FERTILITY AND STERILITY VOL. 71, NO. 6, JUNE 1999 Copyright 1999 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Effect of age on the response of the hypothalamo-pituitary-ovarian axis to a combined oral contraceptive Cheryl Fitzgerald, M.D.,* Max Elstein, M.D., and Jurgen Spona, Ph.D. Department of Reproductive Medicine, University Hospital of South Manchester, Manchester, United Kingdom Received June 10, 1998; revised and accepted February 3, 1999. Supported by Schering AG, Berlin, Germany. Reprint requests: Cheryl Fitzgerald, M.D., Department of Obstetrics and Gynaecology, St. Mary s Hospital, Hathersage Road, Rusholme, Manchester M13 0JH, United Kingdom (FAX: 0161-224-0957). * Department of Obstetrics and Gynaecology, St. Mary s Hospital, Manchester. Institute of Medicine, Law and Bioethics, University of Manchester, Manchester. Department of Obstetrics and Gynaecology, Vienna University and Ludwig Boltzmann Institute for Experimental Endocrinology, Division of Cellular Endocrinology, Vienna, Austria. 0015-0282/99/$20.00 PII S0015-0282(99)00146-6 Objective: To examine the effect of age on the response to treatment with a combined oral contraceptive. Design: Prospective, controlled clinical study. Setting: Reproductive medicine unit in a tertiary care university medical center. Patient(s): Twenty-six healthy female volunteers aged 21 45 years. Intervention(s): After a control cycle, all the women were given a combined oral contraceptive containing 20 g of ethinylestradiol with 75 g of gestodene for three cycles. The women were examined through the posttreatment cycle. Main Outcome Measure(s): Pituitary and ovarian activity was assessed with ultrasound and measurement of ovarian steroids. Result(s): Follicular activity was observed in all treatment cycles, although ovulation was inhibited. Ovarian suppression was maximal in cycle 1. Mean endogenous E 2 levels were lower during cycles 2 and 3 in the older group. Serum FSH levels were higher in the control cycle and on day 28 of the treatment cycles in the older group. Most women ovulated during the posttreatment cycle. Conclusion(s): Combined oral contraceptives did not inhibit all ovarian activity; maximal suppression was seen in cycle 1. Less follicular activity was observed in cycles 2 and 3 in the older group. Raised FSH levels with age reflect increasing ovarian resistance to follicular development. (Fertil Steril 1999;71:1079 84. 1999 by American Society for Reproductive Medicine.) Key Words: Age-related, combined oral contraceptive, fertility decline, follicular development, FSH, pituitary suppression, ovulatory control Treatment with a combined oral contraceptive (COC) does not reduce ovarian activity to zero. During treatment cycles with such preparations, levels of endogenous E 2 and pituitary gonadotropins fluctuate and ultrasound (US) visualization of the pelvis reveals the growth of ovarian follicles, along with changes in uterine size and endometrial thickness (1, 2). Suppression of follicular development as witnessed by US findings and a reduction in endogenous E 2 levels begins soon after the beginning of pill intake. At the beginning of the pill-free interval, the release of pituitary suppression results in immediate follicular recruitment associated with a rise in endogenous serum E 2 levels (3, 4). The follicles observed at this time are dependent on gonadotropins for further development, and thus the COC exerts its action primarily through pituitary suppression (3). Residual pituitary activity is sufficient for preovulatorysized follicles to develop in some women; these have been reported in 2% 30% of women during treatment (5, 6). If a follicle grows sufficiently, it can respond to an artificial gonadotropin surge, resulting in ovulation and a subsequent rise in the progesterone level (7). Many factors control the amount of ovarian activity seen during treatment with a COC, including the dose of steroid, type of progestogen, administration regimen, and individual responsiveness of the woman. Since the introduction of the COC, there has been a gradual reduction in the dose of steroid used to minimize adverse effects. However, lower-dose COCs have been associated with higher levels 1079
