Association Between Cystatin C, Cystatin C Rangeability and Mortality of COVID-19 Patients With or Without Type 2 Diabetes Mellitus: An Opportunistic Retrospective Analysis

Background: Since December of 2019, novel coronavirus (SARS-CoV-2)-induced pneumonia (COVID-19) exploded in Wuhan, and rapidly spread throughout China. Patients with COVID-19 demonstrated quite different appearances and outcomes in clinical manifestations. We aimed to gure out whether risk factors of the cystatin C (CysC) and the CysC rangeability are inuencing the prognosis of COVID-19 patients with or without type 2 diabetes mellitus (T2DM). Methods: 675 T2DM patients and 602 non-T2DM patients were divided into low CysC group, high CysC group and low CysC rangeability group, high CysC rangeability group according to the serum CysC level and the change range of CysC. Demographic characteristics, clinical data and laboratory results of the four groups were collected and analyzed. Results: Our data showed that COVID-19 patients with high CysC level and CysC rangeability had more organic damage and higher mortality rate compared to those with low level or low rangeability of CysC. Furthermore, patients with higher CysC level and CysC rangeability also demonstrated higher blood lymphocytes (lymph), C-reactive protein (CRP), alanine aminotransferase (ALT), aspartate aminotransferase (AST) which may greatly inuence disease progression and poor prognosis of COVID-19. After adjusting for possible confounders, multivariate analysis revealed that CysC ≤ 0.93mg/dl as a reference, CysC (cid:0) 0.93mg/dl were signicantly associated with the risk of heart failure (OR=2.401, 95% CI: 1.118–5.156) and all-cause death (OR=2.734, 95% CI: 1.098-6.811); referring to CysC rangeability ≤ 0, CysC rangeability (cid:0) 0 signicantly associated with all-cause death (OR=4.029, 95% CI: 1.864-8.706). Further grouped by T2DM, these associations were stronger in T2DM than in non-T2DM. Conclusions: It suggests that CysC level and CysC rangeability contribute to clinical manifestations and may inuence the prognosis of COVID-19. The CysC is considered as a potential risk factor of the prognosis of COVID-19. Special medical care and appropriate intervention should be performed in COVID-19 patients with elevated CysC during hospitalization and later clinical follow-up, especially for those with T2DM. on the the peak of higher in


Introduction
In late December 2019, in Wuhan City of China, many patients with unde ned pneumonia were reported to be infected with a novel coronavirus (SARS-CoV-2), formally named coronavirus disease 2019 (COVID- 19) by the World Health Organization (WHO) [1][2][3] . COVID-19 can be mainly spread by respiratory droplets and contact, aerosol transmission, as well as fecal-oral transmission 4 . As with highly infectious, this disease is listed as a class B infectious disease, recommended preventive and control measures similar to those against class A infectious diseases by "the law of the People's Republic of China on the prevention and treatment of infectious diseases". The epidemic continues and the population is expanding, which has brought great challenges to all countries in the world. forms from electronic medical records. The duration from the onset of disease to hospital admission and death was also recorded.

De nition
We assessed the impact of kidney injury severity on admission and changes in renal function during the course of disease were on the prognosis of 1277 patients with COVID-19. Kidney injury severity was measured by CysC level on admission, where the renal function changes were de ned as CysC rangeability, calculated as the difference of CysC level between the time of admission and the highest CysC level testing during hospitalization. We classi ed patients into T2DM and non-T2DM group to eliminate the affection of T2DM itself on kidney function. Diabetics and non-diabetics were hierarchically classi ed into 2 groups according to the median value of CysC on admission and CysC rangeability respectively. Severe in ammation was de ned as highest neutrophil-to-lymphocyte ratio (NLR) during hospitalization higher than 10. Liver injury was de ned as ALT or AST at any time during hospitalization greater than 80U/L. Heart failure was de ned as highest brain natriuretic peptide (BNP) during hospitalization greater than 300pg/ml.

