Systemic hormonal contraception and risk of venous thromboembolism

Abstract

Introduction: The increased risk of venous thromboembolism associated with the use of hormonal contraception is well recognized, but evidence regarding hormonal contraception containing natural estradiol is limited. This study aimed to assess the associations between the patterns of use of different systemic hormonal contraceptives and the risk of venous thromboembolism during 2017-2019.

Material and methods: All fertile-aged women (15-49 years) living in Finland in 2017 and using hormonal contraception in 2017 and their 1:1 age- and residence-matched controls not using hormonal contraception in 2017 (altogether 587 559 women) were selected from the Prescription Centre. All incident venous thromboembolism cases during 2018-2019 and their 4:1 age-matched controls were further analyzed in a prospective nested case-control design to assess the associations between the use (starting, stopping, continuous vs no use) of different hormonal contraception types and venous thromboembolism.

Results: Altogether, 1334 venous thromboembolism cases occurred during the follow-up period (incidence rate 1.14 per 1000 person-years, 95% confidence interval [CI] 1.08-1.20), with an incidence rate ratio of hormonal contraception vs no hormonal contraception use of 1.42 (95% CI 1.27-1.58). Compared with non-use, starting the use of gestodene and ethinylestradiol (adjusted odds ratio [aOR] 2.85; 95% CI 1.62-5.03), drospirenone and ethinylestradiol (aOR 1.55; 95% CI 0.98-2.44), desogestrel and ethinylestradiol (aOR 1.97; 95% CI 0.99-3.92), and transdermal patch releasing norelgestromin and ethinylestradiol (aOR 5.10; 95% CI 1.12-23.16), as well as continuing the use of gestodene and ethinylestradiol (aOR 2.60; 95% CI 1.61-4.21), drospirenone and ethinylestradiol (aOR 1.55; 95% CI 1.02-2.37), cyproterone-acetate and estrogen/ethinylestradiol (aOR 1.66; 95% CI 1.06-2.61), and vaginal ring releasing etonogestrel and ethinylestradiol (aOR 3.27; 95% CI 1.95-5.48) were associated with venous thromboembolism risk. Regarding the type of estrogen, the highest risk was associated with current use (vs non use in the previous 180 days) of ethinylestradiol-containing preparations (aOR 2.20; 95% CI 1.82-2.65), followed by estradiol-containing preparations (aOR 1.39; 95% CI 1.04-1.87) with no risk for progestin-only hormonal contraception. Current use of estradiol-containing preparations was not associated with venous thromboembolism risk after exclusion of cyproterone-acetate and estrogen/ethinylestradiol (aOR 1.05; 95% CI 0.66-1.66).

Conclusions: An increased risk of venous thromboembolism is associated with ethinylestradiol-containing combined preparations. The use of estradiol-containing combined preparations confers only a slightly increased risk, possibly driven by cyproterone-containing combined oral contraceptives, whereas the use of progestin-only contraception is not associated with venous thromboembolism.

Keywords: estradiol; ethinylestradiol; hormonal contraception; nested case-control; progestin-only contraception; venous thromboembolism.

FIGURE 1

Adjusted ORs for hormonal contraceptives. Multivariate model 2 adjusted with the following covariates: marital status, socioeconomic status, education, cancer, and chronic diseases. One substance in the model at a time. Usage of drugs with respect to event day: no = no use in past 180 days; current = use in past 180 days. CHCs, combined hormonal contraceptives; COCs, combined oral contraceptives; EE, ethinylestradiol; ORs, odds ratios.