U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Population. Contraception and Reproduction: Health Consequences for Women and Children in the Developing World. Washington (DC): National Academies Press (US); 1989.

Cover of Contraception and Reproduction

Contraception and Reproduction: Health Consequences for Women and Children in the Developing World.

Show details

4Contraceptive Benefits and Risks

Pregnancy and childbirth carry risks of morbidity and mortality. Although the contraceptives that couples use to avoid pregnancy have their own health risks, they also have substantial noncontraceptive health benefits. Information about these risks and benefits is necessary for informed decision making. Oral contraceptives, for example, not only prevent pregnancy, but they also reduce the risk of endometrial and ovarian cancer and protect against acute pelvic inflammatory disease and ectopic pregnancies. However, oral contraceptives increase the risk of cardiovascular disease. IUDs provide effective contraception but increase the potential for infection in certain high-risk groups. Barrier methods of contraception provide less effective contraception, but they protect against sexually transmitted infections including human immunodeficiency virus (HIV). The importance of the noncontraceptive benefits and risks of contraceptives varies among societies because of variations in the prevalence of the diseases involved.

This chapter reviews evidence on the effectiveness and health consequences of specific contraceptive methods. Our attention is limited to the biological consequences of a method's use, even though each method may have psychological risks and benefits. Our purpose is to provide an account of the direct health consequences of contraceptive use, independent of the effects that fertility control has on health by allowing women to control their fertility. This analysis is particularly important because, in some countries, health officials downplay the health benefits of lower fertility because they fear the adverse health effects of widespread use of modern contraceptives, especially in circumstances in which medical supervision of contraceptive practice is limited.

Our consideration of the effectiveness of contraceptives is based on a recent critical review of the literature by Trussell and Kost (1987). The studies they examined and most of the epidemiologic and clinical studies of the health effects of contraceptives have been carried out in developed countries. We recognize the difficulty of generalizing these results to the special health and cultural situations in the developing world. Furthermore, there are few studies of the effects of various contraceptive methods on the risk of diseases that are generally limited to developing countries. In many cases, the available data pertain to contraceptives that were commonly used in the 1960s and early 1970s and focus on the user population at that time. The research design, the quality of the data, the size of the sample, and the analysis have often been insufficient to allow definitive conclusions. Clearly, more studies conducted in developing countries are needed, and in fact studies sponsored by the World Health Organization are under way. Nevertheless, we regard the available information as a reasonable guide for estimating the risk of pregnancy versus the risks and benefits of contraceptive use in the developing world.

Oral Contraceptives

According to United Nations estimates, oral contraceptives are currently used by nearly 62 million women (United Nations, 1989). Two types of oral contraceptives (OCs) are available: combination OCs, consisting of the hormones estrogen and progestin, and the progestin-only pill (often called the minipill). Combination OCs are used by far more women, and as a result, most epidemiologic studies consider this type, particularly the formulations popular during the 1960s to mid-1970s. OCs prevent pregnancy primarily by inhibiting ovulation, although changes in the cervical mucous and endometrium may also have contraceptive effects. Failure rates associated with OC use are low—roughly 3 percent of women using OCs became pregnant in the first year of use, mainly because of improper or incomplete use (Trussell and Kost, 1987).

Health Benefits

A large cohort study in the United Kingdom has provided clear evidence that OC use decreases the risk of iron deficiency anemia in both current and past users (Royal College of General Practitioners, 1970). The effect is probably caused by the decrease in menstrual flow and consequent increase in iron reserves. This benefit may be especially important in developing countries in which iron deficiency is a problem (Stadel, 1986).

Case-control and cohort studies have found a decreased risk of benign breast disease associated with OC use (Stadel, 1986). The relative risk in women who have used OCs for more than two years compared with nonusers is about 0.6 for fibrocystic disease, 0.3 for fibroadenoma, and about 0.5 for unbiopsied breast lumps. This decreased risk does not persist in former users who have not used OCs for more than one year (Brinton et al., 1981). Since this effect is most likely to be related to the high progestin content of early formulations of the pill, current OC formulations may not decrease the risk of benign breast disease.

Several studies have found that OC use decreases the risk of functional ovarian cysts. This effect is probably due to the suppression of ovulation (Stadel, 1986). Evidence also suggests that OCs protect against uterine fibroids, the protection increasing with the duration of OC use (Ross et al., 1986). While there is still speculation about the mechanism, the protective effect against fibroids may be related to how the effect of circulating estrogens, which may promote the formation of fibroids, is modified by the progestins in OCs.

