David Campo

Background: Personalized medicine requires an understanding of health conditions in real-world settings, but traditional clinical tools are often poorly suited for longitudinal tracking due to logistical hurdles, high participant burden, and data quality issues. While consumer wearables offer a scalable alternative, decentralized clinical trials introduce new challenges in patient recruitment, sustained engagement, and data management. Insomnia serves as a suitable use case for this protocol due to its global prevalence, frequent underdiagnosis, and the well-documented discrepancy between subjective patient reports and objective physiological metrics. Objective: We present a research protocol for large-scale decentralized clinical trials and report its recruitment performance, applied to a longitudinal study on insomnia, called the SLEEPI (Smart Devices–Based Longitudinal Assessment of Patterns of Insomnia) study, using an established base of commercial devices with limited additional manual intervention. This study aims to develop a machine learning model classifying subjective insomnia presence at the participant level against DSM-5 -TR ( Diagnostic and Statistical Manual of Mental Disorders , Fifth Edition, Text Revision) criteria. Methods: Two data streams are collected: subjective, clinically anchored patient reports and objective physiological metrics (eg, heart and respiratory rates) captured via unobtrusive wearable and nearable sensors. This study is executed through a digital protocol designed to conduct a large-scale, longitudinal clinical trial. This protocol replaces the manual interventions of site-based studies with digital tools that manage the full participation process. We detail this protocol from initial recruitment through engagement, adherence, study monitoring, and data management to study completion or withdrawal. The data collected are intended to train and evaluate a machine learning model for subjective insomnia classification. The recruitment objectives were determined by using simulations to determine the minimal sample size for that purpose and by applying a prospective attrition model to calculate the initial contact volume needed to reach that targeted sample. Results: This protocol enables the continuous monitoring of over 10,000 individuals for 26 weeks. Enrollment metrics exceeded projections: initial survey response rates reached 16% (vs 10% projected), the consent rate was 96% (vs 77% projected), and the final inclusion rate was 90% (vs 81% projected). Recruitment opened in February 2025, and the initial wave closed 6 weeks later with 10,823 individuals included. Data collection is projected to end in June 2027. Preliminary analyses have been conducted, and the first results are expected in the summer of 2027. Conclusions: This protocol uses preexisting user bases of connected devices to conduct the trial with limited additional manual intervention. Beyond insomnia, the protocol is intended as a model for generating large-scale, real-world clinical datasets, presenting new opportunities for research in other complex areas such as immunology, mental health, and cardiovascular diseases.

Background: Sleep is essential for overall health and well-being, but assessing sleep architecture is often costly and time-consuming, relying primarily on polysomnography (PSG). While wrist-worn wearables offer alternatives, they face limitations regarding user compliance, such as battery charging and physical discomfort. Nearable devices address these burdens, but they regularly lack rigorous validation, especially in real-world settings. Objective: This study evaluates the accuracy and reliability of the Withings Sleep Analyzer (WSA), a contactless sleep monitoring device, compared to PSG in a home setting using a large and diverse cohort of healthy individuals. Methods: A total of 117 healthy volunteers (69 women; mean 39.9, SD 11.4 years), prospectively recruited from the general population, underwent home-based PSG and simultaneous WSA recording. The study was conducted under free-living conditions, without constraints on substance intake, prebedtime activity, or forced sleep schedules. The main outcomes were the device’s performance in sleep-wake distinction and sleep stage identification using accuracy, kappa, sensitivity, specificity, and the mean absolute error of sleep measures on the entire population and demographic, clinical, and environmental subgroups. Results: WSA demonstrates high sensitivity (93%, 95% CI 92%-94%) for sleep detection and moderate sensitivity (73%, 95% CI 69%-77%) for wakefulness, achieving an overall accuracy of 87% (95% CI 86%-87%) for sleep-wake distinction. The device showed consistent performance across various demographic subgroups, including different age, BMI, mattress, and sleep arrangements (with or without bed partner) categories. Challenges were noted in accurately classifying specific sleep stages, particularly in distinguishing between light and deep sleep, with a mean accuracy of 63% (95% CI 62%-65%) and a Cohen κ of 0.49 (95% CI 0.47-0.51). The WSA tended to overestimate total sleep time (20 min, 95% CI 10 min to 31 min) and light sleep (1 h 21 min, 95% CI 1 h 8 min to 1 h 36 min) while underestimating rapid eye movement (−15 min, 95% CI −23 min to −8 min) and deep sleep (−46 min, 95% CI −59 min to −34 min) durations. Disagreements between expert reviewers were mirrored in part by the WSA’s misclassifications. Participants reported altered perceived sleep quality during the night with the PSG, suggesting discomfort during sleep. Conclusions: Being contactless and placed under the mattress, the WSA offers a promising approach to long-term sleep monitoring in natural home environments. It shows competitive performance in sleep-wake and sleep stage identification compared to other consumer devices. Progress in wearable and nearable devices is necessary to enhance their accuracy to better support the monitoring of populations with impaired sleep, although limited by an imperfect gold standard. This work also emphasizes the importance of using large, diverse, and challenging datasets, as well as the need for a standardized methodology for accurate sleep stage classification.

