This study investigates the relationship between clinical, sociodemographic, and neuropsychological symptoms and serum cytokine concentrations with long-term cognitive and psychiatric outcomes in long-COVID-19 patients. We reassessed 108 adults who survived moderate to severe COVID-19 at two intervals post-discharge (T1, mean 6.9 months; T2, mean 23.5 months). Baseline sociodemographic and clinical data were collected from hospital records, while cognitive and mental health assessments included psychometric tests such as the Hospital Anxiety and Depression Scale (HADS) and Immediate and Delayed Recall Tests from the CERAD Battery. Serum cytokine levels were measured at T1. Generalized Additive Models (GAMs), Elastic Net Regression (NET), and Psychological Network Analysis (PNA) were used to analyze the data. The GAM analysis revealed significant associations between acute COVID-19 severity and Epworth Sleepiness Score with persistent anxiety symptoms at T2. For depression, both WHO severity class and Eotaxin levels were significant predictors. The Anti-inflammatory Index showed a marginally significant relationship with immediate recall, while age was marginally associated with delayed recall performance. In NET, only anxiety was significantly associated with Epworth Sleepiness Score, WHO severity class, and Proinflammatory Index. PNA did not reveal direct connections between cytokines and neuropsychological outcomes in the graphical model. However, centrality measures indicated that the Proinflammatory Index and VEGF were more central within the network, suggesting they might be important components of the overall system. This study provides insights into the complex role of cytokines and inflammation in long-COVID-19 outcomes, potentially aiding in the identification of biomarkers for diagnosis and prognosis.
