1 Epidemiology of Developmental Brain Disorders Department, New York State Psychiatric Institute, New York, NY 10032, 2 Gertrude H.Sergievsky Center, Columbia University, New York, NY 10032, 3 Mailman School of Public Health, Columbia University, New York, NY 10032, 4 Research Foundation for Mental Hygiene, New York State Psychiatric Institute, New York, NY and Graduate School of Arts and Sciences, Columbia University, New York, NY 10032, 5 Research Foundation, Bellevue Womans Hospital, Niskayuna, NY 12309, 6 Department of Obstetrics and Gynecology, Columbia University, New York, NY 10032 and 7 Clinical Genetics and Development, Department of Pediatrics, Columbia University, New York, NY 10032, USA
8 To whom correspondence should be addressed at: Psychiatric Institute, Epidemiology, 722 West 168th Street, Room 1607, New York, NY 10032, USA. e-mail: jkk3{at}columbia.edu
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: epidemiology/FSH/inhibin B/oocyte/trisomy
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It seems reasonable to infer that trisomy results from physiological processes that are intimately tied to chronological age. One candidate process is the age-related decline in the size of the ovarian oocyte pool (oocyte atresia). We hypothesized that the association between chronological age and trisomy reflects a causal connection between the size of the oocyte pool (ovarian age) and trisomy, with risk higher in women with fewer oocytes (Warburton, 1989; Kline and Levin, 1992
). Studies of autopsy and surgical specimens show that the germ cell population, which is largest during fetal development (
7x106), decreases approximately exponentially with chronological age (Block, 1952
; Baker, 1963
; Thomford et al., 1987
; Faddy et al., 1992
; Leidy et al., 1998
). Despite the overall decline in oocyte count with age, counts vary among women of any given chronological age. Variability may reflect differences either in the number of oocytes laid down or in the rate of atresia, each potentially influenced by endogenous or exogenous factors.
Several observations support the idea that the size of the oocyte pool may influence trisomy risk. (i) The rate of aneuploidy is higher in middle-aged mice in which one ovary was removed than in sham-operated mice (although the majority of anomalies are monosomy, not trisomy) (Brook et al., 1984). In women, observations relating oophorectomy to trisomy are inconsistent (Freeman et al., 2000
; Warburton and Kline, 2001
); inferences from both studies are limited by the small number of women with oophorectomies. (ii) Age at menopause may occur earlier in women who had trisomic spontaneous abortions than in women who had chromosomally normal pregnancies (Kline et al., 2000
). (iii) One study (van Montfrans et al., 1999
) suggests that elevated levels of FSHwhich might reflect either diminution of the underlying oocyte pool or impaired function of the developing cohortare more common in women after a Downs syndrome birth than in controls. Another study (Nasseri et al., 1999
), drawing mainly on a sample of women pregnant by assisted reproductive technology, suggests that raised levels of either FSH or estradiol are more common in women with chromosomally abnormal (80% of which were trisomic) versus chromosomally normal spontaneous abortions; the authors do not report associations for the two hormones separately.
An association might arise through various routes. The biological ageing hypothesis focuses on proximity to menopause, postulating that a factor associated with depletion of the oocyte pool, such as hormonal imbalance, influences the likelihood of abnormal chromosomal segregation at ovulation (Brook et al., 1984). The limited oocyte pool hypothesis focuses on the small cohort of follicles that develop during each menstrual cycle (Warburton, 1989
). Like the total pool, the number of developing follicles diminishes with chronological age (Block, 1952
; Gougeon et al., 1994
; Reuss et al., 1996
; Scheffer et al., 1999
). The hypothesis proposes that the chance that an oocyte is available at the optimal stage of maturation decreases as the number of developing follicles decreases; immature or post-mature oocytes are hypothesized to be more liable to abnormal chromosome segregation.
Our study compares indicators of ovarian age among women who had trisomic losses with those among women who had other losses (chromosomally normal or non-trisomic chromosomally abnormal) or with chromosomally normal births. The three primary indicators of advanced ovarian age are low number of antral follicles, low level of dimeric inhibin B, and high level of FSH. The number of antral follicles provides an indirect measure of the size of the underlying oocyte pool. Inhibin B, a product of the cohort of developing follicles (Burger, 2000), reflects their quantity and quality. FSH, a gonadotrophin under negative feedback of two ovarian hormones, estradiol and inhibin (Burger, 1994, 2000), reflects the competence of the developing pool; FSH may also provide an indirect measure of the underlying oocyte pool (Goldenberg et al., 1973
; Klein et al., 2002
). We measured estradiol because some reports suggest that in older regularly cycling women the dominant follicle emerges early, with a consequent early rise in estradiol level (Klein et al., 1996a
, 2000).
