This cross-sectional study involved a total of 200 postpartum individuals, including 150 participants with PGP and 50 matched controls (Figure 1). Participants were recruited through various methods between March 2023 and May 2024, such as posters and brochures displayed at rehabilitation department in Rehabilitation Hospital and Third People’s Hospital affiliated to Fujian University of Chinese Medicine, newspaper advertisements and referrals from communities near the hospital.

The following inclusion criteria were set for the postpartum women with PGP: (1) fulfilled the updated 2011 diagnostic criteria for PGP, which specify pain in either the posterior pelvic girdle at the sacroiliac joint or pubic symphysis on the anterior side of the pelvic girdle (Kanakaris et al., 2011); (2) postpartum women, 20–40 years of age; (3) within 6 months to 5 years after delivery; (4) a Visual Analogue Scale (VAS) pain score of 3–6 cm; and (5) voluntary participation in this trial and provision of informed consent. The exclusion criteria for the PGP group were: (1) pain located between the subcostal and fifth lumbar regions (Casagrande et al., 2015); (2) symptoms associated with pelvic girdle pain already present before delivery; (3) previous surgery on the lumbar spine, pelvic girdle, hip, or other relevant areas; and (4) acute pelvic inflammation, obvious physiological defects, major diseases, or cognitive impairment. In contrast, the asymptomatic group had: (1) no history of prior PGP; (2) no current pain in the lumbopelvic region for at least 6 months; (3) no multiple joint pain in the extremities; and (4) normal gait pattern. All histories, physical examinations, and outcome assessments were performed by the same physiotherapist. Informed written consent was obtained from participants after explaining the detailed procedure. This study was approved by the Research Committee of the Rehabilitation Hospital affiliated with Fujian University of Chinese Medicine (Fujian, China) (ethics approval number: 2023KY-001-01).

Musculoskeletal ultrasound imaging was performed using a Siemens-Sequoia Silver ultrasound system. An experienced physical therapist specializing in women’s health, with over 3 years of experience in musculoskeletal ultrasound conducted the assessments without knowing the participant group. A water-soluble transmission gel was applied to the measurement site, and either the 10L4 (superficial probe) or 5C1 (convex array probe) was placed on the muscle based on the specification outlined in Table 1 (Othman et al., 2023; Nojiri et al., 2021; Taniguchi et al., 2015). Figure 2 illustrates two-dimensional ultrasound imaging of the ES, QL, IM, and RF muscle at rest.

Lumbar muscle (ES, LD, QL) thickness were measured at rest on both sides, comparing the symmetry between left-right side. Given that the skin folds and movement characteristics of these muscles during maximum contraction in the prone position influence the stability of the probe, measurements were only obtained at rest. Pelvic gluteal muscles thickness and left-right symmetry (GMax and GMed) were collected at rest and maximum contraction (MC), while PF muscle thickness was only measured at rest due to the difficulty in obtaining accurate measurements at greater depths during maximum contraction. Hip muscles (IM, RF, and BF) thickness were also measured at rest and maximum contraction on the right side (dominant leg). Given that hip flexion could affect probe stabilization, IM muscle thickness was only captured at rest. Each measurement was repeated three times, and calculate the mean value of resting or maximum contracted muscle thickness. Left-right symmetry (L/R) was determined by comparing muscle thickness at rest between the left and right sides. An L/R ratio closer to 1 indicates greater symmetry, while a ratio >1 suggests greater muscle thickness on the left side, and a ratio <1 indicates greater muscle thickness on the right side. The muscle contraction ratio during maximum contraction was calculated using the following equation (Bashir et al., 2019). A higher contraction ratio (CR) indicates better contraction activity. contraction ratio  CR =   thickness muscle   contracted — thickness   rest thickness   rest

The analysis was conducted using SPSS (Version 27). Data normality was assessed using the Shapiro-Wilk test and Q - Q plots. Continuous variables were reported as mean (SD) or median (first quartile [Q1]−third quartile [Q3]). Student’s t-test or Mann–Whitney U test for continuous data and the chi-square test for categorical data were employed to compare baseline characteristics between the two groups. To compare outcomes in lumbo-pelvic-hip complex muscles thickness and contraction ratio between the two groups, Student’s t-test or Mann–Whitney U test was used. Chi-square test was used to compare the qualitative data, i.e., number and mode of delivery. P < 0.05 was considered statistically significant across the entire statistical analysis.

