Risk of benign gynaecological diseases and hormonal disorders according to responsiveness to ovarian stimulation in IVF: a follow-up study of 8714 women

H. Klip1,2,5, F. E. van Leeuwen1, R. Schats3, C.W. Burger4 and for the OMEGA project group*

1 Department of Epidemiology, Netherlands Cancer Institute, Amsterdam, 2 Research Bureau, Isala klinieken, Zwolle, 3 Department of Obstetrics and Gynaecology, Academic Hospital Free University, Amsterdam and 4 Department of Obstetrics and Gynaecology, Erasmus University Medical Center, Rotterdam, The Netherlands

5 To whom correspondence should be addressed at: Research Bureau, Isala klinieken, PO Box 10500, 8000 GM Zwolle, The Netherlands. e-mail: h.klip{at}isala.nl


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: Over the past decade, attention has been focused increasingly on the long-term health effects of IVF in women. Assuming that hormonal changes due to stimulation regimens for IVF are strongest among ‘high’ responders, we evaluated whether responsiveness to ovarian stimulation in IVF is predictive of the risk of benign gynaecological disorders >12 months after the last IVF cycle. METHODS: A nationwide historical cohort study of women who underwent IVF treatment was conducted. After a median time of 4.6 years following the last IVF treatment cycle, 8714 cohort members completed a health survey questionnaire that inquired about reproductive variables and the occurrence and age at onset of specific medical conditions including uterine leiomyoma, surgically removed ovarian cysts and thyroid disorders. Detailed data on cause of subfertility and IVF treatment were collected from the medical records. Women were included in the ‘high responders’ group when on average >=14 oocytes were retrieved per IVF cycle (n = 1562), in the ‘normal responders’ group when they had a mean number of 4–13 retrieved oocytes (n = 6033), and in the ‘low responders’ group when they had a mean number of 0–3 retrieved oocytes per cycle (n = 1119). RESULTS: Among women with a high response to ovarian stimulation, we found a borderline significantly decreased risk of uterine leiomyoma [relative risk (RR) = 0.6; 95% confidence interval (CI) 0.4–1.0] and surgically removed ovarian cysts (RR = 0.6; 95% CI 0.3–1.0) in comparison with ‘normal responders’. After OHSS, the age-adjusted RRs were 1.8 (95% CI 0.9–3.8) for having surgically removed ovarian cysts and 1.0 (95% CI 0.4–2.2) for uterine leiomyoma (both not significant). CONCLUSIONS: Despite the small number of events observed, highly elevated risks of gynaecological disorders and hormonal diseases in women undergoing IVF treatment can be excluded based on the present data and this follow-up period. Women with a low response to ovarian stimulation tended to have higher risks of benign gynaecological diseases than high responders.

Key words: IVF/long-term follow-up/OHSS/ovarian cysts/uterine leiomyoma


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Although IVF has been a method for treatment of infertility since the early 1980s, the long-term health effects of hormonal stimulation for IVF in women have been investigated in only a few studies (reviewed in Klip et al., 2000Go). As part of the IVF treatment, fertility drugs (FDs) are used during the follicular phase of the menstrual cycle to increase the serum concentration of gonadotrophins, aiming at multiple follicle maturation.

The main focus of recent publications has been on the possible association between the use of FDs and the development of malignancies of the ovary, breast and endometrium (Ron et al., 1987Go; Shu et al., 1989Go; Harris et al., 1992Go; Whittemore et al., 1992Go; Kelsey et al., 1993Go; Rossing et al., 1994Go; Venn et al., 1995Go; Braga et al., 1996Go; Rossing et al., 1996Go; Shushan et al., 1996Go; Mosgaard et al., 1997Go; Parazzini et al., 1997Go, 1998, 2001; Modan et al., 1998Go; Potashnik et al., 1999Go; Venn et al., 1999Go; Klip et al., 2000Go). So far, these studies have shown inconsistent results. However, low statistical power and lack of control for important confounders, such as the cause of subfertility and/or parity, and lack of information about the type of FDs used, have limited previous studies of fertility treatment and cancer risk. Furthermore, the follow-up time after subfertility treatment has been relatively short in most studies.

So far, there is very little information on the long-term effects of IVF on the risk of benign gynaecological diseases (e.g. uterine leiomyoma, surgically removed ovarian cysts, pelvic inflammatory disease and endometriosis) and hormonal disorders (e.g. thyroid gland disorders and diabetes mellitus).

