Psychological and Biochemical Effects of an Online Pilates Intervention in Pregnant Women during COVID-19: A Randomized Pilot Study

The purpose of this study was to analyze the effect of real-time online Pilates exercise during COVID-19 on women’s body composition, blood lipids, and psychological health after childbirth. The participants were 16 pregnant women (24–28 weeks pregnant) enrolled at the C Women’s Culture Center in Seoul, South Korea, classified into online Pilates groups and non-exercise groups (PE, n = 8; CON, n = 8). The online Pilates program was conducted for 8 weeks, twice a week, and 50 min a day using a real-time video chat app. Participants visited the hospital twice for body composition and blood tests. Questionnaires on postpartum depression, sleep disorder, and stress were conducted at 6 weeks and 12 weeks after childbirth. We found a significant difference between groups in body composition. The weight, percentage of body fat, body fat mass, and BMI of the PE group decreased. Blood lipids showed significant differences between the groups in TC, TG, LDL and CRP, while insulin and HDL showed no difference. All blood lipids, insulin, and CRP in the PE group were reduced. There were significant differences between the groups in postpartum depression, sleep disorders, and perceived stress indices performed in the post-test, and the serotonin concentration in the PE group increased. Serotonin levels were significantly correlated with postpartum depression, body fat mass, and body fat rate. Pregnant women’s online Pilates in this study was effective at reducing weight and depression in women after childbirth and should be used to promote women’s mental health during COVID-19.


Introduction
COVID-19 caused great social, economic, and cultural changes worldwide [1]. Many people experienced anxiety and stress due to the pandemic, especially pregnant women, who are at high risk due to the psychological effects of fetal protection [2]. The World Health Organization advised pregnant women with weak immunities not to go out, resulting in decreased physical activity, leading to obesity [3]. Obesity in pregnant women causes diabetes, high blood pressure, and maternal complications, which may cause fetal death [4]. In addition, prenatal care is important because obesity and depression during pregnancy are highly correlated and can cause serious mental illness after childbirth [4]. According to a recent meta-analysis, maternal depression increased by 25.6% and anxiety by 30.5% during the COVID-19 pandemic [5]. Furthermore, the American College of Obstetricians and Gynecologists (ACOG) warned that maternal anxiety had a negative effect on fetal growth and could cause premature birth, placenta previa, and underweight children [6]. Considering that "Corona Blue", which means fatigue, lethargy, and depression caused by COVID-19, has been highly surveyed in women in their 30 s [7], continuous observation and management of pregnant women is warranted. In this regard, Cohen et al. (2020) developed guidelines for mental health support for pregnant women during COVID-19 [8]. Although The study was a randomized controlled experiment, involving 16 pregnant women participants. Participants were randomly chosen by a computer lottery and assigned to the Pilates exercise group (PE; n = 8) or the control group (non-exercise) (CON; n = 8). Assessment staff were blinded to the assignments. Randomized allocation concealment, and implementation was conducted by different staff. Our study was a pilot for verifying the effectiveness of non-face-to-face exercise. Our study complied with the recommendations of the Declaration of Helsinki. In addition, the Ethics Committee of the Korea National Sport University (1263-202109-br-018-01) provided approval for the study.

Participants
The participants were women registered at the 'C' Sports Center in Seoul, Korea, who understood the aim of the study and offered their voluntary participation. The selected participants included pregnant women under the age of 40 years who were at 24-28 weeks of single-fetus pregnancy. They did not have a diagnosis of diabetes or hypertension and did not take any medications. Furthermore, all participants had a BMI < 30, and participants with obesity or physical pain were excluded.

Online Pilates Exercise
All participants in the PE group participated in the exercise using a real-time remote chat application (ZOOM). The PE group used a personal computer or smartphone at home and exercised in real-time, while communicating with the trainer. The Pilates program consisted of a warm-up, main exercise, and cool-down. It was conducted for eight weeks, twice a week, 50 min a day. The exercise intensity was set to 50-60% of the maximum heart rate, which is the appropriate exercise intensity for pregnant women recommended by the ACOG (Figure 1) [6]. The intensity evaluation maintained the self-awareness exercise index (RPE) of 11 to 13 using Borg's scale. Their RPE was monitored to maintain proper strength. The trainer observed the condition of the pregnant women during the sessions and gradually increased the intensity of exercise every three weeks, according to the fitness level ( Table 1). Considering that the participant was a pregnant woman, precautions before movement were sufficiently explained, and in the case of complaints of physical discomfort during the exercise, they could stop the exercise and rest. The CON group did not participate in any exercise for eight weeks.

