6.1 Introduction

Over the past few decades, the average age of both mothers and fathers has risen noticeably in many post-industrial countries. In the United Kingdom (UK), for example, the average age of fathers in 1976 was 29.6 years, whereas by 2016 it had risen to 33.3 years (Figure 6.1). Notably, in the past two decades, the percentage of fathers aged 30 and over in the UK has increased from 59% in 1996 to 66% in 2006 [1]. Similar increases have been reported in the United States (US) [2]. Whilst there are obvious social and demographic reasons for these shifts, such as young men (and their partners) focusing on educational opportunities and career development in their teens and twenties, these trends also present challenges for those working in reproductive medicine. The average age of men and women presenting for infertility and assisted reproduction is also increasing, and management of the “older couple” is therefore becoming increasingly common. This chapter will review the available evidence of the effect of male age on: (i) semen quality, sperm function and sperm DNA damage; (ii) conception and pregnancy outcomes; (iii) outcomes of assisted reproduction; and (iv) the health of offspring.

Figure 6.1 The age of fathers (solid circles) and mothers (open circles) in the UK from 1975 to 2016. Redrawn from data published in [1].

6.2 Effect of Male Age on Semen Quality, Sperm Function, and Sperm DNA Damage

Much has been previously written about the impact of male age on semen quality and/or aspects of sperm function. This section will summarise the relevant literature on how male age may impact on macroscopic measures of sperm quality as observed at semen analysis, functional aspects of sperm, and the quality of sperm nuclear DNA.

With regard to macroscopic and microscopic measures of semen quality, perhaps the largest meta-analysis undertaken to date included semen quality data from 93,839 men taken from a total of 90 studies [3]. Critically, the analysis controlled for a number of important confounding factors that can directly impact on semen analysis parameters, including duration of sexual abstinence (i.e. ejaculates produced after a long abstinence will likely contain fewer motile sperm). The conclusions of the meta-analysis were that there was a statistically significant age-associated decline in many of the macroscopic and microscopic measures of semen quality, including semen volume, total motility, progressive motility, and percentage of sperm with normal morphology. However, no age-related decline in sperm concentration was observed, which suggests that age does not impact on the speed of spermatogenesis, but more upon the “quality” of sperm produced.

Unfortunately, almost no studies have examined the influence of male age on the functional aspects of sperm, such as their ability to undergo acrosome reaction, or bind to the oocyte. This is probably a reflection on how difficult it is to perform such studies as well as poor availability, lack of standardisation, and therefore the lack of use of sperm function tests by diagnostic laboratories. However, it is plausible that the sperm of older men may function less well in vivo or in vitro during assisted reproduction. For example, age-related changes to subtle aspects of sperm movement can be measured by Computer Aided Sperm Analysis (CASA) [4]. Furthermore, a broad spectrum of alterations at the transcriptional and translational level have been described to occur in sperm of older men, which may impact on molecular mechanisms and function [5]. It is recognised that the acrosome reaction is impaired in the spermatozoa from obese men [6]. Since male obesity generally increases as men age, there is some validity in concluding that the sperm of older men may be similarly compromised. But whether this is independent from, or a consequence of, the ageing process is currently unknown.

In comparison to sperm function, much more has been written about the quality of sperm nuclear DNA in older men. For example, a meta-analysis including data from 10,220 men where information about sperm DNA fragmentation was available concluded that the number of sperms with unfragmented DNA (i.e. undamaged) generally decreases as men age [3]. However, given the controversy about sperm DNA fragmentation testing, and the array of tests that can be used to examine this [7], little could be concluded about the type of damage. This is discussed in detail in Chapter 4. It is also uncertain whether male age contributes to the risk of sperm aneuploidy, although age-related alterations in sperm DNA methylation patterns have been reported [8], as well as altered protamine and miRNA expression in the sperm from older men [9]. The clinical significance of this is unclear.

6.3 Effect of Male Age on Conception and Pregnancy Outcomes

Diagnostic semen analysis is itself acknowledged to have limitations in its prediction of male fertility and there are no nomograms to show whether age-related changes in semen quality are likely to be clinically significant. Similarly, with tests of sperm function or measures of sperm genetic integrity, there is little consensus on what these data mean. Recent analysis suggests that a change in semen quality occurs around the age of 40 [4], but it is difficult to be precise and it is more likely that age-related changes to any aspect of semen quality or sperm function are gradual and incremental. For the practising clinician or embryologist it is therefore difficult to conclude whether age-related changes to semen quality, sperm function, or sperm nuclear DNA are truly significant to the probability of natural conception or the success of assisted reproduction. However, given the deterioration in semen quality and increase in sperm DNA damage associated with advancing male age, logic would suggest that male fecundity is likely to reduce as a man gets older. In reality, the effect of male age on fertility is controversial, not least because there is a paucity of studies that have examined this in a robust way.