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
For each woman with a trisomic loss (case) who completed the study, we selected an age-matched control with a chromosomally and anatomically normal live birth 1800 g, no pregnancy loss since the index pregnancy and no known trisomic pregnancy. They were selected from the hospital delivery log of women who delivered during the 713 months preceding the date of selection. Live birth controls were matched to trisomy cases for projected age (± 6 months) at the sonography visit. If a selected control was ineligible for the study or refused to participate, we replaced her. The study was approved by our Institutional Review Boards and all participants gave informed consent. Fieldwork ended in November 2001.
The protocol for live birth controls was identical to the protocol for women with prefetal losses. The interviewer knew the outcome of the index pregnancy, but she did not know the karyotype of the loss (except in about two instances when the subject revealed it). Follicles were counted and hormones assayed without knowledge of any subject characteristics, including the outcome of the index pregnancy.
To obtain valid ovarian age measures, we required: no pituitary disorder or hormonal disorder related to ovarian function, no oophorectomy, no hormonal medication, no pregnancy at the time of ultrasound, no breastfeeding or breastfeeding no more than once per day during the menstrual cycle preceding the study assessments. We required that any diagnosis be current, the report of the diagnostic work-up informative and the clinical symptoms and treatment consistent with the diagnosis. The study reproductive endocrinologist (A.C.K.) reviewed the interview data to determine whether or not a potential participant currently had a condition associated with altered hormone levels.
Participants
Spontaneous abortions
From 435 prefetal losses identified, we set up 269 specimens in culture and karyotyped 244 (Table I). The principal reasons for not setting up a tissue culture were patient refusal of the karyotype offer (17%) and insufficient fetal tissue (15%). Among the 244 women with karyotyped losses, 127 (52%) completed the protocol. The principal reasons for not completing the protocol were: refusal (23%) and ineligibility (25%), primarily due to use of hormonal contraceptives (10%) or pregnancy soon after the index loss (7%). Six women (2%) were excluded because of hormonal conditions and another six were excluded due to use of fertility drugs, although only two had experienced conception delay >1 year prior to the study pregnancy.
|
Analyses exclude 24 women for whom we consider the karyotype of the abortus uncertain: the abortus was karyotyped as a chromosomally normal female, but we could not confirm the validity of the diagnosis by comparison of maternal and fetal DNA (Table I, footnote i). Analyses also exclude repeat study pregnancies to four women who had a trisomic loss on either their first or second pregnancy (Table I, footnote h). The analytical sample thus includes 99 women: 54 with trisomic or hypertriploid losses (cases), 24 with other chromosomally abnormal losses and 21 with chromosomally normal losses (17 male, four female).
Live birth controls
During the first 6 months of the study, we selected live birth controls before we knew the eligibility and consent status of the matched case; as a result, four controls completed the protocol even though their matched cases did not. For most of the study, controls were selected after the case had agreed to the protocol. We selected second or later controls when: (i) a control declined the study or was ineligible or (ii) >2 months elapsed without our having spoken to the first control or (iii) we were unable to scan both ovaries of the first control. During the final 3 months of the study, to speed completion of the fieldwork, we selected additional controls before 2
months had passed. In total, we selected 219 controls; 65 (30%) completed the protocol. The principal reasons for not completing the protocol were refusal (31%) and ineligibility (37%), primarily due to use of hormonal contraceptives (17%) or breastfeeding (10%). Two women (1%) were excluded because of hormonal conditions.
Women who completed the protocol were on average older, though not significantly so, than women who did not. Among the 144 women who completed the eligibility interview, the odds of completing the protocol did not differ with educational attainment, parity, number of prior induced abortions or number of prior spontaneous abortions.
The analytical sample comprises 65 women who completed the protocol. Sixty of the 65 were matched at the time of selection to one of the 54 trisomy cases. An additional five live birth controls were matched post hoc to trisomy cases of the same age.
Characteristics of women with losses, classified by karyotype, and live births
Maternal age was similar for women with trisomic losses and live birth controls; women with non-trisomic chromosomally abnormal losses and women with chromosomally normal losses were younger (Table II). Adjusting for age, none of the demographic or exposure variables (education, ethnicity, obstetric history or smoking) were related to the outcome of the index pregnancy. The majority of women completed the blood and sonography protocols after the second or third menstrual period following the index loss or, for live birth controls, the introductory letter. Protocol menstrual period, day of blood sampling and day of sonography were unrelated to the outcome of the index pregnancy.
|
We videotaped the sonography scan. The sonographer (M.L.R.) identified the optimal scan, converted it to a digitized format and exported it to Matrox Inspector software for follicle counting and measurement. The scans were counted in four batches, each batch including trisomies, other losses and live births. Within each batch, we randomly allocated scans of left and right ovaries and randomly ordered scans. The randomization procedure was unknown to the sonographer. We adjusted for batch number in our initial analyses, but not in the final analyses (see below).