There are a number of reasons why women’s health may be compromised after IVF procedures. The female reproductive tract is unique because the follicular and luteal phases of the menstrual cycle occur in a repetitive cyclic fashion. Disturbance of this balance may result in a disrupted physiological state or various pathological conditions (Abulafia and Sherer, 2000Go). Thus, characteristics directly or indirectly reflecting a woman’s exposure to estrogens and progestagens (e.g. menstrual history, childbearing, exogenous hormone use, such as FDs) may inhibit or promote the development of certain diseases later in life.

We speculated that if hormonal stimulation had any effect on the occurrences of gynaecological diseases and hormonal disorders in later life, this might be most apparent in ‘high’ responders, in particular in women who developed an ovarian hyperstimulation syndrome (OHSS). To evaluate whether responsiveness to ovarian stimulation in IVF is predictive of the risk of benign gyneacological conditions, we evaluated self-reported adverse non-malignant health outcomes according to the average number of retrieved oocytes during IVF treatment in a cohort of women who underwent ovarian stimulation for IVF in The Netherlands.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The OMEGA study
The OMEGA study is a nationwide cohort study of 26 428 women diagnosed with subfertility problems in 12 IVF clinics in The Netherlands between January 1, 1980 and January 1, 1995. The main purpose of the OMEGA study is to examine the late effects of hormone stimulation for IVF in treated women. To be included in the initial cohort, women had to be unable to achieve conception after one or more years of frequent unprotected intercourse and they had to be older than 18 years at the time of their first visit to the fertility clinic. Initially, 19 840 women were included in the exposed (IVF) group, whereas the control group (no IVF) consisted of 6588 women. The present study is restricted to women who received IVF treatment. All participating clinics provided a minimal data set with names, dates of birth and addresses of eligible women, after which women could be traced and informed about the study. All institutional ethics committees of the clinics approved the study procedures.

Follow-up
Given the fact that the subjects’ last visit to the infertility clinic might date back to 1980, extensive tracing techniques were used to obtain current addresses of all women. Cohort members were traced through the Dutch Telephone Service Company (PTT) and searches at municipal resident registries to determine vital status and address (if applicable). According to the study procedures, a health questionnaire survey was conducted among all cohort members, data on subfertility diagnosis and treatment were collected from their medical records, and cancer incidence through 1997 was ascertained through linkage with the population-based Netherlands Cancer Registry (NCR). Results of the cancer incidence analysis will be presented elsewhere.

Data collection
Health questionnaire survey. From the initial 19 840 women in the IVF group, 598 women were not approached because of death, unknown, incomplete or foreign addresses, emigration or specific other reasons related to privacy. A total of 19 242 women received the 23-page study questionnaire, an information letter and a study brochure. An extra form was attached to the questionnaire, asking each participating subject to provide written informed consent for data abstraction from the medical records and linkage with the NCR. The study information letter was signed by the attending gynaecologist or, if he/she had left, the current head of the IVF department. After 4–6 weeks, non-responders received a reminder. Non-responders to the second letter were approached subsequently by telephone (once). Of the 19 242 questionnaires that were mailed, a total of 468 (2.4%) questionnaires were either returned to sender or the women telephoned or wrote to say that they did not want to participate. A total of 13 698 women returned our risk factor questionnaire (response rate 71.2%).

Through the questionnaire, we obtained information on the women’s reproductive history, their history of subfertility treatment, the use of exogenous hormones (FDs, oral contraceptives and hormone replacement therapy), and various lifestyle factors (smoking, alcohol use and physical activity). Furthermore, information was elicited on all hormone-related diseases, specific gynaecological diseases and gynaecological surgical procedures before and after IVF treatment.

The questions on health outcomes were stated as follows: ‘Did you ever have any of the following conditions or diseases?’ with checkboxes for yes/no. If ’yes’, we inquired about the age at diagnosis or the year of diagnosis of the disease. For the purpose of this study, we included the self-reported occurrence of pelvic inflammatory disease, uterine leiomyoma, surgically removed ovarian cysts, endometriosis, several thyroid and adrenal gland-related disorders, and diabetes. Due to imprecision in the measurement of the age and/or year of the event (instead of the exact date of diagnosis), only diseases that occurred at least one calendar year after IVF treatment had been completed were included in the analyses.

Medical records. In each participating clinic, research assistants specifically trained for data collection in this study abstracted detailed information from the medical records on the type of subfertility. Infertility investigations routinely used by IVF clinics and referring gynaecologists included hormone assays, ultrasonography, diagnostic laparoscopy and semen analysis. The cause of infertility was classified as due to tubal factors, male factors, endometriosis, ovarian disorders (including ovulation disorders, polycystic ovary syndrome, premature menopause and hormonal disorders related to ovarian function), other factors (including cervical factors, other uterine abnormalities) or unexplained. Women could have more than one cause of subfertility; unexplained infertility was a unique classification.

