If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Healthcare Administration, Asia University, Taichung, TaiwanDepartment of Medical Research, China Medical University Hospital, Taichung, Taiwan
Department of Healthcare Administration, Asia University, Taichung, TaiwanDepartment of Medical Research, China Medical University Hospital, Taichung, Taiwan
Correspondence to: Shao-Huan, Lan, PhD, School of Pharmaceutical Sciences and Medical Technology, Putian University, 1133 Xueyuan Road, Chengxiang District, Putian, Fujian Province, 351100, PR China.
Purpose: Sleep quality was considered a priority concern facing pregnant women. Conventional wisdom argues that good sleep quality benefits pregnant women and their fetuses. The aim of this study is to assess the effects of a specific exercise program on the sleep quality in pregnant women.
Methods: Searches were executed in seven databases since their inceptions until February 28, 2019, for randomized controlled trials evaluating the effects of an exercise program on the sleep quality and insomnia in pregnant women. A random-effects model was applied for meta-analysis, and odds ratio, mean differences (MDs), and 95% confidence intervals (CIs) are shown as parts of outcomes.
Results: Seven studies were included for meta-analysis. Compared with their not-exercising counterparts, analyses showed that regularly exercising women had significantly enhanced sleep quality, with an odds ratio of 6.21 (95% CI, 2.02–19.11;p = .001; I2 = 80.2%), with a standardized MD of −0.93 (95% CI, −1.19 to −0.67; p < .001; I2 = 30.0%). However, exercising women showed no significant insomnia improvement, with an standardized MD of −2.85 (95% CI, −7.67 to 1.98; p = .250; I2 = 97.0%), relative to their not-exercising counterparts.
Conclusion: This research indicated that exercise has a positive impact on the sleep quality of pregnant women. Despite the aforementioned positive impact on sleep quality, the present study did not find evidence to support that exercise may also improve insomnia for pregnant women.
Sleep disturbance is quite common in pregnant women. About 76% of pregnant women experienced poor sleep quality throughout all trimesters, and 57.3% of pregnant women experience subthreshold insomnia throughout all trimesters [
]. Many pregnant women experience frequent poor sleep quality, nighttime awakening, insomnia, insufficient nighttime sleep, and significant daytime sleepiness during their pregnancy [
]. Based on these associations, the authors believe that the improvement in sleep could presumably result in better pregnancy outcomes. Furthermore, an insomnia drug therapy has several side effects: preterm birth, preterm deliveries, cesarean deliveries, low birth weight, and even delivery of small-for-gestational-age infants [
]. When reviewing individual studies, the effect of physical activity or exercise on sleep quality and insomnia during pregnancy is inconclusive. Therefore, the authors conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) to assess and validate the effects of exercise programs on quality of sleep and insomnia.
Methods
This study used Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols 2015 statements in constructing the structure of this review [
Electronic literature searches were conducted in seven databases MEDLINE, PubMed, Cumulative Index to Nursing and Allied Health Literature, Cochrane, Excerpta Medica database (Embase), Chinese National Knowledge Infrastructure, and Airiti Library since their inceptions through February 28, 2019. Corresponding Chinese terms were used for searching in Chinese databases (Chinese National Knowledge Infrastructure and Airiti Library). The search terms used were “pregnan*” or “gestation” or “prenatal” and “sleep” or “insomnia” and “exercise” or “sport” or “physical activity” or “yoga” or “tai chi.” Manual searches were also performed on retrieved articles for additional references.
Inclusion/exclusion criteria and study selection
A PICOS (Participants, Intervention, Comparison, Outcomes) tool, composed of participant, intervention, comparison, outcomes, and study setting, was used as selection criteria to develop an effective strategy [
]. As per the design, participants were pregnant women, and the intervention was exercise regardless of its types or forms (e.g., aerobic exercise, stretching and relaxation, yoga, or tai chi) versus nonactive intervention (e.g., education or usual care) as a comparison. Quality of sleep or insomnia was the outcome, and for the study setting, only RCTs were assessed.
The following studies were excluded: (1) prospective cohort studies, (2) numerical data not provided or specified on specific tools, and (3) studies with participants reported to have acute complications during the courses of exercises, such as vaginal bleeding, amniotic fluid leakage, or regular painful contractions.
Data extraction
To ensure the objectivity of literature screening, two reviewers (S.-Y.Y. and S.-H.L.) independently screened titles, abstracts, and full-text journal articles. Citations considered potentially relevant to literature with titles or abstracts containing insufficient information were retrieved and further assessed, via applying the PICOS tool, by two independent reviewers (S.-Y.Y. and S.-H.L.) to determine eligibility for inclusion. In cases of disagreement over eligibility for inclusion between the two independent reviewers (S.-Y.Y. and S.-H.L.), a consensus was achieved by discussing and consulting with a third reviewer (S.-J.L.). Two reviewers (S.-Y.Y. and S.-H.L.) independently extracted data from the included studies. The following information was extracted: authors, the year of publication, the number of participants, age of the participants, nationality, pregnant body mass index, gestational age, the information about characteristics of exercise programs, and the information about characteristics of the outcome measurement. A consensus was achieved by discussing and consulting with a third reviewer (S.-J.L.).
