Role of poor semen quality for current infertility and future fertility rates – lessons from the clinic and current population studies

Abstract

In this discussion we shall consider the issues of semen quality and if it has a significant impact on human fertility in Europe in the future? In the extreme case of sterility a man cannot father a child even after medical intervention. The more likely situation is subfecundity causing a longer waiting time to conception. Fecundability is the monthly probability of conception, but does subfecundity cause an overall reduction in fertility if couples just have to wait a little longer to have their children? Period fertility is the number of births in a given calendar year in relation to the number of women of fertile age. Period fertility can be affected negatively by the ‘tempo effect’ whereby couples have the same number of children over their lifetime, but childbirth is postponed. This results in a higher mean age of childbearing. The postponement may be voluntary or due to subfecundity. A reduction in period fertility due to the tempo effect means that the delayed births are lost from the age structure permanently (unless the mean age of childbearing falls again) resulting in fewer children and hence more rapid population ageing. The quantum of fertility is the number of children a woman has over her lifetime. This can also be affected by a longer waiting period because there will be a lower probability of conception in older women, and there will also be an increased risk of discontinuity of the relationship (couples breaking up). There may also be a change in intention perhaps because they are frustrated and stop trying for a pregnancy, or a new job situation or moving to a new location may diminish the desire for children. We must ask the biologists to help explain the apparent seasonal fluctuations in fertility. Data from Austria demonstrate lowest birth rates in the months of October, November and December. Demographers and social scientists would explain the low birth rates at the end of the year by seasonality of sexual activity and different habits, but there may be a biological explanation due to sperm quality, or changing male and female factors with seasons. There may be a reason for the causal relationship between the two phenomena of seasonality of sperm quality and birth rates. Demographically we are facing a decline in fertility which is dramatic because of its intensity and its occurrence over a short period of time. In the past there have been large fluctuations in fertility rates, and when we see dramatic effects we should firstly look for behavioural explanations as being the most likely cause of rapid changes. There is an example which is not widely known outside demographic circles when in 1968 in Japan there was a sudden 15% drop in birth rate in a single year. This could not be explained by any disorder, famine, war, political crisis or economic crisis, and the only explanation was finally found to be astrological. This was the year of the horse and fire which astrologically occurs every 40 years. Boys born in this year are supposed to be bad for their future wives and therefore are unable to get married. The Japanese people were able to prevent children being born in order to circumvent this disadvantage. Birth rate increased again the following year (Biraben, 1968). The recent rise in the age of motherhood is seen in all European countries, but in France it comes after a period of decreasing maternal age. The maternal age for first birth, and all births, in France over the 20th century shows a decline in all births from 1900 to 1975, then a rise to 2000 back to 1900 levels (Daguet, 2002). However, the context has changed because the older maternal age in 1900 was predominantly for subsequent pregnancies whereas in 2000 the age of first birth is now much higher than a century ago. It is hard to believe that contemporary fertility patterns showing temporal and cross-sectional variations could be primarily a result of variation in sperm quality. In particular, the variations in European fertility levels are stunning, changes in fertility have occurred very rapidly over time, and fertility levels vary systematically with proximity determinants such as use of contraceptives, prevalence of marriage and delays of child bearing. Nevertheless, while variation in sperm quality may not be a driving force of contemporary fertility patterns, other biological influences, and the interaction between biological factors and the socioeconomic context, may nevertheless be important for understanding contemporary fertility. Documenting such influences is part of my research into the ‘biodemography of fertility’ that is based on behavioural genetic models and which has showed convincingly that the role of genetic influences on fertility has increased in recent cohorts. For instance, the variation in fertility outcomes in younger cohorts is becoming increasingly determined by genetic factors occurring along two pathways. First, there are explicit biological mechanisms such as genetic variations which influence fecundity. Genetic influences on sperm quality also fall into this category. Second, there are behaviour-mediated biological influences such as individual variation in academic ability resulting in greater variations in education in comparison with previous years when women's involvement in higher or tertiary education was much lower. At present, we cannot disentangle these two pathways. In assessing the future role of these biological factors on fertility outcomes, it is important to note the interactions between biology, changes in behavioural patterns and decision processes. For instance, information about (genetically driven) variation in fecundity may have important impacts on the relevance for this variation on fertility outcomes. In particular, if individuals/couples were aware of their low fecundity at a relatively early age, they could adapt very well. An 18-year-old female who had knowledge of her very low monthly conception probability, and had 20 years to plan for one or two children, could manage very well especially if she had access to assisted reproduction technology including in vitro fertilization (IVF). The problem arises because many people do not find out about their low fecundity until relatively late in life. The implication of subfecundity is thus different for a man who learns about his poor sperm quality at the age of 35 or older when the window of opportunity for achieving desired fertility is markedly reduced. The problem about trends in semen quality is therefore amplified by socioeconomic trends which postpones childbearing towards the end of the female reproductive window, and also for