Ikaro Silva

Infectious threats, like the COVID-19 pandemic, hinder maintenance of a productive and healthy workforce. If subtle physiological changes precede overt illness, then proactive isolation and testing can reduce labor force impacts. This study hypothesized that an early infection warning service based on wearable physiological monitoring and predictive models created with machine learning could be developed and deployed. We developed a prototype tool, first deployed June 23, 2020, that delivered continuously updated scores of infection risk for SARS-CoV-2 through April 8, 2021. Data were acquired from 9381 United States Department of Defense (US DoD) personnel wearing Garmin and Oura devices, totaling 599,174 user-days of service and 201 million hours of data. There were 491 COVID-19 positive cases. A predictive algorithm identified infection before diagnostic testing with an AUC of 0.82. Barriers to implementation included adequate data capture (at least 48% data was needed) and delays in data transmission. We observe increased risk scores as early as 6 days prior to diagnostic testing (2.3 days average). This study showed feasibility of a real-time risk prediction score to minimize workforce impacts of infection.

Rapid and early identification of emergent infections is essential for delivering prompt clinical care. To advance the development of algorithms for the clinical management of infection identification, we performed a vaccination clinical trial to investigate the potential of using vaccination as a model for studying mild inflammation responses associated with different infections (NCT05346302). We collected data at various time points over 4 weeks from blood samples, wearable devices, and questionnaires. Following a 2-week baseline period, 210 healthy participants, aged 18-40 years, were administered either a Pneumococcal Polysaccharide vaccine (PPSV23), Typhoid Vi Polysaccharide vaccine (Typhim Vi), or placebo. In longitudinal analyses of blood biomarkers, we found that CRP was significantly higher at 2 days post-vaccination, whereas basophils, IL-10, IL-12p40, and MIG were significantly higher at 7 days post-vaccination in the PPSV23 group compared to both other groups (all p < 0.05). MIP-1β was significantly lower in the PPSV23 group than in the placebo group, while monocytes and MPV were significantly lower in the Typhim Vi group than in the placebo group at 7 days post-vaccination (all p < 0.05). The PPSV3 group showed a higher inflammatory profile, suggesting that PPSV23 induces a stronger immune response compared to Typhim Vi. The distinct immune responses induced by the two vaccines indicate the potential for utilizing vaccines as models for studying inflammation responses associated with different infectious pathogens.

Inadequate sleep has been associated with an increased risk of mortality and various health issues. We previously conducted a placebo-controlled vaccination trial of healthy adults who were monitored by blood samples, questionnaires, and wearable devices. C-reactive protein (CRP), a systemic marker of inflammation, has been linked to numerous health outcomes, and was found to significantly increase post-vaccination in the trial. In this retrospective study, we investigated that if sleep was associated with an inflammation response triggered by perturbations from vaccine and placebo injections. Plasma hs-CRP levels were measured on the same day as the intervention, prior to the vaccine/placebo administration and two days after the intervention. Associations of sleep duration and CRP levels after vaccine/placebo administration in 188 trial participants were investigated by regression models adjusting for age, sex, body mass index (BMI), comorbidities, vaccination status (vaccination or placebo), and averaged daily steps. We found that shorter wearable-derived Total Sleep Time (TST) and Total Time in Bed (TIB), as well as subjectively assessed sleep duration from the Pittsburgh Sleep Quality Index (PSQI), were independently associated with higher incidence of CRP elevation after vaccine/placebo administration. Our study suggests that sleep deprivation could be a predictor for an increased inflammatory response and highlights a potential application of wearable-derived sleep metrics in public health.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can lead to post-acute sequelae of SARS-CoV-2 infection (PASC), or Long COVID, a chronic multisystemic condition with diverse symptoms and no objective diagnostic test. In this retrospective study, we developed a data-driven method to objectively detect persistent physiological changes using wearable device data in a large cohort of over 12,000 US military personnel. We analyzed physiological data from 663 symptomatic COVID-19 positive cases and 2,513 asymptomatic COVID-19 negative controls. Our method identified persistent physiological changes in 9.4% of COVID-19 positive individuals, most commonly manifesting as elevated nightly heart rate and reductions in some heart rate variability metrics. Our findings demonstrate that wearable technology can be used to objectively detect chronic physiological changes beyond the acute phase of COVID-19 illness. Although our method requires further clinical validation, it could potentially provide objective metrics to help standardize Long COVID diagnosis criteria.