Comparative study on the factors influencing pre-eclampsia symptoms at different pregnancy stages in Bangladeshi women: urban vs. rural perspectives

Abstract

Background

In Bangladesh, pre-eclampsia poses a significant concern, evident in the low attendance (37%) for antenatal care (ANC). Despite efforts to reduce maternal and neonatal mortality, the latest Bangladesh Maternal Mortality Survey (BMMS-2016) indicates limited progress. Access to essential maternal and newborn health services, including ANC, remains constrained, highlighting the challenge of translating service coverage into improved outcomes. A research gap on pre-eclampsia symptoms such as severe headache, blurred vision, high blood pressure, and oedema emphasizes the need for targeted interventions for these symptoms early so that we can reduce the prevalence of pre-eclampsia and therefore maternal mortality in Bangladesh. The aim of this study is to investigate and compare the risk factors for three stages of pre-eclampsia among Bangladeshi women living in urban and rural areas.

Methods

The study utilized BMMS-2016 data, employing statistical analyses, including binary logistic regression, to identify associations. It assessed four pre-eclampsia symptoms prevalence across pregnancy stages, considering factors like maternal age, stillborn births, residency, ANC, healthcare facility delivery, education, and children.

Results

Logistic regression highlights key associations with pre-eclampsia symptoms. Urban mothers aged 36 + face the highest risk during delivery (AOR = 2) and the lowest in rural areas after delivery (AOR = 1.43). Two or more stillborn births increase the risk in urban delivery by 97%. Complete ANC raises odds, notably in urban pregnancy (AOR = 1.5) and rural post-delivery (AOR = 1.16). Skilled ANC providers elevate risks during all stages, with the highest in urban pregnancy (AOR = 1.54) and lowest after rural delivery (AOR = 1.28). Unskilled ANC associates with symptoms only during pregnancy. Healthcare facility delivery increases odds at all stages, particularly in rural delivery (AOR = 1.74) and urban pregnancy (AOR = 1.26). Multifetal gestation raises urban delivery risk (AOR = 2.11). Rural areas show higher chances during both pregnancy and delivery. Higher education in rural pregnancy and 2 to 3 birth order in urban delivery reduce odds of pre-eclampsia symptoms.

Conclusions

Addressing pre-eclampsia symptoms in Bangladesh, especially among urban women, is urgent. Identified risk factors necessitate targeted interventions to enhance ANC and overall maternal health. Advocating findings to policymakers is crucial for effective policies, reducing pre-eclampsia and eclampsia, contributing to lower maternal mortality.

Background

Worldwide, hypertensive disorders of pregnancy, including eclampsia and pre-eclampsia (PE), cause more maternal and newborn deaths than hemorrhages [1]. Both developed and developing countries are affected by the disease, which is a global public health threat [2]. According to WHO, PE is seven times more likely to occur in developing countries than in developed countries [3]. It is estimated that preeclampsia is prevalent in developing countries between 1.8 and 16.7% [4]. However, the fastest growing region for preeclampsia is Africa and Asia, where 9% of maternal deaths are caused by the condition [1]. The Symptoms of preeclampsia constituted the most commonly reported complications, accounting for 37% of all complications [5].

Pre-eclampsia is a pregnancy-related multisystemic condition characterized by the development of hypertension, which often happens during the third trimester [6] with the presence or absence of proteinuria [7]. Depending on this pre-eclampsia is classified as being with or without severe symptoms [6]. Severe pre-eclampsia is categorized according to a specific set of criteria in high-resource nations. This incorporates diastolic blood pressure (dBP) ≥ 110 mmHg or sudden onset severe range blood pressures (sBP) ≥ 160 mmHg, with or without proteinuria. The following specific laboratory findings are also present with severe features such as: (a) pulmonary edema, (b) new-onset headache that is not responsive to medication and cannot be explained by other diagnoses, (c) impaired liver function as evidenced by abnormally elevated blood concentrations of liver enzymes, (d) renal insufficiency (serum creatinine concentration > 1.1 mg/dL or a doubling of the serum creatinine concentration in the absence of other renal disease), (e) pulmonary edema, or (f) visual impairment [6, 8]. This kind of precise pre-eclampsia classification is more practical in high resource settings [9]. On the other hand, neither lab testing nor operational blood pressure monitoring devices are available in many low-resource situations [9]. This makes pre-eclampsia diagnosis more difficult in these situations [10]. When diagnosing pre-eclampsia with severe symptoms, medical professionals frequently rely on a combination of elevated blood pressure and clinical observations in settings with limited resources [11]. Pregnant women frequently arrive emergently with eclamptic seizures, and pre-eclampsia is often misdiagnosed [12]. Presence of pre-eclampsia has a major impact since low-resource settings have high rates of maternal mortality [10].

