Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Wearable sensors enable personalized predictions of clinical laboratory measurements


Vital signs, including heart rate and body temperature, are useful in detecting or monitoring medical conditions, but are typically measured in the clinic and require follow-up laboratory testing for more definitive diagnoses. Here we examined whether vital signs as measured by consumer wearable devices (that is, continuously monitored heart rate, body temperature, electrodermal activity and movement) can predict clinical laboratory test results using machine learning models, including random forest and Lasso models. Our results demonstrate that vital sign data collected from wearables give a more consistent and precise depiction of resting heart rate than do measurements taken in the clinic. Vital sign data collected from wearables can also predict several clinical laboratory measurements with lower prediction error than predictions made using clinically obtained vital sign measurements. The length of time over which vital signs are monitored and the proximity of the monitoring period to the date of prediction play a critical role in the performance of the machine learning models. These results demonstrate the value of commercial wearable devices for continuous and longitudinal assessment of physiological measurements that today can be measured only with clinical laboratory tests.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Overview of the iPOP wearables study.
Fig. 2: Methodology for predicting clinical laboratory measurements from vital signs collected using wearables.
Fig. 3: Predicting clinical laboratory measurements from vital signs collected using wearables.
Fig. 4: Relationship between duration and proximity of monitoring and model accuracy.
Fig. 5: Personalized models improve predictions of clinical laboratory tests from vital sign measurements.

Data availability

Intel Basis watch data are available on the Stanford iPOP site ( and in the Digital Health Data Repository48 ( Data that are unique to this study are included as source data and in the supplementary tables. Source data are provided with this paper.

Code availability

R version 3.3.3 was used with the base packages and the following additional CRAN packages: stats, glmnet, lme4, randomForest and PMA. Custom scripts were used for data analysis and are open source via (, and wearables data pre-processing scripts are available on the Digital Biomarker Discovery Pipeline (


  1. 1.

    Sackett, D. L. The rational clinical examination. A primer on the precision and accuracy of the clinical examination. J. Am. Med. Assoc. 267, 2638–2644 (1992).

    CAS  Google Scholar 

  2. 2.

    Hatala, R. et al. An evidence-based approach to the clinical examination. J. Gen. Intern. Med. 12, 182–187 (1997).

    CAS  PubMed  PubMed Central  Google Scholar 

  3. 3.

    Armbruster, D. & Miller, R. R. The Joint Committee for Traceability in Laboratory Medicine (JCTLM): a global approach to promote the standardisation of clinical laboratory test results. Clin. Biochem. Rev. 28, 105–113 (2007).

    PubMed  PubMed Central  Google Scholar 

  4. 4.

    Sudlow, C. et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 12, e1001779 (2015).

    PubMed  PubMed Central  Google Scholar 

  5. 5.

    Nagai, A. et al. Overview of the BioBank Japan Project: study design and profile. J. Epidemiol. 27, S2–S8 (2017).

    PubMed  PubMed Central  Google Scholar 

  6. 6.

    Vaithinathan, A. G. & Asokan, V. Public health and precision medicine share a goal. J. Evid. Based Med. 10, 76–80 (2017).

    PubMed  Google Scholar 

  7. 7.

    Shcherbina, A. et al. Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. J. Pers. Med. 7, 3 (2017).

    PubMed Central  Google Scholar 

  8. 8.

    Li, X. et al. Digital health: tracking physiomes and activity using wearable biosensors reveals useful health-related information. PLoS Biol. 15, e2001402 (2017).

    PubMed  PubMed Central  Google Scholar 

  9. 9.

    Radin, J. W., Topol, E. & Steinhubl, S. Harnessing wearable device data to improve state-level real-time surveillance of influenza-like illness in the USA: a population-based study. Lancet Digit. Health 2, 85–93 (2020).

    Google Scholar 

  10. 10.

    Nemati, S. et al. Monitoring and detecting atrial fibrillation using wearable technology. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. 2016, 3394–3397 (2016).

