Abstract
Aim
To identify the anatomical morphology of levator ani (LA) in primigravidae at term pregnancy and its natural process of changing after delivery.
Methods
Forty‐one primigravidae (vaginal delivery: 29 women, cesarean delivery in the first stage of labor: 12 women) underwent magnetic resonance imaging (MRI) at full‐term pregnancy, 6 weeks and 10 months postpartum. Three‐dimensional (3‐D) model of LA created from MRI data using Mimics v.21.0 software and source images were assessed to determine the morphology. LA volume (LVOL) was calculated and used as indicator of muscle atrophy.
Results
Decrease of levator hiatus length (LH‐L) was shown in both groups since 6 weeks postpartum. In the vaginal delivery group, the differences in LVOL between time points were significant (p < 0.05), showing a persistent decreasing tendency. Puborectalis attachment width (PAW) on the left was the smallest at 6 weeks postpartum (p < 0.05). LA avulsion and significant 2‐D morphological change after delivery were only observed in this group (p < 0.05); In the cesarean section group, smaller LVOL was found at 6 weeks postpartum comparing with full‐term pregnancy (p < 0.05); Larger levator‐symphysis gap (LSG) and levator hiatus width (LH‐W), smaller PAW were observed in vaginal delivery group comparing with cesarean section group at 6 weeks postpartum (p < 0.05), but none of the values exhibited between‐group differences (p > 0.05) at 10 months postpartum. No other comparisons were considered significant (p >0.05).
Conclusions
Vaginal delivery, or even active labor itself may both lead to LA atrophy. And the morphology of LA is basically similar in different delivery modes at 10 months postpartum once the onset of labor has occurred, even though it changes more complicatedly after vaginal delivery.
Keywords: Levator ani, Magnetic resonance imaging, Pelvic floor anatomy, Primigravida
Introduction
Magnetic resonance imaging (MRI) is one of the optimal choices in studying the morphology of pelvic floor structure due to its high soft‐tissue contrast. Strohbehn et al 1 reported a good correlation between cadaveric dissection and MRI findings. The pelvic floor structures sustain oppression and traction during the period of pregnancy and delivery, which is a well‐known factor for the injury of pelvic floor and the development of pelvic floor dysfunction (PFD) later in life. 2 , 3 , 4
As a crucial component of pelvic floor, Levator ani (LA) composes of pubococcygeus, puborectalis, pubovaginalis and iliococcygeus. Previous MRI study detected LA tears 6 months after vaginal delivery in primiparas. 5 Another descriptive study found LA injury 2–3 days postpartum in primiparous women underwent elective cesarean section or urgent cesarean section after the onset of labor, but the latter had 2.7 times more abnormalities than the former. 6 While those studies described the appearances and abnormalities of LA caused by pregnancy and delivery in 2‐D image, it was not easy to disclose the real 3‐D morphology which may be at the root of the defects that lead to clinical pelvic floor pathology. There were studies that used computer‐assisted 3‐D MR‐based models to correlate imaging findings with the presence of clinical pelvic floor dysfunction in living women. 7 And Muriel K et al 8 assessed the appearance of the LA in postterm nulliparas by 3‐D MRI. However, none of the previous studies have used this technique to evaluate changes of LA morphology before and after delivery. In addition, previous Ultrasonic study revealed that there were no differences considering hiatal dimensions between women with different delivery modes, 9 whereas others found that women with a spontaneous vaginal delivery had a significantly larger hiatal area than those delivered by cesarean section even 2 years postpartum. 10 So controversial issues remain in this field.
Therefore, based on our published results and expriences of establishing 3‐D MRI model to describe female pelvis in a large population of Chinese females at term pregnancy, 11 we assessed the morphology of LA before childbirth and its nature process of changing after normal vaginal delivery or cesarean section in the first stage of labor. Further comparisons were made between delivery modes.
Materials and methods
Study population
This is a prospective observational study that is based on a parent study on pelvis and pelvic floor in primigravidas with singleton and cephalic presentation using MRI originated 3‐D models. This study was conducted from January 2014 to January 2017 at Southern Medical University Nanfang Hospital and approved by the Ethics Committee of the hospital (approval number:032; 2013), with clinical research registration number of ChiCRT‐DDD‐1600975. Written informed consent were obtained before data collection. Primigravidas from the parent study who were scheduled to deliver in our hospital and willing to fully participate in postnatal follow‐up were screened for eligibility during the time of prenatal visit at full‐term pregnancy. The participants were invited for MRI scan at 37–41 weeks' pregnancy, 6 weeks and 10 months postpartum. Forty‐one women finally completed the entire study. Exclusion criteria at the time of participants screening included maternal age younger than 18 or older than 35 years, contraindications to MRI examination, maternal medical complications, medical history of urinary incontinece, genital or rectal prolapse before pregnancy, pelvic floor surgery. After delivery, the clinical data were assessed and further exclusion criteria included fetal weight >4000 or <2500 g, severe perineal laceration, prolonged second stage of labor, instrumental delivery, cesarean section without trailing of labor or in second stage of labor. Those women who were lost to follow‐up, unable to undergo MRI scan as scheduled, delivered in other units for special reasons or pregnant again were also excluded.
