Huang David, Song Yue, Qin Jennifer, Wong Rebecca, Durgana Chantal, Adeleye Amanda, Rinaudo Paolo, Lustig Robert H, Zablotska Lydia B, Cedars Marcelle I
Department of Obstetrics, Gynecology & Reproductive Sciences, University of California San Francisco (UCSF), San Francisco, CA.
Department of Obstetrics, Gynecology & Reproductive Sciences, University of California San Francisco (UCSF), San Francisco, CA.
Am J Obstet Gynecol. 2025 Aug 5. doi: 10.1016/j.ajog.2025.07.043.
Obstetric complications linked to assisted reproductive technology could be confounded by underlying parental factors and were often derived from databases not designed to study these complications. With evolving assisted reproductive technology practices, it is unclear to what extent contemporary assisted reproductive technology independently contributes to obstetric risks.
We aimed to prospectively evaluate obstetric risks associated with contemporary fertility practices, adjusting for parental and assisted reproductive technology parameters.
Prospective enrollment of patients with infertility who subsequently conceived unassisted, with non-in vitro fertilization fertility treatment, or assisted reproductive technology at a single US academic fertility center (September 2017-December 2021). Multiple gestations and surrogacy pregnancies were excluded from analysis. Parental characteristics, treatment parameters, and obstetric complications were assessed from medical records. The main outcome was physician-adjudicated obstetric complications (hypertensive disorders, abnormal placentation, fetal growth restriction, spontaneous preterm birth, and gestational diabetes). Maternal serum placental analyte level (pregnancy-associated plasma protein A) was also assessed by mode of conception. Cumulative incidences of obstetric complications by mode of conception were compared using chi-squared test. Univariable and multivariable log-binomial regression analyses were used to generate crude and adjusted risk ratios and their 95% confidence intervals to assess for associations between putative risk factors and abnormal placentation. Distributions of pregnancy-associated plasma protein A levels were compared using Kruskal-Wallis test, with correction for multiple pairwise comparisons with Dunn's test. All tests were 2-sided and conducted at the 0.05 level of significance.
Of 2332 pregnancies approached for recruitment, 782 enrolled in the cohort, resulting in 656 singleton live births: 92 unassisted, 116 from non-in vitro fertilization fertility treatment, and 448 from assisted reproductive technology. The median (interquartile range) ages of the female patients were 37 (35-39), 36 (34-38), and 38 (35-41) years, respectively (P<.01). Abnormal placentation was the only assessed complication with a higher cumulative incidence in assisted reproductive technology vs nonassisted reproductive technology pregnancies (unassisted and non-in vitro fertilization fertility treatment combined). Thirty-nine patients had an abnormal placentation event in the assisted reproductive technology group, accounting for 8.7% (95% confidence interval, 6.3% to 11.7%) of assisted reproductive technology pregnancies, compared to 3, 1.4% (95% confidence interval, 0.3% to 4.1%), in non-assisted reproductive technology pregnancies (P<.01). Compared to non-assisted reproductive technology pregnancies, adjusting for parental factors, assisted reproductive technology was independently associated with a higher relative risk of abnormal placentation (adjusted relative risk, 6.19 [95% confidence interval, 1.84-20.82], P<.01). Within assisted reproductive technology pregnancies, fresh embryo transfer was the treatment parameter associated with abnormal placentation (adjusted relative risk, 2.07 [95% confidence interval, 1.10-3.92], P=.025). Pregnancy-associated plasma protein A level was particularly lower in fresh embryo transfer-conceived pregnancies (median multiple of the median value of 0.65 [interquartile range, 0.42-1.02 multiple of the median]), compared to non-assisted reproductive technology pregnancies (1.09 multiple of the median [interquartile range, 0.70-1.56 multiple of the median]) (P<.001).
Adjusting for underlying infertility, assisted reproductive technology remained independently associated with abnormal placentation. This risk should be considered during patient counseling and clinical management, particularly when assisted reproductive technology is pursued for elective reasons. The impact of fresh embryo transfer on placentation warrants further investigation.
与辅助生殖技术相关的产科并发症可能会受到潜在的父母因素的混淆，且这些并发症往往来自并非专门设计用于研究这些并发症的数据库。随着辅助生殖技术实践的不断发展，目前尚不清楚当代辅助生殖技术在多大程度上独立导致产科风险。
我们旨在前瞻性评估与当代生育实践相关的产科风险，并对父母因素和辅助生殖技术参数进行调整。
在美国一家学术性生育中心（2017年9月至2021年12月），对随后自然受孕、接受非体外受精生育治疗或辅助生殖技术的不孕患者进行前瞻性招募。多胎妊娠和代孕妊娠被排除在分析之外。从医疗记录中评估父母特征、治疗参数和产科并发症。主要结局是医生判定的产科并发症（高血压疾病、胎盘异常、胎儿生长受限、自发性早产和妊娠期糖尿病）。还通过受孕方式评估了母体血清胎盘分析物水平（妊娠相关血浆蛋白A）。使用卡方检验比较不同受孕方式下产科并发症的累积发生率。采用单变量和多变量对数二项回归分析来生成粗风险比和调整后的风险比及其95%置信区间，以评估假定风险因素与胎盘异常之间的关联。使用Kruskal-Wallis检验比较妊娠相关血浆蛋白A水平的分布，并通过Dunn检验对多个两两比较进行校正。所有检验均为双侧检验，显著性水平为0.05。
在2332例拟招募的妊娠中，782例纳入队列，最终有656例单胎活产：92例自然受孕，116例来自非体外受精生育治疗，448例来自辅助生殖技术。女性患者的年龄中位数（四分位间距）分别为37（35 - 39）岁、36（34 - 38）岁和38（3
