Introduction
The under-five mortality rate is considered one of the most sensitive barometers of child well-being, indicating health and nutrition status in early childhood. It is also used as a critical indicator to assess ongoing child health interventions and a country’s human development level (Usmani & Ahmad, Reference Usmani and Ahmad2017). India has fixed neonatal and child survival as one of the top priorities under Sustainable Development Goal (SDG) 3, but the latest Sample Registration System (SRS) report shows that the neonatal mortality rate still stands at 23 and under-five mortality at 36 per 1000 live births (United Nations Department of Economic and Social Affairs, 2015; Office of the Registrar General and Census Commissioner (2018) Sample Registration System Statistical Reports, 2018). It clearly shows that the country is far from achieving the goals. One can also easily infer from the recent SRS estimates for childhood mortality that infant deaths constitute nearly 89 per cent of the under-five mortality rate, while neonatal deaths constitute nearly 70 per cent of the total infant deaths at the national level (Sample Registration System Statistical Reports for 2018; Office of the Registrar General & Census Commissioner, 2018). Hence, decline in the under-five mortality can be efficiently achieved by significantly reducing the neonatal mortality rate.
Previous studies have clearly established that early initiation of breastfeeding reduces the risk of neonatal mortality and also saves the child from chronic diseases (Sen et al., Reference Sen, Mallick and Bari2020). Studies have also highlighted that breastfeeding initiation within one hour of birth and exclusive breastfeeding in the first six months of life significantly decrease neonatal and under-five mortality (Ballard & Morrow, Reference Ballard and Morrow2013; Oot et al., Reference Oot, Sommerfelt, Sethuraman and Ross2015). The World Health Organization (WHO) classifies the initiation of breastfeeding of newborns within one hour of birth at the rate of 0–29% as poor, 30–49% as fair, 50–89% as good and 90–100% as very good (WHO, 2003). It also says that while breastfeeding is a natural act, it is also a learned behaviour. Breastmilk contains all the nutrients an infant needs in the first 6 months of life, contributes to the child’s immunological defence system and increases its disease resistance. Non-breastfed children in developing countries are six times more likely to die before the age of one month than children who receive at least some breastmilk (WHO, 2013). According to the Pan American Health Organization (PAHO), encouraging appropriate infant feeding practices, such as early initiation of breastfeeding and exclusive breastfeeding for 6 months, is the single most effective preventive intervention for improving the survival and health of children (PAHO, 2011) and is also a key to sustainable development.
Colostrum, which is the first milk produced after birth, is a baby’s first immunization. It has high levels of antibodies, vitamin A and other protective constituents and strengthens the child’s immune system, reducing the chances of death in the neonatal period (Kroeger, Reference Kroeger2003; Edmond et al., Reference Edmond, Zandoh, Quigley, Amenga-Etego, Owusu-Agyei and Kirkwood2006). The evidence drawn from a meta-analysis from over 63 developing countries shows that early initiation of breastfeeding prevents infections among newborns and averts death due to sepsis, pneumonia, diarrhoea and hypothermia; besides, it facilitates sustained breastfeeding (Kramer et al., Reference Kramer, Chalmers, Hodnett, Sevkovskaya, Dzikovich, Shapiro and Shishko2001; Jones et al., Reference Jones, Steketee, Black, Butta and Morris2003; Stanley et al., Reference Stanley, Mei, Gowri, Priscilla, Maqula, Deirdre and Joseph2007). Early and exclusive breastfeeding is crucial in developing countries where limited access to clean water increases the risk of diarrhoeal disease and respiratory tract infections (Aarts et al., Reference Aarts, Kylberg, Hörnell, Hofvander, Gebre-Medhin and Greiner2000; Mihrshahi et al., Reference Mihrshahi, Oddy, Peat and Kabir2008).
Provision of mother’s breastmilk to an infant within one hour of birth is referred to as ‘early initiation of breastfeeding’ (EIBF) and ensures that the infant receives the colostrum, or ‘first milk’. Initiation of breastfeeding as early as possible has become mandatory in baby-friendly initiatives because it serves multiple purposes for the growing baby, the mother and the entire family (Kamath et al., Reference Kamath, Garg, Khan, Jain and Baliga2016).