Several studies in developed and developing countries have found that current or recent OC use reduces the risk of pelvic inflammatory disease (PID), a major cause of female infertility (Stadel, 1986; Gray and Campbell, 1985). These studies have found that OC use lowers the risk by, on average, about 40 percent. Two mechanisms may be operative: OCs may change the cervical mucous such that it prevents pathogenic organisms from ascending into the upper genital tract; or OCs reduce menstrual blood flow, thus decreasing the amount of medium available for bacterial growth (Rubin et al., 1982). Unfortunately, most of the studies of oral contraceptives and PID have been hospital-based, so the results may not apply to women who are asymptomatic or who have PID not requiring hospitalization (Washington et al., 1985). For example, OCs may protect against gonorrhea, an important cause of acute PID that would require hospitalization, whereas other bacterial etiologies that cause less severe PID, such as chlamydia, may receive little or no protection from OC use.

Because they are highly effective at inhibiting ovulation, OCs greatly decrease the risk of ectopic pregnancy. Results from large case-control studies conducted in the United States and developing countries found that current OC users were 10 times less likely to have an ectopic pregnancy than women using no method (Ory and the Women's Health Study, 1981; Gray, 1984). Because the risk of death from ectopic pregnancy is high for women living in rural areas in the developing world, this effect is particularly noteworthy.

Another important benefit from OC use is a reduction in the risk of endometrial and ovarian cancer. Several epidemiologic studies have confirmed reduction of endometrial cancer among users. The Cancer and Steroid Hormone (CASH) study conducted in the United States found a 40 percent reduction in the risk of endometrial cancer, even long after OC use had been discontinued, and the benefit increased with the cumulative duration of pill use (Centers for Disease Control and the National Institute of Child Health and Human Development, 1987a, 1987b). The continued protection the pill provides to former users is not clearly understood, but apparently the carcinogenic effect of estrogen on the endometrium is obviated by the progestin in the pill.

The CASH study also found a 40 percent reduction in the risk of ovarian cancer (Centers for Disease Control and National Institute of Child Health and Human Development, 1987a, 1987b). Other epidemiologic studies have supported these findings. Suppression of ovulation and suppression of secretion of the hormone gonadotropin have both been postulated as mechanisms of this protection. It is noteworthy that there is consistent evidence from independent epidemiologic studies that the pill protects women from endometrial and ovarian cancer. Such consistency suggests true biological effect.

Adverse Health Effects: Cardiovascular Diseases

Cardiovascular diseases are a major cause of death in developed countries, where most research on the association between OC use and cardiovascular diseases has been conducted. These diseases are less common in developing countries, so alteration in their occurrence by OC use may not be as substantial as in industrialized countries.

OC use increases the risk of cardiovascular disease, in particular the risk of venous thromboembolism, myocardial infarction, and stroke (Stadel, 1986; Prentice and Thomas, 1987; Vessey, 1980). The risk of serious illness or death attributable to OC use from adverse cardiovascular effects is concentrated primarily among older women over age 30 and women who smoke cigarettes or have other cardiovascular risk factors. The excess risk of cardiovascular diseases seems to be directly related to both the estrogen and progestin content of the pill. And the risks may be substantially lower with newer low-dose preparations.

Venous thrombosis is the blockage of a vein by a blood clot particle. Thromboembolism occurs when the blood clot moves from a primary site to another, such as to the lungs or the brain. It is a major source of illness that may lead to death. Although the risk of venous thromboembolism is increased for current OC users, the increased risk does not persist among former users and is not related to duration of use (Vessey, 1980). The higher the estrogen content of the OC, the greater is the risk of venous thromboembolism, both for superficial and deep vein thrombosis (Stadel, 1986). The risk of venous thromboembolism among pill users appears to be unrelated to cigarette smoking. Mechanisms underlying increases in venous thromboembolism involve effects of estrogen or blood clotting factors that increase the coagulability of blood.

Myocardial infarction and stroke are much more important causes of mortality attributable to OCs. The risk is strongly influenced by age and by the presence of other cardiovascular risk factors, including cigarette smoking, hypertension, and diabetes. The annual risk of myocardial infarction attributable to current OC use rises from about 4 cases per 100,000 among nonsmoking OC users ages 30 to 39 to 185 cases per 100,000 among heavy-smoking OC users ages 40 to 44 (Stadel, 1986). Current OC use has been found to slightly elevate blood pressure in most women, possibly a contributing factor to the pathogenesis of myocardial infarction and stroke among current OC users. OC use leads to a three- to sixfold increase in the risk of overt hypertension, increasing with a woman's age and duration of OC use. It must be remembered that these risks pertain to use of the relatively high-dose pills of the 1960s and 1970s and their patterns of use in relation to such factors as age and smoking.

Other Possible Health Effects

Metabolic Effects

Metabolic changes may underlie the effects of OCs on myocardial infarction. Estrogens have the apparently desirable effect of increasing HDL-cholesterol (high density lipoprotein) concentration. Depending on type, progestins may either increase, decrease, or have no effect on HDL-cholesterol (for a complete discussion of changes in HDL-cholesterol, see Vessey, 1980). The net effect of different OC formulations on HDL-cholesterol is a function of both the dose of estrogen and the dose and type of progestin (Stadel, 1986).

