Sex differences in the association between autoimmune disease and cardiovascular outcomes: A prospective population-based cohort of 450 000 individuals
A Schuermans, M Woodward, H Su, B N Weber, P Natarajan, M C Honigberg, G Cambridge, B Casadei, J J V McMurray, K Rahimi, W Budts, N ConradAbstract
Introduction
Autoimmune diseases are associated with an increased risk of cardiovascular disease (CVD). Because immune responses, autoimmune disease prevalence, and cardiovascular risk differ substantially between females and males, excess risk of CVD associated with autoimmune diseases may also vary by sex.
Aim
We sought to examine sex differences in the magnitude and patterns of CVD risk associated with autoimmune diseases.
Methods
We used data from the UK Biobank, linked to secondary care records, to assemble a cohort without known CVD at enrolment. We identified 19 autoimmune diseases and cardiovascular risk factors at enrolment; missing data were imputed using multiple imputation by chained equations. Incident cases for 12 individual cardiovascular outcomes, as well as a composite cardiovascular outcome, were ascertained. We examined sex differences in the associations between autoimmune diseases and CVD using sex-stratified Cox proportional hazards models. We present estimates from age-adjusted models and from models additionally adjusted for known cardiovascular risk factors (socioeconomic status, smoking, body mass index, blood pressure, non-HDL cholesterol, diabetes, lipid-lowering and antihypertensive medication use, C-reactive protein, lipoprotein(a), and kidney function). Effect modification by sex was formally tested by including a sex interaction term in fully adjusted models. We quantified the proportion of excess risk attributable to measured risk factors in females and males as the relative log hazard ratio change between age-adjusted and fully adjusted models.
Results
Among 443,709 participants included in the study (56% females; mean [SD] age, 56.0 [8.1] years, 24,847 (5.6%) had at least one autoimmune disease at baseline; 15,158 (6.1%) amongst females. Over a median (IQR) follow-up of 13.6 (12.9-14.3) years, 6,621 participants (26.6%) with and 76,049 (18.2%) without autoimmune disease experienced incident CVD. Autoimmune disease was more strongly associated with incident CVD in females than males. In age-adjusted models for the composite outcome, hazard ratios were 1.63 (1.57-1.69) in females and 1.41 (1.36-1.46) in males; in fully adjusted models, 1.45 (1.40-1.50) in females and 1.25 (1.21-1.30) in males (Pinteraction<0.001). This pattern was observed across all individual outcomes, with the largest between-sex differences observed for peripheral artery disease, heart failure, and infective endocarditis (Figure). Adjustment for measured risk factors attenuated the autoimmune-associated hazard by 30%, with even a smaller attenuation in females (23.9%) than in males (34.4%).
Conclusions
Autoimmune diseases are associated with substantial excess risk for CVD in both sexes, which is only partially explained by known risk factors and varies by outcome. Females had a greater proportional increase in risk related to autoimmune diseases and a smaller proportion of risk explained by conventional factors.