Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events
Background: Wearable fitness trackers generate extensive physiological and activity data, offering potential to monitor health and predict outcomes. Machine learning (ML) techniques applied to these data may enable early identification of adverse health conditions, such as hospitalizations and devel...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-06-01
|
| Series: | American Journal of Preventive Cardiology |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666667725000819 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850151674264420352 |
|---|---|
| author | John Kundrick Aditi Naniwadekar Virginia Singla Krishna Kancharla Aditya Bhonsale Andrew Voigt Alaa Shalaby N.A. Mark Estes Sandeep K Jain Samir Saba |
| author_facet | John Kundrick Aditi Naniwadekar Virginia Singla Krishna Kancharla Aditya Bhonsale Andrew Voigt Alaa Shalaby N.A. Mark Estes Sandeep K Jain Samir Saba |
| author_sort | John Kundrick |
| collection | DOAJ |
| description | Background: Wearable fitness trackers generate extensive physiological and activity data, offering potential to monitor health and predict outcomes. Machine learning (ML) techniques applied to these data may enable early identification of adverse health conditions, such as hospitalizations and development of cardiovascular diseases (CVD). This study aimed to evaluate ML models' ability to forecast the incidence of (1) hospitalizations from any cause and (2) of new diagnosis of CVD, including a composite of heart failure (HF), coronary artery disease or myocardial infarction (CAD-MI), cardiomyopathy (CMP), and atrial fibrillation (AF). Method and Results: Data from 14,157 participants in the All of Us study that included both Fitbit and electronic health record (EHR) information were censored on the date preceding events and analyzed using various ML classifiers for extracted feature data. Performance metrics included accuracy, area under the receiver operating characteristic (AUROC) curve, and F1 scores. Our overall study population was young (median age 54 years), with good representation of women (67%). For hospitalizations, a Random Forest classifier achieved the best performance (AUROC=0.95, accuracy=0.99, F1 score=0.92). For the CVD events, the best prediction model was gradient boosting (AUROC=0.80, accuracy=0.71, F1 score=0.15).Conclusion: ML models applied to Fitbit data demonstrate promise in predicting clinical outcomes with strong performance for predicting all-cause hospitalizations and modest performance for predicting incident CVD. Wearable technology could play a role in risk assessment and patient management. |
| format | Article |
| id | doaj-art-fd2410410ec64d39badcb5f7bed12b09 |
| institution | OA Journals |
| issn | 2666-6677 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Elsevier |
| record_format | Article |
| series | American Journal of Preventive Cardiology |
| spelling | doaj-art-fd2410410ec64d39badcb5f7bed12b092025-08-20T02:26:09ZengElsevierAmerican Journal of Preventive Cardiology2666-66772025-06-012210100610.1016/j.ajpc.2025.101006Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular eventsJohn Kundrick0Aditi Naniwadekar1Virginia Singla2Krishna Kancharla3Aditya Bhonsale4Andrew Voigt5Alaa Shalaby6N.A. Mark Estes7Sandeep K Jain8Samir Saba9Heart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesHeart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesCorresponding author at: Division Chief, Cardiology, Co-Director, Heart and Vascular Institute, University of Pittsburgh Medical Center, 200 Lothrop Street, South Tower E355.6, Pittsburgh, PA 15213, United States.; Heart and Vascular Institute, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA, United StatesBackground: Wearable fitness trackers generate extensive physiological and activity data, offering potential to monitor health and predict outcomes. Machine learning (ML) techniques applied to these data may enable early identification of adverse health conditions, such as hospitalizations and development of cardiovascular diseases (CVD). This study aimed to evaluate ML models' ability to forecast the incidence of (1) hospitalizations from any cause and (2) of new diagnosis of CVD, including a composite of heart failure (HF), coronary artery disease or myocardial infarction (CAD-MI), cardiomyopathy (CMP), and atrial fibrillation (AF). Method and Results: Data from 14,157 participants in the All of Us study that included both Fitbit and electronic health record (EHR) information were censored on the date preceding events and analyzed using various ML classifiers for extracted feature data. Performance metrics included accuracy, area under the receiver operating characteristic (AUROC) curve, and F1 scores. Our overall study population was young (median age 54 years), with good representation of women (67%). For hospitalizations, a Random Forest classifier achieved the best performance (AUROC=0.95, accuracy=0.99, F1 score=0.92). For the CVD events, the best prediction model was gradient boosting (AUROC=0.80, accuracy=0.71, F1 score=0.15).Conclusion: ML models applied to Fitbit data demonstrate promise in predicting clinical outcomes with strong performance for predicting all-cause hospitalizations and modest performance for predicting incident CVD. Wearable technology could play a role in risk assessment and patient management.http://www.sciencedirect.com/science/article/pii/S2666667725000819Heart rateStep countHospitalizationCardiovascular diseasePrediction |
| spellingShingle | John Kundrick Aditi Naniwadekar Virginia Singla Krishna Kancharla Aditya Bhonsale Andrew Voigt Alaa Shalaby N.A. Mark Estes Sandeep K Jain Samir Saba Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events American Journal of Preventive Cardiology Heart rate Step count Hospitalization Cardiovascular disease Prediction |
| title | Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events |
| title_full | Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events |
| title_fullStr | Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events |
| title_full_unstemmed | Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events |
| title_short | Machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events |
| title_sort | machine learning applied to wearable fitness tracker data and the risk of hospitalizations and cardiovascular events |
| topic | Heart rate Step count Hospitalization Cardiovascular disease Prediction |
| url | http://www.sciencedirect.com/science/article/pii/S2666667725000819 |
| work_keys_str_mv | AT johnkundrick machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT aditinaniwadekar machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT virginiasingla machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT krishnakancharla machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT adityabhonsale machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT andrewvoigt machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT alaashalaby machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT namarkestes machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT sandeepkjain machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents AT samirsaba machinelearningappliedtowearablefitnesstrackerdataandtheriskofhospitalizationsandcardiovascularevents |