Effects of previous use of oral contraceptives on early follicular phase follicle-stimulating hormone*

Effects of previous use of oral contraceptives on early follicular phase follicle-stimulating hormone*

Vol. 64, No.4, October 1995 FERTILITY AND STERILITY Printed on acid-free paper in U. S. A Copyright c 1995 American Society for Reproductive Medici...

549KB Sizes 1 Downloads 7 Views

Vol. 64, No.4, October 1995


Printed on acid-free paper in U. S. A

Copyright c 1995 American Society for Reproductive Medicine

Effects of previous use of oral contraceptives on early follicular phase follicle-stimulating hormone*

Robert L. Barbieri, M.D.t Xiaoying Gao, M.D. Huijuan Xu, M.P.H. Daniel W. Cramer, M.D., Sc.D. Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts

Objective: To determine if previous oral contraceptive (OC) use is associated with changes in early follicular phase FSH, LH or E 2 • Design: A cross-sectional study examining determinants of early follicular phase hormone levels. Subjects: Subjects included 106 premenopausal women with a family history of ovarian cancer and 116 premenopausal women without this history who were not taking OCs currently. All subjects completed a structured interview and gave an early follicular phase blood sample. Setting: Gynecologic Epidemiology Center and Familial Ovarian Cancer Research Center. Main Outcome Measures: Follicle-stimulating hormone, LH, and E2 were measured in early follicular phase plasma samples. Results: Recency or length of prior OC use did not affect early follicular phase LH or E2 levels. Length of OC use did not affect FSH levels in all subjects; but lower levels of FSH were observed in women over age 45 who had used OCs for >5 years. Early follicular phase FSH is lower in women with OC use within the past 5 years compared with women with more remote use or who never used OCs, after adjustment for age, smoking, and family history status. Conclusions: Past use of OCs may have a residual effect on basal FSH levels in women not using them currently that depends on recency of use and to a lesser extent duration of prior use. Fertil Steril 1995;64:689-92 Key Words: Oral contraceptives, FSH, early follicular phase, LH, estradiol, smoking

Oral contraceptives (OCs) inhibit ovarian follicular growth and suppress ovulation by decreasing the secretion of LH and FSH (1). Monophasic estrogenprogestin OCs decrease both LH pulse frequency and amplitude (2). The decrease in LH pulse frequency associated with OC use implies that these agents decrease GnRH pulse frequency. The progestagenic

Received September 14, 1994; revised and accepted April 25, 1995. * Supported in part by a grant 90-BW25 from the American Institute of Cancer Research, Washington, D.C. t Reprint requests: Robert L. Barbieri, M.D., Department of Obstetrics and Gynecology, Brigham and Women's Hospital, 75 Francis Street, Boston, Massachusetts 02115 (FAX: 617-7321440). Vol. 64, No.4, October 1995

component of OCs also has a major influence on the pattern of hypothalamic GnRH secretion (3). Although there is understanding of the effects of current OC use on gonadotropin levels, the effects of past OC use on gonadotropins and E2 levels in naturally cycling women are less clear. Data collected in the context of a study designed to examine hormonal levels in women with and without a family history of ovarian cancer also provided information about the effects of prior OC use on hormonal levels and is the topic of this paper. MATERIALS AND METHODS Experimental Subjects

The study protocol was approved by the institutional review board, and all study subjects consented Barbieri et aI. Previous OC use and early follicular FSH



to participate. The subjects in the study were premenopausal women with a family history of ovarian cancer, defined as one primary or two second-degree relatives with ovarian cancer, or women without a family history of ovarian cancer recruited from the general population of greater Boston using town lists. By study design, women currently using OCs were excluded, but past users were included if they had discontinued use ~3 months before study entry. The sample included 106 women with a family history of ovarian cancer and 116 women without this history. Subjects must not have had hysterectomy or bilateral oophorectomy, although subjects who had either partial or complete loss of one ovary were included. Additional factors pertaining to the selection of these subjects and their characteristics have been reported previously (4, 5). Hormone Measurements

