TY - JOUR
T1 - Risks of and risk factors for COVID-19 disease in people with diabetes: a cohort study of the total population of Scotland
AU - McGurnaghan, Stuart J.
AU - Weir, Amanda
AU - Bishop, Jen
AU - Kennedy, Sharon
AU - Blackbourn, Luke A.K .
AU - McAllister, David
AU - Hutchinson, Sharon
AU - Caparrotta, Thomas M.
AU - Mellor, Joseph
AU - Jeyam, Anita
AU - O’Reilly, Joseph E.
AU - Wild, Sarah
AU - Hatam, Sara
AU - Höhn, Andreas
AU - Colombo, Marco
AU - Robertson, Chris
AU - Lone, Nazir
AU - Murray, Janet
AU - Butterly, Elaine
AU - Petrie, John
AU - Kennon, Brian
AU - McCrimmon, Rory
AU - Lindsay, Robert
AU - Pearson, Ewan
AU - Satta, Naveed
AU - McKnight, John
AU - Philip, Sam
AU - Collier, Andrew
AU - McMenamin, Jim
AU - Smith-Palmer, Alison
AU - Goldberg, David
AU - McKeigue, Paul M.
AU - Colhoun, Helen M.
AU - Public Health Scotland COVID-19 Health Protection Study Group
AU - Scottish Diabetes Research Network Epidemiology Group
N1 - *1-1-50*
Acceptance in SAN. ST 12/01//21
AAM: apply 6m embargo
LA task: input all named authors. ST 14/12/20
PY - 2020/12/23
Y1 - 2020/12/23
N2 - Background: The objectives were to i) ascertain the cumulative risk of fatal or CCU treated COVID-19 in those with diabetes and compare it to those without diabetes ii) among those with diabetes to investigate risk factors for, and build a cross-validated predictive model of, fatal or CCU treated COVID-19.Methods: In the total population of Scotland we ascertained all persons who had developed fatal or critical care unit-treated COVID-19 (hereafter F/CCU-COVID-19) between 1st March and July 31st 2020 from the nationwide virology, critical care unit, hospital discharge and register of deaths databases. Among those with F/CCU-COVID-19, diabetes status was ascertained by linkage to the national diabetes register. The cumulative incidence of F/CCU- COVID-19 in those with and without diabetes was compared using logistic regression. Among those with diabetes, data on potential risk factors for F/CCU-COVID-19 were obtained from diabetes register and other linked health administrative databases. Among those with diabetes we tested association of these factors with F/CCU-COVID-19 and constructed a prediction model using stepwise regression and 20-fold cross-validation.Findings: 1082 (0.3%) of all those with diabetes in Scotland (n=319 349) developed F/CCU-COVID-19; the age and sex adjusted odds ratio was [OR] 1.395 95% CI: 1.304-1.494, p<0.001 overall compared to the risk in total population of Scotland without diabetes (n=4081 cases in 5 143 951 persons). The OR was 2.396 (1.815-3.163) in type 1 and 1.369 (1.276-1.468) in type 2 diabetes. Of the 1082 persons with diabetes who developed F/CCU-COVID- 19, 90% were ≥60 years old. Among those with diabetes, adjusted for age, sex, diabetes duration and diabetes type, those who developed F/CCU-COVID-19 were more likely to be male, live in residential care, or live in a more deprived area, have a condition already listed as a COVID-19 risk condition such as heart or lung disease, have retinopathy or reduced renal function, have worse glycaemic control, have a prior diabetic ketoacidosis or hypoglycaemia hospitalisation in the past five years, be on more diabetes and other type of drugs, and to have been a smoker (all p<0.001). Those with F/CCU-COVID-19 were less likely to be on an antihypertensive and had lower systolic blood pressure than those without F/CCU-COVID-19 (both p<0.001). The relationship with BMI was J-Shaped. The cross-validated predictive model of F/CCU-COVID-19 in those with diabetes, retained 11 factors in addition to age, sex, diabetes type and duration and had a C-statistic of 0.85 (95% CI 0.83,0.86).Interpretation: Overall risks of F/CCU-COVID-19 are substantially elevated in those with type 1 and type 2 diabetes compared to the background population. The risk of F/CCU-COVID-19, and therefore the need for special protective measures, varies widely among those with diabetes but can be predicted reasonably well using prior clinical history.Funding: none
AB - Background: The objectives were to i) ascertain the cumulative risk of fatal or CCU treated COVID-19 in those with diabetes and compare it to those without diabetes ii) among those with diabetes to investigate risk factors for, and build a cross-validated predictive model of, fatal or CCU treated COVID-19.Methods: In the total population of Scotland we ascertained all persons who had developed fatal or critical care unit-treated COVID-19 (hereafter F/CCU-COVID-19) between 1st March and July 31st 2020 from the nationwide virology, critical care unit, hospital discharge and register of deaths databases. Among those with F/CCU-COVID-19, diabetes status was ascertained by linkage to the national diabetes register. The cumulative incidence of F/CCU- COVID-19 in those with and without diabetes was compared using logistic regression. Among those with diabetes, data on potential risk factors for F/CCU-COVID-19 were obtained from diabetes register and other linked health administrative databases. Among those with diabetes we tested association of these factors with F/CCU-COVID-19 and constructed a prediction model using stepwise regression and 20-fold cross-validation.Findings: 1082 (0.3%) of all those with diabetes in Scotland (n=319 349) developed F/CCU-COVID-19; the age and sex adjusted odds ratio was [OR] 1.395 95% CI: 1.304-1.494, p<0.001 overall compared to the risk in total population of Scotland without diabetes (n=4081 cases in 5 143 951 persons). The OR was 2.396 (1.815-3.163) in type 1 and 1.369 (1.276-1.468) in type 2 diabetes. Of the 1082 persons with diabetes who developed F/CCU-COVID- 19, 90% were ≥60 years old. Among those with diabetes, adjusted for age, sex, diabetes duration and diabetes type, those who developed F/CCU-COVID-19 were more likely to be male, live in residential care, or live in a more deprived area, have a condition already listed as a COVID-19 risk condition such as heart or lung disease, have retinopathy or reduced renal function, have worse glycaemic control, have a prior diabetic ketoacidosis or hypoglycaemia hospitalisation in the past five years, be on more diabetes and other type of drugs, and to have been a smoker (all p<0.001). Those with F/CCU-COVID-19 were less likely to be on an antihypertensive and had lower systolic blood pressure than those without F/CCU-COVID-19 (both p<0.001). The relationship with BMI was J-Shaped. The cross-validated predictive model of F/CCU-COVID-19 in those with diabetes, retained 11 factors in addition to age, sex, diabetes type and duration and had a C-statistic of 0.85 (95% CI 0.83,0.86).Interpretation: Overall risks of F/CCU-COVID-19 are substantially elevated in those with type 1 and type 2 diabetes compared to the background population. The risk of F/CCU-COVID-19, and therefore the need for special protective measures, varies widely among those with diabetes but can be predicted reasonably well using prior clinical history.Funding: none
KW - Covid-19
KW - diabetes
KW - Scotland
U2 - 10.1016/S2213-8587(20)30405-8
DO - 10.1016/S2213-8587(20)30405-8
M3 - Article
JO - The Lancet Diabetes & Endocrinology
JF - The Lancet Diabetes & Endocrinology
SN - 2213-8587
ER -