Smartwatch-derived sleep and heart rate measures complement step counts in explaining established metrics of MS severity

Tim Woelfle; Silvan Pless; Óscar Reyes; Andrea Wiencierz; Ludwig Kappos; Cristina Granziera; Johannes Lorscheider

doi:10.1016/j.msard.2023.105104

Smartwatch-derived sleep and heart rate measures complement step counts in explaining established metrics of MS severity

Mult Scler Relat Disord. 2023 Dec:80:105104. doi: 10.1016/j.msard.2023.105104. Epub 2023 Oct 24.

Authors

Tim Woelfle¹, Silvan Pless², Óscar Reyes³, Andrea Wiencierz⁴, Ludwig Kappos⁵, Cristina Granziera⁶, Johannes Lorscheider²

Affiliations

¹ Research Center for Clinical Neuroimmunology and Neuroscience Basel (RC2NB), Switzerland; Department of Neurology and MS Center, University Hospital Basel, Switzerland; Translational Imaging in Neurology (ThINk) Basel, Department of Biomedical Engineering, University Hospital, University of Basel, Switzerland. Electronic address: tim.woelfle@usb.ch.
² Research Center for Clinical Neuroimmunology and Neuroscience Basel (RC2NB), Switzerland; Department of Neurology and MS Center, University Hospital Basel, Switzerland.
³ Healios AG, Basel, Switzerland.
⁴ Department of Clinical Research, University Hospital, University of Basel, Switzerland.
⁵ Research Center for Clinical Neuroimmunology and Neuroscience Basel (RC2NB), Switzerland.
⁶ Research Center for Clinical Neuroimmunology and Neuroscience Basel (RC2NB), Switzerland; Department of Neurology and MS Center, University Hospital Basel, Switzerland; Translational Imaging in Neurology (ThINk) Basel, Department of Biomedical Engineering, University Hospital, University of Basel, Switzerland.

PMID: 37913676
DOI: 10.1016/j.msard.2023.105104

Abstract

Background: Passive remote monitoring of patients with MS (PwMS) with sensor-based wearable technologies promises near-continuous evaluation with high ecological validity. Step counts correlate strongly with traditional measures of MS severity. We hypothesized that remote monitoring of sleep and heart rate will yield complementary information.

Methods: We recruited 31 PwMS and 31 age- and sex-matched healthy volunteers (HV) as part of the dreaMS feasibility study (NCT04413032). Fitbit Versa 2 smartwatches were worn for 6 weeks and provided a total of 25 features for activity, heart rate, and sleep. Features were selected based on their pairwise intercorrelation (Pearson |r| < 0.6), test-retest reliability (intraclass correlation coefficient ≥ 0.6 or median coefficient of variation < 0.2) and group comparisons between HV and PwMS with moderate disability (expanded disability status scale (EDSS) ≥ 3.5) (rank-biserial |r| ≥ 0.5). These selected features were correlated with clinical reference tests (EDSS, timed 25-foot walk (T25FW), MS-walking scale (MSWS-12)) in PwMS, and multivariate models adjusted for age, sex, and disease duration were compared.

Results: We analyzed 28 PwMS (68% female, mean age 44 years, median EDSS 3.0) and 26 HV in our primary analysis. The objectively selected features discriminated well between HV and PwMS with moderate disability with rank-biserial r = 0.83 for Total number of steps, 0.51 for Deep sleep proportion, -0.51 for Median heart rate, 0.85 for Proportion very active, and 0.65 for Total number of floors. In PwMS they correlated strongly with the three clinical reference tests EDSS (strongest Spearman ρ = -0.75 for Proportion very active), T25FW (-0.75 for Total number of floors), and MSWS-12 (-0.72 for Total number of floors). Deep sleep proportion and Median heart rate complemented Total number of steps in explaining the variance of reference tests.

Conclusions: Activity, deep sleep and heart rate measures can be derived reliably from smartwatches and contain independent clinically meaningful information about MS severity, highlighting their potential for continuous passive monitoring in both clinical trials and clinical care of PwMS.

Keywords: Digital biomarkers; Digital health technology; Multiple sclerosis; Outpatient monitoring; Smartwatch; Wearable electronic devices.

MeSH terms

Adult
Disability Evaluation
Female
Heart Rate
Humans
Male
Multiple Sclerosis* / diagnosis
Reproducibility of Results
Walking / physiology