The turn of this century marked the fourth decade of oral contraceptive use. Although there has been a wealth of literature that has shown “the pill” to be safe and effective, there continues to be some controversy about some aspects of safety. In addition, because of persistence of nuisance side effects such as abnormal bleeding and nausea, new agents have been introduced during the past several years. This article reviews the current status of the oral contraceptive including new developments, efficacy, major sequelae, selected side effects, and noncontraceptive benefits.
The major new developments in the past two decades include introduction of new progestins, ie, desogestrel and norgestimate, and reduction in the dosage of ethinyl estradiol from 35 μg to 20 μg per pill. More recent innovations have focused on shortening the hormone-free interval. For example, one preparation provides 10 μg of estradiol for 5 days after completion of the 21 days of estrogen–progestin combination dosage. Another preparation awaiting Food and Drug Administration approval provides a continuous combination dosage for 84 days. The major impetus for these changes has been to improve safety and to reduce side effects. Although there is some evidence that the newer progestins may have less androgenicity, desogestel has been embroiled in a controversy about whether its use increases the risk for thromboembolism. There are little data to indicate whether reduction of the estrogen dose to 20 μg is associated with decreased risk of serious sequelae, primarily because of lack of data. Relative to alterations in side effects, as will be presented later, the data are mixed. Finally, shortening the pill-free interval, although helpful for some users, does not appear to be the solution that markedly reduces nuisance side effects. 1
Combined oral contraceptives suppress ovulation by diminishing the frequency of gonadotropin-releasing hormone pulses and halting the luteinizing hormone surge. They also alter the consistency of cervical mucous, affect the endometrial lining, and alter tubal transport. When administered correctly and consistently, they confer a greater than 99% method effectiveness in preventing pregnancy. 2 Unfortunately, problems with compliance, frequently secondary to side effects such as abnormal bleeding, have led to a significantly reduced use-effectiveness. Overall, the use-effectiveness, because of these factors, ranges between 94–97%. 2 Although overall efficacy may not be affected by recent dosage alterations, some data examining ovarian follicle changes suggest that there may be less margin for error with low-dose preparations such that missing pills for 2 or more days may be more likely to result in break-through ovulation. 3
By the 1970s, it had been established that oral contraceptives were associated with various types of cardiovascular risks. As time progressed, additional concerns were raised, particularly risk for breast cancer. Both of these major issues will be examined in this section because with the significant reduction in estrogen and progestin dosages and the introduction of new formulations, many of the original estimates may require modification.
The lowering of the estrogen dose in oral contraceptives is associated with a reduction in risk for venous thromboembolism. For example, a pharmacy-based study demonstrated a dose–response effect with rates of venous thromboembolism ranging from 10 events per 10,000 woman-years for greater than 50-μg estrogen pills to 4.2 events per 10,000 woman-years for less than 50-μg estrogen pills. 4 Little significant new information evolved about oral contraceptive components or dosages until studies published in late 1995 and early 1996 suggested that two of the newer progestins, gestodene and desogestrel, had a greater risk for venous thromboembolism than older progestins such as levonorgestrel. 5,6 As shown in Table 1, use of most oral contraceptives approximately triples one’s risk of venous thromboembolism, although some studies with formulations containing gestodene or desogestrel have shown approximately a seven-fold increased risk compared with nonusers of oral contraceptives. These findings have been discussed in several publications, with concerns being raised about possible sources of bias and confounding and lack of biologic plausibility. 7 It is important to recognize, however, that even with the worst case scenario, the attributable risk annually is approximately 18 additional cases of the disorder per 100,000 users of gestodene or desogestrel containing oral contraceptives compared with nonusers of oral contraceptives. In addition, the mortality caused by venous thromboembolism among oral contraceptive users is quite low. Age does affect mortality; for women age 35 to 44 years, the mortality rates double. In a recent World Health Organization study, obesity and age also were found to be risk factors for venous thromboembolism, although the former is not a consistent finding in all studies. 6 Finally, there is no evidence that smoking or the presence of varicose veins among users of oral contraceptives appreciably affects their relative risk for venous thromboembolism. 6
