Association of the insulin resistance marker TyG index with the severity and mortality of COVID-19

This study retrospectively analysed confirmed COVID-19 from Jan 12 2020, to Feb 13 2020 at Tongji Hospital of Huazhong University of Science and Technology in Wuhan, China. All patients enrolled in this study were diagnosed as COVID-19 according to WHO interim guidance [11]. Patients with missing data on clinical characteristics were excluded. Considering the complexity of the disease, as well as the uncertainty of the course of treatment, 151 patients which had completed medical records and follow up data were included. This study was approved by the Ethics Commission of Tongji Hospital, and all participants provided oral informed consent.

We reviewed the epidemiological, demographic, clinical, and laboratory data from patients’ medical records. Two researchers checked the data collected independently. We collected information on socioeconomic characteristics (age, sex, exposure history), smoking, chronic medical histories (diabetes, hypertension, cardiovascular disease, cerebrovascular disease, chronic kidney disease, chronic pulmonary disease, chronic liver disease), clinical symptoms (fever, cough, expectoration, dyspnoea, chest pain, diarrhea, headache, hypodynamia, anorexia, nausea and vomiting), vital signs (temperature, heart rate, respiratory rate, oxygen saturation, blood pressure), laboratory data (complete blood count, liver and renal function, fasting plasma glucose (FPG), hemoglobin A1c (HbA1c), lipid, high-sensitivity C-reactive protein (CRP), Ferritin, and tumor necrosis factor α (TNFα)), as well as living status during study period.

The TyG index was calculated by the formula ln [fasting triglycerides (mg/dL) × FPG (mg/dL)/2]. The patients were divided into 3 tertiles according to TyG index levels, T1 (n = 50, 7.5 ≤ TyG index ≤ 8.6), T2 (n = 51, 8.7 ≤ TyG index ≤ 9.1), T3 (n = 50, 9.2 ≤ TyG index ≤ 10.7). Diabetes was defined as self-reported medical history of diabetes, and the use of antidiabetic drugs. Hypertension was defined as self-reported history of hypertension, and the use of antihypertensive drugs. Other chronic diseases, such as cardiovascular disease, cerebrovascular disease, chronic kidney disease, chronic pulmonary disease, and chronic liver disease, were diagnosed according to self-reported medical history.

Severity of the disease was staged according to the guidelines for diagnosis and treatment of COVID-19 (trial sixth edition) published by National Health Commission of China on February 18, 2020. Severe case was defined as including one criterion as follow: 1. respiratory rate > 30/min, 2. oxygen saturation ≤ 93%, 3. PaO2/FiO2, 4. Patients developed either with shock, or respiratory failure requiring mechanical ventilation, or combined with the other organ failure admission to intensive care unit (ICU).

We presented continuous variables as mean ± standard deviation (SD) for normally distributed data or median (interquartile range, IQR) for data with skew distribution, and categorical variables as frequencies (percentage,  %). We assessed differences between mild and severe cases using two-sample T test or Mann–Whitney U test depending on parametric or nonparametric data for continuous variables and Chi square test for categorical variables. Comparisons among TyG index groups were assessed by the analysis of variance (ANOVA) or Kruskal–Wallis test. Univariate and multivariate logistic regression analyses were used to evaluate the associations between TyG index and the severity and mortality of COVID-19. The associations of TyG index with severity of COVID-19 in subgroups classified by age (< 60 vs. ≥ 60 years), sex (male vs. female), and medical history of chronic disease (presence vs. absence), including diabetes mellitus and hypertension, were also studied using logistic regression analysis. We used SPSS software, version 19.0 (SPSS, Inc., Chicago, IL, USA) for all analyses with statistically significant for P < 0.05.