For each IVF cycle, the following data were recorded: date, type of IVF treatment [IVF, frozen embryo transfer, ICSI, zygote intrafallopian transfer (ZIFT), gamete intrafallopian transfer (GIFT), IVF without ovarian stimulation, IVF with oocyte donor, IVF with sperm donor, IVF with oocyte acceptor], dosages and type of FDs used in each phase of the menstrual cycle (HMG, FSH, clomiphene citrate, HCG, progesterone and luteal phase support), number of oocytes collected, number of embryos transferred, outcome and complications. Information on OHSS was collected for each IVF treatment cycle with a checkbox for yes/no, according to the criteria of the hospital where the IVF treatment had taken place. Although several classification systems for OHSS have been developed, all participating IVF clinics used the guidelines of the Netherlands Organization of Obstetrics and Gynecology (NVOG) (Navot and Bergh, 1993Go; Simons, 1998Go).

Definition of analytic cohort
For the purpose of the present research question, the study population consisted of women in the IVF group who responded to our questionnaire (n = 13 698). We excluded women who only filled out an informed consent form (n = 270), and all women with <1 year of follow-up (n = 635). Furthermore, since data collection of the individual medical records turned out to be extremely time consuming, and only limited funding was available, we were not able to carry out data collection for 3227 women who did give informed consent. These women were excluded from the present study since medical record data were essential to answer our research questions. When defining the analytical cohort, we selected women based on complete information regarding all IVF treatments (classical IVF treatment procedure, ICSI, GIFT or ZIFT). If a women had undergone only one IVF treatment, she was excluded in the following situations: women who received IVF without ovarian stimulation, donor IVF, acceptor IVF and unknown or other IVF procedures (n = 620), women with missing data on the IVF treatment (n = 53), women for whom the number of oocytes retrieved at IVF treatment was not known (n = 44), and women who had their first (and only) IVF treatment cycle cancelled due to reasons other than impaired follicular reaction in response to exogenous gonadotrophins or imminent risk of OHSS, or due to cancelled cycles for unknown reasons (n = 135). Thus, the total cohort for this analysis consisted of 8714 women.

In Figure 1, the distribution of the average number of oocytes retrieved during all IVF treatment cycles is presented graphically. The average number of oocytes retrieved was used as a proxy measure of the responsiveness of each woman to ovarian stimulation. In order to create three groups of women based on the average number of oocytes retrieved, we created cut-off points at the 15th and 85th percentile. According to this arbitrary, but objective definition, the ‘low’ responders had <4 oocytes at retrieval (cut-off point at the 15th percentile) and the ‘high’ responders had >=14 retrieved oocytes (cut-off point at the 85th percentile). In all other cases, ovarian response was classified as ‘normal’.



View larger version (12K):
[in this window]
[in a new window]
 
Figure 1. Distribution of women according to average number of oocytes retrieved during IVF treatments.

 
Statistical analysis
The questions on ever having suffered from gynaecological disorders or endocrine diseases were grouped together in one section of the questionnaire. Descriptive analyses revealed that a considerable proportion of subjects had reported a history of one or two conditions, but had not completed the checkboxes for all other questions in this group of disorders. As a result, we noted up to 8% of missing observations for some of these disorders. As the proportions of missing observations per disorder were very similar for exposed and non-exposed groups, we interpreted missing observations for these questions to be a ‘no’ if one or more of the other questions in this section was answered in the affirmative. Since the definition of exposure category was based on the numbers of retrieved oocytes over all IVF cycles, only disorders occurring 1 year after the last IVF treatment date were taken into account in the analyses. As a consequence, disorders occurring after previous cycles were not included when patients resumed treatment after resolution of any acute problem.

Frequency tables of disorders of interest were assembled for the three exposed groups according to the mean number of retrieved oocytes, and homogeneity was tested using Pearson {chi}2-tests (Rothman and Greenland, 1998Go). Cox proportional hazards models were used to obtain relative risks (RRs) and 95% confidence intervals (CIs) (Cox, 1972Go). The RRs estimate the risk of specific gynaecological disorders in ‘low’ and ‘high’ responders compared with the risk in ’normal’ responders, adjusting for follow-up time, age at end of follow-up and other possible confounders, as appropriate. The observation time for each woman started 1 year after the last IVF treatment had been completed. Person-years of observation were calculated to the date the questionnaire was returned or the date of the diagnosis of the gynaecological disease under study, whichever came first. In the Cox models, we excluded all women with missing data on confounders. All women who reported a specific condition before or during IVF treatment and 1 year after IVF treatment was finished were excluded from analysis of that particular condition.