The identification of the factor V Leiden mutation in 1993 introduced another consideration about the relation between oral contraceptive use and venous thromboembolism. A single amino acid substitution results in a variant of factor V that is resistant to activated protein-C, which leads to enhancement of clot formation and an increased risk of venous thromboembolism. This mutation, which is known as either factor V Leiden or activated protein-C resistance, has a prevalence in the general population of approximately 5% in US white women, 2.2% in Hispanic women, and 1.2% in black women, making it the most commonly occurring natural anticoagulant deficiency. 8,9 A recent study estimated the risk of venous thromboembolism among oral contraceptive users with factor V Leiden. 8 With the mutation at a prevalence of approximately 5%, the risk for reproductive-aged women who were not using oral contraceptives was 5.7 venous thromboembolism events per 10,000 woman-years. In contrast, among oral contraceptive users with the mutation, the rate increased to 28.5 events per 10,000 woman-years. However the absolute risk of venous thromboembolism among such women is quite low. For example, one investigator determined that screening one million potential users for all known coagulation factor deficiencies or mutations would identify approximately 50 women at risk, but also would result in approximately 62,000 women having false-positive results. 9
Stroke often is divided into two major categories: ischemic or thrombotic and hemorrhagic. Although earlier studies focused most of their attention on estrogen dose, more recent studies of low-dose oral contraceptives have better recognized the role of confounders and have used more rigorous approaches to both design and analysis, such that current risk estimates are probably more accurate.
The risk of thrombotic or ischemic stroke among current users of low-dose oral contraceptives appears to be relatively low. As shown in Table 1, among women age 20 to 24 years, the overall risk among oral contraceptive users is increased by approximately two and one-half-times compared with that of nonusers. 5 There is no evidence that type of progestin influences risk or mortality associated with ischemic stroke. 5,6 The risk of ischemic stroke does appear to be directly proportional to estrogen dose, based on studies completed in developed as opposed to less developed countries. 6 Age appears to be a risk factor relatively independent of oral contraceptive use, with the relative risk of ischemic stroke doubling as women reach age 40 to 44 years. It is unclear in some studies whether this increment actually represents an increased likelihood of the presence of an undetected risk factor such as hypertension. 6 Hypertension, cigarette smoking, and migraine headaches also interact with oral contraceptive use to substantially increase the risk of ischemic stroke. Of importance is the finding in at least three studies that women administered oral contraceptives containing less than 50 μg of estrogen who have their blood pressure checked regularly do not have a significantly increased risk for ischemic stroke. 6
As shown in Table 1, the risk of hemorrhagic stroke in young women is low and is not increased by use of oral contraceptives in the absence of risk factors. 5,6 Although age independently increases the risk in both users and nonusers, there appears to be no relation between the components of oral contraceptives, their dosages, or duration of use and the risk of hemorrhagic stroke. In addition to age, the major risk factors for hemorrhagic stroke are cigarette smoking and hypertension. Smoking increases the risk of hemorrhagic stroke approximately two-fold among non-oral contraceptive users and three-fold among oral contraceptive users. 6 In one study, history of hypertension increased the risk by approximately five-fold among women in Europe who were not oral contraceptive users compared with non-oral contraceptive using women who were normotensive; among women with hypertension in developing countries, the relative risk increased to 9.4. 6 With current oral contraceptive use, in the absence of hypertension, the risk of hemorrhagic stroke is not increased, but with a history of hypertension, the relative risk among users increases to 10.2 in Europe and to 14.2 in developing countries. 6 Overall, for women younger than 35-years-old who do not smoke and who are normotensive, the risk of hemorrhagic stroke is not affected by oral contraceptive use.
