Effects of sanitation practices on adverse pregnancy outcomes in India: a conducive finding from recent Indian demographic health survey

The study of sanitation in India goes as back as 1912 when a book was written on the topic [40]. Although a few books or articles had been written earlier, this was a comprehensive treatise on this issue to the author’s best knowledge. Sanitation in India continues to be inadequate, despite longstanding efforts. House-listing and housing census of India, 2011, revealed that as many as 53% of the households do not have drainage facilities within their premises. The Indian government in the year 2014 launched a nation-wide mission known as “Swachh Bharat Mission” (SBM) to accelerate the efforts to achieve universal sanitation coverage. The mission aims to make India clean by the year 2019. Separate ministries look after the Swachh Bharat Mission in rural and urban areas. In rural areas, the Ministry of Drinking Water and Sanitation is entrusted to look after the sanitation progress, while in urban areas, the Ministry of Housing and Urban Affairs does the same. The Ministry of Drinking Water and Sanitation claims that with the inception of Swachh Bharat Mission on October 2, 2014, it has built nearly 80 million toilets, with the number of toilets increasing every day.

There are very few studies that have directly discussed the effect of sanitation on adverse pregnancy outcomes in India [5]. A survey on a cohort of women from urban slums in New Delhi, Kolkata, and Chennai and rural slums in Hyderabad, Varanasi, and Chandigarh in 1981 recommended that improving sanitation and hygiene would have a positive effect on lowering the adverse pregnancy outcomes including prenatal and neonatal mortality [41].

Although there were a few studies in the Indian context relating the adverse pregnancy outcomes to causes other than sanitation, linking to sanitation are very limited. In a survey undertaken in the Indian state, West Bengal, it was found that exposure to a high concentration of arsenic during pregnancy was associated with a six-fold increased risk of stillbirth [42]. Another study identified different socio-economic determinants of low birth weight by following a community based prospective cohort in 45 villages of Pune district in India [43].

The fourth round of National Family Health Survey (NFHS), also known as Demographic Health Survey (DHS) - India, conducted in 2015–16, was utilized for this study. NFHS is one of the biggest population-based surveys in India which collects information on health and nutrition for each Indian state and union territory. The survey was designed to provide estimates of the prevalence of anemia, HIV, glucose level, and malnutrition. The sampling frame for NFHS-4 was the Census of India 2011. Census Enumeration Blocks (CEBs) and Villages in urban and rural area respectively served as Primary Sampling Unit (PSUs). Two-stage stratified random sampling was used to get the samples separately in urban and rural areas in each of the districts. In rural areas, stratum was a part of approximately six equal substrata, created on the basis of the estimated number of households in each village, and also the percentage of the population belonging to schedule caste and scheduled tribe (SC/ST).

Within each explicit rural sampling stratum, PSUs were sorted according to women literacy rate before PSU selection. All PSUs were selected with probability proportionate to the size of the PSU within each stratum in the first stage. The house listing was done in all selected PSUs, which served as a sampling frame for household selection. If the selected PSU had 300 or more households, then they were segmented into 100–150 households, out of which one was selected through the PPS method. A fixed number of 22 households per cluster/PSU was selected during the second stage. Similarly, CEBs and households were selected in urban areas. Around 628,900 households were selected initially for the survey, out of which 616,346 were occupied. Among the occupied, 601,509 were interviewed successfully with a response rate of 98%. Among the interviewed households, 699,686 women aged 15–49 completed the women’s interview while 112,122 men aged 15–54 completed interview with a response rate of 97% and 92%, respectively. It was noted that men were selected for interview only for the state module that provided state-level estimates, while women’s indicators were analyzed at the district level.

Anthropometry measures including height and weight were measured for children using a standard instrument that included SECA 417 infantometer (for the height of children under 2 years), SECA 213 stadiometer (for the height of children under 24–59 months and adults) and SECA 874 (for weight of other children and adults) in this survey. Health investigators collected blood specimens for anemia from eligible women aged 15–49 years, children aged 6–59 months and men aged 15–54 years. For this, prior consent was taken from them and their parents (for children under 6–59 months). Blood sample was taken from the fingertip, collected in a micro cuvette, and hemoglobin analysis was done. A hemoglobin level below 9 g/deciliter (g/dl) among pregnant women and 7 g/dl among non-pregnant women/men and children was considered as severely anemic (NFHS Report, 2015–16).

