Abstract
During the transition years leading to menopause, the possibility of conception persists, although at a lower rate. Contraceptive choices available to perimenopausal women are as varied as those for their younger counterparts, albeit with some limitations related predominantly to coexisting medical conditions rather than the advancing age itself. In this review, different contraceptive choices pertaining to this age group will be discussed, with a focus on evidence-based data.
Keywords
Natural menopause is the complete cessation of menses secondary to depletion of ovarian follicles. The median age, the age at which 50% of women reach menopause, was determined to be 51.4 years among American women [1]. In addition to environmental factors and ethnicity, genetic factors have been implicated in contributing to menopausal age [2]. While the word ‘perimenopause’ strictly means the period prior to and after the onset of menopause, it has been commonly used to refer to the ‘transition years’ leading to menopause. The exact duration of this stage in the lifecycle, more accurately named ‘pre-menopause’, has been ill defined. While there is suggestion of an accelerated loss of follicles after the age of 38 years [3], the onset of irregular bleeding episodes, after the age of 45 years, was arbitrarily chosen to delineate the beginning of perimenopause in the Massachusetts Women's Health Study [4]. However, in a systematic review addressing the accuracy of the diagnosis in women older than 40 years and with bleeding irregularities, no single symptom or laboratory test was found to be sufficient to rule in or out the diagnosis of perimenopause [5]. Ovulation becomes irregular, menstrual cycles are lengthened or shortened, and menstruation becomes unpredictable. Episodes of severe fluctuation of estrogen production are encountered [6], while hot flashes are experienced by up to 50% of women before menopause [7].
Fecundity, the probability of a live birth per ovulation, progressively decreases [8]. This is not only the result of decreased fecundability (the probability of pregnancy per ovulation), but also of an increased rate of subclinical miscarriages observed with advancing age [9]. This phenomenon reflects a change in the quality of the oocyte rather than the receptive endometrium [10]. In addition, conditions negatively affecting fertility, such as endometriosis and leiomyomas, are not uncommon in women in their fifth decade. Nevertheless, approximately 45% of women aged over 40 years old are capable of carrying a pregnancy to term [11]. The exact fertility or pregnancy potential of women during perimenopause is difficult to calculate on a per-year basis. In a classic study of Hutterite couples practicing no contraception, 13% of women over the age of 45 years were still fertile [12].
In a recent USA survey, 46% of women seeking abortion were not using contraception. Out of these, a third had a perceived ‘low risk’ of pregnancy [13]. However, the attitude towards sexuality does not differ consistently across a woman's lifespan [14]. Consequently, adequate contraception counseling should be an integral part of reproductive healthcare during perimenopause.
Combination oral contraceptives
Since the introduction of combination oral contraceptives (COCs) in the early 1960s, there has been a trend towards lowering the estrogen and progestin dosages and introducing progestins with fewer side effects.
From the early 1980s, pills containing 35 μg or less ethinyl estradiol (EE2) largely replaced the ‘older generation’ EE2 50 μg pills [15].
Under this heading, COC refers to formulations containing 35 μg or less EE2, except if otherwise specified.
Perfect use of COCs is associated with one of the lowest failure rates among all contraceptive options, approximately 0.3% within the first year of use [16].
Combination oral contraceptives & venous events
Data collected during the 1980s, i.e., at the change of the estrogen dose in the marketed COCs in the USA, revealed that the risk of venous thromboembolism (VTE) increases with the estrogen dose in the combination pill [17]. In a UK database search, the risk of VTE among COC users had the most significant association with obesity, followed by smoking, general poor health and asthma [18]. However, it was notable that the risk sharply increases after the age of 39 years. This age relation was consistent with a 1998 WHO report that stated that the risk of VTE among COC users in their fifth decade, 18 per 100,000 woman-years, is twice the risk found when users are in their third decade [19]. However, as mentioned in the same report, the rise in the incidence of VTE with age follows the same pattern as in nonusers.