As shown in Table 1, myocardial infarction is a rare condition among reproductive-aged women. Age influences this risk in an exponential fashion, with the incidence of myocardial infarction increasing to 30 cases per 100,000 annually among women age 40 to 44 years. 5 The presence of other risk factors such as cigarette smoking, hypertension, and diabetes strongly influences the risk of myocardial infarction. The more recent studies of oral contraceptive use and myocardial infarction have shown some increase in risk with oral contraceptive use for those women with risk factors. 6 Thus, as shown in Table 1, although oral contraceptive users overall have some increased risk, it is estimated that 80% of the cases among oral contraceptive users are attributable to cigarette smoking, with the remainder occurring in users with other risk factors such as hypertension or diabetes. There is no increased risk for myocardial infarction associated with increasing duration of oral contraceptive use or with past use. 6 Nonsmoking oral contraceptive users who have regular blood pressure examinations and who do not have diabetes have no increased risk of myocardial infarction. There is no evidence that use of new low-dose oral contraceptives modifies risk for myocardial infraction, although one study has suggested that users of oral contraceptives containing gestodene or desogestrel have a reduced risk for myocardial infarction compared with users of preparations containing the progestin levonorgestrel;10 however, this has not been confirmed in other studies.
BREAST CANCER RISK
For several decades, there has been concern about the possible association between oral contraceptive use and breast cancer. Unfortunately, many epidemiologic studies reported conflicting results, with the discrepancies probably related to how well potential biases and confounding were addressed in the individual studies. To address the concerns with earlier publications, a meta-analysis of most of the better epidemiologic studies in the literature was published in 1996. 11 A total of 54 studies were included in the analysis, representing 53,297 women with breast cancer and 100,239 control patients. To be included, studies had to have had at least 100 women with breast cancer diagnosed, and the original data had to be available for reanalysis. In all, there were 11 cohort studies, 28 case-control studies with population-based controls, and 15 case-control studies with hospital-based control patients. In addition, approximately one-third of the cancers were diagnosed in women younger than 45-years-old, and approximately one-half had cancer diagnosed after 1985. Figure 1 provides the major findings from this meta-analysis. For current users of oral contraceptives, the relative risk of breast cancer compared with never-users was 1.24. As shown, this small increase in risk persisted for approximately 10 years, but the risk essentially disappeared after that time period. In addition, there was no overall effect of oral contraceptive use by dosage, specific formulation, duration of use, age at first use, age at time of cancer diagnosis, or by family history of breast cancer. The comparison of ever-users of oral contraceptives with never-users revealed that the relative risk for tumors that had spread as opposed to localized disease was 0.88. Thus, although oral contraceptive users have a modest increase in risk of breast cancer, the disease tends to be localized. Also, the pattern of disappearance of risk after 10 years coupled with the tendency toward localized disease suggests that the overall effect may represent detection bias or perhaps a promotional effect. For example, if oral contraceptive users undergo more frequent breast examinations because they are required to see a clinician to get their prescriptions refilled compared with users of other methods, the findings may be because of users having a better chance of earlier detection. A promotional effect would occur if the hormones in oral contraceptives, especially estrogen, cause growth of cells that already have been transformed into a neoplasia, leading to earlier detection. Although most of the available data indicate little concern for most oral contraceptive users relative to breast cancer, an historic cohort study suggests that oral contraceptive use before 1975 among women with a first-degree relative with breast cancer increases one’s risk by approximately three-fold. 12
Compliance and Selected Side Effects
