Background: A wide variety of studies demonstrate the influence of the
oocyte source’s age on the success of assisted reproduction treatments; however,
the effect of paternal age has been studied to a lesser extent. Therefore, our
goal was to assess the impact of male age and sperm quality on in vitro
fertilization (IVF) outcomes. Methods: Three hundred ninety-four ova
donation IVF cycles from Ingenes México were retrospectively analyzed. All
ova donors (age range: 18–35 years) underwent a similar IVF stimulation
protocol. The oocytes were aspirated and inseminated by intracytoplasmic sperm
injection (ICSI) using either partner sperm (n = 332, age: 42.4
Male partner’s effects are usually considered negligible during IVF, even though over 40 to 50% of infertility diagnoses are associated with male factor [1, 2]. An increase in paternal age and age-independent degeneration of semen quality can be associated with a decline in viable embryos during IVF [3]. Numerous studies have demonstrated that advanced maternal age is associated with decreased ovarian reserve, oocyte pick-up, fertilization rates, and the production of high-quality embryos [4, 5, 6]. However, concerning paternal age, few studies have examined this association, with some studies demonstrating a significant association between paternal age and the embryo aneuploidy rate [7, 8, 9] as well as deterioration of sperm concentration, motility, and morphology [10, 11, 12, 13, 14]. Therefore, it has been suggested that, under certain circumstances, such as low sperm quality in the presence of advanced paternal age, a couple should consider using donor sperm and PGT-A [15]. This begs the question, when semen quality is considered, does using donor sperm improve IVF outcomes? Furthermore, if the male factor is considered, is this an indication to perform PGT-A? Therefore, this study aimed to compare the embryo aneuploidy rates as well as the implantation and pregnancy rates between paternal sperm and donor sperm, with respect to the sperm source’s age and sperm abnormalities.
We performed a clinical retrospective study by reviewing all medical records from October 2016 to December 2018 at Ingenes México, México City, México. To be included in the study, the infertile patients had to undergo IVF using oocyte donation, and the embryos had to be assessed by PGT-A using Next-Generation Sequencing (NGS). As exclusion criteria, the oocyte recipients were not diabetic or obese, taking medication for high blood pressure or any other medications that would affect IVF. For their male partners, the subjects also had not to be suffering from diabetes, obesity or taking medications (antibiotics and retroviral agents) that would affect the production or quality of their sperm. PGT-A was performed only on embryos with sufficient morphological quality to survive the biopsy and freezing procedures.
All oocyte donors were between 18 and 35 years old, had normal-appearing ovaries
confirmed by transvaginal ultrasound, an antral follicle count
Sperm donors were between 18 and 35 years old, and their samples were only used after being frozen and passed a 6-month quarantine at the Donor Sperm Bank at Ingenes Institute in México City. All donor sperm were diagnosed as normozoospermic (exceeding the minimum acceptable value for each category as determined by the World Health Organization criteria for the Examination and Processing of Human Semen) before cryopreservation [16]. All paternal sperm samples were fresh (not frozen) and ranged in quality. Independent of the sperm source (paternal or donor), before the semen sample was collected, the patient or donor was explained to abstain from sexual activity and drug use for 3 to 5 days. The semen was obtained by masturbation, and in the case of azoospermia, the sample was obtained by Percutaneous Epididymal Sperm Aspiration. All sperm samples were analyzed for volume, concentration, progressive motility, and morphology by phase-contrast microscopy. For concentration and motility, the counting was performed in a Makler Chamber® (Sefi-Medical Instruments, Haifa, Israel) with 20x magnification. The morphological assessment was performed by stained smearing with 100x magnification. The limit values applied for diagnosis in volume, concentration, and motility were according to the World Health Organization 2010 criteria [16]. For morphology, the Kruger criteria were used [17]. Sample preparation was performed according to standard laboratory procedures (washing, density gradients, and swim-up), and diagnosis and sperm capacitation were performed using the World Health Organization criteria [16].
The oocytes were all inseminated by ICSI, and fertilization was judged by
forming two pronuclei and two poplar bodies, 16–18 h after insemination. Embryos
were cultured (37
For PGT-A, we used NGS. According to the manufacturer’s instructions, each
biopsy sample was amplified using the SurePlex DNA amplification system (Illumina
Inc., San Diego, CA, USA). Whole-genome amplification (WGA) products were
quantified using a Qubit 3.0 Fluorometer
(Life Technologies, Carlsbad, CA, USA).
