This study was a hospital-based cohort study conducted at the Affiliated Hospital of Yunnan University between March 2021 and August 2022. We started the study by determining the sample size, considering the different number of groups (k) to be included. For k = 6, the sample size (n) was calculated to be 35.141; for k = 4, the sample size was n= 44.599; and for k = 3, the sample size was 52.40. Subsequently, the study enrolled 150 pregnant patients who underwent routine obstetric ultrasound examinations. Additionally, the study included 110 primiparas who underwent routine pelvic floor ultrasound examinations 42–45 days post-delivery. Additionally, a control group comprising 55 healthy nulliparous women, with no history of abortion, was included into the study. Ethical approval for this study was obtained from the Institutional Review Board of the Affiliated Hospital of Yunnan University (Ethics review No.: 2021164).

The study focused on first-time pregnant patients and primiparas who experienced a single full-term birth.

Participants were excluded if they presented a body mass index (BMI) $>$35 kg/m${}^2$; were unable to perform effective anal contractions and Valsalva maneuvers during examinations; individuals with pelvic and urogenital tumors; a long history of cough and chronic constipation; history of pelvic or gynecologic surgery; or previous pelvic dysfunction diseases such as POP and stress urinary incontinence.

Dietz et al. [13] analyzed and proposed the application of ultrasound in the quantification of female POP. The diagnosis utilized a horizontal reference line positioned behind the lower edge of the pubic symphysis. The vertical distance between the bladder neck and external cervical os to this reference line was measured during maximum Valsalva maneuvers. A diagnosis of cystocele was established if the bladder neck was at or below the reference line. Similarly, uterine prolapse was diagnosed when the external cervical os reached or was below this line. For rectocele, a straight line extending cephalad parallel to the ventral internal sphincter and anal canal served as the reference. A diagnosis was made if the anterior rectal wall bulged anteriorly beyond this line during maximum Valsalva maneuvers.

After bladder emptying, subjects assumed a lithotomy position. To minimize the impact of posture on examination results, consistent angles of hip abduction and knee flexion were maintained throughout the procedure. The examinations were conducted using a SuperSonic Aixplorer ultrasound diagnostic instrument (SuperSonic Imagine Company, Aix-en-Provence, Franch) with a 4–15 MHz high-frequency linear array probe. The probe, placed in the median sagittal plane of the perineum directly above the anus, was rotated approximately 10°–20° clockwise and counterclockwise to locate the left and right PRM. Once the two-dimensional ultrasound displayed complete bilateral PRM muscle bundles, the RT-SWE program was initiated. Young’s modulus for elastography was set to 180 kPa, and a rectangular sampling frame was used to cover the PRM maximally [14]. The region of interest (ROI) for measuring PRM’s elastic modulus was a 5.0 mm diameter circle. Tissue elasticity was expressed in kilopascals (kPa) using color-coded elastography. The mean elasticity (Emean) of PRM was calculated by repeating measurements three times on the anterior part of the bilateral PRM, as depicted in Fig. 1.

Fig. 1.

Real-time shear wave elastography was used for Measurement of the anterior elastic modulus of both the left and right PRM. (A) left PRM Emean; 16.9 kPa, (B) right PRM Emean: 15.3 kPa. PRM, puborectalis muscle; Emean, mean elasticity; kPa; kilopascal.

Using the UGEO WS80A ultrasound diagnostic instrument (Samsung company, Seoul, Korea), a three-dimensional volume probe, coated with a sterile coupler and covered with a protective sleeve, was positioned in the median sagittal plane of the subject’s perineum. This setup ensured a clear and comprehensive display of the pelvic floor tissue structure. To visualize the LAM located behind the rectum, the probe was rotated 90° clockwise in the transverse plane. The thickness of LAM was then determined by averaging measurements taken three times, as depicted in Fig. 2. Patients were instructed to perform a maximum Valsalva maneuver to evaluate the occurrence of POP. During the examination, caution was exercised to prevent compressing of the bulge, which could potentially yield false negative results. Following clear visualization of the pelvic organs, the probe was stabilized, and the 3D imaging mode was activated to capture volumetric data. Measurements were subsequently obtained for the LH anteroposterior (LHAP) diameter, LH lateral diameter (LHLD), LHA, and LH circumference (LHC) during maximum Valsalva maneuver, as illustrated in Fig. 3.

