Rates and determinants of early initiation of breastfeeding and exclusive breast feeding at 42 days postnatal in six low and middle-income countries: A prospective cohort study

A major finding of our study was that the overall rate of EIBF was higher, at 75%,
than has been typically reported in prior studies using DHS survey data 15-18,20-22]. We observed some site-specific variations in EIBF, with the lowest rate observed
in Pakistan. Some variation in rates between sites may have been due in part to health
system-wide disruptions in service delivery (e.g., floods in Pakistan, 2010; health
worker strikes in Kenya, 2012). However, the lower rates of EIBF and EBF observed
in the Pakistan site in the current study have also been noted in prior investigations
23,24]. It is interesting to note that the Pakistan site differs from others within the
Global Network, in that women face many additional risk factors that have been shown,
in previous studies, to interfere with EIBF. These include: higher rates of women
who have no formal education (83%); women with higher parity (47% parity of 2 or more);
later initiation of antenatal care in the 3rd trimester (44%); higher percentage of
babies who required resuscitation (5%); and the lowest rate of babies placed on the
mother’s chest after delivery (7%) 23-26].

Our study also confirmed several factors generally associated with lack of EIBF such
as nulliparity, delivery by caesarean section, the neonate not being put on the mother’s
chest after delivery, multiple births, male gender (Africa and Latin America), low
birth weight, and if the neonate was resuscitated. Our study supported previous research
that delivery by caesarean section is a consistent barrier to EIBF, even in the absence
of any neonatal condition that interferes with early initiation of breastfeeding 16,18,27,28]. This is significant, as it delineates a major interventional target by which to
improve EIBF in resource-limited settings 29], especially given the recent increase in institutional deliveries and caesarean section
rates, particularly in India. There is also a need to reinforce essential newborn
care training and education among health workers and families, with emphasis on immediate
skin to skin contact after delivery and initiation of breastfeeding within the first
hour, especially focusing on low birth weight and premature babies 30-32].

It is unclear why male babies were less likely to have EIBF, but as described elsewhere,
there may be cultural beliefs surrounding the birth of males that discourage immediate
breastfeeding 23,33-35]. Additionally, in some regions, such as Guatemala, cultural factors such as those
related to the belief that colostrum is “dirty” can serve as barriers to EIBF. The
role of nulliparity in lack of EIBF may be related to some interplay between maternal
age, lack of knowledge, and cultural beliefs, but also provides a group that can be
targeted for interventions to improve EIBF rates.

Our study also highlights the importance of EIBF to increase early rates of EBF, at
least through day 42 of life. Edmond et al demonstrated that EIBF has the potential
to save 22% of neonatal deaths and 16% of all infant deaths 8]. Lack of EIBF may, in particular, be related to an increased risk of mortality due
to infection 5,8,36,37]. In our study, lower birth weight is a risk factor for both lack of EIBF and has
also been associated with mortality risk. With increasing survival of lower birth
weight babies in resource limited settings, interventions to improve EIBF rates should
be considered. Our results also highlight the fact that sicker and/or smaller babies
are more likely to have feeding problems overall, including inability to initiate
early breastfeeding, than their heavier and/or healthier counterparts. And yet, the
relationship among birth weight and infant illness is not straightforward; exposure
to breast milk may be even more crucial for reduction of morbidity and mortality outcomes,
including infection, among the most vulnerable newborns 38]. These complex, multi-factorial, bi-directional associations among infant birth weight,
gestational age, delivery complications, maternal characteristics, breastfeeding,
and morbidity and mortality outcomes should be further investigated.

The factors associated with lack of EBF were less consistent across the regions than
the factors associated with failure to achieve EIBF. The only factor that was significant
in all regions for EBF was multiple gestation. However, many of the other factors
examined had a significant relationship in sites in one or more regions and points
to the necessity to understand this issue in the local context. For example, in Guatemala,
several factors including low education, being delivered by a TBA, and starting prenatal
care late were protective against failure to achieve EBF at 42 days. Research to understand
regional differences in EBF is therefore important.

There are several plausible explanations why our EIBF and EBF rates are higher than
in other studies. First, sites in the current study have been part of the Global Network
for a number of years 39-57]. These sites have participated in a variety of cluster-based randomized trials to
improve maternal and neonatal health including: training of community-based health
providers in essential newborn care 46,47]; Emergency Obstetric and Newborn Care 48,49]; Helping Babies Breathe neonatal resuscitation training 50-52]; complementary feeding 53-55] and antenatal corticosteroids 56]. In addition to the MNHR itself, several site-specific efforts have focused on improving
case-finding and reporting for a variety of maternal and newborn outcomes 47,58] as well as improved description and classification of facility-based and lay health
services 59]. As a result of exposure to these maternal and newborn health care initiatives, it
is likely that there is heightened awareness of women, health workers, community opinion
leaders, and family stakeholders in these settings about the importance of EIBF and
EBF for 6 months.

Second, since data on EIBF is collected shortly after birth, and assessment of EBF
occurs on day 42 post-partum, this likely reduces the risk of maternal recall bias,
which may have impacted results from cross-sectional DHS surveys and other similar
studies. We believe that our data are representative of our communities because the
MNHR has high rates of consent, low rates of loss-to-follow up, and well defined variables
such as antenatal, delivery, newborn/infant morbidity and mortality, and maternal
outcomes.

There are several important strengths of our study. First, we prospectively determined
rates of EIBF and EBF at 42 days in a large cohort of women and their babies at 7
sites in 6 low and middle income countries. Second, we prospectively collected data
on barriers to EIBF and EBF in a standardized manner, by trained health workers, over
a four-year period. Limitations of our study include our reliance on maternal recall
about precisely when they initiated breast feeding, although this information is collected
within hours or days of birth, not at variable times over months and years. Our rates
of EBF at 42 days are based on maternal report using a 24 hour recall method; we did
not confirm these maternal reports via observations. In addition, we did not differentiate
between “predominant” and “exclusively” breastfed—it is possible that some women may
have mistakenly reported that their baby was exclusively breastfed when they were
predominantly breastfed. Finally, rates of EIBF and EBF may have been overestimated,
particularly if mothers provided a desirable response to the breastfeeding questions
because they were familiar the data collectors, and/or had a desire to provide socially
appropriate responses 60,61].

A recent systematic review noted the paucity of high-quality data for the “understanding
of the independent or combined effects of early initiation and breastfeeding patterns”
36]. We believe that our large, prospective, population-based study of live born neonates
at seven sites in Africa, Asia, and Latin America adds to the global evidence-base
about risk factors and outcomes of lack of EIBF and EBF. Our study provides an evidence
base for specific barriers, within particular global settings, that should be targeted
by interventions to improve rates of EIBF and EBF.