Exercise effects on maternal vascular health and blood pressure during pregnancy and postpartum: a systematic review and meta-analysis

Abstract

Aims

This systematic review aimed to assess the effects of exercise training during pregnancy and the postpartum period on maternal vascular health and blood pressure (BP).

Methods and results

The outcome of interest was pulse wave velocity (PWV), flow-mediated dilation (FMD), and BP from pregnancy to 1-year postpartum. Five databases, including Ovid MEDLINE, EMBASE, CINAHL, Web of Science, and Cochrane Library, were systematically searched from inception to August 2023. Studies of randomized controlled trials (RCTs) comparing the effects of prenatal or postpartum exercise to a non-exercise control group were included. The risk of bias and the certainty of evidence were assessed. Random-effects meta-analyses and sensitivity analyses were conducted. In total, 20 RCTs involving 1221 women were included. Exercise training, initiated from Week 8 during gestation or between 6 and 14 weeks after delivery, with the programme lasting for a minimum of 4 weeks up to 6 months, showed no significant impact on PWV and FMD. However, it resulted in a significant reduction in systolic BP (SBP) [mean difference (MD): −4.37 mmHg; 95% confidence interval (CI): −7.48 to −1.26; P = 0.006] and diastolic BP (DBP) (MD: −2.94 mmHg; 95% CI: −5.17 to −0.71; P = 0.01) with very low certainty. Subgroup analyses revealed consistent trends across different gestational stages, types of exercise, weekly exercise times, and training periods.

Conclusion

Exercise training during pregnancy and the postpartum period demonstrates a favourable effect on reducing maternal BP. However, further investigations with rigorous methodologies and larger sample sizes are needed to strengthen these conclusions.

Graphical Abstract

Open in new tabDownload slide

Introduction

Pregnancy imposes substantial demands that require intricate maternal physiological adaptations, especially within the cardiovascular system.¹ These haemodynamic changes, such as increased cardiac output, reduced systemic vascular resistance, decreased blood pressure (BP), and the remodelling of the heart and vasculature, are essential to support the heightened metabolic needs of both the mother and the growing fetus.^1–3 However, maladaptation within this system may increase the vulnerability of pregnant women to hypertensive disorders of pregnancy (HDP), with far-reaching implications for maternal and neonatal health.⁴

Research suggests that maintaining optimized BP control following pregnancies can aid in the recovery of the cardiovascular system postnatally, persistently reducing BP and positively impacting cardiac remodelling after childbirth.^5,6 Failure to manage BP during gestation and the postpartum period may significantly cause cardiovascular diseases later in life, particularly in women who experienced HDP during pregnancy.^6,7 Additionally, controlling BP during pregnancy and in the immediate post-natal period not only benefits the mother's long-term cardiovascular health but also crucial in preventing or reversing target organ damage caused by HDP, which tends to persist despite fluctuations in BP levels throughout the perinatal and postpartum phases.⁸ Implementing early strategies during the peripartum period may be necessary to mitigate or reverse high BP and the risk of lingering multiple organ damage associated with hypertensive pregnancies in women.

Exercise intervention has been established as a safe and effective non-pharmacological strategy for managing vascular health and BP alterations in non-pregnant individuals.^9,10 Recent meta-analyses have underscored the positive impact of prenatal exercise training in reducing maternal BP and the associated risk of HDP.^11–16 However, a critical gap exists in the current literature as these analyses have not comprehensively explored BP responses across various maternal health conditions, gestational ages, and specific components of exercise programmes. Additionally, there is a notable absence of evidence attempting to collectively assess the influence of exercise interventions on the vascular health in pregnant women. Therefore, this systematic review and meta-analysis aimed to synthesize the effects of physical exercise training during pregnancy and the postpartum period, compared with non-exercise interventions, on vascular structure [pulse wave velocity (PWV) and brachial arterial diameter (BAD)], endothelial function [flow-mediated dilation (FMD)], and arterial BP [systolic BP (SBP), diastolic BP (DBP), or mean arterial pressure (MAP)].

Methods

This review was performed in accordance with the Cochrane Handbook for Systematic Reviews of Interventions¹⁷ and is reported in keeping with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.¹⁸ The protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO; Registration No. 471449). The data underlying this article will be shared on reasonable request to the corresponding author.

