1 International Health Foundation, Utrecht, 3526 KS, 2 Julius Center for Patient Oriented Research, and 3 Department of Obstetrics and Gynaecology, University Medical Centre, Utrecht, 3508 GA, The Netherlands
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: cohort study/follicular depletion/menopause/oral contraceptive/risk factors
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The few studies that included the effect of OC use in their design found either a delaying effect of OC use on age at menopause (Van Keep et al.1979; Stanford et al., 1987
) or no effect (Brambilla and McKinlay, 1989
; Cramer et al.1995b
; Bromberger et al.1997
; Van Noord et al.1997
). However, most studies failed to exclude women who used hormone replacement therapy (HRT) or OCs for menopausal complaints, leading to a potentially spurious positive association between OC use and menopausal age.
Studies addressing the relationship between the use of high dose OCs and menopausal age are lacking. High dose OCs strongly suppress FSH concentrations because of their high doses of oestrogens and progestins. Other types of OCs contain lower doses of oestrogens and progestins, suppressing FSH in a more moderate way. Therefore, particularly in the case of high dose OC use, one can expect a delay in menopausal age. If it can be confirmed that use of high dose OCs delays age at menopause it is also likely that use of high dose OCs delays the end of fertility. If lower dose OCs also have a delaying effect on menopause, this could be a promising perspective for women who postpone their first pregnancy until their thirties by means of OC use.
In this cohort study, we investigated the effect of OC use on menopausal age. We were able to use a cohort of a large number of women (n = 8701), 3242 of whom reported that they used high dose OCs.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Women were defined as being post-menopausal if their menses had been absent for at least 12 months due to factors other than pregnancy or lactation; all of the other women were defined as pre-menopausal. Women who were still considered pre-menopausal at their initial screening visit were sent a postal questionnaire in 1986, comprising questions about the date of their previous menstrual period and the use of HRT or OCs for menopausal symptoms during the past 4 years. If women had undergone a medical treatment that caused cessation of menstruation, they were asked what kind of treatment this was. They were also asked to keep their menstrual data on a calendar sent along with the questionnaire. Women who were still pre-menopausal were sent a similar questionnaire in 1987, 1989, 1992, 1993 and 1995. Response was about 80% each year. Both women who responded and those who did not respond in early 1986 were sent a questionnaire in late 1987. But women not responding in 1987, 1989 and 1991 were not sent a questionnaire in 1989, 1991 and 1993 respectively. We made special efforts to complete the follow-up for a random sub-sample of 4207 women in 1992. We found the new addresses of women who had moved via municipal administrations. All women who had failed to respond to one of the previous questionnaires were sent a questionnaire again and those still not responding were approached by telephone. After these efforts follow-up for this subsample was 90% complete. For the total cohort the follow-up is 82% complete.
Menopausal age
The date of menopause was the last date of menses, defined retrospectively following 12 months of amenorrhoea. Age at menopause was computed by subtracting the date of birth from the date of the onset of menopause. For women who were post-menopausal at their screening visit (n = 1076), the exact date of menopause was not known. The questionnaire filled out at screening provided data on menopausal age (in years), not the exact date (day, month, year) of their last menstruation. Therefore, only truncated menopausal ages are known for these women. We decided to truncate the menopausal ages of all the women to standardize them.
OC use
OC use is characterized as `ever use of OCs', `duration of use', and `duration of use of high dose OCs'. Ever use was defined as OC use for at least 3 months. Duration of OC use was defined as the sum of years in which a woman used OCs for at least 3 months. In The Netherlands, the first tablets containing oestrogens and progestins (lynestrenol) were introduced in 1962, officially only for treatment of disturbances of the menstrual cycle. Based on market information it was estimated that in the Utrecht region prior to 1972, high dose (>50 µg oestrogen) OCs were predominantly prescribed (Haspels and Kay, 1982). Duration of high dose OC use was defined as the sum of years before 1972 in which a woman used OCs for at least 3 months (van Hoften et al.2000
).
Exclusion and inclusion criteria
Subjects who used HRT or OCs for menopausal symptoms during the 4 years prior to the date of last menstruation or the first year following the last menstruation (n = 3477) were excluded from the analyses, because these women can have withdrawal bleedings caused by the use of OCs or HRT, obscuring the date of the menopause.
Of the women included in the analyses (n = 8701), age at natural menopause was known in 4523 cases (1076 of whom were post-menopausal at their initial screening visit), 4178 women were censored at the age of their last known menstruation because they did not experience a natural menopause. Of these censored women, 2466 women had undergone surgery that caused cessation of menstruation, 1547 were still pre-menopausal when they were lost to follow-up and 165 women were pre-menopausal in the questionnaire from 1995.
Confounders
Smoking, parity, body mass index (BMI), socio-economic status and possibly age at menarche are related to menopausal age, and were therefore considered as potential confounding variables. Current smoking was defined as having smoked cigarettes regularly during the past year and past smoking was defined as having smoked more than a year ago. Socio-economic status is based on the type of health insurance, whether public health insurance, which is government mandated (lower status), civil servants insurance (intermediate status) or private insurance (higher status). Trained medical assistants measured height and weight at the screening. BMI was calculated using the formula [weight (kg)/height (m)2]. The year of birth was also included in the analyses to check for any discrepancies in variable characteristics among the different age groups.
Statistical analyses
KaplanMeier estimates were used to compare failure time distributions for the categories in each risk factor. Log-rank tests were carried out to test the differences in failure time distributions. The endpoint of the analyses was natural menopause (failure). Cessation of menstruation due to other causes and loss to follow-up were censoring events. The biological age of the participating women was on the time axis.
To obtain estimates of the hazard ratio (HR) for each variable and for each variable adjusted for all potential confounders we used Cox's proportional hazard models. These models predict the ratio of incidence rates as the dependent variable: the hazard ratio. This ratio is assumed to be constant over time (proportional hazards) (Rothman, 1986). This assumption of proportional hazards was checked for each variable by making log-minus-log plots. The assumption was found to be justified for all variables.
An HR <1 indicates a decreased risk for an event. In our analyses, this means a decreased risk of entering menopause, meaning a decreased risk of menopause at any given age. An HR >1 indicates an increased risk of entering menopause. Tests for trend were calculated by polynomial contrasts.
To further quantify the effect of OC use on menopausal age, linear regression analyses were performed on the group of women with a natural menopause (n = 4523).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Tests for trend showed the existence of a significant linear trend for years of high dose OC use (P < 0.005) and total years of OC use (P < 0.005) with menopausal age.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It is possible that the effect of high dose OCs is due to the fact that in The Netherlands, from 19621969, OCs were officially only registered to treat disturbances of the menstrual cycle and not for contraceptive purposes. Therefore, it is possible that women who used high dose OCs were predominantly women with signs of early ovarian ageing (irregularity of cycles). These women were expected to enter menopause at an earlier age than women without signs of early ovarian ageing. However, as women who used OCs in the 4 years prior to menopause were excluded from our analyses, we find it unlikely that such a selection will have substantially influenced our results. In addition, it is very likely that physicians often prescribed these OCs for contraceptive purposes, despite the fact that this was officially prohibited until 1969. The population-based design of our study is an advantage, as we did not examine a selected group of women. We were able to collect data from a large number of women (n = 8701), making it easier to find subtle effects of potential risk factors on menopause.
Our cut-off point for high dose OC use is somewhat arbitrary. Before 1972, lower dose OCs were already being prescribed and after 1972, some women were still using high dose OCs. Different manufacturers introduced new types of OCs throughout the years and used different doses and types of oestrogens and progestins in their products. The possibility of mis-classification of women using high dose OCs is present. We were obliged to use this cut-off point because the individual data about the type of OC women used is not known. By using the market information we were able to estimate the effect of long-term use of high dose OCs, despite the fact that the individual data was missing.
