Live birth outcomes after euploid transfer: autologous vs. donor oocyte embryos in patients aged > 35 years.

Peter Lindner: Writing – review & editing, Writing – original draft, Investigation, Conceptualization. Kerry Flannagan: Writing – review & editing, Formal analysis, Conceptualization. Howard J. Li: Writing – original draft. Jerry Wang: Investigation. Luis R. Hoyos: Writing – review & editing, Investigation. Meike L. Uhler: Writing – review & editing, Investigation. Michael Homer: Writing – review & editing, Investigation. Kate Devine: Writing – review & editing. Micah Hill: Writing – review & editing. Phillip Romanski: Writing – review & editing, Supervision, Conceptualization.

The study included a total of 15,013 single euploid FET cycles, which included 13,883 autologous cycles and 1,130 donor oocyte cycles. Demographics and cycle characteristics are shown in Table 1 . The mean age at the time of autologous oocyte retrieval was 37.9 years, whereas donor-derived oocytes were retrieved at a mean age of 26.9 years. A diagnosis of diminished ovarian reserve was observed in 20.9% of the autologous cohort and 76.0% in the donor oocyte cohort. Embryo characteristics were similar between the 2 groups. Embryos were most commonly biopsied and vitrified on day 5 in both groups. Similarly, most embryos transferred were graded as “good” by simplified SART grading scale. Table 1 Demographic and cycle characteristics of the study population. Characteristics Autologous N = 13,883 Donor N = 1,130 Age of oocyte at time of retrieval a (y) 37.9 ± 2.30 26.6 ± 3.09 Female race b  Caucasian 6,961 (50.1%) 631 (55.8%)  African American 1,281 (9.2%) 80 (7.1%)  Hispanic 1,034 (7.4%) 83 (7.3%)  Asian 2,483 (17.9%) 173 (15.3%)  Other 229 (1.6%) 14 (1.2%)  Declined/unknown 1,895 (13.6%) 149 (13.2%) Primary infertility diagnosis b  Male factor 2,572 (19.1%) 57 (5.3%)  Unexplained 2,267 (16.8%) 26 (2.4%)  Anovulation 1,308 (9.7%) 11 (1.0%)  Tubal 951 (7.1%) 21 (1.9%)  Diminished ovarian reserve 2,814 (20.9%) 819 (76.0%)  Endometriosis 393 (2.9%) 18 (1.7%)  Uterine 326 (2.4%) 9 (0.8%)  Other 2,841 (21.1%) 116 (10.8%) Recipient BMI a (kg/m 2 ) 26.5 ± 5.5 25.8 ± 5.2 Recipient gravidity c 1 (0–2) 0–17 1 (0–2) 0–9 Recipient parity c 0 (0–1) 0–6 0 (0–1) 0–5 FET preparation b  Programmed 11,445 (82.4%) 940 (83.2%)  Natural cycle 2,438 (17.6%) 190 (16.8%) Peak endometrial thickness c (mm) 9.6 (8.5–11.2) 9.5 (8.4–11.1) Age of embryo b  Day 5 7,895 (57.8%) 666 (60.6%)  Day 6 5,445 (39.9%) 415 (37.8%)  Day 7 313 (2.3%) 18 (1.6%) Embryo quality b  Good 9,901 (76.4%) 823 (79.1%)  Fair 3,019 (23.3%) 214 (20.6%)  Poor 42 (0.3%) 4 (0.4%) Note: BMI = body mass index; FET = frozen embryo transfer. a Mean ± SD. b n (%). c Median (interquartile range [25%–75%]). Demographic and cycle characteristics of the study population. Note: BMI = body mass index; FET = frozen embryo transfer. Mean ± SD. n (%). Median (interquartile range [25%–75%]). Table 2 presents the clinical outcomes analyzed for the autologous oocyte–derived euploid embryo transfer cycles. Live birth occurred in 55.8% of all euploid FETs in patients aged 35–37 years, the referent age group. There was a linear decrease in the live birth rate as age group increased ( P <.0001) down to 43.4% among patients aged ≥45 years. Although the