The efficacy of non-contraceptive uses for hormonal contraceptives

The combined oestrogen–progestogen oral contraceptive pill (COCP) was first marketed for the treatment of menstrual disturbances in 1957 in the United States.1 With the steadily increasing popularity of "the pill" for contraception, anecdotal evidence began to accumulate for a range of beneficial health effects, and it became widely used (without rigorous supporting evidence) for treating various gynaecological symptoms.2 Over time, it became clear that the COCP could provide health benefits for women in three ways: by providing highly effective contraception, treating some gynaecological symptoms, and preventing some gynaecological and medical conditions.3 In recent years, good quality evidence has begun to accumulate for the non-contraceptive health benefits of some long-acting progestogen-only methods, such as depot medroxyprogesterone acetate4 and the levonorgestrel intrauterine system (LNG-IUS; Mirena, Schering).5

Health benefits associated with use of hormonal contraceptives have not received the same degree of research or publicity as potential adverse effects, and the quality of evidence for such benefits is highly variable. It is probable that different COCPs and long-acting delivery systems (with different steroid combinations, dosages and routes of delivery) will provide varying degrees of benefit.

Conditions that may respond to COCP treatment are listed in Box 1. The condition that responds best is primary dysmenorrhoea. While most of the evidence for this comes from studies with medium-dose (50 g oestrogen) COCPs,9 low-dose (20 g oestrogen) COCPs are likely to have a similar effect.10 The benefit is probably associated primarily with suppression of ovulation. Secondary dysmenorrhoea due to endometriosis or chronic pelvic inflammatory disease may also respond, to a lesser degree, to COCP treatment. One randomised, open-label study reported that the COCP was as effective as a gonadotrophin-releasing-hormone agonist in reducing dysmenorrhoea due to endometriosis, but less effective in reducing deep dyspareunia.11

Menorrhagia due to ovulatory dysfunctional uterine bleeding usually responds well to COCP treatment, whereas the response of menorrhagia caused by other conditions is quite variable.11 The best evidence comes from studies with COCPs containing 50 g oestrogen.7 It is not clear whether the same level of benefit occurs with the lowest-dose pills currently available,14 although a randomised placebo-controlled double-blind trial of COCPs delivering either 20 g or 35 g of oestrogen has shown that these dosages significantly reduce dysfunctional uterine bleeding.8 Low-dose COCPs may have a particularly important role as contraceptives in perimenopausal women.15 The use of COCPs tends to raise haemoglobin levels, especially in women with a convincing clinical history of menorrhagia, and reduce the severity of iron deficiency anaemia.16

The connection between hormonal secretions of the ovary and premenstrual syndrome was suggested as early as 1931,17 and it has been common practice to treat premenstrual syndrome with the COCP to inhibit ovulation. However, no satisfactory controlled studies supporting the effectiveness of this treatment have been published.12

Acne responds well to treatment with most COCPs.13 The mechanism of suppression of acne appears to involve partly a decrease in ovarian secretion of testosterone, partly an increase in the production of sex-hormone-binding globulin, and partly antiandrogenic effects (eg, with cyproterone acetate). Hirsutism is less likely to respond to COCPs, and usually requires substantially higher doses of an antiandrogen.18

The evidence that the COCP alleviates other cyclical symptoms, such as mid-cycle pain, perimenstrual migraine, menses-related epilepsy and rarer symptoms, is limited. Many of these conditions are so uncommon that randomised treatment trials are not feasible. The most effective approach to treating these conditions may be the continuous use of the COCP (ie, with no monthly break) or use of a progestogen-only method that inhibits ovulation, provided that breakthrough bleeding is not a problem.19

Of the newer hormonal methods of contraception, the one that probably has the greatest health benefits is the levonorgestrel intrauterine system. After one year of use, this system, which involves local release of high concentrations of progestogen, produces a dramatic 95% reduction in measured menstrual blood loss in menorrhagia caused by ovulatory dysfunctional uterine bleeding.20 It probably has similar effects in most women with menorrhagia due to anovulatory dysfunctional uterine bleeding (including endometrial hyperplasia), small intramural fibroids and coagulopathies. Recent small-scale studies also indicate substantial symptomatic benefits with adenomyosis and endometriosis.21,22 The menstrual effects of this therapy are discussed in more detail in an article by Hickey and Farquhar in this issue of the Journal (page 625).23

Possible small associations between the COCP and breast or cervical cancer have been given extensive publicity.24 By contrast, the very large degrees of protection afforded COCP users against endometrial and ovarian cancer are much less well known. The long-term risk of ovarian cancer is reduced by 40% after 4 years of COCP use, 54% after 8 years and 60% after 12 years.25 The risk of endometrial adenocarcinoma is reduced by 56% after 4 years of COCP use, 67% after 8 years and 72% after 12 years. Protection against these two forms of cancer continues for many years after discontinuation of COCP use,25 and appears to be related to the progestogenic component of the pill.26 There is also reasonably sound evidence from case–control studies that long-term use of the COCP provides some protection against the later development of uterine fibroids, endometriosis, recurrent ovarian cysts, acute pelvic inflammatory disease, infertility, iron-deficiency anaemia, benign breast lumps, toxic shock syndrome, premenstrual syndrome, acne and hirsutism.27 There is less substantial evidence for beneficial effects in reducing the later incidence of thyroid disease, rheumatoid arthritis, duodenal ulcer and Trichomonas vaginalis infection, and in assisting long-term maintenance of bone mineral density.

Evidence of efficacy of the COCP in preventing other conditions is summarised in Box 2. Limited evidence is emerging to indicate that long-acting progestogen-only contraceptive methods will also offer some degree of protection against many of these conditions. We are confident that longer-term experience with the levonorgestrel intrauterine system will show that it provides substantial protection against endometrial hyperplasia28 and cancer, but has little or no influence on ovarian cancer. The levonorgestrel intrauterine system may thus be an ideal contraceptive method for women who are perimenopausal or have risk factors for conditions such as polycystic ovary syndrome. These situations are commonly associated with anovulation and an increased risk of endometrial hyperplasia.

Data from individual studies as to whether depot medroxyprogesterone acetate and the levonorgestrel intrauterine system protect against acute episodes of pelvic inflammatory disease are conflicting.29-31 Depot medroxyprogesterone acetate appears to help prevent recurrent vaginal candidiasis,32 but it is too early to say what other conditions may be alleviated by long-term use of progestogen-only methods.

Over the past 40 years, information has gradually accumulated to show that a number of non-contraceptive health benefits should be an important part of the decision-making process about COCP use. The same will almost certainly apply to the newer long-acting progestogen-only methods. Awareness of these potential advantages of hormonal contraception may assist women in the increasingly complex matter of choosing a contraceptive method.33