males because there is no evidence for an increasing age gap between husbands and wives. It is unclear to what extent individual couples are informed about their expected fecundity with age, and how such information affects fertility trends. The problematic trends are the delays in childbearing combined with the trend towards subfecundity. Young couples should have more access to information about fecundity relatively early in life so that this could be taken into account to help them make their decisions accordingly. From a gynaecological point of view we must consider female in addition to male issues. A semen sample is a good parameter of male fertility but the process is more complex in the female requiring assessment of hormone levels, transvaginal scans, etc. We require information on factors which we know will affect female fecundity. We have acquired much knowledge on ovaries and fertility in the development of assisted reproduction. A young ovary has a large pool of follicles, but there are very few left by the age of 42 when fertility virtually stops. The polycystic ovary is a specific ovulatory disorder which is possibly due to fetal exposure to androgens. We may see a 32-year-old woman in the infertility clinic whose husband has poor semen quality, but, in addition, an ultrasound scan shows a small ovary looking like a 42-year-old ovary. Although she has normal hormones and menstruation, she is likely to be very subfertile because of oocyte depletion. Many of the mechanisms of the testicular dysgenesis syndrome could perhaps be transferred to similar disturbances in the female. Infertility in women is not just the problem of chlamydia and blocked tubes, ectopic pregnancy, fibroids, etc. but there are many dysfunctions in the ovary associated with poor oocyte quality resulting in an ovulatory fertility problem. We must remember the female in infertile couples. I have looked at the effects of diethylstilboestrol which is the archetypical endocrine disrupter whose effects were first seen in the female (Herbst et al., 1971). For example, altered sex ratio is a parameter which is not applicable to an individual, but affects a population adversely if it varies significantly from the norm. Similarly, a small change in sperm concentration or a small delay in time to conception has little effect on an individual but can have population consequences. In our studies we were able to analyse retrospectively pesticides and phthalates only because we kept biological specimens. Storing samples of urine, and to a lesser extent blood, is easy because they are readily accessible in abundant quantities and we should increase our use of repositories for biological samples in order to look for aetiological factors. Age-specific fertility is under the influence of biological determinants and behavioural determinants. The behavioural determinants are under our free will and are usually stronger than the biological determinants, and act as confounding factors in the general epidemiological approach to characterizing the biological factors. Behavioural determinants can be measured individually in each particular person, but that is rarely possible for the biological determinants. Normally, calendar time is used as a proxy measure for the biological effects in an attempt to interpret the temporal variations. There is little scope for inferring biological determinants from simple descriptive studies of calendar time variations of age-specific fertility and pregnancy rates which will never provide proof nor evidence of disproof. Biological investigators must familiarize themselves in great detail with the behavioural determinants, otherwise they will never be able to control for those confounders. Imaginative and sophisticated epidemiological designs are required to identify the weak signal in the presence of massive behavioural confounding. The aim of the biologist must be to provide the epidemiologists with better criteria for measuring male fertility than sperm counts and pregnancy. There have been attempts to refine methods for diagnosing male fertility, but these were set back by the advent of intra cytoplasmic sperm injection (ICSI) which largely negated the definition of infertility on the basis of semen analysis. However, we must reassess this approach and try to find more robust tools to assess fertility. I suggest that an analysis of oxidative stress is an important factor, and also the integrity of DNA in the germ line which not only affects fertility but is also important for the health and well-being of the children. Oxidative stress and germ line DNA damage are correlated. A man's age is a significant factor in couple infertility. The sperm chromatin structure assay (Evenson et al., 1980, 2002) is a measure of sperm DNA fragmentation which shows that men in their early twenties have about 5% of their sperm population containing fragmented DNA. The levels of DNA fragmentation index (%DFI) increase in a near-linear fashion to age ≤80 (Wyrobek, unpublished data). Regression analysis shows an intersect between age and the 30% threshold DFI at ∼age 47. A highly significant correlation exists between the threshold %DFI of 30% and poorer pregnancy outcome by both in vivo (natural) fertilization and IVF, although the overall effect can be partly overcome by the IVF/ICSI procedures (Henkel et al., 2003, 2004; Bungum et al., 2004; Check et al., 2005; Evenson & Wixon, 2005). The %DFIs for men of similar ages is heterogeneous which may relate to antioxidant levels in semen in keeping with Dr Aitken's data on effects of oxidative stress. Some men in their twenties have very high levels of damage, whereas some men in their sixties have very low levels. There is a general hypothesis that DNA strand breaks are mostly due to ROS activity, which relates to the aging process, and the testis is not exempt. We have observed numerous cases in the reduction of men's sperm DFI values by elimination of exposure to reproductive toxicants, varicocele repair and the use of therapeutic dietary antioxidants. We know very little about the origins of DNA damage in the germ line and exactly what its consequences will be. Animal models have been helpful and we know that the ‘fertilized’ oocyte has tremendous potential for repairing DNA damage, but this is not perfect and persisting defects have the potential for producing mutations for the offspring which will impair development and the health of the offspring. We need more research on detailed mechanisms. Many of the assays in current use to look at sperm function, particularly DNA integrity, are generic assays without giving details of the specific nature of the DNA damage in the germ line. The cause of the damage is multifactorial and poorly understood.