| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Unipath Ltd. (F.M., J.C., P.W.P., J.E.E.), Bedford MK44 3UP, United Kingdom; Unilever Research Colworth (L.J.A.), Bedford MK44 1LQ, United Kingdom; and Parkwood Clinic (S.W.P.), Bournemouth BH7 7DW, United Kingdom
Address all correspondence and requests for reprints to: F. Miro, Unipath Ltd., Stannard Way, Priory Business Park, Bedford MK44 3UP, United Kingdom. E-mail: fernando.miro{at}unipath.com.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A large database of daily urinary concentrations of FSH, LH, and estrone 3-glucuronide (E1G) from 37 regularly menstruating women (median 11 cycles per patient) was used. Initial FSH levels (iFSH) were estimated as the mean value of d 1–5. The day of E1G take-off (ETO) was determined by an algorithm, and accordingly, the FP was divided into early (d 1 to ETO) and late (ETO+1 to LH peak). FP maximum and integrated E1G were calculated.
Subjects were distributed according to their mean iFSH into three categories (
5, >5 to 10, and >10 IU/liter). There was a gradual decrease in FP length with increasing category (15.2 ± 3.8, 14.1 ± 3.6, and 13 ± 2.6 d, respectively; P < 0.0001). A similar effect occurred in early FP (7.5 ± 4, 6.4 ± 3.7, and 5.4 ± 2.7; P < 0.0001); in contrast, late FP was unaffected (7.7 ± 2.1, 7.7 ± 2.1, and 7.6 ± 2.4; P = 0.86). No consistent increase in E1G was found with advancing iFSH category; however, women with mean initial LH higher than 6 IU/liter had significantly elevated maximum (P < 0.0001) and integrated (P = 0.002) E1G.
FP length decreases in parallel with increasing iFSH, with a selective effect on the early FP. Increased FSH does not affect E1G; however, elevated initial LH level was related to higher E1G.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Another early feature of reproductive aging is a reduction in the length of the cycle (15, 16, 17), which is due to the shortening of the follicular phase (FP) (4, 18, 19). During the FP, there is a process of follicle recruitment, growth, and selection as well as synthesis of estrogens. Because FSH plays a central role in this process by stimulating the granulosa cells (20), a rise in FSH might affect FP length, possibly by accelerating follicular development (4). Nevertheless, although there are several studies proposing a relationship between increased FSH as well as reduced FP length with advancing age, a direct association between increasing FSH and shortening FP has not been shown.
The FP can be divided into two distinct stages according to the onset of estrogen take-off (ETO): early FP (from d 1 to ETO) and late FP (from the day after ETO to the day of LH peak). At the early FP, follicles are recruited, and begin to grow, with very low steroidogenic activity. At the late FP, follicles (mainly the dominant one) start to produce estrogens increasingly. A reduction in FP length could in principle affect both the early and late stages. However, if FSH is involved in this reduction, because levels are higher during the early FP, this stage should be more affected than the late FP.
The second stage of the FP is characterized by the production of estrogens. Research on estrogen variation with reproductive aging has produced diverse results (1, 2, 4). Those studies, however, are based exclusively on chronological age. Because hormonal parameters appear to be better indicators of reproductive aging, examining estrogen changes in relation to increasing FSH might be more appropriate.
In this study we investigated the relationship between initial levels of FSH in the menstrual cycle and length of the FP. To examine it in further detail, we divided the FP into early and late, according to the onset of estrone 3-glucuronide (E1G) rise. In addition, we investigated possible changes in E1G production during the FP in relation to increasing FSH. As the source of information, we used a large database of daily urinary hormonal profiles over several menstrual cycles from women with menstrual regularity.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This is part of a large study on changes in reproductive hormones during aging (FREEDOM study: Fertility Recognition Enabling Early Detection Of Menopause), which started in 1998 (Miro, F., S. Parker, L. Aspinall, J. Coley, P. Perry, J. Ellis, manuscript submitted). The study involved, initially, a total of 112 women of different reproductive ages, from women with normal regular cycles to postmenopausal ones. Each woman collected daily urine samples for a variable period of time (6–18 months) and kept daily records of days of menses, medication taken, and other relevant information. The use of exogenous hormones or any medication known to interfere with the secretion of the hormonal markers investigated was an exclusion criterion. In this part of the study, we regarded only those women with regular cycles.
