Abstract
Objective: To identify the risk factors of postpartum hemorrhage (PPH) in women of advanced maternal age (AMA) undergoing natural childbirth and to develop a nomogram model for PPH risk prediction in this population. Methods: This study retrospectively collected data from 220 AMA women who had a natural childbirth at the Third Affiliated Hospital of Chengdu Medical College, Chengdu Pidu District People’s Hospital between March 2020 and May 2023, forming the training cohort. The cohort was categorized into the PPH group and the non-PPH group based on the occurrence of PPH. Clinical data were compared between the two groups. Univariate and multivariate logistic analyses were employed to identify the factors associated with PPH. A predictive model for the risk of PPH in AMA women was developed, and its predictive accuracy was assessed using calibration curves, receiver operating characteristic (ROC) curves, and decision curve analysis (DCA). Additionally, clinical data from 110 AMA women who had a natural childbirth at our hospital between June 2023 and August 2024 were collected, forming the validation cohort. Results: The overall incidence of PPH was 17.58% (58/330), including 39 from the training cohort, and 19 from the validation cohort. Univariate logistic analysis revealed that age, placenta previa, hypertensive disorder of pregnancy (HDP), fetal macrosomia, uterine atony, and scarred uterus were significant risk factors for PPH in AMA women (all P < 0.05). Multivariate logistic model identified age, placenta previa, HDP, uterine atony, and scarred uterus as independent risk factors for PPH in AMA women (all P < 0.05). Based on these independent risk factors, a nomogram model for predicting PPH in AMA women was developed, demonstrating an area under the ROC curve (AUC) of 0.841 (95% CI: 0.773-0.908) in the training cohort and 0.868 (95% CI: 0.767-0.969) in the validation cohort. The calibration curve analysis indicated that the model’s predicted PPH risk in AMA population closely aligned with the actual outcomes, while DCA demonstrated model’s significant clinical utility. Conclusion: The nomogram prediction model developed in this study effectively estimates the risk of PPH in AMA women, offering valuable clinical guidance.
Keywords: Advanced maternal age, postpartum hemorrhage, risk factor, prediction model, nomogram
Introduction
Postpartum hemorrhage (PPH), defined as abnormal vaginal bleeding exceeding 500 mL within 24 h after delivery, is one of the most common complications in obstetrics. If not managed appropriately, it could lead to severe consequence, including disseminated intravascular coagulation (DIC), hemorrhagic shock, and multiorgan failure (MOF), ultimately posing a life-threatening risk to the mother [1]. Statistical data indicates that approximately 25% of global maternal mortalities are associated with PPH [2]. Even among survivors, complications such as puerperal infection, post-transfusion infections, postpartum psychiatric disorders, and Sheehan’s syndrome may still arise, significantly affecting their quality of life [3]. Consequently, effective prediction and prevention of PPH have become global priorities in obstetric research.
Advanced maternal age (AMA) is typically defined as pregnancy in women aged 35 years or older. In recent decades, an increasing number of women worldwide have delayed childbearing. In the United States, the proportion of births to women aged 35-39 years and 40-44 years increased by 5% and 8%, respectively, between 2006-2007 and 2014-2015. By 2015, women aged 35 years or older accounted for 15.7% of all deliveries [4]. In England and Wales, the average age at first childbirth rose from 27.7 years in 1990 to 30.5 years in 2022 [5]. A similar trend has been observed in China [6], where the fertility rate among women aged 35-39 years increased from 10.98% in 2005 to 18.60% in 2015, while the fertility rate among women aged 40-44 years rose from 2.05% to 5.37% over the same period [7]. A recent study in China revealed that the proportion of AMA pregnancies ranged between 10.00% and 20.24% [8,9].
The risk of pregnancy complications is elevated in AMA women during spontaneous delivery, primarily due to the physiological changes associated with the aging process. These changes include declining ovarian function, decreased vascular function, weakened immune system, and diminished uterine contractility [10]. Such physiological changes render older women more susceptible to pregnancy-related complications, including gestational hypertension, gestational diabetes, and PPH. Additionally, impaired placental function and fetal development further elevate maternal and fetal risks. Studies have shown that advanced age significantly increases the likelihood of PPH and increases the risk of poor prognosis [11]. PPH is considered a multifactorial condition, with most cases being preventable through appropriate clinical management. Effective prevention and mitigation of PPH-related adverse outcomes hinges on the early identification of risk factors and timely screening of high-risk individuals, enabling preventive or therapeutic measures based on risk assessment [12]. However, studies and guidelines vary in defining PPH risk factors. Currently, widely used clinical tools for assessing the risk of PPH include the California Maternity Quality Collaboration (CMQCC), the Women’s Health Obstetrics and Neonatal Nurses Association (AWHONN), and the New York Obstetric Bleeding Safety Guide (NYSBOH) [13]. Although these assessment tools incorporate expert opinions and previously identified risk factors, their predictive efficacy is moderate, primarily identifying high-risk obstetric patients undergoing cesarean delivery [13]. Currently, there is a lack of reliable predictive model that accurately screens high-risk AMA women for PPH following natural childbirth.
This study aims to identify the risk factors associated with PPH in AMA women undergoing vaginal delivery and to develop a nomogram-based predictive model for quantifying individual PPH risk. The nomogram serves as a user-friendly risk assessment tool, enabling clinicians to efficiently stratify AMA women based on their PPH risk scores and implement targeted preventive and therapeutic interventions. By facilitating early risk identification and personalized management, this model has the potential to reduce PPH incidence and improve overall maternal and neonatal outcomes.