Statistical Analysis
Continuous variables are presented as the mean ± standard deviation if normally distributed or median (lower quartile, upper quartile) otherwise, and Shapiro-Wilk test was used for normality test. Categorical variables are presented as numbers and percentages. Differences in the parameters among groups were analyzed using student's t-test for normally distributed variables, the Mann-Whitney U test was used for non-normally distributed continuous variables, and the chi-square test was used for categorical variables. Logistic regression models were used in univariate analyses and multivariate analyses to determine the prognostic value of CysC. Multivariate analyses adjusted for signi cant baseline variables and the factors closely related to the outcome of patients with cardiovascular disease, such as age, sex, severe pneumonia, serum albumin, blood glucose, Lg BNP.

Characteristics of hospitalized patients with COVID-19
A total of 1277 COVID-19-associated hospitalized adults from Wuhan city were included. Among the 1277 hospitalized adults, the median age was 63 years, (interquartile range [IQR], 51-70) ( Table 1). 9.4% of COVID-19 patients were diagnosed with severe pneumonia and 48% were males. There were 675 patients with T2DM and 602 patients with no T2DM. 8.5% and 10.2% severe pneumonia were diagnosed in non-T2DM and T2DM patients respectively (Table 1). The total population was divided into 2 groups by T2DM or not. Compared with non-T2DM patients, the value of age, hospital stay, the percentage of heart failure, severe in ammatory, liver injury in the hospital are higher in T2DM patients; The difference between the number of males and severe pneumonia are not signi cant. Refer to blood index, the value of CysC, lgBNP, lymphocyte, neutrophile, NLR, ALT, AST, blood glucose (Glu), alkaline phosphatase (ALP), choline kinase (CK), high-sensitivity C-reactive protein (hsCRP), D-Dimer on the baseline, the peak of NLR, CRP (C-reactive protein) are higher in T2DM group. Other than, the value of procalcitonin (PCT), albumin (ALB), creatinine, cardiac Troponin I (cTnI), creatine kinase isoenzymes (CK-MB) on the baseline, the peak of lymphocyte, neutrophile, PCT, ALT, AST, ALB, creatinine, ALP, CK, CK-MB, cTnI, CysC, hs-CRP, D-Dimer are not different between those 2 groups. It was worth mentioned mentioning that T2DM contributed to all-cause death. (Table 1).
Characteristics of T2DM or not patients grouped by CysC level T2DM and non-T2DM patients were divided into 2 groups by CysC level, respectively. Among the T2DM patients, the CysC high group has a higher percentage of severe pneumonia, heart failure, death, a higher value of age than the CysC low group. The difference between males, severe in ammatory, and hospital stay are not signi cant. Refer to blood index, the value of CysC, lgBNP, PCT, ALB, Glu, cTnI, hs-CRP, D-Dimer on the baseline, the peak of ALB is higher in T2DM group. The value of lymphocyte, neutrophile, Data are reported as mean ± SD, median (IQR) or number and percentage. T2DM, Type 2 diabetes mellitus; BNP, Brain