The 2004 British Society for Population Studies (BSPS) AGM took place in the Graham Wallas Room at LSE on Tuesday 6 July 2004 at 10:45am. All members were welcome to attend. The AGM was followed by a day meeting, which was open to members and non-members.
British Society for Population Studies Day Meeting
Late fertility - how late can you wait?
The day meeting consisted of six papers on different aspects of late fertility. The first three concentrated on medical aspects of late fertility, including artificial reproductive technology (ART). The final three papers covered demographic analyses of late fertility.
Professor Ovrang Djahanbakhch (Barts and the London School of Medicine, Queens Mary's University London) opened the conference by discussing age-related changes in human fertility. He gave an introduction to the biology of ovulation, in particular discussing the role of the lutenising hormone (LH) surge in ovulation.
Ovrang then looked at physiological changes that occur with regards to late fertility, in particular decreased levels of fertility. The number of oocytes a woman has are at a peak at birth and then this number decreases throughout life. The decline is particularly marked from the age of 37 onwards, and this is one of the factors in decreasing fertility at older ages. Ovrang showed that 30 per cent of fertilised eggs are lost within 2 weeks of conception and that only 30 per cent of fertilised eggs are carried to term. The risk of miscarriage increases with a woman's age, and a relationship also exists with male age. Raised risks are seen for women age 35 and over and for men age 40 and over, with couples comprising of a female age 35 and over and male age 40 or over having the highest risk of a miscarriage. There is a debate as to whether these relationships are due to egg quality or implantation problems. Studies looking at the success of in vitro fertilisation (IVF) with a woman's own eggs compared with donor eggs by age show a decline in success rates after age 39 for women using their own eggs, but no such age effect for women where donor eggs were used. This implies that egg quality is important in miscarriage rates.
Ovrang concluded by reminding the conference how far ART has come in 25 years. In 1979 many important things were unknown, including: the timing of ovulation, techniques for obtaining eggs, the size of eggs and the diameter of needle required to extract eggs. By 2004 much improved ARTs are now available, including such techniques as egg/ovarian tissue preparation and ovarian transplantation for cancer sufferers. However, he also wanted to remind those present that technology can't always solve fertility problems and that it is better to seek help earlier, because for example if a woman's ovarian reserve is too low then they are less likely to respond to IVF.
In the discussion following his paper Ovrang was asked that given there is clear variability in the cessation of ovulation between women, whether we any clearer on why this variation exists? In reply Ovrang pointed out that we know age at menarche also varies and that oestrogen is needed for ovulation and this varies within patients as well as between patients. However, the mechanisms that act to cause cessation of menarche and why timing of this varies are unknown and more research is needed on this topic.
Professor Alison Macfarlane (City University) gave a presentation on the Statistics on the use and outcome of techniques for medical management of subfertility in the UK: what we can count, what we don't know and what we need. She discussed the data available and the data needed for monitoring the outcome of the medical management of subfertility. The data sources currently available are civil registration of births and deaths, data from the Human Fertilisation and Embryology Authority (HFEA), prescription data and notifications of abortions to the Department of Health.
Alison looked at multiple birth rates, showing that they have been increasing since the mid-1970s. In particular they have increased amongst women aged 30 and over. The greatest increase has been among the small numbers of births to women aged 45 and over, where the rate has increased 3-4 fold. There was some in decrease in triplet and higher order multiple birth rates in the late 1990s and much more marked decreases in 2002, following guidelines from the Royal College of Obstetricians and Gynaecologists on the number of embryos that should be implanted.
In part to look at these trends further Alison then discussed, in detail, work Lisa Hilder and she have recently completed analysing data from HFEA, to provide backup for the Authority's new Code of Practice issues in January 2004. HFEA was established in 1991 and licences, monitors and regulates many aspects of ART treatments. HFEA maintains a formal register of information about donors, treatments and children born as a result of ARTs, and uses the data from it to monitor trends in practice and its outcome. Alison spoke a little about the data problems they encountered and presented results from HFEA data that they had 'cleaned' (at least to some extent).