Materials and methods
General information
Clinical data from 330 cases of AMA women who had a natural childbirth at the Third Affiliated Hospital Affiliated of Chengdu Medical College, Chengdu Pidu District People’s Hospital between March 2020 and August 2024 were retrospectively collected. The patients were divided into a training cohort (n=220, March 2020-May 2023) and a validation cohort (n=110, June 2023-August 2024) using a 2:1 ratio (Figure 1). This study was approved by the hospital’s ethics committee.
Figure 1.
Flow diagram detailing the selection of samples included in this retrospective analysis.
Inclusion criteria: (1) age ≥ 35 years old; (2) natural childbirth; (3) regular maternal and fetal examinations during pregnancy with normal results; (4) gestational age at delivery ≥ 28 weeks; (5) singleton pregnancy.
Exclusion criteria: (1) individuals with communication barriers or mental disorders preventing cooperation; (2) incomplete medical records; (3) coexisting hematologic disorders; (4) coexisting dysfunction of other major organ systems; (5) significant bleeding tendency due to tumors or other underlying diseases.
Clinical data collection
The following data were collected from each AMA case: age, marital status, educational background, pre-pregnancy body mass index (BMI), gestational age, number of pregnancies, postpartum bleeding within 24 h, placental status (presence of placenta previa), amniotic fluid condition (presence of contamination), hypertensive disorders of pregnancy (HDP), gestational diabetes mellitus (GDM), uterine fibroids, fetal weight (presence of macrosomia), uterine atony, scarred uterus, and the use of assisted reproductive technologies.
Diagnostic criteria for PPH
According to the World Health Organization (WHO), PPH is defined as a cumulative blood loss ≥ 500 mL within 24 h after delivery [14].
Measurement of PPH
Sterile gauze was pre-weighed before use. The gauze was then used to absorb blood during delivery. After absorption, the gauze was re-weighed. Finally, blood loss (mL) was calculated according to the formula: blood loss (mL) = (post-absorption gauze weight (g) - initial gauze weight (g))/1.05 [15].
Statistical analysis
Statistical analyses were conducted using SPSS version 22.0. Kolmogorov-Smirnov normality test was used to assess the normality of data distribution. All continuous variables were normally distributed and presented as mean ± standard (x̅ ± sd). The comparison between groups was conducted using the t-test. The categorical variables were expressed as percentage and compared by χ2 test or Fisher’s exact tests. Univariate and multivariate logistic analysis were conducted to identify risk factors for PPH in AMA. The nomogram were constructed using R 4.4.0 software. Its diagnostic accuracy and clinical utility were evaluated using calibration curves, receiver operating characteristic (ROC) curves, and decision curve analysis (DCA). A two-sided P-value < 0.05 was considered statistically significant.
Results
Incidence of PPH in AMA women
The incidence of PPH among AMA women undergoing natural birth was 17.58% (58/330). In the training cohort, 39 women experienced PPH, while 19 women in the validation cohort developed PPH. Based on the occurrence of PPH, the training cohort was categorized into the PPH group (n=39) and the non-PPH group (n=181), while the validation cohort was similarly categorized into the PPH group (n=19) and the non-PPH group (n=91).
Comparison of clinical data between the training and validation cohort
The comparison of clinical data between the training and validation cohorts revealed no statistically significant differences (P > 0.05) (Table 1).
Table 1.
Comparison of clinical data between the training and validation cohorts [n (%)]/(x̅ ± sd)
| Training cohort (n=220) | Validation cohort (n=110) | t/χ2 | P | |
|---|---|---|---|---|
| Age | 38.51±2.99 | 38.86±2.64 | 1.055 | 0.292 |
| Marital status | ||||
| Unmarried | 4 (1.82) | 3 (2.73) | 0.018 | 0.893 |
| Married | 216 (98.18) | 107 (97.27) | ||
| Educational background | 0.070 | 0.966 | ||
| Primary school | 25 (11.36) | 13 (11.82) | ||
| Middle and high school | 65 (29.55) | 31 (28.18) | ||
| College | 130 (59.09) | 66 (60.00) | ||
| Pre-pregnancy BMI (kg/m2) | 22.46±2.50 | 22.29±2.16 | 0.196 | 0.845 |
| Pregnancy BMI (kg/m2) | 27.85±2.96 | 28.17±2.54 | 0.969 | 0.333 |
| Gestational age (week) | 38.07±2.21 | 38.24±2.36 | 0.637 | 0.524 |
| Number of Pregnancies | 1.210 | 0.546 | ||
| 1-2 times | 165 (75.00) | 88 (80.00) | ||
| 3-4 times | 48 (21.82) | 20 (18.18) | ||
| > 4 times | 7 (3.18) | 2 (1.82) | ||
| Placenta previa | 0.140 | 0.708 | ||
| No | 195 (88.64) | 99 (90.00) | ||
| Yes | 25 (11.36) | 11 (10.00) | ||
| Amniotic fluid contamination | 0.050 | 0.824 | ||
| No | 188 (85.45) | 95 (96.36) | ||
| Yes | 32 (14.55) | 15 (13.64) | ||
| HDP | 0.010 | 0.919 | ||
| No | 181 (82.27) | 90 (81.82) | ||
| Yes | 39 (17.73) | 20 (18.18) | ||
| GDM | ||||
| No | 185 (84.09) | 91 (82.73) | 0.100 | 0.752 |
| Yes | 35 (15.91) | 19 (17.27) | ||
| Uterine fibroids | 0.052 | 0.819 | ||
| No | 190 (86.36) | 96 (87.27) | ||
| Yes | 30 (13.64) | 14 (12.73) | ||
| Fetal macrosomia | 0.087 | 0.769 | ||
| No | 204 (92.73) | 101 (91.82) | ||
| Yes | 16 (7.27) | 9 (8.18) | ||
| Uterine atony | 0.185 | 0.667 | ||
| No | 155 (70.45) | 80 (72.73) | ||
| Yes | 65 (29.55) | 30 (27.27) | ||
| Scarred uterus | 0.202 | 0.653 | ||
| No | 188 (85.45) | 96 (87.27) | ||
| Yes | 32 (14.55) | 14 (12.73) | ||
| Assisted reproduction | 0.191 | 0.662 | ||
| No | 217 (98.64) | 107 (97.27) | ||
| Yes | 3 (1.36) | 3 (2.73) |
Note: BMI, body mass index; HDP, hypertensive disorder of pregnancy; GDM, gestational diabetes mellitus.
