Cerebellar volumes in men with schizophrenia and alcohol dependence
Katarina Varnäs, PhD Department of Clinical Neuroscience, Psychiatry Section, Karolinska Institutet, SE-171 76 Stockholm, Sweden. Email: firstname.lastname@example.org
Abstract The aim of the present study was to investigate the variation in cerebellar morphology in schizophrenia and alcohol dependence. Volumetric measurements of cerebellar structures were performed using magnetic resonance imaging in 17 men with schizophrenia, nine men with alcohol dependence, and 18 healthy men. Schizophrenia patients had smaller volumes of the posterior superior vermis, while alcohol-dependent patients had smaller volumes of both the anterior and the posterior superior vermis compared with controls. The groups were not significantly different with respect to cerebellar hemisphere volumes. The results provide indications for differential morphological abnormalities of the cerebellar vermis in patients with schizophrenia and alcohol dependence.
Findings from magnetic resonance imaging (MRI) investigations have provided support for dysmorphology of the cerebellum, characterized by a smaller vermis and relative preservation of the cerebellar hemispheres, in patients with schizophrenia compared to healthy controls.1–4 It has been hypothesized, based on the pathology of the vermis in neurodevelopmental disorders,5–7 that the smaller volume of this region could be due to abnormal neurodevelopment of the vermis in schizophrenia.1 Because dysmorphology of the vermis is also found in alcohol dependence,8 alcohol consumption may contribute to some of the vermian abnormalities seen in patients with schizophrenia. This pilot investigation was undertaken to examine whether regionally distinct patterns of cerebellar dysmorphology are detectable and different in patients with schizophrenia and in patients with alcohol dependence.
The protocol was approved by the ethics committee at the Karolinska Hospital, Stockholm, Sweden. The work was conducted in accordance with the Declaration of Helsinki. All subjects gave written informed consent.
Seventeen men fulfilling DSM-IV criteria for schizophrenia, nine men fulfilling DSM-IV criteria for chronic alcohol dependence, and 18 healthy men participated in the investigation (Table 1). Schizophrenia patients and controls represented a subset of a previously investigated subject sample,3 and they were selected to match the age range of patients with alcohol dependence (Table 1). Schizophrenia patients with a diagnosis of alcohol abuse or dependence, as evaluated in clinical interviews, were excluded from the study. At the time of investigation none of the alcohol-dependent subjects fulfilled DSM-IV criteria for schizophrenia or major affective disorders, although two of the patients had acomorbid diagnosis of anxiety-related disorders. Alcohol consumption among schizophrenia patients and healthy controls was estimated using the Alcohol Use Disorder Identification Test (AUDIT).9 The alcohol-dependent patients had been abstinent for at least 20 days before the investigation (range 21–91 days; see Agartz et al. for clinical characterization10).
Table 1. Demographic and clinical data of men with schizophrenia, men with alcohol dependence and healthy men
|No. right-handed subjects||11|| ||8†|| ||17|| |
|Age at onset of illness (years)||26.9||16–45||27.3‡||16–42|| || |
|Illness duration (years)||18.9||3–34||17.8‡||6–33|| || |
|Current antipsychotic medication (haloperidol equivalents, mg/day)||3.0|| 0–8.0|| || || || |
|AUDIT total score§||4||0–13||N/A|| ||5|| 1–9|
|AUDIT score, items 1–3§,¶||3||0–10||N/A|| ||4|| 0–9|
The subjects were examined with a 1.5-Tesla GE Signa (General Electric, Milwaukee, WI, USA) system at the MR Research Center, Karolinska Hospital, Stockholm, Sweden. T1-weighted images, using a spoiled gradient/recall acquisition, 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. All scans were judged to be excellent, without obvious motion artifacts, and without gross clinical pathology as evaluated by a neuroradiologist.
MRI data analysis was performed using the BRAINS software (University of Iowa, Iowa City, IA, USA).11,12 Intracranial volumes were defined in segmented MR images using artificial neural networks as described in Magnotta et al.13 Tracing of the cerebellar structures was performed as described previously.3,4 The vermis was manually parcellated into anterior, posterior superior, and posterior inferior subregions in the sagittal planes on segmented images. Cerebellar hemispheres and 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.
Intra-class correlations (ICCs) were used to establish interrater reliability for manual tracing of the cerebellar subregions by two experienced operators on 10 scans. Intra-rater reliability was established on another set of 10 scans. Excellent inter- and intrarater reliabi-lities were seen for volume measurements of the vermian regions and cerebellar hemispheres (all ICCs > 0.95, see Okugawa et al.3). All vermis delineations were performed by the same operator (GO). The time period between manual tracings for patients with schizophrenia and controls, and patients with alcohol dependence was approximately 20 weeks.