Quality assessment of selected studies
The methodological quality of the selected RCTs was evaluated against the Cochrane risk of bias tool. Higgins et al. states that “the risk of bias tool covers six domains of bias: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other bias.” In evaluating any RCTs, each item was depicted as having either a low risk of bias, a high risk of bias, or an unclear risk of bias. The risk of bias was independently evaluated by two authors (S.-Y.Y. and S.-J.L.) by applying the Cochrane risk of bias tool [
The reviewers extracted data in connection with the study characteristics (PICOS criteria) by using the data form. Once verified for their exactness and completeness, the results were then analyzed by applying the Cochrane Collaboration Review Manager (RevMan) software program version 5.4 (Cochrane, London, UK). As for data synthesis of continuous variables, results of the individual studies were calculated as mean difference (MD) or standardized MD (SMD), with 95% confidence intervals (CIs). When the pooled trials used different rating scales, the absolute MD divided by the SMD was applied. For dichotomous variables, results of the individual studies were calculated as odds ratios (ORs), with 95% CIs [
]. The researchers deemed only randomized trials demonstrating clinical homogeneity to be potentially eligible for meta-analysis. As for pooled effects, heterogeneity was tested applying the Breslow–Day test, with p < .10 considered statistically significant. The I2 analysis is a useful statistic to quantify inconsistency: I2 = [(Q - df)/Q] × 100%, where Q is the χ2 statistic and df is its degrees of freedom [
]. The χ2 test was further applied to evaluate and quantify statistical heterogeneity across trials by using the I2 statistic (small, I2< 25%; moderate, I2 between 26% and 74%; and high, I2 ≥ 75%) [
]. If heterogeneity was observed (the value of heterogeneity being higher than or equal to 25%), the authors then applied a random-effects model. If the value of heterogeneity was lower than 25%, the authors applied a fixed-effects model [
A flow chart describing the literature extraction process as well as the criteria for inclusion and exclusion can be seen in Figure 1. As per initial search results, 461 records were identified from relevant databases and their reference lists, with 137 duplicate references being removed. Among the remaining 324 potentially eligible articles, 273 were excluded after screening their titles and abstracts, and in addition, four more non–full-text articles were further removed. During the eligibility screening stage, 40 of 47 full-text articles were excluded for not meeting the inclusion criteria. These remaining seven articles were included for meta-analysis [
Figure 1Flow chart describing literature extraction process. Note. CINAHL = Cumulative Index to Nursing and Allied Health Literature; CNKI = Chinese National Knowledge Infrastructure.
A total of 688 pregnant women were included as participants in seven RCTs. All studies focused on pregnant women without physical complications reported during pregnancy. Participants in one study were pregnant women with depression, whereas another study has involved pregnant women with complaints of insomnia or fatigue. The mean age for participants was around 26.00–32.12 years, with their pregnancy stages covering the first, second, and third trimesters. Of the seven studies included for analyses, three were conducted in China, one in the USA, one in Spain, one in Turkey, and the remaining one in Nigeria (Table 1).
Table 1Main Characteristics of the Included Studies in the Meta-analysis (N = 7).
Among studies included for meta-analysis, participants of three studies practiced yoga, whereas participants in the other four studies practiced multiple forms of exercise, including aerobic exercise, gymnastics training, tai chi, and relaxation exercise. The duration of exercise was between 4 and 16 weeks, with its frequency ranging from 1 to 3 sessions per week and duration for each session lasting as long as 20–60 minutes. In addition to the regular programmed exercise, participants of four studies also performed self-initiated exercise by themselves at home on a daily basis (Table 2).
Table 2Characteristics of Exercise Programs and Outcomes Assessment of the Studies Included in the Meta-analysis (N = 7).
The control group of one study was allocated to a waitlist childcare transportation and that of five studies was implemented with education and another one was in routine care (Table 1).
] reported no difference in insomnia conditions among participants in both the exercise and the no-exercise control group. The characteristics of the included articles are summarized in Table 2.
Risk of bias assessment
Risk of bias assessment is illustrated in Figure 2. The methodological quality of all the included studies was judged to be “moderate” based on the distribution for each item assessed for risk of bias. Publication bias was calculated as Begg and Egger tests (Begg test, p = .042; Egger test, p = .029).
