Pilates is widely acknowledged as an effective exercise for enhancing physical, psychological, and motor functions. However, there remains ongoing debate regarding the effects of Pilates on pregnancy outcomes. Thus, the objective of this systematic review was to evaluate and critically analyze the existing scientific evidence regarding the effects of Pilates on pregnant women.

A thorough search was performed across 7 databases, focusing on articles published up to June 2024. The quality of evidence was assessed using the Cochrane RoB Tool, and STATA 17.0 software was utilized to perform tests for heterogeneity, publication bias, and sensitivity analysis.

(1) A total of 11 studies were included in this review, comprising 538 participants. Among these, 3 studies exhibited a low risk of bias, while 8 studies presented an unclear risk of bias. (2) The meta-analysis results indicated that Pilates exercise resulted in non-significant changes in the Body Mass Index (BMI) of pregnant women (SMD 0.02; 95% CI, -0.27 to 0.31; P = 0.887), the body weight of pregnant women (SMD 0.25; 95% CI, -0.52 to 1.02; P = 0.519), the mode of delivery (RR 1.41; 95% CI, 0.95 to 2.09; P = 0.093), and the Low-density lipoprotein (LDL) levels of pregnant women (SMD − 1.182; 95% CI, -2.443 to 0.079; P = 0.066). However, Pilates exercise during pregnancy was associated with significant improvements in APGAR scores at 1 min (SMD 0.41; 95% CI, 0.08 to 0.74; P = 0.015), a reduction in the duration of labor (SMD − 0.557; 95% CI, -0.915 to -0.198; P = 0.002), and a decrease in the Visual Analogue Score (VAS) score (SMD − 0.76; 95% CI, -1.45 to -0.07; P = 0.032). (3) Sensitivity analysis revealed no significant differences in effect size before and after the removal of any individual study, indicating that the meta-analysis results were stable.

This systematic review and meta-analysis demonstrate that Pilates during pregnancy positively influences the shortening of labor duration, improves the newborn’s immediate physiological status, and alleviates labor pain.

The study was registered on the International Prospective Registry of Systematic Reviews——PROSPERO database (CRD42023393309) on January 30, 2023.

Physical activity(PA) during pregnancy can yield long-term health benefits for both the fetus and the mother. Observed benefits include improved cardio-respiratory fitness, prevention of stress urinary incontinence, reduction of lumbar pain, decreased levels of depression, and better control of weight gain during pregnancy [1]. The American College of Obstetricians and Gynecologists (ACOG) recommends engaging in physical activities as a form of non-pharmacological therapy for at least 150 min per week to promote health [2]. However, evidence indicates that only 32% of pregnant women adhere to the recommended physical activity guidelines, with this figure declining to 12% in the later stages of pregnancy [3].

Insufficient physical activity among pregnant women may stem from a lack of professional guidance regarding the advisability of exercise during pregnancy, as well as the most suitable types, frequencies, intensities, and durations of exercise [4]. Internationally, Pilates is recognized for its series of exercises that enhance strength and flexibility throughout the body, making it a significant form of exercise for improving physical, psychological, and motor functions [5]. Furthermore, there is growing evidence suggesting that Pilates may positively influence various aspects of birthing outcomes and enhance maternal health^[6]. Notably, a study found that pregnant women are more inclined to engage in Pilates, with the frequency of participation in other activities decreasing, while participation in Pilates remained stable [6].

Despite this, there remains controversy and a lack of comprehensive research regarding the effects of Pilates on pregnancy outcomes. Randomized controlled trials examining Pilates in women indicated that it can improve lower limb function, quality of life, and pain management [7], yet pregnant women were not included in this analysis. Consequently, this systematic review and meta analysis aims to assess and critically analyze the existing scientific evidence concerning the effects of Pilates on pregnant women.

The research was executed in alignment with the standards set forth by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (see Additional file 1) [8], ensuring a rigorous approach to systematic reviews. Furthermore, it adhered to the protocols outlined in the Cochrane Handbook for Systematic Reviews of Interventions, version 6.1 [9], which was established in the United Kingdom. To enhance transparency and facilitate further research, this study has also been registered with the International Prospective Registry of Systematic Reviews, commonly referred to as the PROSPERO database, under the registration number CRD42023393309 (available from https://www.crd.york.ac.uk/PROSPERO/display_record.php? RecordID=393309).

