Development and External Validation of a Gait Test Based Diagnostic Model for Detecting Mild Cognitive Impairment

Area of Science:

• Gerontology
• Neurology
• Biomedical Engineering

Background:

• Mild cognitive impairment (MCI) affects a significant portion of the aging population.
• Current screening tools for MCI lack large-scale applicability and efficiency.
• Gait analysis offers a promising, non-invasive method for cognitive assessment.

Purpose of the Study:

• To evaluate the discriminatory power of various gait tests for identifying MCI.
• To develop and validate a novel screening tool for MCI based on gait parameters.
• To address the need for accessible, large-scale MCI screening solutions.

Main Methods:

• A case-control study involving 134 older adults (≥65 years) for model derivation and 70 for validation.
• Participants underwent single-task (ST) and dual-task (DT) gait tests, including cognitive tasks like serial subtractions.
• Statistical analyses involved binary logistic regression, ROC curves, and calibration curves for model development and validation.

Main Results:

• The DT gait test with serial subtractions demonstrated high accuracy (AUC=0.906) in detecting MCI.
• The model showed consistent performance across different demographic groups and robust external validation (AUC=0.913).
• Adjusting for years of education further enhanced the model's predictive capability.

Conclusions:

• Dual-task gait testing, specifically with serial subtractions, is a highly effective method for discriminating MCI.
• This gait-based approach has the potential to become a valuable large-scale screening tool for MCI.
• Early identification of MCI through gait analysis can facilitate timely interventions for cognitive health in older adults.