A number of authors have examined oral contraceptive compliance in various patient populations during the past two decades. In a study of 1,657 oral contraceptive users completed in the United States, women were surveyed at 2 and 6 months after initiating or switching to the use of a new oral contraceptive preparation. 13 Although the results are somewhat hampered by the low rate of return of the questionnaires, they still provide some insight about issues that affect oral contraceptive compliance. Failure to administer one’s pill at the same time every day, having a totally inconsistent pill-administering time, and not fully understanding the package insert information were associated with missing two or more pills per cycle. Administering the pill right before bedtime was associated with the greatest risk of missing pills during a cycle. Spotting and heavy or prolonged menstrual periods also were associated with missing pills during a cycle. Overall, in this study, 47% of oral contraceptive users reported missing one or more pills during a cycle, and 22% missed two or more. Side effects were the most important reason women stopped using oral contraceptives in this study. Thirty-seven percent of women stopped administering their preparation for this reason; this percentage increased to 46% when physicians’ recommendation to discontinue, presumably for some side effect, is included in the total. Bleeding irregularity was the side effect most frequently cited as being associated with oral contraceptive discontinuation. Other side effects cited included nausea, weight gain, mood changes, and breast tenderness. Somewhat disconcerting was the finding that among study participants who elected to discontinue oral contraceptives but did not desire to become pregnant, 19% adopted no method, and approximately 60% selected a less effective method, eg, barrier method, spermicides, or withdrawal. Although the level of response to the study questionnaire needs to be taken into account, approximately 32% of women who were “new start” oral contraceptive users and approximately 17% of women switching to another oral contraceptive had discontinued by the end of the 6-month study. Further, among all women who elected to stop oral contraceptive use, 42% discontinued without first contacting their heath care provider.
Examination of recent data evaluating the frequency of side effects among users of newer formulations indicate that decreasing the estrogen dose improves some symptoms, but not all symptoms. For example, a randomized trial comparing a monophasic 20-μg ethinyl estradiol and norethindrone acetate preparation with a 35-μg ethinyl estradiol triphasic norgestimate preparation demonstrated an incidence of breakthrough bleeding during 6 months of 3.7–13.5% among those using the 20-μg oral contraceptive, compared with a range of 23.5–49.7% among those using the 35-μg oral contraceptive. 14 Another randomized clinical trial comparing a 20-μg ethinyl estradiol plus levonorgestrel pill, a 20-μg ethinyl estradiol with 10 μg added during days 22 through 26 plus desogestrel pill, and a 35-μg ethinyl estradiol plus tricyclic norgestimate preparation demonstrated substantial variation in cycle control, particularly for the 20-μg preparations, but with similar cycle control achieved for all preparations by 6 months of use. 15 This latter trial also evaluated nausea and breast tenderness. Nausea was noted in 20–40% of all users, but by 6 months, approximately 20% of those using 35-μg ethinyl estradiol versus approximately 10% of those using 20-μg experienced nausea. Similarly, breast tenderness was noted in a high percentage of users; by 6 months, approximately 58% of users of 35-μg ethinyl estradiol experienced the symptom, compared with approximately 38% among the users of the 20-μg pill. Thus, it appears that decreasing the dose of estrogen reduces the frequency of estrogen-related side effects but, for the most part, is less effective in controlling unwanted bleeding.
Noncontraceptive Health Benefits
There are a growing number of noncontraceptive health benefits associated with oral contraceptive use. This section reviews some of the major reported benefits, including emerging possibilities such as protection against osteoporosis and colorectal cancer.
PROTECTION AGAINST ECTOPIC PREGNANCY
Although not a noncontraceptive health benefit, protection against ectopic pregnancy still has important public health implications. Based on a number of epidemiologic studies, oral contraceptives convey an approximately 90% reduction in risk of ectopic pregnancy. 16 The likely mechanism is through suppression of ovulation, an effect that obviously prevents all types of pregnancy and that should be present with low-dose oral contraceptives as well.