P value for trend indicated a statistically significant decrease in positive hCG with age ( P <.0001), relative risk estimates indicated a U-shaped association with similar chance of hCG pregnancy at the youngest and oldest ages and lower probability from 41 to 44 years. The clinical pregnancy rate significantly decreased ( P <.0001) as oocyte age increased and when compared with the referent group. Increasing age was also associated with an increasing risk of spontaneous abortion ( P =.0004); however, the risk of a biochemical loss was not significantly different as age increased ( P =.08). Table 2 Pregnancy outcomes among patients undergoing a euploid embryo transfer, autologous cycles, and donor oocyte cycles a . Autologous oocyte cycles 35–37 y N = 6,136 ( Referent ) 38–40 y N = 5,421 41–42 y N = 1,801 43–44 y N = 472 ≥45 y N = 53 P , trend Positive hCG 4,518 (73.9%) 3,941 (72.8%) 1,236 (68.7%) 308 (65.7%) 39 (73.6%) <.0001 Ref 0.99 (0.96–1.01) 0.93 (0.90–0.96) 0.89 (0.83–0.95) 1.0 (0.85–1.17) Biochemical loss 514 (11.4%) 499 (12.7%) 158 (12.8%) 32 (10.4%) 10 (25.6%) .08 Ref 1.11 (0.99–1.25) 1.12 (0.95–1.33) 0.91 (0.65–1.28) 2.25 (1.31–3.87) Spontaneous abortion 523 (11.6%) 522 (13.2%) 177 (14.4%) 49 (15.9%) 6 (15.4%) .0004 Ref 1.15 (1.02–1.28) 1.24 (1.06–1.45) 1.37 (1.05–1.80) 1.33 (0.63–2.79) Clinical pregnancy 3,975 (65.0%) 3,418 (63.2%) 1,071 (59.6%) 276 (58.8%) 29 (54.7%) <.0001 Ref 0.97 (0.95–1.00) 0.92 (0.88–0.96) 0.91 (0.84–0.98) 0.84 (0.66–1.08) Live birth 3,409 (55.8%) 2,838 (52.5%) 878 (48.9%) 224 (47.8%) 23 (43.4%) <.0001 Ref 0.94 (0.91–0.97) 0.92 (0.88–0.92) 0.86 (0.78–0.94) 0.78 (0.57–1.06) Donor oocyte cycles 35–37 y N = 98 38–40 y N = 168 ( Referent ) 41–42 y N = 206 43–44 y N = 268 ≥45 y N = 390 P , trend Positive hCG 61 (62.2%) 0.87 (0.73–1.04) 120 (71.4%) Ref 147 (71.4%) 1.00 (0.88–1.14) 183 (68.3%) 0.96 (0.84–1.08) 259 (66.4%) 0.93 (0.83–1.05) .91 Biochemical loss 13 (21.3%) 17 (14.2%) 22 (15.0%) 26 (14.2%) 40 (15.4%) .49 1.50 (0.78–2.89) Ref 1.06 (0.59–1.90) 1.00 (0.57–1.77) 1.09 (0.65–1.84) Spontaneous abortion 9 (14.8%) 18 (15.0%) 23 (15.8%) 30 (16.4%) 33 (12.7%) .64 0.98 (0.47–2.06) Ref 1.05 (0.60–1.85) 1.09 (0.64–1.87) 0.85 (0.50–1.45) Clinical pregnancy 48 (49.0%) 102 (60.7%) 124 (60.2%) 155 (57.8%) 217 (55.6%) .80 0.81 (0.64–1.02) Ref 0.99 (0.84–1.17) 0.95 (0.81–1.12) 0.92 (0.79–1.07) Live birth 37 (37.8%) 83 (49.4%) 96 (47.1%) 121 (45.1%) 179 (45.9%) .58 0.76 (0.57–1.03) Ref 0.95 (0.77–1.18) 0.91 (0.75–1.12) 0.93 (0.77–1.12) Note: Positive human chorionic gonadotropin (hCG) defined as an hCG level of >5 mIU/mL 10 days after frozen embryo transfer. Pregnancy loss included biochemical pregnancies and spontaneous abortions. Biochemical loss defined as positive hCG with return of hCG level to <5 mIU/mL before the visualization of the pregnancy on ultrasound. Spontaneous abortion defined as loss of a clinical pregnancy before 22 weeks of gestation. Live birth defined as delivery of a viable fetus at > 22 completed weeks of gestation. Totals may not sum precisely because of a small number of outcomes not categorized under live birth, spontaneous abortion, or clinical pregnancy including stillbirths, therapeutic abortions, and ectopic pregnancies. a The models for early loss outcomes were restricted to hCG pregnancies. Live