When all factors are analyzed independently, the presence of hereditary thrombophilias represents the strongest association with VTE. In a recent meta-analysis, the combination of COC use and the homozygous or heterozygous mutation for factor V Leiden, deficiencies of antithrombin protein C or protein S, and elevated levels of factor VIIIc prothrombin, was analyzed [20]. The combination of two risk factors, COC use and a thrombogenic mutation, was found to increase the incidence of VTE by two- to fivefold, depending on the mutation, compared with women with only one risk factor. In addition, the risk of cerebral vein thrombosis, a rare life-threatening complication, is increased among carriers of these thrombogenic mutations, especially factor V Leiden [21]. As at least 1–2% of the population carries one of these thrombophilias, and considering the very low incidence rate of VTE, the positive predictive value for thrombophilia screening prior to initiation of COC is poor. In addition, screening fails to identify over two-thirds of women who will develop VTE while receiving COC [22].
It is noteworthy that the presence of superficial varicose veins, probably more common during perimenopause than in earlier life, does not increase the risk of VTE [23].
The estrogen component of the COC may not be the exclusive culprit in initiating the cascade of events leading to VTE. Two of the third-generation progestins used in some COC preparations, desogestrel and gestodene, were implicated in a higher risk of VTE (two- to three-fold), compared with levonorgestrel [23]. However, these findings could have been compromised by confounding variables such as age and prescription bias. In fact, no comparative data among currently marketed progestins is sufficient to make a recommendation regarding a preferred specific progestin COC component during perimenopause.
Surgery, a risk factor for VTE, is more common with advancing age. Discontinuing COC 6 weeks preoperatively is advisable in order to allow sufficient time for the procoagulant changes to return to normal. The other alternative would be to administer heparin prophylaxis to current users [24].
Combination oral contraceptives & arterial events
When risk factors for myocardial infarction (MI) are adjusted for, COC use alone was estimated to contribute to one MI case/100,000 woman-years [25]. Among all factors, smoking has been cited as having the highest synergistic effect with COC [26]. In addition, all data concur that prior COC use does not increase the probability of MI. It is now safe to state that when currently marketed COC formulations are used, age does not seem to play a role in the development of MI, provided the user is a nonsmoker, nondiabetic and normotensive [26,27].
Stroke, classified as either hemorrhagic or ischemic, has the strongest association with hypertension, smoking and a history of migraine. In normotensive nonsmoking women, a history of migraine acts synergistically with COC and increases by several folds the risk of ischemic stroke [28].
In a review of the articles published between 1970 and 2000, it was concluded that the association between stroke and COC as an independent factor “is tenuous at best and perhaps nonexistent” [29]. As with VTE, the incidence of stroke increases with age, irrespective of COC use [30]. Since a history of migraine headaches, especially with no focal neurological signs, is not uncommon among women of all age groups, it is vital to have clear guidelines on COC use in this population. The American College of Obstetricians and Gynecologists (ACOG) advises consideration of COC in women with migraines to be limited to those younger than 35 years of age [27].
Combination oral contraceptives & metabolic effects
When metabolic effects were evaluated in COC users aged over 40 years, clinically insignificant changes were observed [31]. Although many studies confirmed the safety of COC use among women with controlled diabetes or dyslipidemia, ACOG suggests avoiding such use in women over 35 years of age [27].
During perimenopause, women may be more concerned about possible weight change. Although studies addressing the effect of COC use on weight are controversial, a recent Cochrane review found no evidence of a significant effect of COCs on weight gain [32]. However, no relevant data exists in the perimenopausal age group.
‘Lower’ estrogen combination oral contraceptives
Assuming that most of the above-mentioned potential side effects are related to the estrogen dose, does a further reduction of EE2 below 30–35 μg confer a better safety profile? COCs with EE2 20 μg were shown to have the same efficacy, but poorer cycle control, when compared with the higher estrogen dose formulations, possibly resulting in more frequent discontinuation [33]. However, to date, insufficient comparative data regarding safety is available, especially in the perimenopausal age group. Minesse® (Wyeth-Ayerst Pharmaceuticals, Watson Laboratories, Inc., TX, USA), containing an ultra-low dose of 15 μg EE2 and 60 μg gestodene per tablet, was approved for contraceptive use in EU countries in the year 2000. It is supplied in packages of 24 active pills and four placebo pills. Limited data reveals good efficacy and adequate cycle control with time [34]. Minesse is not currently available in the USA market. Lower estrogen component potency, i.e., the equivalent of 0.625 μg of conjugated estrogens or 2 μg micronized estradiol, is found in combination hormone replacement therapy (CHRT) approved to treat vasomotor symptoms. Unfortunately, CHRT, in its current form, was not found to offer adequate contraception [35].