  11. 11.

    Suzuki, T., Kameyama, K. & Tamura, T. Development of the irregular pulse detection method in daily life using wearable photoplethysmographic sensor. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. 2009, 6080–6083 (2009).

  12. 12.

    Steinhubl, S. R. et al. Rationale and design of a home-based trial using wearable sensors to detect asymptomatic atrial fibrillation in a targeted population: the mHealth Screening To Prevent Strokes (mSToPS) trial. Am. Heart J. 175, 77–85 (2016).

    PubMed  Google Scholar 

  13. 13.

    Hannun, A. Y. et al. Cardiologist-level arrhythmia detection with convolutional neural networks. Nat. Med. 25, 65–69 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  14. 14.

    Avram, R. et al. Predicting diabetes from photoplethysmography using deep learning. J. Am. Coll. Cardiol. 73 (2019).

  15. 15.

    Steinhubl, S. R. et al. Validation of a portable, deployable system for continuous vital sign monitoring using a multiparametric wearable sensor and personalised analytics in an Ebola treatment centre. BMJ Glob. Health 1, e000070 (2016).

    PubMed  PubMed Central  Google Scholar 

  16. 16.

    Chen, R. et al. Personal omics profiling reveals dynamic molecular and medical phenotypes. Cell 148, 1293–1307 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  17. 17.

    Piening, B. D. et al. Integrative personal omics profiles during periods of weight gain and loss. Cell Syst. 6, 157–170 e158 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  18. 18.

    Zhou, W. et al. Longitudinal multi-omics of host-microbe dynamics in prediabetes. Nature 569, 663–671 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  19. 19.

    Schussler-Fiorenza Rose, S. M. et al. A longitudinal big data approach for precision health. Nat. Med. 25, 792–804 (2019).

    CAS  PubMed  Google Scholar 

  20. 20.

    Albanese, M., Neofytou, M., Ouarrak, T., Schneider, S. & Schols, W. Evaluation of heart rate measurements in clinical studies: a prospective cohort study in patients with heart disease. Eur. J. Clin. Pharm. 72, 789–795 (2016).

    CAS  Google Scholar 

  21. 21.

    Kavsaoglu, A. R., Polat, K. & Hariharan, M. Non-invasive prediction of hemoglobin level using machine learning techniques with the PPG signal’s characteristics features. Appl. Soft Comput. 37, 983–991 (2015).

    Google Scholar 

  22. 22.

    Mandal, S. & Manasreh, M. O. An in-vitro optical sensor designed to estimate glycated hemoglobin levels. Sensors (Basel) 18, 1084 (2018).

    Google Scholar 

  23. 23.

    Boucsein, W. et al. Publication recommendations for electrodermal measurements. Psychophysiology 49, 1017–1034 (2012).

    PubMed  Google Scholar 

  24. 24.

    Wilson, P. W. et al. Prediction of coronary heart disease using risk factor categories. Circulation 97, 1837–1847 (1998).

    CAS  PubMed  Google Scholar 

  25. 25.

    Welch, H. G., Chapko, M. K., James, K. E., Schwartz, L. M. & Woloshin, S. The role of patients and providers in the timing of follow-up visits. telephone care study group. J. Gen. Intern. Med. 14, 223–229 (1999).

    CAS  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Reule, S. & Drawz, P. E. Heart rate and blood pressure: any possible implications for management of hypertension? Curr. Hypertens. Rep. 14, 478–484 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  27. 27.

    Palatini, P. et al. Reproducibility of heart rate measured in the clinic and with 24-hour intermittent recorders. Am. J. Hypertens. 13, 92–98 (2000).

    CAS  PubMed  Google Scholar 

  28. 28.

    Bloomfield, H. E. & Wilt, T. J. Evidence brief: role of the annual comprehensive physical examination in the asymptomatic dult. in VA Evidence-Based Synthesis Program Evidence Briefs (Department of Veterans Affairs, 2011).