The sample of the current study included only the married women who gave birth in the past 5 years and their last born child. The total sample of the study includes 26,972 women aged 15–49 years.

The study variables included the mother’s age, body mass index (BMI), education, anemia, and ANC visits during their last pregnancy. The household-level variables included the place of residence, religion, caste, wealth index, access to the toilet, type of toilet, availability of water within toilet premises, and facility of hand wash near the toilet. Study variables for children included low birth weight (LBW), the order of birth (Parity), and death of children of women occurring in the last 5 years. All the independent variables were included in this study after going through the literature and discussion in the introduction. In the Indian context, religion, caste, and place of residence play an important role. It is worth stating that Southern states are more developed as compare to other parts of India.

Mother’s age, education, and anemia level have been given in the data, while BMI was calculated using the women’s weight (in Kg) divided by height (in meter) square (BMI = weight / height2). The BMI for pregnant women was not calculated and marked as missing for the current study (11% of women were pregnant at the time of the survey). BMI is categorized by the World Health Organization (WHO) for South-East Asian countries as underweight (BMI < 18.5), Normal (18.5–25.0) and Overweight/obese (BMI > 25.0). This study used the same categorization. The number of ANC visits has been given in the data and categorized as 0, 1, 2, and 3+ as standard WHO rules, i.e., minimum of three ANC visits should be made during the pregnancy. Characteristics of children such as LBW and parity were determined while the variable related to death of a child was directly available from the data. LBW was determined using the birth weight of the child as given in the data wherein children weigh less than 2500 g were considered as LBW. Parity was determined using the data available as total number of children ever born to a woman.

Wealth index for household (Poorest to Richest), access to toilet, toilet type, availability of water and hand wash near toilet, religion (Hindu, Muslim, and others) caste (Schedule caste-SC, Schedule Tribe- ST, Other Backward Caste-OBC, and others) of household were directly available from the data. The wealth index was based on the number and kinds of consumer goods, the household owned.

The target variable Adverse Pregnancy Outcome (APO) was determined using children who were born with low birth weight or died during the last pregnancy. APO among mother represents that either the child born with low weight or died. The NFHS data only give LBW and death occurred to a woman for her last pregnancy, so we could only add up these two variables in the formation of the variable of adverse pregnancy outcome.

The data was analyzed using bivariate (Chi-square for finding the association) and multivariate techniques (Logistic regression) in SPSS v25. Only those variables were considered in the logistic models, which were found to be associated. Adjusted odds, as well as unadjusted odds, were calculated to measure the difference between the target variable. Appropriate weights were applied to account for the intricate sampling design and respondent attrition as available in the data. To ensure the actual represention of the survey results at the national as well as at the domain level, sampling weights were required. Since the survey has utilized the two-stage sampling in the rural areas and three-stage sampling in the urban areas; weights were calculated using the sampling probabilities separately for each sampling stage. The final weights were normalized to give a total number of weighted cases that equals to the total number of un-weighted cases at the national level. These weights are used in this study to analyze the data and make it generalized. ArcGIS software is also used for generating maps showing the prevalence of adverse pregnancy outcomes and the percentage of households with no access to toilet in India.

Table 1 presents the sample characteristics. Nearly 70% of the women were in the age group 20–29. One in every four women was underweight while 14% were overweight. 38% of the women belonged to the first parity, 33% to second parity, while remaining women belonged to 3 or more parity. Nearly one in every ten women was bereft of any education, while only one in every ten women had higher education. Almost 16% of the women were severely or moderately anemic, while 40% of women were mildly anemic. Around 34% of the women had less than the required (at least three ANC visit) ANC visit, while almost 15% of women had no ANC visit. Almost 40% of the women did not have any access to a toilet, while 16% of the women used a shared toilet.