Comparison of clinical data between the PPH and non-PPH groups in the training cohort
As shown in Table 2, the average age of the PPH group was 40.31±2.53 years, significantly higher than 38.12±2.95 years in the non-PPH group (t=4.300, P < 0.001). Besides, significant differences were observed between the two groups in the proportions of placenta previa (χ2=20.142, P < 0.001), HDP (χ2=7.916, P=0.005), uterine atony (χ2=9.615, P=0.002), and scarred uterus (χ2=17.387, P < 0.001).
Table 2.
Comparison of clinical data between the PPH and non-PPH groups in the training cohort [n (%)]/(x̅ ± sd)
| Variable | PPH (n=39) | Non-PPH (n=181) | t/χ2 | P |
|---|---|---|---|---|
| Age | 40.31±2.53 | 38.12±2.95 | 4.300 | < 0.001 |
| Marital status | 0.040 | 0.842 | ||
| Unmarried | 1 (2.56) | 3 (1.66) | ||
| Married | 38 (97.44) | 178 (98.34) | ||
| Educational background | 0.128 | 0.938 | ||
| Primary school | 4 (10.26) | 21 (11.60) | ||
| Middle and high school | 11 (28.21) | 54 (29.83) | ||
| College | 24 (61.54) | 106 (58.56) | ||
| Pre-pregnancy BMI (kg/m2) | 22.93±1.47 | 22.35±2.67 | 1.313 | 0.191 |
| Pregnancy BMI (kg/m2) | 27.91±3.04 | 27.83±2.96 | 0.150 | 0.881 |
| Gestational age (week) | 38.33±2.06 | 38.01±2.24 | 0.820 | 0.413 |
| Number of Pregnancies | 0.913 | 0.634 | ||
| 1-2 times | 30 (76.92) | 135 (74.59) | ||
| 3-4 times | 7 (17.95) | 41 (22.65) | ||
| > 4 times | 2 (5.13) | 5 (2.76) | ||
| Placenta previa | 20.142 | < 0.001 | ||
| No | 26 (66.67) | 169 (93.37) | ||
| Yes | 13 (33.33) | 12 (6.63) | ||
| Amniotic fluid contamination | 0.114 | 0.736 | ||
| No | 34 (87.18) | 154 (85.08) | ||
| Yes | 5 (12.82) | 27 (14.92) | ||
| HDP | 7.916 | 0.005 | ||
| No | 26 (66.67) | 155 (85.64) | ||
| Yes | 13 (33.33) | 26 (14.36) | ||
| GDM | 0.010 | 0.922 | ||
| No | 33 (84.62) | 152 (83.98) | ||
| Yes | 6 (15.38) | 29 (16.02) | ||
| Uterine fibroids | 0.123 | 0.726 | ||
| No | 33 (84.62) | 157 (86.74) | ||
| Yes | 6 (15.38) | 24 (13.26) | ||
| Fetal macrosomia | 3.279 | 0.070 | ||
| No | 33 (84.62) | 171 (94.48) | ||
| Yes | 6 (15.38) | 10 (5.52) | ||
| Uterine atony | 9.615 | 0.002 | ||
| No | 19 (48.72) | 136 (75.14) | ||
| Yes | 20 (51.28) | 45 (24.86) | ||
| Scarred uterus | 17.387 | < 0.001 | ||
| No | 25 (64.10) | 163 (90.06) | ||
| Yes | 14 (35.90) | 18 (9.94) | ||
| Assisted reproduction | - | 0.445* | ||
| No | 38 (97.44) | 179 (98.90) | ||
| Yes | 1 (2.56) | 2 (1.10) |
Note: PPH, postpartum hemorrhage; BMI, body mass index; HDP, hypertensive disorder of pregnancy; GDM, gestational diabetes mellitus;
Fisher’s exact tests.
Univariate logistic analysis of PPH in AMA women
Variable assignments are listed in Table 3. A univariate logistic analysis was conducted with PPH as the dependent variable and AMA and pregnancy related variables as independent variables. The analysis revealed that age (odds ratio [OR]=1.275, 95% confidence interval [95% CI]: 1.130-1.439, P < 0.001), placenta previa (OR=7.042, 95% CI: 2.901-17.090, P < 0.001), HDP (OR=2.981, 95% CI: 1.360-6.533, P=0.006), fetal macrosomia (OR=3.109, 95% CI: 1.057-9.142, P=0.039), uterine atony (OR=3.181, 95% CI: 1.560-6.488, P=0.001), and scarred uterus (OR=5.071, 95% CI: 2.244-11.462, P < 0.001) were significant factors associated with PPH in AMA women (Table 4).
