Since the birth of the first test-tube baby in 1978, assisted reproductive technology (ART) has become an effective treatment for infertility (1). With the progress of technology and provision of services, an increasing number of infants are born following ART therapy (1, 2). In developed countries, ART pregnancies account for 1.5–5.9% of all births (3–7), while in China, ART pregnancies account for 1.7% and are increasing year by year (2). ART is sometimes defined differently and is usually defined as the application of laboratory or clinical techniques to gametes and/or embryos for reproductive purposes, however it has been broadened to include not only in vitro procedures but also ovarian stimulation with gonadotropins or ovariotropic drugs (8, 9). As these and other reproductive technologies expand, leading to a substantial number of successful pregnancies and births, it is critical for prospective parents to understand the maternal and neonatal outcomes associated with ART.

Several studies have shown that ART pregnancies have an increased risk of multiple pregnancy and adverse pregnancy outcomes, including gestational diabetes, gestational hypertension, placenta previa, preterm birth, operative delivery, low birth weight, birth defects and perinatal mortality (10–16). However, other studies have concluded ART pregnancies do not have increased risks of adverse perinatal outcomes (7, 17–19). The incidences of small for gestational age, preterm birth and cesarean section are similar between ART and naturally pregnancies (13, 20). Nevertheless, pregnancy outcomes in ART pregnancies appear to be generally poorer due to the increased risk of multiple pregnancies. Multiple pregnancy is a post-processing confounding factor, which appears after ART treatment and may confound causal effects. Many previous studies did not adjust for maternal age, BMI and other confounding factors (12, 21–23). It is not clear whether the increased risk of adverse pregnancy outcomes is due to ART itself, multiple births, or potential infertility. At present, opinions are too far apart to reach a consensus.

The present retrospective cohort study was conducted to compare maternal and neonatal outcomes between ART and naturally pregnancies, and in addition to explore the association of ART with adverse pregnancy outcomes by stratifying on birth plurality and maternal age.

A total of 211,606 women were included in this retrospective analysis, of whom 13,604 women conceived by ART as group of ART pregnancy, and 198,002 women conceived naturally without ART as group of naturally pregnancy. Over the past 10 years, the proportion of ART pregnancy has increased each year, reaching a peak in 2020 (9.0%) (Figure 1A). Of the ART pregnancy, the proportion of pregnant women over 35 years old and multiparae increased mildly, while the proportion of women with intrapartum BMI greater than 30 kg/m² decreased slightly. And with the implementation of the universal two-child policy in 2015, the proportion of women with abnormal pregnancy history in ART pregnancy decreased sharply, and then increased rapidly (Figure 1B).

Maternal characteristics between ART and naturally pregnancy is summarized in Table 1. The mean maternal age and intrapartum BMI of women conceived by ART were significantly higher than those of women conceived naturally (standardized difference = 0.547 and 0.309, respectively). And women conceived by ART were more likely to be nulliparae (88.5% vs. 75.4%, standardized difference = 0.345), more likely to have a long or irregular menstrual cycle (15.2% vs. 7.9%, standardized difference = 0.230) and an abnormal pregnancy history (including early abortion, intermediate and late abortion, abnormal development, or ectopic pregnancy, 29.1% vs. 9.9%, standardized difference = 0.500), and more likely to have uterine fibroids (8.3% vs. 5.1%, standardized difference = 0.128). There were no significant standardized differences in maternal height and birthplace between the two groups.

