Withings real world wearables aggregated dataset (2023-2024)

Abstract The integration of digital health technologies into consumer devices has enabled the collection of large-scale real-world data, providing valuable insights into population health trends. This study presents descriptive statistics of key physiological parameters collected from million measurements using Withings connected health devices, including smart scales, sleep analyzers, blood pressure monitors, and wearable sensors during 2023 and 2024. The dataset, stratified by age, body mass index (BMI), and sex, encompasses body weight, pulse wave velocity, sleep duration, blood pressure, and daily steps, offering a comprehensive view of health metrics across diverse demographic subgroups. This work aims to provide a benchmark for researchers and clinicians to compare their populations, identify deviations from expected norms, and assess population health trends. This dataset serves as a resource for standardizing data interpretation and fostering comparisons across different research cohorts, thereby advancing the field of digital health and population health research. Terms of Use Research Only: This dataset is provided exclusively for academic and public-interest research. Non-Commercial: Use of this data by commercial entities or for-profit activities is not permitted. Attribution: You must provide appropriate credit, provide a link to the license, and indicate if changes were made. No Redistribution: Please direct others to the Zenodo DOI to download the dataset rather than redistributing the files yourself. Introduction The integration of digital health technologies into clinical practice and research has led to the collection of large-scale real-world data, offering valuable insights into population health trends (Lechat et al., 2023, 2024; Scott et al., 2023). Withings connected health devices, such as smart scales (Campo et al., 2017; Riveline et al., 2023), sleep analyzer (Edouard et al., 2021), blood pressure monitors (Topouchian et al., 2022; Zelveian et al., 2022; Hakobyan et al., 2023), and wearable sensors (Kirszenblat and Edouard, 2021; Campo et al., 2022), provide continuous and objective measurements of physiological parameters, enabling the study of variations across different demographic and clinical subgroups. However, for researchers and clinicians to interpret these data meaningfully, reference values derived from robust datasets are essential. In this brief data report, we present descriptive statistics—mean, standard deviation, and other relevant distributional metrics—of key physiological parameters collected from a large dataset of individuals using Withings connected health devices. The Withings real world wearables aggregated dataset (WRWWAD) (2023-2024) aggregates more than 65 millions health-related measurements, including body weight, pulse wave velocity, sleep duration, blood pressure, steps, active minutes, and is stratified by age, body mass index (BMI) and sex to facilitate comparative analysis. By making these mathematical parameters available, we aim to provide a benchmark against which researchers and clinicians can compare their own populations. Such reference data are crucial for identifying deviations from expected norms, assessing population health trends, and improving personalized medicine approaches. This dataset serves as a resource for those utilizing at-home health devices in their studies, helping to standardize data interpretation and foster comparison across different research cohorts while removing the need for complex and heavy dataset manipulation. Material and Methods Dataset The biomarkers included within the data are : body weight in kilograms, pulse wave velocity (PWV) in meter/second, sleep duration (per night) in hour, average nightly heart rate in beat per minute, systolic and diastolic blood pressure in millimeter of mercury (mmhg), daily number of steps in raw count. The data presented in this study were collected from users across North America (USA, Canada, and Mexico), Japan, Western Europe (France, Germany, and the United Kingdom) and Southern Europe (Spain, Portugal, Italy and Greece) (Publications Office of the European Union, 2025) from January 1, 2023, to December 31, 2024. To ensure data quality and relevance, several filtering steps were applied. First, we restricted the dataset to individuals aged 7 to 90 years. The lower age limit was chosen to exclude erroneous weight measurements from adults holding infants, which was suggested by the observation of a median weight of 76 kg in users under 7. The upper age limit was set due to the limited number of data points for individuals over 90 and the potential for measurement errors caused by balance issues in this elderly population. Then age was converted to category (07-12, 13-18, 19-25, 26-39, 40-49, 50-59, 60-69, 70-79, 80-90). Second, to address extreme outliers, we excluded the lowest and highest 0.005% of values for each biomarker per year. This stringent filtering removed highly improbable data points while preserving the integrity of the real-world measurements leading to keep 67636261 measurements from 2655433 users across our biomarkers. To reduce the influence of users with disproportionately high measurement counts we computed only the median value per user per year for each biomarker, thus minimizing the impact of measurement frequency on population statistics. This work focuses on users who have at least three weight measurements per year, allowing for the computation of BMI categories. Regarding BMI we used the median BMI per year per user to attribute them to the BMI category. BMI categories were constructed as follows : underweight (<18.5), normal (18.5-24.9), overweight (25-29.9), obese (>= 30). For each stratum (age, sex, area, bmi), we computed a range of descriptive statistics: sample size category (100-, 100-499, 500-999 and 1000+), minimum, maximum, median, first quartile (Q1), third quartile (Q3), interquartile range (IQR), mean, standard deviation (SD), standard error (SE), and confidence interval (CI). We included numerical categories (n_group) for each parameter