of serum gonadotropins (8, 9) and greater pituitary responsiveness to GnRH (10), allowing more follicular development (5). The doses of estrogen (9) and progestogen (11) have been implicated in controlling pituitary and hence ovarian activity. Individual women respond differently to treatment; although many demonstrate good ovarian suppression with repeated pill omission, others experience contraceptive failure in spite of good compliance. It is not yet possible to predict those women in whom pill failure can most readily occur. In 1990, the U.S. Food and Drug Administration (12) issued the following statement: The suggestion that women over 40 who do not smoke should not take oral contraceptives is based on information from older, high-dose pills and on less selective use of pills than is practiced today. An Advisory Committee of the FDA discussed this issue in 1989 and recommended that the benefits of oral contraceptive use by healthy, non-smoking women over 40 years of age may outweigh the possible risks. However, all women, especially older women are cautioned to use the lowest dose of pill that is effective. The safety of COCs as a contraceptive option for healthy, nonsmoking older women is now accepted. Many suggest that it is a particularly appropriate choice for older women because it controls climacteric symptoms, regulates menstruation, and may increase bone mineral density (13). An alteration in hypothalamo-pituitary-ovarian control as women age has been demonstrated (14); in addition, older women respond less readily to fertility-enhancing drugs (15). This study was designed to examine ovarian activity during pill treatment cycles in women of different ages to determine whether different regimens would be more suitable for older women. MATERIALS AND METHODS Thirty-five women aged 18 45 years were recruited for the study from the staff of the University Hospital of South Manchester. All the women were healthy, with no history of subfertility requiring investigation or treatment. During the 6 months before the study, the women had regular menstrual cycles of 25 35 days, did not use oral contraceptives, were not pregnant, and did not breast-feed. The volunteers were asked to use barrier methods of contraception for the duration of the study. All the women gave informed consent to the study, and the local ethics committee approved the study. The women were examined throughout a control cycle to confirm ovulation. Only women in whom ovulation was confirmed by US findings and a raised serum progesterone level of 30 pmol/l in the luteal phase were allowed to enter the study. The women then were started on treatment with a combination of 20 g of ethinyl estradiol and 75 g of gestodene for three cycles. Treatment was started on day 1 of the cycle and taken continuously for 21 days, followed by a 7-day pill-free interval. Any woman who delayed pill intake for more than 12 hours was excluded from the rest of the study. Any woman who did not experience withdrawal bleeding was seen immediately before the next cycle of treatment for measurement of -hcg levels so as to exclude pregnancy. The women were seen during the morning on days 2, 4, 8, 10, 12, 14, 16, 18, 20, 25, and 28 during treatment. At each visit, a transabdominal US scan was performed with a sector scanner through a full urinary bladder, using a 5-MHz transducer (Ultramark 4 Plus; Type Advanced Technology Laboratories, Trevose, PA). All measurements were made from a frozen image on the screen. Follicular diameter was calculated using the mean value of measurements obtained in the longitudinal and transverse directions in the coronal and sagittal planes. If a follicle of 10 mm in diameter was observed, the woman was asked to attend daily for examination. Blood was taken from the antecubital fossa at each visit. The whole blood was centrifuged within 4 hours of collection at 2,500 3,000 rpm for 10 minutes, and the serum obtained was frozen and stored at 20 C until assayed. To reduce interassay variability, samples were assayed only after all volunteers completed the study. The intra-assay and interassay coefficients of variation of all assays were 5% 9%. Levels of LH were determined using materials obtained from Sorin Biomedical (France). Follicle-stimulating hormone assays were performed using materials provided by the Behring Institute (Germany). Levels of 17 -E 2 and progesterone were estimated using materials obtained from Diagnostic Products Corporation (Los Angeles, CA). The sensitivity of the 17 -E 2 and progesterone assays was 8 mg/ml and 0.03 ng/ml, respectively. The assay of 17 -E 2 had a cross-reactivity with ethinyl estradiol of 1.8%. All the women kept a diary throughout the study to record pill-taking, bleeding patterns, and any noticeable side effects. After the three treatment cycles, the women were examined through a follow-up cycle on days 2, 4, 8, and 10 before ovulation and then on days 4, 7, and 10 after ovulation. A US scan and venesection were