The Analysis of Association Between Imaging Biomarkers of Cerebral Small Vessel Disease and Cognitive Impairment: a Shanghai Elderly Community⁃based Cohort

YING Yun-qing,WANG Yi-qing,XIA Yi-wei,WU Dan-hong,WU Wei-wen,CHENG Xin,DONG Qiang

DOI: https://doi.org/10.3969/j.issn.1672⁃6731.2021.10.004

2021-01-01

Abstract:Objective To investigate the associations between imaging markers of cerebral small vessel disease (CSVD) and global cognition and cognitive domains in the community elderly people. Methods From August 2016 to October 2019, 933 healthy community elderly subjects were enrolled from three districts in Shanghai. We used MRI to assess imaging markers of CSVD, including lacunar infarct (LACI), white matter hyperintensity (WMH), cerebral microbleeds (CMBs), enlarged perivascular space (EPVS) and total CSVD score. These subjects underwent Montreal Cognitive Assessment (MoCA) to assess global cognition, Auditory Verbal Learning Test (AVLT) to assess memory function, Trial Making Test (TMT) to assess executive function, Stick Test (ST) to assess visuospatial function, adapted Common Objects Sorting Test (COST) to assess language function. We performed Spearman rank correlation and partial correlation analyses to investigate the association between imaging markers and cognition, and multiple linear regression analysis to validate the relationships. Results A total of 863 subjects underwent head MRI and cognitive assessment. In Spearman rank correlation analysis, LACI (rs=⁃0.076, P=0.025) and periventricular WMH (rs=⁃0.070, P=0.044) was correlated with MoCA. LACI (rs=⁃0.086, P=0.022), basal ganglia EPVS (rs=⁃0.078, P=0.038) and total CSVD score (rs=⁃0.090, P=0.023) were correlated with AVLT. LACI was correlated with TMT (rs=0.099, P=0.004). Periventricular WMH (rs=⁃0.113, P=0.029) and strictly lobar CMB (rs=⁃0.107, P=0.041) were correlated with ST. Centrum semiovale EPVS was correlated with COST (rs=⁃0.150, P=0.004). In partial correlation analysis, LACI was associated with TMT (r=0.072, P=0.038). Deep WMH was associated with ST (r=0.105, P=0.047). Periventricular WMH (r=⁃0.111, P=0.001), CMBs (r=⁃0.088, P=0.015) and total CSVD score (r=⁃0.087, P=0.015) were associated with MoCA. In multiple linear regression analysis, after adjusting sex, age, education, hypertension, diabetes mellitus, hyperlipidemia and smoking history factors, periventricular WMH (βstandardized=⁃0.088, PFDR=0.020) and CMBs at any position (βstandardized=⁃0.078, PFDR=0.040) were independently correlated with MoCA. Conclusions In the elderly community population in China, CSVD may already have an impact on cognition. Periventricular WMH and CMBs are early imaging markers for CSVD⁃related global cognitive impairment.