Risk of venous thromboembolism in people with RA: a population-based study in the UK

Abstract

Objectives: To evaluate the absolute and relative risk of venous thromboembolism (VTE) in individuals with RA, with relation to age, sex, BMI, disease duration and exposure to exogenous oestrogens.

Methods: Individuals with RA, registered with the UK Oxford-RCGP RSC primary care database between 1999 and 2018, were matched 1:4 with individuals without RA. Multivariable-adjusted Cox proportional hazards was used to compare VTE risk, stratified by age, sex, BMI, disease duration and prescription of oestrogen-containing contraceptives or hormone replacement therapy (HRT).

Results: VTE risk was higher in individuals with RA (n = 23 410) than the matched controls (n = 93 640): adjusted hazard ratio 1.46 (95% CI 1.36, 1.56). Absolute risk of VTE increased with age and higher BMI. Compared with the controls, however, the relative excess risk of VTE was higher in younger than older individuals: 18-49 years (2.13; 95% CI 1.62, 2.79); 50-69 years (1.57; 95% CI 1.38, 1.78); ≥70 years (1.34; 95% CI 1.14, 1.60) and higher in individuals with normal vs elevated BMI: <25 kg/m2 (1.66; 95% CI 1.39, 1.98); 25-30 kg/m2 (1.60; 95% CI 1.36, 1.88); >30 kg/m2 (1.41; 95% CI 1.19, 1.68). VTE risk remained elevated irrespective of disease duration, was similar between women prescribed vs not prescribed oestrogen-containing contraceptives and higher for women prescribed HRT than those not prescribed HRT.

Conclusion: Individuals with RA are at increased risk of VTE regardless of age, sex, BMI, disease duration and exposure to exogenous oestrogens. This highlights the need to consider VTE risk in all individuals with RA.

Keywords: RA; epidemiology; obesity; pulmonary embolism; safety; venous thromboembolism.

Figure 1.

Relative excess risk of venous thromboembolism (VTE) in individuals with RA, compared with matched controls, shown for the full study cohort (Panel A), by sex (Panel B), by age group (Panel C) and by BMI (Panel D). Estimates are from multivariable-adjusted Cox proportional hazards models, reported as adjusted hazard ratios with 95% CIs. Covariates included sociodemographic and clinical characteristics, and established VTE risk factors, as detailed in the methods section

Figure 2.

Incidence of venous thromboembolism (VTE) in people with RA, relative to matched controls, by duration of disease. Unadjusted absolute incidence rates are presented per 1000 person-years of exposure, in addition to multivariable-adjusted hazard ratios with 95% CIs comparing VTE risk in individuals with RA and matched controls. Covariates included sociodemographic and clinical characteristics, and established VTE risk factors, as detailed in the Methods section

Figure 3.

Incidence of venous thromboembolism (VTE) in female individuals with RA, relative to matched female controls, comparing individuals prescribed vs not prescribed oestrogen-containing contraceptives (Panel A) or hormone replacement therapy (HRT) (Panel B). Unadjusted absolute incidence rates (IR) for VTE are shown per 1000 person-years of exposure, in addition to multivariable-adjusted hazard ratios with 95% CIs comparing VTE risk in female individuals with RA and matched female controls. Covariates included sociodemographic and clinical characteristics, and established VTE risk factors, as detailed in the Methods section