Benefits of combination oral contraceptives
In addition to preventing unintended pregnancies, the use of COCs by women of perimenopausal age confers well-documented benefits. Vasomotor symptoms and vaginal dryness, sometimes experienced before the actual menopause, respond to COCs [36]. Cycle regulation is another clear benefit during a time where bleeding irregularities are more common [37]. When bothersome uterine bleeding is in the form of menorrhagia, COCs are found to be at least as effective as other medical treatment forms [38]. Considering that hysterectomy is most frequently performed for dysfunctional uterine bleeding and is associated with a 3% risk of severe operative complications [39], the theoretical role of COCs as an alternative to surgery cannot be overlooked.
In the fifth decade of a woman's life endometriosis and leiomyomas, conditions that could be associated with dysmenorrhea, are prevalent. A frequent off-label indication for prescribing COCs continues to be dysmenorrhea, probably supported by many observational studies. In fact, COC use has not been proven to relieve dysmenorrhea when randomized, controlled trials (RCTs) were analyzed in a 2001 Cochrane review [40]. However, in a more recent RCT, a formulation containing EE2 20 μg was found to be effective in relieving dysmenorrhea among adolescent girls [41].
Although many studies concluded that COC use does not increase the risk of developing leiomyomas [27], the effect of such use on the size of existing leiomyomas is uncertain. It is noteworthy that an increase in size was found in the first 2 years after CHRT [42].
COC use does not seem to offer protection against the development of functional ovarian cysts, a more common event during perimenopausal years. Data suggesting such a protection mostly originated from studies using higher EE2 formulations than those currently marketed for contraception [43].
Despite a long-claimed benefit of COCs in the risk of pelvic inflammatory disease, careful evaluation of the available data leads to inconclusive results [44].
Combination oral contraceptives & the bone
While the imbalance between bone formation and bone resorption starts in the mid-twenties, bone loss is accelerated in the early menopausal years, albeit at varying rates [45]. Although COCs have been proven to preserve bone mineral density (BMD) during perimenopause [46], the correlation between BMD in this age group and the risk of fractures is questionable [47]. Indeed, in the large, prospective cohort multicenter ‘Women Health Initiative’ study, prior use of COCs did not protect against fractures in later life [48]. However, a Swedish retrospective case–control study showed that past users of COCs, aged 50–81 years, had a 25% reduction in the risk of hip fractures [49].
Combination oral contraceptives & cancer
The possible association between COCs and breast cancer gains special importance during perimenopause, when the incidence of breast cancer is on the rise. Epidemiological data suggest that COC use, similar to pregnancy, may accelerate the proliferation of an already existent breast cancer nidus. This hypothesis is supported by the fact that although the relative risk (RR) of breast cancer among current users is 1.24, it gradually declines to baseline risk by 10 years after discontinuation [50]. In one retrospective study, breast cancer risk was found to correlate with the progestin potency and estrogen dose (>35 μg), more so in women younger than 35 years [51]. The possibility of declining risk in older women, i.e., during perimenopause, warrants further study.
Burkman and colleagues engaged in a statistics exercise to determine the actual effect of low-dose COCs on breast cancer risk in different age groups [44]. They concluded that this risk, while still a function of the duration of use, diminishes with advanced age. Among women aged 45–59 years, the RR in COC users for 4 years is a nonsignificant 1.02. When prior use totaled 8 years, the RR remains at a low 1.03.
COC use is not contraindicated in women with benign breast disease or those who have first-degree relatives with breast cancer [50]. Interestingly, a retrospective, multicenter matched case–control study on 1311 pairs of women with known deleterious breast cancer, early onset (BRCA)1 and/or BRCA2 mutations revealed that prior COC use for more than 5 years slightly increased the risk of breast cancer in BRCA1 (odds ratio: 1.33), but not BRCA2 carriers [52].