birth outcomes were per transfer. Pregnancy outcomes among patients undergoing a euploid embryo transfer, autologous cycles, and donor oocyte cycles a . Note: Positive human chorionic gonadotropin (hCG) defined as an hCG level of >5 mIU/mL 10 days after frozen embryo transfer. Pregnancy loss included biochemical pregnancies and spontaneous abortions. Biochemical loss defined as positive hCG with return of hCG level to <5 mIU/mL before the visualization of the pregnancy on ultrasound. Spontaneous abortion defined as loss of a clinical pregnancy before 22 weeks of gestation. Live birth defined as delivery of a viable fetus at > 22 completed weeks of gestation. Totals may not sum precisely because of a small number of outcomes not categorized under live birth, spontaneous abortion, or clinical pregnancy including stillbirths, therapeutic abortions, and ectopic pregnancies. The models for early loss outcomes were restricted to hCG pregnancies. Live birth outcomes were per transfer. Results for donor oocyte–derived euploid embryo transfer cycles by age of the female recipient at time of embryo transfer are also shown in Table 2 . Live birth occurred in 49.4% of euploid FETs in the 38–40-year group, the referent group. Live birth in the 35–37-year group occurred in 37.8% of euploid FETs, which decreased compared with the referent group (RR, 0.76; 95% CI, 0.57–1.03). However, across all ages, there was no statistically significant decrease in live birth rate ( P =.58). Live birth occurred in 47.1% (RR, 0.95; 95% CI, 0.77–1.18) of euploid FETs in the 41–42-year old group, 45.1% (RR, 0.91; 95% CI, 0.75–1.12) of euploid FETs in the 43–44-year old group, and 45.9% (RR, 0.93; 95% CI, 0.77–1.12) in the ≥45-year group. Furthermore, there was no statistically significant difference between the age groups for any of the secondary outcomes including positive hCG, biochemical loss, spontaneous abortion, and clinical pregnancy ( P =.81). Clinical outcomes were directly compared between the autologous and donor euploid embryo transfer cycles for each age group as shown in Table 3 . Live birth was statistically significantly different between donor and autologous cycles in the 35–37-year group with donor cycles having a decreased chance of live birth (37.8% vs. 55.8%; RR, 0.68; 95% CI, 0.53–0.87) as shown in Figure 1 . There was no difference in live birth between donor and autologous euploid FETs in the other age groups; in the 38–40-year group (49.4% vs. 52.5%; RR, 0.94; 95% CI, 0.81–1.10); 41–42-year group (47.1% vs. 48.9%; RR, 0.96; 95% CI, 0.83–1.12); 43–44-year group (45.1% vs. 47.8%; RR, 0.95; 95% CI, 0.80–1.11); and ≥45-year group (45.9% vs. 43.4%; RR, 1.06; 95% CI, 0.76–1.47). Similarly, chance of clinical pregnancy was noted to be significantly decreased in donor patients aged 35–37 years compared with autologous cycles (65.0% vs. 49.0%; RR, 0.75; 95% CI, 0.62–0.92). There was no statistically significant difference in clinical pregnancy between cohorts in any of the other age stratifications. Furthermore, outcomes of biochemical loss and spontaneous abortion were not significantly different between cohorts in any of the age stratifications. Table 3 Age category comparison of pregnancy outcomes between autologous and donor euploid embryo transfer cycles a . Outcome