Perhaps the first one to do so used the Avon Longitudinal Study of Parents and Children (ALSPAC). ALSPAC is a world-leading birth cohort study, charting the health of 14,500 families in the Bristol area (www.bristol.ac.uk/alspac/). A study of 6,524 couples from the cohort examined the effect of male age on time to conception (TTC), which was used as an index of fecundity [10]. The results demonstrated a significant association between advancing male age and length of TTC. A logistic regression model was used to take account of other variables that may have had a significant influence (e.g. female age). After adjustment, paternal age remained highly significantly associated with conception within 6 or 12 months (p < 0.0001): the average age of the men who took more than 6 months to impregnate their wives was 31.8 ± 5.75 years compared with 30.8 ± 5.27 years in men who took 6 months or less (p < 0.0001). Men who took more than 12 months were also significantly (p < 0.0001) older (32.6 ± 5.91 years) than men who took 12 months or less to conceive (30.9 ± 5.32 years). In summary, older men took longer to conceive naturally. An alternative concept is that it is a couple’s cumulative age, rather than just paternal age, that defines fertility potential. A retrospective study of 6,188 European women [11] identified a paternal age of 40 years or more to be a risk factor for infertility (TTC>12 months), but only where female age was 35 years or more (odds ratio [OR] 2.99 95% CI: 1.82, 4.91). Similarly, a study of 782 healthy couples observed a decrease in the daily probability of conception where the couple was composed of a woman 35–39 years old and a man in his late thirties or older [12].

As men age, testicular function and metabolism deteriorates [13]. The testis undergoes age-related structural and cellular changes, including loss of volume, narrowing of seminiferous tubules, and a decrease in number of germ cells, Leydig cells, and Sertoli cells. Serum testosterone levels also decrease with age, with resultant effect on the hypothalamic-pituitary-gonadal (HPG) axis. Therefore, whilst age-related changes to semen quality, sperm function, or the quality of sperm DNA may explain decline in male fertility (see above), another plausible explanation is that older men may simply be less interested in sex or suffer increased levels of erectile dysfunction (ED). Both of these will clearly affect TTC and pregnancy rates. For example, in the Massachusetts Male Aging Study (MMAS), sexual function and coital frequency were assessed in 1,290 men. The probability of having severe ED increased three fold and the probability of moderate ED increased two fold between the ages 40 and 70. The annual incidence of ED increased with each decade of age and was 12.4 cases per 1,000 man-years (95% CI: 9.0, 16.9), 29.8 (95% CI: 24.0, 37.0), and 46.4 (95% CI: 36.9, 58.4) for men 40 to 49, 50 to 59, and 60 to 69 years old, respectively [14]. In the same cohort, followed for an average of 9 years, coital frequency was assessed in 1,085 men. After adjusting for baseline sexual function, men engaged in sexual activity on average 6.5 times per month prior to age 40. This frequency decreased by 1 to 2 times per month after age 50 and by another once or twice per month after age 60 [15].

In addition to delay in conception, male age also appears to have an impact on adverse pregnancy outcomes, including pregnancy loss [16]. For example, a prospective study of 5,121 natural conceptions found an adjusted hazard ratio of 1.27 (95% CI: 1.00, 1.6) for first trimester miscarriage for partners of men 35 years and older, compared to those whose partners were younger than 35 years [17]. An analysis of all live and stillbirths in Denmark (1994–2010) identified paternal age as a risk for late pregnancy loss (22+ gestational weeks). Over 75,000 births were fathered by men of more than 40 years old in this large dataset. After meticulous adjustment for maternal age, increasing paternal age was found to significantly increase risk of stillbirth. Compared with offspring of fathers aged 32 years, the risk was 1.23 in offspring of fathers aged 40 years and 1.36 in offspring of fathers aged 50 years [18].

