Associations between previous cesarean section and maternal and neonatal complications: the modification of long inter-pregnancy interval

Introduction

Cesarean section (CS), one of the most common surgeries worldwide, serves as a life-saving intervention when medically indicated [1]. However, CS is also associated with elevated risks of maternal and neonatal complications in the current and even subsequent pregnancies. China has the world’s highest CS rate, with urban women reaching a peak of 64.1 % in 2008 [2]. Notably, China has the world’s highest rate of non-indicated CS, which is a major contributor to the overall high CS rate [3], 4]. Evidences demonstrate that prior CS elevates risks of maternal complications including hysterectomy, abnormal placentation, uterine rupture, premature rupture of membranes (PROM), and postpartum hemorrhage [1], 5], 6]. Moreover, it correlates with adverse neonatal outcomes such as low birth weight, preterm birth, and neonatal intensive care unit (NICU) admission [7].

Before the implementation of the universal two-child policy in 2015, many Chinese women didn’t consider giving birth to a second child. As a result, they often preferred CS over vaginal delivery, without concerning the potential complications of CS in future pregnancies. Nevertheless, following the policy change, a substantial number of multiparous women with a history of CS decided to have a second child. The sudden transition from a one-child to two-child policy has led to at least an additional 90 million multiparous women giving birth again [8]. Notably, many of these women have a long inter-pregnancy interval (IPI), a known risk factor for adverse maternal and neonatal outcomes.

Furthermore, studies suggest prolonged IPIs may mitigate certain CS-associated maternal risks, including uterine rupture, placenta accreta spectrum, maternal perinatal mortality, and primary postpartum hemorrhage [9], [10], [11], [12]. Conversely, a recent Chinese study reports that a long IPI (≥60 months) may reduce maternal risks but elevate risks of neonatal adverse outcomes [13]. Given China’s unique public health landscape – marked by high CS rates and high second-child rate, it is important to clarify the association between previous CS and maternal-neonatal adverse outcomes. Considering both CS and IPIs as potential modifiable factors, determining their population attributable risk percent (PARP) and interactive effects on adverse maternal and neonatal outcomes is critical. Such insights are critical for informing clinical guidelines and healthcare policies to optimize perinatal care in the post-two-child policy era. The universal two-child policy in China provides us with this unique opportunity to do this analysis in a population with high rates of CS and a long IPI.

Materials and methods

Study subjects

The included subjects were multiparous women who delivered singleton babies≥28 weeks between December 1, 2015, and December 1, 2020. The total number of registered pregnant mothers during this period was 14,347. First, primiparous women (n=5,586) were excluded from the cohort. Multiparous women were also excluded if they had twin or multiple births (n=394) or used artificial reproductive technology (n=787). Additionally, multiparous women with important missing information were excluded (n=213), including missing birth weight information (n=84), previous delivery modes (n=52), and IPIs (n=77). Finally, a total of 7,367 multiparous women were included in the analysis, consisting of 4,711 women (63.95 %) with previous CS delivery and 2,656 women with previous vaginal delivery (Figure 1).

Figure 1:

The flowchart of the whole process.

Results

Previous CS and the risks of adverse maternal and neonatal perinatal outcomes

Compared with the previous vaginal delivery group, previous CS was associated with higher risks of uterine-related complications (aOR=1.57, 95 % CI 1.25–1.98), but with lower risks of preterm birth (aOR=0.73, 95 % CI 0.63–0.85), and severe neonatal adverse outcomes (aOR=0.59, 95 % CI 0.46–0.78) (Figure 2). No significant associations were found between previous CS and placental-related complications, PROM, and SGA.

Figure 2:

Logistic regression analysis for adverse maternal and neonatal outcomes associated with previous cesarean section. Potential covariates included maternal age of the current delivery, gravidity (2, 3, >3), parity (2, >2), body mass index, abortion history (yes/no) inter-pregnancy interval, and plus: ^aMode of the current delivery (vaginal/cesarean), gestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no). ^bGestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no). ^cGestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no), mode of the current delivery (vaginal/cesarean), gestational age, birth weight.