Each sonolucency interpreted as a follicle was followed through the scan to identify its maximum size; its largest diameter was measured in the vertical plane using the centimetre scale generated by the ultrasound machine. To assess the reliability of the counting procedure, we counted again the follicles in 40 ovaries (from seven women with trisomic losses, six with other losses, seven with births). The intraclass correlation coefficient (ICC) was 0.92. We evaluated reliability early in the study (after 12 scans were completed) and at the end. The ICC of the first 12 scans (0.83) was excellent (Fleiss, 1986); the ICC of the two later reliability samples (0.97 for the second 12 scans, 0.92 for the final 16 scans) were even better. Analyses use the second count for the 20 women whose scans were counted twice.
Among the 155 women with both ovaries scanned, counts ranged from 2 to 70 (median 15, mean 18.5, SD 13.0). The mean difference between counts in the left and right ovaries was 0.34 (SD = 5.52, range 15 to 28). The correlation between them was 0.70 (P < 0.0001). For the nine women with one ovary scanned, we imputed the total number of follicles in the ovary not scanned from regression equations that included age and the number of follicles in the contralateral ovary. Results were unchanged when we excluded women with imputed counts from the analysis.
Serum hormone levels
Blood samples were processed in a refrigerated centrifuge and, after separation, sera were frozen at 25°C at the study hospital; they were shipped to New York City and stored at 20°C. Dimeric inhibin B was measured by radioimmunoassay (Oxford Bio-Innovation Ltd); FSH and estradiol were measured by solid-phase chemiluminescent enzyme immunoassays (Immulite; Diagnostic Products Co.). For inhibin B, sensitivity was 20 pg/ml; intra- and inter-assay coefficients of variation (CV) were 5.1% and 6.2% respectively. For FSH, sensitivity (the minimum detection limit) was 0.1 mIU/ml; intra- and inter-assay CV were 9.3% and 10.5% respectively. For estradiol, sensitivity was 20 pg/ml; intra- and inter-assay CV were 1.9% and 5% respectively.
Table III shows summary statistics for antral follicle count and hormone levels and their correlations with chronological age. Correlations with count, inhibin B and FSH are in the expected directions. For estradiol, the modest positive correlation is compatible with observations that levels may be elevated during the menopausal transition (Klein et al., 1996b, 2000).
|
Primary analysis
We first examined associations with each of the ovarian age indicators separately. We used stratified multiple linear regression models to test the null hypothesis that, at any maternal age, there is no difference between trisomy cases and any of the three comparison groups in antral follicle count or hormone measures. We estimated the magnitude of differences between trisomy cases and each comparison group and set 95% CI around these. We used a logarithmic transformation for count of antral follicles, ln(1 + count), and each hormone to meet the normal error assumption of least squares regression. Thus, differences correspond approximately to percentage changes in the hormone measures and to percentage changes in 1 + antral follicle count. We checked the results in a complementary analysis using polytomous conditional (matched sample) logistic regression (Breslow and Day, 1980; Levin, 1987
, 1990); for these analyses, we classified each of the ovarian age measures in three categoriesvalues in the lowest 20th percentile, values in the upper 20th percentile and, as the reference group, values between the 20th and 80th percentiles (not reported).
Secondary analyses
We explored whether the strength of associations between trisomy and ovarian age measures varies with chronological age or with trisomy type (trisomy 16, other non-acrocentric trisomies, acrocentric trisomies). To address the possibility that associations with trisomy might be attenuated by the presence of women with prior unkaryotyped trisomies (expected to be 3.8%) among controls, we repeated the primary analyses using only the largest comparison group: live birth controls.
We also estimated associations with two indicators of older ovarian age that combined our primary measures. The first combination, a priori, defined older ovarian age as: follicle count 6 (the lower 12th percentile in our data) or inhibin B
20 pg/ml (the sensitivity level of the assay) or FSH
10 mIU/ml (a level, for the Immulite assay, many clinicians use to indicate poor ovarian reserve). For the second indicator, we summed the z-scores for our three primary measures (count, inhibin B, FSH), reversing the sign for count and inhibin so that higher positive z-scores indicated older ovarian age. We defined older ovarian age as summed z-scores in the upper 14th percentile, corresponding to one or more SD from the mean of the summed z-score. We used conditional maximum likelihood logistic regression to estimate the odds ratios and 95% confidence intervals relating trisomy to each indicator. As in the analogous matched-pairs analysis, the odds ratio is based only on conditionally informative strata in which the case and control are discordant. We report the results from analyses comparing trisomy cases with live birth controls because the two pregnancy loss comparison groups were too small to be informative. Among the 119 women with trisomic losses or live births, 28 (24%) were classified as having older ovarian age by the first definition and 17 (14%) by the second; 15 (13%) were so classified by both definitions.