REDUCED RISK OF OVARIAN CANCER
Although ovarian cancer is relatively uncommon, it has a high case fatality rate. At least 22 case-control studies and three cohort studies have examined the relation between oral contraceptive use and ovarian cancer. 17 All but two of these studies have shown a protective effect for oral contraceptives. There appears to be a 40–80% overall decrease in risk among users, with protection beginning approximately 1 year after initiating use, with a 10–12% decrease annually in risk for each year of use. In addition, protection persists between 15 to 20 years after oral contraceptive discontinuation. There are also data to suggest that oral contraceptive use reduces the risk of ovarian cancer among women who are at high-risk for the disease. A recent study indicated that this protection occurs with the use of the newer low-dose oral contraceptives as well. Preliminary data from one study also suggest that oral contraceptive use may protect women with the BRCA1 or BRCA2 mutation; however, this study is limited by its small sample sizes. 18 The mechanisms by which oral contraceptives may produce their protective effects include suppression of ovulation, thus resulting in a reduced frequency of “injury” to the ovarian capsule, and the suppression of gonadotropins. Although the “gonadotropin theory” is controversial, proponents cite data about how high levels are associated with ovarian cancer in animals and that gonadotropins have been shown to stimulate some human ovarian cancer cell lines.
REDUCED RISK OF ENDOMETRIAL CANCER
Endometrial cancer is somewhat more common than ovarian cancer, and the 5-year survival rates are much better. At least 11 case-control studies and one cohort study have demonstrated that use of oral contraceptives conveys protection against endometrial cancer. 19 Overall, the studies suggest up to a 50% reduction in risk beginning approximately 1 year after initiation of use, that protection increases with duration of use, and that protection persists up to 20 years after oral contraceptive use is discontinued. Protection has been demonstrated for adenocarcinoma, adenosquamous tumors, and adenoacanthomas. The protective effect varies in studies for women with potential risk factors such as obesity and nulliparity. The mechanism of action is likely reduction in the mitotic activity of endometrial cells because of progestational effects. Because current low-dose preparations use highly potent progestins that have the same endometrial effects as older preparations, this protective effect is likely to occur with their use as well.
PROTECTION AGAINS PELVIC INFLAMMATORY DISEASE
A number of epidemiologic studies demonstrate that use of oral contraceptives will reduce the risk of salpingitis by 50–80% compared with the risk to women not using contraception or who use a barrier method. 16 Interestingly, there is no protective effect against the acquisition of lower genital tract sexually transmitted diseases, including Chlamydia trachomatis and Neisseria gonorrhoeae infections of the uterine cervix. The purported mechanisms for protection include progestin-induced thickening and increased viscosity of cervical mucus, so that ascent of bacteria is substantially inhibited. Additional theories include reduced menstrual flow, resulting in less retrograde menstrual flow to the fallopian tubes, an environment less likely to promote growth of bacterial organisms, and possible changes in uterine contractility that retard the ascent of organisms. Because current oral contraceptives exert these effects, it is likely they also will offer protection.
DECREASED INCIDENCE OF BENIGN BREAST DISEASE
The results of several studies reveal a 30–50% decrease in the incidence of benign fibrocystic conditions with oral contraceptive use. 20 Occurrence of fibroadenomas, specifically, is decreased among women younger than age 45. These effects are seen mainly in current and recent long-term users of oral contraceptives. The likely mechanism is through suppression of ovulation and therefore inhibition of the breast cell proliferation that normally occurs in the first half of a menstrual cycle. This effect should continue to be seen with the estrogen doses of 30 to 35 μg, although the effect may not be as significant with the newer 20-μg pills, in which more frequent ovulatory periods occur. Higher doses or increased potency of progestins may improve the risk reduction. Oral contraceptive use also has been demonstrated to decrease the incidence of galactorrhea.
