Microstructural white matter abnormalities in type 2 diabetes mellitus: A diffusion tensor imaging study
Introduction
Diabetes mellitus (DM) is a common metabolic disease characterized by hyperglycemia due to the insufficient availability of, or insensitivity to, insulin. Diabetes is typically associated with structural brain abnormalities and an increased risk for stroke (Folsom et al., 1999), lacunar infarctions (Kobayashi et al., 1997), silent stroke events (Eguchi et al., 2003), and gray matter atrophy (Manschot et al., 2006, Musen et al., 2006). While the relationship between white matter (WM) hyperintensities and diabetes is still controversial, a recent systematic review on diabetes and brain imaging concluded that there is convincing evidence for an association between diabetes and cerebral atrophy and lacunar infarctions (van Harten et al., 2006). It was stated, however, that it is still unclear whether or not type 2 diabetes mellitus (T2DM) is associated with WM hyperintensities as observed with magnetic resonance imaging (MRI). With nine of the 27 WM lesion studies included in their meta-analysis, no association between diabetes and WM hyperintensities was observed in the “vascular cohorts”. By contrast, for the “outpatient cohorts”, there appeared to be a modest association between diabetes and WM hyperintensities. This uncertainty could be partly due to the insufficient sensitivity of conventional MRI modalities (e.g., T1 and T2 maps) to detect subtle brain WM changes or to assess the severity of WM hyperintensities. In the review paper of van Harten et al., 2006, it was also suggested that diffusion tensor imaging (DTI) could be a promising technique for the investigation of WM properties in diabetes patients.
DTI is a unique technique for assessing WM structural properties based on the three-dimensional anisotropic Gaussian diffusion of water molecules (Basser et al., 1994). With DTI the directionality and magnitude of random water movement in tissue can be estimated yielding several quantitative measures, such as the three principle diffusivities (i.e., the eigenvalues of the diffusion tensor: λ1 > λ2 > λ3), mean diffusivity (MD = [λ1 + λ2 + λ3] / 3), transverse diffusivity (TD = [λ2 + λ3] / 2), axial diffusivity (AD = λ1), and the degree of diffusion anisotropy (e.g., the fractional anisotropy: FA) (Pierpaoli and Basser, 1996). Without barriers, water molecules move uniformly in all directions, which results in isotropic diffusion. By contrast, in the presence of barriers, such as cell membranes, nerve fibers, or myelin sheets, the diffusion rate is typically larger in one direction than in another, which is then referred to as anisotropic diffusion (Beaulieu, 2002). Being quantitative in nature, these DTI based measures have been shown to be more sensitive to tissue abnormalities than the typical visual evaluation of WM hyperintensities observed in conventional MRI data (Della Nave et al., 2007, Van Hecke et al., 2008a). To date, DTI studies have revealed WM alterations through measurements of decreased FA and/or increased MD in a variety of conditions, including aging (Hsu et al., 2008, Sullivan and Pfefferbaum, 2007, Van Hecke et al., 2008a), multiple sclerosis (Patel et al., 2007), schizophrenia (e.g., Carpenter et al., 2008), traumatic brain injury (e.g., Caeyenberghs et al., 2010a, Caeyenberghs et al., 2010b), amyotrophic lateral sclerosis (e.g., Sage et al., 2009) and Alzheimer's disease (e.g., Stahl et al., 2007). For an in-depth discussion of DTI, the interested reader is referred to a recent review of Tournier et al. (2011).
Only a few studies have been published that use DTI to investigate WM properties in DM patients. Kodl et al. (2008) demonstrated that DTI can detect microstructural WM abnormalities in subjects with long-standing type 1 DM and they showed that poor performance on selected neurocognitive tests correlated with reduced WM FA. More recently, another study examined the emotional and declarative memory impairment and associated WM microstructural properties in 24 middle-aged and elderly patients with T2DM but without any obvious vascular pathology or psychiatric disorder (Yau et al., 2009). Their findings suggest that T2DM patients show diffuse and predominantly frontal and temporal WM microstructural abnormalities, with extensive involvement of the temporal stem. Also the FA of the temporal stem was found to be associated with immediate memory performance.