Statistical adjustment
All but one set of analysesthose which test whether associations vary with chronological agecontrol for age stratum. Analyses of hormones adjust for day of blood draw because inhibin levels were lower on day 1 than days 24; a similar trend, of lesser magnitude, was present for FSH. In our initial analyses of follicle count, we adjusted for counting batch; however, because adjustment for batch changed the regression coefficients very little (i.e. 2046% of a standard error), we did not adjust for batch in the final models. For each of the four outcomes, we checked the stability of our point estimates against a model that included the maternal demographic and obstetric characteristics, body mass index and smoking and, for follicle count, a model that included batch, transducer and day of sonography. Changes in the point estimates were small. Results were unchanged when analyses were limited to the 49 strata (149 women) in which the ages of all comparison group women were within 6 months of their age-matched trisomy case.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
To explore the possibility that associations varied with maternal age, we adjusted for age as a continuous variable (rather than a stratification variable) and tested whether differences between trisomy cases and each of the three comparison groups varied with age. We checked the validity of the results from the regression analysis by smoothing the data with locally weighted regressions (Cleveland, 1979) for each pregnancy outcome group separately (Figure 1). For follicle count and inhibin B, the pattern of change with age was similar for trisomy cases and the three comparison groups. For FSH, the regression analysis indicated a steeper increase with age among trisomy cases than among live birth controls [
for trisomylive birth difference in ln(FSH) x age = 0.035, SE = 0.014, P = 0.015], but no difference between trisomy cases and either of the pregnancy loss comparison groups [
s for the difference in ln(FSH) x age were 0.006 in comparison with non-trisomic chromosomally abnormal losses and 0.005 in comparison with chromosomally normal losses]. FSH levels tended to be higher among trisomy cases than among live birth controls after the late 30s; the two pregnancy loss comparison groups exhibited patterns similar to those of the trisomy case group. For estradiol, the pattern of change with age did not differ between trisomy cases and any comparison group.
|
|
With respect to combined indicators of ovarian age, among the 54 trisomy cases, 27.8% met our a priori definition of older ovarian age (i.e. follicle count 6 or inhibin B
20 pg/ml or FSH
10 mIU/ml) compared with 20.0% of the 65 live birth controls (adjusted matched-strata odds ratio = 1.4, 95% CI 0.6, 3.4). For the summed z-score indicator of ovarian age, 11.1% of trisomy cases and 16.9% of live birth controls had measures in the 14th percentile or greater (adjusted matched-strata odds ratio = 0.6, 95% CI 0.2, 1.7). Post hoc analyses of the three components of the a priori measure showed: no association with follicle count
6 (stratum-adjusted odds ratio = 1.0, 95% CI 0.3, 2.9); a positive association with inhibin B
20 pg/ml (stratum-adjusted odds ratio = 1.6, 95% CI 0.6, 4.7); a positive association with FSH
10 mIU/ml (stratum-adjusted odds ratio = 2.8, 95% CI 0.5, 14.6).
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Our data do not support the hypothesis. Ovarian age indicators were similar for women with trisomic spontaneous abortions and for women of the same age whose pregnancies ended in non-trisomic chromosomally abnormal loss, chromosomally normal loss or live birth. Results were unaltered by statistical adjustment for several maternal characteristics and exposures or for technical variables (e.g. ultrasound transducer), or when we excluded from the live birth group women with prior spontaneous abortions (which might include undetected trisomies). The sample size was sufficient (power = 0.80, = 0.05, two-tailed) to detect differences between trisomy cases and live birth controls (the largest comparison group) equivalent to 0.52 standard errors of regression. At the median value of each indicator (15 antral follicles, inhibin B 79.5 pg/ml, FSH 4.4 mIU/ml), detectable effect sizes correspond to 4.1 fewer follicles, a 22.3 pg/ml decrease in inhibin B and a 1.1 mIU/ml increase in FSH. Our 95% CIs suggest that a decrease in antral follicle count of >13.2%, a decrease in inhibin B of >6.5% and an increase in FSH of >10.4%all small effectsare unlikely.
Two a priori definitions of ovarian age, which combined data on follicle count, inhibin B and FSH, were unrelated to trisomic loss. Post hoc analyses showed that follicle count 6 (the lowest 12th percentile in our data) was unrelated to trisomy. Inhibin B level
20 pg/ml and FSH level
10 mIU/ml were positively, but not significantly, associated with trisomy.
Estradiol levels were similar for trisomy cases and live birth controls. A single measure of estradiol on days 14, however, may not tap the phenomenon of interest, namely, the earlier emergence of the dominant follicle associated with greater chronological age (Klein et al., 1996a, 2000). (The significantly higher level of estradiol in women with chromosomally normal losses compared to women with trisomic losses may result from chance or may prove informative about the aetiology of chromosomally normal loss.)