Figure 2Assessment of risk of bias of the included RCTs. Note. RCT = randomized controlled trial.
Four studies reported the sleep quality of participants by using the PSQI, with the results of dichotomous variables revealing that the participants in the exercise group showed an obvious sleep quality improvement (OR, 6.21; 95% CI, 2.20–19.11; p = .001; I2 = 82.0%). Two studies reporting the sleep quality of participants, with the results of continuous variables revealing that the participants in the exercise group demonstrated an obvious sleep quality improvement (SMD, −0.93; 95% CI, −1.19 to −0.67; p < .001; I2 = 30.0%) (Figure3).
Figure 3The effects of exercise program on improving sleep quality and insomnia; the effects of aerobic exercise program on improving sleep quality. Note. AE = aerobic exercise; CI = confidence interval; df = degrees of freedom; SD = standard deviation.
Two studies reported participants' insomnia condition by using the Verran and Snyder-Holpern Sleep Scale and PSQI, with the results of continuous variables revealing no significant difference among the participants in either the exercise group or the no-exercise counterpart group (SMD, −2.85; 95% CI, −7.67 to 1.98; p = .250; I2 = 97.0%) (Figure 3).
Of the two studies that conducted integrated analyses on the participants who practiced aerobic exercise, the results of the dichotomous variables showed that the participants in the exercise group also had significant improvement in sleep quality (OR, 4.26; 95% CI, 2.23–8.11; p < .001; I2 = 16.0%) (Figure 3).
Discussion
A previous meta-analysis examining the effects of aerobic exercise was found to have improved the sleep quality of middle-aged women [
]. Similarly, findings derived from our meta-analysis also showed that programmed aerobic exercise can improve the sleep quality of pregnant women. This study has found that the effectiveness of aerobic exercise, when compared with overall exercises, on sleep improvement was least effective. However, opposite results were also discovered when research participants were middle-aged pregnant women [
]. Whether or not exercises of lower intensity such as yoga, tai chi, and relaxation exercise are more effective in improving sleep quality of pregnant women is an interesting topic and warrants further studies to explore and clarify their effectiveness.
Because of different statistical value representations (continuous variables and dichotomous variables) and distinct outcomes used, three articles exploring the effectiveness of practicing yoga on improving the sleep quality and insomnia conditions in pregnant women were included in the meta-analysis. The results of our review cannot conclusively determine the effectiveness of practicing yoga on improving the sleep quality or insomnia conditions in pregnant women. Conventional wisdom has nevertheless endorsed yoga and considered it beneficial in indirectly facilitating a better sleep, as well as a viable intervention in reducing pain, discomfort, and depression in pregnant women [
Assessment of the sleep quality and effects of relaxation exercise on the sleep quality in patients hospitalized in internal medicine services in a university hospital: the effect of relaxation exercises in patients hospitalized.
]. As in our meta-analysis, only one small study (n = 82), addressed the effects of relaxation exercise on sleep quality. The results derived from our analyses could neither substantiate nor repudiate the proposition that relaxation exercise can actually improve the sleep quality of pregnant women. Accordingly, more relevant studies are warranted and encouraged to validate the presumed beneficial effects of yoga and relaxation exercise on improving the sleep quality and insomnia conditions in pregnant women.
According to the meta-analysis, exercise from the second trimester can effectively improve sleep (only one study started exercise from the first trimester) [
]. However, the effect of exercise type and intensity on improving pregnant women's sleep is still not clear in the analysis. Further studies on the impact of exercise type and intensity on sleep in pregnant women at different trimesters are recommended. The results of this study showed that exercise intervention is an assistive way to improve pregnant women's quality of sleep. Moreover, according to the meta-analysis, three studies pointed out that pregnant woman can do yoga, relaxation exercise, and gymnastic exercise at home by themselves. If the researchers can design more exercises that are suitable and safe for pregnant women to do at home, it can improve the pregnant women's quality of sleep more conveniently and effectively.
Exercise has long been claimed to have a positive impact on sleep quality in many RCT-based studies [
]. Putative mechanisms suggested that generally exercise has a positive impact on physiological functions, including depression, anxiety, immune function, body restoration, circadian phase shifting, cytokine concentration, adenosine release, and thermoregulation [
]. As, unfortunately, there are no related studies conducted to determine whether and how exercise benefits pregnant women with comorbidities, interested researchers are encouraged to delve more broadly and deeply into studies designed to explore and determine the exact types or forms of exercise that may improve the sleep quality in women with other clinical conditions.