In this work, we studied the WM microstructural organization in a well-controlled and large (n = 40) cohort of non-hypertensive T2DM patients who do not exhibit significant cognitive deficits. In particular, we investigated whether DTI is able to detect differences in diffusion properties (i.e., FA, MD, TD, and AD) between these patients and healthy controls even when the conventional MRI data of the T2DM patients do not show any significant abnormalities. In addition, we performed a battery of medical tests to examine potential relationships between the DTI indices and clinical parameters. Our results demonstrate that, compared with age-matched non-diabetes controls, T2DM patients show several regions with abnormal diffusion values reflecting DM related changes in microstructural tissue organization. To the best of our knowledge, this is the largest DTI based T2DM study to date (40 patients vs. 97 controls) and the first one that also investigates both regional and global diffusivity measures, including AD and TD, providing a more detailed picture of the underlying microstructural tissue organization in T2DM patients.
Section snippets
Subjects
Forty-two middle-aged T2DM patients (27M/15F) and 100 non-diabetic non-hypertensive controls (54M/46F) were examined by a collaborating endocrinologist based on the guideline criteria of “The American Diabetes Association: clinical practice recommendations 1996”. The patients were selected from our community-based prospective cohort study, which investigates the cardio- and cerebro-vascular risk factors in the general population, or were recruited from a general health screening program. The
Descriptive statistics of clinical measures
Clinical and demographic characteristics of the T2DM patients and the age-matched healthy controls are shown in Table 2. There is no significant difference in age, gender, total cholesterol, LDL, and total ARWMC scores between both groups (assessed with two-tailed independent sample t-tests and Chi-square tests). However, diabetic patients have higher levels of BMI, systolic and diastolic BP, fasting glucose, HbA1c, and triglyceride; and lower HDL levels than controls. As shown in Table 3, for
Discussion
In this investigation, we demonstrated significant differences in the microstructure of brain tissue, as determined by global and voxel based analyses of FA, MD, RD, and AD measures, between T2DM patients and healthy controls. More specifically, we observed a regionally decreased FA in T2DM patients due to a significantly increased TD (with no significant decrease in AD). In addition, for both the global and regional analysis, the observed increase in MD for T2DM patients compared to healthy
Author contributions
Jung-Lung Hsu: wrote manuscript, designed analysis, investigated data, performed analyses.
Yen-Ling Chen: investigated data.
Chyi-Huey Bai: contributed to discussion.
Fu-Shan Jaw: contributed to discussion.
Cheng-Hui Lee: investigated data.
Yuh-Feng Tsai: investigated data.
Chien-Yeh Hsu: contributed to discussion.
Jyu-Gang Leu: investigated data.
Alexander Leemans: contributed to discussion, designed analysis, revised manuscript.
Acknowledgments
This work was sponsored by the Shin Kong Wu Ho-Su Memorial Hospital (SKH-8302-98-DR-19) and the project Care4Me (Cooperative Advanced REsearch for Medical Efficiency) in the framework of the EU research program ITEA (Information Technology for European Advancement).
References (57)
- et al.
Voxel-based morphometry — the methods
Neuroimage
(2000) - et al.
Unified segmentation
Neuroimage
(2005) - et al.
MR diffusion tensor spectroscopy and imaging
Biophys. J.
(1994) - et al.
Quantifying the effects of normal ageing on white matter structure using unsupervised tract shape modelling
Neuroimage
(2010) - et al.
Risk of dementia in diabetes mellitus: a systematic review
Lancet Neurol.
(2006) - et al.
Multiple indicators of age-related differences in cerebral white matter and the modifying effects of hypertension
Neuroimage
(2010) - et al.
Brain–behavior relationships in young traumatic brain injury patients: fractional anisotropy measures are highly correlated with dynamic visuomotor tracking performance
Neuropsychologia
(2010) - et al.
Cardiovascular autonomic neuropathy, autonomic symptoms and diabetic complications in 674 type 2 diabetes
Diabetes Res. Clin. Pract.
(2008) - et al.
Gender differences and age-related white matter changes of the human brain: a diffusion tensor imaging study
Neuroimage
(2008) - et al.
Microstructural white matter changes in normal aging: a diffusion tensor imaging study with higher-order polynomial regression models
Neuroimage
(2010)