Association of Low Back and Pelvic Pain with Mental Health Condition during Pregnancy and Postpartum

Background: Low back and pelvic pain (LBPP) as a physical symptom may cause psychological symptoms during pregnancy and the postpartum period. We examined the association of LBPP with mental health condition in women from pregnancy to puerperium in a prospective study. Methods: This study was conducted from March 2015 to December 2016 in a birth center in Kagawa Prefecture in Japan. We recruited 89 pregnant women who responded to questionnaires in all four stages (first, second and third trimesters and one week postpartum). We designed a self-administered questionnaire including a visual analog scale (VAS) of LBPP and Edinburgh postnatal depression scale (EPDS). Results: There were significant differences in EPDS scores and VAS scores among the 4 stages. The EPDS score at the first trimester showed significant positive correlations with VAS scores at the second and third trimesters, and the EPDS score at the second trimester showed a significant positive correlation with VAS score at the third trimester. The VAS scores at the second trimester showed significant positive correlations with EPDS scores at the third trimester and at one week postpartum. Conclusion: It was shown that the mental health condition at early pregnancy was related to an increase in LBPP at mid-pregnancy and that LBPP at mid-pregnancy was associated with the mental health condition thereafter.


Introduction
Low back and pelvic pain (LBPP) during pregnancy has a great influence on daily life thereafter. We reported that LBPP at the early stage of pregnant was associated with limitation of mobility and difficulty in performing daily activities during pregnancy [1]. Olsson et al. reported that 40% of women at 6 months postpartum had LBPP and that continuation of LBPP in the postpartum period might have an influence on child rearing [2]. It has been reported that pelvic girdle pain during late pregnancy prevented return to work after childbirth in working women at 12 weeks postpartum [3] [4]. Recently, it was reported that one in 10 women with pelvic girdle pain during pregnancy had severe consequences for up to 11 years, and that women with pelvic girdle pain had significantly decreased ability to perform daily activities, a lower level of self-efficacy, and higher levels of anxiety and depression and that they worked significantly fewer hours/week compared to women without pain [5].
With regard to mental disorders in pregnancy and puerperium, it has been reported that a depressive mood observed in early pregnancy decreased at late pregnancy but increased during the postpartum period [6]. Mental disorders in the postpartum period are important social issues in Japan. Death of children due to child abuse has been increasing in Japan, and psychological disease, anxiety about child rearing and burden of childcare are considered to be factors related to child abuse [7]. Also, 60% of pregnant women who committed suicide had postpartum depression, and efforts have been made to improve management of postpartum depression in Japan [8]. Physical symptoms such as LBPP may be linked to psychological symptoms. Regarding the association between low back pain in late pregnancy and depression, Chang et al. reported that initial depression scores predicted increases in pain interference and that depression was associated with an increase in pregnancy-related low back pain interference [9]. Pregnant women with physical health problems such as exhaustion, morning sickness, back pain, constipation and severe headaches or migraines are at increased risk of having depressive symptoms in early pregnancy [10]. It has also reported that postpartum depressive symptoms were 3-times more prevalent in women with lumbopelvic pain than in women without lumbopelvic pain [11]. We hypothesized that LBPP from pregnancy to the postpartum period is related to psychological symptoms and that an increase in LBPP is correlated with levels of psychological symptoms. There has been no longitudinal study on the association between LBPP and mental condition during pregnancy and the postpartum period. In the present prospective study, we investigated longitudinal changes in LBPP in the first trimester, second trimester, third trimester and one week postpartum in order to determine correlations of LBPP in these stages with mental health.

Subjects and Methods
This study was conducted from March 2015 to December 2016 in a birth center Y. Uemura et al. in Kagawa Prefecture in Japan. We distributed questionnaires regarding LBPP to 300 pregnant women at the first trimester. Number of delivery was approximately 500 per year in the hospital that we examined, and sample size was 246 by using permissible errors (5%), reliability (95%) and a ratio of a population (20%). We determined that sample size was 300 considering the number of un- Low back pain during pregnancy is said to be "low back pain", "back pain", "low back pain and pelvic pain", "pelvic girdle pain", "back or pelvic pain" "lumbopelvic pain", "peripartum pelvic pain", "pelvic girdle syndrome", "puregnancy-related lumbopelvic pain (PRLPP)" in various terms. In this study, low back pain caused by pregnancy is defined as including back pain and pelvic pain (LBPP) during pregnancy.

Questionnaire
We designed a self-administered questionnaire consisting of three parts that equired about 20 minutes to complete. Uemura et al. [12] designed the questionnaire for this study. The contents of the questionnaire were shown in detail in our previous report [12]. Briefly, the first part of the questionnaire consisted of questions regarding baseline characteristics such as age, marital status, education and week of pregnancy. The second part of the questionnaire consisted of Scale (EPDS). The EPDS questionnaire is a self-reported questionnaire consisting of ten items scored on a four-point Likert scale (0 -3) that is designed to assess pregnancy and postpartum depression [13]. The EPDS Japanese version showed good internal consistency (Cronbach alpha = 0.78) and test-retest reliability (Spearman correlation = 0.92) [14]. A score equal to or higher than 9 was designated to screen for minor and major depressive episodes, with sensitivities of 75% and 82% and with specificities of 93% and 95%, respectively [14] [15].

