Livebirth rates are influenced by an interaction between male and female partners' age: analysis of 59 951 fresh IVF/ICSI cycles with and without male infertility

The effect of male partner’s age on reproduction is less understood than that of the female. Advanced paternal age (APA) has been found to be associated with neurodevelopmental and psychiatric disorders including autism and schizophrenia in the offspring ( Malaspina et al. , 2001 ; Hultman et al. , 2011 ; Kong et al. , 2012 ; Lan et al. , 2021 ). The association of advanced male age and the risk of embryo aneuploidy is less compelling ( Wyrobek et al. , 2006 ; Yang et al. , 2007 ). Following adjustment for female partner’s age, older male partner’s age has been shown to delay natural conception ( Hassan and Killick, 2003 ) and increase the risk of early pregnancy loss ( Slama et al. , 2005 ). However, the impact of male’s age in relation to female partner’s age on IVF or ICSI outcome remains inconclusive. APA has been reported to be negatively associated with fertilization rates ( Aboulghar et al. , 2007 ; Oluwayiose et al. , 2021 ) and embryo development in IVF with donor oocytes from young women ( Frattarelli et al. , 2008 ; Luna et al. , 2009 ). However, no significant difference in the euploidy rates was found whether the embryos were derived from a male partner younger or older than 38 years, after adjusting for female partner’s age ( Kim et al. , 2019 ). With regard to pregnancy outcomes, a study found declining pregnancy rates with APA in oligospermic men, but not when men are normozoospermic ( Ferreira et al. , 2010 ). Several studies showed that APA lowered the conception rates or live birth rates (LBRs), independent of maternal age: some studies reported the cut-off age to be over 40 years ( de La Rochebrochard et al. , 2006 ; Horta et al. , 2019 ), others reported it to be over 45 years ( Marsidi et al. , 2021 ) or over 50 years ( Morris et al. , 2021 ). In contrast, other older studies failed to find such association ( Spandorfer et al. , 1998 ; Aboulghar et al. , 2007 ). Donor oocyte cycles are the way to minimize the effect of female age on the IVF outcome and several studies were designed to explore the impact of male partner’s age on donor-recipient cycles. Noticeably, most of these studies did not find any influence of paternal age on the pregnancy outcomes even when adjusted for recipient’s age ( Whitcomb et al. , 2011 ; Begueria et al. , 2014 ; Capelouto et al. , 2018 ). A few of them reported a negative influence on LBRs when donor oocytes were inseminated with sperm from men of older age ( Frattarelli et al. , 2008 ; Robertshaw et al. , 2014 ; McCarter et al. , 2021 ). A systematic review including 12 retrospective studies till early 2015 concluded that advancing paternal age was not associated with adverse oocyte donation programme outcome; however, the overall quality of studies was low ( Sagi-Dain et al. , 2015 ). The conflicting results on the effect of male age on IVF outcomes are mainly attributed to confounding factors that may have strong influence on the pregnancy outcomes. Age of the female is probably the most important one. Multivariate analyses after adjusting for the age of the female partner in the majority of the studies reported a negative impact of APA on pregnancy outcomes ( Humm and Sakkas, 2013 ), while no effect was found in most of the studies in young donor oocyte cycles ( Sagi-Dain et al. , 2015 ). There is some suggestion that male partner’s age does not impact the pregnancy outcome when the female partner is young ( Spandorfer et al. , 1998 ; Wu et al. , 2016 ). Severe male-factor problems may affect pregnancy outcome, regardless of the age-related influence. The evidence however is conflicting; e.g. Ferreira et al. (2010) found lower pregnancy rates with APA only in men with oligospermia and there was no difference among men with normal sperm parameters, while the study by Horta et al. (2019) reported lower LBRs with APA in all female partner age-groups in couples with idiopathic infertility. Some studies that reported no association of male age on the LBR performed ICSI cycles on the partner’s oocytes, implying that ICSI could rectify some of the adverse male factors ( Spandorfer et al. , 1998 ; Aboulghar et al. , 2007 ). On the other hand, Horta et al. (2019) did not observe that the method of insemination (IVF or ICSI) was associated with pregnancy outcome in normospermic men. It appears that the impact of sperm is more subtle than woman’s age alone and is influenced by other confounders. To determine the effect of male partner’s age on the LBR and miscarriage rate (MR), we analysed a large national database taking possible confounding factors into account. To minimize the influence of significant underlying factors related to oocytes and sperm, we recruited all intended fresh IVF/ICSI cycles with partner’s sperm due to unexplained or tubal infertility, limiting the female partner’s age to 44 years. Parallelly, we also analysed the cycles including male-factor infertility to observe any effect of semen abnormalities that may or may not be associated with male age.