Along with above stated problem the untreated PE, can lead to eclampsia, a condition characterized by seizures and other complications that can result in maternal mortality or stillbirth [7, 13, 14]. Despite continued improvements in public health, eclampsia and PE remain the second leading cause of maternal mortality (24%) in Bangladesh, following post-partum hemorrhage (PPH) [15]. The most important thing is that Pre-eclampsia and eclampsia can be prevented through proper treatment, routine blood pressure monitoring and urine protein testing during antenatal care (ANC) [16]. In this view, it is crucial to identify the risk factors which deteriorate the condition of PE. Numerous risk factors for PE have been described in Bangladesh, including primigravida, past history of PE, hypertension, obesity, diabetes mellitus, and women who did not receive antenatal care [17, 18].

There are very few studies [19, 20] that have been validated using statistical models that successfully account for potential confounding variables. Pre-eclampsia prevalence and risk factors in Bangladesh’s nationally representative population have not been thoroughly studied on a broad scale. A nationally representative sample allows us to generalize our finding to urban and rural population, which is not possible otherwise [21]. This study focuses on identifying risk factors for pre-eclampsia across three key stages: during pregnancy, delivery, and post-delivery. Additionally, it examines differences between urban and rural populations to determine if disparities exist. Understanding these variations will help guide targeted interventions to address gaps and improve maternal health outcomes in both settings.

Methods

Data source

This study based on secondary data extracted from the latest Bangladesh Maternal Mortality and Health Care Survey (BMMS) [6, 22]. A nationally representative sample of 321,214 ever-married women was surveyed for the study. Ever-married women aged 13 to 49 years were questioned, and any mortality among women of reproductive age was addressed specifically maternal and pregnancy-related deaths. The BMMS-2016 data collection was conducted over six months in field level, beginning August 22, 2016 to February 10, 2017. In the field, eighty-four interviewing teams consisting of one male supervisor, one female editor, five female interviewers, and one field logistic assistant were recruited. The data for the BMMS 2016 was accessed through the University of North Carolina website at https://dataverse.unc.edu/.

Sampling design

The main objective of the BMMS 2016 sampling design was to identify significant representative indicators for both urban and rural populations. This study used a complete list of urban and rural areas in the eight regions i.e., Dhaka, Khulna, Rajshahi, Chittagong, Barisal, Rangpur, Mymensingh, and Sylhet as its sampling frame and ensuring the sample as a country representative. The BMMS-2016 employed a multi-stage selection approach design to obtain representative samples for maternal mortality at the national level. In addition, the BMMS Women’s Long Questionnaire was used in this study to gather data on respondents’ backgrounds, reproduction, child mortality, and family planning from all 321,214 eligible ever-married women (weighted) aged 13 to 49 years. The authors also gathered information on household wealth through the BMMS Household Questionnaire. The detailed sampling procedure is available on the published report [5].

Participants

The main inclusion criteria were ever-married women who had given their last birth in the three years preceding the survey. After dropping the missing and irrelevant observations the final sample size was 32,236.