Library preparation was carried out with the VeriSeq PGS Library Prep Kit
(Illumina Inc.). DNA ‘indexing’ was performed
to simultaneously analyze samples from different embryos using the Nextera XT
96-Index Kit (Illumina Inc.). For library preparation, 5
Oocyte donation indications were failed IVF cycles with own oocytes, low ovarian
reserve, poor oocyte quality, genetic or chromosomal abnormalities transmissible
to offspring, and spontaneous/iatrogenic menopause. Before the thawed embryos
were transferred, estrogen endometrial preparation was performed using an
increasing transdermal dose of estradiol valerate for a period of 15 to 40 days
before the start of luteal support with vaginal progesterone for five days.
Embryos were transferred 2 hours after being thawed. The Reproductive
Specialist/Embryologist determined clinical decisions about which and how many
embryos should be transferred. Depending on the embryo survival rate and specific
clinical conditions related to each case, every decision was made with the
patient’s approval. Clinic pregnancy was considered positive if the
To determine the required sample size, we used the G*Power 3 software (https://stats.idre.ucla.edu/other/gpower/power-analysis-for-one-sample-t-test/). For the comparison between normal paternal and donor semen, using the assumptions that the expected aneuploidy rate is around 20% of ova donor cycles with a standard deviation of 15% and that a clinical difference of 10% would indicate a significant result, alpha = 0.05, beta = 0.05, and a one-to-one ratio, we determine that 60 IVF cycles would be required for each group.
All analyses were carried out with the Statistical Package for the Social
Sciences software v26.0 (IBM, Chicago, IL, USA). The Shapiro-Wilk test assessed
the normality of the data. Differences between categorical data were assessed
with the Chi-Square test. Homogeneity of the variances in parametric data was
determined with Levene’s test. For parametric data, differences between groups
were determined with Analysis of Varianza (ANOVA) with a post hoc
Bonferroni test. For non-parametric data, differences between groups were
determined with the Kruskal-Wallis test with a post hoc Dunn’s test.
Pearson correlation coefficient (r) or Spearman correlation coefficient [6] were
used to determining the association between variables. p-values
We originally identified 457 ova donor cycles. However, using the inclusion and
exclusion criteria, 394 IVF cohorts qualified for the study (Fig. 1). As shown in
Table 1, 15.7% (95% CI: 12.2–19.8%) were from donor sperm, 19.8% (95% CI:
16.0–23.9%) were from parental sperm with normal seminal parameters
(normozoospermic patients), 38.8% (95% CI: 33.8–43.7%) were from parental
sperm that suffered from teratozoospermia, whereas 25.6% (95% CI: 21.6–29.9%)
were from parental sperm that suffered from severe alterations in one or more
seminal parameters (volume, concentration, progressive motility, and morphology).
There was a 17-year difference between the average age of partner sperm (42.4
The study flow chart in line with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.
Category | Total | Donor semen | Paternal semen | p-value | |||
Normozoospermia | Normozoospermia | Teratozoospermia | Severe alterations | ||||
IVF cycles | 394 | 62 | 78 | 153 | 101 | ||
Embryos analyzed | 1447 | 227 | 297 | 558 | 365 | ||
Paternal Age | 39.7 ± 1.0 | 25.0 ± 3.3 |
40.3 ± 7.2 |
42.0 ± 6.5 |
44.7 ± 8.2 |
||
Maternal age | |||||||
Donor | 24.5 ± 3.5 | 24.7 ± 3.2 | 24.9 ± 3.6 | 24.4 ± 3.4 | 24.1 ± 3.7 | 0.467 | |
Recipient | 41.7 ± 4.1 | 41.7 ± 3.1 | 40.5 ± 4.6 |
42.1 ± 4.3 |
42.1 ± 3.9 |
0.031* | |
Biopsy Day | |||||||
Day 5 | 307 (77.9 ± 2.2%) | 51 (82.3 ± 4.8%) | 64 (82.1 ± 4.6%) | 119 (77.8 ± 3.3%) | 73 (72.3 ± 4.4%) | 0.344 | |
Day 6 | 87 (22.1 ± 2.2%) | 11 (17.7 ± 4.8%) | 14 (17.9 ± 4.6%) | 34 (22.2 ± 3.3%) | 28 (27.7 ± 4.4%) | ||
Male etiology | |||||||
Normozoospermia | 140 (35.5 ± 2.3%) | 100.0% | 100.0% | - | - | ||
Teratozoospermia | 153 (38.8 ± 2.5%) | - | - | 100.0% | - | ||
