Socioeconomic Disparities of Low Birth Weight in Rural Indonesia

Background: Poverty is estimated to increase a mother's chances of giving birth to a low birth weight baby (LBW). Objective: This study aimed to analyze the socioeconomic disparities of LBW in rural Indonesia. Methods: Mother aged 15-49 years old who had given birth in the last 5 years in rural Indonesia was used as the unit of analysis. With stratification and multistage random sampling found 19,283 mothers were sampled. The variables analyzed included LBW, socioeconomic, age, marital, education, employment, and antenatal care (ANC). The binary logistic regression test was used at the final stage to determine the existence of disparities. Results: The results show that the mother with the poorer wealth status was 0.686 times more likely than the poorest mother to give birth to an LBW baby. Mothers with middle wealth status were 0.688 times more likely than the poorest mothers to give birth to LBW babies. Mothers with richer wealth status were 0.621 times more likely than the poorest mothers to give birth to LBW babies. Finally, the richest mothers were 0.452 times more likely than the poorest mothers to give birth to LBW babies. The results of this analysis inform that the better the socioeconomic of mothers, the lower the likelihood of giving birth to LBW babies. Apart from socioeconomics, the analysis also found two other predictors of LBW in rural Indonesia, namely education level and ANC visits. Conclusion: It was concluded that there were socioeconomic disparities of LBW in rural Indonesia.


Introduction
Low birth weight (LBW) is a condition where the baby's body weight is less than 2500 grams. This condition can be caused by genetic elements or premature birth signs. 1 LBW is found in 15.5% of births in the world and 90% occurs in developing countries 2,3 . The prevalence of LBW in Indonesia has increased from 2002 to 2013, namely 9.30% to 10.2%. This is linear with the situation of infant mortality due to LBW in Indonesia is still high, namely 29% 4 .
LBW is responsible for infant mortality as much as 75% of total infant mortality. This often occurs in the first week of birth because of the conditions of asphyxia and prematurity 5 . LBW is an important matter for health workers, policymakers, and parents to pay attention to because it can inhibit the baby's brain growth, mental retardation, and the risk of several other infectious diseases. The risk of LBW will be more complex when entering adolescence which can affect the increased risk of coronary heart disease, diabetes, immune system disorders and metabolism 6 .
In general, LBW describes maternal health conditions, maternal malnutrition, irregular access to ANC services, and the mother's low socioeconomic status 5,7,8 . Several studies explain that the factors that cause LBW include maternal factors, nutritional factors, social factors, medical factors, environmental factors, and socio-cultural factors 5 . Indonesia as a country that has a diversity of socio-cultural values and adheres to a patriarchal ideology, which places men as family leaders, who have power and control rights over women [9][10][11] . In the end, the gender ideological context attaches the position of women as companions to husbands, household regulators and the primary caregivers of children, thereby weakening the bargaining position of wives towards husbands in making decisions in the household. This condition ultimately makes women not economically independent so that it can have consequences on access to health services during pregnancy and childbirth 12,13 . This situation is still very strong, especially in rural areas.
Poor households have limitations in providing food for all family members, including the food needs of pregnant women. Insufficient food availability does not only apply in quantity but also quality 14,15 . This condition is a challenge in itself for poor households related to the increased need for macro and micronutrition in pregnant women due to nutritional needs for mothers and fetuses 16 .
In general, the health status of people living in rural areas is often lower than those in urban areas, including access to health service facilities, both for basic and referral services 17,18 . Previous research in Italy and Ethiopia informed that mothers living in rural areas have a higher likelihood of giving birth to LBW babies [19][20][21] . Neighborhood socioeconomic status, wealth index, toilet types, and sources of drinking water were the most significant contributors to prorural inequalities in malnutrition 22 . Based on the background description, the study aimed to analyze the socioeconomic disparities of LBW in rural Indonesia.

Data Source
The study employed secondary data from the 2017 Indonesian Demographic Data Survey (IDHS) as material for analysis. The unit of analysis in this study was mother aged 15-49 years old who had given birth in the last 5 years in rural Indonesia. By using the sampling method stratification and multistage random sampling obtained 19,283 mothers as respondents.

