Association Between Obstetric Violence and Symptoms Suggestive of Postpartum Depression

Background: Postpartum depression is a common condition in the pregnancy and postpartum cycle. The development of this condition is multifactorial and can be inuenced by previous traumas. This study sought to verify whether there is an association between having been exposed to obstetric violence and presenting symptoms suggestive of postpartum depression. Methods: This is a cross-sectional study, with the inclusion of 287 women without complications in childbirth, randomly selected from two maternity hospitals of Porto Alegre, southern Brazil, in 2016. Four weeks after delivery, the postpartum women answered a face-to-face interview about socioeconomic aspects, obstetric history, health history, and childbirth experience (practices and interventions applied) and completed the Edinburgh Postnatal Depression Scale (EPDS). From the perception of women regarding the practices performed in the context of childbirth care, a composite variable was created, using item response theory, to measure the level of obstetric violence. The items that made up this variable were: absence of a companion during delivery, feeling insecure and not welcome, lack of privacy, lack of skin-to-skin contact after delivery, not having understood the information shared with them, and not having felt comfortable to ask questions and make decisions about their care. To dene symptoms suggestive of postpartum depression, reecting on increased probability of this condition, the EPDS score was set at ≥ 8. Poisson Regression with robust variance estimation was used for modeling. Results: Women who experienced obstetric violence had a higher prevalence of symptoms suggestive of postpartum depression (PR 1.55 95% CI 1.07-2.25), as well as those with a history of mental health problems (PR 2.41 95% CI 1.15-5.05), while higher socioeconomic status (A and B) had an inverse association (PR 0.38 95% CI 0.19-0.78). Conclusions: Symptoms suggestive of postpartum depression seems to be more prevalent in women who have suffered obstetric violence, of low socioeconomic status, and with a history of mental health problems. Thus, qualifying care for women during pregnancy, childbirth and postpartum and reducing social inequalities are challenges to be faced in order to reduce the occurrence of postpartum depression.

different forms and methods used for diagnosis, and the characteristics of the populations studied, interfere in the estimation of the prevalence of PPD [5].
In clinical practice, screening is ideally done by asking two questions that investigate whether, in the last month, the woman has been troubled feeling depressed or hopeless; or has had little interest or pleasure in doing daily tasks [4]. If the answer is a rmative for either of the two questions, screening for PPD is recommended by applying the Edinburgh Postnatal Depression Scale -EPDS [6]. This scale is able to identify symptoms that may be associated with depressive conditions. However, the diagnosis of PPD depends on further medical evaluation to de ne the diagnosis [7].
PPD has a multifactorial etiology. Among the agents involved in its causality are the rapid drops in hormone levels, previous negative life events, individual susceptibilities to the development of depressive conditions, low levels of social support, marital instability, domestic violence, as well as work overload, changes in routine and sleep patterns, and the feeling of disability involved in the postpartum period [8,9].
Recent researches investigate the association between mistreatment, disrespect and abuse in childbirth care -situations de ned as obstetric violence -with the development of symptoms and presentation of PPD [10,11]. Obstetric violence is a public health problem, experienced worldwide by many women, and can be de ned as a violation of human rights in a period of women's vulnerability [10,12]. Categorized as a gender-based violence, it can be expressed by verbal disrespect, physical or psychological abuse, discrimination, neglect, lack of privacy, limitation in access to information, and application of unconsented procedures [13][14][15].
The magnitude of obstetric violence varies widely, according to the different surveys conducted, with prevalence ranging from 6-98% [13,16,17]. This variation arises from the heterogeneity in the measurement of the phenomenon and recognition (or non-recognition) of the different practices performed by health professionals as abusive, disrespectful and without an evidence base for their bene ts [16,18].
In the international context, studies show the existence of an association between traumatic experiences in childbirth and higher incidence of post-traumatic stress, anxiety and depression in the early (one week after birth) and late postpartum (04, 06, 12 and 24 weeks postpartum) [19,20]. In this same direction, there are the results of research conducted in Brazil, which demonstrated an association between having suffered disrespect, abuse or violence during childbirth and having PPD [10,21]. When analyzing its subcategories, signi cant association was found between having suffered verbal violence and developing moderate and severe PPD and having suffered physical violence and developing severe PPD.
The harms of PPD are not restricted to the women. Studies show that children of women with this condition are at higher risk of hospitalizations (RR 1.93; CI 95% 1.02-3.64) and mortality in the rst year of life (RR 1.44; CI 95% 1.10 -1.89) [22]. In the long term, these children are twice as likely to develop behavioral disorders, anxiety, depression, concentration de cits, and lower performance in school subjects such as mathematics [23,24]. Therefore, early identi cation of PPD symptoms and screening of women at higher risk for this disorder is essential for diagnosis and management. Considering that obstetric violence seems to be a risk factor for developing PPD, it becomes relevant to further explore this association, making it possible to intervene in order to promote the women's quality of life, including a positive motherhood experience. Thus, the aim of this study is to verify the existence of an association between obstetric violence and symptoms suggestive of PPD in a sample of women four weeks postpartum.