A bstract: Background: Sleep is essential for overall health and well-being, but assessing sleep architecture and quality is often costly and time-consuming, relying primarily on polysomnography (PSG). Wearable and nearable devices offer potential alternatives, but they regularly lack rigorous validation, especially in real-world settings. Objectives: This study evaluates the accuracy and reliability of Withings Sleep Analyzer (WSA), a contactless sleep monitoring device, compared to PSG in a home setting using a large and diverse cohort of healthy individuals. Methods: A total of 117 participants (69 women; 39.9 ± 11.4 years, mean ± std) underwent home-based polysomnography (PSG) and simultaneous WSA recording. Data analysis focused on evaluating sleep-wake distinction and sleep stage identification using standard classification metrics. Results: WSA demonstrates high sensitivity (93%) for sleep detection and moderate sensitivity (73%) for wakefulness, achieving an overall accuracy of 87% for sleep-wake distinction. The device showed consistent performance across various demographic subgroups, including different age, BMI, mattress and sleep arrangements (with or without bed partner) categories. Challenges were noted in accurately classifying specific sleep stages, particularly in distinguishing between light and deep sleep, with a mean accuracy of 63% and a Cohen’s Kappa of 0.49. The WSA tended to overestimate total sleep time (+20 min) and light sleep (+1h21 min) while underestimating REM (−15 min) and deep sleep (−46 min) durations. Disagreements between expert reviewers, particularly between light and deep sleep stages, mirrored in part the WSA’s misclassifications. Participants reported significantly altered perceived sleep quality during the night with the PSG, suggesting potential discomfort during sleep. Conclusions: WSA offers a promising approach to sleep monitoring in natural home environments. Being contactless and placed under the mattress, the WSA’s allows for long-term monitoring of sleep measures. It shows competitive performance in sleep-wake and sleep stage identification compared to other consumer devices. Progress in wearable and nearable devices is necessary to enhance their accuracy to better support the monitoring of populations with strongly impaired sleep, although limited by an imperfect gold standard. This work also emphasizes the importance of using large, diverse, and challenging datasets, as well as the need for a standardized methodology for accurate sleep stage classification.