Available MRI data
MRI was performed with Turbo spin‐echo (TSE) thin‐layer axis for pelvis and LA. A 1.5 T GE Signa magnet (Signa EXCITE HD System, GE Company, American), a torso coil, and an abdominal coil were used. The scanning parameters for TSE included the following: 1.5 T MR TR/TE, 10040/48 ms; slice thickness, 3 mm; inter layer spacing, 0; array, 192 × 288; feld of view, 38 × 38 cm; bandwidth, default; number of excitations, 3; 3.0 T TR/TE, 3000/102 ms; slice thickness, 3 mm; interlayer spacing, 0; array, 512 × 512; feld of view, 34 × 34 cm; bandwidth, default; number of excitations, 4.
Parameter measurement methods and generation of three‐dimensional reconstruction models
The MRI images were saved in the digital imaging and communication in medicine (DICOM) format and imputed into Mimics 21.0 software (Interactive Medial Image Control System, Version 21.0, Materialize Company, Belgium) for 3‐D reconstruction and measurement (Supplementary Figure S1). Alignment images were also measured by the same software to determine the 2‐D Parameters. The contours of LA, as well as pelvis were identified and manually segmented in consecutive axial image layer by layer (Supplementary Figure S2a). The 3‐D models were then calculated and reconstructed in the software using a surface‐rendering algorithm as it was also mentioned in our previous study, 11 , 12 and the levator ani volume (LVOL) could be calculated automatically (Figure S2b,c). Angle measurement tool was selected from the toolbar to measure angular parameters (The cursor will change to a measuring instrument. Click three points with the left mouse button to define the angle. The angle is expressed in degree and shown on the screen) (Supplementary Figure S3). Distance measurement tool was selected from the toolbar to measure linear parameters like distance, length, width, or thickness (Click once the left mouse button to set the first point and click again to fix the other end of the measurement. The distance is expressed in millimeter and shown on the screen) (Supplementary Figure S4). This procedure was performed by a trained gynecologist and senior experienced radiologists who were blinded to all other data.
3‐D MRI parameters
The levator‐symphysis gap and puborectalis attachment width
The levator‐symphysis gap (LSG) was measured as the distance from the middle of the inferior symphysis to the nearest aspect of the pubovisceral muscles on the right and left. The puborectalis attachment width (PAW) is defined as the width of origin on the inner surface of pubic bone of the levator sling on the right and left (Figure 1a,b). 13
FIGURE 1.
Measurements of the levator ani on three‐dimensional model in anterior–posterior top view. The LSG (a) and PAW (b) are shown by the green lines with double arrows on the left and the black lines with double arrows on the right. LSG, Levator‐symphysis gap; PAW, puborectalis attachment width; S, Symphysis.
The levator plate angle
The levator plate angle (LPA) is defined as the angle formed between levator plate and the plate that is perpendicular to the longitudinal axis of the body (Figure 2a). 14
FIGURE 2.
Measurements of the levator ani on three‐dimensional model in different views. (a) The LPA is shown in right lateral view by the blue arrow with most of the pelvis bone removed. The single dotted line represents the plate that is perpendicular to the longitudinal axis of the body (the double dotted line). (b) The ICW is shown in a coronal view by the green lines with double arrows on the left and the black lines with double arrows on the right. (c) The ICA is shown in anterior–posterior top view by the green arrow on the left and the black arrow on the right with most of the pelvis bone removed. (d) The LH‐L and LH‐W are shown in a dorsal lithotomy view by the black lines with double arrows and the green lines with double arrows, respectively. ICA, Iliococcygeus angle; ICW, Iliococcygeus width; LH‐L, Levator hiatus length; LPA, Levator plate angle; LH‐W, Levator hiatus width; S, Symphysis.
The width of the iliococcygeus and The iliococcygeal angle
The width of the iliococcygeus (ICW) was measured on both sides at the level of the ischial spines in a coronal view of the generated model. The iliococcygeal angle (ICA) is the angle between the direction of the interspinous diameter and the iliococcygeal fibers in the axial plane, also measured bilaterally (Figure 2b,c). 15 , 16
The levator hiatus
The levator hiatus length (LH‐L) was measured from the inferior border of symphysis to the curve of the LA. The levator hiatus width (LH‐W) is defined as the widest transverse diameter of the levator hiatus (Figure 2d). 16
2‐D MRI parameters
The puborectalis and iliococcygeus thickness
The puborectalis thickness (PRT) and iliococcygeus thickness (ICT) were measured bilaterally in the the middle part of the muscles, at the level of the inferior border of the pubic symphysis in the axial plane, and mid‐rectum level in the coronal plane respectively (Figure 3a,b). 17
FIGURE 3.