Current OC use has been found to decrease glucose tolerance in most women, although this decrease appears to be small and unrelated to duration of use. This decrease is directly related to the estrogen content of the OCs (Stadel, 1986).

Neoplastic Diseases

The forms of neoplasia that are of greatest concern with the potential effects of OC use are breast cancer, cervical cancer, endometrial cancer, and ovarian cancer. There are two main reasons for the concern. First, these cancers are major causes of morbidity and mortality, particularly breast cancer in developed countries and cervical cancer in some developing countries.1 Second, the breast, the uterus, and the ovaries are endocrine-dependent organs, and a large body of research shows that hormonally related factors, such as age at menarche and age at first birth, affect the risk of developing neoplastic diseases. Thus, any factor that alters hormones requires careful scrutiny as a possible carcinogen or anticarcinogen for these organs. In addition, cervical cancer is caused by the human papiloma virus, and contraception may modify transmission.

Complex methodological problems make the study of possible relationships between OC use and these cancers difficult. Such problems include a possible long latency period and the difficulty of evaluating factors that might alter the effects of OCs, such as age at first pregnancy for breast cancer and the number of sexual partners for cervical cancer. In fact, some studies on breast and cervical cancer among OC users have found no effect on cancer risk and others have suggested increases. Since breast and cervical cancer are two of the most common cancers affecting women, the debate has taken on an urgency unlike that of other health risks. Family planning programs in the least developed countries generally lack the resources to monitor and respond adequately to these cancers. For example, Papanicolaou (Pap) screening, which is routine in developed countries, is not commonly performed in many developing countries. Although OC use clearly provides protection from the development of endometrial and ovarian cancer, its effect on other malignancies is generally unclear.

The relationship between OC use and breast cancer is controversial. The Cancer and Steroid Hormone study, the largest study to date, was conducted in eight regions of the United States from 1980 to 1982 (Centers for Disease Control and National Institute of Child Health and Human Development, 1986). This study found no increased risk of breast cancer among pill users, regardless of length of use or OC formulation. Even groups known to be at high risk, such as women with prior benign breast disease or a family history of breast cancer, nulliparous women, or those who had a late age at first full-term pregnancy, were unaffected by OC use. Controversy centers on long-term OC use, use at an early age, or use before the first full-term pregnancy. One study showed a higher rate of premenopausal breast cancer among women who used ''high-progestin'' OCs before age 25. Another study of women with long-term OC use before the birth of their first child found the risk of breast cancer as much as doubled in some cases (Pike et al., 1983; McPherson et al., 1983; Meirik et al., 1986). Although a subsequent analysis of the CASH data that replicated the analysis made by Pike and McPherson contradicted their findings, a recent analysis of the data from the CASH study suggests that very long-term OC use may decrease the age of onset of breast cancer for a small subset of nulliparous women without an appreciable impact for women in the aggregate (Stadel et al., 1988).

Breast cancer is uncommon among women in developing countries, and premenopausal breast cancer in these populations is rare. While there may be increased risk in small, select subgroups, in the aggregate there is probably no appreciable increase in risk. McPherson et al. (1983) have suggested that any possible risk of breast cancer associated with OC use at early ages may not become apparent until at least 20 years after that use, in which case researchers may not be able to detect such a relationship at the present time. The CASH study has found no increased risk of breast cancer within 10 to 15 years after use, even when use began at early ages (Schlesselman et al., 1988). The preponderance of epidemiologic studies suggest that OCs do not increase the risk of breast cancer, and any increase that may exist for certain subgroups of women is not great. Moreover, the inconsistencies among studies suggest that there may be methodological problems in the investigation of this complex disease.

According to available data, cancer of the cervix is the most frequent malignancy among women in developing countries (Lunt, 1984). No definite causal relationship has been established between OC use and cervical cancer. Some of the major epidemiologic studies conducted have found no increased risk and some have found significantly increased risk, at least in certain subgroups (Piper, 1985; Brinton et al., 1986; Ebeling et al., 1987; Irwin et al., 1988). A large study by the World Health Organization, which included many developing countries, found some indication of increased risk with prolonged OC use (World Health Organization, 1985a), but these studies have serious methodological problems, most notably a detection bias mused by increased Pap screening of OC users compared with nonusers and differences in sexual behavior among users and nonusers of OCs (Piper, 1985; Swan and Petitti, 1982). More recent studies have attempted to address these methodological problems, but the results remain conflicting. While OCs probably do not dramatically increase the overall risk of cervical dysplasia or cancer, long-term OC use or use by specific subgroups of women may increase the risk. Two large British cohort studies have shown a higher incidence of cervical neoplasia among oral contraceptive users (Vessey et al., 1983; Beral et al., 1988). The most important conclusion from the conflict over these results is the importance of annual Pap screening in the prevention of invasive cervical cancer.