Fifty-two per cent of women with karyotyped losses and 30% of potential live birth controls completed the study. Nevertheless, we think that our sample fairly represents the ovarian age of the population from which it was drawn. The principal reason for declining the study (22% of women with losses, 31% of women with births) was inconvenience: it is implausible that the ovarian age of women too busy to enroll differed systematically from the ovarian age of women who participated. Among women ineligible for the study (25% of women with losses, 37% of women with births), the principal reasons were hormonal contraceptive use, pregnancy or, for women with births, breastfeeding. Exclusion for pregnancy may have selectively removed more fertile (younger ovarian age) women from the sample. About the same proportion of women with losses (18/61) and with births (12/57 after excluding controls ineligible due to breastfeeding) became pregnant; thus, if pregnancy is a marker of young ovarian age, both the loss and birth samples overestimate ovarian age. Exclusion of women who used hormonal contraceptives would lead to bias only if contraceptive method relates to ovarian age (unknown by the woman), given chronological age. The remaining exclusionary criteria (hormonal conditions, oophorectomy, fertility medications), rendering 14 women with losses and six women with births ineligible, limit generalizations to women with neither hormonal disorders nor the constellation of circumstances (not all medical) leading to prescription of fertility agents.
A study of particular relevance (van Montfrans et al., 1999, 2002) compared women with Downs syndrome births (n = 118) to controls with at least two children and no history of trisomy (n = 102). Mean FSH was calculated from measurements taken in three cycles; the proportion with elevated FSH (the upper 5th percentile in controls) was identified by the highest of the three levels observed. As in our study, inhibin B and estradiol were unrelated to trisomy. In contrast, mean FSH levels were significantly higher in cases: we estimate that their geometric mean was 9% higher in cases than in controls, whereas we observed a downward shift of
5.5%. Strictly speaking, the results of the two studies do not differ (P = 0.10).
In the Downs syndrome study, the odds ratio relating elevated FSH to Downs syndrome births was 3.0 (95% CI 1.1, 8.6). In our study, comparing trisomy spontaneous abortions with live birth controls, the adjusted odds ratio for FSH 10 mIU/ml (the upper 3.1st percentile in controls) was 2.8 (95% CI 0.5, 14.6). Both odds ratios are compatible with the oocyte pool hypothesis. Inferences derived from our sample are insecure, however, because only five cases and two controls had elevated levels. Furthermore, the more powerful test of differences in geometric means suggests that in our sample FSH is not associated with trisomic pregnancy.
Our study, unlike the Downs syndrome study, included trisomies of most autosomes. Accumulating evidence suggests that, although the majority of trisomies derive from error during meiosis I, the predominant mechanism varies with trisomy type (Hassold and Hunt, 2001). Exploratory analyses of trisomic losses classified in three groupstrisomy 16, other non-acrocentric, acrocentric (including five with trisomy 21)did not suggest older ovarian age in women with acrocentric trisomies.
The exploratory analyses raise the possibility of younger ovarian age in women with trisomy 16 losses than in live birth controls. This result is intriguing because trisomy 16 is unusual: it occurs more often than trisomy of any other chromosome; it increases linearly with maternal age, whereas most other trisomies increase at a faster rate after the mid-thirties than before (Hassold and Chiu, 1985; Risch et al., 1986
).
Two methodological differences require mention. First, the Downs syndrome researchers collected hormone data from three cycles, whereas our data derive from a single cycle. Second, in the Downs syndrome study, residual confounding by chronological age cannot be definitively excluded even though the average age of cases and controls was similar.
A potential limitation of our study is that we did not obtain measures from several menstrual cycles. It is uncertain, however, that replicate measures would have improved our estimate of ovarian age. For antral follicle count, one study of women of proven fertility (Scheffer et al., 1999) suggests only modest improvement from replicate measures: the correlations with chronological age of count from a single cycle and from the mean of two cycles were comparable. Moreover, because more than a third of the sample was aged 4146 years, this study may overestimate intercycle variability in women of reproductive age. For FSH and inhibin B, the literature provides no evidence about intercycle variability in fertile women of reproductive age. For FSH, data from in vitro samples (Scott et al., 1990
; Martin et al., 1996
) suggest that ovarian age might be better classified by the highest of multiple measures rather than the mean. Given the select sample and special circumstances of IVF and the outcomes (ovarian response and clinical pregnancy), it remains unproven that classification by highest FSH level provides the best indicator of ovarian age.
A countervailing strength of our study is that we measured several different indicators of ovarian age. Three criteriareliability, validity and biological plausibilitysupport the assumption that these measures accurately reflect the size of the oocyte pool. The procedure for counting antral follicles was repeatable; the intra- and inter-assay CV for the hormone assays were excellent. The indicators were each associated, in the expected directions, with each other (data not shown) and with chronological age.