Response variables

If women had any of the four symptoms of pre-eclampsia (severe headache, blurred vision, high blood pressure, and oedema) during their last pregnancy, they were considered to have pre-eclampsia symptoms. Therefore, the pre-eclampsia symptoms response variable was coded as 1 if the respondent had any of the four symptoms and 0 if she did not. We determined the variable for three different stages during pregnancy, during delivery, and after delivery. Women were asked if they experienced any of the complications (severe headache, blurred vision, high blood pressure, and oedema) during pregnancy were in pre-eclampsia symptoms during pregnancy stage. Women who experienced any of the four symptoms at the time of delivery were categorized as having pre-eclampsia symptoms during delivery. Women reported experiencing any of the four symptoms after giving birth, they were classified as having pre-eclampsia symptoms in the after-delivery period. The survey questions asked women directly whether they experienced these specific symptoms at each stage (during pregnancy, during delivery, and after delivery). This study focused on four symptoms of pre-eclampsia that were not measured by any skilled or unskilled provider during the interview, the interviewer only asked the respondents if they had these symptoms. The symptoms were self-reported and there was no clinical verification of these symptoms by healthcare providers. Self–reported reproductive morbidity in general and maternal morbidity, in particular, has some limitations, including a lack of association between women’s self–reported and clinically diagnosed problems. So, there is probability of bias such as misclassification and recall bias. The BMMS data collection team used standardized, symptom-specific questions and structured interviews in order to reduce these biases [6]. Moreover, exclusion of other causes such as lack of sleep, anemia, and hemoglobin status were not determined due to lack of data. This may result in residual confounding, where unmeasured factors influence the outcome, potentially biasing the estimated associations [23]. The BMMS survey was cross-section, and we only investigated the association between different factors and pre-eclampsia symptoms.

Exposure variables

The following socioeconomic, demographic, and individual fertility variables were considered as exposure variables in our analysis based on the available literature review: mothers age, parity, education level, place of delivery, previous stillborn, at least 4 ANC visits, access to media, quality ANC, multiple births, household wealth, sex of last-child, place of residence, region, medically trained provider (MTP) [18, 24,25,26,27]. However, the variables “access to media”, “medically trained provider (MTP)”, “quality ANC”, and “all ANC” were not obtained directly from the survey data. Instead, these four variables were created from the survey data’s available information: a mother was considered exposed to media if she usually read newspapers/magazines/watched television or listened to radio at least once a week. Medically trained provider (MTP) includes qualified doctors, nurses/midwife/paramedics, family welfare volunteer (FWV), community skilled birth attendant (CSBA). Quality ANC is defined as ANC, which has all the components at least once during the post-natal period by medically trained provider (MTP). Finally, all ANC means receiving all the ANC component at least once.

Statistical analysis

The BMMS-2016 data were collected from urban and rural areas in Bangladesh. It is possible that this clustering will result in correlated responses. In bivariate analysis, the Chi-square test was used to determine if there was a significant association between the outcome and the exposure variables. In the binary logistic regression analysis, the crude odds ratios (CORs) and adjusted odds ratios (AORs) were calculated by exponentiating the unadjusted and adjusted effects of the covariates obtained from the multiple binary logistic regression model.

The form of the logistic regression models for urban and rural for three different stages of pregnancy can be written as,

$$\begin{array}{l}\Pr \,\left( {{\text{Indicator of having symptoms of pre-eclampsia}}=1} \right)\,\\=\frac{{\exp \,\left( {{X_i}\beta } \right)}}{{1+\exp \,\left( {{X_i}\beta } \right)}}.\end{array}$$

A convenient form of this model can be expressed as:

$$\begin{array}{l}\ln \,\left[ {\frac{{{P_i}}}{{1 - {P_i}}}} \right]\\={\beta _0}+{\beta _1} \times {\text{age}}\,{\text{+}}{\beta _2} \times {\text{education+}}...{\text{+}}{\beta _k} \times {\text{multifetal gestation}}\end{array}$$

,

where ${p_i}=$ indicates the probability of a women any of the four symptoms of pre-eclampsia.