  29. 29.

    Ikeda, N. et al. Effects of submaximal exercise on blood rheology and sympathetic nerve activity. Circ. J. 74, 730–734 (2010).

    PubMed  Google Scholar 

  30. 30.

    Weinberg, A. D. & Minaker, K. L. Dehydration. Evaluation and management in older adults. Council on Scientific Affairs, American Medical Association. J. Am. Med. Assoc. 274, 1552–1556 (1995).

    CAS  Google Scholar 

  31. 31.

    Xiao, H., Barber, J. & Campbell, E. S. Economic burden of dehydration among hospitalized elderly patients. Am. J. Health Syst. Pharm. 61, 2534–2540 (2004).

    PubMed  Google Scholar 

  32. 32.

    Avram, R. et al. Real-world heart rate norms in the health eHeart study. NPJ Digit. Med. 2, 58 (2019).

    PubMed  PubMed Central  Google Scholar 

  33. 33.

    St John, A. & Price, C. P. Existing and emerging technologies for point-of-care. Test. Clin. Biochem. Rev. 35, 155–167 (2014).

    Google Scholar 

  34. 34.

    Londeree, W., Davis, K., Helman, D. & Abadie, J. Bodily fluid analysis of non-serum samples using point-of-care testing with iSTAT and Piccolo analyzers versus a fixed hospital chemistry analytical platform. Hawaii J. Med. Public Health 73, 3–8 (2014).

    PubMed  PubMed Central  Google Scholar 

  35. 35.

    Hall, H. et al. Glucotypes reveal new patterns of glucose dysregulation. PLoS Biol. 16, e2005143 (2018).

    PubMed  PubMed Central  Google Scholar 

  36. 36.

    Zeevi, D. et al. Personalized nutrition by prediction of glycemic responses. Cell 163, 1079–1094 (2015).

    CAS  PubMed  Google Scholar 

  37. 37.

    Parlak, O., Keene, S. T., Marais, A., Curto, V. F. & Salleo, A. Molecularly selective nanoporous membrane-based wearable organic electrochemical device for noninvasive cortisol sensing. Sci. Adv. 4, eaar2904 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  38. 38.

    Emaminejad, S. et al. Autonomous sweat extraction and analysis applied to cystic fibrosis and glucose monitoring using a fully integrated wearable platform. Proc. Natl Acad. Sci. USA 114, 4625–4630 (2017).

    CAS  PubMed  Google Scholar 

  39. 39.

    Whelton, P. K. et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the american college of cardiology/american heart association task force on clinical practice guidelines. J. Am. Coll. Cardiol. 71, e127–e248 (2018).

    PubMed  Google Scholar 

  40. 40.

    Cadmus-Bertram, L., Gangnon, R., Wirkus, E. J., Thraen-Borowski, K. M. & Gorzelitz-Liebhauser, J. The accuracy of heart rate monitoring by some wrist-worn activity trackers. Ann. Intern. Med. 166, 610–612 (2017).

    PubMed  PubMed Central  Google Scholar 

  41. 41.

    Hastie, T. & Fithian, W. Response to ‘Perils of LOO crossvalidation’. (2013).

  42. 42.

    Poldrack, R. The perils of leave-one-out crossvalidation for individual difference analyses. (2012).

  43. 43.

    Friedman, J., Hastie, T. & Tibshirani, R. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 33, 1–22 (2010).

    PubMed  PubMed Central  Google Scholar 

  44. 44.

    Tibshirani, R. Regression shrinkage and selection via the lasso. J. R. Stat. Soc. Series B Stat. Methodol. 58, 267–288 (1996).

    Google Scholar 

  45. 45.

    Breiman, L. Random Forests. Mach. Learn. 45, 5–32 (2001).

    Google Scholar 

  46. 46.

    Witten, D. M., Tibshirani, R. & Hastie, T. A penalized matrix decomposition, with applications to sparse principal components and canonical correlation analysis. Biostatistics 10, 515–534 (2009).

