Identifying Long Covid phenotypes and their association with personal characteristics, healthcare use, and daily life burden: population-based study in Belgium

Long Covid (LC), a multisystem disorder with persistent symptoms > 4 weeks post COVID-19, has impacted healthcare systems substantially. As information is scarce for Belgium, this study aimed to identify LC phenotypes and assess participant characteristics, healthcare use and daily life burden by phenotype. The last survey wave of COVimpact (a longitudinal online cohort study among Belgian adults recruited after SARS-CoV-2 infection) focused on LC with questions on healthcare utilization and perceived daily impact. A latent class analysis (LCA) identified phenotypes, based on symptom type, disease duration, and LC-related disability. Backward multivariable multinomial logistic regression explored predictors of LC class membership. Among 1,840 respondents self-reporting LC, four phenotypes emerged. Class 1 presented mild clinical outcomes, class 2 and 3 presented moderate clinical outcomes, differentiated by symptoms of memory problems and brain fog (class 2), and respiratory problems and muscle/joint pain (class 3), and class 4 included the most severe clinical outcomes. Compared to class 1, being older, female, lower educated, non-European, obese and experiencing moderate to severe acute COVID-19 symptoms predicted higher class severity. Diagnosis, healthcare use and support differed among adults with moderate and severe LC, many reported insufficient healthcare access and major disruption to daily life, including absenteeism from work/school. People with severe LC were the most financially impacted. By distinguishing LC phenotypes, this study enhances understanding of varying healthcare trajectories and daily life impacts. Healthcare planning and support should be more effectively tailored to the specific needs of those affected by LC.