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.

Study question: Does advanced male partner's age impact live birth rates (LBRs) in IVF treatment when female partner's age is factored in?

Summary answer: In fresh IVF cycles LBRs decline with male partner's age ≥40 years when the female partner is aged 35-39 years, irrespective of the presence or absence of male factor; but not when the female partner is <35 years or ≥40 years of age; this decline is not observed in ICSI cycles.

What is known already: Advanced paternal age is associated with declining sperm parameters, impaired embryo development, compromised pregnancy outcomes, and abnormalities in the offspring in IVF/ICSI cycles. However, data on the interaction between maternal and paternal age on IVF outcomes are very limited and inconsistent. No significant effect of male partner's age on pregnancy outcomes has been noted in donor oocyte cycles.

Study design, size, duration: Retrospective analysis of all eligible autologous IVF/ICSI cycles with oocyte retrieval and intended fresh embryo transfer (ET) from the UK's national anonymized registry, published online by the Human Fertilisation and Embryology Authority (HFEA). There were 59 951 cycles that qualified the inclusion criteria in the study period: 1 January 2017 to 31 December 2018.

Participants/materials, setting, methods: 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), tubal (n = 6605), or male infertility (n = 22 905). Treatment cycles with endometriosis (n = 5563), ovulatory disorders (n = 9970), female partner aged >44 years (n = 636), and PGT (n = 280) were excluded. Women were stratified by age in the following groups: <35, 35-39, 40-42, and 43-44 years; male partner's age as <35 (reference group), 35-37, 38-39, 40-42, 43-44, 45-50, 51-55, 55-60, and >55 years as presented by the HFEA. Some age-groups were merged in the analysis to increase the population size. Chi-square test was used to compare binominal data; and multiple logistic regression to find any association between male and female age-groups on live birth adjusting for other confounders that had a significant effect on this outcome.

Main results and the role of chance: LBRs per oocyte retrieval as well as per ET were no different across the male partners' age-groups when the female partners were aged <35 years or in 40- to 44-year age-group, whether male-factor infertility was included or excluded and whether it was IVF or ICSI cycle. However, when IVF was the method of insemination in the female partner's age-group of 35-39 years, LBRs per oocyte retrieval dropped significantly from 27.0% in the male age-group of <35 years (reference group) to 22.9% (P = 0.002), 22.0% (P = 0.006), and 18.8% (P = 0.004) in 40-44, 45-50, and >50 years age-group, respectively in population that included male-factor infertility. Likewise, LBR per retrieval declined 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 in cycles without male infertility. However, there was no impact of male age on LBR in any female partner's age-group when ICSI was performed in either the presence or the absence of male infertility. A similar decline in the LBR per retrieval and per ET was observed in female age-group of 35-39 years in the analyses with IVF and ICSI cycles combined. The inference remained unchanged when only the first treatment cycle was included (per patient analysis) or when single blastocyst transfer cycles were analysed, eliminating the impact of the number and stage of embryo transferred. After adjusting for confounders including male age, female age, number of previous treatment cycles, previous live birth, insemination method (IVF or ICSI), number of embryos transferred, and day (stage) of ET, male partner's age remained significantly associated with LBR in the female age-group of 35-39 years, but not when women were in <35 years or 40- to 44-year age-group, in population including as well as excluding male infertility. Miscarriage rates per single ET trended to rise (non-significantly) in IVF as well as ICSI cycle only when men were over 55 years and female partners aged <40 years, particularly when male infertility was excluded.

Limitations, reasons for caution: Information on ovarian reserve and stimulation protocols was not available. This probably would have had little impact, given the large size of the population studied. The ages of female and male partners were given in groups necessitating taking them as ordinal variable in the regression analysis. Cumulative LBRs could not be determined as the information on subsequent frozen-thawed ET cycles could not be traced and the severity or cause of abnormal semen parameters were not present in the HFEA database. Some age-groups with small number of patients were merged to obtain a reliable result.

Wider implications of the findings: This is the largest clinical data to support the laboratory evidence of the ability of oocytes from young women to reverse the age-related deterioration of sperm quality. As the ageing oocytes lose this reparatory mechanism, the ageing sperm exert a detrimental effect on the LBR. The message of this study is important in counselling of patients and planning out treatment. Further research on interaction between male and female age will increase our understanding of this matter and help to establish whether ICSI procedure is more appropriate for older male partners even when there is no apparent semen abnormality.

Study funding/competing interest(s): No funding was required. There is no competing interest.

Trial registration number: N/A (retrospective analysis).