This study validated the Chinese version of CCAS-s in patients with acute cerebellar infarction and adjusted the pass/fail diagnostic cut-off scores for CCAS-s tests. The Chinese CCAS-s showed acceptable reliability and validity in cerebellar infarction, whereas with relatively lower specificity and sensitivity than those reported in previous studies11. Our analyses further demonstrated that CCAS-s raw scores were correlated with age, education, and fine motor skills.
Our findings found that the Chinese CCAS-s was useful to identify CCAS in patients with cerebellar infarction. The CCAS-s raw score and the number of failure items were significantly different between patients and healthy controls. The high area under curve (AUC) of these indexes suggested a moderate discriminative ability of the Chinese CCAS-s. Patients performed significantly worse than healthy controls on phonemic fluency and semantic fluency supporting the core cognitive features of impaired executive function in CCAS12. The ascending cerebellar projections to the frontoparietal cortex and the feedback loops may be the neural substrates of cerebellar involvement in executive function13. In addition, no group differences for CCAS-s tests were found between isolated and mixed cerebellar infarction, suggesting an independent and crucial role of the cerebellum in cognition.
The Cronbach alpha in this cohort was 0.72, suggesting that the items within the scale are necessary to measure the affected cognitive domains in CCAS and are not redundant. The Cronbach alpha in the present study was consistent with those in previous validation studies (0.74 in the SCA2 cohort and 0.75 in the heterogeneous cerebellar disorders cohort)6,9 but higher than the original study (0.59)4. After removing semantic fluency, go/no-go, and DSB, the Cronbach’s alpha dropped to 0.59. The removed items examined language, executive function, and working memory, which can be evaluated by the remaining items, such as phonemic fluency, category switching, and DSF, leading to a shorter version of the Chinese CCAS-s.
One of the important results of the present study is that we adjusted the cut-off scores for the Chinese CCAS-s tests based on the Youden index. The adjusted cut-off scores for semantic fluency, DSF, DSB, and go/no-go were higher than the original cut-off scores. Based on the Youden cut-off scores, our passing score was 81, which was higher than the original passing score of 724. This discrepancy may in part reflect the different methodologies applied to determine the cut-off scores between studies. The original study intended to maximize selectivity of each test while maintain reasonable sensitivity. The present study, by contrast, determined cut-off scores based on the Youden index, a measure to balance sensitivity and selectivity14. The shorter version of the Chinese CCAS-s and these adjusted cut-off scores should be validated in future larger groups.
Our study revealed that age and education were correlated with the performance of CCAS-s, in line with prior studies6,7,15. The investigation into the relations between age, education, and the performance of CCAS-s, however, have yielded mixed results4, most likely due to demographic differences between studies. The mean age of the patients in our cohort was 51.0 years and the mean education was 10.8 years. By contrast, the patients recruited in the original study aged 19.0 to 64.0 years and had at least a high school diploma. Of note, the present study used structural MRI images to rule out participants with cerebral small-vessel disease or neurodegenerative disease (e.g., Alzheimer’s disease) which could lead to mild cognitive impairment. This further confirmed that the controls included in this study were cognitively healthy and the effects of age and education on the performance of CCAS-s were not driven by incipient cognitive decline. Age and education-dependent reference values of this scale, therefore, should be developed to improve its diagnostic properties.
Our study also indicated that the cut-off scores of CCAS-s tests may differ among cerebellar diseases. The passing score (sum of cut-off scores for each item of the scale) was 81 for our cerebellar infarction cohort in contrast to 95 for the German hereditary ataxia cohort7 and 72 for the original validation cohort4. The severity of cognitive deficits following cerebellar damages may in part reflect the distinct patterns of disrupted connectivity between the cerebellum and cerebrum in these cerebellar disorders and their diverse clinical presentations16. In addition, disease progression may provide a partial explanation for the differences in cut-off scores. The average disease duration of patients with cerebellar degeneration in previous studies was approximately 10 years. Our study, by contrast, only recruited cerebellar infarction patients in the acute phase (average 9 days after disease onset). Future studies in different cerebellar disorders at different clinical stages are warranted to establish syndromes-specific cut-off scores.
This study also explored the associations between cognitive and motor functions in cerebellar infarction. Contrary to findings in cerebellar degeneration5,17, the performance of CCAS-s was not associated with the severity of ataxia in the present study, but related to fine motor function. In cerebellar degeneration, motor function may reflect general disease progression and widespread cerebellar atrophy18. Lesions in cerebellar infarction, by contrast, are focal and may only influence certain aspects of cognitive functions. Fine motor difficulties in cerebellar infarction may reflect deficits in attention and executive function, which therefore, correlated with CCAS-s scores.
Furthermore, our study for the first time examined the reliability and validity of CCAS-s and ACE-III and found that both ACE-III and CCAS-s could identify cognitive deficits in patients with cerebellar infarction. Unlike other commonly used cognitive function screening tools with high memory load (e.g., The Montreal Cognitive Assessment scale (MoCA))4,6, ACE-III has been demonstrated as a sensitive screening tool to detect cognitive impairments in patients with cerebellar damages19. These results also suggested the potential of CCAS-s as a generic screening tool to detect cognitive impairment in clinical setting. Validation studies in patients with mild cognitive impairment and dementia will be one of the future directions.
We are well aware that our study has limitations. First, patients’ cognitive performance was evaluated in the acute phase of stroke. Some patients’ symptoms of dizziness and vomiting led to difficulties in cooperating efficiently when performing the tests and therefore affected their evaluation results. Second, only hospitalized patients were included in this study, where patients with mild cognitive deficits at outpatient clinic were missing. Long-term cognitive performance needs to be examined with a larger sample size and covering patients from both inpatient and outpatient departments in future studies.
The Chinese CCAS-s shows high discriminative ability to identify cognitive deficits in patients with cerebellar infarction. Patients with cerebellar infarction were impaired on language, episodic memory, executive function, and affect. The pass/fail cut-off scores of the Chinese CCAS-s tests were adjusted. Age and education-dependent reference values of CCAS-s and validation studies in other neurological disorders will be important future directions to address.