If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Obstetrics and Gynecology, Akershus University Hospital, Lørenskog, NorwayInstitute of Clinical Medicine, University of Oslo, Oslo, Norway
Department of Obstetrics and Gynecology, Akershus University Hospital, Lørenskog, NorwayInstitute of Clinical Medicine, University of Oslo, Oslo, Norway
Institute of Clinical Medicine, University of Oslo, Oslo, NorwayDepartment of Reproductive Medicine, Oslo University Hospital Rikshospitalet, Oslo, Norway
Does a successful spontaneous pregnancy in the years close to natural menopause depend on age at menopause?
Design
This was a retrospective population-based study of 4157 parous postmenopausal women in Norway, born during the years 1925–1940. Data were obtained by two self-administered questionnaires in the HUNT2 Survey (1995–1997). We calculated the proportions of women who gave birth within 5 years and within 10 years prior to menopause both among all women, and according to categories of age at menopause.
Results
Overall, 2.7% (114/4157) of all women gave birth within 5 years, and 11.7% (487/4157) gave birth within 10 years, prior to menopause. Among women with menopause before the age of 45 years, 23.5% (81/344) gave birth within 5 years, and 55.5% (191/344) gave birth within 10 years, before menopause. Among the women with menopause at the age of 55 years or older, no women (0/474) gave birth within 5 years, and 0.2% (1/474) gave birth within 10 years, prior to menopause.
Conclusions
More than half of the women with menopause before the age of 45 years gave birth within the 10 years before natural menopause, whereas virtually no women with menopause at the age of 55 years or older did. Thus, the length of the sterile interval before natural menopause may vary by age at menopause.
). Postponing pregnancy until the age of 35–40 years may therefore reduce a woman's chance of giving birth, particularly if she will go on to experience early menopause.
The assumption of a fixed sterile interval prior to natural menopause independent of age at menopause is poorly documented. This assumption is mainly based on retrospective data with a comparison of the age distribution at last childbirth in one population with the age distribution of menopause in another (
). In these studies, the difference between mean age at menopause and mean age at last childbirth was 10 years. In addition, the poor success rate of assisted reproduction treatment in women older than 40 years (
). However, none of these studies accounts for the large individual variation in age at menopause. We are not aware of any studies of childbirth in the years close to natural menopause using individual data of age at last childbirth and also of age at menopause.
Using a study population of 4157 postmenopausal women in Norway, born during the years 1925–1940, we calculated the proportions of women who had had a successful pregnancy within 5 years and within 10 years prior to natural menopause, and studied whether the proportions with a successful pregnancy differed by age at menopause.
Materials and methods
Study design and recruitment
We performed a retrospective study using data from a population-based survey in Norway (the HUNT2 Survey). This survey aimed to include all inhabitants aged 20 years or older in the Nord-Trøndelag county of Norway during the years 1995–1997. Details of the HUNT2 Survey are described elsewhere (
Data were collected using two self-administered questionnaires, and only postmenopausal women who had given birth and had answered both questionnaires were eligible for our study. Of all women aged 20 years or older in the county of Nord-Trøndelag, 71% participated in the HUNT2 Survey and answered the first questionnaire. Of these, 87% also completed the second questionnaire that was answered by women aged 20–69 years. Thus, approximately 60% of women in this age group in the North-Trøndelag county answered both questionnaires (24,865 women). Women aged 60–69 years had the highest response rate.
We excluded women who had not given birth (n = 3554), women for whom information about the number of childbirths was lacking (n = 740) and women who had not reached menopause (n = 12,722) (Supplemental Figure). We also excluded women with missing or implausible values for age at last childbirth or age at menopause (n = 460). Additionally, we excluded women who had had both ovaries and/or the uterus surgically removed before natural menopause, or had missing information about age at such surgery (n = 1110). To avoid an overrepresentation of women with early menopause, we excluded menopausal women born after 1940 (n = 2122). Thus, 4157 women born during the years 1925–1940 were eligible to be included in our data analyses.
Study factors
The first questionnaire included questions about socio-demographic factors and health. The second questionnaire included questions about menstruation, childbirth and surgery on the ovaries or uterus. Information about the number of childbirths and age at last childbirth was based on the following questions: ‘How many times have you given birth?’, ‘List the year of each childbirth’ and ‘How old were you at your last childbirth?’
