COVID-19 has significant long-term impacts, including a chronic syndrome known as long-COVID, characterized by persistent symptoms post-recovery. The inflammatory response during acute infection is hypothesized to influence long-term outcomes. This study aimed to identify inflammatory biomarkers predictive of functional outcomes one year after hospital discharge. A prospective cohort study was conducted with 213 COVID-19 patients admitted to ICUs in Southern Brazil between June and November 2020. After exclusions and follow-ups, 109 patients were evaluated for one-year post-discharge. Plasma levels of Th1 (TNF-α, INF-γ, IL-12), Th2 (IL-4, IL-5, IL-6, IL-10, IL-13), and Th17 (IL-17, IL-22) cytokines were measured. Functional outcomes in psychiatric, cognitive, general health, and health perception domains were assessed. Statistical analyses included multivariate regression, regularized partial correlation network analysis, and K-means clustering. We demonstrate that plasma levels of various cytokines, along with demographic and clinical characteristics, can predict four distinct domains of functional outcomes one year following hospital discharge due to COVID-19 and that an hyperinflammatory phenotype was associated with the occurrence of a worse in psychiatric, general health, and health perception domains. The network analysis highlighted complex interconnections among immune markers and clinical variables, elucidating their roles in long-term health. These findings support using biomarkers for patient stratification and indicate potential targets for therapeutic interventions.