Unlocking the silent signals: Motor kinematics as a new frontier in early detection of mild cognitive impairment

Comprehensive Summary

The paper explores whether motor kinematics, or how a person walks or writes, can serve as an objective tool for detecting mild cognitive impairment (MCI) earlier than traditional cognitive tests. This is an editorial piece reviewing a study that used a combination of motor kinematics for walking and writing, then used a computer algorithm to determine how well it could distinguish people with MCI from those without it. The key takeaway of the study is that combining gait + “graphic” handwriting features achieved 74% accuracy (AUC ≈ 0.74) while handwriting alone achieved a comparable yet lower accuracy. The author of the piece suggests that using motor kinematics may be more “real-world” than other biomarkers because it reflects changes in how a person performs in daily activities, rather than underlying pathology. In reviewing the study, the author also notes areas for further research, such as the study only being a snapshot in time and not demonstrating longitudinal prediction, a small sample of 95 participants from a specific geographic area, and not fully using dual task testing of motor kinematics (how well a person can walk or write while simultaneously doing another cognitive task).

Outcomes and Implications

From a clinical perspective, the big draw is that gait and handwriting tests may be easier to use in a primary care, community, or elder care setting than imaging or fluid-based biomarkers, and may help reduce cultural/language biases in certain cognitive tests. If validated, it may help increase screening and allow for early intervention for risk reduction, monitoring, and recruitment of prevention clinical trials; essentially, catching the decline before it’s too late. Of course, the paper is also clear that we would need to validate it on a larger, more diverse population to show kinematic signals can predict MCI or dementia development, but a realistic time frame is “near-to-mid future” rather than immediate. It is more like something we could use in the future after validation, possibly even moving toward wearable, passive monitoring (e.g., smartphones, smartwatches) once accuracy and interpretability are high enough for clinical use.