Gonadotropins and E2 were measured in venous blood specimens taken during the early follicular phase of the menstrual cycle: day 1 (21% of subjects), day 2 (41% of subjects), or day 3 (38% of subjects). Blood was drawn during daytime working hours into heparinized tubes. The plasma was stored at -70°C before assay. Estradiol was measured in unextracted plasma using a solid phase RIA (Diagnostic Products Corp., Los Angeles, CA). Luteinizing hormone and FSH were assayed using an immunoradiometric method with the World Health Organization 2nd International Reference Preparation-hMG as the standard (Diagnostic Products Corp.). All assays were performed in duplicate and averaged. The intraassay and interassay coefficients of variation were <10% and 15%, respectively, for all three assays. Determination of Previous OC Use

Before entry into the study all subjects completed a structured interview and a questionnaire, which included a series of inquiries concerning OC use, including years since discontinuing use and total years of use. Also, details were obtained on current and past smoking and a variety of factors from the reproductive history. Statistical Analysis

For the analysis related to effects of past OC use, subjects with and without a family history of ovarian cancer have been combined. Univariate analyses to determine key variables affecting FSH levels included analysis of variance (F tests) using standard Statistical Analysis System (SAS Institute, Cary, NC) programs. Multiple linear regression analysis was applied to assess the effects of more than one 690

Barbieri et al. Previous DC use and early follicular FSH

study variable on FSH, including status by family history and smoking. RESULTS

Luteinizing hormone and E2 levels did not vary significantly by length or recency of OC use (data not shown). The principal effect of past OC use was observed for FSH levels (Table 1). Women who had used OCs within 5 years of study had lower FSH levels compared with women who never used them or used them >5 years before the study. This difference was apparent in all subjects combined and persisted after adjustment for family history status. However, few women over age 40 in this study had been recent users of OCs, and the effect of recent OC use on early follicular FSH levels was most apparent in women under age 40. Data on FSH levels as they varied by total length of OC use also are shown in Table 1. Overall no significant differences were noted in women who had never used OCs compared with those who had used them <3 years, 3 to 5 years, or >5 years. However, in women over age 45 who had used OCs for >5 years, the mean (±SE) for FSH was 12.9 ± 2.4 mIUlmL (conversion factor to SI unit, 1.00) compared with 18.6 ± 3.6 mIUI mL for women who had never used OCs, which did represent a significant difference (t = 2.38, P = 0.03). Factors that might confound the association between OC use and FSH levels in this data set include age, smoking, and family history status. Table 2 shows regression parameters and P values from a multiple linear regression model with FSH level as the dependent variable and age, family history status, smoking, length of OC use, and recency of OC use as the independent variables. This model was restricted to women <45 because no one over 45 had recent OC use. For simplicity, the independent variables were dichotomized: age «40, ~40), family history of ovarian cancer (yes, no), current smoking (yes, no), length of OC use (~5, <5, or none), and recency of OC use (:'5:5 years, none, or >5 years). Significant variables in this model included age, smoking status, and recency ofOC use. Replacement of age by a continuous variable in this model reduced the P value for recency of OC use to P = 0.12. DISCUSSION

During the menstrual cycle, FSH levels vary considerably in response to ovarian steroid (E 2) and protein (inhibin) signals (6). Follicle-stimulating hormone levels are relatively elevated and fairly constant during cycle days 1 through 5 when circulating E2 and inhibin levels are suppressed. After cycle day 5 and before ovulation, FSH levels are Fertility and Sterility

Table 1 Basal FSH Levels as Affected by Age, Recency of the Length of CC Use* Age Category of OC use None Recency of use >5y 3 to 5 y <3 y Length of use <3 y 3 to 5 y >5y


35 to 39 y


;;",45 y

All subjects

11.2 ± 1.16 (10)

10.6 ± 0.80 (19)

14.6 ± 1.60 (15)

18.6 ± 3.56 (8)

13.1 ± 0.87 (52)

10.8 ± 0.86 (15) 9.6 ± 0.63 (12) 8.3 ± 0.52 (5)

12.3 ± 0.84 (45) 10.1 ± 0.72 (9) 8.8 ± 0.80 (2)

14.2 ± 0.98 (55) 13.5 ± 1.86 (5) 10.8 (1)

15.2 ± 1.81 (21)

13.4 ± 0.57 (136) 10.5 ± 0.57 (26) 8.7 ± 0.46 (8)

