Journal of Endocrinology and Metabolism, ISSN 1923-2861 print, 1923-287X online, Open Access
Article copyright, the authors; Journal compilation copyright, J Endocrinol Metab and Elmer Press Inc
Journal website https://www.jofem.org

Original Article

Volume 10, Number 5, October 2020, pages 118-130


Association of Diabetes and Severe COVID-19 Outcomes: A Rapid Review and Meta-Analysis

Figures

Figure 1.
Figure 1. PRISMA flow diagram showing results of search and study selection.
Figure 2.
Figure 2. (a) Forest plot showing pooled hazard ratio of diabetes associated with composite outcome of patients hospitalized for COVID-19. (b) Forest plot showing pooled odds ratio of diabetes associated with composite outcome of patients hospitalized for COVID-19. Composite outcome = mortality/admittance to ICU/requirement for IMV/clinically diagnosed with refractory, progressive, severe, or critical disease [15, 16, 63]; severe outcome = any except mortality. DM: diabetes mellitus; CO: composite outcome.
Figure 3.
Figure 3. Pooled prevalence proportion of patients with diabetes among patients hospitalized for COVID-19 who experienced composite outcome (severe disease, admittance to ICU, requirement for IMV, or death), sub-grouped into studies set in Asia (China and Korea) and those outside of Asia (USA and France).
Figure 4.
Figure 4. Funnel plot for the evaluation of publication bias.

Tables

Table 1. Study Characteristics
 
Total patients analyzedSettingTimeframeExposureOutcomeAdjustment
Guan WJ et al, 2020 [29]1,590575 hospitals in 31 provinces/autonomous regions/provincial municipalities across mainland ChinaDecember 11, 2019 to January 31, 2020Self-reportComposite endpoint (admission to ICU, IMV, or mortality) vs. no composite endpointAge and smoking status
Hu L et al, 2020 [33]323Tianyou Hospital, Wuhan, Hubei Province, ChinaJanuary 8 to February 20, 2020 with follow-up until March 10, 2020Medical historyProgressive disease vs. non-progressive disease of patients classified as non-severe, severe, and critical at baselineNone
Huang R et al, 2020 [2]2028 designated hospitals in 8 cities of Jiangsu Province, ChinaJanuary 22 to February 10, 2020Medical historySevere vsx non-severe diseaseObesity (BMI > 28), lactate dehydrogenase (LDH) > 250 U/L, ALB < 35 g/L, and CRP > 10 mg/L
Hur K et al, 2020 [35]48610 hospitals in the Chicago metropolitan area, Illinois, USAMarch 1 to April 8, 2020 with follow-up until April 18 2020Medical historyIMV or discharge from hospitalSelected variables of importance from the random forest model
Mo P et al, 2020 [38]155Zhongnan Hospital of Wuhan University, Wuhan, Hubei Province, ChinaJanuary 1 to February 5, 2020Medical recordsRefractory disease (not responding to treatment) vs. non-refractory diseaseVariables identified by univariate analysis
Shi Q et al, 2020 [40]306Renmin Hospital of Wuhan University and Zhongnan Hospital of Wuhan University, Wuhan, Hubei Province, ChinaJanuary 1 to March 8, 2020Medical recordsMortality vs. discharged from hospitalHypertension, cardiovascular disease, chronic pulmonary disease
Simonnet A et al, 2020 [41]124CHRU Lille, Nord - Pas de Calais, FranceFebruary 27 to April 5, 2020Medical recordsIMV vs. non-IMVAge, sex, hypertension, dyslipidaemia, BMI
Targher G et al, 2020 [42]3394 unnamed sites in Zhejiang Province, ChinaJanuary and February, 2020Self-reported or random plasma glucose or HbA1cSevere and critical disease vs. mild and moderate diseaseAge, sex, obesity, hypertension, smoking
Wang L et al, 2020 [46]339Renmin Hospital of Wuhan University, Wuhan, Hubei Province, ChinaJanuary 1 to February 6, 2020 with follow-up until March 5, 2020Medical historyMortality vs. discharged from hospitalComorbidities, complications, age
Wu C et al, 2020 [50]201Jinyintan Hospital, Wuhan, Hubei Province, ChinaDecember 25, 2019 to January 26, 2020Medical recordsMortality vs. discharged from hospitalNot reported
Zhang X et al, 2020 [58]645Various designated hospitals according to the government emergency rule of Zhejiang Province, Zhejiang, ChinaJanuary 17, to February 8, 2020No descriptionSevere vs. non-severe diseaseNot reported
Zhang Y et al, 2020 [59]145Tongji Hospital, Wuhan, Hubei Province, ChinaFebruary 8 to March 21, 2020Medical historyComposite endpoint (admission to ICU, IMV, or mortality) vs. no composite endpointAge, sex, BMI, medical histories of hypertension, cardiovascular disease, and malignancy
Zhou F et al, 2020 [61]191Jinyintan Hospital and Wuhan Pulmonary Hospital, Wuhan, Hubei Province, China29 December 2019 - 31 January 2020Medical recordsMortality vs. discharged from hospitalNone
Zhu L et al, 2020 [62]7,33719 hospitals, Hubei Province, China30 December 2019 - 20 March 2020Medical history or clinical diagnosisMortality vs. discharged from hospitalAge, gender, hospital site on admission, with no adjustment for comorbidities closely related to type 2 diabetes (hypertension, CHD, cerebrovascular disease, and chronic kidney disease)

 

Table 2. Putative Prognostic Factors
 
Uncontrolled hyperglycemiaA retrospective cohort study (n = 570) in the USA investigating blood glucose levels among hospitalized patients with diabetes for acute hyperglycemia found that there is an association. The study reported that hospitalized COVID-19 patients with diabetes and/or uncontrolled hyperglycemia had a higher prevalence of mortality as compared with patients without diabetes or uncontrolled hyperglycaemia [20]. Diabetes was defined as A1c ≥ 6.5% [20]. Uncontrolled hyperglycemia was defined as ≥ 2 blood glucoses (BGs) > 180 mg/dL within any 24-h period [20]. However, the CORONADO prospective cohort study with 1,317 COVID-19 patients with diabetes from 53 hospitals in France, where the HbA1C level of the patients was examined, did not find that long-term glycemic control impacted the severity of COVID-19 disease within the first 7 days of admission to hospital [5].
Insulin useA retrospective study of patients with diabetes (n = 136) and those without diabetes (n = 768) with moderate, severe, or critical COVID-19 in Wuhan, China, noted the use of insulin in patients with diabetes was related to poor prognosis of COVID-19 clinical course [25]. However, in the CORONADO study, insulin use was not associated with a severe prognosis (intubation and/or death on day 7) in a multivariable analysis after adjustment [5]. Insulin use may be a proxy of advanced diabetes in older people with complications such as chronic kidney disease (CKD), rather than a causal factor of COVID-19 severity [5].
Triglyceride and glucose (TyG) index marker for insulin resistanceA retrospective study of 151 patients in Wuhan, China, who were admitted to hospital with moderate to severe COVID-19 found an increasing TyG index to predict increased odds of severe or mortal outcomes from COVID-19 [39].