OCs have been associated with malignant melanoma (skin cancer), but the association is rather weak and possibly confounded by differences in exposure to sunlight (Stadel, 1986). Some studies do suggest an increase within certain subgroups of women, particularly those with long-term use (Bain et al., 1982; Beral et al., 1984; Holly et al., 1983; Ramcharan et al., 1981). Due to the rarity of this malignancy in developing countries, however, the attributable risk is quite low and not very important for public health policy.

Recent case-control studies have found an increased risk of hepatocellular carcinoma (liver cancer) among OC users, largely confined to long-term users (Forman et al., 1986; Neuberger et al., 1986; Henderson et al., 1983). Unfortunately, these studies all had small sample sizes and methodological problems that may have biased the results. Since hepatocellular carcinoma is extremely rare in developed countries, the attributable risk is very low. The disease is a much more common problem in many developing countries, especially where there is a high prevalence of chronic hepatitis B. The interrelationships among OC use, hepatitis B, and liver cancer are not well understood. The World Health Organization is conducting a multicenter case-control study in three developing countries to address the question.

It is clear that OC use increases the risk of hepatocellular adenoma (HCA), a rare, benign tumor of the liver that can cause serious intra-abdominal hemorrhage and death. The case fatality rate is approximately 8 percent (Rooks et al., 1979). The risk attributable to OC use is very low, estimated to be about 2 cases of HCA per 100,000 users per year among women who have used OCs five years or more (Stadel, 1986).

Other Effects

It has been suggested that OC use might accelerate the appearance of gall bladder disease in susceptible women (Royal College of General Practitioners, 1982), although evidence for this hypothesis is limited. Early studies (Boston Collaborative Drug Surveillance Program, 1974; Royal College of General Practitioners, 1982) suggested that the risk of gall bladder disease might be increased in OC users. Recent studies and further analysis of information from British studies, which had first shown an increased risk of gallbladder disease in OC users (Layde et al., 1982; Wingrave and Kay, 1982), have failed to confirm this association.

There have been extensive studies of the effects on pregnancy outcome of hormonal contraceptive use prior to or during pregnancy. Although there are some reports of adverse effects, the majority of studies show no increased risks, and several comprehensive reviews of the literature have concluded that in utero exposure to synthetic steroids at the doses used for contraception does not result in significant deleterious effects on fetal growth or development (Wilson and Brent, 1981; World Health Organization, 1981; Simpson, 1985).

Even at low doses, the estrogen component of combination OCs has been shown to suppress milk volume in lactating mothers. Progestin-only contraceptives, including the minipill and long-acting methods discussed below, do not suppress milk production and can be used by breastfeeding women (World Health Organization, 1981). Although the synthetic hormones of the pill do pass on to the suckling infant, no adverse effects have been observed. Some reports have postulated an association between birth defects and the use of hormonal contraceptives prior to or during pregnancy. However, the majority of studies show no increased risks of deleterious effects on fetal growth or development (wilson and Brent, 1981; World Health Organization, 1981; Simpson, 1985).

Intrauterine Devices

The intrauterine device (IUD), which is inserted and remains in the uterus, prevents conception through several modes of action. IUDs may be medicated or nonmedicated; examples include the inert Lippes Loop, Copper-T (medicated with copper), and Progestasert (medicated with progesterone). The IUD is highly effective, having a failure rate of less than 6 percent in the first year of use. Many failures are due to undetected IUD expulsion (Trussell and Kost, 1987). It appears that new copper IUDs have a much lower failure rate of 1 to 2 percent. Rates of IUD use vary widely among countries. Partly because of its widespread use in China, the IUD is the most commonly used, reversible method of birth control in the world. IUDs are currently used by roughly 79 million women, nearly 58 million of whom live in China (United Nations, 1989).

Because IUDs appear to prevent both intrauterine and ectopic pregnancies, the overall risk of ectopic pregnancy is decreased by IUD use by about 60 percent, according to U.S. and multinational WHO studies (Ory and the Women's Health Study, 1981; Gray, 1984). However, 5 to 15 percent of IUD-associated pregnancies are ectopic, indicating that the IUD is more effective at preventing intrauterine pregnancies. Progesterone-releasing IUDs decrease menstrual blood loss and dysmenorrhea (Hatcher et al., 1988). No other noncontraceptive health benefits to IUD use have been identified.

Major health risks that have been associated with IUD use include pelvic inflammatory disease, tubal infertility, septic abortion, spontaneous abortion, and uterine perforation. The attributable mortality risk is extremely low in the United States, estimated at 1 to 2 deaths per 100,000 users and was mainly due to the now discontinued Dalkon Shield (Ory et al., 1983). Where access to medical facilities is poor and diagnosis and treatment of complications are delayed, mortality rates may be higher.