Data Analysis
In this study, LBW was defined according to WHO terminology, namely a birth weight of fewer than 2,500 grams (or 5.5 pounds), regardless of gestational age. Birth weight was the newborn's first bodyweight measured after birth and should be measured within the first hour of life before significant postnatal weight loss occurs 2 .
Socioeconomic status was based on the wealth quintile owned by a household. Households were scored based on the numbers and types of items they had, from televisions to bicycles or cars, and housing characteristics, such as drinking water sources, toilet facilities, and main building materials for the floor of the house. This score was calculated using principal component analysis. National wealth quintiles were arranged based on household scores for each person in the household and then divided by the distribution into the same five categories, with each accounting for 20% of the population, namely quintile 1 (poorest), quintile 2 (poorer), quintile 3 (middle), quintile 4 (richer), and quintile 5 (richest) 23 .
Apart from socioeconomic, other independent variables analyzed in this study were age group, marital status, education level, employment status, and antenatal care (ANC) visits. The age group was divided 5 years into 7 categories, namely 15-19, 20-24, 25-29, 30-34, 35-39, 40-44, and 45-49. Marital status was divided into 2 categories, namely never in union/divorced/widowed and married/living with a partner. Education level was divided into 4 categories, namely no education, primary, secondary, and higher. Employment status was divided into 2 categories, namely unemployed and employed. Meanwhile, the ANC visits were divided into 2 categories, namely incomplete ANC visits (<4 times) and complete ANC visits (≥ 4 times). The distribution of ANC visits was based on recommendations from the Ministry of Health of the Republic of Indonesia that recommends ANC visits during pregnancy be performed at least 4 times, namely, 1 time in the first trimester, 1 time in the second trimester, and 2 times in the third trimester 24 .
In this study, all the variables analyzed were dichotomous variables, so the Chi-Square test was used to see the relationship between ANC visits and other variables. The binary logistic regression was used in the final stage to determine predictors and see their odds ratios. All statistical analyzes were carried out using SPSS 22 software.

Procedure
The national ethics committee has approved ethical clearance from the 2017 IDHS. The respondents' identities have all been deleted from the dataset. Respondents have provided written approval for their involvement in the study. The author has obtained permission to use the 2017 IDHS data from ICF International through its website: https://dhsprogram.com/data/new-user-registration.cfm. Figure 1 shows the distribution map of LBW by the province in rural Indonesia. The distribution looks random. There is no certain pattern spatially, either in the western, central, or eastern regions.  Figure 2 shows a diagram of the interaction between 3 variables, namely LBW, socioeconomic, and ANC visits in rural Indonesia. It can be seen that for mothers who make complete ANC visits (≥ 4 times) the percentage of LBW tends to be lower following the better socioeconomic status. Furthermore, based on employment status, it can be seen that the employed mothers dominate all socioeconomic categories, except for the poorer and the middle who are dominated by unemployed mothers. Based on ANC visits, all socioeconomic categories are dominated by mothers who make complete ANC visits during pregnancy.

Results
Information on the results of the binary logistic regression test of LBW in rural Indonesia can be seen in Table 2. It can be seen that socioeconomics is one of the strong predictors of LBW in rural Indonesia. Mothers with poorer wealth status were 0.686 times more likely than the poorest mothers to give birth to LBW babies (OR 0.686; 95% CI 0.585-0.805). Mothers with middle wealth status were 0.688 times more likely than the poorest mothers to give birth to LBW babies (OR 0.688; 95% CI 0.571-0.829). Mothers with richer wealth status were 0.621 times more likely than the poorest mothers to give birth to LBW babies (OR 0.621; 95% 0.489-0.788). Finally, the richest mothers were 0.452 times more likely than the poorest mothers to give birth to LBW babies (OR 0.452; 95% CI 0.314-0.651). The results of this analysis inform that the better the socioeconomic of a mother, the lower the likelihood of giving birth to LBW babies. The socioeconomics is a barrier from LBW in rural Indonesia.
Apart from socioeconomic, the analysis also found two other variables as predictors of LBW in rural Indonesia. First, education level. Mothers with secondary education have a probability of 0.608 times compared to no education mothers to give birth to LBW babies (OR 0.608; 95% CI 0.425-0.870). Mothers with higher education have 0.483 times the chance of having no education mothers to give birth to LBW babies (OR 0.321; 95% CI 0.321-0.726). Meanwhile, there is no difference between mothers with primary education compared to mothers with no education. The results of this analysis inform that a better level of education is an inhibiting factor for LBW in rural Indonesia.

