Main findings
In this retrospective observational study, we investigated the relationship between IPI and adverse maternal and neonatal outcomes in subsequent twin pregnancies. Our findings revealed that an IPI of 36–59 months was a protective factor against VLBW or ELBW for at least one twin. No significant associations were found between IPI and other birth outcomes (PTB < 34 weeks, LBW/LBW, LBW/NBW, PPH and MHM). Furthermore, several modifiable risk factors, such as maternal BMI at delivery, gestational weight gain, ART usage, general anesthesia and anemia, along with unmodifiable risk factors, including chorionicity, PPROM, gestational hypertension, preeclampsia, cesarean delivery history, placenta previa, low-lying placenta, and placental adherence, have been identified as potential contributors to adverse maternal and neonatal outcomes.
Interpretations
Our study reported that all women with twin pregnancies had IPIs ≥ 12 months, with only 2.4% having IPIs < 18 months, and 64.8% having IPIs ≥ 60 months. However, two population-based cohort studies in Australia and America illustrated that approximately 4.6–5.6% of women with twin pregnancies experienced extreme short IPIs (< 6 months), 8.5–35% had short IPIs (< 18 months), and 13-23.4% had long IPIs (≥ 60 months) [14, 17]. Thus, it appears that longer IPIs are more prevalent before twin births in Chinese population. The significant changed China’s fertility policy, including the partial two-child policy introduced in November 2013, the universal two-child policy implemented in October 2015 [2], and the universal three-child policy announced in May 2021 [1], may lead to the different IPI distributions.
Numerous studies have been conducted to explore the relationship between IPI and adverse neonatal outcomes [9, 13, 14, 17]. However, these previous studies have primarily focused on neonates with LBW (< 2500 g) [9, 13, 14, 17] and our study have innovatively categorized twin birth weights into three groups: LBW/LBW, LBW/NBW, and VLBW or ELBW for at least one twin. We have first identified an IPI of 36–59 months as a protective factor against VLBW or ELBW in at least one twin. Considering the risks associated with short IPIs, including maternal nutritional depletion [13], cervical insufficiency, infection transmission [10], as well as the risks of long IPIs, such as physiological regression [13], advanced maternal age, and increased use of ART, a moderate IPI (36–59 months) may be optimal. Although prior investigations have generally demonstrated a correlation between IPI and LBW in both singleton [9, 21, 22] and twin pregnancies [14], we found limited evidence supporting such association specifically within the categories of LBW/LBW and LBW/NBW. The discrepancies may be attributed to the inconsistent reference categories, varying LBW definitions, and distinct IPI distributions and population characteristics. Additionally, the association between IPI and PTB has also been widely documented in both singleton [9, 21, 22] and twin pregnancies [14, 17]. However, our group and Weiss A et al. [15] have found no association between IPI and PTB. The varying definitions of PTB, ranging from less than 34 weeks to 37 weeks, may partially explain the inconsistent conclusions.
Regarding the secondary outcomes, several studies have investigated the relationship between IPI and hemorrhage-related events in singleton pregnancies [11, 23, 24]. Garg B et al. found that women with short IPIs had a higher incidence of PPH, which was further associated with an increased risk of blood transfusion [11, 24]. Chen P et al. reported a significant association between long IPIs and PPH [23]. Liu C et al. observed that longer IPI was linked to an increased risk of severe maternal morbidity, including blood transfusion [3, 12, 13]. There are multiple mechanisms that could contribute to these observed associations [12]. A short IPI may be associated with uterine incomplete healing and placental abnormalities, thus leading to greater blood loss [12]. Additionally, a prolonged IPI may impair maternal physiological adaptation to childbirth and increase the risk of preeclampsia, which results in a higher risk of excessive blood loss [13]. Human reproduction selective mechanisms may also play a role in these associations [3]. Considering the extended uterine healing time, higher incidence of pregnancy complications and placental abnormalities, as well as increased usage of ART in subsequent twin pregnancies, we hypothesized that both short IPI (< 24 months) and extremely long IPI (≥ 120 months) might be related to PPH and MHM. However, our findings revealed that the risks of PPH and MHM were comparable across varying IPI groups. The unique distribution of IPIs and similar clinical characteristics may be the reasons. Given the limited existing literature on the relationship between IPI and maternal blood loss-related outcomes, particularly within the context of China’s fertility policy, further research is warranted to elucidate the association between IPI and adverse maternal outcomes among twin pregnant women in China.