A structured approach to inclusion and exclusion criteria was established to enhance the overall quality and applicability of the research results.

Inclusion criteria for this study: (1) The research focused on journal articles that addressed the impact of Pilates exercise during pregnancy. These articles had to be published in both Chinese and English and sourced from any database available from their inception up to June 2024; (2) The study population was restricted to pregnant women who had no history of habitual abortion and did not present any contraindications to exercise during their pregnancy; (3) The intervention group in the study was required to participate in Pilates exercises frequently. In contrast, the control group received standard prenatal care education or participated in alternative physical exercise forms; (4) The study design was randomized controlled trials (RCT).

Exclusion criteria for this study: (1) Any literature that provided incomplete data or whose full text was unavailable; (2) Articles published in non-core Chinese journals, as well as duplicate literature; (3) Studies that were not original research, such as conference proceedings, abstracts, meta-analyses, and systematic reviews.

A comprehensive and systematic search of published studies was meticulously conducted, covering publications up to June 2024. The following electronic databases were utilized for the searches: PubMed (MEDLINE), the Cochrane Library, Embase, EBSCO, Scopus, CNKI (Chinese National Knowledge Infrastructure), and the WanFang database. The search options selected included “Advanced Search” and “All fields”: (“Pilates” OR “Mat-Pilates” OR “Pilates-Based Exercises” OR “Exercises, Pilates-Based” OR “Equipment-Based Pilates” OR “Pilates Training” OR “training, Pilates”) AND (“pregnancy” OR “gestation” OR “maternal” OR “puerperal” OR “parturient” OR “pregnant”). These keywords were searched for using the search formula (see Table 1 ). Additionally, reference lists from pertinent studies and earlier reviews were examined to discover other studies that may be relevant.

Two reviewers (Y.L. and H.L.) conducted the literature search and screening process independently. An extensive literature search was performed across seven key databases, with a customized search strategy designed for each database. The screening process was initially conducted by reviewing the title and abstract, categorizing them as either excluded or possibly included, and any duplicate references within the identified articles were eliminated. Subsequently, the reviewers assessed whether the studies fulfilled the inclusion criteria, which considered methodology, participant characteristics, and interventions. Furthermore, the reference lists of the selected articles as well as relevant review papers were scrutinized to uncover additional potential studies. During the eligibility phase, a third reviewer (L.Z.) was involved in resolving conflicts through a consensus process, and the final set of articles for inclusion in the review was established.

Data extraction was performed by utilizing a standardized template and organizing the findings into Excel spreadsheets. The variables collected include identification information (title, authors, publication date); design of the study, sample size, and characteristics of the population involved (BMI, age, gestational age at delivery); specifics of the intervention (type, method, frequency, and duration of follow-up in terms of gestational weeks); techniques for outcome assessment; and the outcomes reviewed in this analysis. Any discrepancies were addressed through discussion and the attainment of a consensus. In instances where data were not available in the original articles, the study authors were contacted via email to obtain supplementary information.

The results encompassed five primary categories: outcomes related to body mass index (BMI) and body composition (body fat mass, skeletal muscle mass, and percent body fat); biochemical indicator outcomes (levels of high-density and low-density lipoprotein); physical function outcomes (strength of pelvic floor muscle contractions, lumbopelvic stabilization, and disability); pregnancy outcomes (duration of labor, delivery method, neonatal weight, and APGAR scores at 1 and 5 min); Quality of sleep and psychological outcomes (stress and anxiety levels); and quality of life outcomes.

Two independent reviewers (Y.L. and H.L.) conducted an assessment of the risk of bias (RoB) by utilizing the Cochrane RoB Tool version 2 [10]. This tool incorporates a domain-based approach to evaluation, focusing on five fundamental areas: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. For each of these domains, the reviewers categorized the risk of bias as “low risk,” “unclear risk,” or “high risk.” Any discrepancies arising during the quality assessment were resolved by the third author (L.Z.).