Table 3.
Assignment of each factor
| Variable | Assignment |
|---|---|
| Age | Enter the actual value |
| Pre-pregnancy BMI | Enter the actual value |
| Pregnancy BMI | Enter the actual value |
| Gestational age | Enter the actual value |
| Marital status | 0= Unmarried, 1= Married |
| Educational background | 0= Primary school, 1= Middle and high school, 2= College |
| Number of Pregnancies | 0=1-2 times, 1=3-4 times, 2= > 4 times |
| Placenta previa | 0= no, 1= yes |
| Placenta previa | 0= no, 1= yes |
| HDP | 0= no, 1= yes |
| GDM | 0= no, 1= yes |
| Uterine fibroids | 0= no, 1= yes |
| Fetal macrosomia | 0= no, 1= yes |
| Uterine atony | 0= no, 1= yes |
| Scarred uterus | 0= no, 1= yes |
| Assisted reproduction | 0= no, 1= yes |
Note: BMI, body mass index; HDP, hypertensive disorder of pregnancy; GDM, gestational diabetes mellitus.
Table 4.
Univariate logistic analysis of PPH in AMA women
| B | S.E | P | OR (95% CI) | |
|---|---|---|---|---|
| Age | 0.243 | 0.062 | < 0.001 | 1.275 (1.130-1.439) |
| Marital status | ||||
| Unmarried | Reference | |||
| Married | -0.466 | 1.168 | 0.703 | 0.640 (0.065-6.325) |
| Educational background | ||||
| Primary school | Reference | |||
| Middle and high school | 0.067 | 0.638 | 0.916 | 1.069 (0.306-3.734) |
| College | 0.173 | 0.591 | 0.770 | 1.189 (0.374-3.782) |
| Pre-pregnancy BMI (kg/m2) | 0.093 | 0.073 | 0.198 | 1.098 (0.952-1.266) |
| Pregnancy BMI (kg/m2) | 0.050 | 0.077 | 0.513 | 1.052 (0.904-1.223) |
| Gestational age (week) | -0.066 | 0.056 | 0.236 | 0.936 (0.838-1.044) |
| Number of Pregnancies | ||||
| 1-2 times | Reference | |||
| 3-4 times | -0.264 | 0.456 | 0.563 | 0.768 (0.314-1.878) |
| > 4 times | 0.588 | 0.861 | 0.495 | 1.800 (0.333-9.725) |
| Placenta previa | ||||
| No | Reference | |||
| Yes | 1.952 | 0.452 | < 0.001 | 7.042 (2.901-17.090) |
| Amniotic fluid contamination | ||||
| No | Reference | |||
| Yes | -0.176 | 0.522 | 0.736 | 0.839 (0.301-2.335) |
| HDP | ||||
| No | Reference | |||
| Yes | 1.092 | 0.400 | 0.006 | 2.981 (1.360-6.533) |
| GDM | ||||
| No | Reference | |||
| Yes | -0.048 | 0.488 | 0.921 | 0.953 (0.366-2.480) |
| Uterine fibroids | ||||
| No | Reference | |||
| Yes | 0.173 | 0.495 | 0.726 | 1.189 (0.451-3.138) |
| Fetal macrosomia | ||||
| No | Reference | |||
| Yes | 1.134 | 0.550 | 0.039 | 3.109 (1.057-9.142) |
| Uterine atony | ||||
| No | Reference | |||
| Yes | 1.157 | 0.364 | 0.001 | 3.181 (1.560-6.488) |
| Scarred uterus | ||||
| No | Reference | |||
| Yes | 1.624 | 0.416 | < 0.001 | 5.071 (2.244-11.462) |
| Assisted reproduction | ||||
| No | Reference | |||
| Yes | 0.857 | 1.238 | 0.489 | 2.355 (0.208-26.642) |
Note: BMI, body mass index; HDP, hypertensive disorder of pregnancy; GDM, gestational diabetes mellitus; OR, odds ratio; CI, confidence interval.
Multivariate logistic analysis of PPH in AMA women
Variables with P value less than 0.05 in univariate analysis were incorporated into the multivariate analysis, and the results identified age (OR=1.302, 95% CI: 1.132-1.499, P < 0.001), placenta previa (OR=5.295, 95% CI: 1.891-14.827, P=0.002), HDP (OR=3.086, 95% CI: 1.220-7.808, P=0.017), uterine atony (OR=2.597, 95% CI: 1.128-5.976, P=0.025), and scarred uterus (OR=5.593, 95% CI: 2.094-14.944, P < 0.001) as independent risk factors for PPH in AMA women (Table 5).
Table 5.
Multivariate logistic analysis of PPH in AMA women
| B | S.E | P | OR (95% CI) | |
|---|---|---|---|---|
| Age | 0.264 | 0.072 | < 0.001 | 1.302 (1.132-1.499) |
| Placenta previa | ||||
| No | Reference | |||
| Yes | 1.667 | 0.525 | 0.002 | 5.295 (1.891-14.827) |
| HDP | ||||
| No | Reference | |||
| Yes | 1.127 | 0.474 | 0.017 | 3.086 (1.220-7.808) |
| Uterine atony | ||||
| No | Reference | |||
| Yes | 0.954 | 0.425 | 0.025 | 2.597 (1.128-5.976) |
| Scarred uterus | ||||
| No | Reference | |||
| Yes | 1.722 | 0.501 | < 0.001 | 5.593 (2.094-14.944) |
Note: HDP, hypertensive disorder of pregnancy; OR, odds ratio; CI, confidence interval.
