Zayed University Physiological Wellness Dataset 2025\u201d (ZU-PWD \u201925)

Background: Wearable devices have become essential tools for capturing continuous physiological and behavioral signals in real-world environments. However, publicly available datasets that combine multimodal, high-resolution physiological signals with behavioral measures in adults remain scarce, limiting opportunities for reproducible and ecologically valid research.Objective: This study introduces a longitudinal, dataset collected using Empatica Embrace Plus wrist-worn devices, designed to support research in digital health, behavioral science, and machine learning based sensor fusion.Methods: Data were collected from 28 participants over a period of up to 30 consecutive days each, capturing synchronized measures of heart rate, heart rate variability (RMSSD), electrodermal activity, peripheral skin temperature, respiratory rate, tri-axial acceleration, posture classification, step count, metabolic equivalent of task (MET), and sleep stages. The technical validation assessed data completeness, physiological plausibility, and cross-sensor consistency through missing values analysis, range checks, circadian pattern evaluation, and correlation profiling.Results: The technical validation confirmed that physiological signals fell within expected biological ranges for adults, behavioral measures reflected realistic daily activity rhythms, and cross-sensor associations were consistent with established physiological relationships. The missing values varied by modality; activity and metadata fields had higher coverage, while respiration had larger gaps. These results establish the reliability of dataset for diverse applications.Potential Applications: The dataset enables research in digital phenotyping, behavioral health monitoring, physical activity recognition, sleep and circadian rhythm analysis, chronic disease risk modeling, and machine learning driven multimodal data fusion. Its ecological validity and temporal depth make it suitable for both laboratory benchmarking and deployment-oriented modeling.Conclusions: By releasing this dataset with full documentation and validation, we intend to advance open, collaborative, and reproducible research at the intersection of wearable sensing, health analytics, and human behavior modeling.