Asthenoteratozoospermia | 37 (9.4 ± 1.5%) | - | - | - | 37 (36.6 ± 4.8%) | ||
Asthenozoospermia | 19 (4.8 ± 1.1%) | - | - | - | 19 (18.8 ± 3.8%) | ||
Azoospermia | 6 (1.5 ± 0.6%) | - | - | - | 6 (5.9 ± 2.3%) | ||
Cryptozoospermia | 2 (0.5 ± 0.3%) | - | - | - | 2 (2.0 ± 1.4%) | ||
Oligoasthenoteratozoospermia | 21 (5.3 ± 1.1%) | - | - | - | 21 (20.8 ± 4.1%) | ||
Oligoteratozoospermia | 16 (4.1 ± 1.0%) | - | - | - | 16 (15.8 ± 3.6%) | ||
Results are presented as frequency, percent
and standard error, or mean and standard deviation. Differences between groups
were calculated using either the Chi-Square test for categorical data, an ANOVA
with a post hoc Bonferroni test for parametric, continuous data, or the
Kruskal-Wallis test with a post hoc Dunn’s test for non-parametric, continuous
data. A significant difference between the groups (p |
Aneuploidy rates for
embryos developed from donor ova using patient or donor sperm, stratified into
normozoospermic donor (left black column), normozoospermic patient (right black
column), patients with teratozoospermia (white column), and patients with severe
abnormalities (gray column). The column height corresponds to the average
aneuploidy rate, whereas the bar length is the standard error. Differences
between groups were determined with ANOVA with a post hoc Bonferroni
test. Significant differences (p
For patient sperm, there was a weak correlation between paternal age and the
prevalence of semen abnormalities (Pearson’s r
Aneuploidy rates for embryos developed from donor ova using patient or donor sperm, stratified by age. Data were separated into quartiles based on the sources’ age of the donor sperm (A) or paternal sperm (B). Afterward, the paternal sperm was categorized, based on etiology, as either normozoospermic (C), teratozoospermic (D), or severe abnormalities (E). The column height corresponds to the average aneuploidy rate, whereas the bar length is the standard error. Associations between the sources’ age and the aneuploidy rate were determined by calculating Spearman’s rho.
When the IVF outcomes were compared, there was no clinical difference between the study groups with respect to the number of ova captured, the number of ova fertilized, the number of blastocysts formed, or the embryo transfer rate (Table 2). Moreover, the sperm source (donor versus paternal) and the etiology present in the sperm sample (normozoospermic versus teratozoospermic versus severe abnormalities) had no difference in the clinical pregnancy rate and embryo implantation.
Category | Total | Donor semen | Paternal semen | p-value | |||
Normozoospermia | Normozoospermia | Teratozoospermia | Severe alterations | ||||
Ova captured (n) | 14.9 ± 4.7 | 13.1 ± 5.9 |
15.4 ± 4.5 |
15.4 ± 3.7 |
14.9 ± 4.9 | 0.039* | |
Ova Fertilized (n) | 10.0 ± 3.3 | 9.0 ± 4.1 | 10.3 ± 3.0 | 10.1 ± 2.9 | 10.0 ± 3.5 | 0.198 | |
Blastocyst formed (n) | 5.7 ± 2.5 | 5.6 ± 2.5 | 5.9 ± 2.3 | 5.7 ± 2.4 | 5.5 ± 2.9 | 0.703 | |
Embryos Transferred (n) | 1.9 ± 0.6 | 2.0 ± 0.6 | 1.9 ± 0.5 | 1.9 ± 0.6 | 1.8 ± 0.6 | 0.176 | |
Transfer rate (%) | 44.0 ± 39.3 | 41.1 ± 37.9 | 50.0 ± 40.1 | 40.4 ± 37.2 | 47.0 ± 42.9 | 0.299 | |
No transferred | 32 (8.1 ± 1.4%) | 5 (8.1 ± 3.6%) | 4 (5.1 ± 2.4%) | 9 (5.9 ± 1.8%) | 14 (13.9 ± 3.4%) | ||
0.746 | |||||||
Negative | 132 (36.5 ± 2.5%) | 22 (38.6 ± 6.3%) | 23 (31.1 ± 5.4%) | 55 (38.2 ± 4.1%) | 32 (36.8 ± 5.3%) | ||
Positive | 230 (63.5 ± 2.5%) | 35 (61.4 ± 6.3%) | 51 (68.9 ± 5.4%) | 89 (61.8 ± 4.1%) | 55 (63.2 ± 5.3%) | ||
Ultrasound results (n) | 0.711 | ||||||
Biochemical | 10 (4.3 ± 1.3%) | 1 (2.9 ± 2.8%) | 2 (3.9 ± 2.8%) | 3 (3.4 ± 1.8%) | 4 (7.3 ± 3.7%) | ||
1 sac | 143 (62.2 ± 3.3%) | 21 (60.0 ± 8.7%) | 29 (56.9 ± 7.1%) | 59 (66.3 ± 4.8%) | 34 (61.8 ± 6.7%) | ||
2 sacs | 63 (27.4 ± 2.8%) | 9 (25.7 ± 7.6%) | 18 (35.3 ± 6.8%) | 21 (23.6 ± 4.4%) | 15 (27.3 ± 6.1%) | ||
3 sacs | 14 (6.1 ± 1.6%) | 4 (11.4 ± 5.5%) | 2 (3.9 ± 2.7%) | 6 (6.7 ± 2.7%) | 2 (3.6 ± 2.6%) | ||
Results are presented as frequencies,
percent and standard error, or means and standard deviation. Differences