Data from 37 healthy women with regular menstrual cycles (mean length 24–33 d) and no abnormal bleeding were considered. The ages of the volunteers ranged from 30 to 52 yr (median value 44) with a median body mass index of 23.
Hormonal analyses
Urine samples (first morning void) were collected in universal specimen bottles containing sodium azide (0.1% volume) as preservative. Volunteers were asked to keep the samples refrigerated until delivered to the laboratory (on a weekly basis). On arrival, the specimens were aliquoted and stored at 4 C until the analyses were performed.
The samples were analyzed for LH, E1G, and pregnanediol 3-glucuronide by immunoassay, using AutoDelfia (Perkin-Elmer Life Sciences, Cambridge, UK) following in-house established and validated protocols, as previously described (20). FSH was measured using a similar procedure as used for LH. The antibody reagents used in the FSH assay were biotin-labeled anti-
FSH (clone 4882) together with strepavidin-coated microtitration strips and europium-labeled anti-ßFSH (clone 5948). The sensitivity of the FSH assay was 0.17 IU/liter, with intraassay coefficients of variation of 5.1, 2.4, and 1.8% for FSH concentrations of 1.8, 8.2, and 42.9 IU/liter, respectively. The interassay coefficients of variation at the same FSH concentrations were 4.0, 3.4, and 2.4%. The FSH assay had no cross-reactivity to LH, human chorionic gonadotrophin, and TSH (<0.1%).
Data preparation and analyses
The concentrations of FSH, LH, and E1G were adjusted to compensate for urine volume fluctuations as previously described (Miro, F., J. Coley, M. Gani, P. Perry, D. Talbot, L. Aspinall, manuscript submitted). Briefly, a smooth curve was fitted to ln(pregnanediol 3-glucuronide) and the residuals were used to adjust the concentrations of the other hormones. This was achieved by using the SAS/IML Splinec routine with a smoothing parameter of 10 (23) (Miro, F., J. Coley, M. Gani, P. Perry, D. Talbot, L. Aspinall, manuscript submitted). FSH and LH results are expressed as international units per liter and E1G as nanograms per milliliter.
Parameters, algorithms, and statistics
Initial levels of FSH and LH in the follicular phase (iFSH, iLH) were estimated as the mean concentration of FSH or LH over the first 5 d of the cycle, starting from first day of bleeding.
The onset of ETO in the FP was determined by an algorithm that retrospectively detects the first day of a sustained increase in E1G concentration. The algorithm was applied to smoothed-adjusted values of E1G throughout the menstrual cycle and works by measuring the surge size (S) for each day in the cycle having a positive slope, excluding the last 10 d of the cycle (Fig. 1
). Thus, S is based on the change in E1G slope relative to the current E1G level:
|
*, If the left slope is negative, then it is substituted by 0. **, If the E1G concentration is less 4 ng/ml, then it is substituted by 4, where right slope = the slope of the line joining the current day with the day R days later and left slope = the slope of the line joining the current day with the day L days earlier.
The value for L is fixed in the model as 2. However, R depends on the overall length of the cycle. For cycles 30 d or longer, R = 3, and it increases at a rate of 1 for each 5-d increase in cycle length up to 10, for cycle lengths longer than 60 d. This allows for longer cycles tending to have longer sustained E1G rises. Dividing the difference in the slopes by the current E1G level gives greater weight to changes in the slopes at low E1G concentrations. Setting the minimum E1G level to 4 ng/ml avoids undue sensitivity of the algorithm at very low concentrations.
The day of ETO is defined as the first day in the cycle for which S > 0.25 or S > 0.65 x maximum(S). The latter condition provides adaptation to the cycle, ensuring that S is always defined.
FP of the cycle is defined from d 1 of cycle to the day of LH peak and is divided according to the onset of ETO, into:
Early FP from d 1 of the cycle until the ETO day.