Discussion
The findings of the study indicate that the better a mother's socioeconomic level, the lower the chances of having an LBW baby. This means that socioeconomic status is a strong protective factor for LBW in rural Indonesia. LBW can be caused by poor maternal nutritional intake in mothers with low socioeconomic status as found in other previous studies. This information indicates that the poverty inherent in a family affects the availability and diversification of food consumed by pregnant women in the family so that it has an impact on babies born LBW 25 .
Socioeconomic disparities of LBW have also been reported in several studies in various low-income countries. Among them in India, Ethiopia, Bangladesh, and Pakistan 8,[26][27][28] . This finding reinforces the results of a previous systematic review analyzing publications in three electronic databases (PubMed, Scopus and Science Direct), which found that the socioeconomic disparities of LBW existed in 11 studies published in English 29 .
Meanwhile, another study in India found different findings. The results of the analysis in this study found that the relationship between socioeconomics and LBW was not significant. It can occur if a poor pregnant woman can maintain a good nutritional status and avoid potential medical complications during pregnancy so that giving birth to a normal weight baby is possible 30 . This information can also be one of the government's strategies to reduce the prevalence of LBW with food interventions for poor pregnant women.
The results showed that a better level of education is an inhibiting factor for LBW in rural Indonesia. Better education levels are proven to provide a better understanding to mothers so that they can manage their needs and fulfillment efforts. Better education levels are closely related to women's empowerment, which also impacts maternal nutrition and reduces the incidence of LBW 16,31 .
A study in Brazil informed that there was a significant decrease in LBW in neonates born to mothers with higher education levels, and there was a slight increase in mothers with poor education 32 . The results of this study are in line with several previous studies in various countries, which found that a better level of maternal education is a protective factor for LBW 16,26,33 . In several previous studies, it has often been found that a better level of education results in better output on health outcomes [34][35][36][37] . On the other hand, poor education is a barrier to achieving health status to become more qualified 38,39 .
Meanwhile, complete ANC visits are proven to be a protective factor for mothers to deliver LBW babies in rural Indonesia. This finding reaffirms the importance of monitoring the pregnancy to health workers 24 . ANC is considered beneficial for the health of pregnant women. Good ANC quality can reduce the incidence of LBW 40 . ANC provides routine monitoring of height and weight gain, identification of maternal or fetal medical problems, counseling on tobacco or substance use, providing psychosocial support, nutritional advice, and early intervention that can reduce adverse pregnancy output including LBW. Access to high-quality ANC should be emphasized as it not only improves maternal health but also provides opportunities for counseling and other risk assessment related to pregnancy 7 Unlike Indonesia, in Ghana, the local government recommends ANC visits. during pregnancy for 8 times. It was informed that mothers who make ANC visits as recommended by the government (at least 8 visits) are significantly less likely to have LBW babies than mothers who make fewer visits 41 . Similar information was also found in the Indian study. Mothers who have complete ANC visits, mothers who take IFA tablets during pregnancy, mothers with normal BMI, and mothers who are over 20 years of age, are shown to have a lower chance of giving birth to LBW babies 8 . Similar information is also informed in a study by taking samples in several developing countries 3 .
The results of this study provide a clear target for the government in its efforts to reduce the prevalence of LBW in Indonesia. Pregnant women with poor socioeconomic and poor education are the main targets. A policy is needed to encourage poor pregnant women with poor education to routinely make ANC visits according to government recommendations. On the other hand, a food intervention for these specific targets must also be prepared.

Strengths and Limitations
This study has at least 2 strengths. First, the 2017 IDHS used a validated and standardized survey tool to interview survey participants. Second, this study uses loggedverified LBW data, eliminating the possibility of recall bias. Meanwhile, the limitation is that because this is a cross-sectional study, the authors cannot confirm a temporal relationship between exposure and outcome variables. Another limitation is that it did not investigate some of the known risk factors from previous studies for LBW, including dietary and nutritional factors 42,43 , genetic factors, and [44][45][46] , as well as pollution and environmental factors 47,48 . This limitation is more due to limited secondary data from the 2017 IDHS.

Conclusions
Based on the research results, it can be concluded that there were socioeconomic disparities of LBW in rural Indonesia. Besides, two other variables were found that also significantly influence the incidence of LBW in rural Indonesia, namely education level and ANC visits.