Wearable Sensor-Based Daily Life Walking Assessment of Gait for Distinguishing Individuals With Amnestic Mild Cognitive Impairment

Objectives: To characterize gait disorders in patients with amnestic mild cognitive impairments and determine the association between the performance of the gait function and cognition. Methodology: In this study, we enrolled 38 patients with amnestic mild cognitive impairment (aMCI) and 30 cognitively normal individuals (NC). Neuropsychological assessments included tests of memory, executive function, language, and attention. Using an inertial-sensor-based wearable instrument, we collected the gait data dynamically for at least 1 h/day for 2 weeks. The gait parameters included walking velocity, stride length, stride time, cadence, and stride time variability. Results: The aMCI patients had reduced walking velocity and stride length and increased stride time variability compared with the NCs. Walking velocity was positively correlated with long-term memory (r = 0.256, p = 0.043) and executive function (r = 0.490, p = 0.007). Stride length was positively correlated with long-term memory (r = 0.322, p = 0.036), executive function (r = 0.366, p = 0.024), and attention (r = 0.264, p = 0.048). Stride time variability was negatively correlated with long-term memory (r = −0.206, p = 0.038), executive function (r = −0.430, p = 0.022), and attention (r = −0.247, p = 0.015). Conclusion: This study suggested that cognitive impairment-related gait disorders occur (reduced gait speed, gait length, and gait stability) in daily life walking among the aMCI patients. A sensor-based wearable device for gait measurement may be an alternative and convenient tool for screening cognitive impairment.