Autologous Donor Autologous Donor Autologous Donor Autologous Donor Autologous Donor 35–37 y N = 6,136 35–37 y N = 98 RR (95% CI) donor vs. autologous 38–40 y N = 5,421 38–40 y N = 168 RR (95% CI) donor vs. autologou 41–42 y N = 1,801 41–42 y N = 206 RR (95% CI) donor vs. autologous 43–44 y N = 472 43–44 y N = 268 RR (95% CI) donor vs. autologous ≥45 y N = 53 ≥45 y N=390 RR (95% CI) donor vs. autologous P , interaction Positive hCG 4,518 (73.9%) 61 (62.2%) 0.84 (0.72–0.98) 3,941 (72.8%) 120 (71.4%) 0.98 (0.89–1.08) 1,236 (68.7%) 147 (71.4%) 1.04 (0.95–1.14) 308 (65.7%) 183 (68.3%) 1.04 (0.94–1.15) 39 (73.6%) 259 (66.4%) 0.9 (0.76–1.08) .17 Biochemical loss 514 (11.4%) 13 (21.3%) 1.87 (1.15–3.06) 499 (12.7%) 17 (14.2%) 1.12 (0.71–1.75) 158 (12.8%) 22 (15.0%) 1.17 (0.78–1.77) 32 (10.4%) 26 (14.2%) 1.37 (0.84–2.22) 10 (25.6%) 40 (15.4%) 0.60 (0.33–1.11) .20 Spontaneous abortion 523 (11.6%) 9 (14.8%) 1.27 (0.69–2.34) 522 (13.2%) 18 (15.0%) 1.13 (0.73–1.75) 177 (14.4%) 23 (15.8%) 1.10 (0.74–1.64) 49 (15.9%) 30 (16.4%) 1.03 (0.68–1.56) 6 (15.4%) 33 (12.7%) 0.83 (0.37–1.85) .09 Clinical pregnancy 3,975 (65.0%) 48 (49.0%) 0.75 (0.62–0.92) 3,418 (63.2%) 102 (60.7%) 0.96 (0.85–1.09) 1,071 (59.6%) 124 (60.2%) 1.01 (0.90–1.14) 276 (58.8%) 155 (57.8%) 0.98 (0.87–1.12) 29 (54.7%) 217 (55.6%) 1.02 (0.78, 1.32) .09 Live birth 3,409 (55.8%) 37 (37.8%) 0.68 (0.52–0.87) 2,838 (52.5%) 83 (49.4%) 0.94 (0.81–1.10) 878 (48.9%) 96 (47.1%) 0.96 (0.83–1.12) 224 (47.8%) 121 (45.1%) 0.95 (0.80–1.11) 23 (43.4%) 179 (45.9%) 1.06 (0.76–1.06) .02 Note: Positive human chorionic gonadotropin (hCG) defined as an hCG level of >5 mIU/mL 10 days after frozen embryo transfer. Biochemical loss defined as positive hCG with return of hCG level to <5 mIU/mL before the visualization of the pregnancy on ultrasound. Spontaneous abortion defined as loss of a clinical pregnancy before 22 weeks of gestation. Live birth defined as delivery of a viable fetus at > 22 completed weeks of gestation. a The models for early loss outcomes were restricted to hCG pregnancies. Live birth outcomes were per transfer. Figure 1 Live birth among autologous and egg donor euploid embryo transfers stratified by age group. The asterisk(∗) denotes an effect estimate P <.05 between the two arms in the same age group. NS = not significant (denotes an effect estimate that was not statistically significant). Age category comparison of pregnancy outcomes between autologous and donor euploid embryo transfer cycles a . Note: Positive human chorionic gonadotropin (hCG) defined as an hCG level of >5 mIU/mL 10 days after frozen embryo transfer. Biochemical loss defined as positive hCG with return of hCG level to <5 mIU/mL before the visualization of the pregnancy on ultrasound. Spontaneous abortion defined as loss of a clinical pregnancy before 22 weeks of gestation. Live birth defined as delivery of a viable fetus at > 22 completed weeks of gestation. The models for early loss outcomes were restricted to hCG pregnancies. Live birth outcomes were per transfer. Live birth among autologous and egg donor euploid embryo transfers stratified by age group. The asterisk(∗) denotes an effect estimate P <.05 between the two arms in the same age group. NS = not significant (denotes an effect estimate that was not statistically significant).

