Peer-reviewed research powered by real-world wearable data collection across health, medicine, and behavioral science.
This study investigates performance differences between chest-worn and wrist-worn heart rate monitors. Despite significant discrepancies exceeding 10 BPM between devices, all tested monitors effectively detect stress using a machine learning model achieving F1-score above 0.8, also introducing formal verification of classifier rules for model explainability.
This pilot study explores voice characteristics as a digital biomarker for detecting heart failure decompensation. Over two months, 35 stable heart failure patients record daily voice samples via tablet while tracking health status. Results could support early intervention and reduce hospitalizations through passive, non-invasive monitoring.
This pilot study examines differences in daily caloric intake between esports gamers (n=7) and recreational gamers (n=5). Participants tracked diet for seven days using Labfront and ASA24. Results showed no significant difference in calorie intake between groups, despite prior suggestions that esports gamers may be more health-conscious.
This pilot study examines daily caffeine intake among esports gamers (n=7) and recreational gamers (n=5). Participants tracked their diet for seven days using Labfront and ASA24. Despite suggestions that esports gamers may be more health-conscious, results showed no significant difference in caffeine intake between groups.
This study investigates how a dynamic lighting schedule influences neurobehavioral performance, alertness, and attention during a simulated 45-day space mission. Using a psychomotor vigilance task administered multiple times per day, findings suggest dynamic lighting can positively affect cognitive performance.
This 12-week crossover RCT evaluates therapeutic footwear impact on physical activity in older adults with chronic foot pain. OrthoFeet shoes significantly increased daily steps by 698 and calorie expenditure by 177 kcal/day, highlighting the importance of biomechanical features in promoting mobility among older adults.
This study explores wrist-worn fitness trackers to address challenges in remote musculoskeletal monitoring. A model estimating tibia bone loading from fitness tracker data showed errors below 5% when insoles were worn at least 25% of the day, supporting a multi-sensor approach for long-term telerehabilitation.
This study investigates early time-restricted eating where participants consumed all meals within a 6-hour window vs. a 12-hour eating schedule. eTRE improved cardiometabolic markers including lower 24-hour glucose, reduced glycemic variability, and improved insulin sensitivity, but did not impact intestinal nutrient absorption.
This pilot study evaluated virtual pulmonary rehabilitation for post-COVID syndrome. Participants were randomly assigned to video conference-led or self-directed exercise sessions over eight weeks. Both were feasible, safe, and well-received, with video conference participants showing improved sit-to-stand capacity.
This study examined whether library noise levels impact physiological stress in patrons using Garmin Vivosmart 5 wearables to measure HRV and skin temperature. 68 participants were monitored while studying in quiet and loud areas. Results showed no significant difference in stress levels between the two conditions.
This study introduces a Bayesian inference-based data fusion approach to improve heart rate accuracy from wearable sensors in older adults. Using functional data analysis and a Gaussian process model, the method reduces noise and motion artifacts, enhancing real-world continuous physiological monitoring beyond conventional techniques.
This case study examines whether sprint interval training can maintain elevated cardiac output observed during pregnancy. A trained female athlete completed a six-week SIT program postpartum. Results show cardiac output remained higher than pre-pregnancy levels, suggesting SIT is effective for trained athletes regaining cardiovascular fitness postpartum.
This study explores wearable sensors for detecting e-cigarette use in young adults. Using BBI data from smartwatches worn by 12 regular e-cigarette users for 7 days, a second-order polynomial model achieved AUC of 0.76, suggesting potential for timing cessation support messages via mobile health interventions.
This study investigated feasibility of a 4-week protocol using Garmin Vivosmart 4 and Labfront app to collect physiological data and EMA in a racially diverse sample of older adults. Results demonstrated high adherence and satisfaction, suggesting Labfront-supported smartwatch monitoring is viable for longitudinal health research.
The RENEWAL study used smartphone GPS, Garmin smartwatches, and digital diaries to monitor 294 participants over 30 days. To minimize missing data, researchers provided smartwatches as incentives and used automated notifications. The pilot achieved 98% diary completion and 92% location data retention.
This study used Garmin smartwatches and the Labfront Companion App to track daily symptoms and physical activity in people with post-COVID-19 syndrome. Greater dizziness consistently predicted fewer daily steps, while achieving 5,000+ steps was associated with lower fatigue and chest pain the next day.