Figure 1 shows the prevalence of Adverse Preganancy Outcomes in India. The APOs are more prevalent in North India and least pervasive in North-Eastern states of India. In states like Uttarakhand, Madhya Pradesh, Rajasthan, Odisha, Uttar Pradesh, Haryana, and Goa, one in every five pregnancy turned out to be adverse. The national prevalence of 18.44 was better than ten states of the country.

Figure 2 shows the percentage of households having access to toilet facility in India. More than half of the households in Jharkhand, Odisha, Bihar, Chhattisgarh, and Madhya Pradesh did not have access to toilet facility, either within household premises or on shared basis. In India, nearly 37% of households did not have access to the toilet.

Nearly 20% of the women who either did not have toilet facilities within the household or were using shared toilets had gone through adverse pregnancy outcomes. Similarly, the adverse consequences of pregnancy were found to be higher among mothers who did not have availability of water and hand wash facility near the toilet.

We calculated both adjusted as well as unadjusted odds ratios for a better understanding of the association between sanitation and adverse pregnancy outcomes. Results showed that women who did not have access to a toilet within the house had a higher risk of adverse pregnancy outcome. In the multivariable model, no association was detected for adverse pregnancy outcomes among women who did not have access to a toilet and women who used shared toilets. Women who neither had access to a toilet at home nor shared a toilet were assumed to practice open defecation. Open defecation is still a problem by and large, and it occurs not only as a result of the unavailability of latrines but also due to cultural and behavioral reasons. There is a scarcity of research linking sanitation to adverse pregnancy outcomes in India, and it paves a difficulty in assessing our findings in the context of other similar findings. This study provides an insight into comparisons on adverse pregnancy outcomes for women who followed sanitary practices versus those who did not along with other socio-demographic factors.

States like Odisha, Madhya Pradesh, Uttar Pradesh, and Rajasthan which had inadequate sanitation coverage also had a higher number of adverse pregnancy outcomes. There is a complicated relationship between sanitation and adverse pregnancy outcomes as sanitation alone could not be a cause for adverse pregnancy outcomes. Many other factors such as, socio-demographic factors played an imperative role in the occurrence of an adverse pregnancy outcome. In spite of this, the importance of sanitation in the occurrence of an adverse pregnancy outcome cannot be ruled out completely.

Women were ignorant about sanitation practices and their importance in avoiding adverse pregnancy outcomes. We reiterate, on the basis of our findings, that sanitation is one of the crucial factors that help prevent adverse pregnancy outcomes. The scarcity in the studies associating sanitation to adverse pregnancy outcomes points out the neglected dimension of sanitation in the follow up of adverse pregnancy outcome. Improved sanitation is a basic need, and many women in India are deprived of it.

Some factors strongly influenced adverse pregnancy outcomes. These include age of the mother, the educational level of the mother, parity, BMI of the mother, ANC visits, and a few others. These findings emphasise that pregnancies during adolescence as well as in the later stages of the reproductive period were associated with an increased risk of adverse pregnancy outcomes. Mothers in the age group 20–24 had high chances of adverse pregnancy outcome as compared to mothers in subsequent age-group [44]. Many studies have confirmed that the chances a pregnancy during early ages will lead to adverse pregnancy outcomes are higher [45‐47]. The results comprehensively conclude that education is one of the most critical factors in discouraging the outcome of adverse pregnancy. Our study found that education is a crucial determinant of adverse pregnancy outcome in a multivariable model. Few studies suggest that with the increase in mother’s education, the occurrence of adverse pregnancy outcomes decrease [44]. Educational interventions not only have a positive impact on the delivery outcomes but also affect the health of the pregnant woman. Education was a missing link in the lives of Indian women, as nearly one in every four women in the study sample was without any education.

Risks were high among anemic women. Women who were not anemic had fewer chances of having an adverse pregnancy compared to women who were severely anemic. Maternal anemia was associated with an increased risk of preterm births and low birth weight rates [48‐51]. Our findings suggest that underweight women also had a higher chances of facing an adverse pregnancy outcome.

Education on sanitation practices is much required as much as it needs to follow. Pregnant women should be imparted knowledge about the need for sanitation during pre and post-pregnancy period.