10.2 ± 0.72 (19) 8.6 ± 0.45 (7) 10.9 ± 1.10 (6)

10.8 ± 1.16 (23) 12.5 ± 1.32 (17) 12.7 ± 1.07 (16)

15.4 ± 1.54 (22) 12.6 ± 1.14 (20) 14.1 ± 1.92 (19)

18.2 ± 3.79 (4) 16.8 ± 3.74 (7) 12.9 ± 2.36 (10)

12.5 ± 0.76 (68) 12.6 ± 0.84 (51) 13.1 ± 0.91 (51)

* Values are means ± SE, with number of subjects in parentheses.

somewhat lower and E 2 1evels higher. Mter the ovulatory surge, FSH levels are quite depressed coincident with high E2 and inhibin levels (7). Recently, there has been a focus on the early follicular phase (basal) FSH levels as a predictor of ovarian competence. This is supported by data from IVF clinics, indicating that basal FSH in an unstimulated cycle predicts subsequent IVF success as measured by peak E2 retrieved, follicles recruited, oocytes achieved, and clinical pregnancy rate (8). In turn, it can be hypothesized that any factor likely to decrease oocyte quality or number is also likely to raise the basal FSH level, an effect likely mediated through inhibin. Age is the most obvious of such factors and basal FSH levels clearly increase with age such that the effect begins to be detectable in the mid-30s to late 30s and is quite apparent by age 40 (9, 10). On the basis of evidence that smoking is linked to an early menopause (11,12), it might be predicted and has been observed that premenopausal smokers have higher FSH levels (13,14). The latter two studies also investigated the consumption and metabolism of another potential oocyte toxin, galactose or milk sugar, on basal FSH levels but disagreed on which aspect might be more important. Cramer et al. (13) found that women who were carriers of classic or the Duarte variant of galactose-1phosphate uridyltransferase (GALT) had higher

Table 2 Multiple Linear Regression Parameters ± SE of FSH on Selected Variables* Variable Family history (yes versus no) Age (;;",40 versus <40 y) Smoking (yes versus no) Length of OC use (>5 versus :s 5 y) Recency of OC use (:s5 versus >5 y) Intercept * Excludes women;;", age 45. Vol. 64, No.4, October 1995

Parameter 0.14 2.64 2.28 1.08 -2.23 9.86

± ± ± ± ± ±

0.79 0.82 1.01 0.98 1.06 7.45

P value (t-test) 0.864 0.002 0.025 0.271 0.038 <0.005

FSH levels but did not find an effect of galactose consumption. Cooper et al. (14) found no effect of GALT measured quantitatively but found higher FSH levels in women who consumed >6 g/d of galactose. Previous surgery on both ovaries is another obvious factor that affects oocyte number and may affect the success ofIVF (15). However, the five subjects in this study who had had either cystectomy or unilateral oophorectomy did not have significantly elevated FSH (13). In this article we have focused on prior OC use as a predictor of the basal FSH level. Overall there was no effect of length of prior OC use except in the subgroup of subjects who were older than age 45. A stronger association was observed between recency of OC use and basal FSH. We observed that women who had used OCs within 5 years had lower FSH levels compared with women who had never used them or used them > 5 years before the study. The effect was most apparent in women less than age 40 and persisted after adjustment for family history and smoking status. Age and smoking are clearly key potential confounders in examining any association between OC use and basal FSH levels. However, inclusion in the multivariate model of additional variables related to galactose consumption and metabolism (see previous paragraph) did not affect the association between FSH and recency of OC use. A variety of other factors have been examined previously in this data set, including cycle day of study, menstrual symptomatology, pregnancy history, and ovarian surgery, and were not found to predict basal FSH and, hence, unlikely to be confounders in the association between past OC use and FSH (13). The biologic mechanism subserving the relationship between a modest decrease in the early follicular phase FSH and a history of prior OC use is unknown. Oral contraceptive use might influence the hypothalamic-pituitary-ovarian set point for FSH secretion. Alternatively, the use of OCs might decrease the natural rate of follicular loss resulting in Barbieri et aI. Previous OC use and early follicular FSH