6.4 Effect of Male Age on Outcomes of Assisted Reproduction

One of the first studies to robustly investigate the relationship between male age and outcomes of assisted reproduction examined data from the French National IVF Registry, submitted by 59 IVF centres [19]. By examining the IVF outcomes of 1,938 men on the register whose partners were totally sterile with bilateral tubal occlusion or absence of both tubes, they concluded that the risk of failure to conceive following IVF clearly increased with both maternal and paternal age. However, critically, after taking into account the possibility that the paternal age effect may differ according to maternal age (i.e. generally couples tend to partner with someone of a similar age), they were able to show that once men were older than 40 years of age, the risk of failing to conceive doubled if the female partner was 35 to 37 years old (OR 2.0 [95% CI: 1.1, 3.6]) and increased over five fold if she was above 41 years old (OR 5.74 [95% CI: 2.16. 15.23]). This suggested that there was a male age effect independent of other variables, which influenced the outcome of assisted reproduction. A retrospective study of 17,000 stimulated intrauterine insemination cycles also demonstrated a highly significant effect of paternal age. Following correction for maternal age, a significant reduction in pregnancy rate was seen in couples with older men (12.3% per cycle where paternal age <30 years, 9.3% per cycle where paternal age >45 years; p<0.001). Partners of men older than 45 years also had significantly higher miscarriage rates; 32.4% compared with 13.7% in couples with men younger than 30 years (p < 0.001) [20].

Further studies using egg donation cycles to control for the influence of maternal age have also examined the association between male age and the outcome of assisted reproduction, but with mixed conclusions. For example, a study of 1,023 men found that there was a significant increase in pregnancy loss and a decrease in live birth rate and blastocyst formation rate in men above the age of 50, but, interestingly, there was no significant difference in implantation rate, pregnancy rate, or measures of early embryo development [21]. This is in contrast to another study that found no effect of male age on the outcome of 1,083 donor oocyte cycles [22]. It is difficult to see how two studies of similar size and design reached such different conclusions and this perhaps reflects why there remains a lack of clarity in this area. Interestingly, neither of these studies discriminated between cycles where fertilisation was achieved through traditional IVF versus intra-cytoplasmic sperm injection (ICSI) and it is possible that the outcome of these treatments may be different. However, a narrative review examining 10 studies involving over 7,000 IVF and ICSI cycles concluded that there was “insufficient evidence to demonstrate an unfavourable effect of paternal age on assisted reproduction technology (ART) outcomes” [23]. Nonetheless, they also noted that most studies were retrospective and there was significant heterogeneity in design, for example, entry criteria and outcome reporting. Certainly paternal age may not be relevant when ICSI and good-quality oocytes are used. A retrospective study of 4,887 donor oocyte cycles showed no effect of male age on biochemical, clinical, ongoing pregnancy rates as well as take-home baby rates, where egg donors were less than 35 years old [24].

The reality is that there remains a lack of well-designed, prospective studies examining the impact of paternal age on ART outcomes, and the literature continues to be contradictory on this point. In recent years, there has been a shift in focus towards understanding the influence of male age on sperm DNA quality and whether this in itself is a critical factor for reproductive outcome, more specifically miscarriage and early pregnancy loss. It is widely acknowledged that there is a general trend for sperm DNA to show more signs of damage as men get older (see above) and meta-analysis shows that there is a relationship between higher damage and early pregnancy loss, at least by some measures of sperm DNA quality [25]. However, there remains a notable lack of consensus and considerable confusion about the best methods to test for sperm DNA damage, as well as how to interpret results [7], so it is perhaps not surprising that there is no clear answer here also (see Chapter 4). However, it has been proposed that the quality of sperm DNA (however it is measured and assessed) may have an important role in determining the health and prospects of any children born and this will be reviewed in Section 6.5.

6.5 Effect of Male Age on Child Health

Continuous division of the male germ cell line results in an increase in frequency of mutations in spermatozoa across the male life course. For example, a longitudinal study of 78 Icelandic parent-offspring trios demonstrated doubling of paternal mutations every 16.5 years, an average of 2 per year [26]. Although rare, paternal age effect (PAE) disorders are autosomal dominant, and caused by mutations in five genes (FGFR2, FGFR3, HRAS, PTPN11, RET). PAE disorders are characterised by congenital skeletal deformities and retarded growth, cardiac disorders, skin hyperpigmentation, and cancer susceptibility, and include Apert, Crouzon, Pfeiffer, and Muenke syndromes, achondroplasia, Costello and Noonan syndromes, as well as multiple endocrine neoplasia (MEN) type 2 A and type 2B [27].

