High-density lipoproteins and clinical outcomes of COVID-19 in hemodialysis patients: A multicenter, propensity-score matched case-control study

A decline in high-density lipoprotein (HDL) has been observed in the general population of COVID-19 infected patients. However, little research has been devoted to this topic in hemodialysis (HD) patients and there is a general lack of data on the association between the baseline HDL level and COVID-19 outcomes in HD patients. The present study aimed to assess the association between baseline HDL levels and the risk of developing severe outcomes of COVID-19 infection in HD patients. Methods. A total of 428 HD patients aged 55 (44-64) years and a dialysis vintage of 44 (21-76.6) months were enrolled in this multicenter propensity-score matched case-control study. After propensity-score matching (1 :1), 142 COVID-19 infected HD patients were compared with 142 HD patients who had never been infected with COVID-19. Propensity scores were estimated with MedCalc software based on potential confounding variables (age, diabetes, dialysis vintage and cardiovascular events). Baseline HDL levels were obtained from electronic health records of the patients (the last measurements were carried out before the COVID-19 onset). The endpoints of the study were COVID-19-related hospitalization, oxygen maintenance, and death. The data were presented as the median and the interquartile ranges [Me (Q25-Q75)] and compared using the Mann-Whitney or Kruskal-Wallis tests, as appropriate. The Spearmen correlation test was used to assess the association between HDL and other markers. The multivariate logistic regression and the receiver operating characteristic (ROC) analyses were performed to evaluate the ability of HDL for predicting the severity of COVID-19 infection. The Cox regression analysis was used to determine the association between HDL and COVID-19-associated mortality. Results. Among 142 patients infected with COVID-19, 40 patients (28%) patients were hospitalized, 34 patients (24%) needed oxygen supplements and 16 patients (11.3%) died. The baseline HDL level was significantly lower in the COVID-19 patients compared with the non-infected patients (1.22 (0.97-1.59) vs 1.52 (1.08-1.85) mmol/L, p = 0.011). Moreover, the baseline HDL level was significantly negatively associated with CT findings of pulmonary involvement (r = -0.76, p < 0.0001) and had a strong inverse correlation with serum C-reactive protein (r = -0.42, p = 0,0002) and D-dimer (r = -0.65, p < 0.0001). The multivariate logistic regression analysis demonstrated that HD patients with HDL concentration <1.22 mmol/l have 3 times higher risk of COVID-19 infection (OR 2.8; 95% CI 1.2; 6.5, p = 0.003). In the Cox regression analysis adjusted for the patients' age, diabetic status, dialysis vintage and cardiovascular events, low HDL level (<1.22 mmol/L) was found to be associated with COVID-19-related mortality in the HD patients (χ2 = 9.2, p = 0.002). Conclusions. Baseline low HDL-C level was independently associated with COVID-19 morbidity, severity and mortality in HD patients.