The number of transferred embryos affects the likelihood of a clinical pregnancy and multiple birth, and this effect changes with age. Younger women who have three embryos transferred are much more likely to retain them all. For women over age 40 the percentage of embryo transfers that result in a clinical pregnancy is significantly lower than among younger women. Over age 40 there is a low rate of multiple births even with three embryos transfers but transferring three embryos results in an increased likelihood of any birth. The relative risk of having two or more live births is greater for each age group for three embryos transferred compared with two embryos. The data supports the conclusion that the transfer of three embryos should only be recommended for women aged 40 and over, because of their decreased probability of obtaining a clinical pregnancy.
Alison also discussed prescription data, which further illustrated the problems with current data sources for investigating the medical management of subfertility. Overall there has been a decrease in the number of GP prescriptions for fertility drugs, but this could be indicative of a number of different things. It could be that the number of prescriptions has truly dropped, that the number of people using the NHS for fertility treatment has dropped or that there has been an increase in prescriptions from hospitals rather than GPs. Furthermore prescription data do not contain information on whether the drugs are used and there is no link of the data to individuals so outcomes are unknown. Other data sources that may give some useful information include Hospital Episode Statistics, GP data systems and specialist surveys. Currently there is interest in trying to set up a cohort study of multiple births.
Alison then moved on to discuss data needs, which include: data on procedures outside HFE Act, data on private prescriptions, data about the use of any prescriptions, including diagnosis link to outcome, record linkage within the NHS and with data from private sector, and more surveys.
Dr Françoise Shenfield, (University College London Hospitals) presented on the Ethical dilemmas in ART. Francoise began by introducing some of the 'eternal' debates that exist in the field of reproduction. These include: the status of the embryo, embryo research, justice and access to healthcare, and concepts of life and personhood. For example, the legal definition of the embryo differs between countries and these different definitions affect other debates and issues. Recently a European Society of Human Reproduction and Embryology taskforce recommended that the term pre-implantation embryo be used rather than pre-embryo for embryos that have not been transferred.
Francoise then moved on to look at some of the current debates in the field. These include anonymity in gamete donation, gender selection, pre-implantation genetic diagnosis and allied methods, stem cell research and multiple pregnancies. The status of gamete donors is different in different countries. In Sweden, for example, this identification has been compulsory since 1985, whereas in France anonymity is enshrined in law. From next year donor anonymity will be removed in the UK. A study in 2000 in Sweden showed, that even in a country where donor anonymity does not exist, most parents had not informed their children that they were conceived with donor sperm. Studies have shown that there are no differences between children conceived from gamete donation, naturally conceived children or adoptive children in their assessment of school behaviour and family interaction. There are also differences between countries in who is allowed access to donor sperm. For example, single women and women in same sex unions are able to receive sperm donation in Spain, the UK and Belgium but not in France.
Pre-implantation diagnosis (PGD), which may be called 'pre-gravid', is the process of performing genetic analysis on pre-implantation embryos and selecting embryos on the basis of the results to ensure the embryo is not at risk of a serious genetic condition. The dilemmas for PGD are the same as for antenatal screening. The issue of eugenics is often raised with regards to PGD. However, Françoise made the point that the term eugenics should not be applied in this case because eugenics are applied to a population rather than to a single woman and her clinician. Allied methods are the combined use of PGD and other techniques such as HLA matching to select for an embryo whose cord blood containing stem cells could be used to help a sibling with a genetic disease. Françoise believed PGD to be an advancement in the field of ART, so long as people are informed and counselled throughout all of the steps.
There is much debate surrounding the issue of cloning. Cloning for reproduction has aroused strong opposition and is thought by most countries that it should be banned. However, opinions over cloning for therapeutic reasons are more divided. Therapeutic cloning is the use of cloned embryo stem cells to cure disease. This is associated with the current debate about the creation of embryos for research and problems of obtaining oocytes.
Amongst some of the questions that were asked in the discussion was if there is a tension between providing free access to IVF and encouraging women to have babies earlier to avoid the difficulties of subfecundity in later life? Françoise replied that the availability of free IVF doesn't appear to influence age of childbearing, for example IVF is free in France and women not having babies later than in UK where it is not free.