RELIEF FROM MENSTRUAL DISORDERS
Few studies have clearly documented the beneficial effects of current low-dose oral contraceptives on menstrual flow and dysmenorrhea. Recently, a multicenter, randomized, clinical trial demonstrated that women with dysfunctional uterine bleeding using a 35-μg ethinyl estradiol tricyclic oral contraceptive experienced an 81–87% improvement in bleeding during 3 months compared with a 36–45% improvement seen in the placebo group. 21 As a mechanism of action, it is hypothesized that the combination of estrogen and progestin in oral contraceptives essentially stabilizes the endometrium with time. Because the progestins in oral contraceptives are 19-nortestosterone derivatives, they are highly potent in their effect on endometrial activity. Although there is limited clinical trial information about oral contraceptives as treatment for primary dysmenorrhea, a number of anecdotal reports document their effectiveness. 22 Oral contraceptives likely decrease the pain by reducing the prostaglandin content of menstrual fluid, leading to less local endometrial vasoconstriction and ischemia, and by reducing uterine contractile activity.
REDUCED RISK OF UTERINE LEIOMYOMAS
Leiomyomas are the most common pelvic neoplasm and are found in up to 40% of women. One group of researchers noted a reduced risk of uterine fibroids in oral contraceptive users, whereas a large case-control study from Italy found no relation between oral contraceptive use and risk of these tumors. 23 Reduced estrogen states do appear to decrease the incidence of leiomyomas, such that the beneficial effect may be noted only in select populations. Oral contraceptive use also has been shown to reduce menstrual blood loss in women with fibroid uteri.
PROTECTION AGAINST OVARIAN CYSTS
The frequency of ovarian cysts among oral contraceptive users is highly dependent on the steroid dosages in oral contraceptive formulations. Although the data are somewhat inconsistent, there is a clear suggestion that higher-dose contraceptives, eg, 50 μg or greater of ethinyl estradiol, protect against cyst formulation. In contrast, current lower-dose formulations appear to have no effect on ovarian cyst formation. 3 It is important to recognize that many of the recent studies evaluating this issue use ultrasound definitions of cysts as opposed to clinical definitions in which some type of therapy may be indicated. Further, many of these “cysts” identified in such studies regress with time. Thus, it appears that current low-dose oral contraceptives reduce the frequency of ovulation but may still allow early recruitment of follicles that do not progress to the stage of ovulation.
REDUCTION OF ACNE
Two randomized, placebo-controlled, clinical trials conducted during a 6-month period demonstrated that approximately one-half of women demonstrated improvement of their acne while administering a tricyclic product containing the progestin norgestimate. 24,25 Interestingly, approximately 30% of women receiving a placebo preparation also showed improvement. Overall, the response rate to the oral contraceptive in these two studies is similar to that achieved by women using topical agents such as tretinoin or benzoyl peroxide or systemic antibiotics. There are several purported mechanisms of action by which oral contraceptives reduce acne. Oral contraceptives increase the levels of sex hormone-binding globulin, which in turn binds testosterone and suppress gonadotropins; both of these effects reduce levels of ovarian androgens. In addition, it appears that certain progestins also have the ability to inhibit 5-α-reductase, which is needed to convert testosterone to dihydrotestosterone at the level of the pilosebaceous unit.
IMPROVEMENT IN BONE MINERAL DENSITY
Bone mineral density in women peaks between the ages of 20 and 25 years, stays constant for approximately 10 years, and then progressively decreases in the later reproductive years. Nineteen studies have revealed a positive effect on bone mineral density, and 13 studies indicated no effect on bone mineral density among oral contraceptive users. 26 No studies have demonstrated a negative effect on bone mineral density. Most importantly, a population-based control study revealed a 25% reduction in hip fracture risk. 27 It appears that oral contraceptives are most effective during times of low estrogen and with increased duration of use. Estrogens act on bone by increasing calcium absorption, decreasing calcium loss, and directly inhibiting bone reabsorption through inhibition of osteoclasts. Because the estrogen component of oral contraceptives is responsible for these effects, it is unlikely that this benefit would be seen with the progesterone-only “mini-pill.”