This retrospective cohort study included patients aged 35–50 years who underwent a single euploid frozen embryo transfer (FET) between January 2010 and January 2023 at a large US fertility network. Both autologous and donor oocyte–derived embryos were included, analyzing the first embryo transfer only. If a patient had both autologous and donor oocyte FETs, only the first type was considered. An upper age limit of 50 years was selected on the basis of our fertility practices’ protocol, which allows for an embryo to be transferred to patients deemed healthy enough for pregnancy up until their 51st birthday ( 13 ). Autologous transfers beyond 24 months after oocyte retrieval were excluded to ensure that oocyte age and uterine age were similar in the autologous oocyte group. This approach increases our confidence that outcomes in this cohort are representative of older oocyte age. Supplemental Table 1 (available online) reports the mean age of the oocyte at retrieval by age group and stratified by cycle type. Data on demographics, cycle characteristics, and outcomes were obtained from an electronic database. Institutional review board approval was obtained (IRB R2009.1). The primary outcome of live birth was defined as the delivery of a viable fetus >22 weeks of gestation ( 14 ). The secondary outcomes included positive human chorionic gonadotropin (hCG), biochemical loss, spontaneous abortion, and clinical pregnancy. Positive hCG was defined as an hCG level of >5 mIU/mL obtained via serum testing or positive home pregnancy test 10 days after FET. Biochemical loss was defined as an initial positive hCG with return of hCG level to <5 mIU/mL before the visualization of pregnancy on ultrasound. Clinical pregnancy was defined as an intrauterine pregnancy with at least one gestational sac visualized on ultrasound. Spontaneous abortion was defined as the spontaneous loss of a clinical pregnancy <22 weeks of gestation ( 14 ). Autologous oocytes were obtained via ovarian stimulation with injectable gonadotropins with either gonadotropin-releasing hormone antagonist or gonadotropin-releasing hormone agonist for luteinizing hormone (LH) suppression. Treatment cycles were monitored with ovarian follicular measurements using transvaginal ultrasound and serum hormone levels (estradiol, LH, and progesterone) every 1–3 days. When lead follicles reached >17 to 18 mm, a trigger injection (hCG, leuprolide acetate, or both) was given to resume oocyte meiosis. Ultrasound-guided oocyte retrieval was performed 35–36 hours after the trigger per institutional protocol, and mature oocytes underwent insemination via conventional methods or intracytoplasmic sperm injection as indicated. After fertilization, zygotes were placed in extended culture media. Embryos were assessed on days 5, 6, and 7, with trophectoderm biopsy for PGT for aneuploidy performed on the basis of progression, followed by vitrification. The simplified Society for Assisted Reproductive Technology (SART) grading system categorized embryo quality using previtrification morphology. Embryos graded AA or AB were categorized as good; BA, BB, or BC as fair; and CB or CC as poor ( 15 ). Donor oocytes were 21–34 years old and underwent ovarian stimulation as described. Retrieved mature oocytes were fertilized immediately or vitrified for later use. Blastocyst culture followed the same protocol as autologous embryos. Patients with a euploid embryo underwent an FET in a programmed or natural cycle at the discretion of the treating physician. In programmed FETs, patients started supplemental estrogen on cycle day 2–3 and continued for approximately 10–14 days. Transvaginal ultrasound assessed endometrial thickness, measured in the midsagittal plane. Luteal phase support involved either daily intramuscular progesterone or daily vaginal progesterone with intramuscular progesterone every third day once adequate endometrial thickness was achieved ( 16 ). Natural cycle FETs were monitored with transvaginal ultrasound and serum hormone levels for dominant follicle development and endometrial growth. Patients with adequate estradiol levels and endometrial thickness were monitored for an LH surge or given an hCG trigger. Serum hCG was measured 10 days after the FET and, if positive, serially remeasured to evaluate for an adequate increase. Transvaginal ultrasound was performed to confirm the presence of an intrauterine pregnancy between 5 and 6 weeks, with a follow-up ultrasound 1–2 weeks later before discharge to the obstetrician at approximately 8 weeks of gestation. Sociodemographic and cycle characteristics were compared between cohorts using means and standard deviations, medians and interquartile ranges, or numbers and percentages as appropriate. In the primary analysis, patients were categorized by their age at FET, stratified by SART age categories, with the >42-year group further divided into 43–44 and >44 years. The primary and secondary outcomes were assessed among autologous and donor oocyte–derived euploid embryos. Risk ratios (RRs) and 95% CIs comparing outcomes in each age group to the 35–37-year referent group were estimated using modified Poisson regression models fitted with generalized estimating equations. In the donor oocyte subgroup analysis, the live birth rate was lower than expected in the 35–37-year group, and thus, the 38–40-year group was used as the referent to aid in interpretability of the RRs. Models for biochemical loss and spontaneous abortion were fitted only for hCG pregnancies; all other outcomes were fitted for all transfers. In a secondary analysis, autologous and donor oocyte outcomes were directly compared within age categories. Each outcome was analyzed using modified Poisson regression models with generalized estimating equations, incorporating main and interaction terms for transfer age and donor status. Biochemical loss and spontaneous abortion models were restricted to hCG pregnancies, and live birth outcomes were per transfer. Differences in donor-outcome associations by age group were assessed using the Wald test of the interaction term, yielding P values for interaction.