Data analysis was carried out utilizing STATA 17.0 software, while literature management was facilitated with Endnote. In the context of individual trials, the measures of effect for categorical outcomes were quantified in terms of odds ratios (OR), accompanied by their respective 95% confidence intervals (CI). For outcomes measured continuously, the mean difference was adopted as the effect metric, also with the inclusion of its 95% CI. To evaluate the heterogeneity among the results of the different studies, Cochran’s Q test was employed, and the extent of heterogeneity was expressed using the I² statistic. A threshold of I² < 50% and P > 0.1 indicated non-significant heterogeneity, which warranted the application of the Mantel-Haenszel (M-H) fixed-effect model for the analysis. Conversely, when the heterogeneity was deemed significant, the DerSimonian-Laird (D-L) random effects model was utilized. Additionally, a leave-one-out analysis was conducted for variables that showed significant effects to ascertain the influence of specific trials on the overall results. To assess the potential for publication bias, funnel plots were generated, and Egger’s test was employed, where a significance level was set at P < 0.05.

Several major databases were searched for studies from the inception of that database to 30th June 2024. The PRISMA diagram illustrates the progression of studies through various stages of the review process. A total of 243 relevant literature were retrieved. After the removal of 132 duplicate records, 28 studies were assessed based on their title and abstracts. Ultimately, 11 studies were included in this review. The reasons for the exclusion of studies are detailed in Fig. 1.

Literature screening flowchart

Full size image

The characteristics of the study participants, interventions, and findings are presented in Additional file 2. There were differences in the number of study participants in the 11 included papers, which included a total of 538 study participants, with a total of 272 in the intervention group and 266 in the control group. The minimum sample size of literature included was 16 [11, 12] and the maximum was 105 [13]. One study did not report the age of the study population [14] and in one study [11], the mean age of the pregnant women in both the intervention and the control groups was more than 35 years. Additionally, there was also one study [14] of pregnant women with gestational diabetes, and one study [15] of pregnant women with back pain.

From the included studies, there were 9 studies with routine mat Pilates interventions and 2 studies [15, 16] using clinical Pilates intervention methods, the overall length of training intervention and the frequency of exercise were variable among trials: they were all in the range of 4–12 weeks, and two studies [13, 17] had a duration of fewer than 50 min per Pilates session, 40–45 min and 35 min respectively, while the remaining 9 studies had sessions in the range of 50–70 min. 6 papers had routine prenatal care [13, 12,13,14,15, 18, 19], 4 controls had no intervention, and 1 study [20] had a control group that performed upper and lower extremity and trunk muscle strength exercises through elastic bands and self-resistance.

The outcome indicators of the studies covered multiple aspects, 4 studies [11,12,13, 18] measured pregnant women’s BMI and body composition, including indicators of muscle mass, body fat mass, and percentage of body fat; 5 papers documented pregnancy outcome indicators [13, 14, 16, 17, 21], including duration of the first stage of labor, duration of the second stage of labor, gestational age at delivery, pain index, mode of delivery, neonatal quality, and neonatal scores; 2 papers [11, 19] measured pregnant women’s psychosocial indicators, and one paper measured the level of postpartum depression, sleep disorders and stress levels in pregnant women, one article measured anxiety and depression levels in pregnant women with gestational diabetes before and after a Pilates intervention, and one article [12] measured indicators related to muscle damage and inflammation in pregnant women.

The results of the risk of bias assessment are expressed as percentages in Fig. 2, and the review author’s judgments about each risk of bias item for the included studies can be seen in Fig. 3. We rated all the trials as low risk of bias related to Randomization process and Missing outcome data, as they reported explicit randomization method and had no missing outcome data. One study [18] did not report explicit information about Deviations from intended interventions and was judged at unclear risk of bias. We judged 5 studies [12, 16, 17, 19, 21] at unclear risk of Selection of the reported result due to the missing outcome definition. 3 studies [13, 14, 20] were at low risk of bias, 8 studies [11, 12, 15,16,17,18,19, 21] were at unclear risk of bias.

