Fasting blood glucose predicts the occurrence of critical illness in COVID-19 patients: A multicenter retrospective cohort study.

Dear editors,

We read the recent published article by Lin and colleagues in journal of infection with great interest, which reported that serum ferritin as an independent risk factor for severity in COVID-19 patients. Previous studies aimed to identification risk factors of critical or mortal condition of patients with COVID-19 from clinical, laboratory, and radiological characteristics at admission. Early warning of patients who are most likely to develop critical disease will enable informed decisions, and facilitate the provision of timely supportive treatment in advance and thus reduce mortality. In this multicenter retrospective cohort study, we aimed to explore the value of admission fasting blood glucose (FBG) in predicting the occurrence of critical illness among patients hospitalized for COVID-19.

We enrolled a total of 123 laboratory-confirmed COVID-19 patients from three designated hospitals of Wuhan and Guangzhou, China. We collected clinical and laboratory data at hospital admission from medical records. We defined the severity of COVID-19 according to the newest COVID-19 guidelines released by the National Health Commission of China and the guidelines of American Thoracic Society for community-acquired pneumonia. We defined critical illness as a composite of admission to the intensive care unit (ICU), respiratory failure requiring mechanical ventilation, septic shock during hospitalization, or death. The optimal cutoff of FBG for discriminating COVID-19 patients with non-critical and critical illness was determined using receiver operating characteristic (ROC) curve and by maximizing the Youden index. To identify predictors for critical illness, baseline variables with p-value <0.10 in univariable analysis were entered into multivariate logistic regression. In addition, Pearson correlation coefficient was used to determine the correlation between FBG and other laboratory parameters. The retrospective study was approved by an ethics committee of our institution, with a waiver of informed consent.

The proportion of critical patients was 31.7% (39/123). The relatively high critically ill rate seen in this study was related to the fact that the First Affiliated Hospital of Guangzhou Medical University only admitted severe/critical cases transferred from other designated hospitals of Guangzhou. Table 1 illustrates the comparison of baseline characteristics between critical and non-critical COVID-19 patients using Mann-Whitney U or Chi-squared or Fisher's exact test. Approximately 17.1% of patients reported a previous diabetes, which was higher in the critical group (p = 0.025). 69 patients (56.1%) had hyperglycemia with FBG level >6.1 mmol/L. The critical patients had significantly higher fasting blood glucose (FBG) level than the non-critical patients (p<0.001) (Fig. 1 a). Multivariate logistic regression indicated that the FBG on admission was an independent risk factor for critical illness in patients with COVID-19 (Odds Ratio [OR] = 1.249, 95% confidence interval [CI]: 1.032–1.512). After adjusting for previous diabetes, the OR of FBG for predicting critical illness in COVID-19 patients was 1.245 (95% CI: 1.056–1.468). The optimal FBG level for predicting critical illness was ≥6.50 mmol/L, with an area under the curve (AUC) of 0.767 (95% CI: 0.677–0.857) (Fig. 1b). The FBG was positively correlated with white blood cells (WBC) (r = 0.300, p = 0.001), neutrophil (r = 0.360, p < 0.001), and lactic dehydrogenase (LDH) (r = 0.277, p = 0.002); while FBG was negatively correlated with lymphocyte (r = −0.310, p < 0.001), albumin (r = −0.265, p = 0.003), and indirect bilirubin (IBIL) (r = −0.193, p = 0.037) (Fig. 1c-h). After adjusting for WBC, neutrophil, LDH, lymphocyte, albumin and IBIL, the OR of FBG for predicting critical illness in COVID-19 patients was 1.307 (95% CI: 1.105–1.546).

Table 1

Comparison of baseline characteristics between critical and noncritical patients with COVID-19.

