UC San Diego

Long COVID is a chronic condition characterized by symptoms such as fatigue, dyspnea, and cognitive impairment that persist or relapse months after an acute infection with the SARS-CoV-2 virus. Many distinct symptoms have been attributed to Long COVID; however, little is known about the potential clustering of these symptoms and risk factors that may predispose patients to certain clusters. In this study, an electronic survey was sent to patients in the UC San Diego Health (UCSDH) system who tested positive for COVID-19, querying if patients were experiencing symptoms consistent with Long COVID. Based on survey results, along with patient demographics reported in the electronic health record (EHR), linear and logistic regression models were used to examine putative risk factors, and exploratory factor analysis was performed to determine symptom clusters. Among 999 survey respondents, increased odds of Long COVID (n = 421; 42%) and greater Long COVID symptom burden were associated with female sex (OR = 1.73, 99% CI: 1.16-2.58; β = 0.48, 0.22-0.75), COVID-19 hospitalization (OR = 4.51, 2.50-8.43; β = 0.48, 0.17-0.78), and poorer pre-COVID self-rated health (OR = 0.75, 0.57-0.97; β = -0.19, -0.32--0.07). Over one-fifth of Long COVID patients screened positive for depression and/or anxiety, the latter of which was associated with younger age (OR = 0.96, 0.94-0.99). Factor analysis of 16 self-reported symptoms suggested five symptom clusters-gastrointestinal (GI), musculoskeletal (MSK), neurocognitive (NC), airway (AW), and cardiopulmonary (CP), with older age (β = 0.21, 0.11-0.30) and mixed race (β = 0.27, 0.04-0.51) being associated with greater MSK symptom burden. Greater NC symptom burden was associated with increased odds of depression (OR = 5.86, 2.71-13.8) and anxiety (OR = 2.83, 1.36-6.14). These results can inform clinicians in identifying patients at increased risk for Long COVID-related medical issues, particularly neurocognitive symptoms and symptom clusters, as well as informing health systems to manage operational expectations on a population-health level.