For a sample of size n, the likelihood for a logistic regression is given by:

$$\begin{array}{l}L\left( {\beta;y,X} \right)=\prod\limits_{{i=1}}^{n} {p_{i}^{{{y_i}}}} {\left( {1 - {p_i}} \right)^{1 - {y_i}}}\,\\=\prod\limits_{{i=1}}^{n} {{{\left( {\frac{{\exp \left( {{X_i}\beta } \right)}}{{1+\exp \left( {{X_i}\beta } \right)}}} \right)}^{{y_i}}}} {\left( {\frac{1}{{1+\exp \left( {{X_i}\beta } \right)}}} \right)^{1 - {y_i}}}.\end{array}$$

This generates the log likelihood:

$$l\left( \beta \right)=\sum\limits_{{i=1}}^{n} {\left[ {{y_i}{X_i}\beta - \log \left( {1+\exp \left( {{X_i}\beta } \right)} \right)} \right]}.$$

Where,

$$\begin{array}{l}\exp \left( {{X_i}\beta } \right)\\={\beta _0}+{\beta _1} \times {\text{age}}\,{\text{+}}{\beta _2} \times {\text{education+}}...{\text{+}}{\beta _k} \times {\text{multifetal}}\,{\text{gestation}}\end{array}$$

.

Maximizing the likelihood (or log likelihood) has no closed-form solution, so an iterative reweighted least square is used to obtain the estimate of the regression coefficients $\hat {\beta }$. A detailed history of the logistic regression is described by researchers [28, 29]. The entry method is employed in logistic regression modelling. The significance of the parameters was tested by the p-value of their respective Wald test and the level of significance (α) for all of the tests was set to 0.05. The OR and 95% CI were also calculated for interpretation of results. All the analysis was performed using MS-Excel, STATA 17. Moreover, the authors confirmed that all methods were performed in accordance with the relevant guidelines and regulations.

Results

In Fig. 1, an overview of pre-eclampsia symptoms is provided, indicating that urban women experience a higher prevalence of pre-eclampsia symptoms compared to their rural counterparts. where it is highest at During Pregnancy (34.4%), which is more than double in number During Delivery (16.4%), while After Delivery saw a percentage of 14.6. In contract, rural woman faces less symptoms During Pregnancy (29.8%) though it is significantly higher than the other two stages. During delivery, 13.3% rural women experience symptoms, and after delivery, the prevalence is 13.5%.

Fig. 1

Percentage distribution of pre-eclampsia symptoms in different stages

Table 1 represents the different demographic and prenatal care by symptoms of pre-eclampsia. These variables were selected based on mother, child, and household characteristics. Findings depict that symptom of pre-eclampsia were seen the most (47.57%) among those mothers who had multiple births. Followed by 42.87% of mothers who got all ANC components and 41.41% who gave birth at a health facility had symptoms of pre-eclampsia. In contrast, mothers who had not taken ANC from any providers showed the lowest (27.18%) symptoms of pre-eclampsia.

Table 1 Socio-demographic and maternal care factors by symptoms of pre-eclampsia during pregnancy/ delivery or after delivery (n = 32,236)

In Table 2, multiple logistic regression showed significant associations between various factors and the likelihood of pre-eclampsia symptoms at different stages of pregnancy and delivery. Mothers aged 36 and older exhibited the highest odds [AOR = 2, 95% CI: 1.19, 3.36] of experiencing pre-eclampsia symptoms during delivery in urban area and lowest [AOR = 1.43, 95% CI: 1.02, 1.99] in rural area in after delivery stage in contrast with mothers aged 17 or younger. Urban mothers who had experienced two or more stillborn births faced a 97% higher [AOR = 1.97, 95% CI: 1.13, 3.42] risk of developing pre-eclampsia symptoms during delivery compared to those who had not. Notably, women who had one previous stillborn residing in rural areas had an 18% higher [AOR = 1.18, 95% CI: 1.02, 1.36] likelihood of exhibiting pre-eclampsia symptoms after delivery. Moreover, women who received all components of antenatal care (ANC) were more likely to experience pre-eclampsia symptoms compared to those who did not. Urban women experienced the highest [AOR = 1.5, 95% CI: 1.27, 1.77] chances of pre-eclampsia symptoms during delivery, while the lowest [AOR = 1. 16, 95% CI: 1.04, 1.3] odds were observed after delivery in rural settings. ANC received from skilled providers was linked to higher odds of pre-eclampsia symptoms during all stages compared to none, with the highest odds [AOR = 1.54, 95% CI: 1.25, 1.91] occurring in urban areas during pregnancy and the lowest [AOR = 1.28, 95% CI: 1.08, 1.5] after delivery in rural areas. Conversely, ANC from unskilled providers was positively associated with pre-eclampsia symptoms only during pregnancy, in both urban and rural contexts. Additionally, delivering in a healthcare facility was associated with an increased likelihood of pre-eclampsia symptoms at all stages. The highest [AOR = 1.74, 95% CI: 1.57, 1.93] association was observed during the delivery stage in rural areas, while the lowest [AOR = 1. 26, 95% CI: 1.11, 1.44] was during pregnancy in urban areas. Women with multifetal gestation experienced a 111% higher [AOR = 2.11, 95% CI: 1.02, 4.38] chance of pre-eclampsia symptoms in urban areas during delivery compared to those with single pregnancies. Remarkably, rural areas exhibited higher chances of association with pre-eclampsia symptoms during both pregnancy and delivery. Additionally, rural mothers with higher level of education had lower odds of developing pre-eclampsia during pregnancy. Similarly, urban women who had already given birth to 2 or 3 children also experienced reduced odds of pre-eclampsia symptoms during delivery.