The straightforward interpretation of our data is that, contrary to our hypothesis, trisomy risk is unrelated to the size of the oocyte pool. If, indeed, our measures reflect the number of developing follicles at the time of the trisomy conception, our data argue against the limited oocyte pool hypothesis (Warburton, 1989). Neither antral follicle count nor inhibin B was associated with trisomy. Because the hypothesis might apply only when the oocyte pool drops below a critical threshold, we explored whether or not associations changed with chronological age. FSH levels tended to increase more steeply in the late thirties in women with trisomic losses than in women with births. The pattern was not specific to trisomic loss, however. Hence, even among older women, our data do not support the hypothesis.
Implications for the biological ageing hypothesis are less clear. Since developing follicles are thought to be the primary source of intraovarian hormones, our data do not support a role for hormonal imbalance associated with depletion of the follicle pool (Brook et al., 1984). On the other hand, it is possible that the total oocyte pool influences local ovarian or follicular hormone concentrations reflected neither by antral follicle count nor by serum hormone levels.
Aside from the Downs syndrome study (van Montfrans et al., 1999), few epidemiological observations speak directly to the oocyte pool hypothesis. Perhaps the strongest evidence comes from the observation, in our prospective study, that menopause occurred
1 year earlier (95% CI 2.10, 0.18) among women with trisomic pregnancy losses than among women with chromosomally normal pregnancies (Kline et al., 2000
). We inferred that women with trisomic pregnancies had smaller oocyte pools than other women of the same chronological age, so that they were chronologically younger when their pools dropped below the threshold necessary to maintain menstruation.
In light of our current study, we considered alternative interpretations of the link between trisomic pregnancy and earlier menopause. One possibilitythat our finding was spuriouscan be addressed only by replicate studies. Another possibility is that defects in follicular growth, maturation or composition unrelated to the size of the oocyte pool affect both the likelihood of trisomy and age at menopause. For example, a recent study (Pan et al., 2002) shows that blood flow to the ovary decreases monotonically among women who are menstruating, peri-menopausal or post-menopausal.
Research into age-related changes in intraovarian and intrafollicular factors and their relation to non-disjunction is just beginning. Because most data derive from follicles or oocytes exposed to exogenous hormones, interpretation is vulnerable to uncertainty about whether or not exogenous exposure alters the phenomena of interest. Studies which draw on women undergoing IVF have the added complication that findings may not be generalizable to fertile women.
One study (Klein et al., 2000) suggests that age-related differences in serum hormone levels may not be mimicked within the follicle. In the follicular fluid of dominant follicles, levels of nine hormones (including inhibin B and estradiol) and growth factors were similar for older (4045 years) and younger (2025 years) women; four were different. The authors suggest that the increase in the average concentration of vascular endothelial growth factor observed in follicles from older women is linked to meiotic error via diminished oxygen content in the follicular fluid which, in turn, increases the rate of abnormalities of the meiotic spindle in metaphase II oocytes (Van Blerkom et al., 1997
). The latter observation, derived from follicles aspirated after ovarian stimulation for IVF, has unknown implications for fertile women. An interesting feature of the data, however, was that oxygen concentrations of adjacent follicles often differed, demonstrating inter-follicular differences within women, even holding constant the size of the oocyte pool.
At least one study (Volarcik et al., 1998) implicates factors occurring around the time of oocyte maturation rather than around the time of ovulation. Pre-ovulatory follicles obtained from ovaries unexposed to exogenous hormones were matured in vitro in a hormone-containing culture solution. The rate of misalignment of chromosomes on the spindle (congression failure) in meiosis II metaphase divisions was higher in oocytes from women aged
35 years than in oocytes from younger women. In addition, the frequency of non-disjunction for four chromosomes at meiosis I was about twice as high in oocytes from older women as in oocytes from younger women. Congression failure was also observed in an earlier study (Battaglia et al., 1996
) of fertile women from whom follicles were aspirated, after stimulation with hCG, around the time the dominant follicle reached 1516 mm, presumably shortly before ovulation.
Volarcik et al. (1998) postulated that an age-related defect in folliculogenesis, already present by the late antral stage, increases the risk of meiotic error. Since follicles were cultured in a hormone-containing medium, their result raises the possibility that hormonal conditions at the resumption of meiosis may not affect abnormal chromosome segregation. Other postulated mechanisms include changes in proteins that influence chromosome segregation (Koehler et al., 1996
; Watanabe and Nurse 1999
; Jeffreys et al., 2003
) and the level of oxygenation within the follicle (Van Blerkom et al., 1997
). This is not to say, however, that hormones cannot affect the risk of trisomy. For example, female mice exposed to bisphenol A, a man-made estrogenic compound, have shown increased rates of aneuploidy; experimental studies suggest that even short-term exposure during the final stages of oocyte growth may disrupt meiosis (Hunt et al., 2003
).
Our data indicate that factors influencing meiotic error are unrelated to the number of developing follicles and, most likely, unrelated to the size of the underlying oocyte pool. One implication is that FSH, estradiol and inhibin B, at least as measured in serum, may not affect chromosome segregation. Another implication, of particular relevance in the clinical setting, is that chronological age remains the best predictor of trisomy risk.