Age-related genetic and epigenetic changes in spermatozoa are also postulated to cause a variety of other diseases in the resulting offspring. Although “congenital abnormality” is a poorly defined concept, encompassing very heterogeneous conditions both genetically determined and of developmental origin, the association with advanced paternal age (APA) is supported (on the whole) by large data sets. For example, a retrospective cohort study of 5,213,248 subjects (USA birth registrations 1999–2000) assessed the association between paternal age and birth defects [28]. A total of 77,514 birth defects were recorded in this study cohort (1.5%) and multiple logistic regressions were used to estimate the independent effect of paternal age on all birth defects, as well as 21 specific defect groups. In summary, APA was associated with increased risk of heart defects, tracheo-oesophageal fistula, oesophageal atresia, and musculoskeletal abnormalities, as well as Down syndrome and other chromosomal abnormalities. However, the overall risk of birth defect for older fathers was small. A further study examining the associations between paternal age and birth defects examined data from the National Birth Defects Prevention Study (NBDPS), a large multicenter case-control study (over 15,000 cases) designed to investigate genetic and environmental risk factors for major birth defects in the USA [29]. Paternal age was found to be a significant contributory risk factor for certain birth defects, whereas others were corrected for by maternal age. Elevated odds ratios for each year’s increase in paternal age were found for cleft palate (OR 1.02 95% CI: 1.00, 1.04), diaphragmatic hernia (OR 1.04 95% CI: 1.02, 1.06), right ventricular outflow tract obstruction (OR 1.03 95% CI: 1.01, 1.04), and pulmonary valve stenosis (OR 1.02 95% CI: 1.01, 1.04). It is worth noting that whilst the incremental increase in risk per year does not appear large, the difference in odds faced by men several years apart in age is potentially substantial. For example, a 40-year-old father would have twice the odds of having a child with diaphragmatic hernia compared with a child whose father was 20 years old. Moreover, the risk doubles again by age 60. A Danish prospective cohort study of 1,575,521 live-born children born between 1978 and 2004 assessed the relationship between APA and the risk of childhood death in children [18]. Compared to children born to fathers aged 30–34 years, a statistically significant excess risk of death was found for children born to fathers aged 40–44 years and 45+ years old. Increase in paternal age significantly increased the hazard risk of death for children under 5 years old (HR 1.24 [95% CI: 1.0, 1.53] for fathers aged 40–44 years; HR 1.65 [95% CI: 1.24, 2.18] for fathers 45+ years old). This was primarily attributed to fatal congenital anomalies, malignancies, and external causes.

Many studies show a clear epidemiological association between increased paternal age and schizophrenia, autism, and dyslexia. However, although the association is now well established, the underlying mechanisms remain equivocal and are likely due to both: (i) inherited genetic factors in couples where an older male becomes a father; or (ii) de novo genetic changes in paternal gametes that arise as a consequence of ageing [30]. This is not specific to assisted reproduction. A population-based study of all individuals (2,615,081) born in Sweden between 1973 and 2001 showed that children whose fathers were over 45 years old had a 3.45 times increased risk of autism, 13.1 times increased risk of attention-deficit/hyperactivity disorder, twice the increased risk of psychosis, 24.7 times increased risk of bipolar disorder, 2.7 times increased likelihood of suicide attempts, and 2.4 times increased risk of substance use problems, compared with offspring born to fathers 20–24 years old [31].

Finally, APA has also been shown to be associated with negative effects on both maternal health as well as those of the offspring. A recent large population-based cohort study of 40,529,905 live births between 2007 and 2016 in the US showed that higher paternal age was associated with an increased risk of premature birth, low birth weight, and low Apgar score [32]. For example, infants born to fathers aged 45 years or older had 14% higher odds of premature birth (OR 1.14, 95% CI: 1.13, 1.15) and 18% higher odds of seizures (OR 1.18, 95% CI: 0.97, 1.44) compared with infants of fathers aged 25 to 34 years. Gestational diabetes was 34% higher (OR 1.34, 95% CI: 1.29, 1.38) in mothers with the oldest partners. To estimate the contribution of APA, the data was recalculated for a scenario in which all fathers were younger than 45 years. In this model, 13.2% of premature births and 18.2% of gestational diabetes were estimated to be specifically attributable to APA.