors. In the first analysis, we assessed diagnostic difference in absolute volumes of cerebellar hemispheres, anterior vermis, posterior superior vermis, and posterior inferior vermis by using the two-way multivariate analysis of variance (MANOVA), which included diagnosis, gender, and the diagnosis-by-gender interaction.

In the final analysis, the intracranial volume was used to correct for individual differences in head size. We assessed diagnostic differences for relative volumes (ratio of volume of interest to intracranial volume) of cerebellar subregions by using the two-way analysis of variance (ANOVA) model, which included diagnosis, gender, and the diagnosis-by-gender interaction. Differences were considered to be statistically significant if p<0.01 (i.e., 0.05/five regions of interest). Spearman’s correlation was used to test the correlation between volume measures and current dose of medication converted into haloperidol units or age or duration of illness in the patient group.

There were no significant differences in age, age at onset, duration of illness, or dose of medication between the subject groups (Table 1). There were significant differences in intracranial volume between male and female patients with schizophrenia (F=18.7, df=1, 57, p<0.001) and between male and female comparison subjects (F=18.0, df=1, 55, p<0.001). Intracranial volumes did not differ between schizophrenic and healthy male subjects (F=0.06, df=1, 71, p=0.80) or between schizophrenic and healthy female subjects (F=0.12, df=1, 41, p=0.73).

Intraclass correlations (ICCs) were used to establish interoperator reliability for the manual tracing of the cerebellar subregions by two experienced operators (I.A. and G.O.) on 10 scans. Intraoperator (G.O.) reliability was established on another set of 10 scans. Excellent inter- and intraoperator reliabilities were seen for volume measurements of anterior vermis, posterior superior vermis, posterior inferior vermis, and cerebellar hemispheres (all ICCs >0.95).

Table 2 presents the absolute volumes of cerebellar subregions in schizophrenic subjects and healthy subjects. Two-way MANOVA revealed that there was significant difference related to diagnosis (F=10.9, df=4, 109, p<0.001) and gender (F=8.2, df=4, 109, p<0.001), but there was no diagnosis-by-gender interaction (F=0.69, df=4, 109, p=0.60).

Table 3 presents the relative volumes of cerebellar subregions in schizophrenic subjects and healthy subjects. Significant differences were found for total vermis (F=26.3, df=1, 112, p<0.001), anterior vermis (F=13.5, df=1, 112, p<0.001), posterior superior vermis (F=38.7, df=1, 112, p<0.001), and posterior inferior vermis (F=9.04, df=1, 112, p=0.003), with smaller volumes seen in the schizophrenic subjects for both men and women. No significant difference related to diagnosis was found in the cerebellar hemisphere volumes. There were no gender differences for the relative volumes of cerebellar subregions. No significant gender-by-diagnosis interaction was found for any of the regions.

Absolute and relative volumes of the vermal subregions were not significantly correlated with current neuroleptic dosage. Neither was there any correlation between volume measures and age or duration of illness in the patient group.