Construction of nomogram prediction model
A nomogram prediction model was developed based on the independent risk factors identified in the multivariate logistic analysis, including age, placenta previa, HDP, uterine atony, and scarred uterus. Each predictor was assigned a score based on its contribution to the outcome variable. The total score was then used to predict the individual risk of PPH in AMA women (Figure 2).
Figure 2.
Nomogram prediction model for PPH in AMA women. PPH, postpartum hemorrhage; AMA, advanced maternal age; HDP, hypertensive disorder of pregnancy.
Validation of nomogram prediction model
In the training cohort, the area under the ROC curve (AUC) was 0.841 (95% CI: 0.773-0.908), indicating excellent discriminatory performance of the nomogram model (Figure 3A). The Hosmer-Lemeshow test confirmed that the model exhibited good fitness and calibration (P=0.787) (Figure 4A). Decision curve analysis (DCA) revealed that when the threshold probability ranged from 0.05 to 1, the net benefit remained positive, suggesting that the nomogram model holds significant clinical value (Figure 5A).
Figure 3.
ROC curves of the predictive model for PPH in the training (A) and validation (B) cohorts. ROC, receiver operating characteristic; PPH, postpartum hemorrhage.
Figure 4.
Calibration curves of the predictive model in the training (A) and validation (B) cohorts.
Figure 5.
DCA for the predictive model in the training (A) and validation (B) cohorts. DCA, decision curve analysis.
In the validation cohort, the AUC was 0.868 (95% CI: 0.767-0.969), further validating the model’s strong discriminatory ability (Figure 3B). Hosmer-Lemeshow test also indicated good fit and calibration (P=0.178) (Figure 4B). DCA analysis showed that when the threshold probability ranged from 0.05 to 0.80, the net benefit remained positive, further supporting the model’s clinical applicability (Figure 5B). These findings suggest that our nomogram model is a reliable tool for guiding clinical decision-making in assessing PPH risk in AMA women.
Discussion
Postpartum hemorrhage (PPH) is one of the most severe complications of childbirth, posing a significant threat to maternal life. Women of advanced maternal age (AMA) are at a higher risk of PPH compared to those of optimal childbearing age, with incidence of PPH varying across different time periods and regions. In this study, 17.58% of the 330 AMA women undergoing natural childbirth experienced PPH, a rate consistent with previously reported range of 6.1-19.0% [3,16]. This suggests that AMA women face a higher risk of PPH during natural childbirth. Therefore, identifying independent risk factors for PPH and developing an effective prediction model are essential for improving maternal outcomes in AMA patients. Previous studies, both domestic and international, have identified several risk factors for PPH, including age ≥ 35 years, placental abnormalities, pregnancy complications (e.g., HDP and GDM), fetal macrosomia, multiple pregnancies, scarred uterus, and polyhydramnios [17,18]. In this study, we categorized 220 AMA into a PPH group and a non-PPH group. Multivariate logistic analysis identified age, placenta previa, HDP, uterine atony, and scarred uterus as independent risk factors for PPH in AMA women. Furthermore, the nomogram prediction model demonstrated strong predictive accuracy, making it a valuable tool for assessing PPH risk in AMA patients.
In this study, age was identified as an independent risk factor for the development of PPH. Li et al. reported that the incidence of PPH is higher in women over 35 years compared to younger women [19]. This may be attributed to the decline in physiological adaptability associated with advancing reproductive age. Research indicates that ovarian function declines with age, leading to alterations in the structure and number of uterine smooth muscle fibers, resulting in a progressive reduction in uterine contractility [20]. Therefore, AMA women undergoing natural childbirth may struggle to achieve sustained and forceful uterine contractions, resulting in incomplete vascular constriction following placental detachment. This impairment in bleeding control increases the risk of PPH. Luca et al. found that AMA women are twice as likely to experience uterine atony during natural childbirth compared to younger women, significantly increasing the risk of PPH [21]. Uterine atony is one of the leading causes of PPH, as demonstrated in the large international WOMAN’s trial, where it accounted for more than 60% of cases [22]. In this study, uterine atony was also identified as an independent risk factor for PPH in AMA women. Additionally, AMA is frequently associated with an increased incidence of labor abnormalities, particularly uterine atony and insufficient cervical dilation during delivery [23]. Therefore, for AMA women presenting with uterine atony, it is crucial to enhance clinical monitoring and promptly initiate interventions, such as the administration of oxytocin and uterine massage, to reduce the risk of PPH.