![]() |
Acknowledgements |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Battaglia DE, Goodwin P, Klein NA and Soules MR (1996) Influence of maternal age on meiotic spindle assembly in oocytes from naturally cycling women. Hum Reprod 11,22172222.[Abstract]
Block E (1952) Quantitative morphological investigations of the follicular system in women: variations at different ages. Acta Anat 14,108123.[ISI]
Breslow NE and Day NE (1980) Statistical Methods in Cancer Research. Volume I. The Analysis of CaseControl Studies. Vol 32. IARC Scientific Publications, Lyon.
Brook JD, Gosden RG and Chandley AC (1984) Maternal ageing and aneuploid embryosevidence from the mouse that biological and not chronological age is the important influence. Hum Genet 66,4145.[ISI][Medline]
Burger HG (1994) Diagnostic role of follicle-stimulating hormone (FSH) measurements during the menopausal transitionan analysis of FSH, oestradiol and inhibin. Eur J Endocrinol 130,3842.[Medline]
Burger HG (2000) Inhibin and reproductive aging. Exp Gerontol 35,3339.[CrossRef][Medline]
Cleveland W (1979) Robust locally weighted regression and smoothing scatterplots. J Am Statist Assoc 74,829.[ISI]
Faddy MJ, Gosden RG, Gougeon A, Richardson SJ and Nelson JF (1992) Accelerated disappearance of ovarian follicles in mid-life: implications for forecasting menopause. Hum Reprod 7,13421346.[Abstract]
Fleiss JL (1986) The Design and Analysis of Clinical Experiments. John Wiley & Sons, New York.
Freeman SB, Yang Q, Allran K, Taft LF and Sherman SL (2000) Women with a reduced ovarian complement may have an increased risk for a child with Down syndrome. Am J Hum Genet 66,16801683.[CrossRef][ISI][Medline]
Goldenberg RL, Grodin JM, Rodbard D and Ross GT (1973) Gonadotropins in women with amenorrhea. The use of plasma follicle-stimulating hormone to differentiate women with and without ovarian follicles. Am J Obstet Gynecol 116,10031012.[ISI][Medline]
Gougeon A, Ecochard R and Thalabard JC (1994) Age-related changes of the population of human ovarian follicles: increase in the disappearance rate of non-growing and early-growing follicles in aging women. Biol Reprod 50,653663.[Abstract]
Hassold T and Chiu D (1985) Maternal age-specific rates of numerical chromosome abnormalities with special reference to trisomy. Hum Genet 70,1117.[ISI][Medline]
Hassold T and Hunt P (2001) To err (meiotically) is human: the genesis of human aneuploidy. Nat Rev Genet 2,280291.[CrossRef][ISI][Medline]
Hunt PA, Koehler KE, Susiarjo M, Hodges CA, Ilagan A, Voigt RC, Thomas S, Thomas BF and Hassold TJ (2003) Bisphenol a exposure causes meiotic aneuploidy in the female mouse. Curr Biol 13,546553.[CrossRef][ISI][Medline]
Jeffreys CA, Burrage PS and Bickel SE (2003) A model system for increased meiotic nondisjunction in older oocytes. Curr Biol 13,498503.[CrossRef][Medline]
Klein NA, Battaglia DE, Miller PB, Branigan EF, Giudice LC and Soules MR (1996a) Ovarian follicular development and the follicular fluid hormones and growth factors in normal women of advanced reproductive age. J Clin Endocrinol Metab 81,19461951.[Abstract]
Klein NA, Illingworth PJ, Groome NP, McNeilly AS, Battaglia DE and Soules MR (1996b) Decreased inhibin B secretion is associated with the monotropic FSH rise in older, ovulatory women: a study of serum and follicular fluid levels of dimeric inhibin A and B in spontaneous menstrual cycles. J Clin Endocrinol Metab 81,27422745.[Abstract]
Klein NA, Battaglia DE, Woodruff TK, Padmanabhan V, Giudice LC, Bremner WJ and Soules MR (2000) Ovarian follicular concentrations of activin, follistatin, inhibin, insulin-like growth factor I (IGF-I), IGF-II, IGF-binding protein-2 (IGFBP-2), IGFBP-3, and vascular endothelial growth factor in spontaneous menstrual cycles of normal women of advanced reproductive age. J Clin Endocrinol Metab 85,45204525.
Klein NA, Harper AJ, Houmard BS, Sluss PM and Soules MR (2002) Is the short follicular phase in older women secondary to advanced or accelerated dominant follicle development? J Clin Endocrinol Metab 87,57465750.
Kline J and Levin B (1992) Trisomy and age at menopause: predicted associations given a link with rate of oocyte atresia. Paediatr Perinat Epidemiol 6,225239.[Medline]
Kline J, Stein Z and Susser M (1989) Conception to Birth: Epidemiology of Prenatal Development. Oxford University Press, New York.