Information about age at natural menopause was obtained using the following questions: ‘Do you still have menstrual periods?’ (yes/no) and ‘If no, at what age did you have your last menstrual period?’ In the analyses, age at menopause was categorized as: under 45 (early menopause), 45–49, 50–54 and 55 or more years old.
The time interval between last childbirth and menopause was calculated by subtracting age at last childbirth from age at menopause and categorized as childbirth within 5 years (yes/no) and childbirth within 10 years (yes/no) before menopause. The number of previous childbirths was categorized as 0–1, 2–3 and more than - 3.
Statistical methods
We used kernel density estimation to illustrate the age distributions at last childbirth and at menopause. Within the above-defined categories of age at menopause, we calculated the mean number of childbirths, mean age at last childbirth, mean age at menopause and mean time interval between last childbirth and menopause, as well as the standard deviations (SD) of these values. Differences in means between categories were assessed by analysis of variance (ANOVA). A 5% level of statistical significance was chosen for all analyses.
We then calculated the proportions of women who gave birth within 5 years and within 10 years before menopause among all the women, and within the categories of age at menopause. We repeated these data analyses after excluding women who had used oral contraceptives or had undergone sterilization. We also performed separate analyses among women born in 1930 or earlier, as these women had particularly limited access to oral contraceptives or intrauterine devices during their reproductive period.
In supplemental analyses, we calculated the proportions of women with childbirth within 5 years and within 10 years prior to menopause according to the number of previous childbirths, and we repeated these analyses within the categories of age at menopause. All the data analyses were performed using Stata/SE version 14.2 (StataCorp, USA).
Details of ethics approval
The HUNT2 Survey was approved by the Regional Committee for Medical and Health Research Ethics and by the Norwegian Data Protection Authority. All participants signed an informed consent form. The present study was approved by the Regional Committee for Medical and Health Research Ethics on 13 March 2017 (reference number 2017/105 REK South-East D) and by the HUNT Research Centre Review Board on 28 November 2017 (reference number 2017/11178/TRS).
Results
The mean age of the women at data collection was 62.5 years (SD 4.2 years) (Table 1), and the mean number of childbirths was 3.2 (SD 1.3 childbirths). In total, 93.0% of the women had given birth to two or more children. Mean age at last childbirth was 31.7 years (SD 5.0 years), and mean age at menopause was 50.0 years (SD 4.1 years) (Table 1). Of all women, 12.7% had used oral contraceptives and 10.0% had undergone sterilization (in total 20.7%).
Table 1Descriptive characteristics of the study sample; parous women who had undergone natural menopause (n = 4157)
Figure 1A illustrates the distributions of age at last childbirth and age at menopause. In total, 6.3% (262/4157) of the women gave birth at the age of 40 years or older, and very few women gave birth at the age of 45 years or older (0.3%, 12/4157). Most childbirths at the age of 40 years or older were among women who reached menopause at the age of 50 years or older (66.8%, 175/262). Nonetheless, the distribution of age at last childbirth was similar across categories of age at menopause (Figure 1B, Supplemental Table A1). Thus, the proportion of women who had had their last childbirth at the age of 40 years or older was 4.1% (14/344) among women with menopause before the age of 45 years, and 6.6% (175/2670) among women with menopause at the age of 50 years or older. The mean time interval between last childbirth and menopause was 18.3 years (SD 6.3 years), and this time interval increased with increasing age at menopause (ANOVA, P < 0.001) (Supplemental Table A1).
Figure 1Distribution of age at last childbirth and age at menopause presented using kernel density estimation (n = 4157). (A) Age at last childbirth and age at menopause. (B) Age at last childbirth according to categories of age at menopause.
