This was a retrospective case-control study conducted from January 2016 to April 2019. This research has been approved by the IRB of the authors’ affiliated institutions.
All the enrolled women who attended the antenatal clinic in the Department of Gynecology, University Hospital of ** gave written consent. The pregnant women who were diagnosed with LBP were assigned as the case group, and the age-matched healthy pregnant women without LBP were assigned as the control group from the same hospital during the same time. Pregnant women generally need to register at an obstetrics unit in the 12th week of pregnancy. They are examined due to obstetric reasons on 12–14 scheduled dates during the whole pregnancy. Nineteen women with a history of any disease before pregnancy or substance abuse were excluded. Another 32 women also had to be excluded: lost to follow-up/incomplete data (N = 09), feelings of severe and constant fatigue(N = 00), adverse life events during pregnancy and previous pregnancy including unplanned abortion, severe foetal malformations, and dead foetuses due to potential risks for perinatal depression(N = 07) , pregnancy via reproductive medicine (N = 01), loss of close companions/family members/friends during the previous 12 months (N = 08), and severe hypertension and diabetes during pregnancy (N = 07). Finally, a total of 484 pregnant women were enrolled in this study: a case group of 242 pregnant women with LBP and an age-matched control group of 242 pregnant women without LBP. The patients in the case group were further divided into three groups: the LP group, PPP group, and CP group.
Edinburgh Postnatal Depression Scale (EPDS)
The outcome of interest was a positive screen for perinatal depression symptoms using the EPDS . Women who have a consultation in an antenatal clinic in our hospital are routinely administered the EDPS to screen for depression. This scale consists of 10 short questions with a choice of four answers that closely reflects how she was feeling over the past seven days. Scores are recorded as 0, 1, 2, and 3 according to symptom severity. Certain question items (i.e., 3, 2, 1, and 0) are scored in a reverse manner. The EPDS has been studied extensively, and it is thought to be a valid screen for both pre- and postnatal depression [15,16,17]. The EPDS has been widely used for research and for use in the community to screen for pregnancy-related depression with a sensitivity of 86%, a specificity of 78%, and a positive predictive value of 73% . A score ≥ 13 on the EPDS is the recommended cutoff to use for identifying probable major depression perinatally . The EPDS was administered by an experienced psychiatrist through an interview or telephone call. Each woman was evaluated once for this rating in the morning during the third trimester (T1) before delivery and six months (T2) after delivery (Fig. 1). Perinatal depression is represented by a positive screen for both prenatal and postnatal depression .
Description of low back pain (LBP) and its subdivisions
LBP in pregnancy has been defined as a recurrent or continuous pain rating of ≥ 3 for more than one week from the lumbar spine or pelvis . The pain intensity was evaluated with the self-reported scale of 0–10 (0 as no pain to 10 as the worst possible pain) to screen LBP through an interview or telephone call at the same time points as those of the EPDS (Fig. 1). A recurrent or continuous LBP rating of ≥ 3 has a disabling influence on the quality of life , and previous studies have demonstrated that disabling LBP has a close association with depression [6, 7]. The exposures of interest were binary variables about the pain types perinatally (lumbar pain = 1 and posterior pelvic pain = 0 during the data input).
Lumbar pain (LP) was characterized by a history of lumbar back pain before pregnancy, pain drawing with markings above the sacrum in the lumbar spine, a decreased range of motion in the lumbar spine, pain upon palpation of the erector spinae muscle and negative results on the posterior pelvic pain provocation test. PPP was characterized by no history of lumbar back pain before pregnancy, pain drawing with markings in the gluteal area, time- and weight-bearing related to pain deep in the gluteal area, pain-free intervals, free range of motion in the spine and positive results on the posterior pelvic pain provocation test. Combined pain was defined as having both LP and PPP.
During the period of pregnancy and six months postnatally, all the women who experienced LBP would be referred to a multidisciplinary team, which included an obstetrician, orthopaedist, acupuncturist, and physiotherapist. This team, the participants of whom were blinded to the results of the depressive evaluation, identified the pain types according to the characteristics mentioned above. According to the results, treatments would be recommended, including education regarding anatomy and kinesiology, back-strengthening exercises, reducing physical activity, avoiding overloading the pelvis, physiotherapy, manipulation, yoga training, and/or acupuncture. The treatment plan depended on the needs of the particular women and the discomfort level .
Questionnaire about the pregnancy
This questionnaire was designed by the authors, and it collected data that included age, BMI, educational level, annual household income, caesarean delivery, breastfeeding, unexpected sex of the baby, parity, sick leave, large amount of physical demand (twisting/lifting movements) and LBP in the previous pregnancy. It was filled in by the subject before the first assessment (Fig. 1). Participants were asked to give their choices. In accordance with the rule of at least ten events per variable in the analysis, the number of variables had to be limited .
The sample size was calculated to detect a mean difference in pain scores of 0.5 with a standard deviation (SD) of 0.25. The error was set at 0.05, and the power level was set at 90% with additional compensation for a possible dropout rate of 20%. The required sample size was at least 28 patients in each group.
The Kolmogorov-Smirnov test was used to assess the distribution of continuous variables. According to the results of the Kolmogorov-Smirnov test, we used the mean and standard deviation or the median and semi-interquartile range to demonstrate normally distributed and non-normally distributed variables, respectively. Ordinal variables were described as proportions. The Chi-square test was carried out to compare dichotomous variables, and Student’s t-test was used for continuous variables. The Kruskal-Wallis test was used for multigroup comparisons of nonparametric data on the ordinal level. Logistic regression was performed to estimate the odds ratio (OR) and the associated 95% confidence interval (CI) to determine LBP types in perinatal depression. Multivariate logistic models were performed using stepwise elimination of variables of interest from univariate analysis after adjustment for confounding factors. Logistic regression analysis was used to examine the association between depression pre- and postnatally, different types of LBP, and possible confounders. The dependent variable was depression pre- and postpartum. Different LBP types were entered as categorical independent variables (no LBP as a reference). The covariates were suspected if the prevalence of the LBP type and perinatal depression were higher when the risk factor was present and if the crude odds ratios were statistically significant for each association. The power of the sample size was calculated by G*Power (version 3.1, Heinrich-Heine-Universita¨t Du¨sseldorf, Germany). The statistical significance and power analysis were P-values ≤ 0.05 and 0.8, respectively. SPSS version 22 (SPSS; Chicago, IL, USA) was used to perform all analyses.