According to a United Nations International Children’s Emergency Fund (UNICEF) report, only 41% of newborns are breastfed within one hour of birth in South Asia. Several South Asian countries have some of the worst early initiation breastfeeding practices in the world; the rate is low in Pakistan (29%), India (41%), Bangladesh (45%) and Nepal (47%) (UNICEF, 2014). Patel et al. (Reference Patel, Banerjee and Kaletwad2013) found that India ranked the lowest in breastfeeding practices among Afghanistan, Bangladesh, India and Sri Lanka, with only 44% of women breastfeeding their babies within one hour of delivery.
Studies have identified various factors that negatively influence the breastfeeding of children in the early hours of life (Patel et al., Reference Patel, Banerjee and Kaletwad2013). These include lack of knowledge among mothers about the health benefits of early breastfeeding (Ochola, Reference Ochola2008; Nguyen & Hawkins, Reference Nguyen and Hawkins2013) and inadequate antenatal counselling on breastfeeding (Dhandapany et al., Reference Dhandapany, Bethou, Arunagirinathan and Ananthakrishnan2008; Blixt et al., Reference Blixt, Johansson, Hildingsson, Papoutsi and Rubertsson2019; Sehhatie et al., Reference Sehhatie, Mirghafourvand and Havizari2020). Clinical trials and quasi-experimental studies have shown that scientifically appropriate advice with high-level sensitivity and support from front-line health workers can significantly improve early initiation of breastfeeding (Blixt et al., Reference Blixt, Johansson, Hildingsson, Papoutsi and Rubertsson2019; Azimi & Nasiri, Reference Azimi and Nasiri2020; Naroee et al., Reference Naroee, Rakhshkhorshid, Shakiba and Navidian2020). Also, an expectant mother should be trained well before childbirth about the right latching technique and the health benefits of early breastfeeding to the mother and child (Kronborg et al., Reference Kronborg, Harder and Hall2015).
Early initiation of breastfeeding has been shown to be associated with the economic status of the family, mother’s age, education, occupation, place of residence and health care utilization such as antenatal care (ANC) visits and place of delivery (Xu et al., Reference Xu, Binns, Zheng, Wang, Zhao and Lee2007; Senarath et al., Reference Senarath, Dibley and Agho2010; Agho et al., Reference Agho, Dibley, Odiase and Ogbonmwan2011; Tan, Reference Tan2011). Although breastfeeding is quite widespread in India, considerable differences exist in breastfeeding and weaning practices across states and socioeconomic groups (Sen et al., Reference Sen, George and Stalin2002; Dwivedi, Reference Dwivedi2010).
In India there are obvious wide disparities in neonatal mortality rates across states, as noted in the latest five SRS reports (Table 1). Of the nineteen major Indian states, Bihar, Chhattisgarh, Madhya Pradesh, Odisha, Rajasthan and Uttar Pradesh have shown much higher neonatal mortality rates (NMR) than the national average for the past several years. Interestingly, all these six states belong to the Empowered Action Group (EAG) of states that lags in multiple socio-demographic, maternal and child health indicators (IIPS & ICF, 2017). There have been consistent efforts under the National Health Mission to improve infant and child survival in states with high infant and child mortality rates. However, none of these six states has been able to reduce their neonatal mortality rates below the national average, at least in the recent past. According to the available literature, early initiation of breastfeeding does have reducing effects on mortality rates in early infancy. Therefore, it becomes imperative to see whether there is a significant difference in the early initiation of breastfeeding and the factors determining it between the states showing NMR above the national average and the remaining thirteen major states (including the National Capital Territory (NCT) of Delhi). Hereafter, the states showing NMR above the national average may be referred to as ‘high NMR states’ and the rest as ‘low NMR states’. The high NMR states have greater programmatic significance, given that the latest civil registration system report for the year 2018 suggests that they contribute half of the total annual live births in India (Office of the Registrar General, 2018).
Table 1. Neonatal mortality rates in India and its major states, 2014–2018
Source: Office of the Registrar General and Census Commissioner (2014–2018): Sample Registration System Reports for 2014, 2015, 2016, 2017 and 2018.
Many studies have been carried out to understand the factors associated with the practice and advantages of breastfeeding. However, to the authors’ knowledge, no previous study has examined the differentials in the early initiation of breastfeeding and its associated factors segregated by low and high NMR states in the Indian context. The present study formulates a research hypothesis to test whether giving breastfeeding advice regarding EIBF to pregnant women during antenatal care service significantly improves their chances of early initiation of breastfeeding.