Measurements of muscle thickness, appearances of Levator ani, and example of Levator ani avulsion on two‐dimensional images. (a) Coronal plane at the level of mid rectum: ICT (white line). (b) Axial plane at the level of the inferior border of the symphysis: PRT (white line). Avulsion of LA on both sides (white arrowhead) in the coronal plane (c) and axial plane (d). (e) LA with an appearance of “V.” (f) LA with an appearance of “U.” (g) IC reflected as a flat shape (white arrows). (h) IC reflected as a convex shape (black arrows). IC, Iliococcygeus; ICT, Iliococcygeus thickness; LA, Levator ani; PRT, Puborectalis thickness; R, Rectum; S, Symphysis; U, Urethra; V, Vagina.
The appearance and avulsion of levator ani
We classified the appearance of LA into two different types, which were referred to “V” and “U” shape in the axial plane. In the sagittal plane, iliococcygeus is defined as “convex” shape, which manifests as convex upward surfaces of the muscle, or “flat” shape that possesses more flattened ones. LA avulsion was determined if the puborectalis bulk lost completely at the site of its arising from the pubic bone either one or both sides. Final determination of levator avulsion was made only when an abnormality was found in both the axial and the coronal planes and was agreed by two examiners (Figure 3c–h). 8 , 13
Statistical analysis
SPSS22.0 was used for the data analysis. Demographic data and MRI measurements were reported as means and standard deviation or frequencies and percentages. One‐way Repeated Measures Anova or Friedman test were used to compare the changes of MRI measurements progressing from full‐term pregnancy to 10 months postpartum within the two groups after normality of the data was assessed. Bonferroni post hoc tests was used to conduct further multiple comparisons. McNemar tests was used to compare the appearance and LA avulsion between observation points. Independent samples t tests and Chi‐square tests were used to compare the MRI values and appearances of the LA between the two groups at different time points. p < 0.05 was considered statistically significant.
Results
From the beginning, 107 primigravidas from the parent study were screened for eligibility. After selection, 95 primigravidas underwent MRI scan and were considered suitable to receive reassessment once delivered, of whom 41 completed the entire study (29 delivered vaginally, 12 delivered by cesarean section in the first stage of labor). The process of study cohort selection is shown in Figure 4. No significant differences regarding maternal age, BMI, and fetal weight were found between delivery modes (vaginal vs. cesarean delivery: 27.83 ± 2.54: 28.00 ± 2.76 years, p = 0.848; 21.91 ± 2.14: 23.16 ± 3.31 kg/m2, p = 0.156; 3290 ± 320: 3410 ± 470 g, p = 0.806).
FIGURE 4.
Flowchart of study participants.
For woman delivered vaginally, a decrease of LH‐L was observed after 6 weeks postpartum while LH‐W remained unchanged throughout the observation time. LPA decreased after delivery but was insufficient to be statistically significant until 10 months postpartum. A steady and significant decrease of LVOL was also found from full‐term pregnancy to 10 months postpartum. No signs of difference were shown considering puborectalis and iliococcygeus thickness on both sides. An enlargement of LSG on the left side was seen 6 weeks after delivery followed by a decrease since then which was statistically significant on the whole, but failed to showed any within‐group difference after further multiple comparisons and Bonferroni adjustments were conducted. The PAW on the left at 6 weeks postpartum was the lowest by comparison with full‐term pregnancy and 10 months postpartum. PAW and LSG on the right remained unchanged. No difference was observed regarding ICW and ICA (Table 1).
TABLE 1.
Comparison of absolute values of levator ani measurements between different time points in women delivered vaginally
| N = 29 | p value | |||
|---|---|---|---|---|
| Full‐term pregnancy | 6 weeks postpartum | 10 months postpartum | ||
| (mean ± SD) | (mean ± SD) | (mean ± SD) | ||
| Three‐dimensional measurements | ||||
| LSG‐R (mm) | 19.50 ± 3.17 | 19.75 ± 3.20 | 19.42 ± 3.04 | 0.855 |
| LSG‐L (mm) | 18.73 ± 2.99 | 20.32 ± 3.06 | 19.49 ± 3.11 | 0.044* |
| PAW‐L (mm) | 11.29 ± 3.37 | 9.17 ± 3.77** | 10.88 ± 3.51*** | 0.001* |
| PAW‐R (mm) | 12.56 ± 3.14 | 11.57 ± 2.73 | 12.05 ± 2.63 | 0.103 |
| LPA (°) | 50.7 ± 8.67 | 47.42 ± 9.43 | 45.00 ± 7.71** | 0.009* |
| LVOL (cm3) | 34.47 ± 7.14 | 31.95 ± 6.84** | 28.86 ± 5.62**,*** | <0.001* |
| ICW‐L (mm) | 50.87 ± 6.82 | 50.52 ± 5.07 | 51.49 ± 6.15 | 0.902 |
| ICW‐R (mm) | 48.75 ± 5.49 | 50.02 ± 3.95 | 48.71 ± 4.40 | 0.354 |
| ICA‐L (°) | 35.15 ± 7.60 | 36.31 ± 8.23 | 33.55 ± 8.58 | 0.170 |
| ICA‐R (°) | 36.82 ± 8.33 | 36.78 ± 7.21 | 35.32 ± 8.00 | 0.506 |
| LH‐W (mm) | 34.4 ± 3.77 | 36.07 ± 2.61 | 35.19 ± 3.08 | 0.063 |
| LH‐L (mm) | 59.52 ± 6.49 | 59.34 ± 3.06 | 54.83 ± 5.60**,*** | <0.001* |
| Two‐dimensional measurements | ||||
| PRT‐L (mm) | 7.37 ± 1.21 | 7.59 ± 1.13 | 7.04 ± 1.19 | 0.153 |
| PRT‐R (mm) | 6.82 ± 1.41 | 6.59 ± 0.96 | 6.72 ± 1.28 | 0.660 |
| ICT‐L (mm) | 3.38 ± 0.78 | 3.06 ± 0.098 | 3.20 ± 1.08 | 0.061 |
| ICT‐R (mm) | 3.17 ± 1.15 | 2.85 ± 1.12 | 2.94 ± 1.14 | 0.279 |
Note: N value is the number of women analyzed. Data presented as mean ± SD for all the measurements. *p < 0.05 significant for overall comparison. **p < 0.05 vs. full‐term pregnancy. ***p < 0.05 vs. 6 weeks postpartum.