Literature search

The database search strategy was developed by a librarian with knowledge and expertise in conducting searches for systematic reviews. Five electronic databases, including Ovid MEDLINE, EMBASE, CINAHL, Web of Science, and Cochrane Library, were searched from inception up to 20 August 2023, with no language restrictions. The search terms were related to exercise, physical activity, pregnancy, postpartum, PWV, FMD, and BP. To identify further relevant publications, the reference citations of retrieved studies were examined. Conference abstracts were not considered in this review. The complete search strategy is presented in Supplementary material, Table S1.

After identifying and removing duplicates via Covidence systematic review software,¹⁹ two reviewers (K.P. and M.B.) independently screened the title and abstract. Subsequently, the eligibility of the articles was evaluated by assessing the full text. Any disagreements were resolved through consensus or, if needed, with the advice of a third reviewer (J.S.).

Eligibility criteria

Based on the PICOS (participants, intervention, comparison, outcome, and study design) framework, the following inclusion criteria were established: (P) pregnant and postpartum women; (I) participation in pre- or post-natal exercise training lasting at least 4 weeks, with or without co-interventions; (C) absence of a formal exercise intervention; (O) assessment of vascular structure (PWV or BAD), endothelial function (FMD), or arterial BP (SBP, DBP, or MAP); (S) a randomized controlled trial (RCT) design.

Study selection and data extraction

Two reviewers (K.P. and M.B.) independently extracted information from each included study including study characteristics (authors, publication year, study location, and number of participants), maternal health status, gestational age or postpartum period at randomization, exercise characteristics (frequency, intensity, time, type, duration, co-intervention, and comparison), and outcomes (PWV, BAD, FMD, SBP, DBP, and MAP). Disagreements were resolved by discussion. Mean and standard deviation (SD) for the change from baseline of each outcome were extracted or calculated following the Cochrane guidelines outlined in section 6.5.2.2.¹⁷

Quality assessment

The risk of bias assessments was completed independently by two reviewers (K.P. and M.B.) using the criteria outlined in Cochrane Risk of Bias 2 Tool for RCT.²⁰ This tool is structured into five bias domains, including randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. Each domain was rated as having a low, high, or some concerns risk of bias.

Certainty assessment

The certainty of evidence was evaluated independently by two reviewers (K.P. and M.B.) using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) tool.²¹ Evidence from the included studies was downgraded if concerns arose regarding the risk of bias, inconsistency, indirectness, imprecision, or publication bias. Inconsistency was deemed significant when there was high heterogeneity or when only one study was evaluated. Indirectness was observed when studies compared alternative interventions, populations, comparators, or outcomes. Imprecision was determined when the 95% confidence interval (CI) crossed the line of no effect or using the optimal information size (OIS) approach²² if the CI did not overlap with the threshold of interest. Publication bias was detected through the presence of asymmetry in funnel plots.²³

Data analyses

Meta-analyses were conducted using Review Manager 5.4.1 software (Cochrane Collaboration, Copenhagen, Denmark). A random-effects model was used to calculate variance-weighted pooled mean differences (MD) and 95% CI for outcome changes between the exercise training and control groups. The Higgins I² statistic was used to determine statistical heterogeneity across studies and was categorized by low (I² < 25%), moderate (I² 25–75%), and high (I² > 75%).²⁴ If P < 0.05 and I² > 75%, leave-one-out sensitivity analysis was further determined to assess results stability and potential sources of heterogeneity.²⁵

To determine if there were specific effects of prenatal and postpartum exercise on each outcome of interest when three or more studies were pooled, subgroup analyses were performed and segmented as follows: maternal health condition (healthy, sedentary/inactive, overweight/obese, or had HDP), gestational age or postpartum period at randomization (first, second, third trimester, or postpartum), type of exercise (aerobic, resistance, yoga/pilates, or combined exercises), exercise times per week (<120, 120–149, 150–179, or ≥ 180 min/week), exercise duration (<8, 8–14, or >14 weeks), and exercise intensity (low or moderate levels).