In a second analysis, women were classified according to whether or not they had ever been diagnosed with OHSS, during any or all of their IVF treatment cycles. The Cox proportional hazards model was used to quantify further the effects of OHSS on the risk of gynaecological diseases.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A summary of population characteristics of women in the OMEGA cohort according to exposure status is shown in Table I. The large majority of women in the ‘normal’ responders category had tubal infertility (35.6%), male infertility (32.5%) or unexplained infertility (24.5%). Only in 4.1% (n = 245) of women was subfertility due to hormonal factors (including ovulation disorders, polycystic ovary syndrome and premature menopause), while 6.6% of the women were subfertile due to endometriosis. In the ‘low’ responders category, more women had tubal infertility (41.9%), and fewer women had male infertility (24.4%) compared with ‘normal’ responders (two-tailed {chi}2-test, P < 0.001). In the ‘high’ responders category, more women were diagnosed with hormonal factors compared with women in the ‘normal’ responders category (two-tailed {chi}2-test, P < 0.001).


View this table:
[in this window]
[in a new window]
 
Table I. Population characteristics of the OMEGA cohort by response to IVF treatment
 
The so-called ‘low’ responders are likely to suffer from diminished ovarian reserve and consequently have impaired pregnancy prospects due to loss of oocyte quality. In our cohort, women diagnosed with premature ovarian failure had a decreased ovarian responsiveness to gonadotrophin administration (less than four oocytes), whereas women with polycystic ovary syndrome more often had a higher response to ovarian stimulation (Table I).

Significantly more women in the ‘low’ responders group were >=35 years of age at the first visit to the fertility clinic compared with women in the ‘normal’ responders group (two-tailed {chi}2-test, P < 0.001). The proportions of women aged >=35 years at the first visit were 18.1, 8.4 and 3.3% among ‘low’, ‘normal’ and ‘high’ responders respectively. The number of children and the total number of IVF cycles also differed in the three exposure groups. Relatively more women in the ‘low’ responders group were nulliparous at the time the questionnaire was completed and had received four or more IVF treatments cycles, compared with the ‘normal’ responders group (two-tailed {chi}2-test, P < 0.001).

Table II shows the results of Cox regression models that estimated the risk of self-reported occurrence of pelvic inflammatory disease, uterine leiomyoma, surgically removed ovarian cyst, endometriosis, and several thyroid and adrenal gland-related disorders, according to the average oocyte response to IVF treatment. All analyses were adjusted for age at end of follow-up, subfertility diagnosis, number of IVF cycles and parity.


View this table:
[in this window]
[in a new window]
 
Table II. Relative risk of gynaecological and endocrinological disorders among women with a low, normal and high average number of retrieved oocytes
 
A total of 209 women reported the occurrence of uterine leiomyoma after IVF treatment was discontinued. For uterine leiomyoma, the RRs showed, after adjustment for confounders, a slightly decreasing trend with increasing number of oocytes. The RR of uterine leiomyomata was decreased among ‘high’ responders compared with the ‘normal’ responders (RR = 0.6; 95% CI 0.4–1.0).

In all, 118 cases of self-reported surgically removed ovarian cysts were observed, 20 events among ‘low’ responders and 12 events among ‘high’ responders. The mean ages at the self-reported surgery for ovarian cysts was 36.2, 37.1 and 34.8 years among ‘low’, ‘normal’ and ‘high’ responders respectively. For ‘low’ responders, the adjusted RR of self-reported surgically removed ovarian cysts was 1.1 (95% CI 0.7–1.8) compared with the risk among ‘normal’ responders. For ‘high’ responders, the adjusted RR was 0.6 (95% CI 0.3–1.0). The risk of self-reported pelvic inflammatory disease also showed a slightly decreasing trend with increasing number of oocytes increased. No risks were observed for endometriosis for any of the categories.

With regard to endocrinological diseases, we assessed the reported frequency of diseases of the thyroid gland among the three responder categories (Table II). In the ‘low’ responders category, 13 cases were reported during 5012 person years compared with 42 cases during 22 443 person years among the ‘normal’ responders (RRadj = 1.2; 95% CI 0.6–2.5). Among women with >14 oocytes retrieved during IVF treatment, a total of five cases of thyroid disease were observed during 5192 person years resulting in a non-significantly decreased risk of 0.5 (95% CI 0.2–1.2) compared with ‘normal’ responders.