Data Analysis
The correlation of VAS score in one period with that in another period and the correlation of VAS score with EPDS score were analyzed by using Spearman's correlation. The change in the proportions of women with LBPP during the 4 stages of pregnancy and postpartum was evaluated by using Cochran's Q test, and changes in the VAS and EPDS scores during the 4 stages were analyzed by using Friedman's test. The VAS scores as well as the EPDS scores between the different 2 stages were compared by paired t-test, and their p values less than 0.0083 (0.05/6) were considered to be statistically significant according to the Bonferroni's method. P values were two-tailed, and those less than 0.05 were considered to be statistically significant. Statistical analyses of the data were carried out using SPSS version 24 for Windows (IBM Corp., Aromonk, NY).

Results
Baseline characteristics of the subjects are shown in Table 1. Mean age ± standard deviation (SD) of the subjects was 32.7 ± 5.0 years. The subjects included 37.1% primiparous women and 62.9% multiparous women. The proportion of unemployed women was 21.3% at the first trimester. The proportion of women with a past history of LBPP before pregnancy was 47.2%, and the proportion of women with a past history of LBPP in the previous pregnancy was 62.5%.
The proportions of women who complained of LBPP were 66.3% at the first trimester, 79.8% at the second trimester, 86.5% at the third trimester, and 76.4% at one week postpartum. There was a significant difference in the proportion of women with LBPP among the 4 stages (p = 0.003) ( Table 2).
The median (25 -75 percentiles) VAS scores were 1.8 (0.0 -4.1) at the first trimester, 3.0 (0.7 -5.6) at the second trimester, 5.1 (2.7 -6.3) at the third trimester and 2.7 (0.5 -5.0) at one week postpartum. There were significant differences in VAS scores among the 4 stages (p < 0.001) ( Table 2). Significant differences were found in VAS scores between first trimester and third trimester (p < 0.001), second trimester and third trimester (p < 0.001), and third trimester and one week postpartum (p < 0.001). There were significant correlations between VAS scores at the first trimester and second trimester (r = 0.334, p = 0.001), Y. Uemura et al.   between VAS scores at the first trimester and one week postpartum (r = 0.286, p = 0.007), and between VAS scores at the second trimester and third trimester (r = 0.485, p < 0.001).
The median (25-75 percentiles) EPDS scores were 5.0 (4.0 -8.0) at the first trimester, 4.0 (2.0 -7.0) at the second trimester, 5.0 (2.5 -7.0) at the third trimester and 6.0 (2.0 -9.0) at one week postpartum. There were significant differences in EPDS scores among the 4 stages (p < 0.001) ( Table 2). There were significant differences between EPDS scores at the first trimester and second trimester (p = 0.001) and between EPDS scores at the first trimester and third trimester (p = 0.001). The EPDS score at the first trimester showed significant positive correlations with the EPDS scores at the second trimester (r = 0.625, p < 0.001), at the third trimester (r = 0.498, p < 0.001), and at one week postpartum (r = 0.339, p = 0.001). The EPDS score at the second trimester showed significant positive correlations with the EPDS scores at the third trimester (r = 0.645, p < 0.001) and at one week postpartum (r = 0.544, p < 0.001). The EPDS score at the third trimester showed a significant positive correlation with the EPDS score at one week postpartum (r = 0.581, p < 0.001). The proportions of women for whom the EPDS score was more than 9 were 20.2% at the first trimester, 16.9% at the second trimester, 16.9% at the third trimester and 32.6% at the one week postpartum. There were significant differences in the proportions of women with a high EPDS score between the first trimester and second trimester (p = 0.002), first trimester and third trimester (p = 0.011), second trimester and third trimester (p = 0.003), second trimester and one week postpartum (p = 0.016), and third trimester and one week postpartum (p = 0.003).
The EPDS score at the second trimester was significantly associated with the VAS score at the second trimester (r = 0.346, p = 0.001), and the EPDS score at the third trimester was significantly associated with the VAS score at the third trimester (r = 0.216, p = 0.042). The EPDS score at the first trimester showed  Table 3). VAS scores at the first trimester and one week postpartum did not show significant correlations with EPDS scores at any stages.