In this study, the date of menopause for most women was determined prospectively (n = 3447). Relatively few women were already menopausal at their screening visit (n = 1076). For these women age at menopause was determined retrospectively. However, the date of their last menstruation was relatively recent, reducing the possibility of recall errors.
We confirmed all known effects for the other variables included in our analyses (Sowers and La Pietra, 1995; Kato et al.1998
; Hardy and Kuh, 1999
). This point is reassuring regarding the validity of our study. Although we adjusted for possible confounders, the possibility of residual confounding by smoking and socio-economic status cannot be excluded.
Our main finding is that long-term users of high dose OCs (>50 µg oestrogen for 3 or more years) experience menopause at a slightly younger age compared with women who did not use OCs or only used lower dose OCs (50 µg oestrogen). Use of lower dose OCs has no influence on age at menopause.
Eleven or more years of total OC use advanced menopausal age. However, this is probably due to the effect of long-term use of high dose OCs. Our results are in contradiction with the results of most other studies investigating the influence of OCs on menopausal age. These studies either found no effect on menopausal age (Brambilla and McKinlay, 1989; Cramer et al.1995b
; Bromberger et al.1997
; Van Noord et al.1997
) or a postponement of menopause with OC use (Van Keep et al.1979
; Stanford et al.1987
). An investigation into the influence of OC use on peri-menopausal age, indicated the existence of a (nonsignificant) trend for earlier inception of peri-menopause with long-term OC use (Hardy and Kuh, 1999
). This result is consistent with our findings.
The majority of studies, with the exception of just two (Bromberger et al.1997; Hardy and Kuh, 1999
), did not exclude women who used HRT or OC for menopausal symptoms from their analysis. In our study we have excluded this group of women, because dates of natural menopause can be obscured by bleedings caused by OC use or use of HRT.
None of the studies mentioned made a distinction between different types of OCs or took into account a dose-response relation. Some studies did not adjust for other variables influencing menopausal age, such as smoking and parity. Our univariate results also showed no effect of OC use; the effect of OCs became apparent only after adjustment for other variables.
As in previous reports, parity and a high BMI increased age at menopause, whilst smoking decreased age at menopause (Cramer et al.1995a, b
; Sowers and La Pietra, 1995
; Kato et al.1998
; Nagata et al.1998
; Hardy and Kuh, 1999
). We have found a decreased risk of menopause in the cohorts of younger women. This is probably due to an age-related artefact: these cohorts simply have not yet had the same chance of being menopausal as the cohort of older women in this study.
The predominant role of FSH in the late stages of follicular development is extensively investigated and generally accepted (Gougeon, 1996). In contrast, the reason why follicles start to grow after many years of dormancy is still a mystery. Experimental findings and observations clearly show that follicles are able to grow without the presence of gonadotrophins (Kumar et al.1997
; Dierich et al.1998
; Parrott and Skinner, 1999
). It is understandable therefore, that follicle growth until the selectable stage was considered to be a gonadotrophin-independent process (Wandji et al.1997
). However, there is growing evidence indicating that in complex cooperation with other factors, FSH facilitates the initiation of follicular growth and stimulates early follicular development (te Velde et al., 1998b
). Moreover, there is strong evidence that the actual resting follicle pool does not only contain primordial follicles but also intermediate and primary follicles (de Bruin et al., 2001
). Only thereafter does transition to the rapid growth phase commence. The fact that FSH receptors have been demonstrated in the granulosa cells of primary follicles (Oktay et al.1997
) is in agreement with the possible role of FSH during the transition from the resting follicle pool to the growth phase.