We also examined the risk of gynaecological diseases according to whether or not women had ever been diagnosed with OHSS during any or all of their IVF treatment cycles. Of the 8714 women in the cohort, 360 women (4.1%) were diagnosed with OHSS during one of their IVF cycles. As shown in Table III, the age-adjusted RR for having surgically removed ovarian cysts associated with OHSS was slightly increased (RR = 1.7; 95% CI 0.8–3.4). Adjustment for potential confounders increased the risk to 1.8 (95% CI 0.9–3.8). The adjusted RR for uterine leiomyoma was 1.0 (95% CI 0.4–2.2).


View this table:
[in this window]
[in a new window]
 
Table III. Relative risk of gynecological diseases according to history of OHSS (n = 360) during all IVF treatment cycles
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In this report, we assessed, for the first time, the long-term risk of gynaecological disorders and endocrinological diseases following IVF treatment. We decided to estimate disease risk according to the average number of retrieved oocytes, speculating that the risk of gynaecological disorders might be most strongly influenced among women who showed an extreme response to hormonal stimulation. In contrast to our hypothesis, for women with a high response to ovarian stimulation (as compared with ‘normal’ responders), we found borderline significantly decreased risks for self-reported surgically removed ovarian cysts (RR = 0.6), uterine leiomyoma (RR = 0.6) and pelvic inflammatory disease (RR = 0.4) After OHSS, the adjusted RR for having surgically removed ovarian cysts was increased, though not significantly so (RR = 1.8, NS), while no effect was observed on the risk of uterine leiomyoma.

The number of oocytes collected in a stimulated IVF treatment cycle can be used as a proxy measure of the responsiveness of each woman to ovarian stimulation. Patients with a low response to controlled ovarian stimulation fall into two distinct categories: low responders and over-responders. In many cases, a low response to ovarian stimulation is attributable to diminished ovarian reserve. Although high responder patients enjoy a better overall success rate than their low responder counterparts, exaggerated ovarian sensitivity to stimulation substantially increases the risk of severe OHSS (Davis and Rosenwaks, 1995Go). Currently, there is no uniform definition of low and/or high response to ovulation induction available (Lashen et al., 1999Go). From a practical point of view, given a mean fertilization rate of 50–60% in IVF, four oocytes are needed to reach an average of two embryos available for transfer. Hence, we defined low response as the collection of less than four oocytes at retrieval (cut-off point at the 15th percentile) and the ‘high’ responders had >=14 retrieved oocytes (cut-off point at the 85th percentile).

OHSS is a serious and potentially life-threatening complication of IVF. The short-term health risks of OHSS are characterized by enlarged ovaries and extravasation of fluid to the abdominal cavity resulting in ascites, hypovolaemia and haemoconcentration, but the long-term health consequences are not clear (Fauser et al., 1999Go). After ovulation induction, women are at higher risk of OHSS when the serum estradiol is >10 nmol/l (2500 pg/ml) (Navot and Bergh, 1993Go). A unique aspect of our study is that we had a median follow-up time of 4.6 years following the last IVF treatment cycle for a very large group of women who developed OHSS (n = 360). No significantly elevated risk of benign gynaecological disorders was observed, although surgically removed ovarian cysts occurred slightly more often among women who developed OHSS. Since this finding was based on a small number of events, clearly this issue deserves more research.

Although the incidence rate of OHSS seen in our cohort of IVF patients is similar to the incidence rates reported in the literature (Surrey, 1995Go), there are two issues to be discussed concerning misclassification of the women in the two groups (OHSS versus no OHSS). First, information on OHSS in our IVF patients was collected from the medical records. As mentioned previously, all clinics used the criteria of the NVOG (Navot and Bergh, 1993Go; Simons, 1998Go). Although it may be asumed that clinics follow the guidelines, it is unknown how much variability there was in using these criteria. Secondly, women with a cancelled cycle because they were considered at risk of OHSS due to exaggerated follicle growth were not included in the OHSS group since they did not develop OHSS according to the criteria. Due to the small number of cases, we were not able to conduct separate analyses for this specific subgroup. Further research in this area should preferably also examine the risk of specific health outcomes in the group of women with cancelled cycles, in order to separate the effects of preventive management procedures for OHSS.

So far, no reports are available on the risk of benign ovarian cysts after ovulation induction for IVF. In our cohort, we found a higher risk for ovarian cysts in women with a low number of retrieved oocytes. As mentioned before, many ‘low’ responders are likely to suffer from diminished ovarian reserve. In a study by de Boer et al. (2002Go), women with a low response (0–3 oocytes) at their first IVF treatment had an RR of 11.6 (95% CI 3.9–34.7) of having an early menopause compared with women with a normal response.