Natriuretic Peptide; NLR, Neutrophil-to-Lymphocyte Ratio; PCT, procalcitonin; ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; ALB, albumin; Glu, glucose; Cre, creatinine; ALP, alkaline phosphatase; CK, choline kinase; CK-MB, creatine kinase isoenzymes; cTnI, hypersensitive troponin I; hsCRP, high-sensitivity C-reactive protein; DD, D-Dimer; Lymph, lymphocyte; Neutrop, Neutrophil; CRP, C-reactive protein; PCT, procalcitonin; ALT, alanine transaminase; CysC, Cystatin C.
As for the non-T2DM patients, the CysC high group has a higher percentage of males, heart failure, higher value of age than the CysC low group. The difference in severe pneumonia and hospital stay, liver injury, ( Table 2) Characteristics of patients with T2DM or not grouped by CysC rangeability T2DM and non-T2DM patients were divided into 2 groups by CysC rangeability, respectively. Among the T2DM patients, the CysC rangeability>0 group has a higher percentage of severe pneumonia, severe in ammatory, liver injury, a higher value of age than the CysC rangeability ≤ 0 group. The difference between the males, hospital stay, heart failure, and death is not signi cant. Refer to blood index, the value of CysC, lgBNP, lymphocyte, ALT, AST, ALB, cTnI, hs-CRP, D-Dimer on the baseline, the peak of CRP are higher in CysC rangeability>0 groups. The value of neutrophile, creatinine (Cre), PCT, ALP, CK, CK-MB on the baseline, the peak of ALB, lymphocyte, neutrophile, NLR, PCT, ALT, AST, Cre, ALP, CK, CK-MB, c TnI, CysC, hs-CRP, D-Dimer are not different between those 2 groups. (Table 3) In the non-diabetic population, the CysC rangeability>0 group has a higher percentage of severe pneumonia, higher value of age, hospital stay than CysC rangeability ≤ 0. The difference between liver injury, severe in ammatory, heart failure is not signi cant. Refer to blood index, the value of CysC, lgBNP, lymphocyte, ALT, AST, ALB, creatinine, CK, CK-MB, cTnI, hs-CRP, D-Dimer on the baseline, the peak of cre and ALT is higher in CysC rangeability>0 group. The value of neutrophile, PCT, CK, CK-MB, glucose on the baseline, the peak of lymphocyte, neutrophile, CRP, PCT, ALT, CysC, and hs-CRP CK, CK-MB, cTnI, D-Dimer are not different between those 2 groups. (Table 3) The association between CysC and Organ dysfunction, all-cause death classi ed by T2DM Multivariate logistic analysis for organ dysfunction and all-cause death showed that when CysC ≤ and liver injury (OR = 0.986, 95% CI:0.601-1.618) for T2DM patients. Whereas the relationship between CysC>0.93 mg/dl and severe in ammatory, liver injury, or heart failure was not signi cant in the non-T2DM group (Fig. 1).
When goes for sex subgroup of T2DM patients, adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP, there was a signi cantly increased contribution of CysC>0.93 mg/dl to all-cause death (OR = 4.059, 95% CI: 1.045-14.340) in females, but showed no signi cance for males. Whereas referring to CysC ≤ 0.93 mg/dl, CysC>0.93 mg/dl did not show obvious links to severe in ammatory, liver injury, and heart failure both in males and females, (Fig. 2).

