A recent study, powered by centralive, in JMIR Formative Research, explores how smartwatches and machine learning can predict sleep quality for dementia caregivers from diverse populations. With data from 529 days of caregiving activity, the study highlights key insights:
- Impactful Predictors: Factors such as wake-up stress, wake-up heart rate, sedentary seconds, total distance traveled, and sleep duration were strongly correlated with sleep quality.
- Machine Learning Models: Among three algorithms—Random Forest, K-Nearest Neighbor (KNN), and XGBoost—the Random Forest classifier achieved the highest accuracy, with an AUC of 0.86 and an F1 score of 0.87.
- Underrepresented Populations: The study specifically targeted caregivers from diverse ethnic backgrounds, addressing barriers such as cultural adaptation and limited emotional support.
At Centralive, we empower researchers to leverage wearable data and predictive models, providing tools for better health insights and personalized care. Together, we’re creating scalable solutions for mental health and caregiver support.
Authors: Park, Jung In PhD, RN, FAMIA; Aqajari, Seyed Amir Hossein; Rahmani, Amir M. PhD; Lee, Jung-Ah PhD, RN, FGSA, FAAN