A total of 59 951 IVF/ICSI cycles conducted during the study period met the inclusion criteria. The process of inclusion and exclusion of the treatment cycles has been presented in the Flow Chart ( Fig. 1 ). Couples underwent IVF (n = 27 226) or ICSI (n = 32 725) treatment with partner’s sperm followed by fresh ET due to unexplained (n = 31 846) or tubal infertility (n = 6605) or male infertility (n = 22 905). Treatment cycles with endometriosis (n = 5563), ovulatory disorders (n = 9970), and PGT (n = 280) were excluded. The distribution of the number of male patients including per retrieval LBR across the male age-groups in each of female age-groups with linear trend lines and P -values linked with chi-square 2xK table has been presented graphically in Fig. 2 . Flow chart: inclusion and exclusion process. PGT, preimplantation genetic testing; -A for aneuploidy; -M, for monogenic diseases LBR, live birth rate. # Cycles not reported to have any cause of infertility have been classed under ‘unexplained’ infertility. *Some cycles had both tubal and male-factor infertility. Numbers of men and live birth rates (LBR) per oocyte retrieval in each male age-group, under each female partner’s age-group with linear trend line . When the patient’s personal and cycle characteristics were compared between those cycles that achieved a LB and those did not, we observed significantly higher proportion of younger men and women (<35 years) achieving a LB ( P  < 0.0001), while a significantly lower proportion was associated with an LB in men and women of 40 years or older age-groups ( P  < 0.0001) ( Table 1 ). A significantly higher proportion of cycles in the LB group was on the first attempt of IVF/ICSI ( P  < 0.0001) and had had a previously successful LB ( P  = 0.01) ( Table 1 ). IVF was performed less often than ICSI in those couples who had a LB (40.6% versus 47.2%, P  < 0.0001). Significantly more successful cycles (79.9% versus 49.2%, P  < 0.0001) had single ET and were at the blastocyst stage (84.5% versus 62.9%, P  < 0.0001) ( Table 1 ). Comparison of personal and cycle variables between the cycles that resulted in live birth and the cycles that did not. ^ Cycles with male infertility included. LBR per oocyte retrieval declined with women in older age-groups: from 35.3% (8508/24 072) in women <35 years old to 25.8% (6290/24 346) in the female age-group of 35–38 years ( P  < 0.0001), and to 12.2% (1413/11 547) in the age-group of 40–44 years ( P  50 years age-groups compared to those aged <35 years, whether male infertility was included or excluded ( Supplementary Table S1 ). Analysis stratified by women’s age-groups showed no difference in the LBRs per oocyte retrieval and per ET across the male age-groups when the female partners were aged <35 years or when they were older (≥40 years) in IVF as well as in ICSI cycles, whether male factor was included or excluded ( Tables 2 and 3 , Supplementary Tables S2 , S3 , S4 , and S5 ). Only in the male age-group of 35–39 years, LBR per retrieval (not per ET) was significantly lower (drop from 42.0% versus 38.6%, P  = 0.002) in the IVF and ICSI combined cycles without male-factor infertility. There was an apparent rise in LBRs per retrieval with advancing male age-groups when IVF was performed under the female partner’s age-group of <35 years and in ICSI cycles in the female age-group of 40–44 years among the population without male infertility ( Table 3 ), but these differences were not statistically significant. Live birth rate per oocyte retrieval in IVF and ICSI cycles in male and female age-groups—including male infertility. Reference group. Live birth rates per oocyte retrieval in IVF and ICSI cycles in male and female age-groups—excluding male infertility. Reference group. Chi-square 2xK analysis P -value 0.53. Age-groups combined together to increase population size and thereby statistical significance of the differences. Chi-square 2xK analysis P -value 0.05. Analysis of only the IVF cycles showed a decline in LBRs per retrieval with advanced male partners age (≥40 years) when the female partners were at 35–39 years age-group: dropping from 27.0% in the male age-group of 50 years, respectively with male-factor infertility included ( Table 2 ) and from 27.6% in 35 years age-group to 23.5% ( P  = 0.002) and 22.2% ( P  = 002) in 40–44 years and older groups, respectively when male infertility was included ( Table 3 ). However, there was no significant difference in the LBRs across the male age-groups, including in the female age-group of 35–39 years where ICSI was performed as a method of insemination whether in presence or absence of male-factor infertility ( Tables 2 and 3 ). Similar pattern of decline in LBRs per oocyte retrieval as well as per ET was observed within the female partner’s age-group of 35–39 years when both IVF and ICSI cycles were combined together, whether the population with male infertility was included or excluded ( Supplementary Tables S2 and S3 ). Subgroup analyses of the above comparison in only the first