Unlike other modern methods of temporary contraception, the IUD increases the risk of pelvic inflammatory disease (Grimes, 1987). PID is usually, although not always, the result of sexually transmitted diseases (STDs). As a result, much of the risk of PID attributed to IUD use is mainly in women who are at increased risk for developing STDs. In the United States, women using IUD types other than the Dalkon Shield have been found to have about 1.5 to 2.0 times greater risk of PID than women using no method. Corresponding data in developing countries shows a relative risk of 2.3 (Gray and Campbell, 1984). The risk is largely concentrated in the first few months after IUD insertion, because insertion may introduce bacteria into the uterus (Lee et al., 1988).

The presence of PID has been clearly linked to subsequent tubal infertility. Two U.S. case-control studies found that the risk of tubal infertility among nulliparous women who ever used an IUD was double that of nonusers (Daling et al., 1985; Cramer et al., 1985). Apparently, this increased risk of tubal infertility is related to the presence of PID, even if PID is never recognized clinically. However, women who reported having only one sexual partner had no increased risk of tubal infertility associated with IUD use (Cramer et al., 1985). Therefore, in populations in which STDs are a major problem, it may be less advisable to promote IUD use. In countries such as China, however, where STDs are uncommon, the IUD is a safe and acceptable method.

If a pregnancy does occur with an IUD in place, a spontaneous abortion is likely, occurring in 50 percent of cases in which the IUD is left in place and 25 percent of cases in which it is removed (Hatcher et al., 1988). When the IUD is left in place, septic abortion in the second trimester may result and can possibly be fatal to the IUD user.

Perforation of the uterus may occur during IUD insertion but this is relatively rare, probably occurring in less than 1 percent of insertions, and usually is not serious (Hatcher et al., 1988). The risk of perforation is substantially increased among breastfeeding women and women between weeks 1 and 8 after delivery (but less during the first 4 or 5 days postpartum), evidently due to softer uterine musculature (Heartwell and Schlesselman, 1983). In general, it is recommended that the IUD be removed when perforation occurs.

Barrier Methods

Because they may prevent transmission of sexually transmitted diseases, including the human immunodeficiency virus (HIV), a great deal of attention is being focused on spermicides and barrier methods of contraception, principally condoms, diaphragms, and sponges. The United Nations estimates that 48 million women or their partners use these methods, but this number may be growing rapidly (United Nations, 1989). The effectiveness of these methods is highly dependent on user motivation and compliance. As a result, average failure rates tend to be higher than for any other modern method of contraception.

Condoms are a very safe method of birth control, but their effectiveness as a contraceptive and as a disease prophylactic depends on consistent and proper use. Failure rates are estimated to be as high as 12 percent per year in practice (Tressell and Kost, 1987). A number of in vitro studies have demonstrated that latex condoms are effective barriers to herpes simplex virus type 2, chlamydia trachomarls, cytomegalovirus, and HIV. Condoms evidently reduce the transmission of organisms present in the semen, such as Neisseria gonorrhoeae, hepatitis B virus, and Trichomonas vaginalis (Conant et al., 1984; Judson et al., in press; Katznelson et al., 1984; Conant et al., 1986; Stone et al., 1986).

Data regarding in vivo condom use and STDs is limited. Several studies have found a lower frequency of gonorrhea and HIV infection among condom users and/or their partners (Barlow, 1977; Hart, 1974; Hooper et al., 1978; Fischl et al., 1987; Centers for Disease Control, 1987). However, these studies are confounded by the fact that condom users are likely to differ from nonusers in many important characteristics (Feldblum and Fortney, 1988). Still, while the evidence is inconclusive, available data suggest that condoms may be quite effective STD prophylactics (Horsburgh et al., 1987). Their failure to protect is explained more probably by misuse than by product failure (Centers for Disease Control, 1988).

Spermicides are chemical agents that inactivate sperm in the vagina before they can move into the upper genital tract. The contraceptive sponge with spermicides may provide some protection against STDs, although, as with other barrier methods, the effectiveness of this contraceptive is highly dependent on user compliance. Failure rates in the first year of use may be as high as 18 percent among nulliparous women and close to 30 percent among parous women (Trussell and Kost, 1987). Laboratory and clinical evidence suggests that their virucidal effects may inhibit the growth of Neisseria gonorrhoeae (Cowan and Cree, 1973; Singh et al., 1972), herpes simplex virus type 2 (Singh et al., 1976), and HIV (Hicks et al., 1985). Although evidence is sparse, there is some indication that spermicides also protect against cervical cancer, which has been associated with the human papilloma virus (Spring and Gruber, 1985).

The sponge also has attendant health risks. Sponge users may be at increased risk of vaginal candidiasis, because normal bacterial growth is suppressed by certain types of spermicide, which leads to the overgrowth of candida (Rosenberg et al., 1987). There is also an association between the sponge and toxic shock syndrome (TSS), which in severe cases can lead to shock, coma, or death. Sponge users are apparently at 10.5 times greater risk of TSS than women using no barrier method (Schwartz et al., 1989). However, the attributable risk is low, since TSS is an extremely rare disease.