Abbreviations: ICA‐L and ICA‐R, left and right iliococcygeus angle; ICT‐L and ICT‐R, left and right iliococcygeal thickness; LH‐L, levator hiatus length; LH‐W, levator hiatus width; LPA, levator plate angle; LSG‐L and LSG‐R, left and right Levator‐symphysis gap; LVOL, levtor ani volume; ICW‐L and ICW‐R, left and right iliococcygeus width; PAW‐L and PAW‐R, left and right puborectalis attachment width; PRT‐L and PRT‐R, left and right puborectalis thickness.
For woman delivered by cesarean section, LH‐L measured at 10 months postpartum was relatively smaller by comparison with full‐term pregnancy while LH‐W remained unchanged all the time. The change of LPA was not statistically significant after Bonferroni adjustment, even though it decreased about 8° after delivery and kept basically the same level henceforth, showing significant overall difference. The decrease of LVOL was considered statistically significant only between full‐term pregnancy and 6 weeks postpartum. No difference was found regarding ICW, ICA, LSG, PAW, as well as muscle thickness on both sides (Table 2).
TABLE 2.
Comparison of absolute values of levator ani measurements between different time points in women delivered by cesarean section
| N = 12 | p value | |||
|---|---|---|---|---|
| Full‐term pregnancy | 6 weeks postpartum | 10 months postpartum | ||
| (mean ± SD) | (mean ± SD) | (mean ± SD) | ||
| Three‐dimensional measurements | ||||
| LSG‐R (mm) | 18.20 ± 1.61 | 19.62 ± 2.00 | 19.51 ± 1.23 | 0.088 |
| LSG‐L (mm) | 18.50 ± 2.51 | 18.86 ± 1.49 | 19.56 ± 2.58 | 0.455 |
| PAW‐L (mm) | 11.59 ± 2.20 | 11.85 ± 1.76 | 12.30 ± 2.40 | 0.779 |
| PAW‐R (mm) | 12.26 ± 2.06 | 11.63 ± 1.91 | 11.84 ± 2.28 | 0.642 |
| LPA (°) | 57.63 ± 9.95 | 49.11 ± 5.19 | 50.08 ± 8.89 | 0.014* |
| LVOL (cm3) | 34.90 ± 6.73 | 30.27 ± 5.93** | 28.01 ± 4.57** | <0.001* |
| ICW‐L (mm) | 52.08 ± 7.82 | 48.28 ± 8.24 | 50.58 ± 6.30 | 0.123 |
| ICW‐R (mm) | 48.50 ± 4.67 | 48.98 ± 3.33 | 49.49 ± 5.44 | 0.733 |
| ICA‐L (°) | 39.69 ± 10.86 | 39.54 ± 10.07 | 34.77 ± 9.00 | 0.137 |
| ICA‐R (°) | 41.71 ± 8.13 | 39.74 ± 10.61 | 35.44 ± 8.46 | 0.139 |
| LH‐W (mm) | 31.83 ± 2.55 | 33.931 ± 2.19 | 33.57 ± 3.53 | 0.060 |
| LH‐L (mm) | 59.12 ± 7.55 | 57.33 ± 4.47 | 53.84 ± 7.08** | 0.027* |
| Two‐dimensional measurements | ||||
| PRT‐L (mm) | 7.16 ± 0.51 | 7.27 ± 1.08 | 6.98 ± 1.23 | 0.779 |
| PRT‐R (mm) | 7.21 ± 1.00 | 6.72 ± 1.08 | 6.92 ± 0.75 | 0.112 |
| ICT‐L (mm) | 3.13 ± 0.65 | 2.95 ± 0.93 | 3.01 ± 0.68 | 0.637 |
| ICT‐R (mm) | 3.49 ± 0.84 | 3.32 ± 0.51 | 3.24 ± 0.65 | 0.437 |
Note: N value is the number of women analyzed. Data presented as mean ± SD for all the measurements. *p < 0.05 significant for overall comparison. **p < 0.05 vs. full‐term pregnancy. ***p < 0.05 vs. 6 weeks postpartum.