Despite the fact that uterine leiomyoma is reported to be the most common benign gynaecological tumour affecting premenopausal women, no studies have been published on the risk of uterine leiomyoma after ovulation induction for IVF. The aetiology of uterine leiomyoma remains unknown. Associations with reproductive characteristics directly or indirectly reflecting a woman’s exposure to estrogens and progestagens are hypothesized to enhance fibroid growth (Schwartz, 2001Go). In a case–control study by Chen et al. (2001Go) on risk factors for fibroids, elevated odds ratios (ORs) were found women with a menstrual cycle length of >30 days (OR = 1.6; 95% CI 1.1–3.3), and menstrual bleeding for >=6 days (OR = 1.4; 95% CI 1.0–2.0). Furthermore, parous women were at reduced risk compared with nulliparous women (OR = 0.2; 95% CI 0.1–0.3) (Chen et al., 2001Go). In our study, we did observe a slightly increased risk of uterine leiomyoma with a lower number of retrieved oocytes; however, a possible biological pathway is unclear to us. A low number of retrieved oocytes after gonadotrophin stimulation in IVF may be due to early ovarian depletion. This is supported by the fact that in older women fewer oocytes are retrieved after stimulation with gonadotrophins during IVF than in younger women. On the other hand, a low number of retrieved oocytes may also be due to hormonal disturbances not related to ovarian ageing. The observed increased risk of uterine leiomyoma (Chen et al., 2001Go) with long cycle length may be indicative of underlying ovarian pathophysiology unrelated to ageing of reproductive processes, or may be a confounding effect of ovarian ageing. Although the number of events was small, we found no elevated risks for having uterine leiomyoma in our OHSS group.

The strengths and limitations of our study design should be taken into account when interpreting our results. The strengths of our study include the relatively large study size and the medically verified information on subfertility diagnosis and treatment(s). Cohort identification was based on individual medical treatment records from participating IVF clinics; therefore, misclassification of response to IVF is highly unlikely. The cohort was followed for 4.6 years, with almost 96% of all subjects traced (similar for all the three exposure groups).

Because information on disease outcome in our study was collected retrospectively and 28.8% of the women did not respond to the questionnaire, selection and/or reporting bias may have occurred. If women in the ‘low’ or ‘high’ responder groups were more inclined to respond to the questionnaire and/or were more likely to report the occurrence of diseases, this might have led to an over-representation of disorders in the ‘low’ and/or ‘high’ responder group. In two hospitals, we collected data for all eligible women in the cohort (instead of only for women who responded to the health survey questionnaire). When we analysed the number of retrieved oocytes in the first IVF cycle for non-responding women in these two hospitals, the percentages of women with a low, normal and high number of oocytes were 21, 65 and 14% respectively. This is very similar to the proportions of ‘low’, ‘normal’ and ‘high’ responders to IVF treatment among women who completed the questionnaire in these two hospitals (18, 66 and 16% respectively), which renders selection bias by response to IVF treatment unlikely. Misclassification of disease outcomes is a potential problem in our study, as we had to rely on self-report in the questionnaires. When assessing the health outcomes in a cohort, linkage with a national morbidity registry is preferable compared with self-reported information; however, no registries of the disorders under study exist. Furthermore, we consider it unlikely that the validity of reporting of the disorders of interest would differ between the three responder groups. With regard to the analyses, we are aware of the fact that some were based on small numbers of events, and as a result some positive findings may have been due to chance.

Although we had to exclude 3227 women in the present study because medical record data were not collected, it is highly unlikely that this caused any selection bias. The hospitals where we could not perform medical record data collection at the end of the study (due to lack of funding) were randomly determined, since the order of the hospitals was determined by the distance from the abstractors’ town of residence to the hospital.

In summary, after a follow-up of almost 5 years after completion of the last IVF treatment, we found no evidence of an elevated risk of gynaecological disorders in women with a high number of retrieved oocytes at IVF in The Netherlands. However, we did find slightly elevated risks of uterine leiomyoma and surgically removed ovarian cysts for women with a low number of retrieved oocytes. Therefore, this potential association should be explored further in future studies of IVF-exposed populations. Definite conclusions regarding other hormonal disorders require a larger number of cases. Due to the still rather young attained age in our cohort, further follow-up is warranted.


