Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Multicenter Study
. 2024 Dec 21;272(1):83.
doi: 10.1007/s00415-024-12831-1.

Subtypes of cognitive impairment in cerebellar disease identified by cross-diagnostic cluster-analysis: results from a German multicenter study

Qi Liu  1   2 , Kerstin Rubarth  3 , Jennifer Faber  4   5 , Patricia Sulzer  6   7 , Imis Dogan  8   9 , Miriam Barkhoff  4 , Martina Minnerop  10   11   12 , Adam M Berlijn  12   13 , Saskia Elben  10   11 , Heike Jacobi  14 , Julia-Elisabeth Aktories  14 , Dana M Huvermann  1   2   13 , Friedrich Erdlenbruch  1   2 , Raquel Van der Veen  1   2 , Johanna Müller  1   2 , Enzo Nio  1   2 , Benedikt Frank  1   2 , Martin Köhrmann  1   2 , Elke Wondzinski  15 , Mario Siebler  15 , Kathrin Reetz  8   9 , Jürgen Konczak  16 , Frank Konietschke  3   17 , Thomas Klockgether  4 , Matthis Synofzik  6   7 , Sandra Röske  4 , Dagmar Timmann  1   2 , Andreas Thieme  18   19
Affiliations
Multicenter Study

Subtypes of cognitive impairment in cerebellar disease identified by cross-diagnostic cluster-analysis: results from a German multicenter study

Qi Liu et al. J Neurol. .

Abstract

Background: Cognitive and neuropsychiatric impairment, known as cerebellar cognitive affective syndrome (CCAS), may be present in cerebellar disorders. This study identified distinct CCAS subtypes in cerebellar patients using cluster analysis.

Methods: The German CCAS-Scale (G-CCAS-S), a brief screening test for CCAS, was assessed in 205 cerebellar patients and 200 healthy controls. K-means cluster analysis was applied to G-CCAS-S data to identify cognitive clusters in patients. Demographic and clinical variables were used to characterize the clusters. Multiple linear regression quantified their relative contribution to cognitive performance. The ability of the G-CCAS-S to correctly distinguish between patients and controls was compared across the clusters.

Results: Two clusters explained the variance of cognitive performance in patients' best. Cluster 1 (30%) exhibited severe impairment. Cluster 2 (70%) displayed milder dysfunction and overlapped substantially with that of healthy controls. Cluster 1 patients were on average older, less educated, showed more severe ataxia and more extracerebellar involvement than cluster 2 patients. The cluster assignment predicted cognitive performance even after adjusting for all other covariates. The G-CCAS-S demonstrated good discriminative ability for cluster 1, but not for cluster 2.

Conclusions: The variance of cognitive impairment in cerebellar disorders is best explained by one severely affected and one mildly affected cluster. Cognitive performance is not only predicted by demographic/clinical characteristics, but also by cluster assignment itself. This indicates that factors that have not been captured in this study likely have effects on cognitive cerebellar functions. Moreover, the CCAS-S appears to have a relative weakness in identifying patients with only mild cognitive deficits.

Study registration: The study has prospectively been registered at the German Clinical Study Register ( https://www.drks.de ; DRKS-ID: DRKS00016854).

Keywords: Cerebellar cognitive affective syndrome (CCAS); Cerebellar disorders; Cluster analysis; German CCAS-Scale; Subgroups of CCAS.

PubMed Disclaimer

Conflict of interest statement

Declarations. Conflicts of interest: Matthis Synofzik has received consultancy honoraria from Ionis, UCB, Prevail, Orphazyme, Biogen, Servier, Reata, GenOrph, AviadoBio, Biohaven, Zevra, Lilly, and Solaxa, all unrelated to the present manuscript. All other authors declare no financial disclosures or conflicts of interest other than the funding stated concerning the research covered in this manuscript. Ethical approval: This study was approved by the local ethics committees of the participating sites and performed in accordance with the ethical standards laid down in the Declaration of Helsinki of 1964 and its later amendments. Informed consent: Participants’ written informed consent was obtained prior to any study related procedures.