Apart from IPI, other modifiable risk factors of perinatal outcomes in twin pregnancies have been considered. Maternal BMI and gestational weight gain are two important indicators that can be monitored and potentially reflect the nutritional status of pregnant women. It has been illustrated that both maternal BMI and gestational weight gain are positively associated with fetal growth in twins [25]. Inadequate gestational weight gain is associated with a higher incidence of LBW [26, 27]. Similarly, our study revealed that women with lower BMI at delivery were more likely to have neonates with VLBW or ELBW, and those with insufficient gestational weight gain were more likely to have neonates with LBW/LBW. The relationship between maternal BMI and PPH has also been discussed but the results are conflicting [28, 29]. In our study, maternal BMI was found to be associated with MHM, but not with PPH. Despite this, monitoring maternal BMI and gestational weight gain is essential to prevent adverse maternal and neonatal outcomes in twin pregnancies. Furthermore, ART and general anesthesia have been identified as significant risk factors for PPH in most prior studies [30,31,32]. Our findings also indicated that women with ART or general anesthesia were more likely to experience PPH and MHM. This might be because ART usage increases the potential of placenta accreta spectrum [30], placental abruption and gestational hypertension [31], which in turn can lead to greater blood loss. Additionally, women with these conditions are more likely to require emergency surgery under general anesthesia. It has been reported that certain drugs used in general anesthesia can impair uterine contractions, platelet function, and hemostasis, thereby elevating the risk of excessive blood loss [32, 33]. However, the relatively limited sample size of women receiving general anesthesia in our study requires cautious interpretation, and further large-scale studies are needed to confirm this association. Nonetheless, when planning anesthesia protocols for women at higher risk of bleeding, the use of general anesthesia should be approached with caution. More attention should be paid to prevent PPH/MHM in twin pregnancies conceived through ART or managed with general anesthesia. Anemia severity has rarely been considered a potential risk factor for hemorrhage, and our study found that moderate anemia was associated with a higher rate of MHM compared to no anemia. Therefore, it is advisable to correct moderate anemia prior to cesarean section whenever possible.
Although unmodifiable risk factors could not be controlled, the implementation of a risk stratification system would enhance the optimization of available intervention measures and improve therapeutic efficiency, ultimately reducing adverse perinatal outcomes. Specifically, preeclampsia has been associated with PTB (< 37weeks) [34], fetal growth restriction [35] and PPH [32]. Similarly, our study found that preeclampsia was a risk factor of PTB before 34 weeks, LBW/LBW, VLBW or ELBW and MHM. Additionally, we observed a negative association between gestational hypertension and LBW/LBW but a positive association with PPH. These findings underscore the importance of better management of hypertensive disorders and alleviating disease severity may be useful to reduce the risk of adverse maternal and neonatal outcomes. Placenta previa, low-lying placenta, placenta adherence, and a history of cesarean section have been widely recognized as risk factors for hemorrhagic events [7, 32], which is in accordance with our findings. We confirmed a positive associative between placenta previa, low-lying placenta, placenta adherence and PPH, as well as between placenta adherence, cesarean section history, and MHM. Moreover, we observed that placenta previa was related to PTB (< 34 weeks) and LBW/LBW probably due to severe blood loss before delivery. The history of cesarean section was also associated with LBW/LBW, possibly due to earlier delivery age caused by uterine pain from an unhealed scar. These findings highlight the importance of antenatal assessment of placenta location and uterine scar thickness and emphasize the necessity of transferring high-risk patients to tertiary care centers with rapid access to blood products or intensive care units. Furthermore, twin gestations affected by PPROM exhibit significantly shorter latency periods and earlier gestational ages at delivery [36]. In our study, women with PPROM demonstrated a higher incidence of PTB before 34 weeks, LBW/LBW, LBW/NBW and VLBW or ELBW in at least one twin. Preventing the occurrence of PPROM and prolonging latency periods appears important to mitigate adverse neonatal outcomes. Monochorionicity has been identified as a risk factor for lower gestational age and neonatal morbidities [6]. Although our analysis did not reveal a significant association between chorionicity and PTB before 34 weeks, dichorionic diamniotic twin pregnancies exhibited a protective effect against LBW/LBW. Collectively, monitoring women with unmodifiable risk factors including cesarean section history, monochorionic diamniotic twins, PPROM, gestational hypertension, preeclampsia, placenta previa, low-lying placenta, and placental adherence plays a role in preventing adverse maternal and neonatal outcomes.
Strengths and limitations
To the best of our knowledge, comprehensive analyses concentrating on the relationship between IPI and adverse maternal and neonatal outcomes in subsequent twin pregnancies are limited, particularly within the Chinese population. Moreover, our study has systematically categorized twin birth weights into three groups (LBW/LBW, LBW/NBW, and VLBW or ELBW for at least one twin), which enhances our understanding of the role of IPI in fetal growth. However, several limitations should be acknowledged. First, as a retrospective observational study, it lacks critical clinical information such as cervical length, history of cervical insufficiency, number of prior cesarean deliveries, and history of PTB. Although all statistical models adhered closely to the 5–10 events per variable rule to ensure stable estimates [37], the relatively small sample sizes for certain variables and rare outcomes may increase false positive risks and constrain the generalizability of the risk estimates. Second, the data sourced from a single hospital and cohort shaped by China’s unique fertility policies could affect the applicability and generalizability of our findings. Therefore, future multicenter studies and prospective trials with long-term follow-up are warranted to validate and extend these findings.