treatment cycles with IVF and ICSI (per patient analysis) ( Supplementary Table S4 ) or in the first single blastocyst transfer cycles, avoiding the impact of the number and stage of embryo transferred revealed similar findings: a decline in LBRs in men aged ≥40 years within the female age-group of 35–39 years, only when IVF was performed, but not in ICSI cycles ( Supplementary Table S5 ). A rising trend of MRs per single ET was observed in the male age-groups of >55 years compared to that in <35 years whether IVF or ICSI in the female age-groups of <35 years and 35–39 years, particularly among couples without having male-factor infertility. Although the number of patients in the subgroups was very small, the rise in MRs only in men over 55 years and women aged 40–44 years was statistically significant ( P  = 0.02) in IVF cycles ( Supplementary Table S6 ). On univariate analysis, male and female partner’s age, the number of previous attempts, history of previous LB, method of insemination (IVF or ICSI), number of embryos transferred, and day of transferred embryo (cleavage-stage or blastocyst) significantly correlated with LBR ( P  < 0.0001) ( Table 4 ). On multiple regression analysis with the OR adjusted for these confounders, male partners’ age-groups were significantly associated with LB ( P  35 years and <40 years ( Table 4 ). This was also noted when analysed when male-factor infertility was included ( Table 4 ). Association of age-groups and other confounders with fresh cycle live birth. Adjusted for female partner’s age (when more than one age-groups), previous IVF/ICSI cycles, previous live birth, number of embryos transferred, stage of embryos, and the method of insemination.

This was a retrospective analysis of all stimulated IVF/ICSI cycles that met the inclusion criteria from the UK’s national registry on assisted reproduction published online by the Human Fertilisation and Embryology Authority (HFEA) between 1 January 2017 and 31 December 2018. Being a retrospective analysis of anonymized data, this study did not require Ethical Committee approval. Only IVF/ICSI cycles with partner’s sperm that had oocyte retrieval with the intention of fresh embryo transfer (ET) in infertile couples who had no ovulatory disorders or endometriosis were included in this study. The female partners were aged between 18 and 44 years, while there was no upper limit for the male partner’s age. Cycles utilizing donor sperm were excluded. Cycles with preimplantation genetic tests for aneuploidy (PGT-A) or monogenetic defects (PGT-M) were also not included. The cycles cancelled before oocyte retrieval were not reported in the HFEA database. Our primary outcome measure was the effect of male age on LBR per oocyte retrieval as well as per ET, stratified in different age-groups of the female partners. Secondary outcome measures determined MRs, also stratified by male and female age. The outcomes were also analysed by taking IVF and ICSI cycles separately, and both excluding and including male-factor infertility. To compare the LBRs on ‘per patient’ basis, a subgroup analysis was performed taking only the first-ever IVF/ICSI cycles (HFEA database does not provide the identity of individual patients). To try to eliminate the effect of potential confounders (previous failed IVF/ICSI cycle(s), the method of insemination, number and stage of embryo transferred) subgroup analyses with single blastocyst transfer on the first ever treatment cycle were also performed with separate analysis for IVF and ICSI cycles. MRs were calculated on per single ET basis to reduce possible influence of transferring multiple embryos. ‘LBR’ was defined as the birth of a viable neonate according to the HFEA (UK) ( HFEA, 2020 ). MR included biochemical pregnancies/miscarriage as defined by International Committee for Monitoring Assisted Reproductive Technology (ICMART) and the World Health Organization (WHO) revised glossary ( Zegers-Hochschild et al. , 2009 ) plus clinical pregnancies that did not progress beyond 24 weeks after detection of gestational sac or foetal heartbeat on the ultrasound scan ( HFEA, 2020 ). As per the HFEA database, women were stratified by age in the following groups: <35, 35–37, 38–39, 40–42, and 43–44 years; and the male ages were grouped by 60 years. These age-groups were taken as ordinal variables in the regression analysis. To increase the population size in each age-group, female age-group of 35–37 years was combined with 38–39 years, and 40–42 years combined with 43–44 years; likewise, male age-groups of 35–37 and 38–30 years, 40–42 and 43–44 years, and all age-groups above 50 years were merged for LBR comparison. Too few cycles in some of the subgroups ran the risk of giving inflated or deflated figures—if this was suspected, the age-groups were merged together when appropriate. LBRs and MRs in different age-groups were compared with the chi-square test 2x2 tables (Fisher’s exact test in case of small numbers). A chi-square 2xK table was used to examine the trend with the chi-square