Association between maternal antenatal care visits and newborn low birth weight in Bangladesh: a national representative survey

Introduction

Low birth weight (LBW) in newborn babies is defined as a birth weight of less than 2500 grams (< 5.5 lbs).¹ LBW may occur due to intra-uterine growth restriction, pre-term birth (i.e., the baby is born before 37 weeks of gestation) or a combination of both.² Higher prevalence of LBW is a global health concern for childhood mortality and morbidity.³ LBW may cause diarrhea, respiratory illness and mortality in childhood, and chronic noncommunicable diseases in adulthood.^4–6 In addition, LBW has significant associations with cognitive development and decision making.⁷ Every year approximately 20 million babies suffer from LBW globally (i.e., 15-20% of total births).⁸ There is a high discrepancy in the incidence and prevalence of LBW across high income and low-income countries. About 95.6% of the total number of LBW births occur in low- and middle-income countries.^9,10 Research also shows that the rate of LBW is double in South Asia regions compared to the global percentage.¹⁰ According to a most recent study prevalence of LBW was reported 14.5% among Bangladeshi newborns.³

Antenatal care (ANC) is a broad aspect of healthcare including medical procedures and care provided to mothers during pregnancy and is significant to maintaining a healthy pregnancy state and ensuring safe childbirth.¹¹ Low utilization of prenatal and maternal health services can adversely affect birth outcomes. ANC is one of the most effective interventions to protect maternal and child health and if implemented properly it can reduce maternal perinatal mortalities.^12–14 Several studies conducted in developing countries demonstrated that improvement of ANC can significantly reduce the burden of LBW.^15–18 The World Health Organization (WHO) recommends that every woman should receive at least four ANC visits during pregnancy.¹⁹ However, according to 2019 Multiple Indicator Cluster Survey (MICS), only 37% of Bangladeshi pregnant women are receiving at least four ANC services.²⁰

In a previous study, several socio-economic factors were identified as predictors of LBW in Bangladesh, including educational status of the mother, educational status of the family head, number of ANC visits, amount of ANC assistance, location of delivery, skilled birth attendance, mode of delivery, and wealth index.⁹ According to the latest Bangladesh Demography and Health Survey (BDHS) report, 27% of newborns with a LBW had mothers who did not receive any ANC services during pregnancy.²¹ As no significant improvement was noted in reducing LBW rate, it is now considered a priority public health problem. Therefore, further identification of LBW is useful to identify the areas where progression has been made, where additional support is needed, and to implement evidence-based interventions to reduce Bangladesh’s burden of LBW. However, little research has been conducted to identify the association between ANC visits and LBW in Bangladesh. To fill the existing knowledge gap, our present study aimed to assess the association between the number of ANC visits and LBW among the Bangladeshi newborns. Furthermore, this study identified the demographic and socio-economic predictors of LBW (such as; sex of the newborns, maternal education, maternal weight status, etc.) Our study findings will generate evidence which Bangladeshi policymakers and public health managers can use to design interventions packages to ensure ANC coverage to reduce newborns’ LBW in Bangladesh.

Methods

Study design and data source: Our present cross-sectional study is based on the secondary data of Bangladesh Demography and Health Survey (BDHS) 2014 conducted by National Institute of Population Research and Training (NIPROT) in collaboration with ICF International and Mitra and Associates.²² This national representative survey is conducted every three to five years. BDHS data were collected from household (HH) level by following a two-stage stratified cluster sampling method based on the enumeration areas (EAs) and HH samples. In the first stage, EAs were selected and HH were selected from each EA in the second stage. Details of the sampling process, data collection procedures, and the questionnaire used are available in the final report of the 2014 BDHS.²² In our study we used the women’s data file and extracted data only for those women who gave birth in recent years and have complete information on infant’s size at birth.

Study variables: Predictor variables such as sex of the newborn, mode and location of the delivery, place of residence seven administrative divisions, HH wealth index, mother and father’s education and occupation were included. We also included data from the BDHS dataset on mother’s weight (kg) and number of antenatal care visits (no visits, 1-3 visits, and ≥ 4 visits). The outcome variables of our study (i.e., LBW) were based on maternal recall on perceived birth size of the newborn at birth. The BDHS did not measure the actual birth weight of the newborns, so we could not define LBW according to the conventional cut-off value (i.e., birth weight < 2500 grams). During the survey, mothers were asked; ‘When (name of the child) was born, was he/she very large, larger than average, average, smaller than average, or very small?’.²² For our analysis, we categorized responses very small and smaller than average as ‘LBW babies’, and those with very large, larger than average, and average as ‘normal weight babies’. We adopted this method from a previous study where authors used DHS data and validated that a mother’s perception of birth size is a good proxy for birth weight in large nationally representative surveys.²³ A similar method was also used in an earlier study.²⁴

Statistical analysis: The dataset was cleaned (removing missing cases, coding and recoding variables) before formal data analysis. Descriptive statistics were performed to assess the frequency and percentage of the study variables. A chi-square test was performed to compare the prevalence of birth weight and to assess the factors associated. We performed a logistic regression model (unadjusted and adjusted) to assess the association between the study variables and LBW. All statistical tests were considered significant at p<0.05 level. Data were analyzed by using Stata v14.2 (StataCorp, College station, TX, USA). We used the sampling weight provided by the BDHS in the dataset to ensure the acceptability of our results at a national level. We considered the stratified survey design by using Stata ‘svy’ command.

