Social deprivation as a risk factor for COVID-19 mortality among women and men in the UK Biobank: nature of risk and context suggests that social interventions are essential to mitigate the effects of future pandemics

Abstract

Objectives: To investigate sex differences in the effects of social deprivation on COVID-19 mortality and to place these effects in context with other diseases.

Design: Prospective population-based study.

Setting: UK Biobank.

Participants: 501 865 participants (54% women).

Main outcome measure: COVID-19 as the underlying cause of death.

Results: Of 472 946 participants alive when COVID-19 was first apparent in the UK (taken as 1 February 2020), 217 (34% women) died from COVID-19 over the next 10 months, resulting in an incidence, per 100 000 person years, of 100.65 (95% CI 79.47 to 121.84) for women and 228.59 (95% CI 194.88 to 262.30) for men. Greater social deprivation, quantified using the Townsend Deprivation Score, was associated with greater risk of fatal COVD-19. Adjusted for age and ethnicity, HRs for women and men, comparing those in the most with the least deprived national fifths, were 3.66 (2.82 to 4.75) for women and 3.00 (2.46 to 3.66) for men. Adjustments for key baseline lifestyle factors attenuated these HRs to 2.20 (1.63 to 2.96) and 2.62 (2.12 to 3.24), respectively. There was evidence of a log-linear trend in the deprivation-fatal COVID-19 association, of similar magnitude to the equivalent trends for the associations between deprivation and fatal influenza or pneumonia and fatal cardiovascular disease. For all three causes of death, there was no evidence of a sex difference in the associations.

Conclusions: Higher social deprivation is a risk factor for death from COVID-19 on a continuous scale, with two to three times the risk in the most disadvantaged 20% compared with the least. Similarities between the social gradients in COVID-19, influenza/pneumonia and cardiovascular disease mortality, the lack of sex differences in these effects, and the partial mediation of lifestyle factors suggest that better social policies are crucial to alleviate the general medical burden, including from the current, and potential future, viral pandemics.

Keywords: cardiovascular disease; infection; social class.

Figure 1

Penalised spline plots of age and ethnicity adjusted HR (with shaded 95% CI) for the association between Townsend score and death from COVID-19, influenza or pneumonia, and cardiovascular disease (CVD), by sex.

Figure 2

Age and ethnicity adjusted, and multiple adjusted¹, HR (with 95% CI) for women and men, and women to men RHR (with 95% CI), for the association between one-unit higher Townsend score² and death from COVID-19, influenza or pneumonia, and CVD.¹Adjusted for baseline age, ethnicity, systolic blood pressure, diabetes, smoking, body mass index, total cholesterol and history of CVD. ²Townsend scores in the UK Biobank ranged from −6.26 to 11.00, with a median of −2.14. CVD, cardiovascular disease; RHR, ratio of HRs.

Figure 3

Age and ethnicity adjusted HR (95% CI) for women and men, and women to men RHR (with 95% CI), for the association between one-unit higher Townsend score and death from COVID-19, by subgroup.BMI, body mass index; CVD, cardiovascular disease; RHR, ratio of HRs.