linear P -value indicating statistical significance when large variations in LBRs between the male age-groups were observed to be due to small population size, particularly in the female age-group of 40–44 years. Male age <35 years (in any female age-group) served as the ‘reference’ group, against which the LBRs and MRs of male partners in older age-groups were compared. Potential confounders were identified by comparing the occurrence (proportions) of personal and treatment variables in the cycles that resulted in a live birth (LB) and those did not. Significant variables were also noted in the univariate regression analysis. Multiple logistic regression was performed in each female age-group to find the association between male age and LBR adjusting the odds-ratio (OR) for the confounding factors. As the HFEA datasheet presented patient’s age in the above age-groups, not in actual age, each age-group has been used as ordinal variables and the occurrence of LB in the <35 years age-group has served as reference for the logistic regression model. CIs 95% were used to find the statistical significance. A P -value of <0.05 was considered statistically significant. StatsDirect software (version 3.3.5, dated 22 March 2021) ( https://www.statsdirect.com/ ) was used for all regression analysis; Microsoft Power-point was used for graphs.

Our analysis from the UK’s national database revealed an impact of male age on the IVF/ICSI pregnancy outcomes which also depended on the female age and the method of insemination, in the presence or absence of obvious male-factor infertility. There was no difference in the LBRs across the male’s age-groups when the female partners are either young (<35 years) or 40 years and over, regardless of whether the cycles with male infertility were taken into the analysis or not. However, a significant drop in LBRs per oocyte retrieval or per ET was observed in men aged 40 years or older only when IVF was performed in the female partners aged between 35 and 39 years both with and without including the cycles reported to have male-factor infertility. Interestingly, there was no impact of male age on LBR per oocyte retrieval or per ET when ICSI was performed in presence or absence of male-factor infertility. We ran several subgroup analyses to adjust for possible confounders. The aforementioned findings remained the same when only the first treatment cycles (i.e. ‘per patient’ outcome) were analysed thus eliminating repeated inclusion of same personal characteristics and any other factor responsible for repeated failed cycle(s) ( Supplementary Table S2 ), or when the subset of single blastocyst transfer cycles were analysed removing the impact of the transfer of multiple embryos or the stages of transferred embryos ( Supplementary Tables S5 and S6 ). Multivariate analyses, adjusting the ORs for significant confounders including female partner’s age, the number of previous IVF/ICSI cycles, previous LB, insemination methods, number and stage of embryo transferred, also confirmed the above findings in the population with as well as without male infertility ( Table 4 ). The MRs tend to rise in only men over 55 years of age when female partner’s age-group below 40 years in couples without having male-factor infertility. Although not statistically significant (except when male partners were >55 years and female partners were aged between 40-44 years in IVF cycles), and the number of patients was generally low, a near doubling of MRs at these age-groups ( Supplementary Table S6 ) could be clinically important that necessitates further data to validate. Although the negative effect of APA on laboratory and clinical outcomes has been described, the evidence remains contradictory mainly due to some strong confounding factors including maternal age, presence or absence of severe male factor, and the effectiveness of possible corrective measures (e.g. ICSI). In an attempt to remove the impact of substantial male factors and other significant female factors, a recent study on unexplained infertility (n = 269) found no difference in the LBRs and MRs between men of <35 and ≥35 years of age, after adjusting for women’s age ( Elbardisi et al. , 2021 ). In contrast, a clinical trial on couples with idiopathic infertility (n = 2425) found a negative impact of male age on clinical IVF/ICSI outcomes independent of female age; a lower odds of predicting LBRs were noted even when men were <40 years and women were <30 years old ( Horta et al. , 2019 ). Two studies were identified in the literature that had similar findings as ours: one study included couples with tubal infertility only (n = 1938) and found a significantly higher risk of failure to conceive through IVF when men >40 years of age, and women were 35 years or older, as compared to men and women of under 30 years of age ( de La Rochebrochard et al. , 2006 ). There was no decline in conception rate with advancing paternal age when the female partners were <30 years of age and the odds of failed conception climbed