Methods
Data
The present study analysed a unit-level dataset of the National Health and Family Survey (NFHS-4) conducted in 2015–16. The Sample Registration System reports from 2014 to 2018 were used to identify low and high NMR states. The NFHS-4 sample was designed to estimate key socio-demographic and health indicators at the district, state and national levels. It successfully interviewed a total of 699,686 eligible women aged 15–49 from 601,509 surveyed households. Data available in the Child data file containing socioeconomic and demographic information on women, and children born to them, in the 5 years prior to the survey date were analysed. The file provides vital information on date of birth, date of death, birth weight, place of delivery, childhood vaccination and health check-ups, initiation and duration of breastfeeding and other characteristics of the children (IIPS & ICF, 2017). The sample for this study was restricted to each woman’s most recent surviving child, provided that child was born during the 2 years preceding the survey date, with the aim of minimizing recall lapse among mothers. Thus, the eventual sample of children in the analysis was 43,529 and 41,643 for high and low NMR states, respectively.
Outcome variable
The study outcome variable was ‘early initiation of breastfeeding’ (EIBF) of the youngest surviving child born to a woman in the 2 years prior to the survey date. In NFHS-4, the following two questions were posed to women who had had at least one live birth since January 2011: 1) Did you ever breastfeed [Name]?; 2) How long after birth did you first put [Name] to the breast? The second question was only canvassed for the last surviving child. The responses were recorded as either ‘immediately’ or in terms of number of hours/days. The response ‘less than one hour’ was re-coded as ‘immediately’. Hence, early initiation of breastfeeding (EIBF) was defined as a child being put to the breast immediately after birth.
Independent variables
Maternal socio-demographic characteristics and health care seeking behaviours were selected based on the availability of variables in the dataset and their importance to EIBF in the literature (Buckles & Kolka, Reference Buckles and Kolka2014; Kalisa et al., Reference Kalisa, Malande, Nankunda and Tumwine2015; Lakew et al., Reference Lakew, Tabar and Haile2015; Singh et al., Reference Singh, Khan, Carvajal-Aguirre, Brodish, Amouzou and Moran2017; Ndirangu et al., Reference Ndirangu, Gatimu, Mwinyi and Kibiwott2018). Mother’s age at the time of childbirth was grouped as: 15–24 years, 25–34 years and 35–49 years. Place of residence was categorized into rural and urban. Household economic status was measured using wealth quintiles based on the sum of scores on the ownership of household assets and consumer goods, taking into consideration the de jure household population: poorest, poorer, middle wealth, richer and richest. Mother’s education level had five categories: non-literate, literate but below primary, primary but below middle, completed middle, and high school and above.
Other variables found to be associated with EIBF in the literature, and therefore included in the analysis, were: sex of the child (male, female); whether a birth was planned (‘type of birth’) (planned, unplanned); place of delivery (home, health institution); type of delivery (vaginal, Ceasarean); delivery by a trained person (yes, no); mother advised on breastfeeding during pregnancy (yes, no); mother’s exposure to mass media (exposed, not exposed); and religion (Hindu, non-Hindu). There variables were all dichotomous in the study. The index child’s birth order was divided into three groups:1st order, 2nd–3rd order and 4th + order.
Caste/tribe has significant relevance in the Indian context for any social and health indicator, including EIBF. Under the varna system, people are divided into four classes. The lowest rung of this system includes Scheduled Castes/Tribes, who were allowed to do derogatory and scavenging work only. People from these strata lack access to information and health services. The Other Backward Classes (OBC) include those deprived of health services primarily due to social and economic constraints. The ‘other’ category comprises those not included in any of the above categories.
Statistical analysis
Bivariate and multivariate analyses were performed to examine the practice and determinants of EIBF in low and high NMR groups of states (Table 1). First the prevalence of EIBF at individual state, NMR group and national levels was estimated. Thereafter, the prevalence was estimated by the selected variables separately for low and high NMR states. Finally, the chi-squared test was applied to rule out any association, merely by chance, between the EIBF prevalence and a particular background.