The diaphragm (with spermicide), like the condom, if used correctly and consistently, can be an effective contraceptive. Because of inadequate motivation, improper fitting, or inconsistent use, the average failure rate is roughly 18 percent per year (Trussell and Kost, 1987). The diaphragm appears to reduce the risk of gonorrhea, PID, and tubal infertility (Jick et al., 1982; Kelaghan et al., 1982; Cramer et al., 1987). Several studies have shown cervical dysplasia and cervical neoplasia to be less common among users (Wright et al., 1978; Harris et al., 1980; Celentano et al., 1987). Since diaphragms and sponges are almost always used with spermicides, it is difficult to separate the specific effects of each.

As with the sponge, the risk of TSS is significantly increased for diaphragm users (Schwartz et al., 1989). Still, the attributable risk is only about 0.2 percent annually. A less serious, but more frequent, complication associated with diaphragm use is urinary tract infections, occurring 2 to 3 times more often among users than nonusers (Foxman and Frerichs, 1985; Fihn et al., 1985; Vessey et al., 1987).

Long-acting Contraceptives

Several long-acting contraceptive methods have been developed, consisting mainly of injectables and implants. Usage is still relatively low, with just over 6 million women estimated to be using injectables (United Nations, 1989). These methods are highly effective and convenient to use and give protection from pregnancy from one month to five years. All contain a progestin, which may lead to a disturbance of the menstrual cycle.


Two injectable progestins, Depo-Provera (DMPA) and Noristerat (NET), have been approved in over 90 countries worldwide.2 Estimated failure rates in the first year of use are between 0.3 and 0.4 percent, depending on the kind of progestin used (Trussell and Kost, 1987). Injections are usually given every 8 to 12 weeks. Injectables prevent pregnancy by inhibiting ovulation, thickening cervical mucous, and altering the endometrial lining, which inhibits implantation (Liskin and Quillin, 1982).

The relationship between the risk of cancer and the use of injectables, particularly DMPA, remains controversial. The largest epidemiologic study yet published is an ongoing case-control study conducted by the World Health Organization. This study has found no increased risk of breast and endometrial cancer, and an analysis of invasive cervical cancer was deemed inconclusive. Final results concerning breast and cervical cancer are expected in the near future from this study and from a study in New Zealand. These and other studies have been hindered by small sample sizes and short durations of exposure. Animal data suggest that DMPA may increase the risk of breast and endometrial cancer (World Health Organization, 1986a).

Reported metabolic effects of the use of injectables include changes in blood pressure and insulin, cholesterol, and triglyceride levels (Liskin et al., 1987; WHO, 1986b). Various studies of DMPA and NET users have found both increases and decreases in total cholesterol and HDL-cholesterol. The findings are thus inconsistent and none has shown clear clinical significance (Liskin et al., 1987). No studies have been published on the possible associations between DMPA or NET use and the risk of cardiovascular disease. Unlike OCs, injectables appear to have little effect on the coagulation and fibrinolytic systems that affect blood clotting.

Amenorrhea or irregular, unpredictable bleeding episodes are the most commonly reported problems with injectables and the primary reason for terminating use (World Health Organization, 1978; Swenson et al., 1980; World Health Organization, 1987b). One-half to two-thirds of users have no regular menstrual cycles in the first year of use (Liskin et al., 1987). After one year of use, as many as 50 percent of users will be amenorrheac. The occurrence of heavy bleeding is rare, occurring in 0.5 percent of users. Conversely, since bleeding is often lighter than normal, increased hemoglobin levels have been reported (World Health Organization, 1986b).

Injectables appear to have no permanent effect on fertility, although ovulation may be inhibited for four to nine months or more after the last injection (Liskin et al., 1987; Pardthaisong et al., 1980; Affandi et al., 1987). Injectables may protect against PID by causing changes in the cervical mucus (Gray, 1985).

Injectable progestins may protect against endometrial and ovarian cancers. A WHO case-control study found a reduced risk of endometrial cancer in DMPA users, but the sample was quite small and results are inconclusive (World Health Organization, 1986a). There are even fewer data regarding ovarian cancer. However, since injectables prevent ovulation, as do OCs, it is hypothesized that injectables will also decrease the incidence of ovarian cancer; preliminary results from the WHO study support this possibility.


The Norplant subdermal implant system is another highly effective progestational contraceptive. One-inch-long plastic rods are surgically implanted under the skin of the upper arm and are left in place for several years. The progestin levonorgestrel is slowly released and remains effective for three to five years. The implants have a cumulative five-year net pregnancy rate of less than 2 percent in most studies (Segal, 1988).