Trial sequential analysis

Systematic reviews and meta-analyses with limited studies and small sample sizes can increase the risk of false-positive results, potentially leading to biased conclusions.²⁶ To reduce this risk, trial sequential analysis (TSA) for outcome indicators was conducted using TSA software (version 0.9.5.10 Beta, Copenhagen Trial Unit, Copenhagen, Denmark), with the probability of Type I error set at α = 0.05, the statistical power of 80%, and sample size at the required information size (RIS). If the cumulative Z value reached both conventional and TSA cut-off boundaries, it indicated that the corrected results were consistent and could be considered conclusive evidence. However, if the Z curve did not intersect with any boundaries, definitive conclusions could not be drawn.²⁷

Results

Search results

A PRISMA flowchart of the literature search is shown in Figure 1, and a complete PRISMA checklist is provided in the Supplementary material (see Supplementary material, Table S2). The initial search identified 2080 records. A total of 309 duplicate papers were removed, with 1738 articles excluded after title and abstract screening. Full texts of 33 citations were retrieved, and 20 records met the inclusion criteria. The study by Stutzman et al.²⁸ reported two separate samples. Therefore, 19 distinct RCTs^28–46 were included in the qualitative synthesis, and 18 of these^28–44,46 were selected for meta-analyses. A list of 13 full-text records excluded with reasons is presented in Supplementary material, Table S3.

Study characteristics

Out of the 19 studies included, 17 (89.47%)^{29,30,32–46} were parallel RCTs, while 2 (10.53%)^28,31 were factorial designs. These studies were published between 2010 and 2023, in 8 different countries, with 11 taking place in Europe (Spain n = 6, Norway n = 2, Sweden n = 2, and Germany n = 1),^{31–34,37–39,41–43,46} 5 in America (Canada n = 4, Colombia n = 1),^{28,30,40,44,45} and 3 in Asia (Iran n = 2, India n = 1).^29,35,36 In total, 1221 participants were involved, with 621 in the experimental group and 600 in the control group. Maternal health status was addressed in nine studies (47.37%) with physically active participants,^{29,37–40,42–44,46} five studies (26.32%) with sedentary behaviour,^{28,30,32,34,45} three studies (15.79%) involving higher body mass index,^28,31,33 and three studies (15.79%) associated with HDP.^35,36,41 Exercise training was conducted during pregnancy in 17 (89.47%) studies,^{28–30,32–40,42–46} with gestational ages ranging from 8 to 34 weeks. Postpartum exercise training was implemented in two studies (10.53%),^31,41 commencing between 6 and 14 weeks after delivery. Among these studies, no adverse events or other exercise-related injuries were reported in only four studies (21.05%).^33,34,39,42 The most frequently reported outcomes were SBP and DBP (n = 18 studies, 94.74%),^28–44,46 followed by MAP (n = 5 studies, 26.32%),^{30,40,41,44,46} PWV (n = 2 studies, 10.53%),^41,45 BAD (n = 1 study, 5.26%),⁴⁰ and FMD (n = 1 study, 5.26%).⁴⁰ Further details of all included studies are presented in Table 1.

Characteristics of exercise interventions

The exercise interventions used in each trial are summarized in Table 1.

Type of exercise training

During pregnancy, aerobic exercise was performed in six studies,^{28,30,35,40,44,45} resistance training in one study,³⁹ yoga/pilates in two studies,^36,42 and a combination of aerobic and resistance exercises with or without additional coordination/balance training in eight studies.^{29,32–34,37,38,43,46} For postpartum programmes, one study focused on aerobic walking workouts,³¹ and another study incorporated a combination of resistance and ergometer training.⁴¹

Intensity of exercise training

The intensity of prenatal and postpartum aerobic exercises was reported in all except three studies.^{35, 41,46} Monitoring methods included percentages of age-predicted maximum heart rate, HR reserve, and rating of perceived exertion. The most common intensity for pregnancy exercise was moderate to vigorous (n = 5 studies),^{30,33,34,43–45} followed by moderate (n =4 studies),^29,32,37,38 low (n = 1 study),²⁸ and low to moderate (n = 1 study).⁴⁰ One postpartum exercise study used moderate intensity.³¹

Frequency of exercise training

Exercise training frequency ranged from one to five sessions per week. The most common prescription for prenatal programmes was three sessions per week (n = 8 studies),^{29,32,33,36–38,40,43} followed by two sessions (n = 4 studies),^34,39,42,46 three to four sessions (n = 3 studies),^30,44,45 five sessions (n = 1 study),²⁸ and four sessions (n = 1 study).³⁵ Postpartum programmes included sessions conducted once⁴¹ or four ³¹ times per week.

