Type 2 diabetes is recognised to result in worse outcomes for people infected with COVID-19, but the relative risks in those with type 1 and type 2 diabetes and the impact of glycaemic control on severity of infection and mortality are unclear. Recently, however, two cohort studies from NHS England have shed light on this issue.
The first of these studies showed that, of the 23 804 hospital deaths related to COVID-19 in England over a 10-week period up to 11 May 2020, one third occurred in people with diabetes: 7466 (31.4%) in those with type 2 diabetes and 365 (1.5%) in those with type 1 diabetes, despite these two conditions occurring in only 4.7% and 0.4% of the general population, respectively.
After adjustment for age, sex, deprivation, ethnicity and geographical region, people with type 1 diabetes were 3.5 times more likely to die in hospital with COVID-19 than those without diabetes, and people with type 2 diabetes were just over twice as likely (odds ratio, 2.03). These odds ratios declined slightly to 2.86 and 1.81, respectively, when further adjusted for previous admissions related to cardiovascular disease.
The second cohort study, which looked at deaths both in and outside hospital, demonstrated that poor glycaemic control, poor renal function and higher BMI were all associated with COVID-19-related mortality rates. Of those with type 2 diabetes, 6.1% had an HbA1c ≥86 mmol/mol (≥10.0%) according to 2018/19 National Diabetes Audit data. After adjusting for other risk factors, compared with an HbA1c of 48–53 mmol/mol (6.5–7.0%), the hazard ratio (HR) for death was 1.23 in those with an HbA1c of 59–74 mmol/mol (7.5–8.9%), and 1.62 for those with an HbA1c of ≥86 mmol/mol. There was also a small increase in risk associated with an HbA1c <48 mmol/mol. In those with type 1 diabetes, the risk associated with poor glycaemic control was further increased, with a more than doubling of the death rate in those with an HbA1c ≥86 mmol/mol compared with 48–53 mmol/mol. However, the differences for the other HbA1c ranges were not significant.
Overall, amongst people with diabetes in England, since 3 April 2020 there has been a doubling of the death rate compared to that expected at this time of year, translating to between 2500 and 3000 more deaths per week.
The association between BMI and COVID-19 mortality in people with diabetes was U-shaped. Amongst those with type 2 diabetes, compared with a BMI of 25–29.9 kg/m2, the HR was 2.26 in those with a BMI of <20 kg/m2 and 1.46 in those with a BMI ≥40 kg/m2. Impaired renal function was also associated with an increased risk of COVID-19-related mortality. In those with type 2 diabetes and an eGFR of 30–44 mL/min/1.73 m2 (stage 3b CKD), a significantly increased HR of 1.75 for mortality was seen compared to those with eGFR >60, and the mortality rate was five times higher in those with an eGFR of <15 (stage 5 CKD). Many of the factors associated with poor prognosis in COVID-19 confirmed in these and other studies are not modifiable, such as age, ethnicity and deprivation. However, glycaemic control and obesity (which will be discussed further in a future Diabetes Distilled) can be modified with lifestyle and currently available drugs. Since COVID-19 is likely to circulate in the UK and elsewhere for the foreseeable future, these papers should act as an urgent call to action for healthcare professionals to be more cautious when assessing COVID-19 in those with diabetes, CKD and obesity, and to help people with diabetes achieve tighter blood glucose control when appropriate.
The studies can be accessed at https://doi.org/10.1016/S2213-8587(20)30272-2
and https://doi.org/10.1016/S2213-8587(20)30271-0