Steve Smallwood (Office for National Statistics) gave an overview of Late fertility in the United Kingdom: past history and future intentions. Over the last 30 years overall fertility, as represented by the total fertility rate (TFR), has declined. At the same time the contribution of different age groups' fertility to overall fertility levels has changed. Fertility rates of women aged 35-39 and aged 40 and over have doubled over the last two decades, in all of the countries of the United Kingdom except Northern Ireland. The percentage of the TFR that is determined by 30 year olds and over has risen, from 1982 to 2002 it rose from 26 per cent to 44 per cent. Associated with these changes has been increasing mean age at childbearing and at first birth, on both a cohort and period basis. These increases are projected to continue. First birth occurrence-exposure rates (first births divided by childless women) have also shifted on both a cohort and period basis from peaking for women aged in their 20s to a peak at ages in the early 30s.
The average family size of all women increased from 2.0 children per woman for the 1920 cohort to 2.4 for the 1940 cohort. Since then family size has declined to below replacement level at 1.98. However, this data includes childless women, if only women who have had children are analysed then the rises and falls in family size are much smaller.
Steve used information from the General Household Survey to look at fertility intentions for women in England and Wales and Great Britain. The survey asks women if they intend to have any more children, how many children they think they will have and how far into the future their next birth will be. The coverage of questions on number of children and time of next birth has changed over time. This makes comparison over time more difficult.
The GHS data shows that although intended family size has declined since the late 1970s women, on average, still intend to have two children. The exception to this is at age 36-38, where intended fertility is lower because women's intentions more closely reflect their achieved fertility. However, Steve presented evidence that fertility intentions were not good predictors of either fertility levels or trends.
The effect of postponement is seen in the fact that from the 1957-59 cohort to the 1966-68 cohort women still wanted the same number of children, on average, however at every age the younger cohorts have a smaller family size. At younger ages (below age 30) women who already have a child are more likely to intend to have a child within 3 or 3-5 years. GHS data also shows education level has an effect on timing of fertility. At ages under 30 women without higher education intended to have their births sooner than women with higher education.
Steve presented results from a simple study he had done to see whether fertility intentions provided by married women, who remained married, were more accurate predictors of fertility than for all women. This study looked at the intended fertility of a selected cohort in one GHS and the achieved fertility in a later GHS of women in this cohort who had remained married. It showed these women roughly achieved or exceeded their stated fertility intentions.
Professor Francesco Billari (Bocconi University, Milan) presented a paper on Pushing the age limit? Long term trends in 'late' childbearing (based on work in collaboration with Hans-Peter Kohler, Gunnar Andersson and Hans Lundström), this looked at very late fertility and idea of rectangularisation of first births. Demographers have increasingly been looking at extremes in populations, such as super centenarians. However, this has been mostly in the field of mortality rather than fertility. Although the numbers of women having children above age 40 is small, it is interesting to look at these 'extreme' cases.
Postponement of fertility is a general feature of populations in Europe and North America. However, the association of late age at first birth with lower fertility at a macro level is ambiguous. Though there is persuasive micro level evidence that postponing first births reduces total fertility. It seems that 'postponers' intend to have children but often don't achieve their intentions.
There are physiological factors that affect limits to fertility. Although new technologies may partially overcome age limits there is scepticism of realising fertility at old age. Recent modelling work by Leridon has shown that even with ARTs if all women started trying to conceive at age 35 14 per cent would remain childless and this rises to 36 per cent if women started trying at age 40. Various socioeconomic factors also affect fertility at older ages. Social norms and in particular age norms shape the limits of fertility. For example in France a survey showed that 70 per cent of women thought that the age limit to becoming a mother was age 40 or lower. Economic analyses have shown the most economically rational behaviour for women is the postponement of birth(s) until the perceived biological limit of fertility. In particular with recent developments suggesting it may be possible to predict menopause it may be that women could be told their fertility limit and then would apply such neo-classical economics to plan their birth(s).