Duration of exercise training

Total training duration varied between 4 and 30 weeks. Most pregnancy exercise programmes lasted more than 14 weeks (n = 11 studies),^{28,30,32,33,35,37,38,40,43–45} while some lasted equal to or less than 8 weeks (n = 4 studies) ^29,36,42,46 or 9–14 weeks (n = 2 studies).^34,39 For postpartum programmes, the durations were either 12 ³¹ or 24⁴¹ weeks.

Quality assessment

The methodological quality assessment is shown in Supplementary material, Figure S1. Among the included studies, 4 were evaluated as having a low risk of bias, while 11 trials were considered to have a high risk of bias, and 5 trials were rated with some concerns. The primary reason for categorizing these studies as having a high risk or being concerning was the inadequate assurance of sequence generation and allocation confidentiality, assignment to intervention, and handling of missing outcome data, which were addressed in only 30%, 30%, and 45% of the studies, respectively.

Certainty assessment

Table 2 describes the results of the certainty assessment. The overall certainty of evidence for exercise training during pregnancy and the postpartum period was very low. The reasons for downgrading were a serious risk of bias (overall risk of bias was high or some concerns), inconsistency (heterogeneity was high or only one study was included), indirectness (exercise-only interventions and exercise with co-intervention were combined), imprecision (the total sample size was smaller than 50% of the OIS or only one study was included), and publication bias (see Supplementary material, Figures S2 and S3).

Synthesis of the findings of the included studies

Pulse wave velocity

Two studies^41,45 were included, reporting a non-significant difference in central (carotid-femoral) PWV changes between the exercise intervention and control groups (n = 69; MD: −0.12 m/s; 95% CI: −0.43 to 0.18; P = 0.44; I² = 53%; P = 0.15; Table 2). This evidence was deemed to have a very low level of certainty due to a high risk of bias, indirectness, and imprecision.

Brachial artery diameter

Only one study⁴⁰ was included, reporting a significant difference in BAD changes between the exercise intervention and control groups (n = 50; exercise group mean ± SD: 0.31 ± 0.08 mm; control group mean ± SD: 0.24 ± 0.09 mm; P < 0.01; Table 2). This evidence was considered to have a very low level of certainty because of a very high risk of bias and inconsistency.

Flow-mediated dilation

Only one study⁴⁰ was included, reporting a non-significant difference in FMD changes between the exercise intervention and control groups (n = 50; mean ± SD of exercise group: −2.61 ± 2.25%; mean ± SD of control group: −2.20 ± 1.85 m/s; P = 0.490; Table 2). This evidence was deemed to have a very low level of certainty due to very high risk of bias and inconsistency.

Systolic blood pressure

Significant differences in SBP changes were observed between the exercise intervention and control groups (19 records; n = 1170),^28–44,46 with statistically significant heterogeneity (MD: −4.37 mmHg; 95% CI: −7.48 to −1.26; P = 0.006; I² = 99%; P < 0.001; Table 2 and Figure 2). This evidence was considered to have a very low level of certainty due to high risk of bias, inconsistency, and indirectness.

Sensitivity analysis, shown in Supplementary material, Table S4, reveals that the effect size and heterogeneity remained unchanged (MD: −8.38 mmHg; 95% CI: −10.87 to −5.90; P < 0.001; I² = 97%; P < 0.001) after removing eight studies^{28,31,32,37–40,43} that exerted a high degree of influence on the overall effect size. The leave-one-out analysis demonstrated that excluding any of the studies had no noticeable impact on the overall effect size.

The subgroup analysis illustrated significant differences between groups based on maternal health condition, gestational age, type of exercise, weekly exercise times, and training duration. Significantly lower SBP levels were observed in women who were sedentary/inactive (MD −4.80 mmHg; 95% CI: −9.41 to −0.19; P = 0.004), overweight/obese (MD −6.80 mmHg; 95% CI: −10.33 to −3.28; P < 0.001), and had a history of HDP (MD −6.99 mmHg; 95% CI: −7.48 to −1.26; P < 0.001; Figure 3). Similar trends were found in women in the second trimester (MD −6.96 mmHg; 95% CI: −11.38 to −2.54; P < 0.001), third trimester (MD −4.94 mmHg; 95% CI: −5.93 to −3.95; P < 0.001), and postpartum period (MD −7.22 mmHg; 95% CI: −13.61 to −0.84; P = 0.03; Figure 2). Additionally, women engaged in yoga/pilates (MD −8.62 mmHg; 95% CI: −15.81 to −1.43; P = 0.02) or combined aerobic training with other types of exercise (MD −3.62 mmHg; 95% CI: −7.14 to −0.10; P = 0.04; Figure 4), exercised for less than 120 min per week (MD −11.58 mmHg; 95% CI: −14.10 to −9.05; P < 0.001; Supplementary material, Figure S4), and participated for a total of 4 to 8 weeks (MD −8.91 mmHg; 95% CI: −13.71 to −4.12; P < 0.001) or longer than 14 weeks (MD −2.98 mmHg; 95% CI: −5.82 to −0.14; P = 0.04; Supplementary material, Figure S5) showed significant reductions in SBP. However, there were no significant differences between groups based on exercise intensity (see Supplementary material, Figure S6). Interestingly, in the subgroup analysis, exercise interventions based on resistance exercise during pregnancy were associated with higher SBP (MD +2.00 mmHg; 95% CI: 0.65 to 3.35; P = 0.004; Figure 4).

Trial sequential analysis showed that the cumulative Z curve was in the upper part of the window (favours exercise), crossed both the conventional and monitoring boundaries, stayed outside the futility boundary, but did not reach the appropriate RIS (see Supplementary material, Figure S7).

Diastolic blood pressure

Significant differences in DBP changes were observed between the exercise intervention and control groups (19 records; n = 1170),^28–44,46 with statistically significant heterogeneity (MD: −2.94 mmHg; 95% CI: −5.17 to −0.71; P = 0.01; I² = 99%; P < 0.001; Supplementary material, Figure S8). This evidence was deemed to have a very low level of certainty due to high risk of bias, inconsistency, and indirectness.

Sensitivity analysis, shown in Supplementary material, Table S5, revealed that the effect size and heterogeneity remained unchanged (MD: −6.71 mmHg; 95% CI: −9.72 to −3.71; P < 0.001; I² = 99%; P < 0.001) after removing 10 studies^{28–33,37–39,43} s degree of influence on the overall effect size. The leave-one-out analysis demonstrated that excluding any of the studies had no noticeable impact on the overall effect size.

Further subgroup analyses of DBP by gestational age, type of exercise, weekly exercise times, and duration found notable distinctions. Significantly lower DBP levels were observed in women who were in the second trimester (MD −6.09 mmHg; 95% CI: −8.72 to −3.46; P < 0.001), third trimester (MD −1.86 mmHg; 95% CI: −2.47 to −1.25; P < 0.001), and postpartum period (MD −5.43 mmHg; 95% CI: −9.89 to −0.97; P = 0.02, Supplementary material, Figure S8). Similarly, women with HDP (MD −5.57 mmHg; 95% CI: −10.32 to −0.82; P = 0.02, Supplementary material, Figure S9), engaged in aerobic exercise (MD −4.35 mmHg; 95% CI: −7.52 to −1.18; P = 0.007; Supplementary material, Figure S10), exercised for less than 120 min per week (MD −7.99 mmHg; 95% CI: −10.14 to −5.84; P < 0.001; Supplementary material, Figure S11), and participated for a total of 4 to 8 weeks (MD −5.79 mmHg; 95% CI: −11.55 to −0.04; P = 0.05; Supplementary material, Figure S12) also showed a significant reduction in DBP. As observed with the SBP, exercise intensity has not significantly changed DBP during pregnancy and the postpartum period (see Supplementary material, Figure S13).

Trial sequential analysis showed that the cumulative Z curve was in the upper part of the window (favours exercise), crossed both the conventional and monitoring boundaries, stayed outside the futility boundary, but did not reach the appropriate RIS (see Supplementary material, Figure S14).

Mean arterial blood pressure

Significant differences in MAP changes were observed between the exercise intervention and control groups (five records; n = 203),^{30,40,41,44,46} with statistically significant heterogeneity (MD: −4.77 mmHg; 95% CI: −8.15 to −1.39; P = 0.006; I² = 93%; P < 0.001; Supplementary material, Figure S15). This evidence was considered to have a very low level of certainty because of a high risk of bias, inconsistency, indirectness, and imprecision.

Sensitivity analysis, shown in Supplementary material, Table S6, revealed that the effect size and heterogeneity remained unchanged (MD: −6.25 mmHg; 95% CI: −9.46 to −3.04; P < 0.001; I² = 87%; P < 0.001) after removing a study⁴⁰ that exerted a high degree of influence on the overall effect size. The leave-one-out analysis demonstrated that excluding any of the studies had no noticeable impact on the overall effect size.

Mean arterial pressure subgroup analyses and TSA were not performed due to limited data from the included studies.

Discussion

In this systematic review and meta-analysis, data from 20 records involving a total of 1221 women were pooled and synthesized to investigate the effect of exercise training during pregnancy and the postpartum period on maternal vascular health and BP. The included studies exhibited heterogeneity and a very low level of certainty. Statistically significant effects of exercise interventions, compared with the control, were observed for maternal BP. However, the analysis did not reveal significant effect of exercise training on outcomes related to vascular structure and function when compared with the control group.

Physical activity has demonstrated its efficacy in improving vascular function in non-pregnant individuals⁴⁷ and post-menopausal women with hypertension.⁴⁸ Additionally, exercise training has proven to be a vital intervention in mitigating the adverse vascular effects of pregnancy-induced hypertension^14–16 and diabetes mellitus.⁴⁹ Despite these well-established benefits, previous reviews have not thoroughly examined the impact of exercise training on vascular structural and functional changes in pregnant and postpartum women.

The limited evidence available in this review, particularly regarding changes in PWV and FMD in response to exercise intervention, revealed no significant differences between exercise and control groups. The very low certainty findings, constrained by a high risk of bias and inconsistency, raise questions about the potential impact of exercise on central arterial stiffness and endothelial function in large vessels in women. In normal pregnancies, PWV experiences a significant decreases in the second trimester, followed by an increase from the third trimester through immediate post-delivery,⁵⁰ while FMD exhibits an opposite pattern, with an increase during the first two trimesters followed by a significant decrease from 36 weeks onward.⁵¹ The lack of significant findings suggests that exercise training may not have a substantial benefit in terms of lowering PWV and enhancing FMD in women during pregnancy and after giving birth. This could potentially be attributed to several factors, including the relatively short exercise time per week (e.g. 120–150 min/week), overall duration of exercise interventions (e.g. >14 weeks), specific exercise prescriptions used in the studies (e.g. unsupervised aerobic exercise at home), or the characteristics of the study population (e.g. normotensive pregnant women, second trimester).

Another measure of vascular structural alteration, BAD, was assessed in one study during pregnancy, and the exercise group showed a greater increase in BAD.⁴⁰ This may suggest potential benefits of prenatal exercise for peripheral vascular structure in early-term pregnancy. However, the overall certainty of evidence was considered very low due to the high risk of bias and the limited number of available studies for analysis. Caution must be taken in the interpretation of the exercise effect on this variable.

The comprehensive meta-analysis examining BP responses during pregnancy and the postpartum period following exercise interventions revealed a consistent trend of positive influence on maternal haemodynamic. Aggregating results from SBP, DBP, and MAP demonstrated significant reductions in overall BP levels. Even though substantial heterogeneity exists, sensitivity analyses confirm the robustness of these findings, enhancing the credibility of the observed BP changes attributed to exercise interventions. These findings are in line with the results from previous meta-analyses.^11,12 However, it is noteworthy that the present study primarily assessed exercise effectiveness by analysing differences in changes from baseline, unlike prior reviews that mainly relied on post-intervention values¹¹ and did not predominantly draw evidence from RCTs.¹² Moreover, the current meta-analysis included seven recent studies^{29,30,32,35,36,43,44} and postpartum study.

The subgroup analyses also provide valuable insights into the heterogeneity of exercise effects across different maternal characteristics and specific components of exercise programmes, underscoring the need for personalized exercise prescriptions. For instance, women characterized as sedentary/inactive, overweight/obese, or with a history of HDP demonstrated more pronounced reductions in BP following exercise interventions. This implies that the impact of exercise on BP may be influenced by the pre-existing health conditions of pregnant and postpartum women. Moreover, BP responses varied across different gestational ages, with women in the second trimester, third trimester, and postpartum period experiencing differing levels of reduction. This highlights the potential influence of timing in exercise interventions during the course of pregnancy and postpartum recovery. Although the benefits of early postpartum exercise rehabilitation have been demonstrated for maternal outcomes such as postpartum weight reduction and alleviation of depression,^52,53 the presence of common peripartum musculoskeletal issues (e.g. low back or pelvic girdle pain and diastasis recti abdominis) may reduce women's participation, adherence, and physical responses to aerobic and resistance exercise training during the fourth trimester (i.e. from childbirth up to 12 weeks after delivery).^54,55 Therefore, addressing peripartum musculoskeletal disorders as early as possible during the fourth trimester could be a practical measure to enhance women's participation and improve physical benefits from exercise training in the postpartum period.

The type of exercise also played a crucial role, as women engaged in yoga/pilates or combined aerobic training with other exercise types exhibited substantial reductions in BP, while resistance exercise during pregnancy was associated with higher SBP. This underscores the importance of tailoring exercise programmes to the specific type of activity. Lastly, exercise duration and frequency emerged as critical factors, with significant BP reductions observed in women exercising for less than 120 min per week and participating for 4 to 8 weeks or longer than 14 weeks. This suggests that the design of effective exercise interventions should consider both the duration and frequency of sessions.

Regarding the TSA results, it is suggested that additional trials and participants are needed before conclusive statements about the effects of exercise training during pregnancy and the postpartum period on lowering BP responses can be made.

Strengths and limitations

This systematic review and meta-analysis have several strengths. The use of TSA enhanced the precision and clarified the uncertainty surrounding the meta-analysis results, enabling a more comprehensive assessment of their conclusiveness. Furthermore, the inclusion of subgroup analyses provided valuable insights into the potential moderators influencing the effects of exercise on BP. However, there are some limitations to consider. Firstly, the available evidence is based on a relatively small number of studies, and not all outcomes of interest could be analysed due to the limited data. Secondly, there was high heterogeneity in most of the meta-analyses, which could be attributed to differences in exercise protocols and/or participant characteristics. Thirdly, no included studies specifically addressed the effects of medication or morbidity in the cohorts under consideration. Fourthly, only one study³⁴ properly estimated an exercise effect by performing the mean adjusted difference in the change in BP from baseline to after the intervention. Finally, the risk of bias in some studies, particularly those involving no blinding and allocation concealment, may have influenced the overall certainty of evidence.

Perspectives

To advance our understanding and optimize the benefits of exercise for pregnant and postpartum women, future studies should explore tailored exercise prescriptions based on individual characteristics, including maternal health condition, gestational age and/or postpartum period, and exercise preferences. Moreover, a focus on well-designed RCTs with larger sample sizes and rigorous blinding procedures is crucial. These efforts will contribute significantly to our comprehension of the role of exercise in enhancing maternal vascular health and preventing hypertension during and after pregnancy.

Conclusions

The present systematic review and meta-analysis suggest a positive impact of prenatal and postpartum exercise training on reducing maternal BP. However, it is important to interpret these findings with caution, particularly concerning the postpartum period and individuals at risk of or experiencing hypertension. The limited number of studies and small sample sizes of the included trials indicate the need for further investigation. Future research should prioritize well-designed RCTs with larger sample sizes to strengthen these conclusions and provide clearer clinical recommendations.

Supplementary material

Supplementary material is available at European Journal of Preventive Cardiology.

Acknowledgements

The authors thank Jill Boruff for assistance with the literature search. The data underlying this article are available in the article and in its Supplementary material.

Author contribution

K.P. led the literature search, data extraction, and analysis. M.B. contributed to the literature search, data extraction, and analysis. K.P. and M.B. drafted the manuscript. N.D., T.J.-F., M.R., and J.S. provided substantive guidance and revised the manuscript. M.B. supervised all aspects of the review and revised the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

The review was supported through a Heart and Stroke Foundation of Canada Grant-in-Aid (M.B., reference number G-23–0033377). K.P. is supported by a J.A. DeSève scholarship from the CIUSSS Nord-de-l’Île-de-Montréal and the Royal Thai Government scholarship.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

References

Author notes