Our investigation was based on the hypothesis that menopause can be delayed by sparing resting follicles. If FSH stimulates the initiation of growth of these resting follicles, then this would mean that very low concentrations or the absence of FSH would inhibit this initiation and thereby delay menopause. This hypothesis is partly confirmed by studies in which it was observed that FSH-suppressing factors, such as parity and irregular or long menstrual cycles, can delay age at menopause (Cramer et al.1995b; Sowers and La Pietra, 1995
; Van Noord et al.1997
; Kato et al.1998
; Hardy and Kuh, 1999
).
OC use suppresses FSH concentrations, therefore it was hypothesized that long-term OC use, in particular long-term high dose OC use, would postpone menopause to a later age. The results of our analyses contradict this hypothesis: long-term use of high dose OCs does not delay menopause, it actually accelerates the onset of menopause. Our finding may be explained by the findings of other studies (Gougeon, 1996) stating that the absence of FSH increases levels of oxidative free radicals in granulosa cells. This in turn causes the activation of endonucleases, thereby initiating apoptosis of follicles. Our findings may also be explained by the fact that exogenous progesterone can have a damaging effect. It was found that mice who had been exposed to exogenous progesterone showed increased atresia of antral follicles (Telfer et al.1991
). In addition, a study on the methods of initiation of primordial follicle development in rats in vitro (Parrott and Skinner, 1999
) showed that primordial follicles develop into primary follicles in medium with no factor added. Treatment of ovaries with gonadotrophins inhibits this spontaneous development of primordial follicles into primary follicles. If this is also the case in vivo in humans, OC use could accelerate the depletion of the follicle pool by lowering gonadotrophin concentrations, thereby allowing the spontaneous development of primordial follicles into later stages of follicle development. The fact that we did not observe any effect on menopausal age in lower dose OCs might be explained by the fact that lower doses of oestrogens and progestins suppress FSH concentrations in a more moderate way (Copeland, 1993
).
In conclusion, the effect of OC use found in this study is in contradiction with the hypothesis that long-term OC use delays age at menopause. Our results suggest that long-term use of high dose OCs accelerates menopause and that use of lower dose OCs has no effect on age at menopause, although possible bias by residual confounding cannot be excluded.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Brambilla, D.J. and McKinlay, S.M. (1989) A prospective study of factors affecting age at menopause [published erratum appears in J. Clin. Epidemiol. 1990, 43, 537]. J. Clin. Epidemiol., 42, 10311039.[ISI][Medline]
Bromberger, J.T., Matthews, K.A., Kuller, L.H. et al. (1997) Prospective study of the determinants of age at menopause. Am. J. Epidemiol., 145, 124133.[Abstract]
de Bruin, J.P., Nikkels, P.G.J., Bruinsen, H.W. et al. (2001) Morphometry of human ovaries in normal and growth restricted foetuses. Early Hum. Dev. 60, 179192.[ISI][Medline]
Copeland, L.J. (1993) Textbook of Gynecology. W.B.Saunders Company, Philadelphia, p 171
Cramer, D.W., Harlow, B.L., Xu, H. et al. (1995a) Cross-sectional and case-controlled analyses of the association between smoking and early menopause. Maturitas, 22, 7987.[ISI][Medline]
Cramer, D.W., Xu, H. and Harlow, B.L. (1995b) Does `incessant' ovulation increase risk for early menopause? Am. J. Obstet. Gynecol., 172, 568573.[ISI][Medline]
Dierich, A., Sairam, M.R., Monaco, L. et al. (1998) Impairing follicle-stimulating hormone (FSH) signaling in vivo: Targeted disruption of the FSH receptor leads to abberant gametogenesis and hormonal imbalance. Proc Natl Acad. Sci. USA., 95, 1361213617.
