Associations with covid-19 hospitalisation amongst 406,793 adults: the UK Biobank prospective cohort study

Abstract

ABSTRACT OBJECTIVES To identify the sociodemographic, lifestyle, comorbidity and antihypertensive medication associations with the development of hospitalisation with covid-19 in an English population. DESIGN Prospective cohort study SETTING The population-based UK Biobank study was linked to English covid-19 test results. PARTICIPANTS Individuals resident in England and alive in 2020. MAIN OUTCOME MEASURES Cases (n=605) were defined by a positive covid-19 test result conducted between 16 th March and 16 th April 2020, during a restricted testing policy for hospitalised individuals with severe disease. RESULTS A total of 406,793 participants were included. Mean age on 1 st January 2020 was 68 years (range 48 to 85 years). 55% were women. In multivariable models, major independent risk factors for hospitalisation with covid-19 were male sex (odds ratio 1.52; 95% confidence interval 1.28 to 1.81; P<0.001), South Asian ethnicity (2.02; 1.28 to 3.17; P=0.002) or black ethnicity (3.09; 2.18 to 4.38; P<0.001) compared to white ethnicity, greater residential deprivation (1.92 for most deprived quartile compared to least deprived quartile; 1.50 to 2.47; P<0.001), higher BMI (2.04 for BMI >35 compared to <25 Kg/m 2 ; 1.50 to 2.77; P<0.001), former smoking (1.39 compared to never smoked; 1.16 to 1.66; P<0.001), and comorbidities hypertension (1.28; 1.06 to 1.53; P=0.009) and chronic obstructive pulmonary disease (1.81; 1.34 to 2.44; P<0.001). Increased risk was observed with increasing number of antihypertensive medications used rather than any individual class. CONCLUSION Understanding why these factors confer increased risk of severe covid-19 in the population may help elucidate the underlying mechanisms as well as inform strategy and policy to prevent this disease and its consequences. We found no evidence of increased risk with specific classes of antihypertensive medication.