Body Composition Test
All participants were asked to stop food intake and empty their bladder two hours before the body composition measurement, according to the instructions. Height was measured using an automatic height scale (DS-103M, Dong Sahn Jenix Co., Seoul, Korea). After removing all metal accessories, weight (kg), body fat (kg), BMI (kg/m 2 ), skeletal muscle mass (kg), and body fat percent (%) were measured using bioelectrical impedance analysis (In-Body 770, Biospace Co., Seoul, Korea). All participants wore light clothing and stood with their soles in contact with the foot and hand electrodes.

Body Composition Test
All participants were asked to stop food intake and empty their bladder two hours before the body composition measurement, according to the instructions. Height was measured using an automatic height scale (DS-103M, Dong Sahn Jenix Co., Seoul, Korea). After removing all metal accessories, weight (kg), body fat (kg), BMI (kg/m 2 ), skeletal muscle mass (kg), and body fat percent (%) were measured using bioelectrical impedance analysis (In-Body 770, Biospace Co., Seoul, Korea). All participants wore light clothing and stood with their soles in contact with the foot and hand electrodes.

Blood Test
For the blood test, all participants fasted from 10 pm of the day before until 9 am on the day of the test. In addition, 10 mL of blood was drawn from the brachial vein for blood glucose, lipid, and serotonin tests. After 30 min of incubation at the laboratory's temperature, blood was centrifuged (3000 rpm, five minutes) to separate the serum and was immediately taken to Green Cross Laboratories, Inc., to test the total triglycerides (TG), total cholesterol (TC), high density lipoprotein (HDL), low density lipoprotein (LDL), C-reactive protein (CRP), insulin, and serotonin. All the assays were carried out according to the instructions of the manufacturers.

Edinburgh Postnatal Depression Scale (EPDS)
We used the Edinburgh Postnatal Depression Scale to study postpartum depression of the participants. The validity of this scale has been verified for use among Korean mothers [39]. The EPDS is a self-report testing tool developed in the UK that tests postpartum depression by enquiring about depression, anxiety, and suicidal thoughts over the past week. The EPDS consists of ten items rated on a four-point Likert scale ranging from 0 to 3. The lowest total score that can be obtained is 0, whereas the highest is 30. Individuals who score 13 or more are considered at risk of developing depression. The reliability (Cronbach's α) of the EPDS in this study was 0.83, and it was processed using SPSS 24.0 software (IBM, Armonk, NY, USA).

Pittsburgh Sleep Quality Index (PSQI)
We used the Pittsburgh Sleep Quality Index to study sleep quality in the participants after childbirth [40]. PSQI is a self-report questionnaire that evaluates sleep quality and disturbances over the past few weeks. It includes 18 items that are clustered together into 7 primary factors (sleep delay, sleep quality, sleep duration, sleep disability, habitual sleep efficiency, sleep medication use, and daytime dysfunction). Each item is scored from 0 to 3, and the total score ranges from 0 to 21. The total score is obtained by summing the scores of the seven factors. A higher total score indicates lower sleep quality. A total score of ≤5 indicates sound sleep, whereas a score of ≥8 indicates a sleep problem. The reliability (Cronbach's α) of the PSQI in this study was 0.81.

Perceived Stress Scale (PSS)
The Perceived Stress Scale (PSS) was used to measure participant's stress [41]. This 10-item tool has items about perceived stress in the past few weeks. Responses are scored on a five-point scale (0 = never, 1 = rarely, 2 = sometimes, 3 = frequently, 4 = very often). Negative items (nos. 4, 5, 7, and 8) are reverse scored. The total score is 40, with a higher score indicating greater severity of stress. The reliability (Cronbach's α) of the PSS in this study was 0.83.