The incidences of pregnancy and perinatal complications in ART and naturally pregnancy was exhibited in Tables 2, 3. Statistically significant increases were noted in gestational diabetes (28.0%), preeclampsia (5.4%), thyroid-related diseases (13.7%), preterm birth (20.4%), placenta previa (8.5%), postpartum hemorrhage (19.2%) and cesarean section (75.5%) in ART pregnancy, compared to naturally pregnancy (standardized difference > 0.1). The occurring rates of anemia (25.9%), liver-related diseases (5.3%), polyhydramnios (3.4%), oligohydramnios (7.1%) and torsion of cord (3.3%) were also elevated in ART pregnancy, but with no significant difference (standardized difference < 0.1). In contrast, there was a decline in the incidences of premature rupture of membrane (21.7%) and amniotic fluid pollution (I°: 4.2%, II°: 3.5%, III°: 3.6%) in ART pregnancy (standardized difference > 0.1). We also analyzed neonatal outcomes between naturally pregnancy and ART pregnancy (Table 4). The mean birth weight of ART pregnancy was significantly lower than that of naturally pregnancy (standardized difference = 0.343). Moreover, significant rises of incidence were observed in twins or multiples (20.7%) and stillbirth or abnormal development (3.3%) in ART pregnancy (standardized difference > 0.1). No significant difference was noted in macrosomia and fetal distress between the two groups.

Multivariable logistic regression analysis showed that the association between ART and pregnancy outcomes were significant (Table 5). Nearly all the pregnancy complication listed, including gestational diabetes [aOR (95%CI) = 1.39 (1.33–1.46)], preeclampsia [aOR (95%CI) = 1.26 (1.14–1.41)], moderate or severe anemia [aOR (95%CI) = 1.20 (1.08–1.32)], liver-related diseases [aOR (95%CI) = 1.14 (1.03–1.26)], and thyroid-related diseases [aOR (95%CI) = 1.29 (1.21–1.37)], were more likely to occur among women conceived by ART. In terms of perinatal complications, the risk of preterm birth [aOR (95%CI) = 1.61 (1.49–1.74)], placenta previa [aOR (95%CI) = 1.48 (1.37–1.60)], postpartum hemorrhage [aOR (95%CI) = 1.14 (1.08–1.21)], and cesarean section [aOR (95%CI) = 2.84 (2.70–2.99)] were significantly increased, while the risk of premature rupture of membrane [aOR (95%CI) = 0.66 (0.62–0.71)], amniotic fluid pollution [I°: aOR (95%CI) = 0.83 (0.74–0.92); II°: aOR (95%CI) = 0.66 (0.59–0.74); III°: aOR (95%CI) = 0.57 (0.51–0.64)], cord entanglement [aOR (95%CI) = 0.89 (0.85–0.93)], and intrapartum fever [aOR (95%CI) = 0.69 (0.64–0.74)] were significantly decreased in ART pregnancy as compared with naturally pregnancy. For neonatal outcomes, women conceived by ART were more likely to have twins or multiples [aOR (95%CI) = 24.20 (22.43–26.11)], and the risk of stillbirth or abnormal development [aOR (95%CI) = 2.76 (2.39–3.17)] was also significantly increased. Moreover, the risk of macrosomia [aOR (95%CI) = 0.88 (0.80–0.95)] and fetal distress [aOR (95%CI) = 0.66 (0.60–0.73)] were significantly decreased in ART pregnancy (Table 5).

The association of ART with maternal and neonatal outcomes were also evaluated by stratifying on birth plurality and maternal age (Tables 6, 7). When restriction to singletons, the risks of adverse pregnancy outcomes as listed above were reduced. And the effects of ART on the risk of premature rupture of membrane, cord entanglement, intrapartum fever, cesarean section, and stillbirth or abnormal development (ART pregnancy vs. naturally pregnancy) were more pronounced among singleton pregnancies compared with that among pregnancies of twins or multiples, while the effect of ART on the risk of polyhydramnios was more prominent among pregnancies of twins or multiples (heterogeneity test: P < 0.05). When stratified by maternal age, we found the effects of ART on the risk of preterm birth, placenta previa, postpartum hemorrhage and cesarean section (ART pregnancy vs. naturally pregnancy) were more pronounced among women under 35 years compared with that among women over 30 years, while the effect of ART on the risk of polyhydramnios was more prominent among women over 35 years (heterogeneity test: P < 0.05).

