Contemporary hormonal contraception and risk of endometrial cancer in women younger than age 50: A retrospective cohort study of Danish women

Abstract

Objective: To examine the association between contemporary hormonal contraceptives and endometrial cancer risk in women younger than age 50.

Study design: Cohort study of women living in Denmark aged 15-49 years through 1995-2014. National registries provided information about hormonal contraception use, incident endometrial cancer and confounders. Ever, current or recent, and former users of any hormonal contraception were compared with non-users, using Poisson regression to calculate incidence rate ratios (RR) with 95% confidence intervals. Duration, time since last use, tumor-specific and product-specific analyses, and population prevented fraction, were calculated.

Results: During 21.1 million person-years, 549 incident endometrial cancers occurred, with ever users of any hormonal contraception having a reduced premenopausal endometrial cancer risk compared with non-users; RR 0.60 (95% Confidence Interval 0.49 to 0.73). A lower risk of endometrial cancer was seen in all current or recent users of any hormonal contraception; 0.65 (0.52 to 0.83) and combined contraceptives; 0.57 (0.43 to 0.75), but not progestin-only contraceptives; levonorgestrel intrauterine system, LNG-IUS; 0.97 (0.66 to 1.42); other progestin-only contraceptives; 0.61 (0.27 to 1.37). Increased RRs were found for current use of any hormonal, combined contraceptives or LNG-IUS of ≤one year, probably because of protopathic bias. Longer durations of use were associated with significant reductions that became stronger with longer use. Former users of any hormonal contraception continued to benefit from a reduced risk of endometrial cancer >10 years after stopping. There was little evidence of differences in risk reduction by the type of progestin in combined oral contraceptives. Current or recent use of any hormonal contraception was associated with an approximate halving of risk of the most common tumor type I carcinoma, and an increased risk of the rarer sarcoma. Overall the estimated absolute reduced risk of endometrial cancer in ever users of hormonal contraceptives was 1.4 per 100,000 person-years, or approximately one less endometrial cancer for every 71,400 women of reproductive age who used hormonal contraception for one year. Use of hormonal contraception was estimated to prevent 25% of endometrial cancers in this population.

Conclusions: Currently available combined hormonal contraceptives are still associated with enduring protection against endometrial cancer, particularly for type I carcinomas.

Implications: We report substantive evidence of the association between different types of contemporary hormonal contraception and endometrial cancer risk in a national cohort of young Danish women. Currently available combined hormonal contraceptives are still associated with enduring protection against endometrial cancer, particularly for type I carcinomas.

Keywords: Cohort; Combined contraceptives; Endometrial cancer; Hormonal contraception; Progestin-only contraceptives.