Abstract
BACKGROUND: Since the end of the COVID-19 pandemic, the potential roles of thyroid-inflammatory derangements in driving or being associated with the prognosis of COVID-19 remain controversial. We aimed to clarify the association between COVID-19 infection and thyroid dysfunction, and highlight the impacts of subsequent autoimmune thyroid disease (AITD) on the prognosis of COVID-19. METHODS: The retrospective, multicenter, cohort study enrolled 2,339 participants with COVID-19 from three hospitals located in the north, middle, and south regions of Shaan Xi Province, China, between December 2022 and July 2023. 464 non-COVID-19 patients within the same period were supplemented, divided into groups with and without AITD. At hospital admission (baseline), 3- and 6-month follow-ups, we presented a dynamic description and correlation analysis of thyroid-inflammatory-autoimmune derangements in patients with AITD. RESULTS: A total of 2,082 COVID-19 patients diagnosed with AITD and 257 cases without AITD were included in the study, and 464 non-COVID-19 patients were supplemented, dividing into 14 AITD and 450 non-AITD cases. We found that COVID-19 infection was closely associated with thyroid dysfunction (χ (2) = 1518.129, p = 0.000). AITD patients with COVID-19 showed a higher prevalence of symptoms and comorbidities and longer hospital stays at baseline than non-AITD patients with COVID-19 (p = 0.000, p = 0.000, and p = 0.000). The baseline free triiodothyronine (FT3), free thyroxine, and radioactive iodine uptake at 24 h in AITD cases significantly decreased (p = 0.000, p = 0.000, and p = 0.000), while thyroid stimulating hormone, thyroglobulin, reverse triiodothyronine (rT3), and thyroid antibodies varying elevated from the baseline to the follow-up (baseline: p = 0.000, p = 0.000, p = 0.000, p = 0.000, p = 0.000, and p = 0.000; 3-month follow-up: p = 0.000, p = 0.000, p = 0.000, p = 0.000, p = 0.030, and p = 0.000). C-reactive protein, calcitonin, interleukin-6, -8, -10, and tumor necrosis factor-α rose significantly at baseline (p = 0.000, p = 0.000, p = 0.000, p = 0.000, p = 0.000, and p = 0.000) in AITD. Interferon-α and interferon-γ at baseline showed a significant decrease (p = 0.000 and p = 0.000), and remained at low levels after 6 months (p = 0.000 and p = 0.000). FT3 and rT3 were positively and negatively correlated with hospitalization, respectively (r = -0.208 and 0.231; p = 0.000 and p = 0.000). ROC curves showed that FT3 and rT3 had better robustness in predicting severe COVID-19 prognosis (AUC = 0.801 and 0.705). Ordered logistic regression revealed that ORs were 0.370, 0.048, and 0.021 for AITD [(subacute thyroiditis, Grave's disease, and Hashimoto's thyroiditis compared to non-thyroidal illness syndrome (NTIS)] with COVID-19 risk, indicating that NTIS was the predominant risk factor for the severity of COVID-19. CONCLUSIONS: A robust association has been identified, wherein COVID-19 infection is closely associated with thyroid dysfunction, and the subsequent AITD may aggravate the poor prognosis of COVID-19.