Development of a CT-based weighted rating scale for subcortical cerebrovascular disease sensitive to mild clinical symptoms

Research output: Contribution to journalArticle

Abstract

Objective: To devise and validate a CT-based visual rating scale sensitive to subcortical cerebrovascular disease (sCVD) in patients with cognitive impairment. Methods: Three types of vascular lesions were rated separately in 16 brain regions. A unique vascular score was computed weighing the regression coefficients of the three scores in linear models. Thereafter, subcortical vascular classes (SVC) of increasing severity (0-3) were created. Known-group and convergent validity of the SVC was tested in 122 cognitively impaired patients with MMSE of 18 and higher. These patients were grouped according to clinical diagnosis in degenerative, mixed, and vascular. Results: Intraclass correlation coefficients were between 0.86 and 0.94. The subcortical vascular score was 8.0±9.4 in the degenerative, 33.0±17.2 in the mixed, and 36.8±14.8 in the vascular patients (p=0.0001). In particular, none of the degenerative patients had SVC=3, and only 14% had SVC=2. On the contrary, none of the mixed and vascular patients had SVC=0, while about 50% had SVC=3 in both groups. The prevalence of hypertension was increasing with increasing SVC (p for trend=0.02). The performance on tasks evaluating balance, gait, and bradykinesia decreased with increasing SVC severity (p for trend=0.006, 0.01, 0.001, respectively). Conclusion: The scale is a valid tool to estimate the vascular component in patients with cognitive impairment.

Original languageEnglish
Pages (from-to)241-245
Number of pages5
JournalJournal of the Neurological Sciences
Volume203-204
DOIs
Publication statusPublished - Nov 15 2002

    Fingerprint

Keywords

  • Cerebrovascular disease
  • Dementia
  • Rating scale
  • Vascular cognitive impairment
  • White matter

ASJC Scopus subject areas

  • Ageing
  • Clinical Neurology
  • Surgery
  • Developmental Neuroscience
  • Neurology
  • Neuroscience(all)

Cite this