Risk of bias assessment chart (expressed as percentages)

Full size image

the review author’s judgments about each risk of bias item for the included studies

Full size image

BMI and body composition of pregnant women

All of the studies observed pre-intervention BMI in pregnant women, 3 papers [11, 13, 18] reported on the BMI and body composition of pregnant women after the intervention. In the study by Kim [11], the Pilates group showed significant changes in BMI and body fat mass (BFM), skeletal muscle mass (SMM), percent body fat (PBF) in one study [18], and the results of studies [13, 19] showed that pregnant women participating in the Pilates program exhibited lower body weight compared to those in the control group. In the study by Hyun et al. [12], SMM differed significantly (P < 0.05), with the Pilates group exhibiting higher SMM compared to the control group.

Previous studies have indicated that Pilates exercises influence various biochemical indicators, including blood glucose and cholesterol levels. A study [19] examined the impact of Pilates exercise on blood glucose levels in pregnant women with gestational diabetes. The findings indicated a statistically significant difference (P < 0.05) in both fasting and postprandial blood glucose levels between the Pilates and control groups. The results suggest that Pilates significantly improved blood glucose levels in this population. In another study conducted by Kim et al. [11], significant differences were observed between the Pilates and control groups with respect to total cholesterol, LDL cholesterol, and serotonin levels (P < 0.05). Specifically, the Pilates group showed significantly lower levels of total cholesterol and LDL cholesterol, and higher serotonin levels compared to the control group.

Pregnancy outcomes have been a primary focus in previous studies evaluating the effects of Pilates exercise interventions.

(1) Previous studies [14, 17] have focused on the duration of the first and second stages of labor in pregnant women. In Mazzarino’s study [14], no statistically significant difference was observed between the Pilates and control groups in the duration of the first and second stages of labor (P > 0.05). However, another study [17] found a statistically significant difference in the duration of the second stage of labor (P < 0.05), with the Pilates group experiencing a shorter duration compared to the control group.

(2) The types of delivery observed by the intervention group were characterized by normal delivery and cesarean Sect. 4 studies [13, 14, 16, 17] revealed a normal delivery rate of 84.96% of pregnant women in the Pilates group and 60.87% in the control group. Rodríguez’s study [13] found a statistically significant difference in the mode of delivery between the Pilates and control groups (P < 0.05), with a lower percentage of cesarean sections in the Pilates group. However, the remaining three studies [14, 16, 17] did not show statistically significant differences in the mode of delivery between the Pilates and control groups (P > 0.05).

(3) 2 studies [13, 16] measured neonatal weight. In Aktan et al.‘s study [16], there was no statistically significant difference in neonatal weight between the Pilates and control groups (P > 0.05). Conversely, Rodríguez et al.‘s study [13] found a statistically significant difference (P < 0.05), with neonates in the Pilates group weighing less than those in the control group.

(4) As a key scoring indicator for evaluating the health of newborns, 2 studies [16, 17] measured APGAR scores at 1 min and 5 min. The results of Ghandali’s [17] study showed that the difference between APGAR scores at 1 min and 5 min was not statistically significant (P > 0.05) in both the experimental and control groups. In contrast, Aktan’s study [16] reported a statistically significant difference in APGAR scores at 1 min between the Pilates and control groups (P < 0.05), though the difference in APGAR scores at 5 min was not statistically significant (P > 0.05).

In a study conducted by Kim et al. [11], the findings indicated that pregnant women in the Pilates group experienced better sleep quality, lower stress scores, and reduced Edinburgh Postnatal Depression Scale (EPDS) scores compared to the control group. Additionally, Aktan et al. [16] measured anxiety levels using the State-Trait Anxiety Inventory (STAI) and found that pregnant women who participated in clinical Pilates exercise and childbirth training had significantly lower anxiety levels (P < 0.05).

In 2021, Sonmezer et al. [15] demonstrated that Pilates can reduce the level of back pain in pregnant women suffering from back pain, and Nascimento et al. [19] concluded that sexual pain scores were significantly lower in the Pilates group compared to the control group in the sexual pain dimension (P < 0.05). Furthermore, 3 studies [16, 17, 21] assessed pain scores during labor using the VAS score and reported statistically significant differences, with the Pilates group experiencing lower pain scores during labor compared to the control group (P < 0.05).