Figures

Fig. 1
Fig. 1
K-mean cluster elbow graph. The abscissa (x = 2, red line) of the inflection point in the K-means cluster elbow graph indicates the most appropriate number (2) of clusters
Fig. 2
Fig. 2
Relation of clusters to each other and to healthy controls. a The two distinct clusters of patients are shown in comparison to each other. Patients with mild cognitive impairment are shown on the left (cluster 2, red), patients with severe cognitive impairment are shown on the right (cluster 1, blue). b The two clusters of patients are shown in comparison to the cluster of healthy controls (yellow). A substantial overlap between patients in cluster 2 and healthy controls is visible. The results from the principal component analysis revealed that principal component 1 (dimension 1, x axis) accounts for 50% of the variance in the original data, while principal component 2 (dimension 2, y axis) accounts for 18% of the variance in the original data
Fig. 3
Fig. 3
Z-scores of cognitive/neuropsychiatric domains across clusters. The figure shows the mean z-score for each cognitive and neuropsychiatric function for the total group and for the two clusters of cerebellar patients
Fig. 4
Fig. 4
Percentage of patients impaired in each cognitive and the neuropsychiatric domain across clusters. The absolute number and the percentage (in parentheses) of patients with deficits in the four core domains of CCAS as well as the memory domain (which is not part of CCAS, but rather indicates cerebral involvement) is shown. A patient was considered impaired on a certain domain if his/her z-score was one standard deviation or more below the healthy control group’s average
Fig. 5
Fig. 5
Ability of the G-CCAS-S to differentiate patients within each cluster from age- and education-matched healthy controls. ROC curves (red line) of G-CCAS-S are shown for the total sum raw score of cluster 1 (a) and cluster 2 (c) as well as for the number failed items of cluster 1 (b) and cluster 2 (d). The 95% confidence interval for each ROC curve is also displayed in the figure (depicted by blue shadow). For each measure, the Youden Index (YI, black dot) is shown and the values for selectivity (that is: the portion of controls correctly identified as controls/all controls in the respective control group) and sensitivity (that is: the portion of patients correctly identified as patients/all patients in the cluster) for the respective YI are given in parentheses. G-CCAS-S German version of Cerebellar Cognitive Affective Syndrome Scale, ROC receiver operating characteristic, YI Youden Index, AUC area under curve
See this image and copyright information in PMC

References

    1. Ahmadian N, van Baarsen K, van Zandvoort M, Robe P (2019) The cerebellar cognitive affective syndrome—a meta-analysis. Cerebellum (London, England) 18:941–950. 10.1007/s12311-019-01060-2 - DOI - PMC - PubMed
    1. Aita SL, Beach JD, Taylor SE, Borgogna NC, Harrell MN, Hill BD (2019) Executive, language, or both? An examination of the construct validity of verbal fluency measures. Appl Neuropsychol Adult 26:441–451. 10.1590/s1980-57642011dn0501000610.1080/23279095.2018.1439830 - DOI - PubMed
    1. Baron-Cohen S, Wheelwright S, Hill J, Raste Y, Plumb I (2001) The “Reading the Mind in the Eyes” test revised version: a study with normal adults, and adults with Asperger syndrome or high-functioning autism. J Child Psychol Psychiatry 42:241–251. 10.1111/1469-7610.00715 - DOI - PubMed
    1. Beer JS, Ochsner KN (2006) Social cognition: a multi level analysis. Brain Res 1079:98–105. 10.1016/j.brainres.2006.01.002 - DOI - PubMed
    1. Benassi M, Garofalo S, Ambrosini F, Sant’Angelo RP, Raggini R, De Paoli G, Ravani C, Giovagnoli S, Orsoni M, Piraccini G (2020) Using two-step cluster analysis and latent class cluster analysis to classify the cognitive heterogeneity of cross-diagnostic psychiatric inpatients. Front Psychol 11:1085. 10.3389/fpsyg.2020.01085 - DOI - PMC - PubMed

Publication types

Associated data

LinkOut - more resources