SARS-CoV-2 lineage-specific disease symptoms and disease severity in a city in southeastern Brazil

In 2020, Sao Caetano do Sul city, located in the metropolitan region of Sao Paulo State, Brazil, established a web-based platform to provide primary care to suspected COVID-19 patients, integrating clinical and demographic data and sample metadata. Here we describe lineage-specific spatiotemporal dynamics of infections, clinical symptoms, and disease severity during the first year of the epidemic, which included circulation of the poorly characterised Gamma variant of concern. From April 6, 2020, to April 30, 2021, we gathered clinical, demographic, spatial and epidemiological data from the city's platform. We selected and sequenced 879 PCR+ swab samples (8% of all reported cases), obtaining a spatially and temporally representative set of sequences. Daily lineage-specific prevalence was estimated with a moving-window approach, allowing inference of cumulative cases and symptom probability stratified by lineage using integrated data from the platform. Most infections were caused by B.1.1.28 (41.3%), followed by Gamma (31.7%), Zeta (9.6%), and B1.1.33 (9.0%). Gamma and Zeta correlated with larger prevalence of dyspnoea (respectively, 81.3% and 78.5%) and persistent fever (84.7% and 61.1%) compared with B.1.1.28 and B.1.1.33. Ageusia, anosmia, and coryza were respectively 18.9%, 20.3%, and 17.8% less commonly caused by Gamma, whereas altered mental status was 108.9% more common in Zeta. Case incidence was spatially heterogeneous and larger in poorer and younger districts. Our study reveals that Gamma was associated with more severe presentation of the disease, emphasising its role in the heightened mortality levels in Brazil.