National Trends in the Provision of Human Milk at Hospital Discharge Among Very Low-Birth-Weight Infants

Abstract

Importance: Human milk confers important health benefits to very low-birth-weight (VLBW) infants (≤1500 g). The extent to which the use of human milk has changed over time and the factors associated with human milk use nationally in this population are poorly understood.

Objectives: To describe US trends in the provision of human milk at hospital discharge for VLBW infants during the past decade according to census region and maternal race/ethnicity, quantify associations of census region and maternal race/ethnicity with the provision of human milk at hospital discharge, and examine regional and state variations in any provision of human milk at hospital discharge among racial/ethnic groups.

Design, setting, and participants: A cohort study was conducted of 346 248 infants, born at 23 to 29 weeks' gestation or with a birth weight of 401 to 1500 g, who were cared for at 802 US hospitals in the Vermont Oxford Network from January 1, 2008, to December 31, 2017. The US census region was categorized as West, Midwest, Northeast, and South (reference). Maternal race/ethnicity was categorized as non-Hispanic white (reference), non-Hispanic black, Hispanic, Asian and Pacific Islanders, and Native American.

Main outcomes and measures: Any provision of human milk at hospital discharge, defined as the use of human milk as the only enteral feeding or the use of human milk in combination with fortifier or formula.

Results: Of the 346 248 infants in the study (172 538 boys and 173 710 girls), 46.2% were non-Hispanic white, 30.1% were non-Hispanic black, 18.3% were Hispanic of any race, 4.7% were Asian and Pacific Islanders, and 0.8% were Native American. Any provision of human milk at hospital discharge increased steadily among all infants, from 44% in 2008 to 52% in 2017. There were increases across all US census regions and racial/ethnic groups. Any provision of human milk at hospital discharge was higher in the West (among singleton births: adjusted prevalence ratio, 1.32; 95% CI, 1.25-1.39; among multiple births: adjusted prevalence ratio, 1.28; 95% CI, 1.21-1.35) and Northeast (among singleton births: adjusted prevalence ratio, 1.11; 95% CI, 1.04-1.19; among multiple births: adjusted prevalence ratio, 1.11; 95% CI, 1.04-1.19), compared with the South, and was higher among Asian mothers (among singleton births: adjusted prevalence ratio, 1.21; 95% CI, 1.18-1.25; among multiple births: adjusted prevalence ratio, 1.12; 95% CI, 1.09-1.15) and lower among Hispanic (among singleton births: adjusted prevalence ratio, 0.98; 95% CI, 0.96-1.01; among multiple births: adjusted prevalence ratio, 0.88; 95% CI, 0.86-0.91), Native American (among singleton births: adjusted prevalence ratio, 0.64; 95% CI, 0.59-0.70; among multiple births: adjusted prevalence ratio, 0.59; 95% CI, 0.50-0.69), and non-Hispanic black mothers (among singleton births: adjusted prevalence ratio, 0.67; 95% CI, 0.65-0.70; among multiple births: adjusted prevalence ratio, 0.57; 95% CI, 0.54-0.60), compared with non-Hispanic white mothers. These results were robust to adjustment for birth year and infant characteristics. Wide regional and state variations were found in any provision of human milk at hospital discharge.

Conclusions and relevance: Overall prevalence of any provision of human milk at hospital discharge among VLBW infants has steadily increased during the past decade. Disparities by US region and race/ethnicity in the provision of human milk exist and have not diminished over time.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Parker reported serving as a volunteer member of the Mother’s Milk Bank Northeast research advisory board. Ms Greenberg reported receiving compensation from the Vermont Oxford Network for serving as a statistician. Dr Edwards reported receiving grants from the Vermont Oxford Network for serving as the Director of Data Science during the conduct of this study. Dr Ehret reported being the director of global health for Vermont Oxford Network. Dr Belfort reported receiving grants from and serving as a consultant for Miris AB (Uppsala, Sweden) outside the submitted work, and serving as a volunteer member of the Mother’s Milk Bank Northeast research advisory board. Dr Horbar reported receiving compensation from Vermont Oxford Network for serving as Chief Executive and Scientific Officer outside the submitted work. No other disclosures were reported.

Figures

Figure 1.. Prevalence of Any Provision of Human Milk at Hospital Discharge From 2008 to 2017

A, Prevalence according to US region. Increases in estimated prevalence per year were West, 0.7%; Northeast, 1.6%; Midwest, 0.9%; and South, 1.2%. B, Prevalence according to maternal race/ethnicity. Increases in estimated prevalence per year were Asian and Pacific Islanders, 0.8%; non-Hispanic white, 0.9%; Hispanic, 1.1%; Native American, 1.2%; and non-Hispanic black, 1.3%. Linear trends by year: P = .003 for Native Americans and P < .001 for all other groups.

Figure 2.. US Regional Variations in Any Human Milk at Discharge According to Maternal Race/Ethnicity

The vertical lines indicate 95% CIs.

Figure 3.. US State Variation in Any Provision of Human Milk at Discharge According to Maternal Race/Ethnicity

US states and Washington, DC, were ranked from lowest to highest with respect to difference in prevalence of any provision of human milk at discharge among minorities vs white mothers with 95% CIs. A value of 0 represents no difference in the prevalence of any provision of human milk among racial/ethnic groups, a negative value represents a lower prevalence of any provision of human milk among a minority group compared with white mothers, and a positive value represents a higher prevalence of any provision of human milk among a minority group compared with white mothers. Significant racial/ethnic disparities are indicated by an error bar that does not cross the center line. Only states with at least 5 infants in each racial/ethnic category are shown.