Epidemiological evidence have shown a high mortality and morbidity of diabetic patients who develop COVID-19 [4]. Our data also observed a high prevalence of diabetes in COVID-19 patients. Notably, we found that the TyG index, a reliable surrogate marker of IR, were predictive of severe cases and mortality of COVID-19 patients. Previous studies have shown that the TyG index may predict the development of type 2 diabetes mellitus [12] and metabolic syndrome [13]. It has been reported that the TyG index is positively associated with the higher prevalence of subclinical coronary artery disease (CAD) [14], symptomatic CAD and metabolic and behavioral risk factors [15], and arterial stiffness and nephric microvascular damage [16]. Moreover, higher TyG index correlated with increased risk of main adverse cardiovascular and cerebrovascular events (MACCE) in acute ST-elevation myocardial infarction (STEMI) patients, and the TyG index might be a valuable predictor of clinical outcomes in STEMI patients undergoing percutaneous coronary intervention [17]. Moreover, an elevated TyG index was significantly associated with a higher risk of arterial stiffness and nephric microvascular damage. The positive association of TyG index with increased risk of developing severe cases and the risk of mortality in confirmed COVID-19 patients, could be attributed to the roles of IR in metabolic disorder, an altered inflammation profile, and cardiovascular disease risk factors. These findings support the significance of TyG index as a valuable predictor of poorer outcomes for COVID-19 patients, and remind physicians that more intensive attention should be paid to patients with high level of TyG index.

The virus of COVID-19 was classified as beta-coronavirus, and had highly pathogenicity similar to SARS-CoV and middle east respiratory syndrome coronavirus (MERS-CoV), that could cause severe respiratory syndrome in humans [18‐20]. Yet, the fundamental pathophysiology of COVID-19 remains unknown. Previous studies have shown that increased concentrations of proinflammatory cytokines in serum (e.g., IL1B, IL6, IL12, IFNγ, IL10, and TNFα) were induced in patients with SARS [21] and MERS [22]. Considering the large amount of cytokines induced by SARS-CoV and MERS-CoV, a cytokine storm in the body might play a crucial role in pulmonary inflammation and extensive lung damage of patients with COVID-19. In early studies, COVID-19 was also reported to induce increased secretion of T-helper-2 (Th2) cytokines (e.g., IL4 and IL10) that suppress inflammation [4]. We noted that severe patients of COVID-19 had high amounts of inflammatory indicators (Ferritin and TNFα) in severe patients than mild ones, suggesting that the cytokine storm was associated with disease severity.

Consistent with previous reported, older patients were more susceptible to COVID-19, and 70% of infected patients were men [1]. In the present study, similar results were observed in age, yet there is no difference in the prevalence of male or female in total patients with COVID-19. Notably, the present study also emphasized a greater number of men among severe cases compared with mild cases. These findings suggest that older, male patients was more likely to have a worse prognosis. In addition to the elderly and males, we found that patients with a history of diabetes and hypertension are at increased risk of severe COVID-19 and mortality, in agreement with previous studies [4, 23]. Nevertheless, we observed a higher prevalence of diabetes and hypertension than that reported in the above study. The high prevalence of comorbid conditions may be expected for Tongji Hospital being the designated hospital for severe cases.

We found TyG index was closely associated with the mortality of COVID-19, which may be useful to predict the poor outcome for the infected patients. Although we attempted to adjust for many confounders, it is difficult to assess the independently cause of death among the 33 dead cases. The causes of death may be respiratory failure, heart failure, inflammation storm, sepsis, disseminated intravascular coagulation, multiple organ failure, acute kidney injury, malignant arrhythmia, diabetic ketoacidosis, even sudden death, and other uncertain events. Taken together, it is not easy to identify the relationship between TyG index and any cause of death in the world.

This study has several strengths. First, the COVID-19 was first observed in Wuhan China. A large number of patients are suffering with COVID-19 in this city, which provided us an opportunity to evaluate the epidemic affair. Second, we benefited from obtaining accurate and thorough clinical data of patients in Tongji Hospital, which is a designated hospital responsible for the treatments for moderate-severe COVID-19 patients in Wuhan. Third, all authors are first line doctors, who could directly observe participants.

Several limitations in this study should also be addressed. First, considering the complexity of the disease, as well as the limited information on the course of disease, the number of samples is relatively small. Second, compared with published paper, the percentage of severe cases was high in our study. It may be explained by that Tongji Hospital is the designated hospital for moderate-severe cases. Third, most people were hospitalized for treatment of pneumonia, resulting that metabolic disorders were easily overlooked. Thus, we could not observe the association of metabolic control with the prognosis of COVID-19. Fourth, the TyG index, which predicted the death of COVID-19 in this study, might be affected by the change of triglycerides and glucose levels. Other factors, such as inflammation, should also be taken into consideration when predicting the death of a patient.