Cumulative Lactation and Metabolic Disease in African American Women Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 20, 2019
  • Chetwynd, Ellen
    • Affiliation: Gillings School of Global Public Health, Department of Maternal and Child Health
  • Current literature suggests that breastfeeding duration is inversely associated with maternal metabolic diseases such as diabetes and hypertension. These conditions disproportionately impact black women in the United States, yet breastfeeding prevalence is lower for black women. We used the Black Women’s Health Study (n = 59,001) to quantify the association between breastfeeding and hypertension (Aim 1) and type 2 diabetes (T2D) (Aim 2). We introduce the use of a nested case control design to reduce selection and survival bias. Specifically, we used density sampling, frequency matching controls to cases by questionnaire cycle and risk sets for age. Effect measure modification analyses were performed using likelihood ratio testing for age, parity, and body mass index at age 18. In Aim 1, 12,513 hypertensive cases were matched 1:2 with controls (n = 37,539) while Aim 2 matched 4505 cases 4:1 to controls (n = 22,525). Breastfeeding was associated with T2D (OR 0.93, 95% CI: 0.85, 1.01) and weakly associated with incident hypertension (OR 0.97, 95% CI: 0.92, 1.02). The association was strongest for women in their forties for both hypertension (OR 0.92, 95% CI: 0.85, 0.99) and T2D (OR 0.86, 95% CI: 0.74, 1.00). As breastfeeding duration increased, rates of both hypertension and T2D decreased in multivariate models (P for trend <0.01). For both hypertension and T2D, the strongest associations were present for women who had a normal BMI and fewer children. The outcomes estimates for our study were farther from the null than previous studies using Cox proportional hazard models, suggesting a reduction in selection bias. They are also closer to the null than previous studies using cross-sectional or retrospective models, suggesting a reduction in survival bias. Our results extend the existing research by providing outcomes in a high priority group of women previously under-represented, as well as introducing methodology to this area of research that has the potential to reduce bias in future studies on breastfeeding and metabolic health. Our outcomes support the need for inclusive messaging in public health and medical care encouraging breastfeeding for all women, particularly those at risk for T2D or hypertension.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Stuebe, Alison
  • Halpern, Carolyn
  • Rowley, Diane
  • Palmer, Julie
  • Troester, Melissa
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2016

This work has no parents.