Table 2 Estimated adjusted odds ratio (AOR) and 95% confidence interval (CI) for weighted binary logistic regression models for symptoms of pre-eclampsia during pregnancy, during delivery, and after delivery in urban and rural places of Bangladesh (n = 32,236)

Discussion

This study highlights that the risk factors associated with pre-eclampsia symptoms vary notably between urban and rural settings in Bangladesh. Rural women were generally found to have higher odds of experiencing pre-eclampsia symptoms during pregnancy and delivery stages compared to their urban counterparts. Factors such as maternal age, history of stillbirths, type and source of antenatal care, place of delivery, educational level, and parity were all significantly associated with pre-eclampsia symptoms, but their influence differed across regions. These findings emphasize the importance of context-specific maternal healthcare strategies to address the unique needs and risks faced by women in different geographic and socio-demographic settings.

Our study attempts to address important knowledge gaps on the prevalence and risk factors of pre-eclampsia in both rural and urban, Bangladesh through extensive data from the Bangladesh Maternal Mortality and Health Care Survey (BMMS) -2016. We categorize Pre-eclampsia (PE) symptoms into three stages during pregnancy, during and after delivery. And an overall stage termed as “any stage”. This study revealed that prevalence of PE among urban pregnant women was higher in compared to rural pregnant women. In particular, the highest prevalence of PE was found in urban Bangladesh during all three phases of pregnancy. The higher prevalence observed in urban settings could potentially be ascribed to the heightened health consciousness of urban women, which prompts them to seek more medical attention in comparison to their rural counterparts. Furthermore, urban women had greater access to healthcare, they consulted doctors more frequently. A study also supports that notion [30].

According to our findings, the prevalence of preeclampsia among women in Bangladesh is significant. During pregnancy, the prevalence of pre-eclampsia symptoms was 34.4% among urban women and 29.8% in rural women. The overall prevalence at any stage was also higher among urban women (39.0%). In comparison with other regions and countries, a hospital-based study in Sylhet division in Bangladesh found overall prevalence of preeclampsia was 14.4% during pregnancy [31]. The prevalence of preeclampsia was also reported (28%) in our neighboring country India, where the prevalence varied by state or region [32]. The prevalence of preeclampsia in Ethiopia 12.4% [4], 3% in Norway [33], 2.31% in German [34], and 4.8% in Brazzaville Teaching Hospital in the demographic republic of Congo DRC [35]. This study found a higher prevalence of PE than previously reported, possibly due to methodological variation and the use of a nationally representative sample.

The adjusted analysis of this study showed that sociodemographic factors including age and education, are significant factors for PE. Increasing maternal age (36+) was significantly associated with increased risk of preeclampsia during delivery and after delivery. Previous studies shows that Advanced maternal age increase the risk of PE [36,37,38]. This could be explained by the fact that Women’s higher risk of cardiovascular disease with age may stem from reduced vascular compliance due to arterial stiffness and aging uterine blood vessels [39]. Furthermore, the hemodynamic adjustments during pregnancy become increasingly challenging as women age [40].