Statistical Analysis
All data were processed using SPSS 24.0 software. The differences in body composition, blood lipids, CRP, insulin, serotonin, EPDS, PSQI, and PSS between the PE and CON groups were analyzed. Due to the small sample size, nonparametric methods were used in the analyses. Independent t-tests were used to assess the differences in age, height, and weight between the Pilates and control groups. Pearson's chi-square test was used to compare the frequencies of maternal BMI, smoking, and employment occupation between the PE and CON. Differences in the average changes (post-pre) between the two groups were analyzed using the Mann-Whitney U test and changes over time in each group were comparatively analyzed using the Wilcoxon signed rank test. A comparison of the reference values between groups was presented by Cohen's d values. All statistical values were presented as mean and standard deviation, and the statistical significance was set at α < 0.05.

Maternal Characteristics
A total of 16 women were randomized and 16 were excluded. Table 2 shows the general characteristics of the pregnant participants in the study groups. No significant differences (p > 0.05) in maternal characteristics were found between the groups at the baseline.

Effect of Pilates on the Lipid Profiles
After eight weeks of experiment, the PE and CON groups showed significant differences in TC, TG, LDL, and CRP, whereas there were no differences in insulin or HDL The CON group showed significant changes in TC, LDL, HDL, and insulin (TC: z = −2.524, p = 0.012; LDL: z = −2.521, p = 0.012; HDL: z = −2.521, p = 0.012; insulin: z = −2.527, p = 0.012) (Table 4, Figure 2).  Values are presented as mean ± SD (n = 8 per group). Main time effect: * p  0.05, pre-versus post-Pilates period in the within groups. CON: non-Pilates exercise, PE: Pilates exercise, TC: total cholesterol, TG: total triglycerides, HDL: high density lipoprotein, LDL: low density lipoprotein, CRP: C-reactive protein.

Discussion
To address the rising obesity concerns caused by COVID-19, several experts have reported that participating in online exercise programs is effective [42]; however, evidence of the effectiveness of non-face-to-face exercise in pregnant women is limited. Additionally, there is a lack of studies that explore the effect of pregnant women's exercise on obesity and psychological health of postpartum women during the COVID-19 pandemic. To

Discussion
To address the rising obesity concerns caused by COVID-19, several experts have reported that participating in online exercise programs is effective [42]; however, evidence of the effectiveness of non-face-to-face exercise in pregnant women is limited. Additionally, there is a lack of studies that explore the effect of pregnant women's exercise on obesity and psychological health of postpartum women during the COVID-19 pandemic. To address this gap, we conducted online Pilates for pregnant women. Our results that describe the women's body composition, blood lipid, and postpartum psychological health are discussed as follows.

Effect of Pilates on Body Composition
Obesity during pregnancy has a negative physical and psychological effect [42], but there are very few studies on online exercise intervention during the COVID-19 pandemic. The ACOG recommends that pregnant women participate in appropriate levels of physical activity to avoid such stress [5,43]. Previous studies have reported that non-face-to-face exercise reduced systolic pressure, diastolic pressure, and fasting blood sugar, and increased quality of life in adult men and women [44]. A meta-analysis showed that eHealth interventions can prevent the risk of cardiovascular disease by reducing men's weight, BMI, waist circumference, and blood pressure [45]. The results of the present study were similar, with weight, percentage of body fat, body fat mass, and BMI decreasing in the PE group. This is consistent with the studies that show that Pilates is effective at increasing women's lower extremity muscle and grip strength. This also shows that online Pilates during COVID-19 is effective in relieving abdominal obesity [10]. Moreover, it is important to note that the period of intervention in this study was set differently from previous studies (effect of exercise in pregnant women and after childbirth). The weight loss in women after childbirth in this study suggests the importance of prenatal management during the COVID-19 pandemic and supports the findings of previous studies that non-face-to-face exercise has a similar effect as face-to-face exercise [30].