performed at these visits as previously described. For purposes of comparison, the women were split into two groups by age: 18 34 years and 35 45 years. The groups were compared using the Student s t-test, and statistical significance was set at the conventional level of P.05. Analyses also were performed using logistic regression methods to examine various parameters as a function of age. RESULTS During the control cycle, 2 women dropped out of the study because of reasons unrelated to the trial, 5 women were excluded because they failed to ovulate, and 1 woman became pregnant. Twenty-seven women began treatment, and 1 of these dropped out during the first treatment cycle for a reason unrelated to treatment. The remaining 26 women all 1080 Fitzgerald et al. The effect of age on response to a COC Vol. 71, No. 6, June 1999
completed the study, and data collected from these women are presented. Sixteen women were aged 18 34 years and 10 were aged 35 45 years. Some data from the younger women were presented elsewhere in an article comparing the effect of different progestogens in low-dose oral contraceptive regimens (16). No ovulation was seen during any of the treatment cycles, and serum progesterone levels were 2 pmol/l. However, follicular activity was observed on US scan and confirmed by raised serum E 2 levels. Serum E 2 levels were suppressed rapidly after the start of pill intake and reached a trough on day 8 of the cycle, where they remained until the end of the 21 days and then began to rise again. This pattern was seen through all three cycles of treatment and in both age groups. Mean serum E 2 levels were significantly lower in the older age group (P.05) throughout cycles 2 and 3 (Fig. 1B and C), but there was no statistically significant difference in cycle 1 (Fig. 1A). Serum gonadotropin concentrations showed a rising trend with age. A significant correlation was seen between menstrual serum FSH levels and age during the control cycle (P.001) (Fig. 2). During treatment, serum FSH levels were significantly higher on day 28 of all treatment cycles for the older age group (P.001) (Fig. 3). A positive, exponential correlation was observed between day 28 FSH levels and age in all cycles (P.001). A positive correlation was seen between the FSH level on day 2 of the control cycle and the FSH level on day 28 of treatment cycle 3 (P.01). Serum LH levels on day 28 of all treatment cycles also were positively correlated with age, although the relation was less significant than for FSH and reduced with each successive cycle. Recovery of ovarian function during the posttreatment cycle was witnessed in both groups. In the 16 younger women, 15 ovulated within this cycle and 1 had a luteinized unruptured follicle, which appeared as a persistent follicle on US scan associated with a serum progesterone level of 30 pmol/l. Of the 10 older women, 9 ovulated and 1 had a luteinized unruptured follicle. Breakthrough bleeding was defined as bleeding after the seventh day of pill intake but before the pill-free interval in cycle 1 and any bleeding during pill intake in cycles 2 and 3. More than 50% of volunteers described breakthrough bleeding in cycle 1, and this dropped to 10% for cycles 2 and 3. There were no statistically significant differences between the age groups. Side effects were few and minor, with no differences between the groups. DISCUSSION The COC pill is now widely accepted as a safe, reliable form of contraception for healthy, nonsmoking women 35 years of age. At an age when an unwanted pregnancy can have particularly serious physical and psychological consequences, reliable contraception is essential; their efficacy, safety, and noncontraceptive benefits make COCs an appropriate choice. The effects of COCs on the hypothalamopituitary-ovarian axis have been investigated extensively but their effects during the later years of a woman s reproductive life, when menstrual control is changing, have not. This study examined 26 women to see whether their response to such a hormonal preparation changes with age. In an attempt to minimize side effects, there has been a gradual reduction in the steroid content of the COCs marketed. It has been shown that ovarian activity is not reduced to zero during treatment, even with older, high-dose pills (17), but the suppressive effect of the pill on ovarian function is dose-dependent (9, 11). As ovarian activity during treatment increases, there is a greater chance of escape ovulation and thus contraceptive failure. The preparation examined here was a low-dose preparation containing 20 g of ethinyl estradiol and 