Menopause and Hypothalamic-Pituitary Sensitivity to Estrogen
Total Page:16
File Type:pdf, Size:1020Kb
ORIGINAL CONTRIBUTION Menopause and Hypothalamic-Pituitary Sensitivity to Estrogen Gerson Weiss, MD Context The onset of human menopause is thought to be caused solely by ovarian Joan H. Skurnick, PhD failure and oocyte depletion. However, clinical symptoms and certain recent data in Laura T. Goldsmith, PhD perimenopausal women suggest central nervous system involvement. Objective To determine if modifications of hypothalamic-pituitary response to es- Nanette F. Santoro, MD trogen feedback mechanisms occur in older reproductive-age women as a mecha- Susanna J. Park, MD nism of the onset of menopause. Design, Setting, and Participants The Study of Women’s Health Across the Na- Y THE YEAR 2030, MORE THAN 1.2 tion (SWAN) is a multiethnic observational cohort study of the menopausal transition billion women in the world will in 3302 women at 7 US sites. Of the subcohort of 840 women who participated in be at least 50 years old.1 This in- the Daily Hormone Study between 1997 and 1999, 680 women had evidence of lu- creasing proportion of the fe- teal activity. The remaining 160 women (19%) did not have luteal activity and are the Bmale population will be experiencing the subject of this report. menopausal transition with its accom- Main Outcome Measures Daily urinary hormone levels of estrogen and proges- panying physiology and pathophysiol- terone metabolites, luteinizing hormone (LH), and follicle-stimulating hormone (FSH). ogy. Reproductive aging in women and Results Three groups of women were studied: those with estrogen increases with the hormonal changes that occur dur- an LH surge, those with estrogen increases without an LH surge, and those with nei- ing the onset of human menopause have ther. There were no differences in age or ethnicity among the 3 groups of women. been ascribed solely to ovarian failure Women in the third group (no increases) experienced more menopausal symptoms and oocyte depletion.2 However, in other (hot flashes) than did women in the other groups with estrogen increases. In older species, the central nervous system is the reproductive-age women, the frequent existence of anovulatory cycles with estrogen major regulator of age-related reproduc- peaks, equivalent to those that result in LH surges in younger women, yet in which no LH surges occur, indicates failure of estrogen-positive feedback on LH secretion. In tive dysfunction.3 other anovulatory cycles, follicular-phase estrogen levels did not lower LH secretion There are 4 events involving the hy- as occurs in cycles of younger women, indicating decreased estrogen-negative feed- pothalamic-pituitary-ovarian axis that back on LH secretion. control the human menstrual cycle: (1) Conclusion Our findings are compatible with hypothalamic-pituitary insensitivity to The secretion of follicle-stimulating estrogen in aging perimenopausal women. hormone (FSH), responsible for the de- JAMA. 2004;292:2991-2996 www.jama.com velopment of ovarian follicles and pro- duction of estradiol.4 Throughout the cycle, estrogen maintains low gonado- luteinization of the follicle, triggered by sites across the United States designed tropin levels via its negative feedback the LH surge, forming a corpus lu- to enhance understanding of the fac- effect on hypothalamic gonadotropin- teum. This is an ovarian response that tors that influence the health of women releasing hormone and consequently lu- results in progesterone secretion nec- of diverse race and ethnicity.10 The de- teinizing hormone (LH) and FSH se- essary for the establishment of a preg- tails of enrollment have been previ- cretion.5 (2) The FSH-induced increase nancy.9 ously reported.10 Race/ethnicity was in ovarian estrogen secretion to levels How these events may be altered dur- of sufficient strength and duration trig- ing menopausal transition has not been Author Affiliations: Department of Obstetrics, Gy- gering an LH surge (positive feed- well established. necology and Women’s Health, New Jersey Medical 6 School of UMDNJ, Newark, NJ (Drs Weiss, Skurnick, back). (3) The LH surge, a hypotha- Goldsmith, and Park) and Department of Obstetrics, lamic-pituitary response to the estrogen METHODS Gynecology and Women’s Health, Albert Einstein Col- lege of Medicine, New York, NY (Dr Santoro). stimulus. This positive feedback re- The Study of Women’s Health Across Corresponding Author: Laura T. Goldsmith, PhD, De- sponse of estrogen on LH secretion has the Nation (SWAN) is a multiethnic ob- partment of Obstetrics, Gynecology and Women’s Health, New Jersey Medical School of UMDNJ, 185 been used as a test of hypothalamic- servational cohort study of the meno- S Orange Ave, Newark, NJ 07103 (goldsmit@umdnj pituitary function.7,8 (4) Ovulation and pausal transition in 3302 women at 7 .edu). ©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, December 22/29, 