Abstract
In this issue of Breastfeeding Medicine, Ware and colleagues reviewed birth and death records of Shelby County, Tennessee, for a 10-year period in which Shelby County saw a rise of breastfeeding rates due to many state and local efforts. In Shelby County, Ware and colleagues were able to demonstrate that any breastfeeding was associated with a decrease in neonatal mortality among the entire population, including the majority black population, and postneonatal mortality among the non-black population. Shelby County, home to Memphis, has a black population of 53.4%.
During this timeframe studied, 2004–2014, Shelby County's safety net neonatal intensive care unit (NICU) went from having just 22% of infants receiving some breast milk to 88%. We do not know what percentage of these infants were black, but given the disproportionate number of preterm births among black mothers, it is possible that that black infants were over-represented among the neonatal population studied, and may account for the mortality benefit seen in black neonates. There was no mortality benefit among the black population in postneonatal deaths. The authors postulated that the “salutary effects of breastfeeding could also be blunted due to other factors (such as racism, poverty, stress, or violence), or the sample size may have been too small to show a statistically significant difference.”
Chen and Rogan used a case control study of 1988 data and found that breastfeeding is associated with a decrease in postneonatal death in the United States, adjusting and accounting for race and birthweight. 1 They intentionally excluded neonates to reduce the possibility of reverse causality, in which only healthy infants breastfeed and infants too sick to breastfeed do not. Like Chen and Rogan, Ware and colleagues tried to account from reverse causality, but they did so by eliminating deaths under just 7 days, and they also eliminated deaths from malignancy and congenital defects. By intentionally including other neonates, Ware and colleagues were able to capture the effect of the increase in breast milk feeding seen in the Shelby County's main safety net NICU.
My work with my colleagues has been able to show epidemiologic estimates of excess deaths from suboptimal breastfeeding in the United States based on odds ratios from the literature for specific potentially lethal conditions affected by breastfeeding, combined national mortality statistics, and known national breastfeeding rates and mathematical modeling to calculate estimates of U.S. deaths due to suboptimal breastfeeding. 2 Our work has the advantage of being able to show the impact of breastfeeding on maternal mortality rates in addition to pediatric mortality rates.
Ware and colleagues' article adds to the body of work showing that breastfeeding is an important strategy for preventing infant mortality in the United States, which has an alarming rate of infant mortality for an industrialized country, at 5.8 per 1,000 in 2017. 3 Their article also, importantly, calls attention to the alarming racial inequities and infant mortality rates seen here in the United States. In the United States, the infant mortality rate is a stunning 11.4 per 1,000 for non-Hispanic blacks, but was 17.4 in Shelby County.
Ware and colleagues appropriately point out that “When evaluating infant mortality statistics, ‘not breastfeeding’ should be included as a risk factor.” They also note that infant feeding assessment should always be included in the data collection and analysis for infant mortality statistics. We are used to seeing breastfeeding as a metric for developing countries but we see breastfeeding is an important metric for U.S. infant mortality as well.
Many studies have shown that breastfeeding saves lives in low-to-middle income countries, such as the Lancet series, 4 which estimated that scaling up breastfeeding to a near universal level could save 823,000 lives of children up to age 5 years annually. It is important to recognize that breastfeeding is important for infant mortality in the industrialized world as well.
One of the biggest causes of infant mortality is sudden infant death syndrome (SIDS) (following birth defects, preterm birth, and pregnancy complications). Currently, the two demographic groups with the highest rates of SIDS in the world are American Indians/Alaskan Natives and non-Hispanic blacks. 5 Breastfeeding is associated with a markedly reduced risk of SIDS, 6 with any breastfeeding for at least 2 months cutting the risk of SIDS by half, and with breastfeeding for at least 6 months, cutting the risk by nearly two-thirds. Black Americans have markedly lower breastfeeding rates than any other U.S. demographic group, although the gap is starting to close. However, they also have markedly disproportionate rates of preeclampsia, 7 preterm births, 8 and pregnancy-related deaths that are more than three times that of white women. 9 These factors are all related, and cannot be considered in isolation from infant mortality, including SIDS, and breastfeeding. They speak of an urgent need to invest in maternal–child health infrastructure in the areas that serve black communities. Breastfeeding is one part of this greater picture of maternal and child health, but an important one.