This study examined how the Garmin Stress Score relates to mood in 95 healthy young adults over 28 days using EMA and continuous smartwatch data. The Stress Score was linked to higher-intensity positive moods but not negative moods. The authors note the label is misleading as it reflects parasympathetic activity rather than stress.
This study examined HRV changes immediately before and after cannabis use in African American young adults with cannabis use disorder. Results showed a significant decrease in HRV and increase in heart rate after cannabis smoking, indicating acute autonomic changes including heightened sympathetic and reduced parasympathetic activity.
This study implemented a walking program and nurse education initiative for patients starting oral chemotherapy. Nurses showed 13% improvement in knowledge and increased confidence in patient education. Findings highlight the importance of early fatigue management education for advanced-stage patients.
This study evaluated a 30-day physical activity promotion program at a U.S. university during COVID-19. Undergraduate students engaged less with the app and had lower retention than graduate students. Higher app engagement, team affiliation, and social connections were associated with better outcomes.
Using a novel twin-based design, this study investigated how intensive meditation affects brain activity, heart rate, and molecular markers. Twins showed similar physiological and molecular changes across the retreat, highlighting the influence of genetics and the potential of mind-based practices to impact health across multiple biological systems.
This study found that using smartwatches and daily surveys to monitor racially diverse older adults, including those with mild cognitive impairment, is feasible and well-accepted. Participants wore the watch 21 hours/day and completed 94% of daily surveys. Overall, digital health monitoring was highly usable in this population.
This study explores wearable devices to continuously measure clinician wellbeing over seven days alongside traditional questionnaires. Comparing HRV, activity, gait, and sleep quality data, researchers assess whether wearables can provide more objective and comprehensive understanding of clinician health.
This study examined the bidirectional relationship between nicotine vaping and sleep quality in 35 young adults using real-time data over a week. Poorer sleep predicted higher negative mood and vaping cravings the next day, while increased vaping led to lower sleep quality and more light sleep.
The Life Improvement Trial (LIFT) is a randomized, double-blind clinical study testing pyridostigmine and low-dose naltrexone for treating ME/CFS. Conducted at Massachusetts General and Brigham and Womens Hospitals, the trial evaluates impact on symptoms and physiological responses in 160 participants over 13 weeks.
This study explores consumer-grade smartwatches for detecting nocturnal hypoglycemia using machine learning. Analyzing data from 36 adults across 351 nights, a trained ML model achieved strong performance (AUROC 0.78-0.83), suggesting smartwatches could complement continuous glucose monitoring as a cost-effective tool.
This study introduces statistical methods for analyzing heart rate data to identify patterns associated with nicotine vaping. Using singular spectrum analysis on 35 young adult e-cigarette users, researchers found heart rate increases before vaping, highlighting potential for wearable-based mobile health interventions.
This pilot RCT examined effects of Sudarshan Kriya Yoga on HRV, sleep, and mental health in healthy individuals over eight weeks. Results show SKY improves resting HRV, sleep quality, anxiety, and social connectedness, confirming feasibility for real-world data collection using wearable devices.
This study used wearable devices and Labfront to collect real-time physiological data from construction workers in Saudi Arabia, developing deep learning models to predict and prevent heat stress. The attention-based model achieved over 95% accuracy, enabling proactive safety decisions on construction sites.
This study tracks recovery after total knee replacement surgery by combining continuous mobility data from wearable devices with patient-reported outcomes and clinical information. Monitoring daily activity before and after surgery, researchers aim to better understand recovery patterns and improve post-operative rehabilitation.
The ATMOSPHERE project develops and tests mobile AI technology to help predict seizures for people with epilepsy. This feasibility study recruits 60 adults to test and refine the technology, gathering usability feedback to prepare for a larger clinical trial aimed at creating a reliable seizure forecasting tool.
The ExerT1D study tested a virtual exercise program for adolescents with type 1 diabetes combining active video games, diabetes management education, and peer support from young adult T1D coaches. Participants reported high satisfaction, improved self-management skills, and strong peer connections, especially after adding virtual reality.
This study evaluated the accuracy of the Garmin Vivosmart 4 for measuring heart rate and HRV compared to a gold-standard ECG device during daily life. Results showed the wearable was most accurate during rest, sleep, and sedentary postures, but less reliable during movement. Findings offer practical guidance for trusting wearable data in real-world health monitoring.