Among patients aged ≥38 years who transfer a euploid embryo, the live birth rate declines with increasing age when an autologous embryo is transferred but not when a donor oocyte–derived embryo is transferred. Therefore, age-related uterine factor does not explain the decrease in live birth rate with increasing maternal age when transferring euploid embryos. Factors related to the age of the oocyte such as cytoplasmic competency and epigenetics, even in euploid embryos, are more likely to explain these results. This study provides insight into possible mechanisms for pregnancy failures apart from traditional embryo chromosome copy number and supports that other embryonic factors that may additionally be impacted by increased age of the oocyte, such as epigenetics, transcriptomics, and metabolomics, may, in part, explain why up to 50% of euploid embryo transfers do not result in a live birth. These results underscore the complexity of the implantation process and open avenues for exploring a broader spectrum of contributing factors. In our subanalysis that compared autologous with donor oocyte–derived euploid transfers within SART age categories, we observed comparable live birth for all age categories of ≥38 years. Thus, despite the statistically significant decline in live birth rate with increasing maternal age observed in the autologous embryo cohort, the absolute chance of live birth with the transfer of a euploid embryo remains high for these patients even among patients aged ≥45 years. The absence of a decline in live birth rate with increasing maternal age in the donor oocyte cohort suggests that there may be age-related oocyte factors that contribute to some failed euploid embryo transfers, and for these patients, the option of donor oocyte treatment could be explored as a next step before considering the use of a gestational carrier.

The aim of this study was to investigate the potential association between embryonic and nonembryonic factors on the declining likelihood of live birth with increasing maternal age by comparing autologous euploid FET outcomes to donor oocyte–derived euploid FET outcomes. We observed that among autologous euploid embryo transfer cycles, there was a significant linear decrease in live birth rate as maternal age increased, as has been observed in previous autologous euploid embryo cohorts. However, among donor oocyte–derived euploid embryo transfers, chance of live birth remained comparable and did not decline as maternal age increased. Similarly, the chance of spontaneous abortion increased with increasing maternal age in patients using autologous euploid embryos but not in those using donor oocyte–derived euploid embryos. Moreover, we observed that pregnancy outcomes were similar when comparing donor vs. autologous transfers within age groups, indicating that any differences we report with increasing age by donor status were not likely attributable to differences in the donor and autologous patient populations. As maternal age increases, decline in fecundity is primarily attributed to a higher prevalence of aneuploidy and eventually to a depleted pool of oocytes ( 1 , 7 , 17 ). Aneuploidy rates have been studied by observing oocytes from females aged 9–43 years, which indicated that the natural fertility pattern can be represented by an inverse U-curve ( 2 ). Aneuploidy can have multiple potential etiologies including nondisjunction of homologous chromosomes (in meiosis I) or sister chromatids (meiosis II), premature separation of sister chromatids, or reverse segregation ( 17 ). Increasing age is most associated with chromosome segregation errors caused by premature separation of sister chromatids and reverse segregation ( 7 ). It is known that the increase in embryo aneuploidy associated with increasing maternal age is the primary cause of decreasing live birth in these patients; however, even among euploid embryo transfers, there is a small but clear inverse association between maternal age and live birth. There are two plausible explanations for this which are as follows: 1) an increase in abnormal embryonic factors (genetic abnormalities not detected by PGT for aneuploidy, epigenetic changes, or aberrant changes to transcriptomics and metabolomics) or 2) an increase in nonembryonic factors such as uterine vascular supply, maternal endocrine environment, and endometrial receptivity. Our study suggests that the age-associated decrease in live birth rate after euploid FET is secondary to the former given that the live birth rate was only observed to decline as the oocyte age increased. A recent study explored whether the age of the uterus is also associated with live birth by evaluating how maternal age at the time of euploid embryo transfer after autologous oocyte cryopreservation impacts live birth ( 18 ). The investigators observed that live birth was not associated with maternal age at the time of transfer when controlling for maternal age at time of retrieval, day of blastulation, and embryo morphology ( 18 ). We observed similar findings in our cohort of patients undergoing donor oocyte–derived euploid embryo transfer. As donor oocytes are generally retrieved from donors between the ages of 21 and 32 years with a mean age of 26.9 years in our cohort, the comparable live birth outcomes with increasing age of the female recipients using embryos derived from this younger group of oocytes suggest that age-related changes to the uterus are not associated with live birth outcomes. In our cohort patients in the