Differentiating dementia disease subtypes with gait analysis: feasibility of wearable sensors?

There are unique signatures of gait impairments in different dementia disease subtypes, such as Alzheimer's disease (AD), dementia with Lewy bodies (DLB) and Parkinson's disease (PDD). This suggests gait analysis is a useful differential marker for dementia disease subtypes, but this has yet to be assessed using inexpensive wearable technology.

This study aimed to assess whether a single accelerometer-based wearable could differentiate dementia disease subtypes through gait analysis.

80 people with mild cognitive impairment or dementia due to AD, DLB or PD performed six ten-metre walks. An accelerometer-based wearable (Axivity) assessed gait. Data was processed using algorithms validated in other neurological disorders and older adults. Fourteen spatiotemporal characteristic were computed, that broadly represent pace, variability, rhythm, asymmetry and postural control features of gait. One way analysis of variance and Kruskall Wallis tests identified significant between-group differences, and post-hoc independent t-tests and Mann Whitney U's established where differences lay. Receiver Operating Characteristics and Area Under the Curve (AUC) demonstrated overall accuracy for single gait characteristics.

The wearable was able to differentiate dementia disease subtypes (p ≤ .05) and demonstrated significant differences between the groups in 7 gait characteristics with modest accuracy. For reference the instrumented walkway showed 2 between-group differences in gait characteristics.

This study found that a wearable device can be used to differentiate dementia disease subtypes. This provides a foundation for future research to investigate the application of wearable technology as a clinical tool to aid diagnostic accuracy, allowing the correct treatment and care to be applied. Wearable technology may be particularly useful as its use is less restricted to context, making it easier to implement.