A total of 200 participants completed the trial, with a mean age of 32.29 years (SD = 4.65). Table 2 lists the clinical characteristics. Both groups were comparable in terms of baseline characteristics. The majority of postnatal women in both groups had delivered once (PGP: 57%, controls: 46%), and most had undergone vaginal deliveries (PGP: 67%, controls: 62%). In the PGP group, pain was predominantly located on the left side (49%), followed by pain in multiple sides (25%), the anterior pelvic region (14%) and the right side (R: 12%). The mean duration of pain in the PGP group was 12.68 months (SD = 9.82).

Table 3 presents the between-group differences in ES, LD, QL, PF, IM, RF, and BF muscle morphometry between postpartum women with PGP and asymptomatic women. The PGP group exhibited lower ES muscle thickness bilaterally (P< 0.001). The LD muscle was thicker on the right side and asymmetrical in the PGP group (P< 0.05). There were no differences in QL and PF muscle thickness bilaterally or in bilateral symmetry between the two groups (P> 0.05). Regarding lower limb muscle morphometry, a statistically significant difference was noted in the right-side IM muscle thickness (P = 0.02). Furthermore, the PGP group demonstrated reduced RF muscle thickness at maximum contraction (P< 0.001) and a smaller muscle contraction ratio (P = 0.01). However, there were no differences in the BF muscle thickness at rest, maximum contraction, and muscle contraction ratio (P> 0.05).

Figure 3A illustrates the muscle thickness and contraction ratio differences in the GMax. The PGP group exhibited significantly reduced muscle thickness at rest, maximum contraction, and muscle contraction ratio on the left side (P< 0.05). Moreover, the right GMax muscle thickness at maximum contraction and muscle contraction ratio were significantly lower in the PGP group compared to asymptomatic women (P< 0.05). Statistically significant difference was also observed in the left–right symmetry (P< 0.001). However, there was no difference in the right GMax muscle at rest (P = 0.17).

Figure 3B presents the differences in GMed muscle morphometry between the two groups. There were no differences in the resting muscle thickness on the left and right side (P> 0.05). However, the left muscle thickness at the maximum contraction and the muscle contraction ratio were significantly lower in women with PGP compared to asymptomatic women (P< 0.05). The right side of GMed muscle thickness at rest, maximum contraction, and contraction ratio did not exhibit differences between the two groups (P> 0.05).

Our findings revealed that women with PGP had reduced ES on both sides compared to asymptomatic women, while the LD muscle was thicker than the asymptomatic women and more thicker on the right side. To our knowledge, no previous study has specifically examined ES and LD muscle morphology in this specific population, resulting in a lack of direct comparative analyses. In functional anatomy, these muscles cross over to produce movement via attachments at the spine or hip (Fiani et al., 2021). Moreover, the ES, BF, and GMax muscles have the greatest effect on sacroiliac joint stiffness (Farley et al., 2024). In this study, the postpartum PGP group exhibited reduced muscle thickness of ES on both sides, potentially indicating a decreased ability in the force closure function of the pelvis. The current results were similar with the findings of Ronchetti et al. (2008) who used muscle strength test to demonstrate that postpartum women with PGP exhibited reduced strength and endurance of the trunk flexors and extensors compared with pain-free postpartum women. Shadmehr et al. (2012) found a smaller recruitment of ES during active straight leg raise (ASLR) in women with sacroiliac pain compared to controls, while their study participants were not pregnant or postpartum women. In the current study, the LD muscle was thicker in the PGP group. A study in healthy volunteers demonstrated that excessive LD activity during ASLR could be induced by experimental pelvic pain (Palsson et al., 2015). The increased LD muscle thickness in the PGP group observed in this study might be a compensatory action for trunk extensor insufficiency, aiming to maintain pelvic stability during childcare activities. However, due to the limited number of studies on lumbar muscles morphometry in postpartum individuals with PGP, conclusions remain tentative, and further research is necessary to elucidate these findings.