Ford, J.H. and MacCormac, L. (1995) Pregnancy and lifestyle study: the long-term use of the contraceptive pill and the risk of age-related miscarriage. Hum. Reprod., 10, 13971402.[Abstract]
Gougeon, A. (1996) Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr. Rev., 17, 121155.[ISI][Medline]
Hardy, R. and Kuh, D. (1999) Reproductive characteristics and the age at inception of the perimenopause in a British National Cohort. Am. J. Epidemiol., 149, 612620.[Abstract]
Haspels, A.A. and Kay, C.R. (1982) Veiligheidsaspecten bij het pilrecept. In Haspels, A.A. and Kay, C.R. (eds) Amsterdam, Excerpta Medica. Conference Proceeding. pp 59.
Kato, I., Toniolo, P., Akhmedkhanov, A. et al. (1998) Prospective study of factors influencing the onset of natural menopause. J. Clin. Epidemiol., 51, 12711276.[ISI][Medline]
Kumar, T.R., Wang, Y., Lu, N. and Matzuk, M.M. (1997) Follicle stimulating hormone is required for ovarian follicle maturation but not male fertility. Nature Genet., 15, 201204.[ISI][Medline]
Nagata, C., Takatsuka, N., Inaba, S. et al. (1998) Association of diet and other lifestyle with onset of menopause in Japanese women. Maturitas, 29, 105113.[ISI][Medline]
Oktay, K., Briggs, O. and Gosden, R.G. (1997) Ontogeny of follicle-stimulating hormone receptor gene expression in isolated human ovarian follicles. J. Clin. Endocrinol. Metab., 82, 37483751.
Parrott, J.A. and Skinner, M.K. (1999) Kit-Ligand/Stem Cell Factor induces primordial follicle development and initiates folliculogenesis. Endocrinology, 140, 42624271.
Richardson, S.J. (1993) The biological basis of the menopause. Baillières. Clin. Endocrinol. Metab., 7, 116.[ISI][Medline]
Rothman, K.J. (1986) Modern Epidemiology 7th edn, Little, Brown and Company, Boston. pp. 295296.
Sowers, M.R. and La Pietra, M.T. (1995) Menopause: its epidemiology and potential association with chronic diseases. Epidemiol. Rev., 17, 287302.[ISI][Medline]
Stanford, J.L., Hartge, P., Brinton, L.A. et al. (1987) Factors influencing the age at natural menopause. J. Chronic. Dis., 40, 9951002.[ISI][Medline]
Telfer, E., Gosden, R.G. and Faddy, M.J. (1991) Impact of exogenous progesterone on ovarian follicular dynamics and function in mice. J. Reprod. Fertil., 93, 263269.[Abstract]
te Velde, E.R., Dorland, M. and Broekmans, F.J. (1998a) Age at menopause as a marker of reproductive aging. Maturitas, 30, 119125.[ISI][Medline]
te Velde, E.R., Scheffer, G.J., Dorland, M. et al. (1998b) Developmental and endocrine aspects of normal ovarian aging. Mol. Cell. Endocrinology, 145, 6773.[ISI][Medline]
van Hoften, C., Burger, H., Peeters, P.H.M. et al. (2000) Long term oral contraceptive use increases breast cancer risk in women over 55 years of age: The Dom cohort. Int. J. Cancer, 87, 591594.[ISI][Medline]
Van Keep, P.A., Brand, P.C. and Lehert, P. (1979) Factors affecting the age at menopause. J. Biosoc. Sci. Suppl., 3755.
Van Noord, P.A.H., Dubas, J.S., Dorland, M. et al. (1997) Age at natural menopause in a population-based screening cohort: the role of menarche, fecundity, and lifestyle factors. Fertil. Steril., 68, 95102.[ISI][Medline]
Waard de, F., Collette, H.J.A., Rombach, J.J. et al. (1984) The Dom Project for the Early Detection of Breast Cancer, Utrecht, The Netherlands. J. Chronic. Dis., 37, 144.[ISI][Medline]
Wandji, S.-A., SraeH, V., Nathanielsz, P.W. et al. (1997) Initiation of growth of baboon primordial follicles in vitro. Hum. Reprod., 12, 19932001.[Abstract]
Submitted on November 6, 2000; accepted on April 26, 2001.