Effect of Pilates on the Lipid Profiles
Exercise and diet control are essential to quickly reduce increased blood lipids during pregnancy. In a related study, eight weeks of Pilates training reduced cholesterol and LDL in obese women [46]. The ACSM reported that 150 min or more of aerobic exercise per week is required to lose weight, and that the risk of adult disease can be reduced when combined with resistance exercise [32,33]. In this study, there were significant differences in TC, TG, LDL, and CRP between the PE group and the CON group. These results are consistent with the studies that show that real-time online high-intensity interval training in obese women (BMI > 25 kg/m 2 , and abdominal circumference > 85 cm) reduces visceral fat and LDL during COVID-19 [47], and is effective in preventing hyperlipidemia and diabetes [48]. Significant differences in CRP between groups in this study were confirmed and exercise during pregnancy was effective at relieving inflammation, recovering quickly, and reducing the weight of women after childbirth. On the other hand, there was no significant difference between the groups with insulin. These results are believed to be different to those that often appear in maternal exercise studies. Kampmann (2019) suggests that women during pregnancy or after childbirth have greater hormonal changes than the general population, and related blood lipid imbalances may appear [49]. Furthermore, considering the research results that women's age, environment, job, and parenting level increase stress hormones after childbirth, concerns about newborns during COVID-19 may interfere with the function of insulin and appetite control hormones. However, it is meaningful that the participants of this study included women with similar hormone secretion (corresponding to 24-28 weeks of gestation). In addition, significant intergroup effects in blood lipids suggest that the exercise protocol of this study is suitable for online exercise intensity of pregnant women.

Psychological Effects of Pilates
Depression can negatively affect the quality of life and one's daily life [50]. In particular, more than 80% of postpartum women experienced depressive symptoms, and parenting stress increased significantly after COVID-19 [5]. Although the pathophysio-logical mechanism for depression has not yet been fully identified, serotonin reduction and obesity have a high correlation, and the severity of the disease can be determined [21]. According to previous studies, exercise is recommended as a nonpharmacological treatment for depression. Moreover, serotonin and stress-related hormone functions are improved by directly stimulating brain neurotransmitters [19]. Studies also show that low intensity exercises, such as walking for less than 150 min per week, as well as moderate intensity exercises, can prevent depression [51]. In this study, there were significant differences between groups in EPDS, PSQI, and PSS of women after childbirth. However, there was no significant difference in serotonin between the two groups, but it increased significantly in the PE group. These results are consistent with studies that show that depressive symptoms improved in older adults who participated in underwater exercise [52], and online exercise helps reduce anxiety and stress during COVID-19 [53]. In addition, as a result of analyzing the association between the three psychological factors and serotonin, a high correlation with EPDS appeared, and the basis for the preceding study is presented (Table 5). Finally, there was a high correlation between body composition and serotonin in body fat mass and body fat rate (Table 6). In summary, we found that exercise during pregnancy during the COVID-19 period alleviated weight and depression after childbirth and had a positive effect on serotonin levels. In addition, when observed by experts, online exercise was found to be safe and effective for pregnant women's health. However, despite these positive results, this study has several limitations. First, this study cannot be generalized, due to the small number of participants. Second, we did not limit or modify the diet of the women. Since it was difficult to recruit pregnant women during the COVID-19 pandemic, we did not want to limit or modify their meals, which could potentially lead to increased stress. To compensate for this issue, the study participants were asked to write a daily diet diary, and a professional nutritionist checked it every Sunday. In future studies, a more specific and clear effect will be obtained if the effect of online exercise is verified with more participants. Online exercise has physical and emotional effects when personal characteristics are considered and implemented as real-time exercise. Therefore, the physical and psychological effects of online exercise conducted in this study present new exercise guidelines for pregnant women in infectious disease environments.

Conclusions
The purpose of this study was to investigate the effects of an eight-week real-time online Pilates program in pregnant women during COVID-19 on body composition, lipid metabolism, psychological health, and serotonin after childbirth. We found a significant difference between the groups in body composition in terms of weight, and the weight, percentage of body fat, body fat mass, and BMI of the PE group decreased. Blood lipids showed significant differences between the groups in TC, TG, LDL, and CRP, while insulin and HDL showed no differences. In the PE group, all blood lipids, insulin, and CRP were reduced. Finally, there were significant differences between the groups in depression, sleep disorders, and stress indices, and an increase in serotonin within the PE group. Therefore, real-time online Pilates with pregnant women, as examined in this study, is effective for weight control and depression relief for postpartum women and should be used as a tool to promote women's mental health during the pandemic. Informed Consent Statement: Informed consent was obtained from all participants involved in the study.