youngest age category, 35–37 years, we observed a significantly decreased live birth rate with the use of donor oocyte–derived embryos compared with autologous embryos. Although this may reflect random error due to a smaller number of patients in this age category using donor oocytes, as expected, another explanation is that there may be inherent differences in patients who need to use donor oocyte treatment at such a young age that may portend a worse prognosis. These patients are more likely to represent those with unusually poor oocyte quality or premature ovarian insufficiency, which is most commonly caused by genetic disorders, autoimmune disorders, and iatrogenic causes such as chemotherapy and radiotherapy, all of which are nonembryonic exposures—conditions that often necessitate the use of donor oocytes. Although the oocyte quality in these cases is addressed through donor gametes, these same conditions can also negatively impact the uterine environment, potentially compromising implantation or pregnancy maintenance ( 19 ). A recent systematic review of the literature on advanced paternal age and in vitro fertilization outcomes found mixed results. Although most studies reported that paternal age over 50 years was associated with decreased fertilization and blastulation rates, these effects were not consistently observed at younger paternal ages ( 20 ). The mean age of the male sperm source in each age-stratified group is presented in Supplemental Table 2 and were all <50 years old. Male partner age increased alongside female partner age and was similar between the autologous and donor oocyte groups. This parallel trend supports the conclusion that the observed decline in live birth rates with advancing maternal age in autologous euploid transfers is attributable to oocyte age. Supporting this, the donor oocyte group—despite also having an increase in the age of male partners alongside female partner’s age—did not show a corresponding decline in live birth rates. A recent retrospective cohort analysis of euploid FETs demonstrated that day 7 blastocysts have a lower likelihood of resulting in live birth, even when controlling for aneuploidy status ( 21 ). To explore whether blastulation timing contributed to our findings, Supplemental Table 3 presents the distribution of transferred embryos by day of blastulation (day 5, 6, or 7), stratified by age group and cycle type (autologous vs. donor). We found that very few day 7 embryos were transferred across all age groups and cycle types. However, the proportion of day 6 transfers increased with age in the autologous cohort but was not associated with age in the donor cohort, indicating that blastulation timing may be one of the oocyte age–related factors (or a proxy of one of these factors) that impacts transfer success. In addition, published studies suggest that poor embryo morphology is associated with lower live birth rates ( 15 ). Supplemental Table 4 shows the distribution of transferred embryo morphology (poor, fair, or good), stratified by age and cycle type. Very few poor-quality embryos were transferred, and this remained consistent even with increasing age in both the autologous and donor groups. This supports the conclusion that the observed decline in live birth rates with increasing age is not driven by an increased use of poor-quality embryos in older patients. Furthermore, most transferred embryos were graded as good quality, regardless of age group or cycle type, reinforcing that embryo morphology does not fully explain the age-related decline in live birth rate among euploid embryos in autologous cycles. This study has several strengths and limitations to acknowledge. First, the study is well powered; the sample size of 15,013 euploid embryo transfers is larger than the combined number of embryo transfers included in the largest meta-analysis on this topic (approximately 11,000 transfers pooled from 7 separate studies) ( 6 ). This allowed us to perform a robust analysis separately for autologous embryos and donor oocyte–derived embryos and to make comparisons within SART age groups. Of note, the sample size in some age categories are smaller, particularly for the age of 35–37 years using donor oocyte, and thus, definitive conclusions cannot be made regarding differences in outcomes among patients in this age category. Additionally, this study was designed as a multisite study including patients from clinics spread throughout the United States. There is likely inherent variability in some of the laboratory and clinical protocols used by each clinic that we were unable to account for this study. However, the geographic distribution of the included study population does increase the external validity and generalizability of our results.

P.L. has nothing to disclose. The views expressed in this paper are those of the authors and do not necessarily reflect the offical policy or position of the Department of the Navy, Department of the Air Force, Department of the Army, Department of Defense, or the United States Government. K.F. is an employee of US Fertility, which provided the data for this manuscript. H.J.L. has nothing to disclose. J.W. has nothing to disclose. L.R.H. reports honorarium from the Midwest Reproductive Symposium board review course; Board Member of the Midwest Reproductive Symposium; and stock options from the US Fertility. M.L.U. has nothing to disclose. M.Ho. has nothing to disclose. K.D. has nothing to disclose. M.H. has nothing to disclose. P.R. has nothing to disclose.