the observed decrease in early follicular phase FSH. Further research will be necessary to clarify the mechanisms responsible for our observation. We believe such research would be valuable in clarifying how prior OC use may reduce the risk for ovarian cancer. Numerous studies have show that OC use reduces ovarian cancer risk and a residual effect may persist for some time (16). Residual effects on gonadotropin levels associated with past use of OCs could be a mechanism to explain the protective effect of OCs on ovarian cancer risk based on clinical and experimental evidence that gonadotropins are important in the pathogenesis of ovarian cancer (5). REFERENCES 1. Mishell DR, Kletzky OA, Brenner PF, Roy S, Nicoloff J. The effect of contraceptive steroids on hypothalamic pituitary function. Am J Obstet Gynecol 1977; 138:60-74. 2. Hemrika DJ, Slaats EH, Kennedy JC, de Vries Robles-Korsen TJM, Shoemaker J. Pulsatile luteinizing hormone secretion during the first and fourth cycle on two different oral contraceptives containing gestodene. Acta Endocrinol (Copenh) 1993; 77:420-6. 3. Hemrika DJ, Slaats EH, Kennedy JC, de Vries-RoblesKorsen TJM, Shoemaker J. Pulsatile luteinizing hormone patterns in long term oral contraceptive users. J Clin Endocrinol Metab 1993; 77:420-6. 4. Cramer DW, Muto MG, Reichardt JKV, Xu H, Welch WR, Valles B, et al. Characteristics of women with a family history of ovarian cancer. I. Galactose consumption and metabolism. Cancer 1994;74:1309-17. 5. Cramer DW, Barbieri RL, Muto MG, Kelly A, Brucks JP, Harlow BL. Characteristics of women with a family history of ovarian cancer. II. Follicular phase hormone levels. Cancer 1994; 74:1318-22.


Barbieri et al. Previous DC use and early follicular FSH

6. Schneyer AL, Slust PM, Whitcomb RW, Sprengel R, Crowley WF. Precursors of alpha-inhibin modulate follicle stimulating hormone receptor binding and biological activity. Endocrinology 1991;129:1987-22. 7. Roseff SJ, Bangah ML, Kettel LM, Vale W, Rivier J, Burger HG, et al. Dynamic changes in circulation inhibin levels during the luteal-follicular transition of the human menstrual cycle. J Clin Endocrinol Metab 1989;69:1033-9. 8. Toner JP, Philput CB, Jones GS, Muasher SJ. Basal follicle stimulating hormone level is a better predictor of in vitro fertilization performance than age. Fertil Steril 1991;55: 784-91. 9. Reyes FI, Winter JSD, Faiman C. Pituitary-ovarian relationships preceding the menopause. I. A cross sectional study of serum follicle stimulating hormone, luteinizing hormone, prolactin, estradiol and progesterone levels. Am J Obstet Gynecol 1993; 129:557 -64. 10. Lenton EA, Sexton L, Lee S, Cooke ID. Progressive changes in LH and FSH and LH: FSH ratio in women throughout reproductive life. Maturitas 1988; 10:35-43. 11. Jick H, Porter J, Morrison AS. Relation between smoking and age of natural menopause. Lancet 1972; 1:1254-5. 12. Kaufman DW, Slone D, Rosenberg L, Mittinen OS, Shapiro S. Cigarette smoking and age at natural menopause. Am J Public Health 1989;70:420-2. 13. Cramer DW, Barbieri RL, Xu H, Reichardt JKY. Determinants of basal follicle stimulating hormone levels. J Clin Endocrinol Metab 1994; 79:1105-9. 14. Cooper GS, Hulka BS, Baird DD, Savitz DA, Hughes CL Jr, Weinberg CR, et al. Galactose consumption, metabolism, and follicle-stimulating hormone concentrations in women of late reproductive age. Fertil Steril1994;62:1168-75. 15. Hornstein MH, Barbieri RL, McShane PM. The effects of previous ovarian surgery on the follicular response to ovulation induction in an in vitro fertilization program. J Reprod Med 1989;34:277 -81. 16. Gross TP, Schlesselman JJ. The estimated effect of oral contraceptive use on the cumulative risk of epithelial ovarian cancer. Obstet Gynecol 1994;83:419-24.

Fertility and Sterility