BACKGROUND Sleep is essential for overall health and well-being, but assessing sleep architecture and quality is often costly and time-consuming, relying primarily on polysomnography (PSG). Wearable and nearable devices offer potential alternatives, but they regularly lack rigorous validation, especially in real-world settings. OBJECTIVE This study evaluates the accuracy and reliability of Withings Sleep Analyzer (WSA), a contactless sleep monitoring device, compared to PSG in a home setting using a large and diverse cohort of healthy individuals. METHODS A total of 117 participants (69 women; 39.9 11.4 years, mean std) underwent home-based polysomnography (PSG) and simultaneous WSA recording. Data analysis focused on evaluating sleep-wake distinction and sleep stage identification using standard classification metrics. RESULTS WSA demonstrates high sensitivity (93%) for sleep detection and moderate sensitivity (73%) for wakefulness, achieving an overall accuracy of 87% for sleep-wake distinction. The device showed consistent performance across various demographic subgroups, including different age, BMI, mattress and sleep arrangements (with or without bed partner) categories. Challenges were noted in accurately classifying specific sleep stages, particularly in distinguishing between light and deep sleep, with a mean accuracy of 63% and a Cohen's Kappa of 0.49. The WSA tended to overestimate total sleep time (+20 min) and light sleep (+1h21 min) while underestimating REM (-15 min) and deep sleep (-46 min) durations. Disagreements between expert reviewers, particularly between light and deep sleep stages, mirrored in part the WSA's misclassifications. Participants reported significantly altered perceived sleep quality during the night with the PSG, suggesting potential discomfort during sleep. CONCLUSIONS WSA offers a promising approach to sleep monitoring in natural home environments. Being contactless and placed under the mattress, the WSA's allows for long-term monitoring of sleep measures. It shows competitive performance in sleep-wake and sleep stage identification compared to other consumer devices. Progress in wearable and nearable devices is necessary to enhance their accuracy to better support the monitoring of populations with strongly impaired sleep, although limited by an imperfect gold standard. This work also emphasizes the importance of using large, diverse, and challenging datasets, as well as the need for a standardized methodology for accurate sleep stage classification.

Background: Atrial fibrillation affects approximately 4% of the world’s population and is one of the major causes of stroke, heart failure, sudden death, and cardiovascular morbidity. It can be difficult to diagnose when asymptomatic or in the paroxysmal stage, and its natural history is not well understood. New wearables and connected devices offer an opportunity to improve on this situation. Objective: We aimed to validate an algorithm for the automatic detection of atrial fibrillation from a single-lead electrocardiogram taken with a smartwatch. Methods: Eligible patients were recruited from 4 sites in Paris, France. Electrocardiograms (12-lead reference and single lead) were captured simultaneously. The electrocardiograms were reviewed by independent, blinded board-certified cardiologists. The sensitivity and specificity of the algorithm to detect atrial fibrillation and normal sinus rhythm were calculated. The quality of single-lead electrocardiograms (visibility and polarity of waves, interval durations, heart rate) was assessed in comparison with the gold standard (12-lead electrocardiogram). Results: A total of 262 patients (atrial fibrillation: n=100, age: mean 74.3 years, SD 12.3; normal sinus rhythm: n=113, age: 61.8 years, SD 14.3; other arrhythmia: n=45, 66.9 years, SD 15.2; unreadable electrocardiograms: n=4) were included in the final analysis; 6.9% (18/262) were classified as Noise by the algorithm. Excluding other arrhythmias and Noise, the sensitivity for atrial fibrillation detection was 0.963 (95% CI lower bound 0.894), and the specificity was 1.000 (95% CI lower bound 0.967). Visibility and polarity accuracies were similar (1-lead electrocardiogram: P waves: 96.9%, QRS complexes: 99.2%, T waves: 91.2%; 12-lead electrocardiogram: P waves: 100%, QRS complexes: 98.8%, T waves: 99.5%). P-wave visibility accuracy was 99% (99/100) for patients with atrial fibrillation and 95.7% (155/162) for patients with normal sinus rhythm, other arrhythmias, and unreadable electrocardiograms. The absolute values of the mean differences in PR duration and QRS width were <3 ms, and more than 97% were <40 ms. The mean difference between the heart rates from the 1-lead electrocardiogram calculated by the algorithm and those calculated by cardiologists was 0.55 bpm. Conclusions: The algorithm demonstrated great diagnostic performance for atrial fibrillation detection. The smartwatch’s single-lead electrocardiogram also demonstrated good quality for physician use in daily routine care. Trial Registration: ClinicalTrials.gov NCT04351386; http://clinicaltrials.gov/ct2/show/NCT04351386

