Abstract
BACKGROUND: The relationship between severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) viral dynamics during acute infection and the development of long coronavirus disease 2019 (COVID-19), or "long COVID," is largely unknown. METHODS: Between October 2021 and February 2022, 7361 people not known to have COVID-19 self-collected nasal swab samples for SARS-CoV-2 reverse-transcription polymerase chain reaction testing every 24-48 hours for 10-14 days. Participants whose first known SARS-CoV-2 infection was detected were surveyed for long COVID in August 2023. Their slopes of viral clearance were modeled using linear mixed effects models with random slopes and intercepts, and the relative risk (RR) of long COVID based on viral slopes was calculated using a log binomial model, adjusted for age, symptoms, and variant. Sex-based interaction terms were also evaluated for significance. RESULTS: A total of 172 participants were eligible for analyses, and 59 (34.3%) reported long COVID. The risk of long COVID with 3-4 symptoms (adjusted RR, 2.44 [95% confidence interval, .88-6.82]) and ≥5 symptoms (4.97 [1.90-13.0]) increased with each unit increase in slope of viral clearance. While the probability of long COVID increased with slowed viral clearance among women, the same relationship was not observed among men (interaction term: P = .02). Acute SARS-CoV-2 symptoms of abdominal pain (adjusted RR, 5.41 [95% confidence interval, 2.44-12.0]), nausea (3.01 [1.31-6.89]), and body aches (2.58 [1.26-5.30]) were most strongly associated with long COVID. CONCLUSIONS: We observed that slower viral clearance rates during acute COVID-19 were associated with increased risk and more symptoms of long COVID . Early viral-host dynamics appear to be mechanistically linked to the development of long COVID.