The effect of Pilates exercise on pelvic floor muscle (PFM) contraction strength can be revealed by Dias’ study [20]. The results indicated that Pilates exercise did not lead to a significant increase in PFM strength, with no statistically significant difference observed between the experimental and control groups (P > 0.05).

E. Sonmezer et al. [15] investigated the effect of Pilates exercise on lumbopelvic fixation and dysfunction in pregnant women with low back pain. The results showed that, following 8 weeks of Pilates exercise intervention, there was a significant difference in the level of lumbopelvic stability and dysfunction between the Pilates and control groups (P < 0.05).

In 2017, Rodriguez-Diaz was the first to focus on the effects of Pilates exercise on grip strength, situs flexibility of the right and left legs, dorsal lordosis, and lumbar lordosis in pregnant women [13]. And results showed that after 8 weeks of Pilates intervention, the Pilates group demonstrated significantly greater grip strength, improved hamstring flexibility, and reduced spinal curvature compared to the control group (P < 0.05).

We found that Sonmezer and Mazzarino observed quality of life scores in outcomes. In a study by Sonmezer measured the quality of life by the Nottingham Health Profile (NHP), and no positive effect of Pilates on the quality of life of pregnant women was found [15]. In contrast, Mazzarino et al. evaluated quality of life using the Short Form Health Survey (SF-12) and reported a statistically significant difference in total scores, with the Pilates group showing higher scores compared to the control group (P < 0.05) [14], due to the use of different measurement scales, no further comparative analysis was conducted.

The meta-analysis focused on BMI, body weight, APGAR scores, the mode of delivery, the duration of labor, labor pain, and the LDL of pregnant women because they were the most studied variables in the selected studies.

BMI and body weight of pregnant women

The results of BMI analysis show that Pilates exercise led to non-significant decreases in the BMI of pregnant women (SMD 0.02; 95% CI, -0.27 to 0.31; P = 0.887; Fig. 4). In the meta-analysis, significant heterogeneity was observed among the studies [11, 13, 18] (I² = 37.7%, P = 0.201). Therefore, a random effects model was used to account for this variability.

Forest plot of Pilates on the BMI of pregnant women

Full size image

2 studies [12, 13] reported on the weight of pregnant women following the intervention. Due to significant heterogeneity among the studies, a random effects model was employed (I² = 54.7%, P = 0.138). The results revealed that Pilates exercise led to non-significant changes in body weight of pregnant women (SMD 0.25; 95% CI, -0.52 to 1.02; P = 0.519; Fig. 5).

Forest plot of pilates on the body weight of pregnant women

Full size image

(1) Neonatal weight

2 studies [13, 16] reported neonatal weight. A fixed-effects meta-analysis using Standard Mean Difference (SMD) showed Pilates exercise led to insignificant decreases in the neonatal weight (SMD − 0.22; 95% CI, -0.15 to 0.11; P = 0.196; Fig. 6). No heterogeneity was observed across the studies (I² = 0.0%, P = 0.435).

Forest plot of pilates on the neonatal weight

Full size image

(2) APGAR scores

The studies by Ghandali [17] and Aktan [16] reported inconsistent findings regarding APGAR scores at 1 min, while results for APGAR scores at 5 min were consistent, showing that Pilates did not affect these scores at 5 min. A random effects meta-analysis using SMD showed Pilates exercise led to significant increases in the APGAR scores at 1 min (SMD 0.41; 95% CI, 0.08 to 0.74; P = 0.015; Fig. 7). No heterogeneity was observed across the studies (I² = 0.0%, P = 0.797).

Forest plot of pilates during pregnancy on the APGAR score at 1 min

Full size image

(3) The mode of delivery

4 studies [13, 14, 16, 17] reported on the mode of delivery of pregnant women. A random effects meta-analysis using Relative Risk (RR) showed Pilates exercise led to non-significant changes in the mode of delivery of pregnant women (RR 1.41; 95% CI, 0.95 to 2.09; P = 0.093; Fig. 8). High and significant heterogeneity was observed across the studies (I² = 77.9%, P = 0.004).