One previous meta-analysis study addressing relationships between exercise and insomnia in middle-aged women has found that exercise has a nonsignificant decrease in insomnia severity in pregnant women [
]. Similarly, findings from our meta-analysis also showed that programmed exercise during pregnancy has a nonsignificant decrease in insomnia severity in pregnant women. Insomnia is a very complex medical condition subjected to influences from a multitude of factors, e.g., comorbid status, medical therapies, negative life events, family status, social relationship problems, and employment conditions, among other factors. Exercise during pregnancy alone may not be enough to neutralize or counterbalance other negative factors contributing to insomnia [
]. In this meta-analysis, only two studies included pregnancy exercise, one study combined yoga and tai chi as pregnancy exercise, and the other practiced aerobic exercise. More studies are warranted in the future, for example, studies adopting new approaches to measure the effectiveness of exercise on insomnia, potential confounding factors, and specific quantitative assessment such as polysomnography.
Of the seven studies included, only one research study has addressed the issues of exercise intensity and warning signs for pregnant women to discontinue exercise. Pregnant women are advised to adhere to those exercise guidelines strictly to ensure their safety during the courses of exercising. The American College of Obstetrics and Gynecology recommends that all women with uncomplicated pregnancies should, after consulting with their providers, routinely engage them in aerobic and strength-conditioning exercise. An exercise program of moderate intensity aimed at achieving optimal effectiveness should be for at least 20–30 minutes per day on most days, if not every day of the week [
]. After consultation with and evaluation by an obstetrician, racquet sports, strength training, and running or jogging may be safe for pregnant women who practiced these exercises regularly before pregnancy [
]. For moderate-intensity exercise, perceived exertion should be rated, 13–14 (“somewhat hard” but not reaching the “hard” scale) on the Borg Scale of Perceived Exertion. Applying the “talk test” is another effective way to evaluate exertion. As long as a woman can have a conversation while exercising, there is no overexerting exercise [
Of all studies included in the meta-analysis, only two articles addressed body weight changes in the pregnancy process. A study by Rodriguez-Blanque et al. [
] has shown that sleep quality was better for the exercising pregnant women with a normal body weight during their pregnancy, yet it was not statistically significant when compared with their obese counterparts in the control group. As most studies did not monitor body weight changes throughout the entire courses of pregnancy, the question of whether or not an abnormal body weight increase during pregnancy may influence sleep quality and insomnia is still unknown. As such, future researchers are recommended to address the factor of body weight increase during pregnancy in their studies, hopefully to explore the relationships between exercise during pregnancy and the degree of sleep quality improvement more clearly.
Conclusions
This systematic review study attempted to assess and validate effects of exercise on pregnant women. Seven studies involving 688 participants were included in our study for systematic review and meta-analysis. The findings from our study showed that exercise of 4–16 weeks in duration significantly improved the sleep quality in pregnant women. Most of the studies included in the reviews and analyses were of low quality in terms of evidence strength owing to a lack of participants or blinding of personnel. Heterogeneity of effects was high across these studies. The high heterogeneity was caused in part by distinct exercise types pregnant women practiced during pregnancy, as shown by the results derived from subgroup analyses. Going forward, better-designed studies involving a greater number of participants are warranted to validate presumed benefits of exercise in pregnant women in the future.
Although publication bias was detected across these studies, the fact that the OR value related to improved sleep quality as a result of exercise during pregnancy was as high as 6.21 allows the authors to infer that exercise indeed can contribute to sleep quality improvement in pregnant women. Ideally, the OR value should be lower than 6.21. Authors therefore expect more future studies, including invalidity studies, to be undertaken in a bid to mitigate publication bias of meta-analyses.
Clinical implications
The findings derived from this meta-analysis and the much broader systematic review confirm with certainty that exercise can be seen as a safe, nonpharmacological approach to improve the sleep quality in pregnant women. Positive findings from this study were nevertheless somewhat offset by the fact that few high-quality studies on this subject were currently available. To make more definitive conclusions regarding the potential effects of exercise in pregnant women and perhaps their actual mechanisms, more clearly defined and rigorously designed large-scale RCTs are warranted in the future.
Conflict of interest
The authors have no conflicts of interest to disclose.
Acknowledgments
The authors wish to express heartfelt gratitude to their family for their support and sacrifice in allowing them sufficient space and time to complete this study. The authors' appreciations also go to Associate Professor Hung-En Liao of Asia University, Taiwan, and Dr. Lin Long-Yau of Chung Shan Medical University Hospital, Taiwan, for their subject matter expertise in women health that offered valuable guidance and recommendations for this research. The authors also thank the fellow PhD students and candidates of Asia University, Taiwan, who have shared with them so many brilliant research ideas and rendered much needed support in conducting this study.
References
Mindell J.A.
Cook R.A.
Nikolovski J.
Sleep patterns and sleep disturbances across pregnancy.
Assessment of the sleep quality and effects of relaxation exercise on the sleep quality in patients hospitalized in internal medicine services in a university hospital: the effect of relaxation exercises in patients hospitalized.