Consumption of breast milk, formula and other non-human milk by children aged under two years: analysis of 86 low and middle income countries

Background Consumption of breast milk substitutes (BMS) by children aged under six months in low and middle income countries (LMICs) is directly driven by country income and family wealth. Multi-country investigations on the consumption of BMS by older children (6-23 months) are lacking. Methods Using data from 86 nationally representative surveys carried out in LMICs from 2010 onwards, we analyzed the prevalence of continued breastfeeding at one and two years, and frequency of consumption of formula and other non-human milk by age in months. Indicators were estimated through 24-hour dietary recall. Absolute and relative wealth indicators were used to describe within- and between-country socioeconomic inequalities. Results were stratified by country income groups. Findings Breastfeeding declined sharply as children became older in all LMICs, especially in upper-middle income countries. Formula consumption peaked at six months of age in low and lower-middle income countries, and at around 12 months in upper-middle income countries. Consumption of formula at any age higher in children from wealthier families in all countries, while breastfeeding was more common among poor children. Multilevel linear regression analysis showed that consumption of formula was positively associated while breastfeeding was negatively associated with absolute national income. Factors at country level explained a substantial proportion of overall variability in formula use and breastfeeding. Interpretation Infant and young child feeding practices vary strongly according to wealth, both within and between countries. Breastfeeding falls sharply as children become older, especially in wealthier families living in upper-middle income countries; this is also the group with highest formula consumption at any age. Country-level factors play an important role in explaining BMS consumption by all family wealth groups in LMICs, suggesting that formula marketing at national level may be partly responsible for the observed differences.