Discussion
In our study, we found that the higher CysC level (CysC 0.93 mg/dl) and CysC rangeability 0 was independently associated with mortality among COVID-19-associated hospitalized adults with T2DM, and CysC rangeability seems to contribute more risk for hospitalized death than a single CysC level. More importantly, our study also suggests that sex could affect this association, when classi ed by sex, this association is stronger in female patients. Notably, we also observed these associations in patients with generally mild pneumonia, and all case of death was contributed by T2DM.
To the best of our knowledge, there was no evidence to con rm the association between CysC and allcause death of COVID-19-associated hospitalized adults and as well as the T2DM and gender's hybrid effect in present studies. CysC, the endogenous biomarkers of renal function, can indirectly assess glomerular ltration rate (GFR) 15 , is produced in nucleated cells, and is uneasily in uenced by decreased muscle mass, acute and chronic illness, and nutritional status than creatinine 16 . Moreover, over the past decade, serum CysC has been extensively studied and found to be a sensitive serum marker of the GFR and a stronger predictor than creatinine as a risk factor of cardiovascular events and all-cause death in generally older patients 16 . Some studies suggested that high levels of serum CysC were associated with severity and poor prognosis of cardiovascular disease 17 , neurodegeneration 18 , CKD 19 , acute kidney injury 20 , chronic liver disease 21 , critically ill intensive care unit (ICU) patients 22 , as well as the general population 23 . Thus, CysC may be a serum molecular marker for predicting poor prognosis of various acute and chronic diseases. Since the COVID-19 outbreak, more and more research has been devoted to discovering its clinical features and organ damage. Some studies have shown that heart, liver, and renal function are impaired in novel coronavirus infected patients 8,[24][25][26] . What's more, the outcome of COVID-19 patients is associated with organ damage. An observative study from South Korea showed that COVID-19 patients with severe AKI had fatal outcomes including death 27 . Those studies only observed renal function injury based on creatinine urea, but not noted the more stable biomarker of CysC. To the best of our knowledge, our study is the rst to focus on CysC and CysC rangeability to observe the impact on organ function and mortality of COVID-19 patients.
Researchers have also observed that novel coronavirus-COVID-19 can directly infect the kidney and cause renal impairment. In Victor G' studies, SARS-CoV-2 viral load in organs and tissues were rst quanti ed, of which autopsy samples obtained from 22 patients who had died from COVID-19, and found that the highest levels of SARS-CoV-2 copies per cell were detected in the respiratory tract, and lower levels in kidney, heart, liver, brain, and blood. These ndings indicate kidneys are among the most common targets of SARS-CoV-2 28 . RNA enrichment for transmembrane serine protease 2 (TMPRSS2), angiotensin-converting enzyme 2 (ACE2), and cathepsin L (CTSL) are considered to facilitate SARSCoV-2 infection. Whereas that RNA of genes are enriched in multiple kidney-cell types from fetal development through adulthood. This enrichment may promote SARS-CoV-2-associated kidney injury 29,30 . CysC is recognized as one of the most reliable renal function parameters in the general population. Thus, when Infected with novel Coronavirus-COVID-19, kidney injury can be re ected in the level of CysC. A recent study concerned about imaging progression on chest CT at the rst week and CysC levels from COVID-19 patients and showed that patients with higher CysC levels with COVID-19 showed more progressive lesions in lung CT, in which only the predictive value of CysC level at admission to the disease progression was observed, rather than dynamic changes 31 . And the severity of COVID-19 is closely related to death. Our study not only found that CysC levels were associated with death in hospital from COVID-19 but also observed the changes in CysC levels during hospitalization were more signi cant in predicting death.
We also found that T2DM more easily caused deaths in our study population, which were consistent with the previous research that T2DM was not only associated with suffering COVID-19 but also an independent risk factor for severe pneumonia and poor prognosis 32 . T2DM can be de ned as the phenotype of hyperglycemia that manifests a group of common metabolic disorders. The metabolic disorder related to T2DM causes secondary pathophysiological changes of multiple organ tissues and results in various complications, which contribute to the morbidity and mortality associated with the T2DM 33 . The pathophysiology of SARS-CoV-2 remains incompletely understood, but evidence demonstrates that SARS-CoV-2 can trigger hyper in ammation in certain organ and emerges tissue tropism, which shared with the same pathologies for chronic in ammation and multi-organ damage in T2DM 33 . Although the intensely acute response of hyperglycemic, in ammatory, and tissue damage in SARS-CoV-2 infection, it is still mirrored by T2DM pathophysiology, which characterized by impaired glycemic regulation, chronic, low-grade in ammation condition, and slowly progressive multi-organ damage, diabetic microvascular (neuropathy, brain, chronic kidney diseases) and macrovascular (cerebrovascular disease) complications. Acute COVID-19-associated adverse response may overlay on glucose instability, preexisting in ammation, and multi-organ damages in T2DM patients, eventually aggravate outcomes 33 . The level of CysC increased in COVID-19 infected patients complicated with acute kidney injury, which also would be an indicator for early detection of diabetic nephropathy in type 2 T2DM. Thus Monitoring changes in CysC may be important for assessing the progression of COVID-19 patients. In the gender subgroup analysis of diabetes, CysC levels at admission were strongly associated with death in females but not for males. Although the relationship between CysC rangeability and death was not sex-speci c, a greater contribution was observed in women than in men, possibly because the levels of CysC were statistically signi cantly higher in males, compared to age-matched females 34  In summary, CysC contributes to clinical manifestations and may in uence the progression and prognosis of COVID-19 patients, with an increased risk of severe complications and death for subjects with T2DM. Thereby, more attention needs to be paid to patients with a high-level CysC as well as CysC rangeability in clinical treatment and later clinical follow-up, especially for those with T2DM. However, the exact mechanisms underlying CysC -related changes in the course of COVID-19 are not established yet and needs to be resolved in the future.