Ethical approval: As this study was based on a secondary source of data, authors did not require any IRB approval. However, Ethical approval for BDHS surveys have been obtained from ICF international and informed consent was received from the participants before data collection. The lead author received authorization from the DHS program for using the relevant dataset for this analysis. Details of ethical approval and data privacy are available at: https://dhsprogram.com/methodology/Protecting-the-Privacy-of-DHS-Survey-Respondents.cfm.

Results

Socio-demographic characteristics: Complete data of 4,235 (weighted) mother-child pairs were included in our study. Among the newborns the majority were male (52.9%), delivered by vaginal delivery (75.8%) and delivered at home (61%). Among the participants 74.2% belonged to rural areas and 21.8% belonged to HH with poor wealth index. Among the mothers; almost half (48.2%) completed secondary education and 16% of them were either overweight or obese. Of the participants, 20.9% of the mothers did not receive any antenatal care services during pregnancy, whereas only 31.2% of them received four and/or more ANC services (Table 1).

Prevalence of LBW by socio-demographics: The overall prevalence of LBW among the newborns was found to be 19.3% (95% CI: 17.8-20.9). A higher prevalence was observed among the females (22%), delivered by vaginal delivery (20%) at home (20.4%). Prevalence of LBW was varied across the seven administrative divisions in Bangladesh. Sylhet division had the highest burden of LBW (27.8%), whereas Rangpur division experienced the lowest prevalence (12.7%) in Bangladesh. A higher prevalence of LBW was also observed among the children whose mothers did not have any formal education (26.4%), were underweight (25.1%) and did not receive any antenatal care service during pregnancy (25.1%). The prevalence was also higher in those whose fathers were not employed (42.5%) (Table 1).

Factors associated with LBW: We found a significant positive association of LBW with (i) sex of the newborn (χ² = 17.41, p = 0.003); (ii) division (χ² = 41.98, p ≤ 0.001); (iii) maternal education (χ² = 27.92, p ≤ 0.001); (iv) maternal weight status (χ² = 30.23, p ≤ 0.001); (v) fathers’ occupational status (χ² = 13.87, p = 0.045); and (vi) maternal antenatal care visits during pregnancy (χ² = 23.15, p ≤ 0.001). However, we did not find any association between LBW and the mode of delivery, location of delivery, place of residence, wealth index, maternal occupation, and father’s educational status (Table 1).

Association between study variables and LBW: The association between socio-demographics and LBW are presented in Table 2. Female newborns had a 1.4 times higher likelihood of being LBW [AOR = 1.42, 95% CI: 1.13-1.78] compared to male newborns after adjusting sex of the newborn, division, mother’s education, mother’s weight status, and fathers’ occupation. Newborns living in Sylhet division had a higher likelihood [AOR = 1.82, 95% CI: 1.32-2.49] of being LBW compared to those who came from Barisal division. We found that the higher the maternal education, the lower the chances of having LBW newborns. Furthermore, we noted that overweight or obese mothers were 42% less likely [AOR = 0.58, 95% CI: 0.42-0.81] to have a LBW baby compared to those who were underweight. We found that the occupation of the father plays a major role in a newborn’s birth weight. Those fathers who were employed had a 71% less chance of having a LBW newborn compared to those who were unemployed [AOR = 0.29, 95% CI: 0.11-0.75] (Table 2).

Association between antenatal care visits and LBW: Among the mothers who received antenatal care services 1-3 times during pregnancy had 35% less possibility of having LBW babies compared with who did not receive any service at all [COR = 0.65, 95% CI: 0.50-0.85]. The association remained significant after adjusting with sex of the newborn, division, maternal educational status, mother’s weight status and fathers’ occupation [AOR = 0.74, 95% CI: 0.55-0.99]. Mothers who received four or more antenatal care services had a 37% less possibility of having LBW babies [COR = 0.63, 95% CI: 0.49-0.81]. However, the association was not significant in the adjusted model [AOR = 0.81, 95% CI: 0.61-1.09] (Table 2).

Discussion

We investigated the association between ANC and LBW among newborns in Bangladesh. We found that the number of ANC visits during pregnancy are significantly associated with LBW of the newborn. Additionally, we found one in every five newborns in Bangladesh are born with LBW. Being female, residing in Chittagong and Dhaka division, having a mother with no formal education, having an underweight mother, and a father who is unemployed were identified as independent predictors of LBW in Bangladeshi newborns.

