This is the first comprehensive retrospective analysis of cumulative live birth rates stratified by age in women > 42 using the SART CORS. This study is an analysis of the first 7 years (2014–2020) of linked ART cycle data and CLBR from one of the largest standardized and validated data sets available in the United States.

It has been reported that CLBR for women 42 or younger generally plateau by the 4th to 5th ART cycle [14,15,16,17]. According to our analysis of the SART CORS data set, CLBRs plateau after fewer and fewer cycles for women ≥ 43. Our findings indicate that women 43 and 44 years old can be informed that CLBR is ~ 9% and 7%, respectively, however, it is unlikely that that additional live births will occur beyond three retrieval cycles. For 43 year old women, the CLBR after three retrievals was ~ 8.5% and by the fifth retrieval the CLBR reached a maximum of 9.7%. Similarly, for those 44 years old, the CLBR reached 7.5% by the third retrieval cycle with a maximum CLBR noted by cycle five at 8.6%. While a statistical difference was noted between cycles three and five for both these ages, the impact of each additional retrieval cycle was well less than 1% for each retrieval beyond the third. Therefore, the clinical utility of additional retrievals for 43 and 44 year old women is questionable.

Women 45 and 46 years old should be informed that CLBR is approximately 4% and 3%, respectively, and that each subsequent retrieval beyond the second (4% and ~ 3% respectively) affects the CLBR by < 0.5% to yield a maximum CLBR for these ages of ~ 5% and 3.6% respectively. CLBR using autologous eggs are minimal for women ≥ 47 years old, ranging from 1–2% and do not change appreciably beyond the initial cycle. The per cycle LBR is 1% or less for women over the age of 46. These observations are consistent with the fact that at age 43, blastocyst aneuploidy rates approach 85% and continue to increase with advancing age [18, 19].

Our findings reveal a clear trend of decreasing number of embryos transferred and blastocyst transfers with increasing age > 42 (Fig. 1). These factors, as indicated by our data in Table 2, emerge as the most influential predictors of live birth. Consequently, it is not surprising that the CLBR for women aged 47 or older remains relatively unchanged after the first treatment cycle. These findings are reinforced by the elevated incidence of aneuploidy observed in blastocyst biopsies from women over 44 [19]. However, it is noteworthy that preimplantation genetic testing (PGT) was associated with decreased odds of live birth in this specific population. This may be due to the observed lower utilization of PGT in older age groups and the limited availability of embryos for testing. For women 43 and older, we noted that markers of ovarian reserve did not appear to be predictive of live birth despite prior findings looking at aggregated age group data (> 42) suggesting the contrary [20]. This is likely due to the fact that AMH values are so low in these advanced age groups that it loses its predictive ability.

The current analysis of this large dataset helps frame expectations for maximum CLBR beyond age 42. Our data suggests that women > 42 years old reach a maximum CLBR after 1–5 cycle retrieval cycles, depending on their specific age. For those who do not yield transferable embryos, the value of IVF with autologous eggs could be questioned after fewer attempts. Because all transfer attempts are linked to a single retrieval, each retrieval may yield multiple transfer attempts and therefore number of “cycles” to achieve a live birth may be underrepresented in this study. Accordingly, our findings add credence to the need for early counseling regarding alternatives to autologous ART (donor egg or adoption) when it comes to the decision-making process attempts at autologous ART for women ≥ 44 years of age.

The CLBRs presented here are based on cycle starts. The relatively high cancellation rates reported in the setting of autologous eggs beyond age 40 predict lower CLBR [9]. The ASRM defines futility as ≤ 1% probability for pregnancy [21]. Considering the data presented here, autologous IVF can be considered futile beyond the first attempt at ages 46 and 47 and with essentially any attempt for women older than 47. This is consistent with the findings of Gunnala et al. who examined IVF cycles in women ≥ 45 [9]. They reported that approximately 40% of cycles could either not start due to findings at baseline or were cancelled prior to retrieval. For those going to transfer, LBR was 3.4%. Others have reported on the futility of LBRs in women 45 years and older [10, 11, 22]. Our findings are more comprehensive than, but remain consistent with a prior report using the SART CORS database [16].

Diminished ovarian reserve is an inevitable part of the human experience. While there have been anthropological rationalizations for the process, they do not console those who are faced, late in life, with the desire to conceive [23, 24]. We were left with a precipitous decline in fecundity during the fifth decade of life [25]. Contemporary population-based data for specific ages are lacking, but grouped data mirror our findings. In 2022, the National Center for Health Statistics reported that the annual provisional birth rate for women between 40-44 was 12.5 births per 1000 women compared to 1.1 births per 1000 women between ages 45-49 [26, 27]. While these data do not allow for a direct comparison to ART outcomes, they demonstrate that birth rates in the general population and the ART population become more comparable beyond age 45. Currently there remain no clear and proven therapeutic options for clock reversal, ovarian regeneration, or oocyte replenishment. What remains, is the application of a technology that has proven success for youthful individuals to a population with little demonstrable benefit [8]. We provide a representative guide, derived from a large and validated contemporary data set, to allow women in their mid- to late-forties to make informed decisions about their reproductive options. Patient centered care requires a clear understanding of the likelihood of the desired health outcome. It is incumbent on care providers to counsel, and for patients to be fully informed, when the probability of live birth is exceedingly low. Such counseling requires data and neither speculation nor a potentially biased opinion. Our data allow for the provision of such patient centered care and the discussion of realistic patient and physician expectations in the context of women age 43 and older.

Our study is limited by its retrospective nature. Prospective controlled trials assessing reproductive outcomes at the twilight of reproductive potential are rare and limited. Morgia et al. conducted an RCT to evaluate different COH regimens and noted low implantation rates for women > 40, however, women > 43 were not included [28]. Such center specific reporting is bound to be limited by small sample size and lack of power. The paucity of controlled IVF trials for extremely advanced maternal age is likely due to limited resources and the relatively small percentage of patients in these age groups. Reporting is thereby limited to large retrospective datasets. Our study is strengthened by its sample size, the largest involving these age ranges to date, and its reliance on a validated national data set of contemporary ART practices. Using the SART CORS database, we were able to discreetly examine the potential for cumulative live birth, arguably the most meaningful IVF outcome, at the terminal boundary of reproductive life.

We show that beyond age 45 live birth outcomes become quite rare. Meager odds may not dissuade some, but it is incumbent upon providers to be able to provide realistic likelihoods for desired goals. Vaguely stating low odds or providing wide ranges cannot suffice as satisfactory when stakes and costs are as high as they are for a single ART cycle. There remain circumstances when cost to the patient is not an issue, as when insurance coverage is guaranteed no matter the prognosis. In such situations, when treatment becomes futile, there is a clear need for objective data as presented here. As noted by ASRM, not offering treatment in the context of futility is an appropriate expression of professional integrity [21]. Stopping treatment may run counter to the patient’s desire. However, cessation of treatment may remain justified because continuing can prevent patients from clearly addressing other options. This is an area that needs to be consistent across an IVF center and not appear arbitrary [21].