between groups were calculated using either the Chi-Square test for categorical
data, an ANOVA with a post hoc Bonferroni test for parametric, continuous data,
or the Kruskal-Wallis test with a post hoc Dunn’s test for
non-parametric, continuous data. A significant difference between the groups
(p |
In the present study, we assessed the age-dependent effects on semen quality between partner and donor samples with respect to embryo implantation. Independent of the semen source, we found no difference in the implantation rate. Moreover, the semen source’s age did not affect the aneuploidy rates if the sperm was diagnosed as normozoospermic. This was a clinical retrospective study, which can only show an association and give credence to future hypotheses for experimental research.
Aneuploid embryos are associated with implantation failure [20, 21], where the ovum is associated with a more significant influence on the aneuploidy rate [22]. Here, using donor ova, the expected aneuploidy rate is between 20 and 30% [23, 24], about 25% of the analyzed embryos presented with aneuploidies. This rate was independent of the sperm source (paternal versus donor) and etiology (Fig. 2). Similar studies by Dubey et al. and Mazzilli et al. confirmed this result as they also observed that sperm morphological alterations do not affect the embryo aneuploidy rate [24, 25, 26]. The proposed explanation suggests that using ICSI still delivers an uncompromised genome and sperm alterations are independent of the sperm chromosome composition. Nevertheless, Coban et al. [27] demonstrated that the severity of the sperm defect, as quantified by Kruger’s criteria, can significantly affect the aneuploidy rate, which is also confirmed by Magli et al. [28]. Here, we observed a prevalence of sperm morphological defects in the analyzed samples (teratozoospermia etiology); however, none of the etiologies correlated with the types of aneuploidies. Still, a sub-analysis was not performed because some of the groups had few samples and significant heterogeneity.
Numerous studies demonstrate the maternal effect age has on aneuploidy rates [4, 29]. Nevertheless, the age of the sperm source has also been shown to have an effect on increasing embryo aneuploidies as well as reducing IVF success [7, 8, 9]. The effect of the semen source’s age is still under debate [30]. Here, we observed no correlation in semen donor samples nor with partner samples. Although the samples were stratified by etiology into normal and abnormal sperm, only the teratozoospermia group positively associated the partner’s age with the aneuploidy rate; moreover, there was a weak correlation between paternal age and the presence of sperm abnormalities. This could be an artifact as the prevalence for specific sperm etiologies is increased with age, and the aneuploidy rate is associated with sperm etiology [27, 31, 32, 33]. Even though our study does not confirm the latter effect, there is a suggested aneuploidy rate difference between certain types of sperm etiologies [27, 28, 34]. Magli et al. [28] demonstrated that the non-obstructive azoospermia and oligoasthenoteratozoospermia groups presented with more chromosomally compromised embryos than the normozoospermia group using fluorescence in-situ hybridization and, as mention above, Coban et al. [27] confirmed this effect. A review by Caseiro et al. [35] nicely explains the association between oligoasthenoteratozoospermia and embryo aneuploidies. Lastly, teratozoospermia was shown to be associated with higher rates for trisomies and monosomies [34]. Still, here, the sample size was not sufficient to demonstrate this effect.
The contribution of spermatozoa to embryos aneuploidies in IVF has been the topic of several debates. Garcia-Ferreyra et al. [9] showed that the aneuploidy rates from ova donation cycles determined a high incidence of aneuploidies when men are older than 50 years. Here, no significant difference was found between the ages of the males (paternal or donors) concerning aneuploidy rates in the analyzed embryos.