Like injectables, the most common side effect of implants is disturbance of the menstrual cycle. Episodes of abnormal bleeding diminish with duration of use but, unlike injectables, the implants can be removed if there are extreme complications. Norplant users are generally protected from ectopic pregnancy since ovulation is suppressed. Transient ovarian cysts occur in a small percentage of women using Norplant, although the cysts eventually regress (Salah et al., 1987; Diaz et al., 1987). Permanent infertility appears not to be a problem (Sivin et al., 1983; Diaz et al., 1987; Affandi et al., 1987). Several studies have shown that fecundity quickly returns after the implants are removed. No changes have been found in liver function, carbohydrate metabolism, blood coagulation, blood pressure, or body weight (Liskin et al., 1987). Of particular importance in the use of implants is the very low blood level of progestogen, which is much lower than with other steroid contraceptives.


Sterilization is the most widely used contraceptive method in the world. More than 108 million women and 41 million men have undergone sterilization procedures (United Nations, 1989). Sterilization is safe and highly effective; most of the health risks are associated with poor anesthetic or surgical technique.

Pregnancy identified after tubal sterilization may result from conception before sterilization or from unsuccessful sterilization. Failure rates, which vary by method of tubal occlusion, surgical expertise, and patient characteristics, are overall estimated to be between 2 and 4 per 1,000 in the first year of use (Trussell and Kost, 1987). When female sterilization failure occurs, ectopic gestation is more likely than intrauterine gestation, but the absolute likelihood of ectopic pregnancy is actually lower than that associated with use of no method or even IUDs.

Tubal sterilization is usually performed via an abdominal incision. A vaginal approach offers the advantage of producing no visible scar, but such a procedure increases the risk of pelvic infection and thus is used less frequently. The fallopian tubes may be blocked by tying (with or without removal), by coagulation, using unipolar or bipolar current, or by mechanical occlusion with silastic bands or clips. All procedures except conventional laparotomy can be safely performed using local anesthesia, thus avoiding the hazards inherent in the use of general anesthesia.

Studies suggest that tubal sterilization is a remarkably safe surgical procedure. The case-fatality rate has been reported as low as 4 per 100,000 procedures in U.S. hospitals (Peterson et al., 1982) but as high as 19 per 100,000 procedures in Bangladesh (Grimes et al., 1982). Most deaths are caused by complications related to use of anesthesia, even when general anesthesia is not used. Deaths have occurred from hemorrhage and thermal injury as well (Peterson et al., 1983). Reports regarding nonfatal complications vary. In general, such studies indicate that major morbidity is uncommon and varies by surgical approach, anesthetic technique, and tubal occlusion method.

No important long-term negative physiological effects of tubal sterilization have been reported in the literature. Much concern has focused on menstrual abnormalities, the so-called post-tubal syndrome, which was identified by a number of studies prior to 1980. These early studies had methodological problems; better designed, more recent studies have found no evidence of a post-tubal syndrome. When menstrual changes did occur, about as many women experienced improvement in symptoms as experienced a deleterious change (Bhiwandiwala et al., 1983). Many of the observed changes were attributable to cessation of OC or IUD use. Studies have found conflicting results on the question of an increased risk of hysterectomy following sterilization. It has been postulated that any observed correlation may be explained by the fact that, once a woman has been sterilized, either she or her physician may more quickly resort to surgical management of any gynecologic problem.

Male sterilization, or vasectomy, is the cutting or occluding of the vas deferens to prevent sperm transport. Although safe, simple, and highly effective, vasectomy is not popular in most countries. Most users reside in the United States, the United Kingdom, China, and India. Access to services and motivational factors have been cited as reasons for the generally low level of use. Few studies report any pregnancies after vasectomies and, of those that do, most have reported failure rates below 1 percent, with most failures attributable to unprotected intercourse shortly after vasectomy or spontaneous rejoining of the vas (Trussell and Kost, 1987).

The procedure consists of isolating the vas deferens, then occluding it by ligation (the most common approach), coagulation, or clip application. Local anesthesia without premedication is most often used. The risk of death attributable to vasectomy is extremely low. The Association for Voluntary Surgical Contraception has recorded only two vasectomy-related deaths associated with over 160,000 procedures in programs it supported (Ross et al., 1985).

Research has consistently failed to identify long-term health risks attributable to vasectomy. In contrast to animal findings, at least six epidemiologic studies in humans, including a large study in China, have indicated that the risk of myocardial infarction is not increased in the 10 years following vasectomy (Goldacre et al., 1978, 1979; Walker et al., 1981; Petitti et al., 1982; Massey et al., 1984; Perrin et al., 1984). Possible relationships between vasectomy and prostatic disease have been examined (Sidney, 1987; Ross et al., 1985). With the exception of one recent study (Honda et al., 1988), no association between vasectomy and prostatic disease has been found, and a plausible alternative explanation for the results was made by the authors of that study. Still controversial is the relationship between vasectomy and subsequent genito-urinary tract diseases, such as kidney stones (urolithiasis). One recent report has found a 70 percent increased risk of kidney stones among men who had undergone vasectomy (Kronmal et al., 1988). Other studies have found no relationship, but the possibility warrants further evaluation.

