Task-induced brain activity patterns in type 2 diabetes: a potential biomarker for cognitive decline

Abstract

Patients with type 2 diabetes demonstrate reduced functional connectivity within the resting state default mode network (DMN), which may signal heightened risk for cognitive decline. In other populations at risk for cognitive decline, additional magnetic resonance imaging abnormalities are evident during task performance, including impaired deactivation of the DMN and reduced activation of task-relevant regions. We investigated whether middle-aged type 2 diabetic patients show these brain activity patterns during encoding and recognition tasks. Compared with control participants, we observed both reduced 1) activation of the dorsolateral prefrontal cortex during encoding and 2) deactivation of the DMN during recognition in type 2 diabetic patients, despite normal cognition. During recognition, activation in several task-relevant regions, including the dorsolateral prefrontal cortex and DMN regions, was positively correlated with HbA1c and insulin resistance, suggesting that these important markers of glucose metabolism impact the brain's response to a cognitive challenge. Plasma glucose ≥11 mmol/L was associated with impaired deactivation of the DMN, suggesting that acute hyperglycemia contributes to brain abnormalities. Since elderly type 2 diabetic patients often demonstrate cognitive impairments, it is possible that these task-induced brain activity patterns observed in middle age may signal impending cognitive decline.

© 2014 by the American Diabetes Association. Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered.

Figures

Figure 1

Regions activated (red) and deactivated (blue) during the encoding (A) and recognition (B) tasks in control subjects (CO) and type 2 diabetic patients (T2DM). Group differences between control subjects and diabetic patients and between diabetic patients with PG ≥11 mmol/L and <11 mmol/L. Age was controlled for in these analyses. See Table 2 for region lists. Statistical parametric brain activation maps within and between groups were computed using one-sample t tests for within-group and unpaired two-sample t tests for between-group analyses with a significance threshold of P < 0.05 (corrected). A z statistic threshold value of 2.3 was applied using the standard cluster-based thresholding Gaussian Random Field theory for inference provided by FSL (25).