Physiological fluctuations in white matter are increased in Alzheimer's disease and correlate with neuroimaging and cognitive biomarkers

NEUROBIOLOGY OF AGING

卷:37

Article

Makedonov, Ilia^1,2  Chen, J. Jean^2,3,4  Masellis, Mario^5,6  MacIntosh, Bradley J.^2,4,6 

[1] Univ Toronto, Fac Med, Toronto, ON, Canada

[2] Sunnybrook Res Inst, Heart & Stroke Fdn Canadian Partnership Stroke Re, Toronto, ON M4N 3M5, Canada

[3] Univ Toronto, Baycrest Ctr Geriatr Care, Rotman Res Inst, Toronto, ON, Canada

[4] Univ Toronto, Dept Med Biophys, Toronto, ON, Canada

[5] Univ Toronto, Sunnybrook Hlth Sci Ctr, Dept Med, Div Neurol, Toronto, ON, Canada

[6] Sunnybrook Res Inst, Brain Sci Res Program, Toronto, ON M4N 3M5, Canada

关键词: Alzheimer's disease;    Neurodegeneration;    Physiological noise;    Pulsatility;    Resting-state functional MRI;    White matter;   

DOI  :  10.1016/j.neurobiolaging.2015.09.010

来源: Elsevier

【 摘 要 】

The objective of this study was to determine whether physiological fluctuations in white matter (PFWM) on resting-state functional magnetic resonance images could be used as an index of neurodegeneration and Alzheimer's disease (AD). Using resting-state functional magnetic resonance image data from participants in the Alzheimer's Disease Neuroimaging Initiative, PFWM was compared across cohorts: cognitively healthy, mild cognitive impairment, or probable AD. Secondary regression analyses were conducted between PFWM and neuroimaging, cognitive, and cerebrospinal fluid biomarkers. There was an effect of cohort on PFWM (t = 5.08, degree of freedom [df] = 424, p < 5.7 x 10(-7)), after accounting for nuisance effects from head displacement and global signal (t > 6.16). From the neuroimaging data, PFWM was associated with glucose metabolism (t = -2.93, df = 96, p = 0.004) but not ventricular volume (p < 0.49) or hippocampal volume (p > 0.44). From the cognitive data, PFWM was associated with composite memory (t = -3.24, df = 149, p = 0.0015) but not executive function (p > 0.21). PFWM was not associated with cerebrospinal fluid biomarkers. In one final omnibus model to explain PFWM (n = 124), glucose metabolism (p = 0.04) and cohort (p = 0.008) remained significant, as were global and head motion root-mean-square terms, whereas memory was not (p = 0.64). PFWM likely reflects end-arteriole intracranial pulsatility effects that may provide additional diagnostic potential in the context of AD neurodegeneration. (C) 2016 Elsevier Inc. All rights reserved.

【 授权许可】

Free