Abbreviations: ICA‐L and ICA‐R, left and right iliococcygeus angle; ICT‐L and ICT‐R, left and right iliococcygeal thickness; ICW‐L and ICW‐R, left and right iliococcygeus width; LH‐W, levator hiatus width; LH‐L, levator hiatus length; LPA, levator plate angle; LSG‐L and LSG‐R, left and right Levator‐symphysis gap; LVOL, levtor ani volume; PAW‐L and PAW‐R, left and right puborectalis attachment width; PRT‐L and PRT‐R, left and right puborectalis thickness.
No LA avulsion was observed at full‐term pregnancy for all the 41 women or at any postpartum time points in cesarean section group, while it was 13.79% (4/29) at 6 weeks postpartum and 6.90% (2/29) at 10 months postpartum in vaginal delivery group with a significant within‐group difference (p = 0.015). LA shape was characterized as “U” in 36.59% and “V” in 63.41% at full‐term pregnancy, and no signs of difference were shown between time points regardless of delivery modes. The iliococcygeus manifested as “convex” shape in 75.61% of all women before delivery (31/41), while in vaginal delivery group it was 72.41% (21/29) at full‐term pregnancy, 41.83% (12/29) at 6 weeks postpartum and 75.86% (22/29) at 10 months postpartum respectively, showing significant difference between the latter two time points (P = 0.002). No difference was found when referring to cesarean delivery group (Table 3).
TABLE 3.
Comparison of appearance and avulsion of Levtor ani between different time points in women delivered vaginally and by cesarean section
| Vaginal delivery (N = 29) | Cesarean delivery (N = 12) | |||||||
|---|---|---|---|---|---|---|---|---|
| 6 weeks postpartum | p | 6 weeks postpartum | p | |||||
| Iliococcygeus appearance | Flat | Convex | Flat | Convex | ||||
| Full‐term pregnancy | Flat | 2 (6.90%) | 6 (20.69%) | 0.289 | 0 (0.00%) | 2 (16.67%) | 0.063 | |
| Convex | 15 (51.72%) | 6 (20.69%) | 5 (41.67%) | 5 (41.67%) | ||||
| 10 months postpartum | Flat | 7(24.14%) | 0 (0.00%) | 0.002* | 3 (25.00%) | 0 (0.00%) | 0.500 | |
| Convex | 10 (34.48%) | 12 (41.38%) | 2 (16.67%) | 7 (58.33%) | ||||
| Levtor ani appearance | V | U | V | U | ||||
| Full‐term pregnancy | V | 16 (55.17%) | 2 (6.90%) | 0.289 | 8 (66.67%) | 0 (0.00%) | 1.000 | |
| U | 6 (6.90%) | 5 (17.24%) | 1 (8.33%) | 3 (25.00%) | ||||
| 10 months postpartum | V | 21 (72.41%) | 5 (17.24%) | 0.219 | 9 (75.00%) | 0 (0.00%) | 1.000 | |
| U | 1 (3.45%) | 2 (6.90%) | 0 (0.00%) | 3 (25.00%) | ||||
| Avulsion of Levtor ani | No | Yes | No | Yes | ||||
| Full‐term pregnancy | No | 25 (86.21%) | 4 (13.79%) | 0.125 | 12 (100.00%) | 0 (0.00%) | – | |
| Yes | 0 (0.00%) | 0 (0.00%) | 0 (0.00%) | 0 (0.00%) | ||||
| 10 months postpartum | No | 25 (86.20%) | 2 (6.90%) | 0.015* | 12 (100.00%) | 0 (0.00%) | – | |
| Yes | 0 (0.00%) | 2 (6.90%) | 0 (0.00%) | 0 (0.00%) | ||||
Note: *Significant at 0.05 level (two‐tailed).
There were a few significant between‐group differences in LA values at 6 weeks postpartum. Except for noticeable wider LH‐W, lager LSG as well as smaller PAW both on the left side were observed in vaginal delivery group by comparison with cesarean section group (LH‐W: p = 0.017; LSG: p = 0.047; PAW: p = 0.024). By the time of 10 months postpartum, none of the values exhibited any between‐group differences (p > 0.05). Classifications of muscle appearance did not differ between the two groups after delivery regardless of observation times (p > 0.05).
Discussion
To our knowledge, this is the first study that used MRI‐based 3‐D reconstruction model to prospectively evaluate the effects of first pregnancy and childbirth on LA morphology, both qualitative and quantitative morphometrics were included in the analysis. There are some key results of our study. First, decreased LVOL was observed in both groups after childbirth. Second, even though vaginal delivery has more complicated impacts on LA than cesarean section in the first stage of labor, there were no between group differences considering any of the morphological parameters at 10 months postpartum.