COCs, possibly by suppressing ovulation, have been found to decrease the risk of epithelial ovarian cancer [53]. In one of the earliest publications addressing this issue, users of COCs containing EE2 50 μg were found to be 40% less likely to develop epithelial ovarian cancer [54]. This effect, found among women aged 20–54 years, persisted for up to 15 years after discontinuation. A year 2000 case–control study concluded that the ovarian cancer protective effect attributed to COC does not change with lower-dose formulations [55]. Having a first-degree relative with epithelial ovarian cancer does not seem to affect the risk reduction in one study [56]. In addition, limited data suggest that this protective effect may extend to carriers of BRCA1 and BRCA2 mutations as well [57].
Reduction in the risk of endometrial cancer has been similarly demonstrated in COC users. Literature review revealed that the protective effect, in the magnitude of 40–80%, persists for many years after discontinuation, and may be correlated with duration of use [58].
Less conclusive evidence exists regarding a possible decrease in colorectal cancer incidence in COC current or recent users. In a 2001 meta-analysis, the calculated pooled RR was 0.82 for all COC use, independent of the duration [59].
There has been a continuous debate regarding the association of cervical intraepithelial neoplasia or cervical cancer and use of COC. To date, a causal relationship has not been established. Many confounding variables are present, most notably the effect of condom and other barrier contraceptive use, and the number of sexual partners, altering the risk of transmission of human papilloma virus (HPV). However, a recent meta-analysis found that, even after adjusting for barrier-method use, there is a steady increase in the RR of carcinoma in situ and invasive cancer of the cervix among COC users, as a function of duration, reaching 2.2 after 10 years of use in all women, and 2.5 in HPV-positive women [60]. As of the writing of this manuscript, COC product labels include a warning about a possible risk of cervical neoplasia.
Extended combination oral contraceptive regimens
Practitioners have frequently prescribed COC in longer active pill cycles (>3 weeks), for various reasons. In 2003, the US FDA approved Seasonale® (Barr Laboratories, NY, USA), an extended COC regimen, with 84 active pills (Levonorgestrel 0.15 mg/EE2 30 μg) and seven placebo pills. The hormonal ingredients are identical to those found in Nordette®. By reducing the number of bleeding episodes from 12 to 4 per year, this method may appeal to perimenopausal women. In fact, in a study conducted in the Netherlands, the proportion of women who preferred having less frequent cycles while on COC increased with age [61].
The combination contraceptive patch
In 2002, Ortho Evra® (Ortho-Mcneil Pharmaceuticals, NJ, USA), a transdermal patch designed to release EE2 20 μg and norelgestromin 150 μg daily, was approved for contraception by the US FDA. The skin patch, worn on the lower abdomen, buttocks or upper body, is changed once/week for 3 consecutive weeks. Withdrawal bleeding is expected during the patch-free week. A review of experience with its use revealed efficacy, tolerability and side effects to be comparable to ‘traditional’ COC, with possibly improved compliance [62].
In 2005, the US FDA, based on recent unpublished studies comparing the mean pharmacokinetic profiles of the patch with a ‘typical’ oral contraceptive containing EE2 35 μg, issued a product label warning. The bolded warning specifically states that women who use Ortho Evra are exposed to approximately 60% more total estrogen in their blood than if they were taking a typical daily birth control pill. However, a more recent study showed the incidence of nonfatal VTE, known to be a function of estrogen exposure, to be comparable between Ortho Evra and Orthocyclen® (Ortho-Mcneil Pharmaceuticals, NJ, USA) users in women aged 15–44 years [63]. Nevertheless, in view of this warning, perimenopauasal women may not be ideal candidates for this contraceptive method until more data are available.
The combination contraceptive ring
NuvaRing® (Organon USA Inc., NJ, USA), a thin, flexible, controlled-release vaginal ring contraceptive, was introduced in the USA in 2002 after receiving US FDA approval. Measuring 54 mm in outer diameter and 4 mm in cross-section, it is designed to release EE2 15 μg and etonogestrel 120 μg daily for 3 weeks, thus preventing ovulation. After 3 weeks, the ring is removed for 1 week, during which time menses occurs. Efficacy, tolerability and safety profiles are comparable to COC [64]. While the systemic exposure to the progestin is similar to that of a COC containing desogestrel 150 μg (e.g., Marvelon®), the estrogen exposure is approximately half that of the COC pill, and a third that of Ortho Evra [65]. This makes it a theoretically attractive contraceptive option for women of advanced reproductive age. NuvaRing can be easily inserted and removed by the woman herself, once every 3 weeks, as no exact position in the vagina is needed for effective functioning. As it is not a barrier method, it is not expected to protect against sexually transmitted diseases (STDs).
Progestin-only pill
Often referred to as the ‘mini-pill’, the progestinonly pill (POP) prevents pregnancy predominately by altering the cervical mucus and endometrium, while its ovulation suppression effect is variable [66]. The 1-year pregnancy rates range from 0.5–5% [66]. As it is taken continuously, no regular withdrawal bleeding is expected. Instead, irregular bleeding occurs among almost half of users [67]. This is a significant drawback, making the method unpopular among many users. While studies regarding its safety have not been of the magnitude of COC studies, currently available information supports the theoretical assumption that serious potential adverse events related to estrogen use, such as VTE, stroke, and MI, are not increased with POP [68]. This finding may be enough to consider POP as a viable perimenopausal contraceptive option. The possibly ‘suboptimal efficacy’ may be balanced theoretically by the decreased fecundity in this age group. Irregular spotting may be remedied by adding a small dose of estrogen, such as the one used in postmenopausal estrogen replacement therapy, if appropriate [67]. At present, limited available data suggests that current POP use does not affect BMD [69].
In a cohort study enrolling more than 100,000 women, past exclusive use of POP (i.e., without COC) was not found to increase the risk of breast cancer [70]. By contrast, recent and current exclusive POP users had an overall 50–60% increased risk of breast cancer, compared with nonusers. However, the risk becomes nonsignificant in the 40–49 year age group, albeit with a wide confidence interval. Although this is the only prospective study designed to evaluate the effect of POP and COC separately, the findings, based on a small number of breast cancer cases among POP users, may have limited statistical power.
A POP containing desogestrel 0.075 mg has been approved for contraceptive use in many countries (Cerazette®, Organon USA Inc., Rose-land, NJ, USA). What differentiates it from other POPs is a much lower failure rate, approaching that of COCs, probably due to better ovulation suppression [71]. Bleeding patterns and adverse events are comparable to other POPs. As of the writing of this manuscript, the manufacturer has not applied for US FDA approval.
Injectable progestins
Depot medroxyprogesterone acetate (DMPA), administered intramuscularly in 150 mg doses every 3 months, is another contraceptive choice in perimenopausal women who present a contraindication to estrogen use. In this age group, the slow return to fertility may not be an issue, while improvement in vasomotor symptoms is an added benefit [72]. The potential bone loss following prolonged use, possibly due to estradiol suppression, has been the subject of a long debate, although numerous studies have documented the reversal of this bone loss following discontinuation of DMPA, or with simultaneous estrogen add-back therapy [69]. However, discontinuation among perimenopausal women would occur at a time when ‘natural’ accelerated bone loss takes place. Nevertheless, two studies found no negative effect on bone density among a small number of users of this age group [73,74]. In November 2004, the US FDA issued a black box warning on the long-term use of DMPA, stating that bone loss might be irreversible with use of more than 2 years. The current ACOG recommendation is to “exercise caution” in prescribing DMPA for perimenopausal women [27].
Notwithstanding some past conflicting reports possibly affected by surveillance bias, currently-available data does not link, with a sufficient degree of certainty, DMPA use to breast cancer development, especially among long-term users [75]. Nevertheless, the possibility of occult cancer growth acceleration during DMPA use mandates yearly surveillance at all ages.
Depo-subQ provera 104™ (DMPA-SC, Pfizer, NY, USA), marketed in a prefilled syringe containing medroxyprogesterone acetate 104 mg for subcutaneous administration once every 3 months, received US FDA approval for use as a contraceptive in December 2004. It was also approved as a treatment for endometriosis-related pelvic pain 4 months later. Preliminary noncomparative data from short-term studies suggest that efficacy, bleeding pattern and safety profile are similar to DMPA [76]. The theoretical advantages of a lower serum progestin level as well as the long-term effect on bone mass are not yet established.
The progestin subdermal implant
In 2004, Implanon® (Organon USA Inc., NJ, USA), a single subdermally implantable rod, 4 cm in length and 2 mm in diameter, containing the progestin etonogestrel, was approved for contraception by the US FDA for up to 3 years. This method was shown to be safe, highly effective, and rapidly reversible [77]. In a 2-year multicenter study, most users had infrequent bleeding episodes, an expected phenomenon, while a sizeable 36% withdrew due to vrious ‘adverse events’ led by bleeding irregularities, emotional lability (6%) and weight gain (3%) [77]. One study specifically evaluated the use of Implanon in a small number of women aged over 35 years and found no age-specific side effects [78].
Intrauterine devices
Currently, two intrauterine device (IUD) types are approved for contraceptive use in USA: a copper-containing IUD (Paragard®), and a progestin-coated IUD designed to release 20 μg of levonorgestrel (LNG) per day (Mirena®). Despite a historically exaggerated fear of salpingitis associated with IUD use, an extensive review showed that the small risk of salpingitis is limited to a brief period following insertion [79]. After the first month, users do not have any increased risk of ascending pelvic infection while using the device. Paragard, introduced in 1988 and approved for 10 years of use, boasts an efficacy equivalent to tubal sterilization [16]. While new onset of menorrhagia and cramps may constitute reasons for early discontinuation [80], a woman who is satisfied with copper IUD insertion while in her 40s may not need another contraceptive method throughout her reproductive years. ‘Irregular bleeding episodes’, should they occur long after insertion, are not cause for removal of the copper IUD, as this bleeding pattern is not uncommon in perimenopause.
The LNG IUD, introduced to the US market in 2000 and approved for 5 years of use, has demonstrated a better contraceptive efficacy than copper IUD, countered by a lower continuation rate [81]. A major advantage is the reduction of menstrual flow, menstrual frequency and blood loss. Consequently, LNG IUD has been used outside the contraceptive domain to successfully treat menorrhagia, whether idiopathic or secondary to adenomyosis or leiomyomas, thus presenting a viable alternative to hysterectomy [82]. When vasomotor symptoms are experienced during perimenopause, LNG IUD may serve as the progestogen component, should one elect to supplement with estrogen replacement therapy [83]. In a group of women with a mean age of 44 ± 7 years, LNG IUD was found to be safe in respect to lipid metabolism, blood pressure and liver function tests [84]. A small sample study found no difference in forearm BMD between long term users of LNG IUD, copper IUD or controls [85]. In addition, data from a postmarketing surveillance study enrolling more than 17,000 users suggested there is no increased risk of breast cancer in any age group [86].
Major reasons for patient dissatisfaction with this method relate to the known side effects of progestogen. Significantly higher rates of amenorrhea, skin and hair conditions, and headache were observed with the steroid IUD than with the copper-releasing IUD [87].
Frameless IUD
In a frameless IUD, the plastic T-shaped frame is missing. GyneFix® (Contrel, Ghent, Belgium) is a copper frameless IUD consisting of small copper sleeves threaded on a suture string. Anchored to the uterine fundus by an experienced healthcare provider, it is claimed to be associated with a lower rate of menorrhagia due to the absence of the inert material [88]. Currently, GyneFix is not being marketed in the USA.
Fertility awareness-based methods
In a recent review, fertility awareness-based methods were found to be associated with low efficacy when used for contraception [89]. Particularly in the perimenopausal age group, these methods, in addition to ‘the rhythm method’, cannot be sufficiently relied upon, in view of the irregularity of ovulation.
Barrier methods
Advantages for barrier methods include absence of serious side effects and protection from STDs. The relatively high failure rate may be overcome, theoretically, by combining two methods, such as a diaphragm and a condom [16].
Femcap®, available since 2002, is a silicone rubber cervical cap claimed to be easier to fit and remove due to its sailor's hat shape. It was found to carry a higher failure rate than the diaphragm in a review of small studies [90].
Lea's shield®, another recently introduced cervical cap available in one uniform size, is designed with a valve allowing drainage of cervical secretions. Efficacy is improved by adding a spermicide [91].
While both the cervical cap and the sponge display notably lower efficacy in parous compared with nulliparous women [16,91], any barrier method may be adequate during perimenopause, particularly in those with low coital frequency.
Postcoital contraception
Often referred to as ‘emergency contraception’ or ‘the morning-after pill’, postcoital contraception has been prescribed, for a long time, to women following unprotected intercourse or method misuse. Although it historically consisted of an off-label use of COC pills 12 h apart [92], Plan-B® (Barr Pharmaceuticals, NJ, USA) is currently the only USA-marketed, US FDA-approved medication product for postcoital contraception. It is packaged as two pills, each containing LNG 0.75 mg, with instructions to take the first pill as soon as possible within 3 days of intercourse, and the second pill 12 h later. There is a reported efficacy of 85% in all ages, implying that 85% of potential pregnancies are prevented, and not that 15% of users get pregnant. Different mechanisms have been proposed for the method of action, none of which include the possibility of interruption of an implanted pregnancy [92]. No teratogenic effects were linked to use in case of failure.
Although a combined panel of the US FDA's Advisory committees concluded, in 2003, that the safety of the levonorgestrel-only emergency contraception regimen has been sufficiently demonstrated to warrant approval for nonprescription status, Plan-B is still only available on a prescription basis as of the writing of this manuscript. While not advocated as a primary ‘ongoing’ contraceptive method, this option is underutilized, partly due to limited consumer awareness [92,93], and arguably limited availability.
Among women seeking abortion, who had not been using contraception, 26.5% reported the pregnancy to be the result of ‘unexpected sex’ [13]. Using a conservative estimate of 75% effectiveness, postcoital contraception would have prevented an estimated 51,000 terminated pregnancies in the year 2000 in USA [13]. In a 2005 ACOG technical bulletin, it was noted that, despite lack of data, progestin-only emergency contraception may be made available to all women, including those with contraindications to the use of COC [93].
Finally, it is noteworthy that IUD insertion, up to 5 days after intercourse, is 99% effective in preventing pregnancy [92].
Female sterilization
Female sterilization is the most popular contraceptive method among women in their 40s. Failure, estimated at around 0.5% [16], is not method-dependent, as previously suggested [94]. In addition to its benefit in decreasing the risk of ovarian cancer [94], it is an attractive method in the perimenopausal age group, especially given that poststerilization regret has an inverse relationship with a woman's age [95].
In November 2002, the US FDA-approved Essure PBC® device (Conceptus Inc., San Carlos, Mexico), consisting of small metallic implants placed in the ostia of the fallopian tubes under hysteroscopic guidance, for female sterilization. This can be performed in the office under local anesthesia. In one study comparing hysteroscopic to laparoscopic sterilization, the hysteroscopic device placement, while successfully completed in only 81%, was associated with a shorter hospital stay and better patient satisfaction [96].
When to stop contraception
Determining the onset of menopause may not be straightforward, especially in the presence of regular bleeding episodes in case of COC use, or amenorrhea while receiving POPs. In one study, serum follicle stimulating hormone (FSH):LH ratio over 1, or estradiol below 20 pg/ml on the seventh day of the COC pill-free interval were deemed appropriate for diagnosing menopause, while FSH alone was not [97]. This was consistent with findings from an earlier study suggesting that, in menopausal women, estradiol is always decreased by the first week of the COC pill-free interval, while simultaneously measured FSH may occasionally remain in the premenopausal range [98]. In the case of DMPA use, a menopausal woman would have FSH level in the menopausal range when measured late in the contraceptive cycle [99].
Future perspective
Recently introduced hormonal contraceptives have particularly benefited perimenopausal women (
Hormonal contraceptives introduced in the 21st century, with particular benefits to women in perimenopause.
COC: Combination oral contraceptives; DMPA: Depot medroxyprogesterone acetate; EE2: Ethinyl estradiol; LNG: Levonorgestrel; POP: Progestin-only pill.
Late-phase clinical trials of potential hormonal contraceptives containing molecules already existing in previous preparations, but presented in lower doses or extended regimens, are ongoing. These include 84-day cycle and 1-year cycle combination formulations of LNG 0.1 mg and EE2 20 μg per tablet [100].
‘Miniature IUDs’, releasing LNG 14 μg/day (instead of LNG 20 μg released by Mirena), are also undergoing late-phase clinical trials. Major advantages include a much lower incidence of progestin-related side effects, while maintaining an efficacy similar to the LNG 20 μg IUD [101].
The substitution of EE2 by 17-β estradiol, or the addition of estradiol valerate to an ultra-low dose of EE2 (10 μg) in COC preparations, is being studied, especially in combination with the progestin dienogest. The potential advantages of the alternative estrogens concerning the hemostatic and metabolic systems, compared with EE2, have not been validated in large trials yet [100,102].
Novel progestins have been synthesized, and are in trial in COC preparations. These molecules have been designed to have no androgenic or estrogenic actions and to be closer in activity to progesterone, while being absorbed orally. Such characteristics may limit the metabolic ill effects of ‘older’ progestins [103]. In addition, a metered-dose transdermal spray containing nestorone (a progestin with poor oral absorption) is undergoing early phase trials for contraceptive use.
Mifepristone, the first synthetic progesterone antagonist, was synthesized more than two decades ago and has been used for early pregnancy termination in many countries. When used in daily doses of 2 or 5 mg in a small number of women, it was found to have a high contraceptive efficacy, mostly by suppressing ovulation [91].
The majority of users remain amenorrheic, but not hypoestrogenic, throughout the duration of use [91]. In view of the acceptable safety profile, more research is in progress regarding its use in contraception.
In addition, different progesterone agonists and antagonists, steroidal and nonsteroidal, administered orally, vaginally, or via an intrauterine system, are in early clinical phase evaluation for use as contraceptives [100].
Excutive summary
Perimenopause has been arbitrarily defined as the period starting at the onset of irregular bleeding episodes in the fifth decade of life and ending at menopause. No single laboratory test is sufficient for diagnosis.
The fecundity potential of a woman gradually declines with advancing age, due to decreased fecundability and an increased spontaneous abortion rate per conception. Nevertheless, contraceptive counseling should be practiced until menopause.
Combination oral contraceptive (COC) use is safe in women during perimenopause, except in the following: current smokers, those with cardiovascular diseases including hypertension, those with metabolic disorders including diabetes, known carriers of a thrombogenic mutation and those with a history of migraine.
During perimenopause, the risk of breast cancer secondary to the effect of COC use is negligible.
Well-established benefits of COC use include a risk reduction of ovarian and endometrial cancers.
Recognized benefits of COC use during perimenopause include improvement of menorrhagia, dysfunctional uterine bleeding and vasomotor symptoms.
COC with lower ethinyl estradiol (EE2) doses, as well as the NuvaRing®, may present theoretical advantages in their safety profiles, compared with COC with EE2 30 μg, when used by perimenopausal women.
Perimenopausal women are not ideal candidates for use of Ortho Evra®, the combination contraceptive patch, due to possibly higher estrogen blood levels compared with COC.
Progestin-only contraceptives are considered safe during perimenopause.
DMPA long-term use has the risk of bone loss that may not be regained if the method is discontinued at menopause.
Presently, there is no universal opinion regarding a possible link between progestin-only contraceptives and breast cancer.
Levonorgestrel (LNG) intrauterine device (IUD) may be used as a therapeutic modality for gynecological bleeding problems of benign etiology.
Both copper and LNG IUDs are excellent, yet underutilized, contraceptive choices for the perimenopausal woman in view of the long duration of action, high efficacy and excellent safety profile.
Perimenopausal women may use barrier methods for contraception, especially when there is a need for sexually transmitted disease protection.
The contraceptive efficacy of fertility awareness methods is further compromised during perimenopause, due to irregular ovulation.
There are no contraindications for progestin-only postcoital contraception in any reproductive age group.
Poststerilization regret is less common in women with advanced reproductive age than their younger counterparts.
European products that may benefit perimenopausal women if introduced in the USA include a COC with EE2 15 μg, a progestin-only contraceptive with desogestrel 75 μg and a frameless copper IUD.
Extended COC regimens of lower estrogen dose and IUDs with a lower LNG release rate are in late-phase clinical trials.
Early phase clinical trials include COC with different estrogens, and novel progestins that may limit thrombotic and metabolic adverse events.
Mifepristone, a synthetic progesterone antagonist, has shown promise for use as a contraceptive in view of good efficacy and proven safety profile.
Other progesterone agonists and antagonists are being considered for contraceptive use.