FLARE: The Interplay Between Diabetes and SARS-CoV-2

Many people are saying...diabetes makes COVID-19 worse

Introduction

Case series detailing risk factors for severe COVID-19 consistently identify diabetes mellitus (Grasselli et al. 2020; Goyal et al. 2020). The CDC has advised that patients with diabetes are at higher risk for severe COVID-19 illness. Does having diabetes make COVID-19 presentations more severe? Relatedly, does having COVID-19 make diabetes presentations more severe?

---

Does diabetes make COVID-19 presentations more severe?

During past viral epidemics, including those caused by SARS-CoV-1, H1N1 influenza, and MERS-CoV, diabetes has been associated with increased mortality and morbidity (J. K. Yang et al. 2006; Kulcsar et al. 2019; Allard et al. 2010). Initial studies emerging from China have reported high prevalence of comorbidities including diabetes among patients admitted with COVID-19. The table below compares the rates of diabetes in the general population across different countries to rates of diabetes in large case series of patients with COVID-19. Briefly, across all studies, the prevalence of diabetes in COVID-19 is generally higher than that in the general population. Of note, most studies to date have not differentiated between type 1 diabetes and type 2 diabetes.

Furthermore, patients with diabetes appear to be at higher risk not only for contracting SARS-CoV-2, but also for developing severe COVID-19 and succumbing to the illness. As illustrated in Figures 1 and 2 below, across several studies reviewed by Singh et al., patients with diabetes demonstrated substantially higher rates of severe COVID-19 and death (Singh, Gupta, and Misra 2020).

Many additional reports have suggested an association between diabetes and severe COVID-19. Wu et al., in an analysis of 201 patients, found that diabetes was associated with development of acute respiratory distress syndrome (ARDS), with a hazard ratio of 2.34 (Wu et al. 2020). In a systematic review of 13 studies, 3,027 patients with COVID-19 were analyzed to determine risk factors for critical or lethal illness, revealing a 3.7-fold increased odds of diabetes in the critically ill (Zheng et al. 2020). In an even larger systematic review, nearly 6,500 patients (representing 30 studies) were analyzed to identify statistically significant associations between diabetes and severe COVID-19 (RR 2.5), ARDS (RR 4.6), death (RR 2.1), and composite poor outcome (RR 2.4), though the authors note the possibility of confounding by age and hypertension (I. Huang, Lim, and Pranata 2020).

In an effort to eliminate potential confounders from the analysis, an observational study by Gou et al. divided patients with COVID-19 into groups of those without diabetes (n=26) and those with only diabetes and no other comorbidities (n=24). However, the median age of patients with diabetes was 61, compared to the median age of 32 for patients without diabetes, making the comparison very difficult (Guo et al. 2020). Thus, the increased prevalence of patients with diabetes in COVID-19 cohorts has revealed an association between these two diseases, but no causal link has been established. There are many possible confounders including differences in age and socioeconomic status, as well as concurrent comorbidities such as hypertension and obesity.

How might diabetes exacerbate COVID-19?

Although causality has yet to be definitively established, several mechanisms have been proposed to explain how diabetes and hyperglycemia might lead to more severe viral respiratory illnesses in particular. Hyperglycemia can reduce neutrophil degranulation, impair complement activation, and impair phagocytosis, all of which can increase severity of viral infections. Elevated blood glucose levels can also directly increase glucose concentrations in airway secretions (Philips et al. 2003) and, based on in vitro studies of influenza, may thereby promote viral replication (Kohio and Adamson 2013). Other studies have suggested that hyperglycemia can damage cells, increase concentrations of toxic intracellular by-products of the glycolytic pathway, and ultimately contribute to the cytokine and inflammatory response, worsening prognosis of acute viral infections (Gentile, Strollo, and Ceriello 2020). As a result, during the current COVID-19 pandemic, many have advocated for better monitoring and management of blood sugar in patients with COVID-19 (Klonoff and Umpierrez 2020; Ma and Ran 2020; Zhou and Tan 2020).

Another hypothesis is that diabetes might regulate COVID-19 severity by modulating the expression of angiotensin-converting enzyme 2 (ACE2), a cell-surface receptor for SARS-CoV-2 (Hoffmann et al. 2020; Lu et al. 2020). Interestingly, acute hyperglycemia has been shown to upregulate ACE2 expression in the kidneys of leptin receptor-deficient mice (Wysocki et al. 2006) and the serum, liver, and pancreas of non-obese diabetic mice (Roca-Ho et al. 2017). Some anti-diabetic drugs, including liraglutide and pioglitazone, have also been associated with ACE2 upregulation in animal studies (Muniyappa and Gubbi 2020; Romaní-Pérez et al. 2015). It remains unknown whether ACE2 levels in humans are likewise regulated or whether such regulation would alter susceptibility to SARS-CoV-2 infection.