REDUCED RISK OF COLORECTAL CANCER
There is growing epidemiologic evidence that oral contraceptives may protect women from having colorectal cancer develop in later life. There have been several case-control studies and at least one cohort study that have demonstrated a protective effect against colorectal cancer development in ever-users of oral contraceptives approaching 40–50%. 28 In some studies, this protective effect appears to be directly proportional to duration of use, although this has not been a consistent finding in all studies. Further, there are other studies that show no effect of oral contraceptive use on the development of this tumor. The mechanism of action for this possible protective effect is essentially unknown, although some hypothesize that oral contraceptives produce this protection by reducing bile acid production and concentration, by altering colonic flora, by having a direct effect on colonic mucosa, or by having some type of tumor-suppressor function. Because most of these findings are related to use of higher-dose oral contraceptives, it is unclear whether users of today’s preparations will experience any protective effect.
REDUCED RISK OF RHEUMATOID ARTHRITIS
Two studies published before 1982 suggested a 40–50% reduction in the risk of premenopausal rheumatoid arthritis in oral contraceptive users. More recently, a meta-analysis evaluating a dozen studies found no conclusive evidence of a protective effect of oral contraceptive use, whereas one other report purported potential prevention of polyarthritis with current oral contraceptive use. 29 The cause of the disease is unclear, although most classify it among the autoimmune disorders. How sex hormones in therapeutic doses interact with a disorder with this type of cause is unknown.
During the past four decades, oral contraceptives have remained a safe and effective method of birth control. Reductions in the estrogen and progestin dosages have significantly decreased the incidence of cardiovascular complications. The association between oral contraceptives and breast cancer appears to be primarily because of detection bias or possibly a promotional effect. Despite the changes in formulation, the problems related to side effects have not been totally solved. Because compliance and successful use is strongly affected by side effects, improvement in this area is probably the biggest challenge faced by developers of oral contraceptives. It is also clear that there are a growing number of significant noncontraceptive benefits that accrue in oral contraceptive users. Unfortunately, many women do not know about these benefits. Thus, one of the issues that providers need to continue to address is how to provide better information about oral contraceptives and contraception in general to patients.
1. Sulak PJ, Cressman BE, Waldrop E, et al. Extending the duration of active oral contraceptive pills to manage hormone withdrawal symptoms. Obstet Gynecol. 1997; 89: 179–183.
2. Trussell J, Kost K. Contraceptive failure in the United States: A critical review of the literature. Stud Fam Plan. 1987; 18: 237–283.
3. Young RL, Snabes MC, Frank ML, et al. A randomized, double-blind, placebo-controlled comparison of the impact of low-dose and triphasic oral contraceptives on follicular development. Am J Obstet Gynecol. 1992; 167: 678–682.
4. Gerstman BB, Piper JM, Tomita DK, et al. Oral contraceptive estrogen dose and the risk of deep venous thromboembolic disease [see comments]. Am J Epidemiol. 1991; 133: 32–37.
5. International Federation of Fertility Societies. Consensus conference on combination oral contraceptives and cardiovascular disease. Fertil Steril. 1999; 71: 1S–6S.
6. WHO Working Group. Cardiovascular Disease and Steroid Hormone Contraception. Geneva: World Health Organization, 1998.
7. Lewis MA. The epidemiology of oral contraceptive use: A critical review of the studies on oral contraceptives and the health of young women. Am J Obstet Gynecol. 1998; 179: 1086–1097.
8. Vandenbroucke JP, Koster T, Briet E, et al. Increased risk of venous thrombosis in oral-contraceptive users who are carriers of factor V Leiden mutation [see comments]. Lancet. 1994; 344: 1453–1457.
9. Winkler UH. Blood coagulation and oral contraceptives. A critical review. Contraception. 1998; 57: 203–209.
10. Lewis MA, Heinemann LA, Spitzer WO, et al. The use of oral contraceptives and the occurrence of acute myocardial infarction in young women. Results from the Transnational Study on Oral Contraceptives and the Health of Young Women. Contraception. 1997; 56: 129–140.
