Assessment of Prolonged Physiological and Behavioral Changes Associated With COVID-19 Infection

Introduction: Long-term COVID symptoms marked by autonomic dysfunction¹ and cardiac damage² following COVID-19 infection have been noted for up to 6 months after symptom onset,³ but to date have not been quantified, to our knowledge. Previous studies have found that wearable data can improve real-time detection of viral illness⁴ or discrimination of individuals with COVID-19 vs other viral infections.⁵ Wearable devices provide a way to continuously track an individual’s physiological and behavioral metrics beginning when healthy (ie, before infection), during the course of infection, and recovery back to baseline. In this cohort study, we aimed to examine the duration and variation of recovery among COVID-19–positive vs COVID-19–negative participants.

Methods: DETECT (Digital Engagement and Tracking for Early Control and Treatment) is a remote, app-based, longitudinal research study enrolling adult participants from all over the US and collecting their wearable data to better understand individual changes associated with viral illness, including COVID-19. All participants provided informed consent electronically. The protocol for this study was reviewed and approved by the Scripps Office for the Protection of Research Subjects. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Results: For this analysis, our study population consisted of 234 COVID-19–positive individuals (mean [range] age, 45.3 [18-76] years; 164 women [70.9%]) and 641 COVID-19–negative symptomatic individuals (mean [range] age, 44.7 [19-75] years; 455 women [71.1%]). Individuals with COVID-19 took longer to return to their RHR (Figure, A and B), sleep (Figure, C and D), and activity (Figure, E and F) baselines compared with symptomatic individuals who were COVID-19 negative. This difference was most marked for RHR, with COVID-19–positive individuals initially experiencing a transient bradycardia followed by a prolonged relative tachycardia that did not return to baseline, on average, until 79 days after symptom onset. Step count and sleep quantity returned to baseline sooner than RHR at 32 and 24 days, respectively. During recovery, individuals with COVID-19 experienced different trajectories in the return of their RHR to their normal compared with COVID-19–negative individuals (Figure, B). A small subset of COVID-19–positive participants (32 participants [13.7%]) maintained an RHR more than 5 beats per minute greater than their baseline RHR that did not return to their normal for more than 133 days. During the acute phase of COVID-19, individuals in this group reported higher frequencies of cough (27 participants [84.4%] vs 57 participants [55.3%] in the <1 beat per minute group and 57 participants [57.6%] in the 1-5 beats per minute group), body ache (20 participants [62.5%] vs 42 participants [40.8%] in the <1 beat per minute group and 35 participants [35.4%] in the 1-5 beats per minute group), and shortness of breath (9 participants [28.1%] vs 9 participants [8.7%] in the <1 beat per minute group and 6 participants [6.1%] in the 1-5 beats per minute group) compared with the other groups (Table).

Discussion: To our knowledge, this is the first study to examine longer duration wearable sensor data. We found a prolonged physiological impact of COVID-19 infection, lasting approximately 2 to 3 months, on average, but with substantial intraindividual variability, which may reflect various levels of autonomic nervous system dysfunction or potentially ongoing inflammation. Transient bradycardia has been noted in a case study⁶ approximately 9 to 15 days after symptom onset, which was also seen in our population. Our data suggest that early symptoms and larger initial RHR response to COVID-19 infection may be associated with the physiological length of recovery from this virus.