Fig. 2.

Measurement of the thickness of LAM of bilateral at rest. On the right side of the thickness of LAM including D1, D2, D3; On the left side of the thickness of LAM including D4, D5, D6 (D1: 0.64 cm, D2: 0.65 cm, D3: 0.63 cm; D4: 0.65 cm, D5: 0.64 cm, D6: 0.62 cm). LAM, levator ani muscle.

Fig. 3.

Measurement of LH parameters in maximum Valsalva condition (C1: LHC measurement; A1: LHA measurement; D1: LHLD measurement; D2: LHAP measurement). LH, levator hiatus; LHC, LH circumference; LHA, LH area; LHLD, LH lateral diameter; LHAP, LH anteroposterior.

For sample size calculation, the Power package in R language was employed. Statistical analyses were conducted using SPSS version 21.0 (IBM Corp, Armonk, NY, USA). Count data were represented by the number of cases (n) and analyzed using the Chi-square ($\chi {}^2$) test. The homogeneity of variance and normal distribution were expressed as mean $\pm$ standard deviation (SD) ($\overline{x}$ $\pm$ s). Comparisons between groups were performed using the independent sample t-test, while the paired t-test was used for within-group comparisons. For comparisons involving three or more groups, a one-way analysis of variance (ANOVA) or Welch’s ANOVA test was applied. Spearman’s correlation analysis was utilized to assess relationships between variables. The receiver operating characteristic (ROC) curve was plotted, and the area under the curve (AUC) was calculated. p 0.05 was considered statistically significant.

The general clinical data of all participants included in the study are presented in Table 1. The study comprised a total of 315 subjects. Statistically significant differences were observed in height, weight, and BMI among the study groups (p 0.05). In the early postpartum period, a significant difference in POP was noted between the vaginal delivery group and the cesarean section group (p 0.05). However, no significant difference was found in the duration of postpartum time (p $>$ 0.05). Age did not exhibit significant differences among the groups (p $>$ 0.05).

The study analyzed the thickness of LAM and the elasticity values of the bilateral PRM among the control group, early pregnancy group, middle pregnancy group, and late pregnancy group. This analysis was conducted using one-way ANOVA or Welch’s ANOVA test and the results are detailed in Table 2. The findings revealed that with the progression of gestational weeks, the elasticity values of the bilateral PRM increased, the thickness of LAM decreased. This variation was statistically significant (p 0.05). However, between the control group and the early pregnancy group, no significant difference was observed in terms of the thickness of LAM and the elasticity of the PRM (p $>$ 0.05). Furthermore, the study noted no significant difference in the thickness of LAM and the elasticity of PRM between the left and right sides (p $>$ 0.05).

This section of the study focused on comparing the thickness of LAM and the elasticity values of the PRM among the three study groups: the vaginal delivery group, the cesarean section group, and the control group. The comparative data is presented in Table 3. The findings indicated that the thickness of LAM and the elasticity value of the PRM in primiparas during the early postpartum period were significantly lower compared to those in the control group. Moreover, within the primipara group, those who had undergone vaginal delivery exhibited lower thickness of LAM and elasticity values of the PRM compared to those who had undergone cesarean sections. These differences were statistically significant (p 0.05). Additionally, it was observed that there was no significant difference in the thickness of LAM and the elasticity values of PRM between the left and right sides (p $>$ 0.05).