A variance inflation factor (VIF) diagnostic test statistic was applied to assess multicollinearity among the chosen independent variables. Multicollinearity occurs when two or more predictors in the model are highly correlated (that is, when the value of VIF>5; Yoo et al., Reference Yoo, Mayberry, Bae, Singh, He and Lillard2014), and can provide redundant information regarding the effects of an independent variable on the response. The interpretation of the results is often misleading in the presence of high VIF. The VIF of all the study predictors was less than 5 in the low as well as the high NMR group indicating that there was no multicollinearity among the predictors used in the multivariate binary logistics regression analysis to examine their independent effects on EIBF.
Finally, a multivariate logistic regression analysis was conducted to examine each background characteristic’s independent influence on EIBF practice separately for high and low NMR states. Adjusted odds ratios (AORs) were generated and tested for statistical significance using p-values and 95% confidence intervals (95% CIs). STATA 13.1 (Stata Corp, 2013) was used for all the statistical analyses. In order to take into account the survey design, the svyset and svy prefix commands available in STATA 13.1 were used (Korn & Graubard, Reference Korn and Graubard1990).
Results
Differentials in early initiation of breastfeeding
Table 2 shows a clear demarcation in the early initiation of breastfeeding between high and low NMR states. Only 35% of children born in the 2 years prior to the survey date in high NMR states were put to the breast immediately after birth compared with 52% in low NMR states. The low proportion of children put to the breast immediately after birth in Uttar Pradesh (27%), Rajasthan (29%), Madhya Pradesh (36%) and Bihar (37%) pulled down the overall proportion in the high NMR group of states. In this group, the level of EIBF was found to be highest in Odisha (71%), followed by Chhattisgarh (49%). On the other hand, the EIBF level varied from 32% each in Punjab and the National Capital Territory (NCT) Delhi to 69% in Assam in the low NMR group.
Table 2. Prevalence of early initiation of breastfeeding among surveyed children under 2 years in high and low NMR states, India, 2015–16
Source: authors’ estimates, NFHS-4 (2015–16). State weights were used to get state-level prevalence, but the state-specific samples (n) were unweighted. National weights were used to generate group and national-level prevalence and sample (n).
Socio-demographic and maternal health care situation in high and low NMR settings
Table 3 shows the percentage distribution of newborns across background characteristics that have a direct bearing on early initiation of breastfeeding in the high and low NMR states. The high NMR states had a lower proportion of first order births (33% vs 43%), a higher proportion of home deliveries (24% vs 11%), a lower proportion of Caesarean deliveries (10% vs 27%) and a higher proportion of deliveries not assisted by trained personnel (21% vs 10%) than the low NMR states. Only 35% of mothers received advice on breastfeeding during the last 3 months of pregnancy in the high NMR states compared with nearly half (49%) of mothers in the low NMR states. The proportion of mothers not exposed to any kind of mass media was higher (39%) in the high NMR states than the low NMR states (15%). Other unfavourable background characteristics of newborns in the high NMR group included: belonging to rural areas (82%), belonging to the poorest wealth quintile (37%) and born to non-literate mothers (39%); the corresponding figures for newborns in the low NMR group were 64%, 13% and 16%.
Table 3. Percentage distribution of surveyed children under 2 years by background characteristics in high and low NMR states, India, 2015–16
Source: authors’ estimates, NFHS-4 (2015–16). Sample and percentage distributions generated using national weights.
Factors associated with early initiation of breastfeeding
Table 4 shows the prevalence of early initiation of breastfeeding by socio-demographic, economic and maternal care characteristics separately for low and high NMR states. Early initiation of breastfeeding was much higher for low NMR states than high NMR states across all the selected characteristics. The chi-squared test indicated that most of selected characteristics were significantly associated with early initiation of breastfeeding, except for mother’s age at birth and caste/tribe in the low NMR group of states and sex of child in both groups.
Table 4. Estimated level of early initiation of breastfeeding among surveyed children under 2 years by background characteristics in high and low NMR states, India, 2015–16
Source: authors’ estimates, NFHS-4 (2015–16); ns=not significant.
In the high NMR group, breastfeeding initiation immediately after birth was least common (22%) among children whose mothers underwent a Caesarean section (C-section). Early initiation of breastfeeding was relatively high among children whose mothers were advised on breastfeeding during ANC visits or those who belonged to Scheduled Tribes (42% each). In the case of the low NMR group, the highest proportion of early initiation of breastfeeding (57%) was observed among children whose mother had a vaginal (normal) delivery, whereas the lowest proportion was among those who underwent a C-section (40%). The level of EIBF was also low among children belonging to the richest wealth quintile (47%), followed by those born of an unplanned pregnancy (48%).