Discussion
In the present study, we showed that the EPDS score at the first trimester was decreased at the second and third trimesters followed by an increase at one week postpartum. Recently, EPDS has been used as an indicator of postpartum depression in pregnant women [6] [10] [16]. Cut-off values of EPDS vary in different countries [10] [17]. It was reported that the proportions of women with depression for whom the EPDS score was more than 10 were 20.0% at the first trimester, 19.6% at the second trimester, 17.4% at the third trimester, 17.6% at childbirth and 11.1% at 3 months postpartum [16]. It has been shown that the unadjusted prevalence of depression in women for whom the EPDS score was more than 12 were 14.1% in early pregnancy, 10.4% in late pregnancy and 8.1% in the postpartum period [18]. It has been reported that a high EPDS score in Health women with depression in early pregnancy continued to increase over the perinatal period, indicating a worsening of mood over that time, although women for whom the EPDS score was less than 12 in early pregnancy experienced an improvement of mood along with a decrease in EPDS scores throughout pregnancy and the postpartum period [19]. Women in early pregnancy are likely to be physically and psychologically unstable and depressive. Since the physical condition of pregnant women becomes stable with advance of gestation and since a stable physical condition leads to a stable mental condition, the EPDS score in early pregnancy decreases in mid-and late pregnancy. Although the EPDS score has been reported to decrease in the postpartum period in other countries, an increase in the EPDS score in postpartum women has been reported in Japan. A score of 9 points is used as a cut-off value of EPDS in Japan [14]. Ishikawa et al. reported that the mean EPDS scores (proportions of women for whom the EPDS score was more than 9) were 4.5 (14.3%) in early pregnancy, 3.7 (11.8%) in late pregnancy and 4.5 (16.4%) at 5 days postpartum, indicating that the EPDS score increases in the postpartum period [6]. Our results are in line with the results of that study. The reason for increase in women with a high EPDS score at postpartum may be an isolated child-rearing environment since support and care for postpartum women are insufficient due to an increase in women of advanced maternal age [20].
We showed that the EPDS score in early pregnancy was correlated with EPDS scores in mid-and late pregnancy and at one week postpartum, suggesting that the mental health condition in early pregnancy is associated with the mental health condition thereafter. It has been reported that 83% of women with persistent depressive symptoms were identified by EPDS assessment at 12 weeks of gestation [21]. Given that the mental health condition during pregnancy has an influence on the mental health condition after delivery, it is important to evaluate the EPDS score in early pregnancy as a predictive indicator of the EPDS score after delivery. Continuous support throughout pregnancy and in the postpartum period may be needed for women for whom the EPDS score in early pregnancy is more than 9.
On the other hand, we showed that the VAS score in early pregnancy increased toward late pregnancy, followed by a decrease in the postpartum period.
The change in the VAS score in the present study is consistent with changes found in previous studies [22] [23] [24]. Support for LBPP should be given to women from mid-pregnancy to the postpartum period, during which the VAS score is increasing.
Interestingly, we found significant mutual correlations between VAS and EPDS at the second and third trimesters. It has been reported that patients with chronic lumbar pain who had a high depression scale score had a high VAS score of pain and that there was an association between depressive level and intensity of pain [25]. Apter  Thus, evaluation of mental status is needed for pregnant women with a high VAS score for LBPP to provide health guidance during the mid-and late periods of pregnancy. The reason for no significant association being found between EPDS and VAS scores at the first trimester may be the low VAS score at the first trimester, and the reason for no significant association being found between EPDS score at one week postpartum and VAS might be that various factors other than LBPP affect EPDS score at one week postpartum.
Also, EPDS scores during early and mid-pregnancy were correlated with VAS score at mid-pregnancy, and VAS score at mid-pregnancy was correlated with EPDS scores at late pregnancy and in the postpartum period. Mental health conditions at early and mid-pregnancy are associated with LBPP at mid-pregnancy, and LBPP at mid-pregnancy is associated with mental health conditions thereafter. Thus, there is a close mutual relationship between mental symptoms and LBPP as a physical symptom during pregnancy and in the early postpartum period. Women with depression are likely to have chronic pain. It has been reported that preoperative depression before Cesarean section was associated with chronic post-surgical pain at 3 months postpartum [27]. Also, Borges et al. reported that preoperative anxiety increases the risk of moderate-severe postoperative pain in women who received Cesarean section [28]. Based on these results, an unstable mental condition is considered to be a factor enhancing subsequent pain. An unstable mental state at early pregnancy may be a factor enhancing LBPP at a later stage of pregnancy. Stability of mental health at early pregnancy may prevent exacerbation of LBPP at mid and late pregnancy. On the other hand, it has been reported that perineal pain at 3 -5 days postpartum predicted depressive symptoms at 3 months postpartum [29], suggesting that physical pain influences the mental health condition thereafter. Exacerbation of pain may be involved in psychological instability. It is important to evaluate mental health by the EPDS score as well as the VAS score for LBPP in early pregnancy and to prevent LBPP in pregnancy and in the postpartum period.
This study has a strength since the study is a prospective design. However, a limitation of this study is that the sample size is relatively small. The characteristics of women without LBPP should be clarified by an increase in the sample size. Further prospective study from the first trimester to three months postpartum may be needed.

Conclusion
In conclusion, the mental health condition in early pregnancy is related to an increase in LBPP at mid-pregnancy, and LBPP at mid-pregnancy is associated Health with the mental health condition thereafter. Stability of mental health in early pregnancy may be important for the management of LBPP during pregnancy and in the postpartum period.