| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Twin Research Unit, St. Thomas’ Hospital, London, United Kingdom SE1 7EH
Address all correspondence and requests for reprints to: Dr. Harold Snieder, Twin Research Unit, St. Thomas’ Hospital, Lambeth Palace Road, London, United Kingdom SE1 7EH. E-mail: h.snieder{at}umds.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The reproductive life of one in five women ends prematurely in developed countries because they have a hysterectomy before they reach their natural menopause (20). Rates of surgical menopause are increasing in these countries, and it is unclear whether genetic factors are involved in the etiology of its two main indications: menorrhagia and fibroids.
These latter questions can be addressed by the analysis of data on genetically informative subjects, of which twins are the most useful. To date, no twin studies on age at menopause have been performed.
The main aim of this study was to estimate the relative importance of genetic and environmental influences on the timing of the natural menopause and on hysterectomy and its principal clinical indications by using quantitative genetic modeling of data from a large cross-sectional sample of unselected female twins. We further examined whether a genetic effect on age at natural menopause may have been mediated by a genetic effect on age at menarche. By using survival analysis approaches, we focused on potential sources of bias arising from censored observations from premenopausal twins and the possible influence of confounding by smoking, alcohol use, social class, body mass index (BMI), or use of the contraceptive pill.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The subjects were 628 twin pairs (age range, 19–76 yr)
ascertained from the general population through a national media
campaign in the United Kingdom (21). Table 1
shows the composition of the sample
with respect to menopausal status (pre/pre, pre/post, and post/post
twin pairs) and the age range of the three subgroups. Zygosity was
determined by standardized questionnaire, and DNA fingerprinting was
used for confirmation (21). Demographic and gynecological information,
including information on operations and indications for surgery, was
obtained by a standardized nurse-administered questionnaire.
|
). Menopausal age was assessed as 1) the recalled age at the
last regular period in 699 individuals and as 2) the age of starting
hormone replacement therapy (HRT) to treat menopausal symptoms in 16
subjects with a history of hysterectomy. It was not possible to
determine menopausal age with precision in an additional 282
postmenopausal subjects, of whom 242 had had a hysterectomy before
becoming menopausal, 38 had started HRT before their last regular
period, and 2 had missing data for other reasons. Background to twin analysis
Twin methodology makes use of the fact that MZ twins share identical genotypes, whereas DZ twins are no more alike genetically than siblings, sharing, on the average, 50% of their segregating genes. A higher MZ than DZ intraclass correlation (r) provides a first impression of the magnitude of genetic influence, which is based on the classic formula to estimate heritability: h2 = 2(rMZ - rDZ) (22). Model-fitting analysis of twin data, however, has some major advantages over the classic twin methodology (23, 24). It allows a more extensive separation of the observed phenotypic variance into its genetic and environmental components: additive genetic variance (A), dominance genetic variance (D), shared (or common) environmental variance (C), and specific (or unique) environmental variance (E), which also contains measurement error. A general univariate genetic model can be represented by the following linear structural equations: Pi = hAi + dDi + cCi +eEi (Eq I) and VP = h2 + d2 + c2 + e2 = VA + VD + VC + VE (Eq II), where P is the phenotype of the ith individual, scaled as a deviation from zero; A, D, C, and E can be conceived of as uncorrelated latent factors with zero mean and unit variance; h, d, c, and e are regression coefficients of the observed variable on the latent factors; and VP is the phenotypic variance. The squared regression coefficients are equal to the (unstandardized) variance components. Dividing each of these components by the total variance, VP, yields the different standardized components of variance, e.g. the heritability (h2 = VA/VP). In twin studies, the effects of D and C are confounded, which means that they cannot both be included in the same univariate model. Estimates of the genetic variation in the quantitative genetic model as applied here represent the influence of the sum of several genes on the trait (i.e. polygenic). Monogenic (or major gene) inheritance cannot be evaluated in the present twin design, even if it is present.
Analytical approach
Model-fitting procedure. Estimates of variance components
and their confidence intervals were obtained by quantitative genetic
model fitting. The significance of A, C, and D was tested by removing
them sequentially in specific submodels, eventually leading to a model
that gives the most parsimonious fit to the data, i.e. a
model in which the pattern of variances and covariance is explained by
as few parameters as possible. Submodels were compared with the full
model by hierarchic 
2 tests. The difference between
minus twice the log-likelihood (-2lnL) for a submodel and that of the
full model is approximately distributed as 2, with
degrees of freedom equal to the difference of the number of estimated
parameters in the full model and the number of estimated parameters in
the submodel.
As stated, menopausal age could not be determined with precision in a
considerable number of people (see Table 1). Standard methods of
maximum likelihood model fitting, which require complete information on
all individuals, could thus not be applied to this dataset. To derive
quantitative estimates of genetic and environmental variance components
of the age at menopause, we therefore fitted models to the raw data
using normal theory maximum likelihood (25, 26). This method allowed us
to use the information provided by unpaired observations. Besides the
complete data (post/post pairs in Table 1), data from the 185 single
individuals for whom menopausal age was known precisely for themselves
but not for their cotwin [post/pre, post/?, and post/hysterectomized
(hyst) pairs in Table 1] could be incorporated in the model
fitting.
To estimate the genetic and environmental influences on hysterectomy
and on its main indications, menorrhagia and fibroids, we used survival
analysis and genetic modelling of dichotomous (yes/no) data as
described by Neale and Cardon (23) using data from all 628 twin pairs,
i.e. regardless of their menopausal status. Akaike’s
information criterion (AIC = 2 - 2 df) was used to
evaluate the fit of the genetic models. The model with the lowest AIC
reflects the best balance between goodness of fit and parsimony.
Censorship. Some inherent aspects of the data could
potentially have biased the modelling results for age at menopause and
hysterectomy. The age at ascertainment for MZ twins was higher than
that for DZ twins (Table 2). Therefore,
cotwins of menopausal MZ twins had greater opportunity to have become
menopausal or have undergone a hysterectomy themselves at the time of
ascertainment than cotwins of menopausal DZ twins. This characteristic
of the data, which for age at menopause was confirmed by the
significant 1-yr difference in MZ compared to DZ pairs
(2[1] = 6.62; P < 0.025) and the
higher number of DZ pairs discordant for menopausal status (post/pre
pairs; Table 1), could theoretically introduce a spurious genetic
effect.
|
For the analysis of age at menopause, censored observations from two
types of twin pairs were included in addition to the complete data
(post/post pairs; Table 1). In these pairs the menopausal age was known
with precision in the postmenopausal twin, but the cotwin was either
still premenopausal (post/pre pairs; Table 1) or had a hysterectomy
before reaching menopause (post/hyst pairs; Table 1). The actual age
and the age at hysterectomy were used as censored observations for
premenopausal and hysterectomy twins, respectively.
For the analysis of hysterectomy, pairs were included if at least one twin of each pair had undergone a hysterectomy and the age at hysterectomy was known (98 MZ and 125 DZ pairs). Observations were censored in case a cotwin of a twin with hysterectomy had remained without hysterectomy until the age at ascertainment.
To investigate whether a genetic effect on age at menopause may have been restricted to either early or late menopause, menopausal cotwins were classified into early and late menopausal groups using the median age of 49 yr as a cut-off. The excess risk of becoming menopausal for MZ compared with DZ twins was subsequently compared for the younger and older menopausal groups by means of a Cox proportional hazards model.
Confounding. A possible genetic influence on age at menopause could be due to a genetic effect on age at menarche, which was measured as the recalled age at the first period. To assess the potential confounding effect of age at menarche, we used standard maximum likelihood modelling to first determine whether age at menarche was influenced by genetic factors itself.
A Cox proportional hazards model was used to further explore confounding in the data. The following covariates were included in the model: current smoking (yes/no), alcohol use (measured on a seven-point scale), age at menarche, body mass index (weight/height2), social class (I–II), and use of contraceptives (ever/never) (10, 11, 12, 13, 14, 15).
Statistical software. All model fitting was carried out with Mx (25), a software package specifically designed for the analysis of genetically informative data. The Kaplan-Meier survival analysis and Cox proportional hazards model were performed using STATA (27).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
shows general characteristics of twin pairs of which at
least one of the pair was postmenopausal. The MZ pairs were, on the
average, 3 yr older than the DZ pairs at the time of ascertainment. The
median and mean ages at menopause were 1 yr greater in MZ compared to
DZ twins. Age at menarche was similar for MZ and DZ twins (median of 13
yr), as were most of the other characteristics.
Intraclass correlations for menopausal age in the 151 MZ and 114 DZ
twin pairs for which we had complete data were 0.58 and 0.39,
respectively, suggesting an important genetic influence. This was
confirmed by the results of model fitting, which included the 185
unpaired observations (Table 3). Shared
environmental and dominance genetic effects did not contribute
significantly to the explanation of the data; they could be dropped
from the model without a significant worsening of fit. A model
specifying additive genetic (A) and unique environmental (E) variance
components gave the most parsimonious explanation of the data, yielding
a heritability estimate of 63% [95% confidence interval (CI),
0.53–0.71; Table 5].
|
|
; by Wilcoxon’s test:
2[1] = 34.56; P < 0.0001). Five years
after their cotwins had undergone the menopause, 86% of MZ twins,
compared with only 55% of DZ twins, had become menopausal
themselves.
|
Intraclass correlations for age at menarche were 0.61 and 0.18 for MZ and DZ twin pairs, respectively. The MZ correlation is more than twice the DZ correlation, which is an indication of dominance genetic effects. This was confirmed by the quantitative genetic modelling, which showed that a model that included A, D, and E showed the best fit to the data. The total contribution of genetic factors to age at menarche was estimated to be 45%, 37% due to dominance and 8% due to additive genetic effects. No correlation was observed between age at menarche and age at menopause (r = 0.032).
The results of the Cox proportional hazards model showed that the genetic effect on age at menopause remained highly significant after adjustment for a range of covariates, the MZ vs. DZ hazard ratio decreased only minimally from 1.85 (95% CI, 1.45–2.36; P < 0.001) to 1.84 (95% CI, 1.34–2.51; P < 0.001) after adjusting for current smoking, alcohol use, age at menarche, BMI, social class, and use of contraceptives in the second twin. Current smoking was the only variable that had an independent effect, bringing forward age at menopause by 1.6 yr (P < 0.001).
Genetic modelling of dichotomous data showed that hysterectomy was also
under genetic influence, with a heritability of 0.59 (95% CI,
0.43–0.72; Tables 4 and 5). The genetic effect was confirmed by
survival analysis (Fig. 2; by Wilcoxon
test: 2[1] = 5.17; P = 0.023). Our
data showed that estimates for the two main indications for
hysterectomy, fibroids and menorrhagia, were similar, with
heritabilities of 0.69 (95% CI, 0.49–0.83) and 0.55 (95% CI,
0.24–0.78) for fibroids and menorrhagia, respectively (Tables 4 and 5).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A major difficulty in this cross-sectional study was the high prevalence of hysterectomy and HRT use, which compromised obtaining precise estimates of the menopausal age. In our study a precise age at menopause was only available in 68% of the total of 1050 subjects. By fitting a model to the raw data we made optimal use of the available information, because unpaired observations could be included. The best-fitting model yielded the above-mentioned heritability estimate of 63%. Age at menopause was obtained by self-report of the age at the last regular period. Although self-report can introduce error, the alternative would have been to perform a highly impractical longitudinal twin study. Random measurement errors due to inaccuracies in this recalled age enlarge the estimate of e2 in the genetic model fitting. It can thus be argued that if recall errors had any effect on our results, it would have resulted in an underestimation of the genetic effect.