Two factors – type of delivery and advice on breastfeeding during ANC visits – showed significant difference with respect to the reference group in the low as well as the high NMR groups (16 and 15 percentage points respectively), but the latter created a stronger difference in the prevalence of EIBF (10 percentage points) in the high NMR group than in the low NMR group (3 percentage points). Interestingly, no rural–urban and religious differentials appeared in the proportion of children put to the breast immediately after birth in the low NMR group, whereas rural and Hindu children had an advantage over their counterparts by 4 and 6 percentage points, respectively, in the high NMR group.
Determinants of early initiation of breastfeeding in low and high NMR settings
Table 5 illustrates the results of the logistic regression analysis, predicting the adjusted odds ratios (AOR) of immediate initiation of breastfeeding. Type of delivery (vaginal/Caesarean) and place of delivery (home/institution) had large independent effects on EIBF in both NMR groups of states. Caesarean section delivery, compared with vaginal delivery, reduced the odds of EIBF by half (AOR=0.475, 95% CI=0.439–0.514 in the high NMR group; AOR=0.494, 95% CI=0.470–0.519 in the low NMR group). Children born in health institutions had significantly higher adjusted odds of early breastfeeding than those born at home: 30% (AOR=1.297; 95% CI=1.193–1.410) in the high NMR group and 41% (AOR=1.412; 95% CI=1.280–1.558) in the low NMR group.
Table 5. Adjusted odds ratios (AOR) for early initiation of breastfeeding among surveyed children under 2 years in high and low NMR states, India, 2015–16
Source: authors’ estimates, NFHS-4 (2015–16).
Ref.=reference category; AOR=Adjusted Odds Ratio; ns=not significant.
Newborns born to mothers who had received advice on breastfeeding from frontline health workers during the last trimester of pregnancy had 1.4 times (AOR=1.410; 95% CI=1.351–1.471) and 1.2 times (AOR=1.117; 95% CI=1.071–1.166) higher odds of having EIBF in the high and the low NMR groups respectively. In both groups of states, children born of an unplanned pregnancy had at least a 12% lower chance of EIBF than those born of a planned pregnancy. Births of orders 2–3 had 12–15% higher odds of being breastfed early than did first births.
In the case of the high NMR group, two other potential programme variables, namely the type of person who performed the delivery and mother’s mass media exposure, had a considerable influence on EIBF, with 18% (AOR=1.176; 95% CI=1.078–1.282) and 9% (AOR=1.086; 95% CI=1.031–1.143) higher odds, respectively, compared with their counterparts in the reference group. In the high NMR settings, children born to mothers aged 25–34 years had 11% higher odds (AOR=1.107; 95% CI=1.054–1.162) of being breastfed early than did children born to mothers aged 15–24 years.
Mother’s educational attainment and household wealth quintile were two cogent socioeconomic characteristics associated with early initiation of breastfeeding. Mothers who had completed middle level of schooling yielded the maximum independent influence on EIBF in both NMR settings (AOR=1.189, 95% CI=1.113–1.258; and AOR=1.333, 95% CI=1.1245–1.426, in the high and low NMR groups, respectively). On the other hand, the adjusted odds of EIBF decreased significantly with increasing household wealth quintile, as seen from the AOR being only 0.68 times that of the poorest quintile for the richest quintile in the high NMR group. In the case of the low NMR group, children born in the richest households had significantly lower odds of receiving EIBF (AOR=0.770; 95% CI=0.698–0.849) than their counterparts from the poorest households.
In both groups of states, non-Hindu children had more disadvantages in EIBF compared with Hindu children, and more so in the high NMR group (AOR=0.876; 95% CI=0.824–0.932). Caste/tribe background of a newborn played a significantly role in being breastfed early in the high NMR group but not in the low NMR group. Babies born to all other caste groups had significantly lower odds of EIBF than babies born in Scheduled Tribes. In contrast, in the low NMR group, a rural newborn had 14% lower odds of having EIBF (AOR=0.864; 95% CI=0.823–0.907) compared with its urban counterparts.