11. Cancer CGoHFiB. Breast cancer and hormonal contraceptives: Collaborative reanalysis of individual data on 53,297 women with breast cancer and 100,239 women without breast cancer from 54 epidemiological studies. Lancet. 1996; 347: 1713–1727.
12. Grabrick DM, Hartmann LC, Cerhan JR, et al. Risk of breast cancer with oral contraceptive use in women with a family history of breast cancer. JAMA. 2000; 284: 1791–1798.
13. Rosenberg MJ, Waugh MS. Oral contraceptive discontinuation: A prospective evaluation of frequency and reasons. Am J Obstet Gynecol. 1998; 179: 577–582.
14. Sulak P, Lippman J, Siu C, et al. Clinical comparison of triphasic norgestimate/35 μg ethinyl estradiol and monophasic norethindrone acetate/20 μg ethinyl estradiol. Cycle control, lipid effects, and user satisfaction. Contraception. 1999; 59: 161–166.
15. Rosenberg MJ, Meyers A, Roy V. Efficacy, cycle control, and side effects of low- and lower-dose oral contraceptives: A randomized trial of 20-μg and 35-μg estrogen preparations. Contraception. 1999; 60: 321–329.
16. Peterson HB, Lee NC. The health effects of oral contraceptives: Misperceptions, controversies, and continuing good news. Clin Obstet Gynecol. 1989; 32: 339–355.
17. Hankinson SE, Colditz GA, Hunter DJ, et al. A quantitative assessment of oral contraceptive use and risk of ovarian cancer. Obstet Gynecol. 1992; 80: 708–714.
18. Ursin G, Henderson BE, Haile RW, et al. Does oral contraceptive use increase the risk of breast cancer in women with BRCA1/BRCA2 mutations more than in other women? Cancer Research. 1997; 57: 3678–3681.
19. Schlesselman JJ. Risk of endometrial cancer in relation to use of combined oral contraceptives. A practitioner’s guide to meta-analysis. Hum Reprod. 1997; 12: 1851–1863.
20. Charreau I, Plu-Bureau G, Bachelot A, et al. Oral contraceptive use and risk of benign breast disease in a French case-control study of young women. Eur J Cancer Prev. 1993; 2: 147–154.
21. Davis A, Lippman J, Godwin A, et al. Triphasic norgestimate/ethinyl estradiol oral contraceptive for the treatment of dysfunctional uterine bleeding. Obstet Gynecol. 2000; 95: S84.
22. van Hooff MH, Hirasing RA, Kaptein MB, et al. The use of oral contraceptives by adolescents for contraception, menstrual cycle problems, or acne. Acta Obstet Gynecol Scand. 1998; 77: 898–904.
23. Burkman RT. Management of the fibroid uterus. Adv Obstet Gynecol. 1996; 3: 103–128.
24. Redmond GP, Olson WH, Lippman JS, et al. Norgestimate and ethinyl estradiol in the treatment of acne vulgaris: a randomized, placebo-controlled trial. Obstet Gynecol. 1997; 89: 615–622.
25. Lucky AW, Henderson TA, Olson WH, et al. Effectiveness of norgestimate and ethinyl estradiol in treating moderate acne vulgaris. J Am Acad Dermatol. 1997; 37: 746–754.
26. DeCherney A. Bone-sparing properties of oral contraceptives. Am J Obstet Gynecol. 1996; 174: 15–20.
27. Michaelsson K, Baron JA, Farahmand BY, et al. Oral-contraceptive use and risk of hip fracture: A case-control study [see comments]. Lancet. 1999; 353: 1481–1484.
28. Fernandez E, La Vecchia C, Franceschi S, et al. Oral contraceptive use and risk of colorectal cancer. Epidemiology. 1998; 9: 295–300.
29. Spector TD, Hochberg MC. The protective effect of the oral contraceptive pill on rheumatoid arthritis: An overview of the analytic epidemiological studies using meta-analysis. J Clin Epidemiol. 1990; 43: 1221–1230.