Some other characteristics of the data may still have biased the model-fitting results. The higher age at ascertainment in MZ compared to DZ twins may have given cotwins of menopausal MZ twins or cotwins of MZ twins with hysterectomy a greater opportunity to have undergone the menopause or hysterectomy themselves. This could have caused a higher concordance among MZ twins, thereby introducing a spurious genetic effect. Application of survival analysis allowed us to include censored data and thus account for the effect of imbalances in their distribution between the zygosities. Survival analysis demonstrated a genetic effect on both age at menopause and hysterectomy, thereby making a large effect of bias on the model-fitting results unlikely. The genetic effect on age at menopause was not restricted to early menopause alone, because the increased risk in MZ compared with DZ twins was significant for those with an early and a late menopause.
The exclusion from the analysis due to hysterectomy of 23% of the total number of twins could in principle have led to biased results if an association between determinants of hysterectomy and menopause had existed. However, two previous studies on determinants of age at menopause imply that individuals undergoing hysterectomy would otherwise have had a normal age at natural menopause (11, 28). Had there been an association between (the genetic effect on) hysterectomy and age at menopause, exclusion of people with hysterectomy in our analysis could only have underestimated the genetic influence on age at menopause.
Although indications for surgery are of a heterogeneous nature, the finding that hysterectomy shows a major genetic component is unlikely to be an artifact, as the genetic influences on the two main indications of hysterectomy, menorrhagia and fibroids, were of similar sizes. Both indications are of unclear etiology, but are likely to have hormonal imbalance as part of the pathogenesis (29). Although the diagnoses were based on recall, it is unlikely that this resulted in systematic bias. Confirmation of these results in a larger study with more validated clinical and pathological information could be of great interest in understanding these processes.
An early age at menarche might imply a faster depletion of the follicle reserve and thus an earlier age at menopause (12). The genetic influence on age at menopause could, therefore, have simply been the result of a genetic effect on age at menarche. The total heritability of age at menarche was estimated to be 45%. The greater part of the genetic variance was due to dominance, confirming findings from large (i.e. sufficiently powered) Australian and Finnish twin studies (2, 4). Directional dominance effects are often found for fitness traits. This explanation also has an intuitive appeal, as the age at which a female begins to menstruate will certainly affect her reproductive fitness (3). Although both onset and cessation of menstruation were under genetic control, no correlation between age at menarche and age at menopause was observed. This lack of relationship and the influence of dominance on age at menarche, but not on age at menopause, indicate that there is no overlap between the genetic and environmental mechanisms determining the two events.
In an earlier study of the same twin pairs we have shown that, besides age at menarche, other possible confounders, such as BMI, smoking, alcohol use, and social class, may show higher MZ than DZ correlations as well (30). The possibility that a genetic effect on age at menopause was due to these confounders seems unlikely, as adjustment had no influence on the excess risk to develop menopause in MZ compared to DZ twins. Only smoking had a significant independent effect. The observed advance in age at menopause of 1.6 yr was in accordance with the literature (11, 16, 17).
To our knowledge our study is the first to show that individual differences in the timing of the menopause are largely due to genetic factors. Earlier studies suggested a link between premature ovarian failure and the gene that causes galactosemia (31), and there is considerable support for X-chromosomal errors as a factor in early menopause (18, 32). Recent studies by Cramer et al. (18) and Torgerson et al. (19) showed family history to be a predictor of menopausal age, but recall bias could have affected the results.
An extensive body of literature is devoted to the question of which processes govern the onset and aging of the reproductive system. The start of menstruation has been linked to the accumulation of a critical amount of body fat (33). Findings by Kaprio et al. (4), who observed that BMI (a measure of obesity) and age at menarche had a substantial proportion of their genetic effects in common, support the idea that the genetic regulation of the onset of menarche is related to the accumulation of body fat. A recent study by Chebab et al. (34) in which prepubertal female mice were injected with leptin suggests that this hormone, secreted from adipose tissue, acts as the signal that triggers female puberty. Aging of the reproductive system, determined by the rate at which the reserve of follicles depletes, has not only been linked to factors within the ovary itself, but also to alterations in neuroendocrine activity (35, 36, 37). In a recent review of the evidence, Wise et al. (38) conclude that both the ovary and the brain are key pacemakers in determining the age at menopause. Our results suggest that these pacemakers are largely under genetic control.