This part of the study involved Welch’s ANOVA test of LHA, LHLD, LHAP and LHC in primiparas with different delivery modes during the early postpartum period, as demonstrated in Table 4. The results revealed that the LHA, LHLD, LHAP, and LHC values in both the vaginal delivery group and the cesarean section group were significantly higher than those in the control group (p 0.05). Additionally, it was observed that these values were notably higher in the vaginal delivery group compared to the cesarean section group, with this contrast being statistically significant (p 0.05). Spearman correlation analysis revealed a positive correlation between LHA, LHAP, LHLD, LHC, and POP during the early postpartum period among primiparas who underwent vaginal deliveries (r${}_{\text{s}}$ = 0.671, 0.476, 0.318, 0.489, p 0.05), with LHA showing the strongest correlation with POP. The correlation between LHAP and POP was found to be stronger than that between LHLD and POP. Similarly, for primiparas who underwent cesarean sections, a positive correlation was observed between LHA, LHAP, LHLD, LHC, and POP (r${}_{\text{s}}$ = 0.634, 0.539, 0.620, 0.608, p 0.05). with LHA once more demonstrating the strongest correlation with POP. In this instance, it is noteworthy that the correlation between LHLD and POP exhibited greater strength compared to that between LHAP and POP, as detailed in Tables 5,6.

This segment of the study focused on the early postpartum diagnosis of POP in both the vaginal delivery and cesarean section groups. The incidence of POP was significantly higher in the vaginal delivery group at 63.6% (35/55) compared to 40% (22/55) in the cesarean section group, with the difference being statistically significant (p 0.05). ROC curves were plotted for LHA,the thickness of LAM and elasticity of PRM, and their combined use in diagnosing POP within both groups, as depicted in Figs. 4,5. The AUC values, Youden index, cutoff values, sensitivity, and specificity for each of these parameters and their combination were calculated, as presented in Table 7.

Fig. 4.

ROC curves of LHA, elastography of PRM, the thickness of LAM, and their combination for the diagnosis of POP in the vaginal delivery group. ROC, receiver operating characteristic; LHA, LH area; PRM, puborectalis muscle; LAM, levator ani muscle; POP, pelvic organ prolapse.

Fig. 5.

ROC curves of LHA, elastography of PRM, the thickness of LAM, and their combination in the diagnosis of POP in the cesarean section group. ROC, receiver operating characteristic; LHA, LH area; PRM, puborectalis muscle; LAM, levator ani muscle; POP, pelvic organ prolapse.

In the vaginal delivery group, the AUC values for LHA,the thickness of LAM and the elasticity of PRM, and their combination were 0.903, 0.689, 0.706, and 0.944, respectively. Correspondingly, in the cesarean section group, the AUC values were 0.873, 0.691, 0.760, and 0.912, respectively. The combined diagnostic approach yielded the highest AUC in both groups.

The optimal cutoff values established for LHA, elasticity of PRM and the thickness of LAM were 23.945 cm${}^2$, 22.20 kPa, and 0.595 cm in the vaginal delivery group and 19.465 cm${}^2$, 22.45 kPa, and 0.615 cm in the cesarean section group, respectively.

Sensitivity analysis revealed that LHA was not much difference in detecting POP in the vaginal delivery group (85.7%) compared to the cesarean section group (86.4%). and the sensitivity of PRM elasticity was higher in the vaginal delivery group (77.1%) compared to the cesarean section group (59.1%). The thickness of LAM sensitivity was also found to be higher in the vaginal delivery group compared to the cesarean section group (80.0% vs. 59.1%). The combined diagnosis sensitivity of vaginal delivery (88.6%) and cesarean section (100%) was the highest.

In terms of specificity, the vaginal delivery group showed the highest for LHA and its combination (90.0% each), followed by the PRM elastography at 60.0%. In the cesarean section group, PRM elasticity showed the highest specificity (87.9%), followed by LHA (84.8%). The lowest specificity was observed for the combined approach at 69.7%.