Discussion
This study highlights that the practice of breastfeeding immediately after birth is low in India, irrespective of an area’s NMR level. The estimates of prevalence of EIBF in 2015–16 indicate that only 44% of children received breastmilk immediately after birth, which is similar to the rates documented for India in the past (Patel et al., Reference Patel, Banerjee and Kaletwad2013; UNICEF, 2014). India as a whole lags behind Bangladesh (45%) and Nepal (47%) in early initiation of breastfeeding, and rates are dismally low (35%) in states with high NMR levels. The burgeoning literature on the linkages between early initiation of breastfeeding and newborn survival and well-being (Horta et al., Reference Horta, Victora and World Health2013; Radmcher & Adamkin, Reference Radmacher and Adamkin2017) suggests that it is imperative for these states to improve their levels of early initiation of breastfeeding. Some states in the low NMR group, namely NCT Delhi, Punjab, Jharkhand, Andhra Pradesh and Haryana, also need to make substantial improvements in EIBF. The findings indicate that further research is needed in the case of Odisha, where neonatal mortality is high despite a relatively high level of early initiation of breastfeeding (71%). Here, other factors could affect neonatal mortality, such as cord care, birth defects, environmental and sanitation factors, treatment-seeking in the case of acute respiratory and diarrhoeal infections and availability of quality emergency newborn care services (Yadav et al., Reference Yadav, Menon, Mitra, Allarakha and John2020).
The study analysis revealed that having a C-section delivery and receiving advice on breastfeeding during pregnancy had the largest independent effects on the odds of early initiation of breastfeeding across all backgrounds in high and low NMR settings. However, these two factors differed in the direction of their effects, the former suppressing the chance of EIBF significantly, and the latter improving it with equal intensity. Some previous studies have documented the unnecessary and unethical practice of C-section and its huge out-of-pocket burden on families in India and neighbouring developing countries (Mohanty & Srivastava, Reference Mohanty and Srivastava2013; Joshi et al., Reference Joshi, Angdembe, Dae, Ahmed, Faruque and Ahmed2014). The present study suggests that the unnecessary use of C-section delivery negatively impacts early initiation of breastfeeding. A C-section delivery may make a child more susceptible to life-threatening infections, especially when the delivery takes place in poor hygienic conditions and without basic health infrastructure (Hanson & Korotkova, Reference Hanson and Korotkova2002; Patel et al., Reference Patel, Banerjee and Kaletwad2013; Buckles & Kolka, Reference Buckles and Kolka2014; Ekubay et al., Reference Ekubay, Berhe and Yisma2018). However, the widespread practice of prescribing formula milk for neonates is not only limited to C-section deliveries but is also common in the case of normal deliveries in for-profit private institutions (Karim et al., Reference Karim, Billah, Chowdhury, Zaka, Manu, Arifeen and Khan2018). A recent study conducted using Ethiopian Demographic and Health Survey data found that the odds of EIBF were significantly lower (AOR=0.10; 95% CI=0.06–0.15) in the case of normal deliveries compared with C-section deliveries in Ethiopia (Woldeamanuel, Reference Woldeamanuel2020).
The hypothesis that mother receiving advice on breastfeeding during the last trimester of the pregnancy has a significant effect on the early initiation of breastfeeding was confirmed by the results of this study. That providing counselling or support to mothers makes a positive change in their breastfeeding behaviour has been found to be true in both developing and developed country settings in separate clinical trials (Dhandapany et al., Reference Dhandapany, Bethou, Arunagirinathan and Ananthakrishnan2008; Blixt et al., Reference Blixt, Johansson, Hildingsson, Papoutsi and Rubertsson2019; Sehhatie et al., Reference Sehhatie, Mirghafourvand and Havizari2020). Hence, improvements in the proportion of mothers advised on breastfeeding during ANC visits in India from the current one-third to the entire pregnant women’s cohort could substantially increase early breastfeeding initiation, particularly in high NMR settings.