This study identified placenta previa as a significant risk factor for PPH in AMA women. Placenta previa occurs when the placenta implants in the lower segment of the uterus partially or completely cover the internal cervical os. Due to the dense vascular network and abundant blood supply in this region, the risk of severe hemorrhage is heightened during delivery, as these blood vessels are more susceptible to stretching or rupture. Furthermore, placenta previa is associated with an increased risk of abnormal placental implantation, including placenta accreta or placenta increta, which can result in incomplete placental separation and uterine atony in the lower segment, thereby significantly elevating the risk of PPH. Dang et al. reported that placenta previa is a significant contributor to severe PPH and may even necessitate hysterectomy [24]. A meta-analysis further indicated that among patients with severe PPH, 32.3% had placenta previa, and 33.8% had abnormal placental implantation [25]. AMA is associated with age-related changes in the uterine muscle layer, leading to decreased contractility and impaired uterine recovery following placental detachment, further increasing the hemorrhage risk. Sahu et al. demonstrated that the incidence of placenta previa rises with advanced maternal age, a trend closely linked to endometrial aging and structural changes in the uterus [26]. Ultrasound is the primary diagnostic tool for placenta previa. By assessing the uterine artery pulsatility index (UtAPI) and resistance index (UtA-RI), it is possible to predict the depth of placental invasion and the risk of severe hemorrhage [27]. Clinically, the bleeding risk associated with placenta previa in AMA women should be evaluated using standardized protocols to optimize pregnancy outcomes.
This study identified a strong association between HDP and the risk of PPH in AMA women. HDP disrupts coagulation mechanisms during pregnancy, leading to alterations in coagulation factors and platelet dysfunction, which in turn weakens the vascular ability to regulate contraction and dilation. Following placental separation, the inability of blood vessels to constrict promptly increases the risk of prolonged hemorrhage [28]. Cagino et al. found that HDP increases vascular fragility, leading to microvascular rupture and increased vascular permeability, which exacerbates bleeding [29]. Furthermore, research has shown that the incidence of HDP is higher in older pregnant women compared to their younger counterparts, likely due to age-related declines in vascular elasticity, impairing vascular adaptation to hypertension [30]. As a result, the management of labor in AMA women with HDP should focus on the PPH risk reduction. Strategies such as blood pressure control and the use of hemostatic agents should be employed to minimize hemorrhage risk and ensure the safety of both mother and infant.
This study also identified a strong association between a scarred uterus and an increased risk of PPH in AMA women. A scarred uterus typically results from previous cesarean sections or intrauterine surgeries. Multiple studies have demonstrated that a scarred uterus increases the risk of uterine rupture during labor, thereby elevating the likelihood of PPH [31,32]. This increased risk is likely due to the thinning of the uterine wall in the scarred region, as well as the dense and fragile blood vessels. Wang et al. [33] noted that a scarred uterus not only compromises the structural integrity of the uterus but may also impair the contractile function of the uterine muscle, further enhancing the risk of PPH. Given that AMA women often have a higher incidence of scarred uterus due to prior surgeries, a comprehensive risk assessment should be conducted before labor, and appropriate PPH prevention strategies should be implemented [34].
This study presents a nomogram-based predictive model for assessing the risk of PPH in AMA women. Among existing PPH risk assessment tools, the California Maternal Quality Care Collaborative (CMQCC) scoring system is widely used [35]. However, it primarily relies on medical history and placental abnormalities, lacking individualized risk stratification, which limits its accuracy in predicting PPH risk among AMA women. Recent advancements in machine learning-based prediction models, including random forests, logistic regression, and deep learning, have demonstrated improved predictive accuracy [36]. Nevertheless, these models require extensive datasets and offer limited clinical interpretability, restricting their practical implementation in routine obstetric care. In contrast, the nomogram model developed in this study provides a visually intuitive and quantitative approach to PPH risk assessment in AMA women, enabling clinicians to formulate individualized management strategies. For low-risk cases, routine obstetric care with continuous labor monitoring is sufficient, minimizing unnecessary medical interventions. For high-risk cases, proactive preparation, including blood transfusion readiness and enhanced perinatal care, is essential to mitigate adverse maternal and neonatal outcomes.
In summary, this study systematically identified key independent risk factors for PPH in AMA women, including age, placenta previa, HDP, uterine atony, and scarred uterus. Utilizing these factors, a nomogram-based PPH risk prediction model was developed, alongside a stratified management strategy tailored to varying risk levels. This approach aims to enhance clinical decision-making and optimize intervention strategies to mitigate PPH-related complications in AMA pregnancies. However, this study has several limitations. First, the relatively small sample size and the fact that the data were primarily drawn from a single medical institution may limit the external validity of the findings. To enhance the generalizability of the results, future research should adopt a multi-center and large-sample study design. Moreover, while this study has identified several key risk factors for PPH, other potential contributors, such as maternal lifestyle, psychological health, and genetic predispositions, were not fully explored. Future studies should incorporate these additional factors to achieve a more comprehensive analysis, further confirming the model’s applicability and reliability across diverse populations.
Conclusion
Age, placenta previa, HDP, uterine atony, and scarred uterus are independent contributors to PPH risk in AMA. The nomogram, incorporating these factors, demonstrates robust discrimination and high accuracy in predicting PPH among AMA women, enabling targeted interventions for high-risk individuals to enhance maternal safety.
Disclosure of conflict of interest
None.