Pregnancy represents an independent risk factor for POP, as it leads to increased mechanical load on the pelvic floor and hormonal changes that affect the hardness of the LAM [15]. Pregnancy-induced changes in pelvic floor structure pose a potential risk for postpartum POP. Gachon et al. [16] suggested that women with higher ligament laxity were at higher risk of anal sphincter injury. The increase in pelvic floor muscle hardness during pregnancy is viewed as a protective physiologic mechanism against perineal trauma and muscle injury. This increase in muscle fiber length and change in muscle hardness occur specifically in the pelvic floor muscle group, including PRM, coccygeal muscle, iliococcygeal muscle, and pubococcygeal muscle. When there is an abnormal increase in PRM hardness, it is essential to be alert regarding potential damage of pelvic floor function [5]. The decrease in vaginal wall hardness during pregnancy may be a physiologic process to adapt to vaginal distension during delivery [17]. The increase of pelvic floor muscle hardness and the decrease of vaginal wall stiffness combined protect the pelvic floor structures and avoid muscle injuries during labor.

Our study observed no significant difference in PRM hardness and the thickness of LAM between the left and right sides during pregnancy, suggesting an equal effect of abdominal-pelvic pressure on bilateral LAM. In early pregnancy, the hardness of PRM and the thickness of LAM were similar to the control group, likely due to lower fetal pressure exerted on pelvic support structures, limited uterine enlargement within the pelvic cavity, and cellular changes occurring in pelvic floor muscles and ligaments under hormonal influence. As gestational weeks progressed, there was a significant increase in the hardness of PRM, accompanied by thinning of the LAM. By late pregnancy, PRM hardness peaked at its highest values (L-PRM 41.93 $\pm$ 8.28 kPa, R-PRM 42.96 $\pm$ 9.16 kPa), while the thickness of LAM reached its lowest level (L: 0.55 $\pm$ 0.04 cm, R: 0.56 $\pm$ 0.04 cm). These findings suggest a progressive increase in tension within the pelvic floor muscles and supporting structures, necessary to accommodate the growing abdominal and pelvic cavity pressure. The mechanical strain on the uterus leads to a consequential impact on the LAM, consistent with the findings by Shafik et al. [18], who noted a gradual increase in LAM activity from 12 to 36 weeks of gestation.

The POP of 50 late pregnant women in this study was followed up for 42–45 days after delivery. Among 20 pregnant women with PRM elastic modulus surpassing the cutoff value of 43.45 kPa, the positive rate of POP was 85% (17/20). In contrast, among the remaining 30 cases whose average elastic modulus was lower than 43.45 kPa, the positive rate of POP was 26.7% (8/30) ($\chi {}^2$ = 16.333, p 0.001). This suggest that PRM elasticity during pregnancy can serve as a risk factor for early postpartum POP. Kruger et al. [15] proposed that excessive LAM hardness during pregnancy could make LAM more susceptible to damage during delivery. Therefore, it is crucial to closely examine abnormal changes in PRM stiffness during pregnancy to detect possible prenatal LAM damage early. This study underscores the importance of careful examination during pregnancy, especially when abnormal PRM stiffness is observed. This study highlights the need for personalized clinical recommendations for preventing postpartum PFD and to develop relevant plans, including episiotomy or choice of delivery mode to mitigate PFD such as perineal tears, POP, and stress urinary incontinence.

In this study, the hardness of bilateral PRM decreased and the LAM became thinner in the early postpartum period for primiparas. The change in PRM hardness and LAM thinness was more pronounced in vaginal deliveries compared to cesarean deliveries, indicating that different delivery modes exert varying impacts on pelvic floor muscles.

During vaginal delivery, the fetal head directly acts on the birth canal, resulting in the thinning and deformation of LAM after extreme stretching. Exceeding the maximum pressure threshold of the pelvic floor structure can result in irreversible damage or even tear of pelvic floor muscles and ligaments, ultimately leading to PFD. In addition, local ischemia and hypoxia occur after the vessels and nerves that feed LAM are affected during delivery, leading to reduced contractility of LAM. Xie et al. [19] suggested that changes in LAM function may occur earlier than structural abnormalities in POP. Vaginal delivery itself can impact the hardness of LAM. During the fetal head descent, the perineal distension peaks, causing the internal structure of LAM (e.g., the number and type of muscle fibers, the spatial arrangement of muscle bundles, and the actin-myosin isomers) to be irreversibly damaged. The biomechanical properties of LAM undergo changes, and its ability to resist abdominal pressure is weakened after Valsalva maneuver. This results in the occurrence of POP [19]. Studies have shown that even if 2D ultrasound does not detect obvious LAM damage, changes in the biomechanical properties of LAM will also affect the size of the LH, the expansibility, and the severity of POP [20].