Characteristics	Noncritical (n = 84)	Critical (n = 39)	p-value
Age (years)	58.5 (49.3–66.0)	68.0 (61.0–78.0)	0.001
Sex, n (%)
Male	44 (52.4)	26 (66.7)	0.137
Female	40 (47.6)	13 (33.3)
Comorbidities
Hypertension, n(%)	18 (21.4)	15 (39.5)	0.047
Coronary heart disease, n(%)	9 (10.7)	5 (12.8)	0.765
Diabetes, n(%)	10 (11.9)	11 (28.2)	0.025
Chronic liver diseases, n(%)	5 (6.0)	0	0.178
Chronic lung diseases, n(%)	8 (9.5)	4 (10.3)	1.000
Others, n(%)	22 (26.2)	12 (30.8)	0.597
Laboratory findings
WBC (× 109/L)	5.6 (4.3–7.1)	6.0 (4.4–9.4)	0.127
Neutrophil (× 109/L)	3.7 (2.7–5.5)	5.1 (3.2–8.7)	0.010
Lymphocyte (× 109/L)	0.9 (0.7–1.4)	0.7 (0.4–1.0)	0.002
LDH (U/L)	205.5 (161.8–279.8)	395.5 (271.8–538.8)	<0.001
Hemoglobin (g/L)	127.0 (117.0–144.0)	137.0 (122.0–146.0)	0.245
Platelet (g/L)	190.0 (139.0–274.5)	164.0 (124.0–206.0)	0.047
Albumin (g/L)	36.7 (32.0–40.0)	31.9 (29.3–36.2)	0.001
AST (U/L)	25.5 (18.0–36.0)	40.0 (27.0–59.0)	<0.001
ALT (U/L)	23.0 (16.0–37.3)	27.0 (20.0–41.0)	0.186
DBIL (μmol/L)	3.6 (2.8–5.0)	4.2 (3.2–6.9)	0.071
IBIL (μmol/L)	7.2 (4.5–10.2)	5.6 (4.5–7.4)	0.126
TBIL (μmol/L)	11.2 (7.9–14.9)	9.8 (8.0–15.3)	0.755
APTT (s)	33.7 (31.2–36.6)	35.4 (31.2–39.9)	0.197
PT (s)	13.3 (12.5–14.3)	13.9 (12.9–15.1)	0.061
d-dimer (μg/ml)	0.3 (0.1–0.7)	0.7 (0.2–7.2)	0.003
Creatinine (μmol/L)	70.0 (57.3–82.0)	80.0 (67.0–100.0)	0.002
hs-CRP (mg/L)	12.0 (1.8–34.8)	36.1 (32.1–36.9)	<0.001
Procalcitonin (ng/ml)	0.08 (0.05–0.12)	0.2 (0.1–0.4)	<0.001
FBG (mmol/L)	5.7 (5.0–6.9)	7.4 (6.5–11.9)	<0.001
NTproBNP (pg/mL)	97.3 (31.0–213.3)	497.4 (127.0–843.8)	<0.001

Note: Data were number (percentage) or median (interquartile range).

Abbreviations: WBC, white blood cells; LDH, lactate dehydrogenase; AST, aspartate aminotransferase; ALT, alanine aminotransferase; TBIL, Total Bilirubin; DBIL, Direct Bilirubin; IBIL, indirect bilirubin; APTT, activated partial thromboplastin time; PT, prothrombin time; hs-CRP, high-sensitivity C-reactive protein; FBG, fasting blood glucose; NTproBNP, N-terminal portion of proBNP.

Fig. 1

Admission FBG of COVID-19 patients and its correlation with disease severity. (a) FBG level in noncritical and critical COVID-19 patients (p < 0.001). (b) ROC curve of FBG for discriminating COVID-19 patients with non-critical and critical illness. (c–h) correlation between FBG and WBC, neutrophil, LDH, lymphocyte, albumin and IBIL, respectively.

Abbreviations: FBG, fasting blood glucose; ROC, receiver operating characteristic; WBC, white blood cells; lactic dehydrogenase; indirect bilirubin. .