Late FP from the day after ETO to the day of the LH peak (Fig. 2).
|
The integrated E1G during the FP was calculated as the sum of adjusted E1G concentration in the smooth profiles, over the period from the day of E1G take-off and up to the day of maximum E1G. From this value the minimum E1G for each volunteer was subtracted to correct for the baseline E1G in each subject (Fig. 2
).
Differences between categories were determined using one-way ANOVA. Results are expressed as mean ± SD.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
20 ng/ml). Most of these cycles were unusually short in length (mean 19 d long), although there were two cycles longer than normal (41 and 44 d). Because initial FSH levels were considerably lower in these cycles, compared with the rest of the cycles within the same woman, it was judged that the elevated levels of estrogen affected FSH production and thus were excluded from the analyses. Overall, 386 cycles were analyzed (median 11 per volunteer). Variation in total FP length in relation to iFSH levels
To determine the precise relationship between FSH and FP length, we distributed the subjects into three categories according to their mean iFSH levels. Three categories allows a robust enough statistical analysis while providing a perception of gradual variation. The groups were established taking the 33.3 and 66.6 centiles of the distribution as guidance, which were 5.1 and 10.4, respectively.
Low: iFSH 5 IU/liter or less (mean 3.6, SD 1.5), n = 12 cases, 125 cycles. Overall, the mean number of cycles per subject with iFSH less than 5 IU/liter was 83% (SD 13.7), with 0% above 10 IU/liter.
Medium: iFSH more than 5, 10 IU/liter or less (mean 7.2, SD 3.6), n = 15 cases, 153 cycles. Overall, the mean number of cycles per subject with iFSH more than 5, 10 IU/liter or less, was 64% (SD 20), with 22% (SD 15) below, and 14% (SD 9.5) above 5 IU/liter.
High: iFSH more than 10 IU/liter (mean 11.4, SD 4.9), n = 10 cases, 108 cycles. Overall, the mean number of cycles per subject with iFSH more than 10 IU/liter was 60% (SD 15), with 37% (SD 15) more than 5, 10 IU/liter or less, and 3% (SD 5.6) less than 5 IU/liter.
The mean and range of age for each category was: 41 yr for low (30–52); 44.5 for medium (39–50); and 44 for high (30–52). These three categories correspond to three stages in reproductive aging (1, 2a, and 2b) used in a parallel study (Miro, F., S. Parker, L. Aspinall, J. Coley, P. Perry, J. Ellis, manuscript submitted).
Variation in the length of early and late FP in relation to iFSH
The length of the FP was compared among the three resulting categories of subjects with respect to overall FP, early FP, and late FP. With regard to overall FP length, we found a progressive reduction in the length from the low FSH category to high (Fig. 3) of 15.2 ± 3.8, 14.1 ± 3.6, and 13.0 ± 2.6 d, respectively (P < 0.0001).
|
To investigate this further, we determined the distribution of the length of the early and late FP in relation to each category (Fig. 3). The mean length of the early FP showed a similar trend to that of the total FP length, with lengths of 7.5 ± 4.0, 6.4 ± 3.7, and 5.4 ± 2.7 d (P < 0.0001), respectively, for low to high FSH. In contrast, there was no difference in the length of the late FP among the three categories, i.e. 7.7 ± 2.14, 7.7 ± 2.1, and 7.6 ± 2.4 d (P = 0.86).
Variation in FP estrogen production in relationship to iFSH
To investigate the relationship between iFSH and E1G production, we compared the maximum FP values of E1G as well as integrated E1G among the three categories. As shown in Table 1, there was no consistent tendency toward increasing E1G levels in parallel with FSH. Women with medium initial levels of FSH had higher levels of E1G; however, there was no statistical significance in this difference (P = 0.072). Integrated and maximum E1G was very similar in the groups of low and high FSH, whereas the medium FSH group had slightly higher values (P = 0.021).
|
). |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To better understand our finding, we examined the FP length in further detail. Because the day of ETO provides a suitable reference of the emergence of the dominant follicle (24, 25), the FP was divided into early (before) and late (after ETO). Although a reduction in FP length could in principle involve both stages, we found that this FSH-related shortening selectively affects the early FP, with no apparent effect on the late FP.
