Previous studies demonstrated cognitive impairments in spinocerebellar ataxia type 3 (SCA3/MJD); however, there is no consensus about the cognitive domains affected and the correlation with structural brain abnormalities. We investigated the neuropsychological profile and 3T-MRI findings, including high-resolution T1-images, diffusion tensor imaging and magnetic resonance spectroscopy of 32 patients with SCA3/MJD and 32 age-, gender- and educational level–matched healthy controls. We reviewed patients’ clinical history and CAG repeat length, and performed assessment and rating of ataxia (SARA)-Brazilian version and the neuropsychiatric inventory. Patients presented worse performance in episodic and working memory and Beck inventories (depression and anxiety). SCA3/MJD patients had a reduction of gray matter volume (GM) in the cerebellum, putamen, cingulum, precentral and parietal lobe. A positive correlation was identified between the cognitive findings and GM of temporal, frontal, parietal, culmen and insula. We observed positive correlation between the brainstem′s fractional anisotropy and digit span-forward. The following cerebellar metabolite groups (measured relative to creatine) were reduced in patients: N-acetyl-aspartate (NAA), NAA + N-acetyl-aspartate-glutamate and glutamate + glutamine (Glx). We found a positive correlation between Corsi’s block-tapping task forward with Glx; semantic verbal fluency with phosphorylcholine and glycerophosphorylcholine; digits span-forward with NAA. The cognitive impairments in SCA3/MJD are associated not only with cerebellar and brainstem abnormalities, but also with neuroimaging evidence of diffuse neuronal and axonal dysfunction, particularly in temporal, frontal, parietal and insular areas.
Spinocerebellar ataxia type 3 Cognitive deficits Voxel-based morphometry Diffusion tensor imaging Magnetic resonance spectroscopy
Supplementary material 1 Statistical maps of positive correlation between gray matter and the cognitive findings; a gray matter (GM) areas correlated with Rey Auditory Verbal Learning Test (RAVLT)-coding; b GM areas correlated with RAVLT-delayed recall; c GM areas correlated with RAVLT-recognition; d GM areas correlated with Raven′s progressive matrices; e GM correlated with Corsi block tapping task; f GM correlated with Digits span- forward; g GM correlated with semantic verbal fluency. (TIFF 527 kb)
Supplementary material 2 Positive correlations areas between the cognitive findings and gray matter areas; a gray matter (GM) areas correlated with Rey Auditory Verbal Learning Test (RAVLT)-coding; b GM areas correlated with RAVLT-delayed recall; c GM areas correlated with RAVLT-recognition; d GM areas correlated with Raven′s progressive matrices; e GM correlated with Corsi block tapping task; f GM correlated with Digits span-forward; g GM correlated with semantic verbal fluency. Color bars represent the t value. (TIFF 342 kb)