Among all women, 2.7% (114/4157) gave birth within 5 years prior to menopause, and 11.7% (487/4157) gave birth within 10 years prior to menopause. These proportions increased with decreasing age at menopause (Table 2). Among the 8.3% (344/4157) who reached menopause before the age of 45 years, 23.5% gave birth within 5 years, and 55.5% gave birth within 10 years, prior to menopause. Among the women who reached menopause at the age of 55 years or older, no women gave birth within 5 years, and only one woman gave birth within 10 years, prior to menopause. We found a similar pattern among the women who had not used oral contraceptives or undergone sterilization (Table 2). Among the women who were born in 1930 or earlier and had reached menopause at the age of 55 years or older (114/1199), no women gave birth within the 5 years, and only one woman gave birth within the 10 years, before menopause (Table 2).
Table 2Proportions of women with childbirth within 5 years and within 10 years prior to menopause among all women, and according to categories of age at menopause
Model 1 All women in the study sample, born 1925–1940 (n = 4157)
Model 2 Women with oral contraceptive use or sterilization excluded (n = 3298)
The proportion of women who had undergone childbirth within 10 years prior to menopause decreased with each remaining year to menopause (Figure 2), and only 1.0% (42/4157) of all women gave birth within the 2 years before menopause. However, among the women with menopause before the age of 45 years, 10% (34/344) gave birth within 2 years prior to menopause.
Figure 2Proportions of women who gave birth within each year during the 10 years prior to menopause among all women, and according to categories of age at menopause (n = 4157).
Additionally, the number of previous childbirths was related to childbirth in the years close to menopause. The higher the number of previous childbirths, the higher the proportion of women reporting childbirth within 5 years and within 10 years prior to menopause (Supplemental Table A2). This pattern was most prominent among women with menopause before the age of 45 years. In fact, 82.1% of the women with menopause before the age of 45 years and with more than three previous childbirths gave birth within 10 years prior to menopause.
Discussion
In this retrospective population-based study of 4157 parous women, born during the years 1925–1940, more than half of the women with menopause before the age of 45 years gave birth within 10 years prior to menopause, whereas this was true for less than 1% of the women with menopause at the age of 55 years or older.
We performed a study of women in a population with many children and with limited access to modern contraceptive methods during most of their reproductive period. Age at last childbirth and age at menopause may have been erroneously reported by some women in our study. However, studies report that recall for pregnancies and childbirths, and also for age at menopause, is fairly accurate (
). We have little reason to believe that possible erroneous reporting of age at menopause is related to age at last childbirth, and unsystematic errors in reporting may underestimate rather than overestimate associations (
). Nevertheless, some women may have been misclassified. We excluded women who stopped having menstrual bleeding because of surgical removal of both ovaries and/or the uterus. In addition, some women may have had medical conditions or treatments that could have caused early menopause (
). In additional analyses, we identified 55 women who had been diagnosed with cancer prior to menopause, and 844 women with an autoimmune disease (rheumatoid arthritis, ankylosing spondylitis, hypothyroidism and/or asthma). After excluding these women, we found virtually no changes in our results (data not shown).
We are aware of no previous population-based studies of childbirth in the years close to menopause, using individual data. Previous studies have compared the distribution of age at last childbirth in one population with the distribution of age at menopause in another population (
). One of these studies compared the distribution of age at last childbirth in a 19th-century Canadian natural fertility population with the distribution of age at menopause in a Dutch population of women born during the years 1911–1925 (
). The shapes of the distribution of age at last childbirth and age at menopause were almost identical, and the mean difference between the age distributions was approximately 10 years. A study from recent times of women who had undergone ovarian stimulation treatment found that women who responded poorly had a higher risk of reaching menopause within 10 years after treatment compared with women who responded well, independent of age at treatment (
). Based on these studies, it has been suggested that there is a fixed interval between the onset of subfertility and menopause, independent of age at menopause (
). Few women reach menopause before the age of 45 years. Thus, a possible shorter subfertile interval prior to menopause among women with early menopause than among women with late menopause will not easily be detected unless these groups of women are studied separately.
The number of oocytes that a woman is born with makes up her stock throughout life. Atresia of the ovarian follicles has already started before birth and continues throughout the reproductive lifespan. Menopause is assumed to occur when fewer than 1000 follicles remain (
). We found that many women who reached menopause before the age of 45 years gave birth close to menopause. Hence, they were fecund and had functional oocytes close to menopause. This observation could suggest that they had a high rate of follicle atresia from the time of last pregnancy until menopause. In fact, among women with menopause before the age of 45 years, the proportion of women with childbirth close to menopause was highest among the women with many previous childbirths. This observation suggests a high rate of ovarian follicle atresia prior to early menopause. Rapid ovarian follicle atresia prior to early menopause is also supported by a recent Dutch study of 111 women diagnosed with premature ovarian insufficiency (
). In that study, the median time interval from the last conception to the final menstrual period was 4 years.