    Acknowledgements
 
We are greatly indebted to the participants of the OMEGA project. This study would not have been possible without the efforts of all women who participated. We are especially grateful to the research assistants M.Schippers, E.J.de Boer. I.M.Versteegden, S.Braak, A.H.W.van den Belt-Dusebout, G.M.Plas, I.van Gils and I.Verburg for abstracting data from the medical files in the participating hospitals. We would also like to thank the medical registries of the participating clinics for making patient selection possible; and all attending physicians for providing access to their patients’ medical files. This study was supported by grants from the Dutch Prevention Fund (28-2540) (40%), the Dutch Ministry of Health (15%), Organon Netherlands BV (8%) and internal funding from The Netherlands Cancer Institute (37%).


    FOOTNOTES
 
* The OMEGA project group included the following persons: M.Kortman, MD and E.R.te Velde MD PhD (University Medical Center, Utrecht), N.Macklon MD PhD (Erasmus University, Medical Center, Rotterdam), C.A.M.Jansen MD PhD (Diaconessenhuis, Voorburg), R.A.Leerentveld MD PhD (Isala klinieken, Zwolle), W.N.P.Willemsen MD PhD (Academic Hospital Nijmegen, St Radboud), R.Schats MD PhD (Academic Hospital, Free University, Amsterdam), N.Naaktgeboren PhD and F.M.Helmerhorst MD PhD (Leiden University Medical Center), R.S.G.M.Bots MD PhD (St Elisabeth Hospital, Tilburg), A.H.M.Simons MD (Academic Hospital, Groningen), H.V.Hogerzeil MD PhD (Academic Medical Center, Amsterdam), J.L.H.Evers MD PhD (Academic Hospital, Maastricht) and P.A.van Dop MD PhD (Catharina Hospital, Eindhoven). Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Abulafia, O. and Sherer, D.M. (2000) Angiogenesis of the ovary. Am. J. Obstet. Gynecol., 182, 240–246.[ISI][Medline]

Braga, C., Negri, E., La Vecchia, C., Parazzini, F., Dal Maso, L. and Franceschi, S. (1996) Fertility treatment and risk of breast cancer. Hum. Reprod., 11, 300–303.[Abstract]

Chen, C.R., Buck, G.M., Courey, N.G., Perez, K.M. and Wactawski-Wende, J. (2001) Risk factors for uterine fibroids among women undergoing tubal sterilization. Am. J. Epidemiol., 153, 20–26.[Abstract/Free Full Text]

Cox, D.R. (1972) Regression models and life tables. J. R. Statist. Soc. Ser. B, 334, 187–202.

Davis, O.K. and Rosenwaks, Z. (1995) In vitro fertilization. In Keye, W.R., Chang, R.J., Rebar, R.W. et al. (eds), Infertility: Evaluation and Treatment. W.B. Saunders Co., Philadelphia, pp. 759–771.

deBoer, E.J., ten Tonkelaar, I., te Velde, E.G., Burger, C.W., Klip, H. and van Leeuwen, F.E. (2002) Low number of retrieved oocytes at IVF treatment is predictive of early menopause. Fertil. Steril., 77, 978–985.[CrossRef][ISI][Medline]

Fauser, B.C., Devroey, P., Yen, S.S., Gosden, R., Crowley, W.F. Jr., Baird, D.T. and Bouchard, P. (1999) Minimal ovarian stimulation for IVF: appraisal of potential benefits and drawbacks. Hum. Reprod., 14, 2681–2686.[Free Full Text]

Harris, R., Whittemore, A.S. and Itnyre, J. (1992) Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case–control studies. III. Epithelial tumors of low malignant potential in white women. Collaborative Ovarian Cancer Group. Am. J. Epidemiol., 136, 1204–1211.[Abstract]

Kelsey, J.L., Gammon, M.D. and John, E.M. (1993) Reproductive factors and breast cancer. Epidemiol. Rev., 15, 36–47.[ISI][Medline]

Klip, H., Burger, C.W., Kenemans, P. and van Leeuwen, F.E. (2000) Cancer risk associated with subfertility and ovulation induction: a review. Cancer Causes Control, 11, 319–344.[CrossRef][ISI][Medline]

Lashen, H., Ledger, W., Lopez-Bernal, A. and Barlow, D. (1999) Poor responders to ovulation induction: is proceeding to in-vitro fertilization worthwhile? Hum. Reprod., 14 , 964–969.