Methods
Study design, population and power calculation A cross-sectional study, with inclusion of postpartum women who gave birth in two large maternity hospitals (one public and one private) of Porto Alegre, Rio Grande do Sul (RS), was conducted. The women were randomly selected, by drawing, in services responsible for approximately 25% of the 30,268 deliveries that occurred in the state capital city in 2016.
All women living in Porto Alegre who gave birth to full-term newborns in the two participating maternity hospitals were eligible. Women or newborns with unfavorable outcomes at delivery (death or admission to intensive care) or who presented formal contraindication for breastfeeding were excluded from the study, to avoid biases in the measurement for perception of obstetric violence, postpartum depression, and other outcomes of interest in the research that originated this study [25,26]. Women living in areas at risk for home visits were also excluded to preserve the safety of the research team.
For the current study, the power calculation was performed prospectively, considering the sample of 287 women, to meet the objective of identifying the association between obstetric violence and symptoms suggestive of postpartum depression. The power calculated to identify an odds ratio equal to 2.5, in a model adjusted for age, skin color, education, socioeconomic status, living with a partner, parity, mental health problems, alcohol use, and pregnancy planning, considering a signi cance level of 5%, was equal to 85.3%. The power calculation was performed in the SAS Studio software.

Data collection
Data collection occurred between January and August 2016. Every day, all women who had given birth in the previous 24 hours and met the inclusion criteria received a number that was used for the draw. Each day, two women from the public maternity hospital and one from the private hospital were included in the study until the intended sample was reached. This proportion aimed to ensure a reasonable representation in relation to the use of public and private services, described in the literature as being around 70% and 30%, respectively, at national level [27,28].
In the period from 31 to 37 days after delivery, an interview was conducted at the home or, rarely, in another place at the woman's preference, to apply a structured questionnaire, which was speci cally designed for this study, based on the previous experience of the researchers and the guiding documents of childbirth care in Brazil [29,30]. Women who were not found for the interview, after at least three attempts of telephone contact and one in person, were considered a loss.
The interviews were conducted after a pilot study that indicated the need for minor semantic adjustments to the questionnaire. The eld team was composed of 12 interviewers trained for the job. Weekly meetings were held with the eld team, seeking greater uniformity in data collection.