This retrospective, hospital-based cohort study including 13,604 ART pregnancies and 198,002 naturally pregnancies was conducted in Nanjing, China from 2011 to 2020. The study showed the widespread application of ART in China, with the proportion of ART pregnancies increasing year by year in the past decade, and confirmed the increased risks of several adverse pregnancy outcomes in ART pregnancies. We found a 24.2-fold increase in the incidence of multiple births in ART pregnancies compared to naturally pregnancies, then stratified the analysis by birth plurality, suggesting multiple births are indeed an important factor leading to adverse pregnancy outcomes.

In the present study, the increased risks were found in ART pregnancy compared with naturally pregnancy: gestational diabetes (1.39-fold), preeclampsia (1.26-fold), moderate or severe anemia (1.20-fold), liver-related diseases (1.14-fold), thyroid-related diseases (1.29-fold), preterm birth (1.61-fold), placenta previa (1.48-fold), postpartum hemorrhage (1.14-fold), cesarean section (2.84-fold), and stillbirth or abnormal development (2.76-fold), which were largely consistent with the findings of previous studies (10, 28–32). Although these risks were reduced when restriction to singletons, significant differences remained. Some studies have suggested that infertility is one of the risk factors for adverse pregnancy outcomes (33). However, infertility factors cannot fully explain the associations. For infertile women, women conceived with ART had an increased risks of adverse pregnancy outcomes compared with women conceived with non-ART (34). Therefore, some researchers believe that the increased risks of adverse outcomes after ART conception are mainly related to ART manipulation factors (6), which is due to the addition of many non-physiological operations by ART. For example, the type of ovulation induction drugs used in the early stage, the composition of the culture medium, the storage time in the culture medium, the freezing and dissolution process of the embryo, polyspermic fertilization, and the hormone level at the time of implantation, all play an important role in the occurrence of adverse pregnancy outcomes (35). Other studies have pointed out that different methods of ART may lead to different types of adverse pregnancy outcomes (36). At the same time, the longer and more times of ART treatment, the greater the harm to women and their offspring (36). In addition, ART pregnancies may be more closely monitored than naturally pregnancies, which partly explains the higher incidence of adverse pregnancy outcomes in ART pregnancies (37). However, current studies and evidence cannot fully elucidate the mechanism by which ART increases the risk of adverse pregnancy outcomes, and the specific mechanism needs further research.

The main advantage of this study was the large sample size of pregnancy, which allowed us to conduct further subgroup analysis with enough power. And the data obtained from HIS database by using NLP technique is of high quality. However, there are some limitations in this study. First, the population we studied was limited to one city in eastern China (Nanjing). Therefore, we should be cautious in generalizing our findings to other regions. Second, we did not collect information on the ART form. The more intricate and invasive the ART form used, the more likely it was to cause adverse pregnancy outcomes. And the records of pre-pregnancy BMI, baseline endocrine level, causes of infertility, ovarian stimulation protocols, and quality of transferred embryos lacked in our database, were not included in this study. Third, the retrospective design of this study could not assess a causal relationship between ART and adverse pregnancy outcomes. These limitations should be considered in future studies.

Women conceived by ART were at increased risks of several adverse pregnancy outcomes compared with women conceived naturally. Multiple pregnancies due to multiple embryos transferred could partly explain the increased risks. The transfer of single embryo of high quality should be promoted. However, ART singleton pregnancy still showed higher risks of several adverse pregnancy outcomes compared with naturally pregnancy, suggesting ART itself is also significantly correlated with pathological pregnancy. Therefore, policies related to ART indications should be strictly formulated to reverse the high rate of ART pregnancy. Given our findings, prenatal and intrapartum monitoring should be strengthened, and neonatal outcomes should be closely observed for ART pregnancy. And more research should be conducted to further clarify whether the increased risk of adverse pregnancy outcomes is due to ART itself, multiple births, or potential infertility.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by the institutional review board of Women’s Hospital of Nanjing Medical University (2020KY-011). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

JW initiated, conceived, and supervised the study. WT and LH did data collection and performed the data analysis. All authors approved the final format of the submitted manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.