Forest plot of pilates on the mode of delivery

Full size image

(4) The duration of labor

As shown in Figs. 2 and 7 studies [14, 17] recorded the duration of labor. Due to non-significant heterogeneity in studies, a fixed-effects model was applied (I² = 0.0%, P = 0.656). The results indicated that Pilates exercise led to significant decreases in the duration of labor (SMD − 0.557; 95% CI, -0.915 to -0.198; P = 0.002; Fig. 9).

Forest plot of pilates on the duration of labor

Full size image

(5) the VAS scale

Labor pain was measured using the VAS scale in 2 studies [16, 17]. A random effects meta-analysis using SMD revealed that Pilates exercise led to significant decreases in the VAS scores (SMD − 0.76; 95% CI, -1.45 to -0.07; P = 0.032; Fig. 10). High and significant heterogeneity was observed across the studies (I² = 72.1%, P = 0.058).

Forest plot of pilates on the VAS scale of pregnant women

Full size image

The LDL of pregnant women

2 studies [11, 12] showed consistent findings regarding the effect of Pilates on high density lipoprotein (HDL) levels in pregnant women, but their results on LDL levels were inconsistent. Specifically, Kim et al. [11] demonstrated that Pilates effectively reduced LDL levels in pregnant women. A random effects meta-analysis using SMD showed Pilates exercise led to non-significant decreases in the LDL of pregnant women (SMD − 1.182; 95% CI, -2.443 to 0.079; P = 0.066; Fig. 11). High and significant heterogeneity was observed across the studies (I² = 61.5%, P = 0.107).

Forest plot of pilates on the LDL of pregnant women

Full size image

Publication bias and sensitivity analysis

Funnel plots and Egger’s test were applied to analyse the publication bias, and neither the visual inspection of funnel plots (see Fig. 12) nor Egger’s test results revealed any signs of publication bias (see Table 2). Moreover, we conducted a sensitivity analysis (for these outcomes included literature number exceed 2) to evaluate the robustness of our findings. The results indicated that the effects of Pilates exercise on BMI and mode of delivery in pregnant women were not significantly effected by the exclusion of any single study. This suggests that our conclusions are relatively stable (see Tables 3 and 4).

Publication bias test funnel plot

Full size image

The effect of Pilates exercise during pregnancy on maternal body weight remains a topic of conflicting views in current studies. Meta-analysis results indicate that there was no statistically significant difference in maternal BMI between the Pilates and control groups (SMD 0.02; 95% CI, -0.27 to 0.31; P = 0.887), suggesting that Pilates exercise does not impact maternal BMI and supporting the notion that it is not associated with weight gain during pregnancy. A study comparing maternal characteristics of obese and normal pregnant women revealed that excessive weight gain in pregnancy is linked to higher leptin concentrations early in pregnancy, as well as elevated fetal leptin levels [22]. This finding suggests a potential mechanism through which Pilates interventions could influence maternal body composition by potentially reducing maternal leptin levels, thereby mitigating excessive fat accumulation during pregnancy and possibly preventing obesity development. It’s important to note that none of the literature reviewed in this study assessed the dietary intake of pregnant women, which could also influence BMI and body weight outcomes during pregnancy.

It is widely acknowledged that blood lipid metabolism changes during pregnancy to support normal fetal growth and development, with mildly elevated lipid levels considered normal in pregnant women [23]. In a related study, eight weeks of Pilates training reduced cholesterol and LDL levels in obese women [24], potentially mitigating mild lipid disturbances in pancreatic β-cells by compensating for increased insulin resistance. However, excessive lipid levels can surpass the compensatory capacity of pancreatic islet β-cells, leading to gestational diabetes. The results of this study showed no statistically significant (SMD − 1.182; 95% CI, -2.443 to 0.079; P = 0.066) increase in LDL levels in pregnant women in the Pilates and control groups, which was insufficient to demonstrate that Pilates exercise during pregnancy effectively reduces LDL levels. Consistent with the findings of a study by Vasconcelos et al. [25], indicating that LDL levels are influenced by factors beyond exercise intervention.

Neonatal weight

Neonatal weight serves as a crucial indicator of the intrauterine environment and maternal-infant health. The results of this study indicated that there was no statistically significant difference in neonatal weight between the Pilates and control groups (SMD − 0.22; 95% CI, -0.55 to 0.11; P = 0.196), suggesting that Pilates exercise during pregnancy does not lead to a reduction in neonatal weight. In a prospective randomized controlled study by Clapp et al., infants born to women who engaged in high-intensity exercise from the first to the second trimester had an average birth weight 460 g higher than infants born to control women, indicating a potential association between high-intensity exercise and increased risk of macrosomia [26]. Conversely, a meta-analysis of 9 randomized controlled trials demonstrated that moderate-intensity exercise (35–90 min per session, 3–4 times per week) among healthy pregnant women did not affect neonatal weight or increase risks related to preterm birth or neonatal outcomes. This suggests that while exercise intensity during pregnancy may influence neonatal weight, the moderate-intensity Pilates regimen was insufficient to produce a significant impact on neonatal weight.

The APGAR score assesses a newborn’s immediate physiological status postnatally, evaluating reflexes, respiration, muscle tone, and skin color, with scoring practices varying across countries [27]. A meta-analysis examining exercise interventions during pregnancy reported that a combination of aerobic, body sculpting, resistance, strength, and flexibility exercises improved APGAR scores at 1 min but had no significant effect at 5 min [28]. In line with these findings, the meta-analysis from our study revealed a statistically significant difference in 1-minute APGAR scores between the Pilates and control groups (SMD 0.41; 95% CI, 0.08 to 0.74; P = 0.015), indicating that Pilates exercise during pregnancy enhances the newborn’s tolerance during delivery. However, there was no statistically significant difference in 5-minute APGAR scores between the two groups, suggesting inconclusive evidence regarding Pilates’ impact on the newborn’s adaptation to the new environment post-birth. These results align with prior research suggesting that pregnant women engaged in Pilates may experience improved [29].

The duration of labor

Numerous studies have documented the impact of exercise interventions during pregnancy on the duration of labor and their significant association with maternal and infant health outcomes [30]. Prolonged first stage labor can potentially compress the fetal head in the birth canal, leading to decreased APGAR scores or stillbirth, while an extended second stage labor increases risks of perineal tear, cesarean section, and PFM injury. During the second stage of labor, maximal contraction of the diaphragm increases intrauterine pressure, facilitating the descent of the fetal head through the PFM [31]. The results of a meta-analysis in this study indicated a statistically significant difference in the time to the second stage of labor between the Pilates exercise group and the control group (SMD − 0.557; 95% CI, -0.915 to -0.198; P = 0.002). This finding is consistent with the results reported by Salvesen [32], which suggest that Pilates exercises can shorten the duration of the second stage of labor. This effect may be attributed to its impact on the central nervous system and muscle function. Pilates training during pregnancy has been shown to improve muscle control, and flexibility, as well as enhance pelvic and trunk stability [32].

The mode of delivery

The findings from the study suggested that Pilates exercises may enhance the likelihood of normal labor, there was no statistically significant difference in the mode of delivery between the Pilates and control groups (RR 1.41; 95% CI, 0.95 to 2.09; P = 0.093). This outcome could potentially be attributed to the relatively short duration and lower volume of Pilates exercise interventions administered during pregnancy. Previous research suggests that the influence of exercise on the mode of delivery can vary based on the type and frequency of exercise [33, 34]. For instance, aerobic endurance and resistance training performed 3 days per week have not significantly affected the mode of delivery [35], whereas exercise interventions exceeding 50 h throughout pregnancy have been associated with reduced rates of cesarean delivery [36].

The management of labor pain typically involves the use of painkillers and anesthetics, which effectively reduce maternal pain [37]. However, their administration can pose risks to both the mother and fetus [38, 39]. Non-pharmacological methods for pain relief are generally considered safer and less harmful to both the mother and fetus, offering a beneficial alternative for mitigating pain during labor. A meta-analysis demonstrated a statistically significant difference in pain scores between Pilates and control groups during labor (SMD − 0.76; 95% CI, -1.45 to -0.07; P = 0.032), indicating that Pilates exercises can effectively alleviate labor pain. This finding is particularly relevant given that labor pain can be exacerbated by muscle separation and emotional stress associated with uterine stretching [40]. Furthermore, exercise has been shown to reduce labor pain among pregnant women [41].

Pregnancy can diminish PFM strength and contribute to herniated discs in women. These factors, combined with maternal hormonal changes during pregnancy, can lead to pelvic floor dysfunction [42]. Pelvic floor exercises have been demonstrated to reduce the duration of the second stage of labor in approximately one-eighth of pregnant women [43]. Additionally, Pilates exercises have shown greater effectiveness compared to other exercise modalities in reducing herniated discs among women [44]. The findings of this study further indicate that clinical Pilates training during pregnancy can enhance pelvic muscle strength and lumbar pelvic stability in comparison to standard care. However, due to the limited number of randomized controlled trials (RCTs) and the lack of long-term follow-up, the extent of these benefits is somewhat constrained. Further studies are needed to substantiate the impact of Pilates on lumbar pelvic stability in pregnant women.

A systematic review investigating the impact of prenatal physical activity on anxiety and depression during pregnancy and the postpartum period revealed that achieving at least 644 MET-minutes/week of exercise (equivalent to 150 min of moderate-intensity activities such as brisk walking, water aerobics, stationary cycling, or resistance training) can lead to moderate reductions in prenatal depression and anxiety [28]. 2 studies [16, 19] highlighted that moderate-intensity Pilates interventions led to increased metabolic energy expenditure, reduced sympathetic tone, and improved modulation of the sympathetic nervous system. These interventions also showed potential in alleviating prenatal depression and anxiety, as well as enhancing sleep quality in pregnant women. Nevertheless, given the limited research in this area, future studies should focus on exploring the psychological impacts of Pilates during pregnancy.

Strengths and limitations

This systematic review and meta-analysis has several strengths: (1) The study focused exclusively on the health benefits of Pilates for pregnant women and adhered strictly to the PRISMA reporting guidelines for systematic reviews. (2) Systematic searches were conducted across seven databases, and the study incorporated data from research published in Chinese. (3) The study thoroughly examined the effects of Pilates exercise during pregnancy, and the subsequent meta-analysis provides quantitative evidence to support these findings.

This review also has some limitations: (1) This systematic review focused exclusively on searching Chinese and English databases, and the exclusion of studies published in other languages may potentially lead to incomplete inclusion and publication bias. (2) Since the data included in this systematic review were sourced from study groups across various countries, there was unavoidable heterogeneity in the Pilates intervention program, including variations in gestational week, as well as differences in the intensity and duration of the Pilates interventions.

(3) Due to limitations in the information reported by the included studies, this study did not account for the effects of other potential confounders, such as dietary factors, psychological aspects, and sleep patterns.

BMI:

Body Mass Index

PFM:

Pelvic floor muscle

BFM:

Body fat mass

SMM:

Skeletal muscle mass

PBF:

Percent body fat

SMD:

Standard Mean Difference

OR:

Odds ratios

CI:

Confidence intervals

EPDS:

Edinburgh Postnatal Depression Scale

STAI:

The State-Trait Anxiety Inventory

NHP:

The Nottingham Health Profile

LDL:

Low-density lipoprotein

VAS:

The Visual Analogue Score

HDL:

High density lipoprotein

Download references

Y.L., L.Z., and L.L. conceptualized and designed this study. H.L. and Y.L. conducted database searches, data extraction, and analysis. Y.L. and Y.R. wrote the manuscript. L.Z. and X.D. revised the manuscript. All authors critically reviewed and approved the manuscript as submitted. Liuwei Zhang ([email protected]) and Lin Lin ([email protected]) are co-corresponding authors of this manuscript.

Correspondence to Liuwei Zhang or Lin Lin.

Not applicable.

Not applicable.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions