Miscarriage is the most prominent adverse outcomes associated with AMA. Rates rise from 15% to 20% among women aged 35–39 to about 40% at 40–44, and up to 70% in women aged 45 or older (8, 11). Oocyte aneuploidy, with chromosomal abnormalities constituting the predominant cause of early pregnancy loss, is considered the principal underlying mechanism, particularly when native oocytes are used (5). The risk of gestational diabetes also increases substantially with age. Women aged 40 or older face a three- to six-fold higher risk compared with younger counterparts (13), with prevalence approaching 20% in women over 50 (7). This reflects both age-related alterations in glucose metabolism and the higher baseline prevalence of metabolic syndrome in older women. Hypertensive disorders of pregnancy, particularly preeclampsia, represent another major concern. Prevalence increases from 3%–4% in younger mothers to 5%–10% by age 40, and up to 30% in women over 50 (9, 10). Vascular stiffness, endothelial dysfunction, and impaired placental adaptation are believed to be central mechanisms. AMA also raises the risk of placental complications, including placenta previa, placenta accreta, and placental insufficiency. These conditions increase maternal morbidity through hemorrhage and surgical complications while compromising fetal growth and oxygenation (12). Cesarean delivery rates rise sharply with maternal age. While some are elective, many procedures are prompted by medical indications such as abnormal placentation, malpresentation, and labor dystocia. Among women over 45, more than half of deliveries occur via cesarean section (6, 21). Preterm birth is more common in pregnancies affected by AMA, frequently resulting from medically indicated early delivery due to maternal or fetal complications. Stillbirth risk also rises substantially: at 39 weeks, women over 40 face a stillbirth risk comparable to that of women under 30 at 41 weeks (11). Finally, infants born to AMA mothers are roughly twice as likely to require admission to a neonatal intensive care unit (NICU), reflecting complications related to both prematurity and placental dysfunction (4, 6, 10, 12).

ART have transformed infertility care, yet their benefits are tempered by clinically meaningful risks. Some complications stem from the underlying causes of infertility, while others appear attributable to ART procedures themselves (3, 14), making their relative contributions difficult to disentangle. Conditions, such as endometriosis, congenital uterine anomalies, or diminished ovarian reserve independently predispose to adverse obstetrics outcomes regardless of mode of conception. At the same time, ART-specific processes, including ovarian hyperstimulation, gametes and embryos manipulation, and embryo transfer mechanics, may exert additional influence (13). Pregnancies achieved through IVF, particularly those involving donor oocytes, carry a two- to threefold higher risk of preeclampsia (22). Proposed mechanisms include altered maternal immune tolerance to paternal or donor antigens, as well as epigenetic changes induced during ART procedures. Placental abnormalities, including placenta previa and aberrant implantation, occur more frequently in IVF pregnancies (15). It has been suggested that uterine instrumentation and the placement of embryos within the uterine cavity may alter implantation dynamics.

Importantly, elevated risks are observed even in singleton IVF pregnancies. Compared with spontaneously conceived singletons, IVF singletons have approximately a 1.5-fold increased risk of preterm birth and low birth weight, independent of multiple gestations (3, 23). Cesarean delivery rates are also markedly higher among ART pregnancies (14). While a subset of cases may reflect patient or provider preferences, many are driven by elevated rates of obstetric complications requiring surgical delivery (13). Finally, newborns conceived via ART are also more likely to require NICU admission. This vulnerability reflects a constellation of factors, including prematurity, low birth weight, and higher rates of congenital anomalies (15).

When AMA and ART intersect, their associated risks appear not merely cumulative but amplified. This elevated risk profile is not limited to oocyte aging and is evident even among recipients of donor oocytes. For example, preterm birth rates rise disproportionately when both factors coexist, driven by higher rates of both spontaneous preterm labor and medically indicated early delivery (17, 18). Likewise, cesarean delivery becomes more likely, reflecting the combined burden of age-specific complications and ART-associated obstetric risk (14, 16). Maternal complications such as hypertensive disorders, gestational diabetes, and placental pathologies are also magnified in this dual-risk group (19, 22). Neonatal outcomes are similarly affected: infants born to older women conceiving through ART are more likely to be preterm, have low birth weight, and require intensive neonatal care (15, 16).

Paradoxically, some ART techniques mitigate certain risks while simultaneously introducing new ones. Use of donor oocytes, for instance, reduces miscarriage risk associated with oocyte aneuploidy but appear to increase the likelihood of preeclampsia through immune-mediated pathways (22). In support of these observations, a recent large-scale analysis of more than 33,100 good-quality single-embryo transfer cycles involving donor oocytes concluded that live birth rates declined after maternal age 40, whereas implantation failure and pregnancy loss increased after ages 39 and 43 years, respectively (24).

In summary, the elevated obstetric risk observed when AMA and ART co-occur likely reflects the convergence of three interacting mechanistic axes (i) oocyte-intrinsic aging, (ii) maternal vascular, placental, and immune alterations, and (iii) ART-related procedural and hormonal exposure. Framing the evidence within these mechanistic categories helps explain why risks may appear additive in some settings and potentially synergistic in others. A central biological feature of AMA is the progressive increase in oocyte aneuploidy, which strongly contributes to implantation failure, early pregnancy loss, and reduced live birth rates. This risk is thought to arise from age-related deterioration of meiotic competence (e.g., reduced chromosomal cohesion, spindle assembly errors) and compromised oocyte quality, including mitochondrial dysfunction, altered energy metabolism, and epigenetic drift (25–27). In autologous ART cycles, these oocyte-intrinsic mechanisms remain the dominant drivers of implantation failure and miscarriage. Importantly, these pathways primarily explain early reproductive endpoints (fertilization, embryo viability, early loss), but do not fully account for the later gestational complications that persist even when younger donor oocytes are used. Beyond oocyte quality, AMA is associated with maternal vascular and endometrial changes that can impair placentation, including endothelial dysfunction, increased arterial stiffness, and altered decidualization and spiral artery remodeling (28). These abnormalities predispose to abnormal trophoblast invasion and suboptimal uteroplacental perfusion, contributing to hypertensive disorders, fetal growth restriction, and placental complications (29). Age-related immune remodeling further compounds this risk. The concept of “inflammaging”, a shift toward low-grade systemic inflammation and immune dysregulation, may disrupt the finely balanced immune tolerance required at the maternal-fetal interface (30, 31). Advanced maternal age has been shown to alter T-cell subsets, including reductions in regulatory T-cells at the maternal–fetal interface, which may impair tolerance and contribute to adverse outcomes (32). Perturbations in regulatory T cells also influence uterine natural killer cell function and decidual vascular remodeling, processes critical for spiral artery adaptation and placental development and implicated in preeclampsia-spectrum disorders (33). Broader immune dysregulation characterized by imbalanced cytokine profiles, NK cell activation, and altered macrophage polarization has also been associated with preeclampsia (34). A particularly illustrative scenario is donor oocyte pregnancy. Although donor oocytes substantially reduce miscarriage risk by mitigating aneuploidy, they introduce greater fetal-maternal antigenic disparity (35). The increased allogeneic challenge may amplify maternal immune tolerance with downstream effects on placental development, endothelial activation, and hypertensive disease (36). Multiple cohort studies have demonstrated a 2–3-fold increase in preeclampsia risk in donor oocyte pregnancies, particularly among women of advanced age (37), and meta-analyses confirm a broader association between ART and hypertensive disorders of pregnancy (38, 39).

ART itself may add risk through both indirect and direct mechanisms. Controlled ovarian stimulation can create supraphysiologic steroid environments (particularly in fresh embryo transfers) that may adversely affect endometrial receptivity and early placentation. Differences between fresh and frozen embryo transfer (FET) cycles also appear relevant: in programmed FET cycles, the absence of a corpus luteum eliminates physiologic secretion of vasoactive/angiogenic mediators (e.g., relaxin and other luteal products), which may influence maternal vascular adaptation and contribute to hypertensive complications in susceptible individuals (40, 41). In addition, embryo culture conditions, micromanipulation, cryopreservation, and transfer technique have been proposed to influence placental gene expression and epigenetic regulation, providing biologically plausible pathways by which ART may interact with age-related vascular immune signaling vulnerability (42).

Taken together, these pathways support a coherent model in which AMA contributes to oocyte-intrinsic risk (particularly in autologous cycles) and a baseline predisposition to vascular, immune and placental maladaptation, while ART introduces additional hormonal and procedural exposures that may further perturb implantation, placentation, and maternal-fetal immune tolerance. This integrated framework helps explain the persistence of elevated adverse outcomes even in singleton pregnancies and the distinctive risk profile of donor oocyte gestations, characterized by lower miscarriage rates but higher preeclampsia risk, and underscore the need to manage AMA + ART pregnancies as inherently high risk, warranting vigilant intensified surveillance and proactive clinical management (7).

Infertility has a substantial genetic component in both women and men. In women, rare Mendelian conditions accounts for 30%–40% of infertility cases (44) while approximately 5% exhibit FMR1 triplet repeats expansions or X-chromosome abnormalities (45). Heritable disorders such as endometriosis and polycystic ovary syndrome further contribute to female infertility. In males, chromosomal abnormalities, including Klinefelter syndrome or deletions within the azoospermia factor gene, are common causes of infertility (46). A recent genome-wide study involving 1.5 million individuals identified 22 genetic loci linked to female infertility, three to male infertility and novel variants associated with reproductive hormones, FSH, LH, estradiol and testosterone, levels, in men and women (47).

The trend toward AMA and the increasing use of ART have been linked to adverse obstetric and perinatal outcomes, with genetic factors expected to play an increasingly important role, in reduced fertility and negative reproductive outcomes (48). Reproductive genomics offers a pathway toward personalized reproductive medicine. Advances in sequencing technologies coupled with access to large-scale phenotype and genotype datasets now allow the integration of individual genomic data into clinical care. Accurate interpretation requires well-phenotyped, representative reference genomes. The Medical Genome Project, for example, uses exome sequencing of phenotyped control individuals to accelerate variant identification and facilitate the discovery of novel disease-causing genes (49, 50). Similarly, establishing a reference genome derived from healthy, fertile, phenotyped women is crucial to support personalized approaches in ART.

Clinically established genomic tools are already applied at multiple stages of ART. Preimplantation genetic testing (PGT) for aneuploidy (PGT-A), monogenic diseases (PGT-M), and structural rearrangements (PGT-SR) is widely used to assess embryo genetic status prior to transfer (26, 51), reducing the risk of miscarriage and genetic disease and supporting personalized ART protocols. Clinical exome sequencing in individuals undergoing ART and sperm donor has uncovered a substantial burden of pathogenic or likely pathogenic variants, emphasizing the growing relevance of reproductive counseling and PGT-M (52, 53). Large cohort studies further demonstrate that PGT-M, particularly when combined with comprehensive chromosome screening, is associated with improved clinical outcomes, including higher live birth rates (54). Compared to morphological evaluation alone, preimplantation genetic testing for aneuploidy (PGT-A), including non-invasive assessment using cell-free DNA from spent embryo culture medium enhances live birth and ongoing pregnancy rates per embryo transfer (55). Preconception carrier screening (56–59), is another well-established tool, although access remains limited by cost, availability to genetic counseling, infrastructures, and underrepresentation of diverse populations in genomic databases (58).

Emerging or investigational approaches include non-invasive PGT using cell-free DNA from embryo culture media or blastocoel fluid to assess chromosomal status without biopsy (60). Broader genome-wide and epigenomic analyses, including DNA methylation and transcriptomic profiling, are under investigation to improve infertility diagnosis, optimize embryo selection, and elucidate ART-associated molecular changes (61, 62). The use of polygenic risk scores is emerging to explore complex traits such as multifactorial infertility or ovarian reserve depletion (63). Although large-scale genome-wide association studies are identifying reproductive trait loci, translation into actionable clinical metrics in ART is not yet established (63). While promising, these investigational tools require further validation before routine clinical implementation.