This led into Francesco discussing the concept of rectangularisation of the first birth curve. Rectangularisation is the process whereby life course events become compressed into a smaller age range, so there is a lower variability in the ages that events occur at. For example in fertility this would be if 100 per cent of women were childless to age 40 and then all had children. Rectangularisation has occurred with mortality, as mortality rates at younger ages have declined and now nearly all mortality occurs at older ages. The opposing view to rectangularisation is that the diminishing impact of social norms and increasing heterogeneity of preferences and/or economic constraints may imply a de-standardisation of fertility, with higher variability.
Francesco looked at the evidence for rectangularisation of the births and first births, using Sweden as an example because of the long time series available for their birth data. At the beginning of the twentieth century there were a high absolute and relative number of births at age 40 and over and at age 45 and over. The numbers declined over the century until the 1970s when they started to increase again. Occurrence-exposure rates, for first births, show that since the 1970s at ages immediately close to 40 (40-43) fertility has been increasing, whilst fertility above age 45 is not clearly expanding, although extreme cases are becoming more visible due to ARTs. The cohort data shows the same pattern as this period data. There is no clear evidence that rectangularisation of first births is actually occurring. During the following discussion Francesco's stated that he believed rectangularisation to be more of a reference point rather than something that is actually occurring. Also in the discussion it was pointed out that births are not equivalent to deaths, as you can have more than one, therefore it may be better to think of a rectangularisation of motherhood rather than first births.
Laurent Toulemon (Institut National d'Études Démographiques) presented the last paper, which was titled Who are the later parents?. This looked at the characteristics of late mothers (over age 35) with regards to age, parity, union history and education.
Laurent presented information about the recent changes in French fertility
The increasing number of late births are a result of changes in the fertility trends (delay in first births, decreases in fertility rates at young ages and increases in fertility at old ages), but are also, in part, a reflection of the changing age structure of the population. There are increasing numbers of women at older ages and therefore the number of births at older ages will increase, in part, due to this.
Laurent talked about the relationship between later fertility and lower fertility and the macro and micro evidence for a relationship. There is strong evidence on a micro level for a relationship but Laurent presented two pieces of evidence that show there is not necessarily a real relationship between age at childbearing and fertility level. In France the increases in mean age at childbirth, from the 1970s, have not been associated with a decline in the probability of moving on to a next birth. In fact parity progression ratios have remained very stable over the last three decades, at all parities. Laurent also showed an international comparison that indicates there is little association between increasing mean age at first birth and lower fertility. The European countries that have had the largest increases in mean age at first birth are not the ones that have also had the largest declines in total fertility. This led Laurent to comment that you can't infer a macro relationship from micro level data. At the end of the presentation Laurent and Francesco had a discussion about the use and reliability of micro and macro data. Francesco made the point that micro level data is useful as reproductive decisions are taken at an individual level and such data is needed to inform couples about their fertility. Whilst Laurent argued that macro level data is needed to look at a country level over a long time period and that to base such views on micro data would result in misleading projections.
Laurent also presented work he has done looking at the characteristics of older mothers. The proportion of births that are occurring to women aged 35 or older has been increasing for all birth orders, but the biggest increase has been for first order births. Also older mothers (age 35 and over) are becoming more like younger mothers with regards to birth order characteristics. Previously old mothers were having high order births but these have now decreased and old mothers are more likely to be having a second or first order birth. The changes that have occurred since the 1960s with regards to birth order are more specific for women than they are for men.
There has been an increase in births that are first births for the union but not first births for the mother. Also late births are increasingly occurring in second unions, and this trend is even more pronounced for men. However, the increase in fertility in second unions is not specific to older parents. With regards to educational achievement women with the highest fertility rate over age 35 are women with a degree, previously this used to be women with no education. However, there are also more women in the population with a degree, so part of this change is due to a change in the structure of the population.
The conference ended with a discussion, which moved on to talking about the consequences and implications of late fertility. A further conference on this topic may be organised for 2005.
Report by Jessica Chamberlain
Office for National Statistics