subset in Table to help researchers assess data reliability. These comprehensive statistical summaries are detailed in Table. Ethics Data presented here are aggregated and then fully anonymized. Withings follows the European GDPR (European Parliament, 2016) law and the French CNIL recommendations concerning data management and processing (CNIL - France, 2024). HDS/ISO/HIPAA certifications and compliances can be found on Withings’ data security page. Conclusion This dataset can serve as a valuable resource for standardizing data interpretation and fostering comparisons across different research cohorts while using Withings devices. By making these data freely available, we hope to facilitate further research and contribute to the broader understanding of population health dynamics. Future research could build upon these findings by exploring the longitudinal trends and potential causal relationships between these biomarkers and health outcomes. Additionally, integrating these normative values into clinical practice could enhance patient monitoring and personalized healthcare strategies while comparing patient population with specific medical characteristics. Limitations Although the profiles align with global expectations, there may be socio-demographic population biases related to Withings users. While we are confident in the overall robustness of the data, we encourage researchers to review these data carefully when working with underage or underweight groups, as these categories typically contain less data. While being computed, the adult BMI metric is not adapted to the underage category. Be careful on that stratum if not excluded from any analysis. Conflict of interest Withings is the manufacturer and developer of the devices used to generate data and this study was conducted by Withings. Thus, the author declares a potential conflict of interest, as the findings of this study may have implications for the acceptance and use of the company’s products and services. Author contributions B.V., A.C and P.L equally: Conceptualization, Methodology, Software, Formal Analysis, Writing—Original Draft Funding This work was entirely supported by Withings. Acknowledgments We’d like to thank all the Withings collaborators that work on our product. References Campo, D., Elie, V., de Gallard, T., Bartet, P., Morichau-Beauchant, T., Genain, N., et al. (2022). Atrial Fibrillation Detection With an Analog Smartwatch: Prospective Clinical Study and Algorithm Validation. JMIR Form. Res. 6, e37280. doi: 10.2196/37280 Campo, D., Khettab, H., Yu, R., Genain, N., Edouard, P., Buard, N., et al. (2017). Measurement of Aortic Pulse Wave Velocity With a Connected Bathroom Scale. Am. J. Hypertens. 30, 876–883. doi: 10.1093/ajh/hpx059 CNIL - France (2024). Practice guide GDPR. Available at: https://www.cnil.fr/sites/cnil/files/2024-03/cnil_guide_securite_personnelle_ven_0.pdf Edouard, P., Campo, D., Bartet, P., Yang, R.-Y., Bruyneel, M., Roisman, G., et al. (2021). Validation of the Withings Sleep Analyzer, an under-the-mattress device for the detection of moderate-severe sleep apnea syndrome. J. Clin. Sleep Med. JCSM Off. Publ. Am. Acad. Sleep Med. 17, 1217–1227. doi: 10.5664/jcsm.9168 European Parliament (2016). Regulation - 2016/679 - EN - gdpr - EUR-Lex. Available at: https://eur-lex.europa.eu/eli/reg/2016/679/oj/eng (Accessed February 7, 2025). Hakobyan, Z., Zelveian, P., Topouchian, J., Hazarapetyan, L., and Asmar, R. (2023). Validation of the Withings BPM Core Device for Self-Blood Pressure Measurements in General Population According to the Association for the Advancement of Medical Instrumentation/European Society of Hypertension/International Organization for Standardization Universal Standard. Vasc. Health Risk Manag. 19, 391–398. doi: 10.2147/VHRM.S413195 Kirszenblat, R., and Edouard, P. (2021). Validation of the Withings ScanWatch as a Wrist-Worn Reflective Pulse Oximeter: Prospective Interventional Clinical Study. J. Med. Internet Res. 23, e27503. doi: 10.2196/27503 Lechat, B., Loffler, K. A., Reynolds, A. C., Naik, G., Vakulin, A., Jennings, G., et al. (2023). High night-to-night variability in sleep apnea severity is associated with uncontrolled hypertension. NPJ Digit. Med. 6, 57. doi: 10.1038/s41746-023-00801-2 Lechat, B., Naik, G., Appleton, S., Manners, J., Scott, H., Nguyen, D. P., et al. (2024). Regular snoring is associated with uncontrolled hypertension. NPJ Digit. Med. 7, 38. doi: 10.1038/s41746-024-01026-7 Publications Office of the European Union (2025). Western Europe - EU Vocabularies. EU Vocab. Available at: https://op.europa.eu/en/web/eu-vocabularies/concept/-/resource (Accessed February 19, 2025). Riveline, J.-P., Mallone, R., Tiercelin, C., Yaker, F., Alexandre-Heymann, L., Khelifaoui, L., et al. (2023). Validation of the Body Scan®, a new device to detect small fiber neuropathy by assessment of the sudomotor function: agreement with the Sudoscan®. Front. Neurol. 14. doi: 10.3389/fneur.2023.1256984 Scott, H., Lechat, B., Guyett, A., Reynolds, A. C., Lovato, N., Naik, G., et al. (2023). Sleep Irregularity Is Associated With Hypertension: Findings From Over 2 Million Nights With a Large Global Population Sample. Hypertens. Dallas Tex 1979 80, 1117–1126. doi: 10.1161/HYPERTENSIONAHA.122.20513 Topouchian, J., Zelveian, P., Hakobyan, Z., Gharibyan, H., and Asmar, R. (2022). Accuracy of the Withings BPM Connect Device for Self-Blood Pressure Measurements in General Population - Validation According to the Association for the Advancement of Medical Instrumentation/European Society of Hypertension/International Organization for Standardization Universal Standard. Vasc. Health Risk Manag. 18, 191–200. doi: 10.2147/VHRM.S350006 Zelveian, P., Topouchian, J., Hakobyan, Z., Asmar, J., Gharibyan, H., and Asmar, R. (2022). Clinical Accuracy of the Withings BPM Connect for Self-Blood Pressure Measurements in Pregnancy and Pre-Eclampsia: Validation According to the Association for the Advancement of Medical Instrumentation/European Society of Hypertension/International Organization for Standardization Universal Standard. Vasc. Health Risk Manag. 18, 181–189. doi: 10.2147/VHRM.S351313