Interactions between previous CS and IPIs

In our study, the number of multiparous women with a long IPI was 3,702 (50.25 %). Stratified analyses by IPIs (<60, ≥60 months) showed consistent trends between previous CS and risks of uterine-related complications, preterm birth, and severe neonatal adverse outcomes across both subgroups (Figure 3, Supplementary Table S1). Previous CS had similar trends in higher risks of uterine-related complications (aOR=1.62, 95 % CI 1.03–2.55, aOR=1.94, 95 % CI 1.30–2.90, respectively) and in lower risks of preterm birth (aOR=0.78, 95 % CI 0.62–0.97, aOR=0.65, 95 % CI 0.54–0.79, respectively) and severe neonatal adverse outcomes (aOR=0.60, 95 % CI 0.42–0.94, aOR=0.54, 95 % CI 0.38–0.73, respectively) in subgroups with IPIs<60 and ≥ 60 months.

Figure 3:

Stratified analysis to analyze for the associations between previous cesarean section and adverse maternal and neonatal outcomes in subgroups with different inter-pregnancy intervals (IPIs). Potential covariates included maternal age of the current delivery, gravidity (2, 3, >3), parity (2, >2), body mass index, abortion history (yes/no) and plus: ^aMode of the current delivery (vaginal/cesarean), gestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no). ^bGestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no). ^cGestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no), mode of the current delivery (vaginal/cesarean), gestational age, birth weight.

A synergistic biological interaction was observed between previous CS and a long IPI for placental-related complications (RERI=0.32, 95 % CI 0.05–0.58; AP=0.39, 95% CI 0.03–0.76). Conversely, an antagonistic interaction was found for preterm birth (RERI=−0.35, 95% CI -0.68 to −0.01; AP =-0.09, 95 % CI -0.68–0.03). No significant interactions were detected for other outcomes (Table 2).

Table 2:

Interactions between the long inter-pregnancy interval (≥60 months) and previous cesarean section delivery.

Outcomes	RARI (95 % CI)	AP (95 % CI)
Uterine-related complicationsa	0.01 (-0.49, 0.50)	0.32 (-0.31, 2.47)
Placental-related complicationsb	0.32 (0.05, 0.58)	0.39 (0.03, 0.76)
Premature rupture of membranesb	−0.03 (−0.31, 0.25)	−1.74 (−0.32, 0.25)
Small for gestational ageb	−0.56 (−1.24, 0.11)	−0.46 (−0.92, 0.03)
Preterm birthb	−0.35 (−0.68, −0.01)	−0.09 (-0.68, −0.03)
Severe neonatal adverse outcomesc	−0.43 (−0.97, 0.09)	−0.42 (−0.91, 0.06)

1. RARI, relative excess risk caused by interaction; AP, attributable proportion. The reference group is the group with both IPI<60 months and previous vaginal delivery. Potential covariates included maternal age of the current delivery, gravidity (2, 3, >3), parity (2, >2), body mass index, abortion history (yes/no) and plus: ^aMode of the current delivery (vaginal/cesarean), gestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no). ^bGestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no). ^cGestational diabetes mellitus (yes/no), hypertensive disorders in pregnancy (yes/no), mode of the current delivery (vaginal/cesarean), gestational age, birth weight.

The PARPs for previous CS were 27.57 % for uterine-related complications and 1.80 % for placental-related complications (Table S2).

Discussion

In the present study, the findings suggest that CS is associated with a higher risk of adverse maternal outcomes in the Chinese population. However, a long IPI does not attenuate the harm of previous CS to mothers. Although previous CS appears to be associated with a lower risk of adverse neonatal outcomes, a negative impact of a long IPI is observed.

Previous CS was found to confer a higher risk of uterine complications [1], 19], 20]. Consistent with previous studies, our findings also showed that previous CS was associated with increased risks of uterine-related complications. Indeed, uterine rupture rates have been reported to be nearly 40-50-fold higher among mothers with CS history than ones without CS history [21]. This indicated that previous CS was an independent risk factor for uterine-related complications. Notably, our study estimated that 27.57 % of uterine-related complications could be prevented by avoiding previous CS in the general population.