In analyzing the PRM hardness and the thickness of LAM of the cesarean section patients, it was found that the PRM hardness of these patients was lower compared to healthy nulliparous women, and the LAM was thinner. The changes of PRM hardness and LAM thickness of patients who underwent cesarean section are mainly affected by pregnancy, and the delivery process has little effect on pelvic floor muscles. Cesarean section appears to offer some protective effect on the elasticity and contractile function of the PRM, whereas vaginal delivery causes more noticeable damage to the elasticity and contractile function of the PRM.

The rate of POP was found to be higher in the vaginal delivery group (63.4%) compared to the cesarean section group (40%). This suggests that vaginal delivery is more likely to result in POP. Béchard et al. [21] proposed that the mode of delivery significantly correlated with early postpartum POP. They found that primiparas who delivered vaginally had a prevalence of POP 5 times higher than that of those with cesarean delivery, with the symptoms affecting the anterior, middle, and posterior compartments, and a high proportion of urinary symptoms.

LH parameters can effectively evaluate postpartum pelvic floor function injury. In this study, all LH parameters of primiparas in the early postpartum period were higher than those of the healthy nulliparous group. For women who delivered vaginally, LH parameters were higher than those who delivered via cesarean section. The findings of this study suggest that the enlargement of the uterus and fetal growth exert mechanical pression on the pelvic floor structure, causing stretching, deformation, and relaxation of pelvic floor muscles and ligaments. The increase in progesterone and serum relaxin during pregnancy induces relaxation of pelvic floor muscle fibers and a decline in their contraction function, resulting in LH dilation.

The occurrence of POP and LH expansion in cesarean section is mainly affected by the increase of abdominal pelvic pressure and LH expansion during pregnancy. However, since the fetus does not pass through the birth canal and does not directly cause crush injury to the pelvic floor structures during cesarean delivery, the pelvic organ downward displacement distance and the degree of LH expansion are typically less than those in a vaginal delivery.

Women who delivered vaginally not only face increased mechanical load during pregnancy, but also endure mechanical compression from abdominal pressure and fetal head descent during delivery. This leads to extreme swelling of the perineum and potential injury or tearing of the pelvic floor muscles. During postpartum, due to irreversible tensile deformation of LAM, the shape of the LH changes, reducing support for pelvic organs and increasing the likelihood of POP. Therefore, symptoms of POP and LH expansion tend to be more pronounced in women with vaginal delivery than in those with cesarean delivery.

In both vaginal delivery and cesarean delivery primiparas during the early postpartum period, parameters such as LHA, LHAP, LHLD, LHC showed a positive correlation with POP. The correlation between LHA and POP in the two groups was the highest, and the correlation between LHA and POP in the vaginal delivery group was higher than that in the cesarean delivery group (0.671/0.634). This is consistent with the conclusion of Handa et al. [22]. In women with a vaginal delivery, the LAM relaxation expands to both sides, and the LH expansion reduces the support capacity of the pelvic organs. In addition, in women who undergo a vaginal delivery, the Valsalva maneuver tends to be more effective, and the sensitivity of POP increases when the abdominopelvic pressure increases. When the fetus bypasses the birth canal via cesarean section, the damage to LAM and pelvic floor structure is less severe, resulting in a lower correlation between LH parameters and POP than in a vaginal delivery. The current study found that LH parameters and muscle strength are potentially important markers for identifying women at higher risk of prolapse. Future research is still needed to determine whether interventions aimed at reducing hiatus size or increasing muscle strength could be effective preventive strategies to reduce the incidence of POP.