    PubMed  PubMed Central  Google Scholar 

  47. 47.

    Hardoon, D. R. S.-T.J. Sparse canonical correlation analysis. Mach. Learn. 83, 331–353 (2011).

    Google Scholar 

  48. 48.

    Bent, B. et al. The digital biomarker discovery pipeline: an open source software platform for the development of digital biomarkers using mHealth and wearables data. J. Clin. Transl. Sci. 5, E19 (2021).

Download references


The work was supported by grants from the National Institues of Health (NIH) Common Fund Human Microbiome Project (HMP) (1U54DE02378901), an NIH National Center for Advancing Translational Science Clinical and Translational Science Award (UL1TR001085), and the Chan Zuckerberg Initiative Donor-Advised Fund (an advised fund of Silicon Valley Community Foundation; 2020-218599). JD, LK, RR and JH were supported by the Mobilize Center grant (NIH U54 EB020405). SMS-FR was supported by an NIH Career Development Award (no. K08 ES028825). JD is a MEDx Investigator and a Whitehead Scholar. The Stanford Translational Research Integrated Database Environment (STRIDE) is a research and development project at Stanford University that aims to create a standards-based informatics platform supporting clinical and translational research. We thank the participants who generously gave their time and biological samples for this study. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Department of Veteran Affairs.

Author information




J.D., L.K., T.H. and M.P.S. conceived and designed the study; J.D., L.K., J.H., T.H., M.P.S. and S.D. carried out the methods; J.D., R.R., X.L. and A.B. curated the data; J.D., L.K., R.R. and D.W. analyzed the data; J.D., L.K., J.H., S.M.S.-F.R. and M.P.S. wrote the manuscript; and M.P.S., S.D. and J.D. obtained funding.

Corresponding authors

Correspondence to Jessilyn Dunn or Trevor Hastie or Michael P. Snyder.

Ethics declarations

Competing interests

M.P.S. is a cofounder of Personalis, SensOmics, Qbio, January AI, Filtricine, Protos, and NiMo, and is on the scientific advisory board of Personalis, SensOmics, Qbio, January AI, Filtricine, Protos, NiMo, and Genapsys. All other authors have no competing interests.

Additional information

Peer review information Nature Medicine thanks Hugo Aerts and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Michael Basson was the primary editor on this article and managed its editorial process and peer review in collaboration with the rest of the editorial team.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Wearables temperature variations and extended modeling results.

a, Variations in wRTemp over course of the day. b, R statistics based on LOOCV for all tests from Fig. 3b. c, R statistics based on K-fold CV for all tests from Fig. 3b.

Extended Data Fig. 2 Model accuracy changes over time based on window of historical data from an individual.

a, Lasso regularized regression using features calculated using different windows of wearable device monitoring. b, Accuracy of the HCT cVS mixed effects models over time for two example patients that were monitored between 2.5–5 years at Stanford hospital with >50 HCT observations at separate clinic visits. The HCT cVS mixed effects models demonstrate that the model accuracy changes over time, and particularly with a dramatic health event like a myocardial infarction (ICD code I21.4) (red vertical line) or a life-threatening ED visit (blue vertical line; CPT code 99285).

Extended Data Fig. 3 Increasing amounts of personalized data open up new study and model possibilities.

a, Summary of different biomedical data collection modalities and the typical amount of data they result in. b, Demonstration of how the amount and modality of data collection (longitudinal continuous vs. discrete measurements) constrain the type and complexity of models that can be built from the data.

Supplementary information

Source data

Source Data Fig. 1

This file contains a table of data underlying Fig. 1d,e.

Source Data Fig. 3

This file contains a table of data underlying Fig. 3a,d.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Dunn, J., Kidzinski, L., Runge, R. et al. Wearable sensors enable personalized predictions of clinical laboratory measurements. Nat Med 27, 1105–1112 (2021).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing