ABSTRACT
Aims: Previous UK population research identified multiple risk factors for increased covid-19 mortality in people with type 2 diabetes but it is unclear if these are general to respiratory infections or specific to covid-19. We aimed to compare risk factors associated with death from covid-19 (pre-vaccination roll-out) and pneumonia. Methods: In UK routine primary care data (CPRD), we followed adults with type 2 diabetes from 01/09/2019-31/ 01/2020 (pneumonia mortality cohort n = 609,079) and 01/02/2020-31/ 10/2020 (covid-19 mortality cohort n = 587,933). Multivariable Cox proportional hazards models were used to identify risk factors in each cohort. Results: We observed 2,690 deaths (0.5%) due to covid- 19, and 1,612 deaths due to pneumonia (0.3%). For covid- 19 mortality, we replicated previously reported risk factor associations for male sex, older age, higher deprivation, higher BMI, renal impairment, previous stroke and cardiovascular disease. These features were also associated with higher pneumonia mortality. A differential effect was observed for ethnicity: compared to people of white ethnicity, black and south Asian groups had higher covid-19 mortality (adjusted hazard ratio [aHR] 2.07 [95%CI 1.81-2.38], p < 0.001, and 1.50 [1.33-1.70], p < 0.001 respectively), but lower pneumonia mortality (aHR 0.43 [95%CI 0.31-0.60], p < 0.001, and 0.54 [0.43-0.68], p < 0.001 respectively). Higher HbA1c was a stronger risk factor for covid-19 mortality than pneumonia mortality (aHRs [95%CI] HbA1c >86 vs 48-53 mmol: 1.30 [1.09-1.54], p = 0.004 for covid- 19, 1.10 [0.86-1.42], p = 0.442 for pneumonia). Conclusions: In type 2 diabetes, clinical risk factors for covid-19 and pneumonia mortality are largely similar, but non-white ethnicities have disproportionately higher risk of covid-19 mortality compared to lower risk of pneumonia mortality, which needs further exploration.
ABSTRACT
Aim: Diabetes has consistently been shown to increase the risk of poor covid-19 outcomes. Rather than a simple additive effect of diabetes and age-related risk, recent studies suggest a disproportionately higher relative mortality risk in younger people. Better understanding the interaction between age and diabetes could help inform complex prioritisation decisions around covid-19 vaccination. Methods: We triangulate evidence on heterogeneity of diabetes effect by age on covid-19 mortality from large UK studies. Two population-based studies (OpenSAFELY [n = 17,278,392, 8.8% diabetes] and QCOVID [n = 6,083,102, 7.0% diabetes]), report age-specific hazard ratios (HR) associated with diabetes for covid-19 mortality. We also examine age-specific HRs in severe covid-19 (n = 19,256 critical care patients in England, 18.3% diabetes). To aid interpretability, we translate risk estimates into covid-age;the additional years of covid-19 mortality risk added to an individual's chronological age if diabetes is present. Results: Additional covid-19 mortality risk associated with diabetes is markedly higher in younger than older people across all studies. This reflects the higher relative risk associated with diabetes in younger age groups (HRs for diabetes >5 in ages <50 in OpenSAFELY and QCOVID). For a person aged 40 with diabetes, additional mortality risk is equivalent to 20 years of chronological age, meaning risk is similar to that of a person without diabetes aged 60. For a person aged 70 with diabetes, additional mortality risk from diabetes is equivalent to an additional 5 years, so their covid age is 75. Conclusion: The disproportionate covid-19 mortality risk in younger people with diabetes should be considered to ensure they are appropriately prioritised for vaccination.