References
- 1.Patek K, Friedman P. Postpartum hemorrhage-epidemiology, risk factors, and causes. Clin Obstet Gynecol. 2023;66:344–356. doi: 10.1097/GRF.0000000000000782. [DOI] [PubMed] [Google Scholar]
- 2.Nigussie J, Girma B, Molla A, Tamir T, Tilahun R. Magnitude of postpartum hemorrhage and its associated factors in Ethiopia: a systematic review and meta-analysis. Reprod Health. 2022;19:63. doi: 10.1186/s12978-022-01360-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Song Z, Wang X, Zhou Y, Wang Y, Zhang D. Development and validation of prognostic nomogram for postpartum hemorrhage after vaginal delivery: a retrospective cohort study in China. Front Med. 2022;9:804769. doi: 10.3389/fmed.2022.804769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sheen JJ, Wright JD, Goffman D, Kern-Goldberger AR, Booker W, Siddiq Z, D’Alton ME, Friedman AM. Maternal age and risk for adverse outcomes. Am J Obstet Gynecol. 2018;219:390.e1–390.e15. doi: 10.1016/j.ajog.2018.08.034. [DOI] [PubMed] [Google Scholar]
- 5.Birth characteristics in England and Wales - Office for National Statistics. Birth Charact Engl Wales 2022. 2024 [Google Scholar]
- 6.Li YH, Wang YP, Dai L, Zhou GX, Liang J, Li Q, Zhu J. The trend of national advanced maternal age woman proportion in hospital-based surveillance. Zhonghua Yu Fang Yi Xue Za Zhi. 2009;43:1073–1076. [PubMed] [Google Scholar]
- 7.National Bureau of Statistics of China. CHINA STATISTICAL YEARBOOK 2016|China YearBooks. 2016 [Google Scholar]
- 8.Zhang X, Liu KB, Liu FJ, Yang HJ, Yu Y. Clinical characteristics and pregnancy outcomes of pregnant women in different age groups in Beijing. Zhonghua Fu Chan Ke Za Zhi. 2018;53:452–458. doi: 10.3760/cma.j.issn.0529-567x.2018.07.003. [DOI] [PubMed] [Google Scholar]
- 9.Liu Q, Wang XX, Zhang YK, Li JH, Wang L. Correlation between pregnancy-induced hypertension and age in pregnant women from Hebei province, 2016. Zhonghua Liu Xing Bing Xue Za Zhi. 2018;39:1270–1273. doi: 10.3760/cma.j.issn.0254-6450.2018.09.024. [DOI] [PubMed] [Google Scholar]
- 10.Zhou Y, Yin S, Sheng Q, Yang J, Liu J, Li H, Yuan P, Zhao Y. Association of maternal age with adverse pregnancy outcomes: a prospective multicenter cohort study in China. J Glob Health. 2023;13:04161. doi: 10.7189/jogh.13.04161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.American College of Obstetricians and Gynecologists’ Committee on Clinical Consensus-Obstetrics; Gantt A Society for Maternal-Fetal Medicine; Metz TD, Kuller JA, Louis JM Society for Maternal-Fetal Medicine; Cahill AG, Turrentine MA American College of Obstetricians and Gynecologists. Obstetric care consensus #11, pregnancy at age 35 years or older. Am J Obstet Gynecol. 2023;228:B25–B40. doi: 10.1016/j.ajog.2022.07.022. [DOI] [PubMed] [Google Scholar]
- 12.Neary C, Naheed S, McLernon DJ, Black M. Predicting risk of postpartum haemorrhage: a systematic review. BJOG Int J Obstet Gynaecol. 2021;128:46–53. doi: 10.1111/1471-0528.16379. [DOI] [PubMed] [Google Scholar]
- 13.Kawakita T, Mokhtari N, Huang JC, Landy HJ. Evaluation of risk-assessment tools for severe postpartum hemorrhage in women undergoing cesarean delivery. Obstet Gynecol. 2019;134:1308–1316. doi: 10.1097/AOG.0000000000003574. [DOI] [PubMed] [Google Scholar]
- 14.WHO Recommendations for the Prevention and Treatment of Postpartum Haemorrhage. Geneva: World Health Organization; 2012. [PubMed] [Google Scholar]
- 15.Al Kadri HMF, Al Anazi BK, Tamim HM. Visual estimation versus gravimetric measurement of postpartum blood loss: a prospective cohort study. Arch Gynecol Obstet. 2011;283:1207–1213. doi: 10.1007/s00404-010-1522-1. [DOI] [PubMed] [Google Scholar]
- 16.Ortuanya KE, Eleje GU, Ezugwu FO, Odugu BU, Ikechebelu JI, Ugwu EO, Eke AC, Awkadigwe FI, Ezenwaeze MN, Ofor IJ, Okafor CC, Okafor CG. Prophylactic tranexamic acid for reducing intraoperative blood loss during cesarean section in women at high risk of postpartum hemorrhage: a double-blind placebo randomized controlled trial. Womens Health. 2024;20:17455057231225311. doi: 10.1177/17455057231225311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhang Y, Chen L, Zhou W, Lin J, Wen H. Nomogram to predict postpartum hemorrhage in cesarean delivery for twin pregnancies: a retrospective cohort study in China. Front Med. 2023;10:1139430. doi: 10.3389/fmed.2023.1139430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wang Y, Xiao J, Hong F. A risk prediction model of perinatal blood transfusion for patients who underwent cesarean section: a case control study. BMC Pregnancy Childbirth. 2022;22:373. doi: 10.1186/s12884-022-04696-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Li J, Yan J, Jiang W. The role of maternal age on adverse pregnancy outcomes among primiparous women with singleton birth: a retrospective cohort study in urban areas of China. J Matern Fetal Neonatal Med. 2023;36:2250894. doi: 10.1080/14767058.2023.2250894. [DOI] [PubMed] [Google Scholar]