The LA is a single muscle composing of different functional and morphological components, with the supportive component that is horizontally oriented and the sphincteric component that is vertically oriented. 18 Our study mainly focused on the aspect of anatomical morphology which can be better assessed with the help of MRI. The LPA in both groups became less steeper at 10 months postpartum than before, and thus the support provided for pelvic organs to be in the right positions at vertical orientation was enhanced. In addition, while difference existed at 6 weeks postpartum between the two groups considering measurement of levator hiatus, it did not persist till 10 months, at which time none of the other between‐group comparisons of values or morphological descriptions were considered statistically different, except for LA avulsion which was only seen in vaginal delivery group. And the ICW and ICA were relatively unchanged after delivery regardless of delivery mode. As a result, it means that vaginal delivery is likely to have a relatively similar impact on LA morphology after self‐recovery comparing with cesarean section in the first stage of labor.
Moreover, 3‐D MRI parameters can be used as quantitative indicators in evaluating LA injury. Previous study found greater LSG and smaller PAW in women with LA avulsion compared with women with intact LA. 13 And we also observed enlargement of LSG and decrease of PAW at 6 weeks postpartum in woman delivered vaginally, which recovered to antepartum level at 10 months postpartum. And it's worth mentioning that the changes were only observed on the left in vaginal delivery group. As a result, the left arm of levator sling is possibly more likely to sustain injury caused by vaginal delivery. We speculate this might be correlated with the mechanism of labor as left occiput anterior (LOA) is the most common type of fetal position during engagement and descent. Therefore the left arm of levator sling sustains more mechanical stress and provides more strength to facilitate the process of internal rotation when the fetal head reach pelvic floor.
On the other hand, under the help of 3‐D MRI reconstruction software, we were able to calculate LVOL. As we did not find any of the previous relevant ultrasonic studies that had explore volume of pelvic muscles, a few 3‐D MRI studies discussed the differences of LVOL between different populations as well as its clinical significance. Vikky C et al calculated the volume of levator ani and obturator internus to investigate whether there were any evidences of significant age‐related muscle atrophy in nulliparous women. 19 Lennox Hoyte at el 7 indicated that the volume was smaller in women with prolapse and genuine stress incontinence comparing with asymptomatic volunteers. Therefore, it is likely that decrease in LVOL plays a role in the development of PFDs. But none of the others explored its changing process correlated with pregnancy and delivery. Decreased volume was found in our study both at the time of 6 weeks and 10 months after vaginal delivery comparing with full‐term pregnancy, which was considered as an indication of LA atrophy. Furthermore, woman delivered by cesarean section in the first stage of labor sustained LA atrophy as well. MRI study by Sebastien Novellas et al also suggests that active labor during the time of a cesarean induces early lesions of the pelvic floor muscles. 6 As most of the previous studies revealed that the recovery of pelvic floor after uncomplicated vaginal delivery occurred during the the first months postpartum, 20 , 21 and the LVOL did not show any signs of restoration at 10 months postpartum regardless of delivery mode in our study, we speculate that LA atrophy is probably irreversible. The etiology of those findings might be explained by neurophysiological and anatomical studies which have repeatedly demonstrated that vaginal childbirth, or even just the attempt at vaginal delivery may have a negative effect on nervous structures supplying the pelvic floor muscles. 22 , 23 , 24 And the impairment of nerves may further lead to LA atrophy, which is confirmed by a experimental study conducted by Lisa M. Pierce et al 25 that performing bilateral transection of the LA nerve caused LA atrophy in female squirrel monkey. While previous Ultrasonic study used muscle thickness to make quantitative assessment of LA atrophy after delivery, they also mentioned that evaluation of the condition would likely require MRI. 26 In our study, The thickness of puborectalis and iliococcygeus were measured followed by conducting within‐group and between group comparisons, but no changes were found regardless of delivery mode. We recommend LVOL calculated by MR‐based 3‐D model as a objective and practical indicator that can better reflect LA atrophy.
In conclusion, parameter measurement on MR‐based 3‐D model is a direct and comprehensive tool and provides more morphological details about changes of LA before and after childbirth. We found that vaginal delivery, or even active labor itself may both lead to LA atrophy. And the morphology of LA is basically similar in different delivery modes after self‐recovery once the onset of labor has occurred. A limitation of our study stems from the relatively small sample size. And dynamic MRI assessment of pelvic floor was not conducted in our study. Moreover, we did not set observation groups that included women received selective cesarean section or at second stage of labor. And the follow‐up period is not long enough to determine the long‐term effects of childbirth. Further study is necessary to investigate the exact correlation between pelvic floor morphology and clinical symptoms.
Conflict of interest
The authors declare no conflicts of interest.
Supporting information
Appendix S1: Supplementary figures on methodology.
Figure S1. The initial interface for 3D reconstruction after MRI images saved in DICOM format were imputed into the software. L: left side; R: right side; V: ventral side; D: dorsal side; DICOM: digital imaging and communication in medicine.
Figure S2. Illustration and steps of establishing three‐dimensional model of Levator ani and pelvis based on MRI images. (a) Example of identifying and manually segmenting contours of levator ani (yellow contour) and pelvis (white contours) on axial plane. (b) The interface for calculating 3D models using a surface‐rendering algorithm once the operation of “calculate part” is performed. (c) Example of a reconstructed 3D model (left‐anterior view) and automatic calculation of levator ani volume (The levator ani volume is 25003.71 mm3 as it is shown in the information bar). LA: levator ani.