From an evolutionary perspective, the genetic influence on the current distribution of menopausal ages in the population may be the result of a slow directional selection favoring earlier menopause and longer postmenopausal lifespans [the concept of the helpful grandmother (1)] and stabilizing selection against a menopause so early that it would have resulted in a reduction in the total number of offspring. Similar examples of stabilizing selection for quantitative traits resulting in high levels of genetic variability were described by Weiss (39).
The combined evidence of both the genetic model fitting and the survival analysis provided convincing evidence for the importance of genetic factors in determining the timing of age at menopause. Genes that predispose to early or late menopause may induce metabolic and endocrine changes that lead to an enhanced risk of cardiovascular disease, osteoporosis, and a number of reproductive cancers (40). Genes affecting the age of menopause may through this intermediary effect be partly responsible for the genetic effect on these diseases, which has to be considered in future studies of their genetic origins. From a practical point of view, information on early menopause in other family members may have implications in clinical decisions relating to family planning, hysterectomy, and institution of HRT.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received October 28, 1997.
Revised February 23, 1998.
Accepted March 11, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. D. Henderson, L. Bernstein, B. Henderson, L. Kolonel, and M. C. Pike Predictors of the Timing of Natural Menopause in the Multiethnic Cohort Study Am. J. Epidemiol., March 21, 2008; (2008) kwn046v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.J. Silber, M. DeRosa, J. Pineda, K. Lenahan, D. Grenia, K. Gorman, and R.G. Gosden A series of monozygotic twins discordant for ovarian failure: ovary transplantation (cortical versus microvascular) and cryopreservation Hum. Reprod., February 18, 2008; (2008) den032v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. A. Rocca, J. H. Bower, D. M. Maraganore, J. E. Ahlskog, B. R. Grossardt, M. de Andrade, and L. J. Melton III Increased risk of parkinsonism in women who underwent oophorectomy before menopause Neurology, January 15, 2008; 70(3): 200 - 209. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Rocca, J. H. Bower, D. M. Maraganore, J. E. Ahlskog, B. R. Grossardt, M. de Andrade, and L. J. Melton III Increased risk of cognitive impairment or dementia in women who underwent oophorectomy before menopause Neurology, September 11, 2007; 69(11): 1074 - 1083. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Kevenaar, A. P.N. Themmen, F. Rivadeneira, A. G. Uitterlinden, J. S.E. Laven, N. M. van Schoor, P. Lips, H. A.P. Pols, and J. A. Visser A polymorphism in the AMH type II receptor gene is associated with age at menopause in interaction with parity Hum. Reprod., September 1, 2007; 22(9): 2382 - 2388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Livshits, B. S. Kato, S. G. Wilson, and T. D. Spector Linkage of Genes to Total Lean Body Mass in Normal Women J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3171 - 3176. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Livshits, B. S Kato, G. Zhai, D. J Hart, D. Hunter, A. J MacGregor, F. M K Williams, and T. D Spector Genomewide linkage scan of hand osteoarthritis in female twin pairs showing replication of quantitative trait loci on chromosomes 2 and 19 Ann Rheum Dis, May 1, 2007; 66(5): 623 - 627. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R.G. Gosden, S.A. Treloar, N.G. Martin, L.F. Cherkas, T.D. Spector, M.J. Faddy, and S.J. Silber Prevalence of premature ovarian failure in monozygotic and dizygotic twins Hum. Reprod., February 1, 2007; 22(2): 610 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Ralston and B. de Crombrugghe Genetic regulation of bone mass and susceptibility to osteoporosis Genes & Dev., September 15, 2006; 20(18): 2492 - 2506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Bittner Menopause and Cardiovascular Risk: Cause or Consequence? J. Am. Coll. Cardiol., May 16, 2006; 47(10): 1984 - 1986. [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Stewart and C. C. Morton The genetics of uterine leiomyomata: what clinicians need to know. Obstet. Gynecol., April 1, 2006; 107(4): 917 - 921. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. S. Kok, K. M. van Asselt, Y. T. van der Schouw, P. H.M. Peeters, and C. Wijmenga Genetic studies to identify genes underlying menopausal age Hum. Reprod. Update, September 1, 2005; 11(5): 483 - 493. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.O. Okolo, C.C. Gentry, C.W. Perrett, and A.B. Maclean Familial prevalence of uterine fibroids is associated with distinct clinical and molecular features Hum. Reprod., August 1, 2005; 20(8): 2321 - 2324. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Murabito, Q. Yang, C. Fox, P. W. F. Wilson, and L. A. Cupples Heritability of Age at Natural Menopause in the Framingham Heart Study J. Clin. Endocrinol. Metab., June 1, 2005; 90(6): 3427 - 3430. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.A. Hefler, C. Grimm, G. Heinze, C. Schneeberger, M.W. Mueller, A. Muendlein, J.C. Huber, S. Leodolter, and C.B. Tempfer Estrogen-metabolizing gene polymorphisms and age at natural menopause in Caucasian women Hum. Reprod., May 1, 2005; 20(5): 1422 - 1427. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. L. Sedlmeyer, C. L. Pearce, J. A. Trueman, J. L. Butler, T. Bersaglieri, A. P. Read, P. E. Clayton, L. N. Kolonel, B. E. Henderson, J. N. Hirschhorn, et al. Determination of Sequence Variation and Haplotype Structure for the Gonadotropin-Releasing Hormone (GnRH) and GnRH Receptor Genes: Investigation of Role in Pubertal Timing J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 1091 - 1099. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Kok, N. C. Onland-Moret, K. M. van Asselt, C. H. van Gils, Y. T. van der Schouw, D. E. Grobbee, and P. H.M. Peeters No association of estrogen receptor {alpha} and cytochrome P450c17{alpha} polymorphisms with age at menopause in a Dutch cohort Hum. Reprod., February 1, 2005; 20(2): 536 - 542. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. K. Jacobsen, I. Heuch, and G. Kvale Age at Natural Menopause and Stroke Mortality: Cohort Study With 3561 Stroke Deaths During 37-Year Follow-Up Stroke, July 1, 2004; 35(7): 1548 - 1551. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Hunter, D. Hart, H. Snieder, P. Bettica, R. Swaminathan, and T. D. Spector Evidence of altered bone turnover, vitamin D and calcium regulation with knee osteoarthritis in female twins Rheumatology, November 1, 2003; 42(11): 1311 - 1316. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Lobo Early ovarian ageing: a hypothesis: What is early ovarian ageing? Hum. Reprod., September 1, 2003; 18(9): 1762 - 1764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Nikolaou and A. Templeton Early ovarian ageing: a hypothesis: Detection and clinical relevance Hum. Reprod., June 1, 2003; 18(6): 1137 - 1139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Horiuchi, C. E. Finch, F. Mesle, and J. Vallin Differential Patterns of Age-Related Mortality Increase in Middle Age and Old Age J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2003; 58(6): B495 - 507. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. K. Jacobsen, I. Heuch, and G. Kvale Age at Natural Menopause and All-Cause Mortality: A 37-Year Follow-up of 19,731 Norwegian Women Am. J. Epidemiol., May 15, 2003; 157(10): 923 - 929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Kok, K. M.v. Asselt, Y. T.v. d. Schouw, D. E. Grobbee, E. R.t. Velde, P. L. Pearson, and P. H.M. Peeters Subfertility reflects accelerated ovarian ageing Hum. Reprod., March 1, 2003; 18(3): 644 - 648. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Gorai, K. Tanaka, M. Inada, H. Morinaga, Y. Uchiyama, R. Kikuchi, O. Chaki, and F. Hirahara Estrogen-Metabolizing Gene Polymorphisms, But Not Estrogen Receptor-{alpha} Gene Polymorphisms, Are Associated with the Onset of Menarche in Healthy Postmenopausal Japanese Women J. Clin. Endocrinol. Metab., February 1, 2003; 88(2): 799 - 803. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