- 20.Carducci ME, Izbizky G. Advanced maternal age as a risk factor for adverse maternal and perinatal outcomes. Rev Fac Cienc Medicas Cordoba Argent. 2024;81:24–39. doi: 10.31053/1853.0605.v81.n1.41447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Luca AM, Carvalho JCA, Ramachandran N, Balki M. The effect of morbid obesity or advanced maternal age on oxytocin-induced myometrial contractions: an in vitro study. Can J Anaesth J Can Anesth. 2020;67:836–846. doi: 10.1007/s12630-020-01615-6. [DOI] [PubMed] [Google Scholar]
- 22.WOMAN Trial Collaborators. Effect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post-partum haemorrhage (WOMAN): an international, randomised, double-blind, placebo-controlled trial. Lancet Lond Engl. 2017;389:2105–2116. doi: 10.1016/S0140-6736(17)30638-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Cao J, Xu W, Liu Y, Zhang B, Zhang Y, Yu T, Huang T, Zou Y, Zhang B. Trends in maternal age and the relationship between advanced age and adverse pregnancy outcomes: a population-based register study in Wuhan, China, 2010-2017. Public Health. 2022;206:8–14. doi: 10.1016/j.puhe.2022.02.015. [DOI] [PubMed] [Google Scholar]
- 24.Dang X, Xiong G, Fan C, He Y, Sun G, Wang S, Liu Y, Zhang L, Bao Y, Xu J, Du H, Deng D, Chen S, Li Y, Gong X, Wu Y, Wu J, Lin X, Qiao F, Zeng W, Feng L, Liu H. Systematic external evaluation of four preoperative risk prediction models for severe postpartum hemorrhage in patients with placenta previa: a multicenter retrospective study. J Gynecol Obstet Hum Reprod. 2022;51:102333. doi: 10.1016/j.jogoh.2022.102333. [DOI] [PubMed] [Google Scholar]
- 25.Fan D, Xia Q, Liu L, Wu S, Tian G, Wang W, Wu S, Guo X, Liu Z. The incidence of postpartum hemorrhage in pregnant women with placenta previa: a systematic review and meta-analysis. PloS One. 2017;12:e0170194. doi: 10.1371/journal.pone.0170194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sahu SA, Shrivastava D. Maternal and perinatal outcomes in placenta previa: a comprehensive review of evidence. Cureus. 2024;16:e59737. doi: 10.7759/cureus.59737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.DeBolt CA, Rosenberg HM, Pruzan A, Goldberger C, Kaplowitz E, Buckley A, Vieira L, Stone J, Bianco A. Patients with resolution of low-lying placenta and placenta previa remain at increased risk of postpartum hemorrhage. Ultrasound Obstet Gynecol. 2022;60:103–108. doi: 10.1002/uog.24825. [DOI] [PubMed] [Google Scholar]
- 28.Irfan A, Haider SH, Sheikh SM, Larik MO, Abbas M, Hashmi MR. Evaluation of antihypertensives for post partum management of hypertensive disorders of pregnancy: a systematic review and meta-analysis. Curr Probl Cardiol. 2024;49:102584. doi: 10.1016/j.cpcardiol.2024.102584. [DOI] [PubMed] [Google Scholar]
- 29.Cagino KA, Wiley RL, Ghose I, Ciomperlik HN, Sibai BM, Mendez-Figueroa H, Chauhan SP. Risk of postpartum hemorrhage in hypertensive disorders of pregnancy: stratified by severity. Am J Perinatol. 2024;41:2165–2174. doi: 10.1055/a-2297-8790. [DOI] [PubMed] [Google Scholar]
- 30.Newman RA, Naqvi M, Levian C, Smithson SD, Esakoff T. Risk of preterm delivery in very advanced maternal age parturients utilizing in vitro fertilization. Am J Perinatol. 2024;41:1532–1537. doi: 10.1055/a-2157-3106. [DOI] [PubMed] [Google Scholar]
- 31.Madhuri MS, Jha N, Pampapati V, Chaturvedula L, Jha AK. Fetomaternal outcome of scarred uterine rupture compared with primary uterine rupture: a retrospective cohort study. J Perinat Med. 2023;51:1067–1073. doi: 10.1515/jpm-2023-0018. [DOI] [PubMed] [Google Scholar]
- 32.Wang S, Hu Q, Liao H, Wang K, Yu H. Perinatal outcomes of pregnancy in women with scarred uteri. Int J Womens Health. 2023;15:1453–1465. doi: 10.2147/IJWH.S422187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wang S, Zhu C, Su J, Zhu Z, Ma W, Wang S. Construction and application of a prediction model for the delivery outcome of women with scarred uterus based on ultrasonic parameters. Am J Transl Res. 2023;15:264–272. [PMC free article] [PubMed] [Google Scholar]
- 34.Zhang Y, Ding W, Dai X, Wang H, Cheng Y, Dai J, Zhu X, Xu X. Burden of multiple high-risk factors in pregnancy before and after the universal two-child policy in Chinese women: an observational study. J Glob Health. 2024;14:04134. doi: 10.7189/jogh.14.04134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Wen T. Interpretable machine learning predicts postpartum hemorrhage at time of admission. Am J Obstet Gynecol MFM. 2024;6:101355. doi: 10.1016/j.ajogmf.2024.101355. [DOI] [PubMed] [Google Scholar]
- 36.Dogru S, Ezveci H, Akkus F, Bahçeci P, Karanfil Yaman F, Acar A. Artificial intelligence in predicting postpartum hemorrhage in twin pregnancies undergoing cesarean section. Twin Res Hum Genet. 2025:1–7. doi: 10.1017/thg.2024.48. [DOI] [PubMed] [Google Scholar]