75 g of gestodene. Ovulation was inhibited in all treatment cycles of the 26 women examined, but follicular activity was seen. Ovarian suppression was maximal in treatment cycle 1 in both age groups, as has been demonstrated with a preparation containing 30 g of ethinyl estradiol and 75 g of gestodene (18). In this cycle, no difference was observed in the ovarian response of the two age groups. However, in cycles 2 and 3, more follicular activity was seen on US scan and significantly higher mean endogenous E 2 levels were measured in the younger age group, suggesting that older women have a lower threshold for ovarian suppression. Changes are seen in menstrual cycle control 5 10 years before menopause, particularly a rise in basal serum gonadotropin levels (14, 19), so greater ovarian suppression during treatment with a COC may be expected. Previous investigators failed to show any age-related difference in response to treatment with an oral contraceptive, although the preparation used was a higher-dose preparation (10). The more profound suppression observed in the older age group in this study suggests that the chance of breakthrough ovulation occurring (e.g., after incorrect pill intake) would be further reduced. In addition, total circulating estrogen, exogenous ethinyl estradiol, and endogenous E 2 levels are lower, and thus the incidence of serious side effects such as thrombosis should be minimized. The reversibility of the COC as a contraceptive method is of great importance, and it is essential to establish that its use in older women is not associated with a delay in return to fertility. Fertility declines with age, and any delay in the return of ovulatory cycles would be a particular problem. These data showed that 9 of 10 women 35 years of age ovulated normally in the immediate posttreatment cycle, comparing favorably with the younger group. Bracken et al. (20) found that conception delay after discontinuation of oral contraceptive administration is related to the dose of estro- FERTILITY & STERILITY 1081
FIGURE 1 (A), Mean serum E 2 level throughout cycle 1 in women of different ages. (B), Mean serum E 2 level throughout cycle 2. (C), Mean serum E 2 level throughout cycle 3. F 18 34 years of age; 35 45 years of age. 1082 Fitzgerald et al. The effect of age on response to a COC Vol. 71, No. 6, June 1999
FIGURE 2 Serum FSH levels during the control cycle as a function of age (trend line shown). gen, a finding supported by these data, in which a rapid return to ovulation was seen with this low-dose preparation. In this study, basal FSH levels were significantly, positively correlated with age in the control cycle, as has been demonstrated elsewhere (19); older women require higher levels of FSH to induce folliculogenesis. It was reassuring to observe that on discontinuation of treatment, rebound FSH levels correlated with those seen in the control cycles and produced a satisfactory ovarian response. Some data also suggest that long-term COC use protects against spontaneous abortion (21), giving further reassurance to the older population of users. Side effects reported by both age groups were few and minor, and did not differ between the groups. Very-low-dose pills are associated with more breakthrough bleeding because the dose of exogenous steroid is insufficient to stabilize the endometrium in a cycle in which endogenous E 2 levels are very low. Because of the lower levels of circulat- FIGURE 3 Mean serum FSH levels throughout all three cycles of treatment in the two age groups. 18 34 years of age; 35 45 years of age. FERTILITY & STERILITY 1083
ing E 2 in the older group, breakthrough bleeding may have been expected to be more of a problem, but this did not occur. Cycle control with this preparation was very good, with no statistically significant difference seen between the two groups. Previous reports have shown very good cycle control with gestodene-containing products (22), and these findings may reflect this. This study clearly demonstrated an alteration in the response of the hypothalamo-pituitary-ovarian axis of older women to a low-dose COC preparation. Ovarian suppression was greater in the women 35 years of age than in the younger group and was demonstrated clearly by lower serum E 2 levels in the older group in treatment cycles 2 and 3. Because of the release of pituitary suppression during the pill-free interval, FSH levels rise; these were significantly higher in the older group and correlated with pretreatment values. After treatment, ovulatory function returned rapidly, and no difference was seen in the return of ovulation