Furthermore, no significant differences in PF and QL muscle thickness were observed between the two groups. This may be attributed to changes in muscle tone rather than thickness. Wójcik, M et al. (2023) using a review of journal databases, found that increased lumbar and PF muscle tone was strongly associated with lumbopelvic pain. Therefore, it is very important to relax the PF and lumbar muscle. Additionally, the PGP group exhibited greater asymmetry, with the right LD muscle demonstrating increased thickness compared to the left side. This finding contrasts with the results of Bashir et al. (2019), who reported a significant decrease in LD muscle thickness when compared to the painful side in subjects with sacroiliac joint dysfunctions. While in this present study, a tendency for LD muscle thickening on the contralateral side of the painful area (predominantly on the left side in PGP group, 49%). These discrepancies may be partially attributed to participant selection, as their study did not focus specifically on postpartum women. Anatomically, in the posterior oblique chain, the GMax is connected via fascia to the sacroiliac joint and the contralateral LD (Gibbons, 2017). As most subjects in this study experienced left-sided pain, muscle spasms may inhibit the GMax on the painful side, potentially leading to compensatory hypertrophy of the LD muscle on the non-painful side. In summary, ES muscle strengthening and LD lengthening should be considered when developing individualized exercise prescriptions for postpartum PGP.

In this study, the PGP group displayed thinning in bilateral GMax and a lower contraction ratio on both sides compared with asymptomatic women. Additionally, left-right asymmetry of muscle at rest was observed in the PGP group. This result is similar with previous reports by Bashir et al. (2019), which found that GMax muscle thickness decreased significantly among subjects with sacroiliac joint dysfunctions on the pain side. Bagwell et al. (2020) demonstrated that the GMax plays a crucial role in pelvic force closure and dynamic stability of the low back and pelvis, and disuse of the GMax muscle could increase joint loads and potentially contribute to pain during and after pregnancy. Furthermore, Shadmehr et al. (2012) utilized electromyography (EMG) and observed a significant difference in tonicities of the BF, GMax, and ES muscles between patients with sacroiliac joint dysfunction and healthy controls. Ludwig et al. (2024) found that the GMax muscle had the strongest correlation with changes in pelvic tilt in healthy individuals. Moreover, the GMax muscle is morphologically densely linked to the pelvic floor via strands of connective tissues investing the adjacent muscles, which also has an impact on pelvic floor dysfunction (Siess et al., 2023). In the current study, postpartum individuals with PGP exhibited notably reduced GMax muscle thickness on both sides compared to asymptomatic women, suggesting potential laxity in muscle contraction function. As reduced muscle thickness might affect the biomechanical properties of the joint by altering mechanical stress and load transfer (Bashir et al., 2019), it may lead to the occurrence of chronic pain. Therefore, it is recommended to emphasize GMax muscle strength in exercise training for this population. Further investigation of its morphology and activation during tasks is also essential.

Regarding the GMed muscle, the PGP group exhibited reduced muscle thickness at maximum contraction and a lower contraction ratio on the left side. The findings of this study is similar with previous research conducted on patients with low back pain and individuals with no history of pain. Aboufazeli et al. (2018) employing the same ultrasound method, observed a reduction in GMed muscle thickness change in patients with low back pain. Similarly Cooper et al. (2016), using manual muscle testing, discovered that GMed muscle weakness was a common symptom in participants with chronic non-specific low back pain. In conjunction with the GMax muscle results, the left gluteus muscle was suppressed, and contractile dysfunction was more pronounced in patients with PGP. Considering the crucial role of GMed muscle as a pelvic stabilizer during gait and its strong correlation with musculoskeletal dysfunction diseases (Fenato et al., 2021), strengthening the GMed muscle on the painful side is a recommended protocol for exercise prescription in postpartum women with PGP. Furthermore, future research investigating the influence of PGP on gait biomechanics, utilizing motion analysis and EMG systems, is warranted.