Does COVID-19 worsen diabetes control?

Based on the observational data detailed above, patients with diabetes do appear to be at an increased risk for severe COVID-19. What is less clear is whether having COVID-19 makes manifestations of diabetes more severe, such as increased presentations of diabetic ketoacidosis (DKA) or significant increase of insulin resistance. In a recent review in Lancet Diabetes and Endocrinology, the authors shared their observation of high insulin requirements in patients with severe COVID-19, and their personal view that the extent of insulin resistance in patients with diabetes seems disproportionate compared with critical illness caused by other conditions (Bornstein et al. 2020). However, these are simply the impressions of the authors, not primary data. Li et al. recently reported that COVID-19 infection may cause ketosis and ketoacidosis. Out of 658 hospitalized patients with confirmed COVID-19, 42 presented with ketosis. 15 of the 42 patients with ketosis had diabetes (only one with type 1 diabetes) and 3 out of the 15 cases developed diabetic ketoacidosis (DKA). They concluded that COVID-19 may induce DKA in patients with type 2 diabetes (Li et al. 2020). Unfortunately, the sample size of this study was small and there was no proposed mechanism as to how COVID-19 may predispose to DKA.
 
Acute infection is a known trigger for DKA, typically in the setting of concurrent dehydration and insensible losses. The acute cytokine release and inflammatory reaction triggered by infection is also known to contribute to insulin resistance and beta-cell dysfunction (Hasnain et al. 2014; Frydrych et al. 2017). COVID-19 may predispose patients to severe hyperglycemia and DKA like any other acute infection, or COVID-19 may uniquely increase this risk, particularly in type 2 diabetes. 
 
There is a hypothesis that coronaviruses cause transient beta-cell dysfunction, leading to acute hyperglycemia and relative insulin deficiency, predisposing to DKA. In a study on SARS coronavirus (SARS-CoV-1) by Yang et al. detected ACE2 expression in the endocrine pancreas. Clinically, thirty-nine patients with SARS who had no history of type 2 diabetes and who received no glucocorticoid therapy were compared to their healthy siblings during a 3-year follow up period. More than 50% of the patients (20 out of 39) developed diabetes (diagnosed by fasting glucose) within 2 weeks of their hospitalization for SARS, but only two (5%) remained diabetic after three years of recovery. Given the expression of ACE2 in the pancreas (Figure 3), they proposed that binding of SARS-CoV-1 to ACE2 allows for entry into islet cells and subsequent damage causing acute diabetes (J.-K. Yang et al. 2010). Based on what we know about the similarity of binding proteins and cellular entry between SARS-CoV-1 and SARS-CoV-2 (Hoffmann et al. 2020; Lu et al. 2020), SARS-CoV-2 may have a similar effect on pancreatic islet cells in patients with COVID-19. In patients with known diabetes, the effects of beta-cell dysregulation may lead to acutely worsening hyperglycemia, higher insulin needs, and ketoacidosis, while those without known diabetes may present with new-onset disease. This hypothesis will be thoroughly tested as we continue to learn about the presence of SARS-CoV-2 in different tissues from pathology reports.

The proposed relationship between ACE2 and SARS-CoV-2 suggests the possible existence of a feedback loop in which: (1) SARS-CoV-2 virus infects the lung and pancreas via ACE2; (2) infection leads to severe COVID-19 illness and severe hyperglycemia, respectively; (3) severe hyperglycemia may, in turn, lead to upregulation of ACE2 in various organs, (4) upregulation of ACE2 leads to further viral entry and further inflammation (Brufsky 2020).

---

Conclusions

Observational data suggests that patients with diabetes develop severe COVID-19 more often than patients without diabetes. Whether SARS-CoV-2 has intrinsic properties that affect the pathophysiology of glycemic regulation in hosts and makes diabetes more severe in patients with COVID-19 is yet to be determined. It is possible that a vicious cycle develops, where the virulence properties of SARS-CoV-2 and acutely impaired glycemic control in patients with diabetes directly contribute to the morbidity and mortality of patients with COVID-19. 

Future studies are needed to determine the importance of and parameters for glycemic control in this specific population. It may not be that COVID-19 patients require an alternative management strategy to currently accepted blood glucose management protocols (e.g. NICE-SUGAR Study Investigators et al. 2009), but rather that clinicians should practice with heightened awareness of potential complications from diabetes and SARS-CoV-2.

---

View all COVID-19 Updates

Sign up for FLARE updates