. Author manuscript; available in PMC: 2019 Dec 1.

Published in final edited form as: J Pediatr. 2018 Aug 30;203:47–54.e4. doi: 10.1016/j.jpeds.2018.07.066

Table V.

Multivariable models of the association between breastfeeding vs formula feeding and the infant intestinal microbiome, stratified by infant race/ethnicity

	β adjusted for demographic and dietary variables (95% CI)	African American (n = 120), Adjusted β (95% CI)	Hispanic (n = 80), Adjusted β (95% CI)	White (n = 50), Adjusted β (95% CI)	Breast milk and race/ethnicity interaction P value
Bifidobacterium	0.56 (0.12, 1.00) P = .01	0.38 (−0.37, 1.13) P = .32	0.52 (−0.14, 1.19) P = .12	0.50 (−0.57, 1.56) P = .36	.10
Lactobacillus	3.50 (2.14, 4.86) P < .001	6.35 (3.99, 8.71) P < .001	2.53 (0.14, 4.91) P = .04	3.08 (−0.66, 6.82) P = .11	.09

Significant (P <.05) results are in bold.

Results shown for negative binomial regression analyses. All models adjusted for treatment, mode of delivery, age, race/ethnicity of child, infant solid food, and first principal component of maternal diet (high in vegetables and low in processed meats and deep fried food).