Recent literature cites breastfeeding pain, perception of lack of sufficient milk, inappropriate latching and the child’s inability to suckle as prominent barriers to the early initiation of breastfeeding in India (Wagner et al., Reference Wagner, Chantry, Dewey and Nommsen-Rivers2013; Kronborg, et al., Reference Kronborg, Harder and Hall2015; Naroee et al., Reference Naroee, Rakhshkhorshid, Shakiba and Navidian2020). India’s programme managers at the local level should first assess the need for counselling and build the capacity of frontline workers to offer high-quality counselling and support to women on early initiation and exclusive breastfeeding. This study also indicated that first-time mothers need additional support to encourage them to start breastfeeding early. Programmes aimed at increasing the prevalence of EIBF among mothers should aim to promote colostrum intake and 6-month exclusive breastfeeding and discourage the use of formula milk. Nulliparous pregnant women require special attention because the odds of putting the child to the breast immediately after birth were lower among primiparous mothers than others in both NMR settings. Other studies have made similar observations (Kronborg et al., Reference Kronborg, Harder and Hall2015; Woldeamanuel, Reference Woldeamanuel2020).
Another significant finding of this study was that women who had non-institutional deliveries had higher odds of delaying breastfeeding compared with those who had institutional deliveries – 30% higher in the high NMR group and 41% higher in the low NMR group. Other studies have also found that institutional delivery has a positive and independent effect on early initiation of breastfeeding (Edmond et al., Reference Edmond, Zandoh, Quigley, Amenga-Etego, Owusu-Agyei and Kirkwood2006; Senarath et al., Reference Senarath, Fernando and Rodrigo2007). This study also shows that delivery by a trained person can also improve EIBF by 18% in high NMR settings. Therefore, achieving universalization of institutional and trained person assisted delivery, especially in high NMR states, is expected to yield higher returns on EIBF.
Son preference is a long-standing social and public health issue in India. Prior studies have documented that son preference remains instrumental in childhood mortality and fertility preferences, and varies in intensity across the country (Arnold, Reference Arnold2008; Ahmed & Ray, Reference Ahmed and Ray2014). However, neither of the NMR groups in this study showed any sex difference in EIBF. This finding is supported by a previous Indian study, which found no sex difference in colostrum breastfeeding too (Dwivedi, Reference Dwivedi2010). Furthermore, the present study indicated that the children of unplanned births had 13–17% less chance of being put to the breast early compared with those of planned pregnancies in both NMR groups. Unplanned births have a high risk of early death, and a plausible pathway for this could be the delayed initiation of breastfeeding.
Mother’s education did not predict a clear trend of effects on EIBF in the high NMR group but there was a general effect of increasing education being associated with a higher odds of EIBF. It could be because the proportion of no or low educational attainment among women aged 15–49 years was much higher (54%) in the former than the latter group (28%). Therefore, this study hints that a minimum educational attainment of up to middle level is needed to enhance mothers’ self-efficacy in early initiation of breastfeeding and to help them realize its advantages for the child’s as well as their own health (Amstrong & Reilly, Reference Armstrong and Reilly2002; Chudasam et al., Reference Chudasama, Patel and Kavishwar2009; Lamberti et al., Reference Lamberti, Walker, Noiman, Victora and Black2011). In the absence of such a threshold, increasing the proportion of women exposed to mass media from 39% to universal exposure could greatly help build the required capacity to avoid the delay in early initiation of breastmilk, particularly in the high NMR group.
The study also found that household wealth was significantly associated with putting children to the breast early. The probability of early breastfeeding initiation decreased as households moved from the poorest to the richest quintile in the high NMR group and from the richer to the richest quintile in the low NMR group. This pattern may be related to the availability of alternative food and formula food to well-off children and the type of advice and counselling received by mothers from health providers. The myth of body disfiguration after breastfeeding among economically more prosperous women may also be a reason (Joshi et al., Reference Joshi, Angdembe, Dae, Ahmed, Faruque and Ahmed2014; Ogbo et al., Reference Ogbo, Dhami, Awosemo, Olusanya, Olusanya and Osuagwu2019). Previous studies have also shown an association between household economic status and breastfeeding practices in India and Bangladesh (Dwivedi, Reference Dwivedi2010; Joshi et al., Reference Joshi, Angdembe, Dae, Ahmed, Faruque and Ahmed2014). The higher EIBF among the poorer households could also be due to their exposure to the government-run Integrated Child Development Scheme (ICDS), aimed at poor women, predominantly those in rural areas, to promote safe motherhood and child survival. With their participation in the ICDS, poor mothers may be taking full advantage of the scheme in the absence of their own resources (Tiwari et al., Reference Tiwari, Mahajan and Lahariya2009).