progressively as the paternal and or maternal age advanced ( de La Rochebrochard et al. , 2006 ). The other study also demonstrated no significant difference in the implantation rates as well as pregnancy rates between men of advancing age-groups when female ages were of <30 years as well as between 35 and 38 years, but pregnancy rates declined with increasing male age when female age was between 30 and 34 years ( Wu et al. , 2016 ). Thus, our findings are mainly in line with de La Rochebrochard et al. and Wu et al. indicating no impact of APA when female partners were young, and in contrast with Horta et al. (2019) who reported a significant decline in the probability of LB in men over 40 years in all female age categories. The data from Wu et al. as well as those of our study show no influence of APA on pregnancy outcomes when maternal age is advanced. Unlike the study by de La Rochebrochard et al. and Horta et al. that excluded male-factor infertility; the study by Wu et al. was done on unselected population (including male infertility), and the findings of this study were the same as ours. However, Wu et al. (2016) had a smaller study population (n = 9991), did not report on LBR’s, the upper limit of female age was only 38 years and male uppermost age-group was ≥42 years. These studies except that of Wu et al. excluded male-factor infertility from the study population to try to avoid the impact of severe sperm abnormality which may overshadow any subtle age-related effect ( de La Rochebrochard et al. , 2006 ; Horta et al. , 2019 ). Our findings with tubal and unexplained infertility were no different from the population when male infertility was added. The finding that ICSI may correct age-related sperm defects is in the line with a few older studies that found no impact of paternal age when ICSI was performed ( Spandorfer et al. , 1998 ; Aboulghar et al. , 2007 ). Our study gives evidence from a large national database that male age does not affect IVF outcomes (LBR) when female partners are young (<35 years). This can explain why, except in a few studies ( Frattarelli et al. , 2008 ; Robertshaw et al. , 2014 ), most of the clinical trials and a systematic review of available publications concluded that male age has no impact on the outcome of donor oocyte treatment ( Sagi-Dain et al. , 2015 ), given that the donors are usually young women with good prognosis. We are the first to endorse the earlier suggestion from a smaller study by Wu et al. that advancing men’s age does not worsen the LBRs when women are older (≥40 years, in our study). Basic science studies have reported reduced size of the testes, arteriosclerotic changes, thickening of tunica, and a reduction in the Leydig cell count with APA ( Plas et al. , 2000 ). Reduced semen volume and sperm motility have been associated with advanced age ( Kidd et al. , 2001 ; Whitcomb et al. , 2011 ). A delay in natural conception has been shown to be significant when the male partner is over 40 years old ( Ford et al. , 2000 ; Hassan and Killick, 2003 ). APA has also been associated with an increased incidence of spontaneous miscarriages ( Slama et al. , 2005 ). Age-related epigenetic changes in the sperm have been suggested to occur in animals but have not been confirmed in humans ( Ashapkin et al. , 2023 ). Despite the above evidence, the impact of male partner’s on LBR appears to be more subtle than that of advanced female partner’s age ( Humm and Sakkas, 2013 ). The incidence of significant sperm chromosomal defects including aneuploidy, diploidy, or y-chromosome micro-deletion is rare and is usually manifested in the form of severe oligoasthenoteratozoospermia or azoospermia in men of any age. The main age-related factor present in males even with apparently normal semen parameters is sperm DNA damage. Unlike as with maternal age, the embryo aneuploidy rate does not rise significantly with APA ( Wyrobek et al. , 2006 ; Dviri et al. , 2020 , 2021 ). A meta-analysis found a weak clinical association between sperm DNA damage and IVF/ICSI outcomes ( Collins et al. , 2008 ). However, the predictive value of sperm DNA fragmentation testing depends on the accuracy of the different tests available, as well as on various intrinsic factors including, whether it is single or double-strand DNA damage, the proportion of sperm with DNA damage, the extent of damage in each sperm, and importantly, the ability of oocyte to repair sperm DNA damage ( Sakkas and Alvarez, 2010 ). It is the reparatory mechanism within the oocyte and other compensatory processes such as adjustment (elongation) of the chromosome telomeres in response to ageing sperm ( Eisenberg, 2011 ) which tend to interact, affecting the relationship between paternal and maternal age on the IVF/ICSI outcome complex. Studies on an animal model showed that sperm of moderate to high sperm DNA damage can fertilize oocytes but subsequent embryo development is affected by the advancing age of female mice ( Horta et al. , 2020 ). Our study and most of those dealing with donor