Statistical Aspects
The outcome variable of this study was symptoms suggestive of PPD. To identify these symptoms, the EPDS instrument was used, a self-administered questionnaire with 10 items, which considers the seven days preceding the interview. Each item has four possible answers, with a score associated to symptom severity. This score ranges from 0 to 3 (0: no change and; 3: signi cant mood change). The aspects evaluated by means of the EPDS are: ability to laugh, to nd things funny, thoughts about the future, feelings (guilt, anxiety, worry, panic, overload and unhappiness), the desire to cry and to do harm to oneself [7]. The instrument has scores ranging from zero to 30, according to validation carried out in Brazil, values close to zero indicate low or no risk of PPD, while values close to or greater than 10 indicate susceptibility to PPD [31].
Considering the severity of symptoms for women and children, and the outcome of interest in this study (symptoms suggestive of PPD), and not of possible depression, which is usually based on scores between 11 and 13 [17,31], we chose to use a score ≥8 points of the EPDS. This cut-off point increases the sensitivity of the instrument and allows symptomatic women, albeit with a lower score, to be identi ed. Previous research has already suggested that a score ≥8 points performs better on diagnostic tests for depressive symptoms [32] The choice of this cut-off point is also justi ed by the validation of the instrument (with score ≥8) in the general population, in a scenario with sociodemographic, cultural, and climatic characteristics very similar to those of the present study. This study showed that the ≥8 cut-off point had a sensitivity of 80% and 84.4% and a speci city of 87% and 81.3%, for the general population and women, respectively [33].
The exposure of interest was obstetric violence, measured by means of a binary variable that categorized postpartum women for having or not suffered obstetric violence, constructed from the latent trait Level of Obstetric Violence (Nível de Violência Obstétrica -NVO). This measure, which was de ned with mean zero and standard deviation 1, was based on an instrument composed of nine items calibrated by the twoparameter logistic model of Item Response Theory (IRT) [34] The items that composed the NVO were: not having had a companion during the prepartum, labor, and postpartum period, not having understood the information provided by professionals, not having had privacy during labor, not having felt comfortable to ask questions and participate in decisions about their care, not having felt welcomed and safe at the delivery environment, and not having had immediate skin-to-skin contact with the baby. The cut-off point of the NVO scale de ning whether or not she had experienced obstetric violence was 0.5 standard deviation above the mean. The full presentation of the development of this measure has been described in a paper that is currently submitted for publication.
Other variables were used for adjustment in this study to reduce confounding biases: age, skin color, education, socioeconomic status, living with a partner, mental health problems, alcohol use, last planned pregnancy, and parity.
As these data are from a cross-sectional study and the objective was to estimate the magnitude of the association (prevalence ratio) between the predictor obstetric violence and the outcome symptoms suggestive of PPD, the Poisson regression model with robust variance estimation was used for the analyses. The software used for the analyses were SAS Studio and SPSS 21.

Results
Of the women drawn, 379 were eligible to participate in the study. Of these, 287 were interviewed. There were 25 (6.6%) refusals, and 67 (17.7%) were lost due to failure in contacting to schedule the interviews.
The women not interviewed differed in terms of education and skin color, showing less education (p<0.01) and a higher prevalence of white skin color compared to those interviewed (p=0.032). Table 1 presents the characteristics of the women interviewed regarding sociodemographic factors, obstetric and health history, and childbirth care. The sample was composed predominantly of women aged between 20 and 34 years, white, with high income and education, and who resided with their partner. Slightly more than half of the women had planned their last pregnancy. Regarding protagonism and care in childbirth, 15.1% of the women did not feel comfortable asking questions and participating in decisions, and 32% did not have skin-to-skin contact with their babies. The prevalence of symptoms suggestive of PPD, considering the cut-off point ≥8 was 28.6%. Table 2 shows the analysis of association of sociodemographic, health, and obstetric factors with the highest frequency of symptoms suggestive of PPD. These factors were subsequently included in the multivariable model of association between the exposure variable (obstetric violence) and the outcome.  Table 3 shows the association between obstetric violence and symptoms suggestive of PPD, with adjustment for confounding factors. The association between obstetric violence and symptoms suggestive of PPD was signi cant both in the crude analysis (PR 1.581; 95% CI 1.09-2.29) and in the adjusted models. In Model 6, with a greater number of adjustment variables, a PR of 1.55 (95% CI 1.07-2.27) was found, with little variation in the summary measure and con dence interval after different adjustment models (Table 3).