Nonetheless, concerning IPIs after CS, the conclusions of studies are not consistent and conclusive. Sarah Cunningham et al. found that normal IPI (24–59 months) was associated with the lowest frequency of uterine rupture, while a long IPI (≥60 months) was not significantly linked to uterine rupture risk in mothers with previous CS [22]. Another study even found that a long IPI (≥60 months) was associated with a decreased risk of uterine rupture (aOR=0.83, 95 % CI 0.69–1.00) [23]. In contrast, the present study found that the prolonged IPI after previous CS did not reduce the risks of uterine-related complications. Unlike prior researches [23], we observed no antagonistic biological interaction between a long IPI and previous CS, suggesting that the adverse effects of previous CS on uterine-related complications might persist irrespective of IPI duration. Previous literature reports showed that previous CS was associated with the risk of placental-related complications, including placenta previa and placental abruption [24], 25]. Long IPI might indirectly increase placental abruption risk by elevating the likelihood of eclampsia and preeclampsia [26]. Notably, although our study didn’t detect a significant association between previous CS and placental-related complications (possibly due to sample size limitations), a positive interaction was observed between previous CS and a long IPI for these outcomes. This suggested that the combination of a long IPI and previous CS might have a synergistic effect on increasing placental complication risks.

Previous studies have indicated that certain obstetrical complications or serious maternal conditions determine the need for CS [18]. Jane Sandall et al. reported that exposure to a previous medically indicated CS delivery, including preeclampsia or maternal hyperglycemia, increased the risk of severe acute maternal morbidity in the current pregnancy, such as hemorrhage, hysterectomy, or uterine rupture [1], 27], 28]. Similarly, our findings also showed an increased risk of uterine-related complications in previous CS population with a high rate in non-medical indicated CS. These findings indicated that uterine-related complications are more associated with previous CS itself rather than with prior medical conditions. Moreover, compared with the reference group, the previous CS groups of different maternal ages and IPIs all showed higher trends of uterine-related complications. This highlighted the importance of reducing non-essential CS rates to prevent maternal complications in subsequent pregnancies.

Most prior studies reported that previous CS was associated with higher or neutral neonatal complication risks [7], 29]. In contrast, our findings indicated that previous CS appeared to be associated with lower risks of preterm birth, and severe neonatal adverse outcomes. According to previous studies, women’s socioeconomic status and the availability of healthcare services were associated with increased CS use [30]. In China, women who preferred CS perceived it to be safer for both the baby and themselves and free from pain and anxiety, despite the cost of CS being twice as high as vaginal delivery [31]. Additionally, CS delivery entailed longer maternal hospital stays and increased drug usage compared to vaginal delivery [32], 33]. Therefore, women with better financial conditions were more willing or able to afford the expense of non-medically indicated CS. After the implementation of the universal two-child policy in 2015, women with a history of non-indicated CS who wanted to give birth again usually had better social welfare benefits, health insurance, and maternity care. They had better nutrition, paid more attention to, and received more obstetric care during pregnancy to monitor fetal growth. Thus, pregnancies with previous CS seemingly appear to be particularly associated with lower risks of adverse neonatal outcomes. Similar findings have been reported in another study in the Chinese population [34]. However, it should be noted that these benefits may not be directly attributed to previous CS but rather to the effects of improved antenatal care. Therefore, better antenatal care can partly prevent neonatal complications associated with previous CS. However, our study’s reliance on data from one southwest China city might introduce bias to the results. Socioeconomic levels, healthcare access, maternal cares across China were different. So, future multi-center studies with geographically diverse data were needed to verify these associations in wider populations.

Along IPI was associated with elevated odds of preterm birth and SGA in a previous study [35]. However, the present study demonstrated that previous CS was associated with decreased risks of preterm birth and severe neonatal adverse outcomes and antagonistic effects were observed between previous CS and a long IPI for preterm birth. It should be awareness that although previous CS was associated with lower risks of adverse neonatal outcomes, prolonged IPIs might negatively modify this association. We considered that in essence, the long IPI might offset the benefits of improved antenatal care behind CS.

Several limitations existed in the present study. First, the hospital-based data were collected from a single city in southwest China, which might not represent the national general population. Second, we lacked accurate data on maternal socioeconomic status and prenatal healthcare to support our assumptions regarding the reduced risks of neonatal complications. Finally, the observational study design only allowed for association analysis rather than causal inference.

Conclusions

In a population with a high rate of previous CS, previous CS was associated with increased risks of adverse maternal outcomes but decreased risks of certain adverse neonatal outcomes. Prolonged IPIs might not diminish the adverse effects of previous CS in mothers, but might adversely harm the neonates.