Several studies have reported the correlation between the aneuploidy rates in embryos derived from abnormal sperm. Coates et al. [36] determined that the aneuploidy incidence was significantly higher in embryos obtained from sperm with oligozoospermia etiology than those in embryos obtained from sperm with normal male factor, independent of the ova source. However, Mazzilli et al. [26] compared five different male etiologies (normozoospermic patients, patients with moderate male factor, patients with oligoasthenoteratozoospermia/cryptozoospermia, patients with obstructive azoospermia, and patients affected with non-obstructive or secretory azoospermia) for their effects on fertilization, embryonic development, aneuploidies incidence, and gestational results. They found that early embryonic development was impaired in patients with severe male factors with respect to fertilization rates and development potential; however, the euploidy rate and blastocyst implantation potential were independent of the sperm quality. They concluded that the male factor would not be an indication when suggesting Preimplantation Genetic Testing (PGT) for an IVF cycle. In the present study, no significant differences were found regarding embryonic development results (fertilization and blastocyst formation rate) as well as with pregnancy and implantation rates, when compared to patients with normal or altered male factor. Our result is in agreement with Mazzilli et al. [26].
Due to the oocyte’s contribution to the aneuploidy rate, under certain circumstances, such as advanced maternal age, embryos are analyzed by PGT-A. In the present study, we focused on evaluating the impact of the male factor in terms of age and seminal quality in exclusively ova donation cycles. Although no significant correlation was found between age and aneuploidy incidence nor between seminal quality and aneuploidies incidence, we did observe an effect of a male factor on aneuploidy rates. Therefore, it would be beneficial for these patients to consider PGT-A.
This study has a few limitations. First, we did not quantify or qualify the type of embryo aneuploidies. Only euploid embryos were transferred. However, specific aneuploidies could be generated and, through the embryo’s auto-correcting mechanisms, produce a viable euploid fetus [37, 38]. Second, the severity of the sperm deformities was not quantified in samples with teratozoospermia diagnosis or other severe alterations. As mentioned above, the effects of sperm with abnormal etiology are associated with higher aneuploidy rates. Third, only the donor semen samples were frozen until use, whereas patient samples were fresh. The effect of cryopreservation on sperm’s DNA integrity is still unclear and under debate [39]. Fourth, the reasons for oocyte donation were not considered, such as recurrent implantation failure, low reserve, or inability to produce viable embryos, to name a few. Underlining the mechanism associated with each pathology could introduce bias to the results. Fifth, all embryos underwent PGT-A using NGT; therefore, these results are limited to this group. Lastly, this is a clinical retrospective study in which we can show an association but not a causal relationship.
Here, we examine if the sperm source’s age affects the aneuploidy rate determined by NGS-PGT-A, which it does not; however, with the teratozoospermia diagnosis, there was a weak association. Therefore, it would be prudent to perform PGT-A when advanced age and teratozoospermia etiology are concurrent. Lastly, low-quality semen did not affect embryo implantation. Overall, these results promote the use of PGT-A.
BMI, Body-mass index; FSH, Follicle-stimulating hormone; GnRH, Gonadotropin-releasing hormone; hCG, Human chorionic gonadotropin; HSA, Human serum albumin; HTF, human tubal fluid; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization; NGS, Next-Generation Sequencing; PCR, Polymerase chain reaction; PGT-A, Preimplantation Genetic Testing for Aneuploidies; WGA, Whole-genome amplification.
JP and ELB conceived and designed the study; JP, HS, and JC performed the experiments, data collection as well as monitored the patients; JP and ELB analyzed the data; JP and ELB wrote the manuscript and critically analyzed it; ELB supervised the study and was responsible for the funding. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.
In this retrospective study, no written informed consent was needed besides the general consent for data utilization; nevertheless, the study was conducted following the Declaration of Helsinki. The Ethics Committee of the Institute Ingenes México approved this study (approval number: ISF120516).
We want to express our gratitude to the study participants and the medical staff at Ingenes Institute. We thank Dr. Leonardo M. Porchia for drafting the manuscript and Ing. Lucero Cervantes for her editorial assistance.
The study was funded by Consejo Nacional de Ciencia y Tecnología (Conacyt grant number: 250768) to ELB. The funding source did not have a role in the study design, collection, analysis, and interpretation of data and no part in the report’s writing and submitting the paper for publication.
The authors declare no conflict of interest.