Does Pre-existing Diabetes Correlate with Long COVID-19 in Europe? Evidence from the Analysis of the Survey of Health, Ageing and Retirement in Europe’s Corona Surveys

Background. A substantial proportion of those infected with COVID-19 are presenting with persistent symptoms, referred to as long COVID-19. Emerging evidence suggests that the presence of pre-existing chronic conditions, such as diabetes, may increase the risk of long COVID-19. Objectives. To investigate whether having pre-existing diabetes increases the risk of developing long COVID-19 in the population of middle-aged and older adults (≥50 years old) in Europe, while assessing if this relationship can be accounted for or is modified by the known long COVID-19 and diabetes risk factors (age, sex, hospitalization, pre-existing hypertension, and weight status). Methods. A population-based longitudinal prospective study involving a sample of respondents aged 50 years and older () with probable or confirmed COVID-19 infection from 27 countries that participated in both waves 7 and 8 of the Survey of Health, Ageing and Retirement in Europe and its 2020 and 2021 Corona Surveys. Logistic regression modeling was performed. Results. Overall, 66.8% of the respondents affected by COVID-19 infection reported at least one long COVID-19 symptom; 55.2% were female, and the average age was 64.6 years; 13.2% had pre-existing diabetes. Respondents with pre-existing diabetes had significantly higher odds of developing long COVID-19, compared to those without diabetes (; 95% ). This relationship remained significant (; 98% ) after adjusting for sex ( for females; 95% ), hospitalization for COVID-19 illness (; 95% ), pre-existing hypertension (; 95% ), and overweight (; 95% ) and obese (; 95% ) weight status. The effect of pre-existing diabetes on the risk of long COVID-19 is moderated by age; it was highest at the age of 50 (; 95% ), and then, it declined with age. Conclusions. There is a relationship between pre-existing diabetes and long COVID-19, even after controlling for literature-based confounding factors, with age having a moderating effect on this relationship.