This study conducted a detailed analysis of the ROC curve and the diagnostic efficacy of LHA, PRM Elasticity, the thickness of LAM, and their combination in diagnosing POP in both vaginal delivery and cesarean section groups. It was found that the AUC for LHA in the vaginal delivery group was higher than in the cesarean section group (0.903 vs. 0.873). The thickness of LAM showed no significant difference between the two groups (0.689 vs. 0.691). The highest AUC was seen in the combined diagnosis of the three parameters (0.944 in the vaginal delivery group and 0.912 in the cesarean section group). Sensitivity was highest in the combined diagnosis approach (88.6% in the vaginal delivery group, and 100% in the cesarean section group).

Vaginal delivery leads to more evident dilation of the LH and weaker pelvic floor muscle strength, increasing the likelihood of POP compared to cesarean section. Although cesarean section theoretically protects the pelvic floor tissue structures, it entails iatrogenic trauma to the abdomen and uterus. The incision itself can negatively impact pelvic floor function and limit postpartum activities. Furthermore, complications such as incision infection, scar tissue formation at the incision site, and endometriosis can hinder the recovery of pelvic floor function.

Béchard et al. [21] found that 12 months after delivery, the pelvic floor function and quality of life of women who had delivered vaginally or by cesarean section were evaluated. They found that women who delivered vaginally scored higher than those who delivered by cesarean section (56:55), indicating a better quality of life. Schiavi et al. [23] reported that following natural vaginal tissue repair treatment in postpartum POP patients, improvements were observed in scores related to prolapse quality of life, pelvic floor disability index and pelvic floor impact questionnaire. These changes demonstrated that the quality of life was significantly improved after treatment, and the proportion of patients with regular sexual activity increased. Dyspareunia was reduced, indicating that those receiving the necessary reparative treatment after childbirth can significantly improve their quality of life.

In this study, TP-3DUS combined with RT-SWE was used to quantitatively and qualitatively evaluate the pelvic floor function of pregnant patients and primiparas in the early postpartum period. This evaluation encompassed parameters such as pelvic floor muscle strength, LAM thickness, LH size, POP and downward displacement distance. The aim was to provide exploratory data for clinical practice on the changes of PRM elasticity and LAM thickness during pregnancy. This data may help clinical identification of high-risk parturients, assisting in the guidance of delivery mode selection. It may also provide personalized suggestions for the prevention of postpartum PFD, such as considering episiotomy, choosing appropriate delivery mode, avoid the occurrence of perineal tears, POP, Stress Urinary Incontinence (SUI) and other forms of PFD. We evaluated the efficacy of PRM elasticity, LHA, the thickness of LAM, and the combination of the three parameters in the diagnosis of early postpartum POP by comparing the differences of pelvic floor structure and pelvic floor muscle stiffness in primiparas with different delivery modes, so as to provide personalized suggestions for the treatment of early postpartum POP.

• At present, there are few research reports on pelvic floor function during pregnancy. In this study, quantitative analysis of pelvic floor muscles during pregnancy was performed to provide clinical exploratory data on the changes of PRM elasticity and LAM thickness during pregnancy.

• In this study, the stiffness and thickness of pelvic floor muscles, the LHA and the combined diagnosis of POP were analyzed in the early postpartum period of women with vaginal delivery and cesarean section. By detecting the changes of PRM elasticity and the thickness of LAM, the adverse factors of poor cooperating women with Valsalva maneuver and the influence of subjective factors of examination were effectively compensated.

• The stiffness and thickness of pelvic floor muscle were only analyzed in the first-, second- and third-trimester of pregnancy, and the specific gestational weeks were not available.

• No prolonged follow-up.

• The pelvic floor function after pelvic floor rehabilitation training was not evaluated.

• Longitudinal studies are recommended to assess PRM hardness during pregnancy and to refine the gestational weeks for better prediction of prenatal LAM damage.

• Investigation into PFD-related genes could provide genetic insights into the occurrence of POP.

• Explore the potential relationship between genetic factors and ultrasound indicators to enhance understanding and diagnosis of POP.