Figure S3. Example of measuring angular parameter in a coronal view (iliococcygeus angle: the angle value is 43.73°as it is shown on the screen). L: left side; R: right side.
Figure S4. Examples of measuring linear parameters. (a) Example of width measuring in a coronal view (iliococcygeus width: the width value is 50.72 mm as it is shown on the screen). (b) Example of thickness measuring on an axial plane (puborectalis thickness: the thickness value is 6.53 mm as it is shown on the screen). (c) Example of length measuring in a lithotomy view (levator hiatus length: the length value is 57.06 mm as it is shown on the screen). (d) Example of distance measuring in an anterior–posterior top view (levator‐symphysis gap: the distance value is 18.45 mm as it is shown on the screen). L: left side; R: right side.
Acknowledgments
This research was funded by National Natural Science Foundation of China (81370736). The authors wish to thank the help from all participants enrolled in Department of Gynecology and Obstetrics, Nanfang Hospital, Southern Medical University, and the support of Department of Radiology.
The work was carried out at Nanfang Hospital, Southern Medical University.
Data Availability Statement
Data available on request from the authors.
References
- 1. Strohbehn K, Ellis JH, Strohbehn JA, DeLancey J. Magnetic resonance imaging of the levator ani with anatomic correlation. Obstet Gynecol. 1996;87(2):277–85. [DOI] [PubMed] [Google Scholar]
- 2. Blomquist JL, Carroll M, Muñoz A, Handa VL. Pelvic floor muscle strength and the incidence of pelvic floor disorders after vaginal and cesarean delivery. Am J Obstet Gynecol. 2020;222(1):62.e1–8. [DOI] [PubMed] [Google Scholar]
- 3. Urbankova I, Grohregin K, Hanacek J, Krcmar M, Feyereisl J, Deprest J, et al. The effect of the first vaginal birth on pelvic floor anatomy and dysfunction. Int Urogynecol J. 2019;30(10):1689–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Lien KC, Mooney B, DeLancey JOL, Ashton‐Miller JA. Levator ani muscle stretch induced by simulated vaginal birth. Obstet Gynecol. 2004;103(1):31–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lockhart ME, Bates GW, Morgan DE, Beasley TM, Richter HE. Dynamic 3T pelvic floor magnetic resonance imaging in women progressing from the nulligravid to the primiparous state. Int Urogynecol J. 2018;29(5):735–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Novellas S, Chassang M, Verger S, Bafghi A, Bongain A, Chevallier P. MR features of the levator ani muscle in the immediate postpartum following cesarean delivery. Int Urogynecol J. 2010;21(5):563–8. [DOI] [PubMed] [Google Scholar]
- 7. Hoyte L, Schierlitz L, Zou K, Flesh G, Fielding JR. Two‐ and 3‐dimensional MRI comparison of levator ani structure, volume, and integrity in women with stress incontinence and prolapse. Am J Obstet Gynecol. 2001;185(1):11–9. [DOI] [PubMed] [Google Scholar]
- 8. Boreham MK, Zaretsky MV, Corton MM, Alexander JM, McIntire DD, Twickler DM. Appearance of the levator ani muscle in pregnancy as assessed by 3‐D MRI. Am J Obstet Gynecol. 2005;193(6):2159–64. [DOI] [PubMed] [Google Scholar]
- 9. Araujo CC, Coelho SSA, Martinho N, Tanaka M, Jales RM, Juliato CRT. Clinical and ultrasonographic evaluation of the pelvic floor in primiparous women: a cross‐sectional study. Int Urogynecol J. 2018;29(10):1543–9. [DOI] [PubMed] [Google Scholar]
- 10. Falkert A, Willmann A, Endress E, Meint P, Seelbach‐Göbel B. Three‐dimensional ultrasound of pelvic floor: is there a correlation with delivery mode and persisting pelvic floor disorders 18–24 months after first delivery? Ultrasound Obstet Gynecol. 2013;41(2):204–9. [DOI] [PubMed] [Google Scholar]
- 11. Liao KD, Yu YH, Li YG, Chen L, Peng C, Liu P, et al. Three‐dimensional magnetic resonance pelvimetry: a new technique for evaluating the female pelvis in pregnancy. Eur J Radiol. 2018;102:208–12. [DOI] [PubMed] [Google Scholar]
- 12. Chen J, Chen C, Li Y, Chen L, Xu J, Liu P. Impact of radical hysterectomy on the transobturator sling pathway: a retrospective three‐dimensional magnetic resonance imaging study. Int Urogynecol J. 2018;29(9):1359–66. [DOI] [PubMed] [Google Scholar]
- 13. Zhuang RR, Song YF, Chen ZQ, Ma M, Huang HJ, Chen JH, et al. Levator avulsion using a tomographic ultrasound and magnetic resonance–based model. Am J Obstet Gynecol. 2011;205(3):232.e1–8. [DOI] [PubMed] [Google Scholar]