Kline J, Kinney A, Levin B and Warburton D (2000) Trisomic pregnancy and earlier age at menopause. Am J Hum Genet 67,395404.[CrossRef][Medline]
Koehler KE, Hawley RS, Sherman S and Hassold T (1996) Recombination and nondisjunction in humans and flies. Hum Mol Genet 5,14951504.[Abstract]
Lamb NE, Freeman SB, Savage-Austin A, Pettay D, Taft L, Hersey J, Gu Y, Shen J, Saker D, May KM et al (1996) Susceptible chiasmate configurations of chromosome 21 predispose to non-disjunction in both maternal meiosis I and meiosis II. Nat Genet 14,400405.[ISI][Medline]
Leidy LE, Godfrey LR and Sutherland MR (1998) Is follicular atresia biphasic? Fertil Steril 70,851859.[CrossRef][ISI][Medline]
Levin B (1987) Condtional likelihood analysis in stratum-matched retrospective studies with polytomous disease states. Commun Statist B 16(3),699718.
Levin B (1990) The saddlepoint correction in conditinal logistic likelihood analysis. Biometrika 77,275285.
Martin JS, Nisker JA, Tummon IS, Daniel SA, Auckland JL and Feyles V (1996) Future in vitro fertilization pregnancy potential of women with variably elevated day 3 follicle-stimulating hormone levels. Fertil Steril 65,12381240.[ISI][Medline]
Nasseri A, Mukherjee T, Grifo JA, Noyes N, Krey L and Copperman AB (1999) Elevated day 3 serum follicle stimulating hormone and/or estradiol may predict fetal aneuploidy. Fertil Steril 71,715718.[CrossRef][ISI][Medline]
Pan HA, Cheng YC, Li CH, Wu MH and Chang FM (2002) Ovarian stroma flow intensity decreases by age: a three-dimensional power doppler ultrasonographic study. Ultrasound Med Biol 28,425430.[CrossRef][ISI][Medline]
Reuss ML, Kline J, Santos R, Levin B and Timor-Tritsch I (1996) Age and the ovarian follicle pool assessed with transvaginal ultrasonography. Am J Obstet Gynecol 174,624627.[ISI][Medline]
Risch N, Stein Z, Kline J and Warburton D (1986) The relationship between maternal age and chromosome size in autosomal trisomy. Am J Hum Genet 39,6878.[ISI][Medline]
Scheffer GJ, Broekmans FJ, Dorland M, Habbema JD, Looman CW and te Velde ER (1999) Antral follicle counts by transvaginal ultrasonography are related to age in women with proven natural fertility. Fertil Steril 72,845851.[CrossRef][ISI][Medline]
Scott RT, Jr, Hofmann GE, Oehninger S and Muasher SJ (1990) Intercycle variability of day 3 follicle-stimulating hormone levels and its effect on stimulation quality in in vitro fertilization. Fertil Steril 54,297302.[ISI][Medline]
Thomford PJ, Jelovsek FR and Mattison DR (1987) Effect of oocyte number and rate of atresia on the age of menopause. Reprod Toxicol 1,4151.[Medline]
VanBlerkom J, Antczak M and Schrader R (1997) The developmental potential of the human oocyte is related to the dissolved oxygen content of follicular fluid: association with vascular endothelial growth factor levels and perifollicular blood flow characteristics. Hum Reprod 12,10471055.[CrossRef][ISI][Medline]
vanMontfrans JM, Dorland M, Oosterhuis GJ, van Vugt JM, Rekers-Mombarg LT and Lambalk CB (1999) Increased concentrations of follicle-stimulating hormone in mothers of children with Downs syndrome. Lancet 353,18531854.[CrossRef][ISI][Medline]
vanMontfrans JM, van Hooff MH, Martens F and Lambalk CB (2002) Basal FSH, estradiol and inhibin B concentrations in women with a previous Downs syndrome affected pregnancy. Hum Reprod 17,4447.
Volarcik K, Sheean L, Goldfarb J, Woods L, Abdul-Karim FW and Hunt P (1998) The meiotic competence of in-vitro matured human oocytes is influenced by donor age: evidence that folliculogenesis is compromised in the reproductively aged ovary. Hum Reprod 13,154160.[Abstract]
Warburton D (1989) The effect of maternal age on the frequency of trisomy: change in meiosis or in utero selection? Prog Clin Biol Res 311,165181.[Medline]
Warburton D and Kline J (2001) Maternal age and aneuploidy: a re-examination of the limited oocyte pool hypothesis. Salisb Med J 101(Suppl),5164.
Watanabe Y and Nurse P (1999) Cohesin Rec8 is required for reductional chromosome segregation at meiosis. Nature 400,461464.[CrossRef][ISI][Medline]
Submitted on July 26, 2003; accepted on January 15, 2004.