2004—Vol 292, No. 24 2991 Downloaded From: https://jamanetwork.com/ on 10/01/2021 MENOPAUSE AND HYPOTHALAMIC-PITUITARY SENSITIVITY TO ESTROGEN self-determined by study participants dence of luteal activity based on a vali- times the mean baseline level (of 4 con- and was obtained by asking the follow- dated algorithm for pregnanediol gluc- secutive days starting 5 days earlier), ing open-ended question: “How would uronide.12 The algorithm locates the 5 (3) baseline mean plus 3 standard de- you describe your primary racial or eth- nadir days of pregnanediol glucuro- viations of the 4-day baseline levels, (4) nic group?” The responses then were nide in the follicular phase using mov- baseline mean plus 2 standard devia- categorized as Caucasian (white), Af- ing averages throughout the cycle. A tions of levels on days 2 through 6 af- rican American, Chinese, Japanese, or 3-fold increase in pregnanediol gluc- ter the peak, and (5) 0.8 times the maxi- Hispanic. This study was approved by uronide concentrations above this na- mum level in that cycle. In addition, the all of the sites’ institutional review dir for at least 3 consecutive days was LH surge was required to culminate in board, and written informed consent considered evidence of luteal activity. a drop to no more than 1.5 times base- was obtained from each participant. The cycles of these women have been line within 6 days following the peak. A subcohort participated in the Daily reported previously.12 The remaining Cycles were classified by these algo- Hormone Study (DHS) from 1997 to 160 women (19% of the total 840 wom- rithms as falling into 1 of 3 distinct pat- 1999. The women in the DHS in- en) did not have luteal activity. One terns: (1) both estrogen increase and LH cluded 257 Caucasian (white) women, woman could not be subclassified due surge (coincident within 2 days), (2) 175 African American women, 152 Chi- to missing data points. The 159 remain- estrogen increase only, and (3) nei- nese women, 170 Japanese women, and ing women are the subjects of this re- ther. Visual inspection of the cycle data 86 women of Hispanic origin. The port. plots by 2 observers (G.W., J.H.S.) re- cohort has been described previ- All women in the SWAN DHS study vealed that 20 cycles had apparent es- ously.11,12 Inclusion criteria were age 42 had levels of FSH equal to or greater trogen increases that were missed by the to 52 years; an intact uterus and at least than those previously demonstrated in defining algorithm due to an increase one ovary; at least one menstrual pe- younger women throughout the en- too early in a short cycle to establish a riod in the prior 3 months; no use of tire cycle.12 This has been described pre- baseline or a slow decline. Four cycles sex-steroid hormones in the previous viously and is due to the decreased se- with an algorithmically determined es- 3 months; and not being pregnant. DHS cretion of ovarian inhibin in older trogen increase were reclassified to “nei- enrollees completed a daily collection women.14,15 Decreased gonadotropin se- ther.” Thus, of 159 cycles, 29 were clas- of morning voided urine for an entire cretion, as is found in some anovula- sified as “both,” 32 as “estrogen increase menstrual cycle ending in bleeding or tory cycles in premenopausal young only,” and 98 as “neither.” These cycle to 50 days, whichever came first. Dur- women in their teens to 30s, clearly did classifications were based solely on hor- ing the cycle that they collected daily not occur in these women. mone levels and not age or meno- urine specimens, DHS enrollees also Conceptually, an estrogen increase pausal symptoms. completed a daily diary, a question- is a high level compared with base- Analysis of variance was conducted naire in which they answered, once a line, in absolute terms and relative to to compare cycle classification groups day, whether they had experienced observed variability, followed by sub- on women’s ages and body mass in- within the preceding 24 hours any stantial decline. The specific criteria dex; 2 tests were conducted to com- trouble sleeping and any hot flashes or used here were adapted from previ- pare groups on ethnicity, reason for night sweats. ously established definitions for mid- ending collection, and experience of Urinary LH, FSH, the estradiol uri- reproductive age women.13 An estro- symptoms during the cycle. Rank- nary metabolites estrone conjugates gen increase was defined as an E1c level sum tests were conducted to compare (E1c), and the progesterone urinary me- of at least (1) 50 pg/mg creatinine, (2) E1c levels of groups 1 and 2 by cycle tabolite pregnanediol glucuronide were twice the baseline level (the mean of 5 day and to compare the 3 groups on measured using chemiluminescent as- consecutive days starting 9 days ear- women’s percentage of cycle days with says as described previously.11,12 Con- lier), and (3) 3 standard deviations of symptom occurrence.