Declarations
Data availability statements The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
The study protocol complies with the Declaration of Helsinki and was approved by the Institutional Review Board of the Wuhan Third Hospital Foundation. Because of the study's retrospective design, informed consent was waived in cases of discharge, of impossibility of contact with patients and in case of death. The privacy and anonymity of the data collected was guaranteed in accordance with current regulations.

Figure 1
The association between CysC and organ dysfunction, all-cause death classi ed by diabetes mellitus; † a, the association between CysC and severe in ammation classi ed by diabetes mellitus; b, the association between CysC and liver injury classi ed by diabetes mellitus; c, the association between CysC and heart failure classi ed by diabetes mellitus; d, the association between CysC and all-cause death classi ed by diabetes mellitus; ‡ Severe in ammation was de ned as highest NLR during hospitalization higher than 10. Liver injury was de ned as ALT or AST at any time during hospitalization greater than 80U/L. Heart failure was de ned as highest BNP during hospitalization greater than 300pg/ml. § Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. ¶ DM, diabetes mellitus; CysC, cystatin C; NLR Neutrophil-to-Lymphocyte Ratio; ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; BNP, Brain Natriuretic Peptide.

Figure 1
The association between CysC and organ dysfunction, all-cause death classi ed by diabetes mellitus; † a, the association between CysC and severe in ammation classi ed by diabetes mellitus; b, the association between CysC and liver injury classi ed by diabetes mellitus; c, the association between CysC and heart failure classi ed by diabetes mellitus; d, the association between CysC and all-cause death classi ed by diabetes mellitus; ‡ Severe in ammation was de ned as highest NLR during hospitalization higher than 10. Liver injury was de ned as ALT or AST at any time during hospitalization greater than 80U/L. Heart failure was de ned as highest BNP during hospitalization greater than 300pg/ml. § Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. ¶ DM, diabetes mellitus; CysC, cystatin C; NLR Neutrophil-to-Lymphocyte Ratio; ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; BNP, Brain Natriuretic Peptide.

Figure 2
The association between CysC and organ dysfunction, all-cause death in diabetes population classi ed by sex. † a, the association between CysC and organ dysfunction, all-cause death in male diabetes population; b, the association between CysC and in female diabetes population. ‡ Severe in ammation was de ned as highest NLR during hospitalization higher than 10. Liver injury was de ned as ALT or AST at any time during hospitalization greater than 80U/L. Heart failure was de ned as highest BNP during hospitalization greater than 300pg/ml. § Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. ¶ DM, diabetes mellitus; CysC, cystatin C; NLR Neutrophil-to-Lymphocyte Ratio; ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; BNP, Brain Natriuretic Peptide.

Figure 2
The association between CysC and organ dysfunction, all-cause death in diabetes population classi ed by sex. † a, the association between CysC and organ dysfunction, all-cause death in male diabetes population; b, the association between CysC and in female diabetes population. ‡ Severe in ammation was de ned as highest NLR during hospitalization higher than 10. Liver injury was de ned as ALT or AST at any time during hospitalization greater than 80U/L. Heart failure was de ned as highest BNP during hospitalization greater than 300pg/ml. § Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. ¶ DM, diabetes mellitus; CysC, cystatin C; NLR Neutrophil-to-Lymphocyte Ratio; ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; BNP, Brain Natriuretic Peptide.

Figure 3
The association between CysC rangeability and all-cause death classi ed by DM. † CysC rangeability (renal function changes) calculated as the difference of CysC level between the time of admission and the highest CysC level testing during hospitalization. ‡ Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. §DM, diabetes mellitus; BNP, Brain Natriuretic Peptide.

Figure 3
The association between CysC rangeability and all-cause death classi ed by DM. † CysC rangeability (renal function changes) calculated as the difference of CysC level between the time of admission and the highest CysC level testing during hospitalization. ‡ Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. §DM, diabetes mellitus; BNP, Brain Natriuretic Peptide.

Figure 4
Page 33/33 The association between CysC rangeability and all-cause death in DM population. † CysC rangeability (renal function changes) calculated as the difference of CysC level between the time of admission and the highest CysC level testing during hospitalization. ‡ Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. §DM, diabetes mellitus; BNP, Brain Natriuretic Peptide.

Figure 4
The association between CysC rangeability and all-cause death in DM population. † CysC rangeability (renal function changes) calculated as the difference of CysC level between the time of admission and the highest CysC level testing during hospitalization. ‡ Adjusted by age, severe pneumonia, serum albumin, blood glucose, Lg BNP baseline. §DM, diabetes mellitus; BNP, Brain Natriuretic Peptide.