and thus potential fertility between the groups. Low-dose COCs offer a safe, reliable, and reversible method of birth control to women of all ages. These data suggest that in older women, there is greater ovarian suppression and thus a reduced chance of contraceptive failure, side effects are few, and the return to potentially fertile cycles after discontinuation of treatment is rapid. These findings are particularly reassuring to this group of women, who should be encouraged to use such low-dose preparations. Acknowledgments: The authors thank the staff of the Reproductive Medicine Unit for their help in data collection and the volunteers without whom the study would not have been possible. References 1. Spona J, Elstein M, Sullivan H, Ludicke F, Muller U, Dusterberg B. Shorter pill-free interval in combined oral contraceptives decreases follicular development. Contraception 1996;54:71 7. 2. Spona J, Feichtinger W, Kinderman C, Wunsch C, Brill K. Inhibition of ovulation by an oral contraceptive containing 100 g levonorgestrel in combination with 20 g ethinyloestradiol. Contraception 1996;54:299 304. 3. Killick SR, Eyong E, Elstein M. Ovarian follicular development in oral contraceptive cycles. Fertil Steril 1987;48:409 13. 4. Van der Spuy ZM, Sohnius U, Pienaar CA, Schall R. Gonadotropin and oestradiol secretion during the week of placebo therapy in oral contraceptive users. Contraception 1990;42:597 609. 5. Tayob Y, Robinson G, Adams J, Nye M, Whitelaw N, Shaw RW, et al. Ultrasound appearance of the ovaries during the pill free interval. British Journal of Family Planning 1990;16:94 6. 6. Van der Vange N. Ovarian activity with low dose oral contraceptives. In: Chamberlain G, ed. Butterworth s contemporary obstetrics and gynaecology. London: Butterworth, 1988:315 26. 7. Killick SR. Ovarian follicles during oral contraceptive cycles: their potential for ovulation. Fertil Steril 1989;52:580 2. 8. Spellacy WN, Kalra PS, Buhi WC, Birk SA. Pituitary and ovarian responsiveness to a graded gonadotropin releasing factor stimulation test in women using a low-estrogen or a regular type of oral contraceptive. Am J Obstet Gynecol 1980;137:109 15. 9. Mall-Haefeli M, Werner-Zodrow I, Huber PR, Edelman A. Oral contraception and ovarian function. In: Runnebaum B, Rabe T, Kiesel L, eds. Female contraception. Berlin/New York: Springer-Verlag, 1988: 1474 5. 10. Scott JA, Brenner PF, Kletzy OA, Mishell DR. Factors affecting pituitary gonadotropin function in users of oral contraceptive steroids. Am J Obstet Gynecol 1978;130:817 21. 11. Croxatto HB, Diaz S, Pavez M, Brandeis A. Estradiol plasma levels during long-term treatment with Norplant subdermal implants. Contraception 1988;38:465 75. 12. Food and Drug Administration. Correspondence. Jan 30 1990. Data on file. 13. Corson SL. Oral contraceptives for the prevention of osteoporosis. J Reprod Med 1993;38:1015 20. 14. Fitzgerald CT, Seif MW, Killick SR, Elstein M. Age related decline in the female reproductive cycle. Br J Obstet Gynaecol 1994;101:229 33. 15. Avrech OM, Royburt M, Sabah G, Zukerman Z, Pinkash H, Amit S, et al. The initial flare-up induced by gonadotropin releasing hormone agonist may serve as a predictor of ovarian response in the current IVF-ET treatment cycle in normogonadotropic women aged 40 48 years. J Assist Reprod Genet 1996;13:395 400. 16. Fitzgerald CT, Feichtinger W, Spona J, Elstein M, Muller U, Williams C. A comparison of the effects of two monophasic low dose oral contraceptives on the inhibition of ovulation. Adv Contracept 1994;10: 5 18. 17. Elstein M, Bristow PG, Jenkins M, Kirk D, Miller H. Effects of a low dose oestrogen contraceptive on urinary excretion of luteinizing hormone and ovarian steroids. Br Med J 1974;1:11 3. 18. Thomas K, Vankreiken L. Inhibition of ovulation by a low dose monophasic contraceptive containing gestodene. Am J Obstet Gynecol 1990;163:1404 10. 19. Lee SJ, Lenton EA, Sexton L, Cooke ID. The effect of age on the cyclical patterns of plasma LH, FSH, oestradiol and progesterone in women with regular menstrual cycles. Hum Reprod 1988;3:851 5. 20. Bracken MB, Hellenbrand KG, Holford TR. Conception delay after oral contraceptive use: the effect of estrogen dose. Fertil Steril 1990;53: 21 7. 21. Ford JH, MacCormack L. Pregnancy and lifestyle study: the long-term use of the contraceptive pill and the risk of age related miscarriage. Hum Reprod 1995;10:1397 402. 22. Dusterberg BG, Brill K. Clinical acceptability of monophasic gestodene. Am J Obstet Gynecol 1990;163:1398 404. 1084 Fitzgerald et al. The effect of age on response to a COC Vol. 71, No. 6, June 1999