oocytes provide clinical evidence that good quality oocytes from young women are capable of repairing the DNA damage resulting from ageing sperm ( Menezo et al. , 2010 ). Healthy oocytes have been identified with high expression of DNA repair genes ( Menezo et al. , 2010 ). Male age affects reproductive outcomes when this reparative process falls short ( Gonzalez-Marin et al. , 2012 ). Ageing oocytes are associated with a diminished expression of DNA reparatory genes ( Liu and Keefe, 2008 ; Horta et al. , 2020 ) resulting in declining LBRs and rising MRs with APA. The reasons are more obscure as to why advanced male age, in the absence of major semen abnormalities, does not cause further reduction in the LB outcome above a critical female age (40 years in our study). APA has not been shown to increase the embryo aneuploidy rate ( Dviri et al. , 2020 ). It is possible that, increasing aneuploidy from ageing oocytes dominates pregnancy outcomes in this cohort of women overshadowing any additional effect of increasing sperm DNA damage not rectified by ageing oocytes. This is applicable both in terms of LBR and MR. To our knowledge, this is the first study to report evidence from a large national database on the effect of male partner’s age on LBRs as well as MRs from fresh IVF/ICSI cycles and its interaction with female partner’s age. This enabled us to incorporate adequate number of patients in most of the subgroups to obtain a reliable dataset. However, some subgroups with smaller patient numbers had to be combined together or a chi-square analysis on a 2xK table was done to observe the trend when large variation in percentage was suspected to be due to small subgroup population. Our data provide strong clinical evidence supporting the basic science studies describing the capacity of good quality oocytes to repair genomic defects of ageing sperm and a failing reparatory mechanism with advancing maternal age. Our novel finding of having no additional worsening effect of ageing sperm on LBR over and above a certain age of the female partners opens up future areas of research. This is the first study of its kind to analyse LBR per single blastocyst transfer in order to eliminate the impact of the number and stage of embryos transferred on the pregnancy outcomes, while LBRs per ovum pickup assessed the effect of sperm on fertilization and embryo development. The abolition of the impact of male age on the LB outcome when ICSI was performed due to male or non-male factor infertility could be the basis of recommending ICSI procedure in male partners of advanced age in the future. Further studies are needed to explore and confirm this important therapeutic benefit. Two important limitations of this database were: (i) unavailability of data on some personal and cycle characteristics, e.g. BMI of both the partners, ovarian reserve, or stimulation dose, (ii) the ages of the male and female partners are presented in groups, rather than in individual ages. Due to a strict confidentiality policy, HFEA was unable to provide too many details of the treatment cycles. Although in a large population size like ours, the missing information is less likely to have a major influence on the main outcomes, we call for further large studies taking all the confounders into account to validate our findings. To try to overcome the issue of not having the actual age, we have used each age-group as ordinal variable in the regression analysis. The actual magnitude of the effect of male age could be influenced by the severity and underlying causes of semen abnormality, over and above inherent age-related effect if any. Detail information of male-factor infertility was not available in the HFEA database. Future studies should be structured with actual age and BMI of both the male and female partners and should set the inclusion–exclusion criteria based on aetiology of sperm abnormalities. Unfortunately, the HFEA database gives the number of oocytes and embryos in groups, thus laboratory outcomes namely fertilization or blastulation or embryo utilization rates could not be analysed. Finally, the absence of data from subsequent frozen-thawed ET cycles precluded us from analysing the cumulative LB outcomes. To conclude, the data analysed from a large national database demonstrate declining LBRs per oocyte retrieval as well as per ET with advancing paternal age but not when female partners are younger than 35 years or older than 40 years and above. This applies particularly in IVF but not ICSI cycles, indicating the benefit of performing ICSI in this group of patients even in absence of male infertility. Younger (presumably good quality) oocytes seem to be capable of repairing age-related sperm damage, while significant age-related oocyte factors seem to overshadow any subtle sperm factor related to APA. Our data are of value in counselling couples, particularly of disparate ages. The scientific basis for the interaction between ageing oocyte and sperm and the value of performing ICSI needs further research.