To correct for individual differences in head size, adjusted volumes were obtained by dividing absolute volumes by intracranial volumes. First, three separate one-way anovas were performed in order to control for group differences in age, AUDIT scores and intracranial volume. There was a borderline significant group difference in age. In order to control for the influence of age on vermian volumes, a general linear model was estimated with diagnostic group as categorical predictor and age as continuous predictor and with vermian subregion as within-group factor (with three levels). After a significant interaction was found, Tukey'spost-hoc test was used to perform specific comparisons between the groups within each subregion of thecerebellum. Finally, a two-way anova with repeated measures, with left versus right cerebellar hemisphere as the two levels of the repeated factor, was performed to assess group differences in volume of the cerebellar hemispheres (vermis excluded).
The groups were not significantly different with respect to intracranial volume (F2,41 = 0.493, P = 0.614, Table 2). Patients with schizophrenia and healthy subjects did not consume significantly different amounts of alcohol (P > 0.10), as estimated by AUDIT. With respect to age there was a borderline significant group difference (F2,41 = 2.87, P = 0.058), with higher age among patients with alcohol dependence. The general linear model yielded a significant interaction effect for vermian subregion × diagnostic group (F2,41 = 11.38, P < 0.000). There was no significant vermian subregion × age interaction (P = 0.308). Tukey's post-hoc test revealed that patients with alcohol dependence had significantly smaller volumes of anterior vermis than patients with schizophrenia (P < 0.001) and controls (P < 0.001), and that alcohol-dependent patients had significantly smaller volumes of posterior superior vermis than controls (P < 0.01). It also revealed that schizophrenia patients had significantly smaller volumes of posterior superior vermis than controls (P < 0.011). There were no significant differences between left and right cerebellar hemisphere and no significant interaction between diagnostic group and cerebellar hemisphere volumes.
Table 2. Relative volumes of cerebellar regions and intracranial volume in men with schizophrenia, men with alcohol dependence and healthy men
|Intracranial volume (mL)||1525||98||1532||107||1561||122|
| Cerebellar hemispheres, left||4.10||0.30||3.99||0.20||3.95||0.44|
| Cerebellar hemispheres, right||4.12||0.31||3.96||0.22||4.00||0.42|
| Cerebellar vermis||0.53||0.05||0.42||0.08||0.61||0.07|
| Anterior vermis||0.27||0.04||0.18||0.05||0.28||0.04|
| Posterior superior vermis||0.12||0.02||0.11||0.02||0.17||0.03|
| Posterior inferior vermis||0.14||0.02||0.13||0.02||0.16||0.02|
The results indicated smaller volumes of cerebellar vermis in men with alcohol dependence and schizophrenia. Alcohol-dependent patients showed smaller volumes of the anterior and posterior superior vermian subregions, while patients with schizophrenia showed volume reductions confined to the posterior superior subregion.
In a previous investigation of cerebellar volume in patients with schizophrenia and controls significantly smaller volumes were found for all subregions of the vermis with the most prominent volume reduction of the posterior superior vermis.4 In the present study of a subset of the previously investigated subjects no differences were found in the volumes of the anterior or the posterior inferior subregions of the vermis.
Our findings are consistent with results of Joyal et al. who found selective volume reduction of the posterior superior vermis in patients with schizophrenia, with more widespread effects in subjects with a dual diagnosis of schizophrenia and alcohol dependence.2 However, another investigation found smaller vermis in patients comorbid for schizophrenia and alcohol dependence, but not in schizophrenia patients without a diagnosis of alcohol dependence.14 Moreover, in contrast to our findings two morphometric studies found dysmorphology of the vermis in schizophrenia confined to the anterior segment.1,15 The divergent results obtained in these studies could possibly reflect inclusion of different patient populations with respect to variability within the schizophrenia disorders or different levels of alcohol consumption.
The results indicating smaller volume of the anterior vermian subregions in alcohol-dependent patients are in agreement with a previous MRI study of the vermis in alcohol dependence.8 In contrast to this study, however, we found no indications of smaller cerebellar hemisphere volumes in patients with alcohol dependence compared to controls. In the previous study cerebellar hemisphere volume deficits were limited to gray matter tissue. This could have accounted for the discrepant results, as in the present study cerebellar volume estimates included both gray and white matter tissue.
Given previous findings in patients with a dual diagnosis of schizophrenia and alcohol dependence, it is possible that alcohol consumption may contribute to the dysmorphology of vermis as detected in patients with schizophrenia in the present study. However, the similar pattern of alcohol consumption among schizophrenia patients and controls in the present study, would argue against such an explanation.
In summary, the results indicate smaller volumes of cerebellar vermis in alcohol dependence and schizophrenia. The data provide some support for differential morphological abnormalities of the vermian subregions in patients with schizophrenia and alcohol dependence.
This study was supported by the Fredrik and Ingrid Thuring Foundation, the Swedish Medical Research Council (K2004-21X-15078-01A, 2003-5845 and K2003-21KX 14645-01A) and the Wallenberg foundation, Sweden. Monica Hellberg, Margareta Gard-Hedander and Else-Britt Hillner are acknowledged for invaluable assistance in the recruitment and handling of patients throughout the study.