Low levels of education have been linked to negative outcomes during pregnancy and labor, since they are linked to increased susceptibility and a lack of control over decisions. Higher educated women are better able to get, understand, and apply health information, which includes knowledge about things like appropriate nutrition during pregnancy, antenatal care (ANC), birth spacing [41] as well as sign and symptoms of PE [42]. Population-based cohort study of 46,618 singleton birth in southern Sweden (1999–2009) revealed a possible link between a higher risk of pre-eclampsia and lower educational attainment [43]. Another study found that Pre-eclampsia was twice as likely to occur in women with only a primary education or no education at all as in those with a secondary or higher degree [44]. The Present study also revealed that rural participants with higher education experienced reduced odds of preeclampsia symptoms during pregnancy.

A universally acknowledged and emphasized benefit of antenatal care is long-standing. The importance of proper antenatal care in determining a better pregnancy outcome is widely recognized [45]. The health advantages to women who receive ANC visits are greater for both the mother and the baby. According to a systematic review, women who do not receive any antenatal care are almost three times more likely to develop preeclampsia or eclampsia [46]. This trend is consistent with a large body of research showing a link between the lack of ANC visits and a higher risk of preeclampsia and eclampsia [47, 48]. However, recent studies show opposite result as in present study higher prevalence of pre-eclampsia observed among pregnant women who visited at least 4 ANC and received all ANC component. A study in Zanzibar found that women receiving ANC services are more suffering from severe form of preeclampsia [49]. A previous study also got the similar findings [50]. Possible explanation could be that more than two-thirds of pre-eclamptic women visited their antenatal clinic more than four times, and there was a positive association between antenatal visits and pre-eclampsia risk [51]. In context to this, a study in Bangladesh revealed that pregnant women only visit medical facilities when experiencing health difficulties or complications during pregnancy [52].

However, the current study also unveiled that there is a complex and varying relationship between the likelihood of PE and ANC provider, Place of Delivery. Higher odds of PE were observed among the urban women who receive mentioned pregnancy related services. However, we were unable to locate a study that took this particular variable into account, therefore we were unable to directly compare the results with our own. A study was conducted in Bangladesh where Preeclampsia was not specifically discussed, but risk factors for pregnancy-related complications were examined in slum and non-slum settings. It was found that women who gave birth in a medical facility, and received care from medically trained providers had more complications than their counterparts [53].

The risk of preeclampsia is increased in twins or multiple pregnancies. In this study PE was found to be more prevalent in twin pregnancies in compared to single pregnancies. When twin pregnancies occur, PE develops early in the pregnancy and is far more common than when singleton pregnancies occur [54]. There is a direct correlation between unfavorable pregnancy outcomes and twin pregnancies [55]. This correlation may arise from the bigger placenta of twin pregnancies, which exposes the mother to a greater area of placental perfusion damage [56].

Our study is based on cross-sectional data from BMMS-2016, but longitudinal research is needed to better understand the progression of pre-eclampsia symptoms over time and their long-term effects on maternal health. Although we included several key variables in our analysis, unmeasured confounders may still influence the results. Factors such as broader socioeconomic conditions or variations in access to healthcare services could also impact pre-eclampsia outcomes. Despite the large sample size, the findings may have limited generalizability across all regions of Bangladesh due to unexamined cultural and healthcare system differences.

Conclusions

Maternal mortality is a major concern in developing countries like Bangladesh and preeclampsia is one of the leading causes of maternal mortality. Findings demonstrated various socio-demographic and maternal factors were associated with preeclampsia symptoms among Bangladeshi women. In any effort to reduce maternal mortality, mothers should be prioritized so that they have adequate access to health care services, especially in rural areas. Improving mothers’ education and household affluence can help to reduce pre-eclampsia symptoms and, as a result, maternal fatalities. More effective care during pregnancy and delivery is required for mothers carrying multiple fetuses. Appropriate care must be accessible for women giving birth for the first time, which can be accomplished by education and ready access to antenatal, intrapartum, and postnatal care from medically qualified clinicians.