Background: Physical activity (PA) is a modifiable lifestyle factor that can be targeted to increase energy expenditure and promote weight loss. However, the amount of PA required for weight loss remains inconsistent. Wearable activity trackers constitute a valuable opportunity to obtain objective measurements of PA and study large populations in real-life settings. Objective: We aim to study the associations of initial device-assessed PA characteristics (average step counts and step count variability) and their evolution with 6-month weight change. Methods: We analyzed data from 26,935 Withings-connected device users (wearable activity trackers and digital scales). To assess the initial PA characteristics and their 6-month changes, we used data recorded during the first and sixth 30-day periods of activity tracker use. For each of these periods, we used the monthly mean of daily step values as a proxy for PA level and derived the monthly coefficient of variation (CV) of daily step values to estimate PA level variability. Associations between initial PA characteristics and 6-month weight change were assessed using multivariable linear regression analyses controlled for age, sex, blood pressure, heart rate, and the predominant season. Restricted cubic spline regression was performed to better characterize the continuous shape of the associations between PA characteristics and weight change. Secondary analyses were performed by analyzing the 6-month evolution of PA characteristics in relation to weight change. Results: Our results revealed that both a greater PA level and lower PA level variability were associated with weight loss. Compared with individuals who were initially in the sedentary category (63%, users whose PA level CV ranged from 51% to 63%, 40% to 51%, and was ≤40%, had a 0.19-kg, a 0.23-kg, and a 0.33-kg greater decrease in weight, respectively (95% CI −0.38 to −0.01, −0.41 to −0.04, and −0.53 to −0.13, respectively). We also observed that each 1000 steps/day increase in PA level over the 6-month follow-up was associated with a 0.26-kg (95% CI −0.29 to −0.23) decrease in weight. No association was found between the 6-month changes in PA level variability and weight change. Conclusions: Our results add to the current body of knowledge that health benefits can be observed below the 10,000 steps/day threshold and suggest that not only increased mean PA level but also greater regularity of the PA level may play important roles in short-term weight loss.

BACKGROUND Physical activity (PA) is a modifiable lifestyle factor that can be targeted to increase energy expenditure and promote weight loss. However, the amount of PA required for weight loss remains inconsistent. Wearable activity trackers constitute a valuable opportunity to obtain objective measurements of PA and study large populations in real-life settings. OBJECTIVE We aim to study the associations of initial device-assessed PA characteristics (average step counts and step count variability) and their evolution with 6-month weight change. METHODS We analyzed data from 26,935 Withings-connected device users (wearable activity trackers and digital scales). To assess the initial PA characteristics and their 6-month changes, we used data recorded during the first and sixth 30-day periods of activity tracker use. For each of these periods, we used the monthly mean of daily step values as a proxy for PA level and derived the monthly coefficient of variation (CV) of daily step values to estimate PA level variability. Associations between initial PA characteristics and 6-month weight change were assessed using multivariable linear regression analyses controlled for age, sex, blood pressure, heart rate, and the predominant season. Restricted cubic spline regression was performed to better characterize the continuous shape of the associations between PA characteristics and weight change. Secondary analyses were performed by analyzing the 6-month evolution of PA characteristics in relation to weight change. RESULTS Our results revealed that both a greater PA level and lower PA level variability were associated with weight loss. Compared with individuals who were initially in the sedentary category (63%, users whose PA level CV ranged from 51% to 63%, 40% to 51%, and was ≤40%, had a 0.19-kg, a 0.23-kg, and a 0.33-kg greater decrease in weight, respectively (95% CI −0.38 to −0.01, −0.41 to −0.04, and −0.53 to −0.13, respectively). We also observed that each 1000 steps/day increase in PA level over the 6-month follow-up was associated with a 0.26-kg (95% CI −0.29 to −0.23) decrease in weight. No association was found between the 6-month changes in PA level variability and weight change. CONCLUSIONS Our results add to the current body of knowledge that health benefits can be observed below the 10,000 steps/day threshold and suggest that not only increased mean PA level but also greater regularity of the PA level may play important roles in short-term weight loss.