Discussion
Maternal depression is one of the most common mental health problems during pregnancy and childbirth, with important effects on women, children, and their families. Negative experiences related to childbirth have been associated with the occurrence of psychiatric problems, such as depression and post-traumatic stress [19,20].
Our study identi ed 28.6% of women having symptoms suggestive of PPD using the cutoff point of ≥8 on the EPDS). We identi ed a signi cant association between obstetric violence and higher frequency of symptoms suggestive of PPD. Even after adjustment (for socioeconomic variables, mental health history, parity and pregnancy planning), women who suffered obstetric violence had a 50% higher prevalence of symptoms suggestive of PPD. This association possibly stems from the vulnerability related to childbirth, and the frustrations regarding the woman's expectations about the moment, the divergence between what is expected and the experience of abuse; humiliation and mistreatment trigger the trauma, which with the postpartum hormonal changes are exacerbated and favor the development of PPD.
Other studies investigating this association correlate the development of depressive symptoms with women's feelings of lack of control, not being provided with information at birth, experiencing physical pain, humiliation and abandonment, not being cared for properly and undergoing procedures without consent, as well as being frustrated with their expectations at birth and concerned about their child's health [35][36][37].
It is important to note that the method of measuring obstetric violence adopted for this study differs from other surveys, because there is no standardized way to measure the latent variable, an issue also documented by other authors [21]. Added to this, the use of a lower cut-off point in the EPDS (≥8) for de ning symptoms suggestive of PPD, seeking greater sensitivity, limits the comparison of this study with other surveys.
The association between having suffered violence in childbirth and developing PPD was observed in other Brazilian settings [17,21,38]. Research conducted in the southernmost region [17] showed that having suffered verbal violence increases the chance of developing moderate PPD by more than 50% In addition to obstetric violence, the present study identi ed low socioeconomic status and personal history of mental health problems (self-reported by the women) as factors associated with symptoms suggestive of PPD. These aspects have also been cited in other studies investigating factors associated with PPD or likely PPD conditions [10,11]. The authors suggest that women with a history of mental health problems have speci c personal characteristics that may in uence their perception of the care they receive [10,11]. Brown skin color, alcohol abuse, unplanned pregnancy, multiparity (3 or more) have been associated with likely cases of PPD in other investigations [10,11]. Education and marital status were signi cant in a study assessing postpartum emotional disorders [39], but were not signi cant in our study.
The course of PPD is variable [40], and may have complete remission, chronicity or relapses in subsequent pregnancies -a fact that negatively impacts the quality of life of women and the mother-baby binomial relationship, interfering in the cognitive, emotional and social development of the child [24].
These aspects should be taken into account by professionals who care for women in the pregnancypostpartum period, in order to reduce the trivialization/normalization of this condition.
This study uses a latent variable, called level of obstetric violence, constructed through a set of items, using IRT in modeling. The method, besides being statistically robust, evaluates each item of the measurement instrument according to its severity and discrimination capacity, allowing each one to have a different weight (importance) in estimating the level of violence. By using the composite variable -NVO (Level of Obstetric Violence) -it is possible to measure the impacts of the absence of a companion during labor, delivery and postpartum, of the women not having felt the childbirth environment welcoming, safe and private, of not having had skin-to-skin contact with their babies immediately after childbirth, and of not having understood information and had autonomy during childbirth.
Another potentiality of the present study refers to the methodological rigor in its conduction, with continuous quality control and face-to-face interviews four weeks after delivery in the homes of postpartum women, aspects that increase the methodological quality, once they reduce potential biases related to the intense sensations, exhaustion and lack of time to process the facts that occurred in the immediate postpartum period and the relativization and forgetfulness related to the passing of several months after the event.
As weaknesses of the study, we can point out the high number of losses, with respective reduction of the sample effectively investigated, which may have hidden the association of other exposure variables with the outcome. The use of a lower cut-off point (≥8), less frequently used in research, limits the comparison of the ndings of this study with those of other investigations and increases the proportion of false positives (1-speci city). However, in choosing this cut-off point, the highest sensitivity was valued, in search of a group of women at higher risk of developing PPD, since it is a screening and not a diagnostic tool.

Conclusions
Women who experienced violence in childbirth had a 55% higher prevalence of symptoms suggestive of PPD. The personal history of mental health problems more than doubled this prevalence, while the higher socioeconomic status reduced the prevalence by more than 60%. Therefore, efforts to qualify childbirth care and minimize the occurrence of obstetric violence, still very present in the reality of maternity hospitals worldwide, are necessary to reduce the occurrence of postpartum depression and thus prevent the various negative outcomes resulting from this condition.