Regarding the lower limb muscles, the iliacus muscle thickness on the right side was significantly greater in the PGP group compared to the control group. To our knowledge, this study is the first to examine the characteristic changes of the iliacus muscle in postpartum individuals with PGP. Vleeming et al., 's 2008 pelvic girdle guidelines recommend palpating the iliopsoas muscle in patients with PGP, but do not provide specific characteristics (Vleeming et al., 2008). Anatomically, the iliacus muscle joins the psoas muscle to form the iliopsoas muscle, which is surrounded by the dense iliac fascia (Siccardi et al., 2024). The fascia covering the iliopsoas muscle forms multiple fascial connections, linking the muscle to various viscera and muscle regions (Bordoni et al., 2023). Our findings regarding the iliacus muscle is similar with the work of Fujitani et al. (2021), who discovered higher EMG activity in the iliacus muscle of participants with recurrent lower back pain compared to healthy controls. Clinically, iliac muscle stiffness can often be detected through palpation. Using ultrasonography Miyachi et al. (2021), found a significant positive correlation between anterior pelvic tilt position and iliopsoas muscle thickness. Based on these findings, we hypothesize a potential relationship between changes in iliac muscle thickness and increased pelvic tilt. Radiographic studies have indicated that individuals with sacroiliac joint pain often exhibit increased anterior pelvic rotation during ASLR compared with healthy individuals (Mens et al., 1999). Future study could incorporate pelvic imaging methods to analyze the relationship between pelvic tilt angle and iliac muscle characteristics in postpartum women with PGP.

In term of the RF and BF muscle, our results showed that the RF muscle thickness at maximum contraction and contraction ratio were significantly smaller in the PGP group than in the asymptomatic group, while no significant differences were observed in the BF muscle morphometry between the two groups. Anatomically, the RF muscle originates from the anterior inferior iliac spine and inserts into the patellar ligament above the patella. The RF muscle is connected to the pelvis via the anterior inferior iliac spine, and its contraction and relaxation have an effect on the position and movement of the pelvis. Tightness or shortening of the RF muscle may lead to anterior pelvic tilt, increasing lumbar lordosis and sacral tilt angle, potentially causing low back pain and pelvic girdle pain (Gibbons, 2017). In combination with other results from this study, postpartum women with PGP may exhibit tense contraction of the iliac muscle while the RF muscle may be inhibited due to pain. A previous study by Yoo (2013) demonstrated that strengthening exercises targeting anterior pelvic tilt muscles (ES, IM, RF) can restore pelvic tilt angles. Therefore, exercise training for individuals with PGP should also focus on strengthening the contraction ability of RF muscle.

Although the current study did not find any difference in the BF muscle, previous research using EMG methodology had different findings regarding BF muscle activity. Shadmehr et al. (2012) observed reduced recruitment of the BF muscle during the ASLR test in women with sacroiliac pain compared to healthy controls. Conversely Palsson et al. (2015), noted excessive BF muscle activity in healthy volunteers during an experimental pain-inducing ASLR test. Two studies by Bussey and Milosavljevic (2015); Bussey et al. (2020), they reported that patients with PGP appear to employ a symptom-led strategy for bracing the innominate through opposing tension in the BF muscle and external oblique during movement of the affected side, potentially increasing mechanical stress on the sacroiliac joint and exacerbating pain. And patients with PGP exhibited longer muscle onset latencies in the BF muscle under visual occlusion (Bussey et al., 2020). While these studies did not focus on postpartum women, the significant impact of BF muscle activity on lumbopelvic control cannot be disregarded. Moreover, a prospective study demonstrated that BF muscle delay during single-leg lift in pregnancy was a significant predictor for the development of PGP in late pregnancy (Aldabe et al., 2020). Therefore, further investigation on the influence of PGP on BF muscle activity using EMG systems in postpartum women is warranted.

This study has several limitations that should be acknowledged. First, the research included only individuals with mild to moderate PGP. Future studies could explore whether those with severe PGP exhibit different performance characteristics. Second, due to the limited number of studies investigating muscle morphometry and contraction changes in the LD, IM, RF and BF muscle of postpartum women with PGP, the conclusions about these muscles remain tentative. Third, some muscle thickness measurements were only collected at rest. Future research could incorporate EMG method to better observe muscle changes during various functional tasks, and what sort of the muscle asymmetry associated with pain might also be warranted.