TIMING OF INITIATION OF BREASTFEEDING AND EARLY-NEWBORN SEPSIS: EVIDENCE FROM RURAL BANGLADESH

Background and objectives: In Bangladesh, neonatal deaths decreased by 46% during 1990-2014. However, it remains high (28.0/1000 live-births), with early-neonatal deaths (19.3/1000) contributing over 65% of neonatal deaths. Neonatal sepsis is the cause of 24% neonatal deaths, while only 36% women start to breastfeed their newborn within 1st hour of birth. Early initiation of breastfeeding has a positive impact on neonatal health. We aim to explore the effect of timing of breastfeeding initiation on early-newborn sepsis in rural Bangladesh. Methods: We used data from a large community based trial in Bangladesh where pregnant women were enrolled from 2013-2015 covering 30,646 newborns. Sepsis was defined using neonatal danger signs reported by "The Young-Infants Clinical Science Study Group". Timing of initiation of breastfeeding was categorized as within 1hr, 1-24 hr,24-48hr, and more than 48hr of birth. Analysis includes descriptive statistics, risk attribution and binomial logistic-regression while adjusting for clustering effects of the trial design, and other maternal and infant characteristics. Results: 29,172 live-births had information on initiation of breastfeeding. Of these, 4,121(14.1%) neonates had signs of sepsis till 7th day after birth. Two-thirds of the newborns [68.3%,(n=19,913)] had breastfeeding initiated within 1hr of birth. Mean time to initiation was 3.8hrs and by the end of 48hrs all but 3.5% neonates were breastfed. Breastfeeding within first hour was lower among women who delivered through caesarean section [46.5% vs 72.8%,(p