Background
SARS-CoV-2-host dynamics during acute infection appear to be important in the development of long COVID. However, prior studies were limited by cohort size and infrequent sampling during acute infection, and no study has addressed the association of viral rebound and long COVID. Here we present data from a large cohort study that tested people without COVID-19 on a daily or every-other-day basis to catch the onset and clearance of SARS-CoV-2 RNA from the upper respiratory tract with unprecedented resolution, facilitating a more definitive investigation of early virus-host dynamics and their association with long COVID.
Methods
Between Oct. 2021 and Feb. 2022, 7,361 people self-collected nasal specimens for SARS-CoV-2 RT-PCR testing every 24 or 48 hours for 10 or 14 days in the Test Us At Home studies. In August 2023 (18-22 months after enrollment), 244 participants with first SARS-CoV-2 infection and >1 positive test were asked whether they had ever had long COVID, defined as new COVID-associated symptoms present for 3+ months post-infection. The relative risk of developing long COVID was calculated using a log binomial model with modeled viral clearance slopes as the exposure adjusted for age, number of symptoms, and SARS-CoV-2 variant. Viral rebound was defined as testing positive, then negative, then positive again. In the viral rebound analysis, only participants with ≥3 tests were included (n=215). Logistic models were used to examine associations among symptoms, demographics, viral rebound, and long COVID.
Results
Slopes of viral clearance were able to be modeled for 172 participants, of whom 59 (34%) ever had long COVID. Participants who developed long COVID had significantly longer times from peak viral load to viral clearance than those who never developed long COVID (10.0 days [95% CI: 9.25-10.8] vs 8.7 [95% CI: 8.28-9.01]). Slopes were significantly associated with four post-acute symptoms: fatigue (aRR: 2.9), brain fog (aRR: 4.94), shortness of breath (aRR: 5.05), and gastrointestinal symptoms (aRR: 54.6) (Figure). Of the 215 participants assessed for viral rebound, 77 (36%) ever had long COVID and 75 (35%) experienced viral rebound. The development of long COVID was associated with 13-fold higher odds of having experienced rebound during acute infection (OR: 13.1; 95% CI: 6.79-26.2).
Conclusions
We demonstrate here that slow viral clearance and viral rebound during acute infection are associated with long COVID.