Wearable Sensors Predict Exertion Without Changes in Pulse Rate for Frail End-Stage Kidney Disease Patients

End-Stage Kidney Disease (ESKD) patients require careful exertion monitoring during prehabilitation to ensure safety and efficacy. Traditional prehabilitation typically relies on heart rate to estimate physical exertion; however, this is problematic for ESKD patients due to their lower physical capacity and the high prevalence of beta-blocker prescriptions, which prevent typical heart rate responses during exercise. In this study, we investigated whether multi-modal wearable sensors could better predict the Rating of Perceived Exertion (RPE) in (N=13) ESKD patients and (N=13) healthy older adult controls. We collected continuous physiological data—including pulse rate, electrodermal activity, skin temperature, and accelerometer data—using a commercial smartwatch during prehabilitation sessions. Using individualized machine learning pipelines, we compared regression models built solely on pulse rate against comprehensive multi-feature models. We found distinct differences between the groups: baseline pulse-rate models performed substantially worse for the ESKD cohort, whereas they maintained predictive validity for controls. Conversely, multi-feature models significantly improved predictive performance for ESKD participants, achieving a Root Mean Square Error of 1.33. Feature selection confirmed that ESKD models relied heavily on physiological data beyond pulse rate to compensate for the lack of cardiovascular variability. These results demonstrate that heart rate alone is an insufficient biomarker for this population, and that remote exertion monitoring for ESKD patients requires distinct, multi-modal predictive algorithms rather than those designed for healthy adults.

Authors

Institutions

Publication Details

Journal
medRxiv
Published
2026-11-15
DOI
https://doi.org/10.64898/2026.11.12.26367919
Primary Topic
Non-Invasive Vital Sign Monitoring
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Wearable Sensors Predict Exertion Without Changes in Pulse Rate for Frail End-Stage Kidney Disease Patients

Elizabeth C. Lorenz, Momona Yamagami, Farzan Sasangohar, Atiya F. Dhala et al.
medRxiv
Non-Invasive Vital Sign Monitoring
preprint

Wearable Sensors Predict Exertion Without Changes in Pulse Rate for Frail End-Stage Kidney Disease Patients

Elizabeth C. Lorenz, Momona Yamagami, Farzan Sasangohar, Atiya F. Dhala, Mikayla E Deehring, H Philo Katzman, Rebecca Isac
preprint en

Abstract

End-Stage Kidney Disease (ESKD) patients require careful exertion monitoring during prehabilitation to ensure safety and efficacy. Traditional prehabilitation typically relies on heart rate to estimate physical exertion; however, this is problematic for ESKD patients due to their lower physical capacity and the high prevalence of beta-blocker prescriptions, which prevent typical heart rate responses during exercise. In this study, we investigated whether multi-modal wearable sensors could better predict the Rating of Perceived Exertion (RPE) in (N=13) ESKD patients and (N=13) healthy older adult controls. We collected continuous physiological data—including pulse rate, electrodermal activity, skin temperature, and accelerometer data—using a commercial smartwatch during prehabilitation sessions. Using individualized machine learning pipelines, we compared regression models built solely on pulse rate against comprehensive multi-feature models. We found distinct differences between the groups: baseline pulse-rate models performed substantially worse for the ESKD cohort, whereas they maintained predictive validity for controls. Conversely, multi-feature models significantly improved predictive performance for ESKD participants, achieving a Root Mean Square Error of 1.33. Feature selection confirmed that ESKD models relied heavily on physiological data beyond pulse rate to compensate for the lack of cardiovascular variability. These results demonstrate that heart rate alone is an insufficient biomarker for this population, and that remote exertion monitoring for ESKD patients requires distinct, multi-modal predictive algorithms rather than those designed for healthy adults.

medRxiv
Houston Methodist (US), Baylor College of Medicine (US), Houston Methodist Research Institute (US), Rice University (US), Texas A&M University (US)
Non-Invasive Vital Sign Monitoring
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon