April 19, 2022Each year, approximately 3,400 babies die suddenly and unexpectedly in the United States. The risk of sudden unexpected infant death syndrome (SUID) varies across demographic groups. Trying to make sense of these disparities, Melissa Bartick, assistant professor of medicine at Harvard Medical School and MPH student at Harvard TH Chan School of Public Health, looked at one possible explanation — breastfeeding — in a study co-authored with Henning Tiemeier, Sumner Professor and Esther Feldberg in Maternal and Child Health.

Q: What are some of the SUID disparities among demographic groups, and why is lack of breastfeeding considered a possible explanation for these disparities?

A: SUID is the leading cause of post-neonatal death in the United States, which has one of the highest infant mortality rates among wealthy nations. Among non-Hispanic and Native American/Alaskan black infants, the SUID rate is more than twice the national average of 0.91/1,000 live births; among Hispanic infants, it is well below average; and rates among Asian American infants are among the lowest in the world.

There are many reasons why not breastfeeding could put infants at higher risk. Normally, throughout history, mothers and babies slept together, and the mother nursed the baby through the night. After a feed, a baby rolled over onto his back. Both mother and child would be easily awake and their sleep cycles would be synchronized, so if something threatened the child’s airway, the mother would wake up and be able to respond. In the age of formula feeding, however, infants sleep more soundly, sleep alone, and can numb their stomachs, which can threaten their breathing and increase the risk of death. Additionally, non-breastfed infants have a higher risk of infections, and it is thought that this may also play a role. Given that breastfeeding is less common in black infants, who have a higher rate of SUID, and more common in Hispanic infants, who have a lower rate of SUID, we wondered if breastfeeding could explain that.

Q: How did you go about investigating this topic and what did you find?

A: We looked at four years of US birth certificates from the National Center for Health Statistics — more than 13 million certificates in total — along with their related death certificates. Birth certificates list breastfeeding status at discharge as well as a host of other characteristics about the mother, including whether she smoked, marital status, education level and insurance status, as well as characteristics about the baby, such as whether the baby was born premature or low birth weight.

Our study found that breastfeeding rates differed across demographic groups: while it was 82% overall at hospital discharge, it was 69.5% for non-Hispanic black infants, 75% for Native American/Alaska Native infants, 90% for Asian infants and 87% for Hispanic infants.

We also found that, overall, not breastfeeding at birth was clearly linked to an increased risk of SUID. Even after adjusting for 12 different characteristics of mothers and infants we considered, there was still a 14% increased chance of SUID in infants who were not breastfed. This result reinforces previous findings by considering many additional maternal and infant risk factors and still showing an association between non-breastfeeding and SUID risk.

What we found surprising was that, for non-Hispanic black infants, not breastfeeding explained very little of the disparity in SUID risk – only 2.3% of the higher risk was due to not breastfeeding. And among Hispanic infants, their high breastfeeding rates explained very little of their lower SUID risk — just 2.1 percent. We thought that not breastfeeding would further explain the racial/ethnic disparity in SUID risk. To our knowledge, this is the first study to examine whether breastfeeding is actually responsible for the disparity.

Q: What do you take away from these results?

A: Well, there are two issues here. One takeaway is that other social or structural risk factors likely play a much larger role than breastfeeding in explaining the sudden infant death syndrome disparity. These factors could include child poverty, food and housing insecurity, and lack of generational wealth, which are legacies of historical trauma and structural racism that affect black women more than Hispanic women. The other issue is that our study only looked at breastfeeding initiation, but not longer durations of breastfeeding, so this limitation should be kept in mind.

Of course, it is important to support breastfeeding as it helps to reduce child mortality overall. But to reduce disparities, it is important to address the social determinants of health that require more policy solutions. Otherwise, we cannot really reduce the very large disparities in infant mortality rates in the United States.

Michael Blanding