Modan, B., Ron, E., Lerner-Geva, L., Blumstein, T., Menczer, J., Rabinovici, J., Oelsner, G., Freedman, L., Mashiach, S. and Lunenfeld, B. (1998) Cancer incidence in a cohort of infertile women. Am. J. Epidemiol., 147, 1038–1042.[Abstract]

Mosgaard, B.J., Lidegaard, O., Kjaer, S.K., Schou, G. and Andersen, A.N. (1997) Infertility, fertility drugs and invasive ovarian cancer: a case–control study. Fertil. Steril., 67, 1005–1012.[CrossRef][ISI][Medline]

Navot, D. and Bergh, P.A. (1993) Ovarian hyperstimulation syndrome: a practical approach. Amsterdam, The Netherlands. Proceedings of the 8th Reinier de Graaf Symposium.

Parazzini, F., Negri, E., La Vecchia, C., Moroni, S., Franceschi, S. and Crosignani, P.G. (1997) Treatment for infertility and risk of invasive epithelial ovarian cancer. Hum. Reprod., 12, 2159–2161.[Abstract]

Parazzini, F., Negri, E., La Vecchia, C., Moroni, S., Polatti, A., Chiaffarino, F., Surace, M. and Ricci, E. (1998) Treatment for fertility and risk of ovarian tumors of borderline malignancy. Gynecol. Oncol., 68, 226–228.[CrossRef][ISI][Medline]

Parazzini, F., Pelucchi, C., Negri, E., Franceschi, S., Talamini, R., Montella, M. and La Vecchia, C. (2001) Use of fertility drugs and risk of ovarian cancer. Hum. Reprod., 16, 1372–1375.[Abstract/Free Full Text]

Potashnik, G., Lerner-Geva, L., Genkin, L., Chetrit, A., Lunenfeld, E. and Porath, A. (1999) Fertility drugs and the risk of breast and ovarian cancers: results of a long-term follow-up study. Fertil. Steril., 71, 853–859.[CrossRef][ISI][Medline]

Ron, E., Lunenfeld, B., Menczer, J., Blumstein, T., Katz, L., Oelsner, G. and Serr, D. (1987) Cancer incidence in a cohort of infertile women. Am. J. Epidemiol., 125, 780–790.[Abstract]

Rossing, M.A., Daling, J.R., Weiss, N.S., Moore, D.E. and Self, S.G. (1994) Ovarian tumors in a cohort of infertile women. N. Engl. J. Med., 331, 771–776.[Abstract/Free Full Text]

Rossing, M.A., Daling, J.R., Weiss, N.S., Moore, D.E. and Self, S.G. (1996) Risk of breast cancer in a cohort in infertile women. Gynecol. Oncol., 60, 3–7.[CrossRef][ISI][Medline]

Rothman, K.J. and Greenland, S. (1998) Modern Epidemiology. Lippincot-Raven, Philadelphia.

Schwartz, S.M. (2001) Invited commentary: studying the epidemiology of uterine leiomyomata—past, present and future. Am. J. Epidemiol., 153, 27–29.[Abstract/Free Full Text]

Shu,X.O., Brinton,L.A., Gao,Y.T. and Yuan, J.M. (1989) Population-based case–control study of ovarian cancer in Shanghai. Cancer Res., 49, 3670–3674.[Abstract]

Shushan, A., Paltiel, O., Iscovich, J., Elchalal, U., Peretz, T. and Schenker, J.G. (1996) Human menopausal gonadotropin and the risk of epithelial ovarian cancer. Fertil. Steril., 65, 13–18.[ISI][Medline]

Simons, A.H.M. (ed.) (1998) NVOG-richtlijn: het ovarieel hyperstimulatiesyndroom. [Dutch]. Nederlandse Vereniging voor Obstetrie en Gynaecologie, Utrecht, 11, 1–8.

Surrey, E.S. (1995) Hyperstimulation syndrome. In: Keye,W.R., Chang,R.J., Rebar, R.W. et al. (eds), Infertility: Evaluation and Treatment. W.B. Saunders Co., Philadelphia, pp. 145–153.

Venn, A., Watson, L., Lumley, J., Giles, G., King, C. and Healy, D. (1995) Breast and ovarian cancer incidence after infertility and in vitro fertilisation. Lancet, 346, 995–1000.[ISI][Medline]

Venn, A., Watson, L., Bruinsma, F., Giles, G. and Healy, D. (1999) Risk of cancer after use of fertility drugs with in-vitro fertilisation. Lancet, 354, 1586–1590.[CrossRef][ISI][Medline]

Whittemore, A.S., Harris, R. and Itnyre, J. (1992) Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case–control studies. II. Invasive epithelial ovarian cancers in white women. Collaborative Ovarian Cancer Group. Am. J. Epidemiol., 136, 1184–1203. [Abstract]

Submitted on July 2, 2002; resubmitted on October 3, 2002; accepted on May 14, 2003.