Among women with menopause in their fifties, very few gave birth within the 10 years before menopause. This finding could be explained by an inability to give birth, lack of or infrequent sexual intercourse, or use of methods to prevent pregnancy and childbirth. We studied separately women who had not used oral contraceptives or undergone sterilization. In these analyses as well, very few women with menopause in their fifties gave birth within 10 years prior to menopause. Unfortunately, we had no information about the use of other birth control methods. Some women may have had an induced abortion at an advanced reproductive age, despite restricted legal access to pregnancy termination in Norway before 1979. However, the pregnancy termination rates (
) in women older than 45 years have been very low. These observations suggest that women with menopause in their mid-fifties may have a long sterile interval prior to menopause.
In addition to decreased number of ovarian follicles, oocyte quality also decreases with age (
). It is possible that the recruitment of ovarian follicles and the selection of a high-quality oocyte for ovulation is better in younger than in older women. Additionally, the meiotic divisions of the oocyte during ovulation or during the completion of the meiosis during the fertilization may fail more often in older than in younger women. Mitochondrial dysfunctions are assumed to play an important role (
). The low birth rates after the age of 45 years in women who still have menstrual periods may therefore suggest that the remaining oocytes are of insufficient quality for successful fertilization or for normal embryonic development.
In addition, impaired fallopian tube motility and lower receptivity of the endometrium could reduce the chances of successful pregnancy at a high age (
). Furthermore, successful conception is not determined by only the fecundity of the woman. A high age of the woman's male partner may reduce semen quality and thereby the couple's chance of a successful spontaneous pregnancy (
Among women with menopause before the age of 45 years, more than 50% had their last childbirth within the 10 years before menopause. This was true for less than 1% of the women with menopause at the age of 55 years or older. Thus, the length of the sterile interval prior to natural menopause may vary by age at menopause.
Acknowledgements
This work was supported by the South-Eastern Norway Regional Health Authority (grant number 2016112 to MSG) and by the Norwegian Cancer Society (grant number 6863294-2015 to E.K.B). The Nord-Trøndelag Health Study (HUNT Study) is a collaboration between the HUNT Research Centre (Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology [NTNU]), Nord-Trøndelag County Council, Central Norway Regional Health Authority and Norwegian Institute of Public Health.
The proportions of women with childbirth within 5 years and within 10 years prior to menopause according to number of previous childbirths, in all women and according to categories of age at menopause.
Supplemental FigureFlow chart of the study sample of 4157 women in the HUNT2 Survey, Norway, 1995–1997.
Marthe Sørli Gottschalk is currently a PhD student at the Department of Obstetrics and Gynecology, Akershus University Hospital, Norway. She obtained her MD degree at Pecs Medical School, Hungary and her International Master of Public Health at Hebrew University, Israel. Her research interests are reproductive epidemiology, female fertility and menopause.
Key message
This study showed that more than 50% of the women who reached menopause before the age of 45 years had had a successful pregnancy within 10 years prior to natural menopause. Among women with menopause in their mid-fifties, less than 1% had had a successful pregnancy within the 10 years before natural menopause.
Article info
Publication history
Published online: March 28, 2019
Accepted:
March 18,
2019
Received in revised form:
March 6,
2019
Received:
January 7,
2019
Declaration: The authors report no financial or commercial conflicts of interest.
Marthe Sørli Gottschalk is currently a PhD student at the Department of Obstetrics and Gynecology, Akershus University Hospital, Norway. She obtained her MD degree at Pecs Medical School, Hungary and her International Master of Public Health at Hebrew University, Israel. Her research interests are reproductive epidemiology, female fertility and menopause.
30346-3&id=gr1.jpg){kind=link}
30346-3&id=gr2.jpg){kind=link}
30346-3&id=mmc3.jpg){kind=link}