Relationships among cortical activation, cognition, and blood biomarkers in two types of dementia determined using functional near-infrared spectroscopy

Nairong Ruan ¹ Ming Liang ¹ Yuehong Liu ¹ Xi Mei ^1,2* Chengying Zheng ^1*

• ¹ Affiliated Kangning Hospital of Ningbo University, Ningbo, Zhejiang Province, China
• ² Ningbo Kangning Hospital, Ningbo, China

Objective: The most prevalent types of dementia in older adults are Alzheimer disease (AD) and Lewy body dementia (LBD), and they have overlapping clinical symptoms. We aimed to define amounts of cortical activation and to identify indicators of brain function to better distinguish between these types of dementia and aid diagnosis using functional near-infrared spectroscopy (fNIRS). Methods: Oxygenated hemoglobin (HbO) concentrations in the brains of patients with AD and LBD were detected using fNIRS. Brain function was assessed using a verbal fluency task (VFT). Resting-state and task-state cortical activations were investigated to determine differences between AD and LBD. Blood samples were analyzed to identify relevant biomarkers. The clinical and HbO variables were compared between AD and LBD. Functional connectivity at rest and correlations between HbO variables and blood biomarkers were analyzed. The sensitivity and specificity of the parameters for differentiating the dementias were evaluated using areas under the receiver operating characteristic (ROC) curves (AUCs). Results: This study recruited 28 inpatients with AD and 25 with LBD. Mean HbO concentrations did not significantly differ in the resting state (p > 0.05), whereas functional connectivity significantly differed between the groups. Mean HbO concentrations during the VFT, were significantly lower in the left temporal, right dorsolateral prefrontal, and right temporal (p = 0.011) cortices of the AD, than the LBD group. Blood amyloid-β (Aβ) 42 levels were significantly higher in the AD group (p = 0.023), whereas significantly more α-synuclein was expressed in the LBD group (p = 0.012). Correlation analysis of cognition-related blood biomarkers with HbO concentrations associated higher plasma Aβ 42 level with lower HbO concentrations in the right pre-motor and supplementary motor cortex and higher glial fibrillary acidic protein (GFAP) levels in the lower right pars triangularis at rest. Levels of the blood biomarker Aβ significantly and negatively correlated with HbO concentrations in the right temporal cortex during the VFT. The AUC was significantly higher for the combination of multiple fNIRS indicators compared with individual cognitive or blood indicators (AUC = 0.9314). Conclusion: The characteristics of HbO measured using fNIRS can help distinguish AD from LBD in older adults.