Our study suggested that elevated FBG at admission is a crucial risk factor for critical illness in COVID-19 patients, although most patients (82.9%) had no previous diabetes, which was consistent with previous studies. , After adjusting for previous diabetes, the OR of FBG level was not significantly changed. The proportion of hyperglycemia was higher than a previous report of 47.2%. which may be due to the higher rate of critically ill patients we included. Except for well-known diabetes, elevation of FBG level at admission could also due to stress hyperglycemia. Stress hyperglycemia has been often found in patients without diabetes, which is more concerning in clinical practice. Stress hyperglycemia may be induced by a decrease of both insulin secretion and the worsening of insulin resistance. it may produce organ damage by inducing endothelial dysfunction and thrombosis through the glycation process and oxidative stress generation. Evidences have showed that critical COVID-19 cases exhibit characteristics of severely impaired immune system, systemic inflammatory reactions and cytokine storm. Our study indicated that admission FBG level was positively correlated with inflammatory biomarkers, such as WBC and neutrophil and negatively correlated with immune state biomarker—lymphocyte.

In conclusion, this present study demonstrated that a high level of admission FBG was an independent risk factor for developing critical illness of COVID-19 patients. Admission FBG was associated with systemic inflammatory reactions and immune state. As a convenient and easy-to-detect marker, blood glucose level can be obtained and monitored in clinical settings. Continuous glucose monitoring is particularly necessary in COVID-19 patients with elevated admission FBG. Glucose control helps prevent and control infections and their complications. Therefore, well-controlled blood glucose may lead to an improved outcomes of patients with COVID-19.

Declaration of Competing Interest

The authors declare no competing interests.

Table 2

Risk factors associated with critical illness in univariable and multivariable logistic regression analysis.

	Univariable		Multivariable
	OR (95% CI)	P value	OR (95% CI)	P value
Age (years)	1.047 (1.016–1.079)	0.003	1.029 (0.966–1.095)	0.379
Comorbidities
Hypertension, n(%)	2.292 (1.000–5.252)	0.050	1.331 (0.291–6.097)	0.713
Diabetes, n(%)	2.907 (1.113–7.596)	0.029	0.935 (0.151–5.782)	0.942
Laboratory findings
WBC (× 109/L)	1.126 (0.998–1.270)	0.053	0.866 (0.095–7.911)	0.899
Neutrophil (× 109/L)	1.168 (1.032–1.323)	0.014	1.253 (0.136–11.557)	0.842
Lymphocyte (× 109/L)	0.252 (0.099–0.644)	0.004	0.698 (0.055–8.868)	0.782
LDH (U/L)	1.007 (1.004–1.011)	<0.001	1.004 (0.997–1.010)	0.265
Platelet (g/L)	0.994 (0.988–0.999)	0.025	0.994 (0.983–1.006)	0.314
Albumin (g/L)	0.886 (0.820–0.957)	0.002	1.062 (0.862–1.308)	0.571
AST (U/L)	1.036 (1.014–1.058)	0.001	1.022 (0.985–1.060)	0.251
PT (s)	1.261 (1.041–1.528)	0.018	1.010 (0.697–1.466)	0.956
d-dimer (μg/ml)	1.226 (1.057–1.423)	0.007	1.016 (0.704–1.464)	0.934
Creatinine (μmol/L)	1.027 (1.009–1.045)	0.003	1.009 (0.975–1.045)	0.601
hs-CRP (mg/L)	1.018 (1.001–1.035)	0.035	0.990 (0.968–1.013)	0.410
Procalcitonin (ng/ml)	2.689 (1.016–7.118)	0.046	2.167 (0.289–16.253)	0.452
FBG (mmol/L)	1.317 (1.134–1.529)	<0.001	1.249 (1.032–1.512)	0.022
NTproBNP (pg/mL)	1.002 (1.001–1.003)	0.002	1.001 (1.000–1.002)	0.211

Note: Baseline variables with p-value <0.10 in univariable analysis were entered into multivariate logistic regression.

Abbreviations: OR, Odds Ratio; WBC, white blood cells; LDH, lactate dehydrogenase; AST, aspartate aminotransferase; PT, prothrombin time; hs-CRP, high-sensitivity C-reactive protein; FBG, fasting blood glucose; NTproBNP, N-terminal portion of proBNP.