Multimodal Wearable Biosensing for Early Physiological Stress Detection: A Simulated Digital- Health Validation Study

This article presents a simulated empirical study in biomedical engineering and digital health. The synthetic dataset contained 725,760 synchronized five-minute windows after quality screening. A simulated 21-day wearable dataset was generated for 120 adult profiles using heart-rate variability, skin temperature, electrodermal activity, step intensity, sleep duration, and sweat-conductivity proxies. A multimodal stress score was compared with heart-rate- only and activity-only baselines. Signals were sampled at different frequencies and synchronized into five-minute windows. Edge inference latency, missingness, calibration drift, and battery consumption were also modeled. The multimodal model achieved the highest simulated macro-F1 (0.91) and reduced false stress alerts during exercise compared with heart-rate-only detection. Sensor dropout degraded performance most strongly when electrodermal activity and sleep history were simultaneously missing. Edge inference remained below 150 ms per window under the selected lightweight model. All numerical values and outcomes are synthetic and are used only to demonstrate a plausible journal-style research design. No real participants, companies, cities, laboratories, utilities, or field sites were studied. The article combines a distinct analytical structure with topic-specific metrics, visuals, and limitations so that the five papers in the issue remain independent in subject and presentation.