Addressing the complexity and heterogeneity of fertility disorders, requires both genetic and non-genetic approaches. The integration of diverse data types with emerging technologies, such as artificial intelligence and machine learning, offers potential to optimize ovarian stimulation protocols, predict the likelihood of success following elective oocyte cryopreservation, and refine embryo selection to improve clinical outcomes (64). Collectively, these advances underscore the critical role of genetic testing in infertility evaluations and support its incorporation into personalized reproductive management strategies.

The data generated during routine clinical care and captured in sources such as electronic health records (EHRs), patient-reported outcomes, administrative claims, perinatal registries, and ART registries, known as Real-World Data (RWD), can be translated into Real-World Evidence (RWE) when analyzed using fit-for-purpose epidemiological designs and transparent reporting standards (65). RWE is increasingly recognized as a complementary evidence stream that can extend beyond the constraints of randomized clinical trials (RCTs), particularly when trials are infeasible, underpowered for rare outcomes, or poorly representative of the complex, comorbid profiles typical of AMA (66, 67).

In reproductive medicine, and especially in the AMA + ART population, the core clinical questions frequently concern: (i) Comparative effectiveness between competing strategies (e.g., fresh vs. frozen embryo transfer, elective single embryo transfer, donor oocytes, adjuvant treatments); (ii) Safety across multiple maternal and neonatal endpoints (including low-frequency but high-impact outcomes), and (iii) Longitudinal outcomes for both mother and offspring. These questions are difficult to address comprehensively in classic RCT frameworks due to sample size requirements, ethical constraints, rapid evolution of ART practice, and the need for long follow-up (67). Actually, RWD can be particularly valuable in reproductive health, where biological, social, and demographic factors intersect. For instance, AMA and ART risks may be amplified by comorbidities such as hypertension, diabetes, or obesity, conditions unevenly represented in clinical trials but readily captured at scale in healthcare databases.

RWD has been instrumental in evaluating outcomes that are either uncommon or require very large denominators. For instance, large registry-linked analyses have compared perinatal outcomes after frozen vs. fresh embryo transfer, including sibling-comparison designs that partially control for stable maternal factors and strengthen causal interpretation (68). These real-world comparative analyses are directly relevant to AMA, where clinical decisions are frequently individualized and where balancing maternal hypertensive risk, placental complications, prematurity, and neonatal growth patterns may differ across treatment strategies.

An example of generation of evidence from RWD, that has already yielded novel insights in women's health, is the Andalusian Collaborative Environment for Advancing Predictive Care (69). This environment allows researchers carrying out ethical and secure research with clinical data from the Andalusian Health Population Database (BPS) within a trusted research environment, to generate evidence under strict data protection and GDPR compliance, ensuring scalability and reproducibility. Currently, BPS integrates detailed longitudinal clinical data from more than 15 million residents of Andalusia, which provides an idea of the sample sizes available for studies. In this environment. large-scale population-based studies have shown how menopause significantly modifies the trajectory of chronic diseases, with postmenopausal women exhibiting disproportionate increases in hypercholesterolemia, osteoporosis, hypertension, and anxiety disorders compared to men of similar age, using a cohort of more than 550,000 patients (70). These findings highlight how hormonal transitions can be studied in real-world populations, providing actionable knowledge for prevention and management strategies. Similar approaches can be applied to study the combined effects of AMA and ART, generating robust, population-level evidence that complements genomic studies and clinical trials.

Similarly, predictive modeling using RWD has advanced early detection strategies. Our group recently demonstrated that machine learning applied to population health databases can identify women at elevated risk of ovarian cancer well before diagnosis, paving the way for more personalized surveillance and preventive interventions (71). Such approaches are directly relevant to AMA and ART, where early recognition of maternal risks could inform counseling, treatment choices, and follow-up intensity. In the ART setting, RWD can be leveraged to compare outcomes of autologous vs. donor oocyte cycles, fresh vs. frozen embryo transfers, or different stimulation protocols in real-world practice. Importantly, large-scale databases can also track maternal and offspring outcomes beyond the perinatal period, addressing current gaps regarding long-term cardiovascular or metabolic risks after ART pregnancies.

There is an urgent need for dedicated guidelines for pregnancies in AMA women undergoing ART. RWD can play a central role in closing this gap by:

Providing comparative effectiveness evidence between ART modalities in older women