This paper indicated that the number of ANC visits have a significant relationship with the birth weight of newborns. This finding is consistent with earlier studies conducted in different developing countries such as Colombia, Mexico, Nepal, and Brazil.^15–18 Regular ANC visits give the opportunity for health care workers to manage the pregnancy and enable mothers to receive a variety of services such as treatment of pregnancy-induced hypertension, tetanus immunization, and micronutrient supplementations etc.²⁵ These measures play a vital role in improving pregnancy and neonatal outcomes.²⁶ However, some studies also suggest that both adequacy and quality of ANC services are crucial for mothers to get the full beneficial effects of ANC visits and protect maternal and newborn health.²⁷ The quality of ANC is measured by three factors: the number of visits, timing of initiation of care, and inclusion of all recommended components of care.²⁸ According to our results, receiving ANC services influences the reduction of LBW more than the number of visits. We observed that mothers who received 1-3 visits had 35% less chance of having a LBW baby and the chances increased only 2% for those who received four or more ANC services which was the recommended amount according to WHO.¹⁹ The reasons behind this finding is not clear. One possible explanation could be mothers who are receiving any ANC services are provided with adequate services and messages. As several socio-economic factors restrict mothers from receiving four or more ANC services there should also be more information available during their first or second visit.²⁹ After adjusting the potential cofounders, we found that mothers who received 1-3 visits had 26% less chance of having a LBW newborn and the chances decreased by 7% for those who received four or more ANC services. Although the government of Bangladesh created several initiatives (such as launching community clinics) to increase the coverage of ANC services and ensure 4+ ANC visits for mothers, still one in every three mothers are not receiving any ANC visits during their pregnancy which is an alarming statistic.²⁰ Being a lower-middle income country, where there is a lack of resources and inadequate health services, improving health service seeking behavior remains a major challenge in Bangladesh.³⁰ Because our study findings indicated ensuring ANC visits can significantly reduce the burden of LBW rather than a higher number of visits, ensuring 100% coverage of at least one ANC visit should be the first and foremost priority of the government of Bangladesh in reducing LBW prevalence. The focus can then turn to ensuring four or more visits; however, the quality of the visit is essential.

In our study, we found that female newborns had 1.4 times higher odds of being LBW compared to male newborns. Similar findings were reported in previous studies.^31,32 However, a study in Zimbabwe found opposite results showing that female newborns were less likely to be born with LBW.³³ Little is known about the sex differences in the likelihood of being LBW. Newborns from Chittagong and Dhaka division had a higher possibility of being LBW. Being the capital (Dhaka) and commercial capital (Chittagong) city of Bangladesh, adequate healthcare facilities and civic benefits are more assessable there compared to other regions. However, increased industrialization and environmental pollution in those regions may adversely affect the birth outcome, which is supported by previous findings.^34,35

Furthermore, we found an inverse relationship with mothers’ education and LBW of the newborn which is consistent with the earlier findings.³⁶ It has been suggested that lower levels of education may limit understanding of the importance of ANC.²⁹ Similarly, to educational status, the possibility of having a LBW baby has been shown to significantly decrease with the increased weight status among mothers. This would mean that the nutritional status of the mother is crucial for delivering a healthy newborn. Intra-uterine growth can be restricted due to the poor nutritional status of the mother which may lead to LBW of the baby as mother and fetus need to compete for optimal nutrition.³⁷ We also found that newborns whose fathers are employed had a lower risk of LBW compared to those whose fathers are unemployed. Being employed (indicating economic stability) usually means a father can care for his pregnant partner by providing nutritious foods, allowing adequate time for rest and quality health care which ultimately contributes to lowering the risk of LBW.

Our present study has its limitations. A major limitation is that the BDHS data used relied on mothers’ perception of child’s size to define the LBW. However, earlier evidence suggests that mothers’ perception is a valid proxy for birth weight.²³ Secondly, causality cannot be explained since data was cross-sectional in nature. Maternal recall bias may also underestimate or overestimate the findings. We advise caution while interpreting the results. Future large scale observational studies are recommended to establish the causality. Additionally, variables were limited to common observable characteristics and the study is not able to describe the entire story of the LBW in Bangladesh. A mixed method study (quantitative followed by qualitative) may be useful. However, the study highlighted the modifiable socio-demographics which play a major role in newborns birth weight. Regardless, this study highlighted the importance of 100% coverage of ANC services in reducing LBW in Bangladesh through nationally representative data which is the major strength of the study. As this study is based on a national representative sample results can be generalized at national level.

Conclusion

Maternal antenatal care visits have a significant role in newborn birth weight in Bangladesh. Increasing the coverage of antenatal services and enabling mothers to receive quality antenatal services may substantially contribute in reducing the burden of low birth weight in Bangladesh.