- 14. Li N, Cui C, Cheng Y, Wu Y, Yin J, Shen W. Association between magnetic resonance imaging findings of the pelvic floor and de novo stress urinary incontinence after vaginal delivery. Korean J Radiol. 2018;19(4):715–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. LI R, Song Y, Ma M. Relationship between levator ani and bony pelvis morphology and clinical grade of prolapse in women. Clin Anat. 2015;28(6):813–9. [DOI] [PubMed] [Google Scholar]
- 16. Singh K, Jakab M, Reid WMN, Berger LA, Hoyte L. Three‐dimensional magnetic resonance imaging assessment of levator ani morphologic features in different grades of prolapse. Am J Obstet Gynecol. 2003;188(4):910–5. [DOI] [PubMed] [Google Scholar]
- 17. Dierick F, Galtsova E, Lauer C, Buisseret F, Bouché AF, Martin L. Clinical and MRI changes of puborectalis and iliococcygeus after a short period of intensive pelvic floor muscles training with or without instrumentation. Eur J Appl Physiol. 2018;118(8):1661–71. [DOI] [PubMed] [Google Scholar]
- 18. Singh K, Reid WM, Berger LA. Assessment and grading of pelvic organ prolapse by use of dynamic magnetic resonance imaging. Am J Obstet Gynecol. 2001;185(1):71–7. [DOI] [PubMed] [Google Scholar]
- 19. Morris VC, Murray MP, DeLancey JOL, Ashton‐Miller JA. A comparison of the effect of age on levator ani and obturator internus muscle cross‐sectional areas and volumes in nulliparous women. NeurourolUrodyn. 2012;31(4):481–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Zhou B, Zhang H, Yuan J, Bu C, Lai W. Pelvic floor assessment using magnetic resonance imaging after vaginal delivery and elective caesarean delivery. Int Urogynecol J. 2021;32(11):3023–9. [DOI] [PubMed] [Google Scholar]
- 21. van de Waarsenburg MK, Verberne EA, van der Vaart CH, Withagen MIJ. Recovery of puborectalis muscle after vaginal delivery: an ultrasound study. Ultrasound Obstet Gynecol. 2018;52(3):390–5. [DOI] [PubMed] [Google Scholar]
- 22. Snooks SJ, Swash M, Mathers SE, Henry MM. Effect of vaginal delivery on the pelvic floor: a 5‐year follow‐up. Br J Surg. 1990;77(12):1358–60. [DOI] [PubMed] [Google Scholar]
- 23. Allen RE et al. Pelvic floor damage and childbirth: a neurophysiological study. Br J Obstet Gynaecol. 1990;97(9):770–9. [DOI] [PubMed] [Google Scholar]
- 24. Wallner C, Maas CP, Dabhoiwala NF, Lamers WH, DeRuiter MC. Innervation of the pelvic floor muscles: a reappraisal for the levator ani nerve. Obstet Gynecol. 2006;108(3 Pt 1):529–34. [DOI] [PubMed] [Google Scholar]
- 25. Pierce LM, Coates KW, Kramer LA, Bradford JC, Thor KB, Kuehl TJ. Effects of bilateral levator ani nerve injury on pelvic support in the female squirrel monkey. Am J Obstet Gynecol. 2008;198(5):585.e1–8. [DOI] [PubMed] [Google Scholar]
- 26. Dixit P, Shek KL, Dietz HP. How common is pelvic floor muscle atrophy after vaginal childbirth? Ultrasound Obstet Gynecol. 2014;43(1):83–8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1: Supplementary figures on methodology.
Figure S1. The initial interface for 3D reconstruction after MRI images saved in DICOM format were imputed into the software. L: left side; R: right side; V: ventral side; D: dorsal side; DICOM: digital imaging and communication in medicine.
Figure S2. Illustration and steps of establishing three‐dimensional model of Levator ani and pelvis based on MRI images. (a) Example of identifying and manually segmenting contours of levator ani (yellow contour) and pelvis (white contours) on axial plane. (b) The interface for calculating 3D models using a surface‐rendering algorithm once the operation of “calculate part” is performed. (c) Example of a reconstructed 3D model (left‐anterior view) and automatic calculation of levator ani volume (The levator ani volume is 25003.71 mm3 as it is shown in the information bar). LA: levator ani.
Figure S3. Example of measuring angular parameter in a coronal view (iliococcygeus angle: the angle value is 43.73°as it is shown on the screen). L: left side; R: right side.
Figure S4. Examples of measuring linear parameters. (a) Example of width measuring in a coronal view (iliococcygeus width: the width value is 50.72 mm as it is shown on the screen). (b) Example of thickness measuring on an axial plane (puborectalis thickness: the thickness value is 6.53 mm as it is shown on the screen). (c) Example of length measuring in a lithotomy view (levator hiatus length: the length value is 57.06 mm as it is shown on the screen). (d) Example of distance measuring in an anterior–posterior top view (levator‐symphysis gap: the distance value is 18.45 mm as it is shown on the screen). L: left side; R: right side.
Data Availability Statement
Data available on request from the authors.