Functional MRI Signal Fluctuations: A Preclinical Biomarker for Cognitive Impairment in Type 2 Diabetes?


Type 2 diabetes mellitus (T2DM) is associated with macrovascular and microvascular complications as well as an increased risk for cognitive impairment (CI), ranging from mild memory impairment to fully developed dementia (1). Both vascular dementia (VD) and Alzheimer disease (AD) dementia are increased in patients with T2DM. However, it is not clear if T2DM directly causes CI and dementia or if the two disorders share a common pathophysiological process (2). Insulin resistance, hyperglycemia, and hypoglycemia have all been implicated as risk factors for CI in T2DM (3). However, their relative importance remains uncertain. The possibility that altered brain insulin action plays a role is consistent with human data showing that AD is associated with decreased insulin receptor expression in the brain (4), impaired insulin signaling (5), and decreased insulin levels in the cerebrospinal fluid (CSF) (6). Furthermore, insulin delivery directly into the hippocampus enhances spatial memory in nondiabetic rats, whereas this response to intrahippocampal insulin is impaired in diet-induced obese rats (7). In small clinical trials, intranasal insulin delivery appears to improve memory function in patients with CI and AD (8). Functional MRI, based on the blood oxygenation level–dependent (BOLD) contrast mechanism, has become an important tool to investigate the brain’s neurophysiologic response to specific stimuli or cognitive tasks. In contrast, resting-state functional MRI (rs-fMRI) measures spontaneous low-frequency oscillations in the BOLD signal (9). This spontaneous brain activity at rest is altered in VD (10) and AD (11) patients, and to a lesser extent in people with CI (11); however, few studies have examined the resting-state brain activity in T2DM (12,13). Musen et al. (13) compared rs-fMRI in T2DM and age-matched diabetic control subjects with no structural brain abnormalities or CI. Using a seed approach to measure brain functional connectivity during the resting state, the authors demonstrated that T2DM patients had decreased brain connectivity, which was associated with homeostasis model assessment of insulin resistance (HOMA-IR), a measure of insulin resistance. Xia et al. (12) found similar results using the rs-fMRI amplitude of low-frequency fluctuation (ALFF) approach. The ALFF approach avoids the need to select a seed region and allows assessment of the whole brain by measuring signal fluctuations over time at the voxel level (14). The hope is that such rs-fMRI signals might be a biomarker for identifying T2DM patients at risk for CI before structural brain abnormalities and/or CI can be identified. Although it is not known if CI is caused by abnormal oscillations in functional activity or if the oscillations are merely a CI biomarker, if such a tool for early CI risk detection was available, it could eventually be used for preventive dementia interventions. In this issue, Cui et al. (15) examine the rs-fMRI spontaneous low-frequency signal oscillations and their local smoothness in T2DM patients compared with an age-, sex-, and education-matched control group. The authors used the ALFF approach defined above and also the regional homogeneity (ReHo) measure that assesses the ReHo of these oscillatory signals in a small window around each voxel (14). There were no structural brain abnormalities and both T2DM and control subjects showed higher ALFF and ReHo values in the posterior cingulated cortex and the precuneus and medial prefrontal cortex compared with global mean values. However, T2DM subjects exhibited decreased resting brain fluctuations in the occipital lobe, calcarine cortex,

DOI: 10.2337/db13-1685

Extracted Key Phrases

1 Figure or Table

Cite this paper

@inproceedings{BelfortDeAguiar2014FunctionalMS, title={Functional MRI Signal Fluctuations: A Preclinical Biomarker for Cognitive Impairment in Type 2 Diabetes?}, author={Renata Belfort-DeAguiar and R. Todd Constable and Robert S . Sherwin}, booktitle={Diabetes}, year={2014} }