Two different explanations for the age-related reduction in follicular phase have been proposed: accelerated follicular growth (26) and earlier start of follicular growth (27). At the present, the former explanation has received more experimental support, and our results are consistent with it. Some researchers have found evidence of faster follicular development (4, 26), and even earlier achievement of dominance (28), in women of advanced reproductive age. Moreover, these authors too have found indications that the shortening of the follicular phase affects selectively events occurring early in the follicular phase (28).
Our results here, based on a large number of cases, agree with the concept that the cause of the reduction in the length of the follicular phase involved in reproductive aging is accelerated follicular growth (4, 26) and that this affects selectively early events in the FP (28). In addition, our approach detects a gradual relationship between increasing levels of FSH and reduction in FP length. Our approach too, differs in fundamental ways from that used before. We use urine instead of blood as the analytical medium, our reference for reproductive aging is the level of FSH instead of age, and our analysis of the substages of the FP is based on the onset of ETO instead of the FSH peak that follows it.
The concept of accelerated follicular growth receives further support from the finding that granulosa cells in the growing follicles from women exposed to ovarian hyperstimulation with exogenous FSH present increased mitotic index, compared with spontaneous cycles (29). Consistently, too, granulosa cells from women over 40 yr of age (which usually present elevated FSH) also have higher mitotic index, compared with younger women (30).
The steadiness in the length of the late FP found was somehow expected because FSH levels decline as estrogens rise (31). Nevertheless, early exposure to FSH might affect later aspects of follicular development, perhaps by increasing cellular numbers or a more advanced degree of cell differentiation. This, however, appears not to be the case.
Considerable evidence has shown a gradual decline in natural fertility throughout life, with a pronounced decline starting early in the fourth decade (32). This is well reflected by results from artificial insemination (33, 34). Statistical research indicates the existence of a 6-d fertile window in the cycle, during which nearly all pregnancies occur (35, 36, 37). This fertile window contains the day of ovulation and the 5 d before. Because ovulation occurs approximately 24 h after the LH peak, the fertile window thus corresponds mostly with the late FP in our study. A recent ample investigation concluded that although fertility declines with age, there is no evidence of reduction in the length of the fertile window (38). This is consistent with constancy in the late FP.
It remains to be elucidated whether the effects of higher FSH on the early events of the FP have an impact on fecundity. Inhibin B is a glycoprotein with suppressive effects on FSH secretion. Because inhibin B is produced by the newly recruited follicles from the beginning of the cycle (6, 7, 8, 9, 10, 11, 12), it is conceivable that keeping circulating levels of FSH relatively low might be desirable to ensure favorable conditions for follicular development. Some authors have suggested that the resulting increase in FSH might be a compensatory mechanism to maintain normal levels of estrogens (39). Further research is required to establish this point, and there might be important implications for the treatment of natural decline in fertility with increasing age.
Because, in addition to follicular growth, FSH stimulates estrogen production in granulosa cells, we investigated changes in estrogen production related to increased FSH. We found no consistent elevation in E1G (integrated or maximum) with increasing FSH.
Because ovarian estrogen synthesis involves two different cell types, stimulated by two gonadotropins, FSH and LH (40, 41, 42, 43), we considered a possible relationship between increasing iLH and E1G. Although we did not see a gradual rise in E1G with increasing LH, women with the highest iLH had higher values of E1G. These findings are difficult to interpret because early in the FP, granulosa cells have very little, if any, responsiveness to LH (44, 45, 46). However, a comparison between iLH with those on d 7–9 (data not shown) reveals a strong correlation (R2 = 0.54, P < 0.0001). Interestingly, in ovarian hyperstimulation, increasing the LH dose during mid-FP, with a fixed FSH dose, results in a significant increase in plasma estradiol (47, 48, 49).