Place of residence of mothers was only associated with early initiation of breastfeeding in the low NMR group of states, but the positive independent effects of urbanization (14%) on early initiation of breastfeeding were not in the same direction as found by Woldeamanuel (Reference Woldeamanuel2020) in Ethiopia (29% lower in the urban area).
Non-Hindu children and children other than those from Scheduled Tribes in the high NMR groups were significantly more likely to be deprived of early breastfeeding. In the case of religion, a particularly large sect of the Hindu religion portrays mother–child affection and bonding through pictures and religious folkloric stories of the mother breastfeeding the newborn child. It is likely that there is less stigma and inhibition associated with breastfeeding among Hindus, which may be one of the reasons behind their higher odds of early breastfeeding initiation. Tribal children have traditionally had an edge over others in initiating early breastfeeding, perhaps due to the unavailability of alternative food to young children in tribal communities. However, Khedkar (Reference Khedkar2020) found that in the Korku tribe of Madhya Pradesh a newborn child is culturally put to the breast after two days of birth and in the first two days of life only cow’s buttermilk is applied to child’s lips.
This study infers that higher levels of early breastfeeding initiation in India could be achieved by tweaking current maternal and child health care programmes and their promotional strategies. States with high NMR should be given priority as early initiation of breastfeeding is directly linked with child survival, especially as 40% of all neonatal deaths occur on the first day of life and 56% during the first 3 days (ICMR Young Infant Study Group, 2008). Achieving universal early initiation of breastfeeding could be one of the most cost-effective ways of achieving SDG-3 at the national and sub-national levels in India.
The study’s core strength is the generalizability of its findings as it used nationally representative population-based survey data. To the author’s knowledge, this is the first study to assess the potential determinants of early initiation of breastfeeding in low and high NMR settings in India. The study findings could help programme managers at the national level to identify unique and common priorities of the states to improve EIBF according to their neonatal mortality situation. The study shows that minor tweaks to ongoing maternal and child health care programme strategies could significantly improve early breastfeeding initiation irrespective of the neonatal mortality status of a state. However, the study had its limitations. First, the study filtered children born in the 2 years prior to the survey in order to avoid recall lapses among mothers. However, some women might not have accurately reported when the child was first time put to the breast. Secondly, in general, large-scale household surveys only collect information for children surviving on the day of the survey, which poses a challenge of self-selection bias in studying factors affecting early initiation of breastfeeding, especially in the high NMR settings.
In conclusion, this study has established that early initiation of breastfeeding is found to be low in India as a whole, and that there was a wide gap between low and the high NMR settings. It also suggests that early breastfeeding initiation can be improved significantly by ensuring access to quality counselling and support on breastfeeding to all women and enhancing mass media exposure, especially to non- or less-literate women. There should be a focused strategy to improve early breastfeeding among women who have Caesarean sections. Furthermore, especially high NMR states should focus on increasing institutional and skilled personnel assisted deliveries, and exposure to mass media messaging on the health benefits of early initiation of breastfeeding to mother and child. The Pradhan Mantri Surakshit Matritva Abhiyan (PMSMA) – a health programme launched by the Government of India in 2016 – aims to assure free-of-cost comprehensive and quality antenatal care to all pregnant women on the 9th of every month. This study recommends including an extensive session on the necessity of the early initiation of breastfeeding during women’s monthly PMSMA visits to health facilities. This will significantly increase the chance of early initiation of breastfeeding, and therefore child health and survival, in India. Finally, this study recommends that national schemes like the ICDS and India’s Reproductive, Maternal, Newborn, Child and Adolescent (RMNCH+A) health strategy should urgently redesign their priorities on the early initiation of breastfeeding in laggard states.
Acknowledgments
The authors are grateful to three anonymous reviewers and the Associate Editor of the journal for their technical and editorial comments and inputs on the earlier draft of the paper. The analysis and inferences drawn in this research article are solely those of the authors and do not necessarily reflect the official policy or position of their respective affiliated institutions or their governing bodies.
Funding
No funding was received for this study.
Conflicts of Interest
The authors have no conflicts of interest to declare.
Ethical Approval
This work used the secondary dataset of the NFHS-4, which is available in the public domain and is fully anonymized for further research. Therefore, ethical approval was no required for this study. Nevertheless, the NFHS-4 survey sought the necessary ethical approval from the Ethical Review Board of the International Institute for Population Sciences, Mumbai (IIPS & ICF, 2017).