Traditional Methods

Traditional methods of contraception include periodic abstinence or rhythm, withdrawal, douche, or complete abstinence. Unsupplemented breastfeeding on demand postpones the onset of ovulation and may thus also be considered a form of contraception. It is difficult to measure the use of these methods, since they may be practiced without being called contraception. The United Nations reports that over 77 million women rely on one of these methods (United Nations, 1989). Periodic abstinence and withdrawal are much less effective than most of the modern methods already discussed, with failure rates around 15–20 percent in the first year of use (Trussell and Kost, 1987).

That breastfeeding can provide considerable protection against pregnancy is well documented (see Hatcher et al., 1988, for a review). Pregnancy rates in populations depend on breastfeeding prevalence and practices. Hatcher et al. (1988:117) conclude: "Breastfeeding can be an effective method of fertility control for a population, but breastfeeding effectiveness is unpredictable for the individual woman, particularly with western patterns of breastfeeding and supplementation."

Periodic abstinence or rhythm is based on awareness of variation in the woman's fecundity over the menstrual cycle using the calendar, basal body temperature, and/or the character of cervical mucus. Rhythm has no health risks or noncontraceptive benefits for a woman. There may be an increased risk that an old, rather than a fresh, egg will be fertilized, possibly leading to a higher risk of fetal wastage or birth defects. Animal studies have shown that aged gametes may be associated with increased early abortions and increased birth defects, and equivocal, limited data suggest an increase in spontaneous abortions. But studies on humans have been unconvincing, either to support or discount the possible effects (Hatcher et al., 1988; Kambic et al., 1988).

It is uncertain how frequently coitus interruptus (withdrawal) is used worldwide. There are no known biological side effects. Douching and other means of cleaning out the vagina after intercourse have been used to prevent conception ever since it was understood that ejaculation into the vagina caused pregnancy. Not only is the method highly ineffective for contraception, but it also greatly increases the risk of vaginal infection. Douching has been associated with an increased risk of PID, although the relationship may not be causal. A case-control study found that women who douched frequently had 4.4 times the risk of ectopic pregnancy (Chow et al., 1985).

Dimensions of New Research

Clearly, no modern method of contraception is completely free of health consequences, whether adverse or beneficial or both. Oral contraceptives, which increase the risk of a variety of cardiovascular problems, also protect against PID, ectopic pregnancy, and two cancers of the reproductive system. Barrier methods of contraception, which may reduce the transmission of sexually transmitted diseases, are also associated with an increased risk of pregnancy. Sterilization, while generally an extremely safe procedure, can be dangerous if improperly performed.

Priorities for further safety studies should be determined by the incidence of serious disease in a country. For example, where liver cancer is already problematic, contraceptive research should focus on the impact of contraceptive methods on this disease. At the same time, research must respond to case reports that are particularly unusual. A finding that 9 out of 10 cases of a rare disease were all using the same method of contraception would indicate the need for further study. These decisions are far from simple. The pervasive concerns and worries of a population or a government cannot be ignored, even when empirical data negate their importance. Still, we are left with a number of questions. At what level of incidence does an epidemiologic study become necessary? What level of risk is acceptable for the continued marketing of a specific method? How important are discomforting but nonfatal side effects?

Ongoing research to test new variations of existing contraceptive methods as well as the development of new methods must be continued. The long-term effects of most methods can be determined only after many years of use, a situation that mandates repeated and protracted study. Cohort studies are needed to evaluate the overall pattern of mortality and morbidity related to contraceptive use, and case-control studies are needed to evaluate the contraceptive-related risks for specific diseases. Moreover, ongoing surveillance of the use of all hormonal contraceptives in both developed and developing countries is crucial.

By way of conclusion, it is appropriate to put the various risks of contraceptive use into perspective. Due to the uncertainty associated with the various health risks for each method of contraception and the methodological complexities inherent in such analyses, no definitive overall risk can be calculated by method (see Ory et al., 1983, for estimates of risks). However, in developing countries, where maternal mortality is high, and diseases associated with contraception such as myocardial infarction are uncommon, there is no questions that contraception is safer than pregnancy and childbirth.

Any decision regarding contraceptive use must be based not only on the noncontraceptive risks and benefits, but also on the efficacy of the method. Each individual's life situation and the level of risk particular to his or her characteristics must be considered as well. Finally, the life consequences of childbearing for the mother and child must also be considered. We now turn to the health consequences of controlled fertility for children, again with special consideration of high-risk categories.



Approximately 6 percent of British women and 9 percent of American women develop cancer of the breast during their lives (Schlesselman et al., 1988).


Neither DMPA nor NET has been approved for use in the United States. See Richard and Lasagne (1987) for a review of the debate on approval.

Copyright © National Academy of Sciences.
Bookshelf ID: NBK235069


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...