Research on variation in estrogens with age has yielded conflicting results. Initial work found a decrease in estradiol (1) or excreted estrogens (50) throughout the cycle with increasing age. Other studies, although, did not find such decrease (51, 52), and indeed some even found increased circulating FP estradiol with increasing age, concomitant with a rise in FSH and LH (2). More recent studies have produced results in one or the other direction. One reported a brief decrease in preovulatory estradiol in women 36–38 yr old, followed by an increase thereafter (53). A comparison between women 40–45 and 20–25 yr old found higher serum FSH, shorter FPs, and elevated estradiol in the former (4). Interestingly, those older women had an earlier onset of ETO. Finally, there are reports of mild, although not significant, increase in FP estradiol with age (49, 54).
There is a high degree of correlation between plasma estradiol and urinary estrogens throughout the cycle, in particular E1G (55, 56, 57), and there are a few studies on age-related changes in urinary estrogens. One such study reported a substantial increase in urinary estrone during the FP in older women, who also had higher FSH (55). Other researchers, however, found an increase in excreted estradiol, estrone, and estriol with increasing age, although significant differences were found only for estradiol (58).
In summary, these results support the concept that the FP of the menstrual cycle shortens concomitantly with increasing iFSH levels. The relation between increasing FSH and reduction in FP length appears to be gradual and affects selective events occurring prior to the rise in estrogens. Also, the rise in FSH does not seem to have a significant impact on estrogen synthesis in the late FP. The finding of increased FP estrogens with relatively elevated iLH levels might explain some discrepancies reported in the literature regarding increased estrogens and aging and requires further research.
|
Footnotes |
|---|
Abbreviations: E1G, Estrone 3-glucuronide; ETO, E1G take-off; FP, follicular phase; iFSH, initial FSH level; iLH, initial LH level; S, surge size.
Received October 2, 2003.
Accepted April 5, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-hydroxylase and C17–20 lyase activity. Endocrinology 111:1201–1208[Medline]
This article has been cited by other articles:
 |
|
 |
|
  H.G. Burger, G.E. Hale, D.M. Robertson, and L. Dennerstein A review of hormonal changes during the menopausal transition: focus on findings from the Melbourne Women's Midlife Health Project Hum. Reprod. Update, November 1, 2007; 13(6): 559 - 565. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Hale, X. Zhao, C. L. Hughes, H. G. Burger, D. M. Robertson, and I. S. Fraser Endocrine Features of Menstrual Cycles in Middle and Late Reproductive Age and the Menopausal Transition Classified According to the Staging of Reproductive Aging Workshop (STRAW) Staging System J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3060 - 3067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. van der Steeg, P. Steures, M. J. C. Eijkemans, J. D. F. Habbema, P. G. A. Hompes, F. J. Broekmans, P. X. J. M. Bouckaert, P. M. M. Bossuyt, F. van der Veen, B. W. J. Mol, et al. Predictive Value and Clinical Impact of Basal Follicle-Stimulating Hormone in Subfertile, Ovulatory Women J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2163 - 2168. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Meyer, S. L. Zeger, S. D. Harlow, M. Sowers, S. Crawford, J. L. Luborsky, I. Janssen, D. S. McConnell, J. F. Randolph, and G. Weiss Characterizing Daily Urinary Hormone Profiles for Women at Midlife Using Functional Data Analysis Am. J. Epidemiol., April 15, 2007; 165(8): 936 - 945. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. MacGregor, A. Frith, J. Ellis, L. Aspinall, and A. Hackshaw Incidence of migraine relative to menstrual cycle phases of rising and falling estrogen Neurology, December 26, 2006; 67(12): 2154 - 2158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Miro and L.J. Aspinall The onset of the initial rise in follicle-stimulating hormone during the human menstrual cycle Hum. Reprod., January 1, 2005; 20(1): 96 - 100. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Miro, S. W. Parker, L. J. Aspinall, J. Coley, P. W. Perry, and J. E. Ellis Origins and Consequences of the Elongation of the Human Menstrual Cycle during the Menopausal Transition: The FREEDOM Study J. Clin. Endocrinol. Metab., October 1, 2004; 89(10): 4910 - 4915. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
