Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Filters applied. Clear all
Multicenter Study
. 2018 Jul 2;172(7):e181161.
doi: 10.1001/jamapediatrics.2018.1161. Epub 2018 Jul 2.

Association of Exposure to Formula in the Hospital and Subsequent Infant Feeding Practices With Gut Microbiota and Risk of Overweight in the First Year of Life

Affiliations
Free PMC article
Multicenter Study

Association of Exposure to Formula in the Hospital and Subsequent Infant Feeding Practices With Gut Microbiota and Risk of Overweight in the First Year of Life

Jessica D Forbes et al. JAMA Pediatr. .
Free PMC article

Erratum in

  • Error in Figures.
    JAMA Pediatr. 2018 Jul 1;172(7):704. doi: 10.1001/jamapediatrics.2018.2205. JAMA Pediatr. 2018. PMID: 29971408 No abstract available.

Abstract

Importance: The effect of neonatal and infant feeding practices on childhood obesity is unclear. The gut microbiome is strongly influenced by feeding practices and has been linked to obesity.

Objective: To characterize the association between breastfeeding, microbiota, and risk of overweight during infancy, accounting for the type and timing of supplementary feeding.

Design, setting, and participants: In this study of a subset of 1087 infants from the prospective CHILD pregnancy cohort, mothers were recruited between January 1, 2009, and December 31, 2012. Statistical analysis was performed from February 1 to December 20, 2017.

Main outcomes and measures: Feeding was reported by mothers and documented from hospital records. Fecal microbiota at 3 to 4 months (from 996 infants) and/or 12 months (from 821 infants) were characterized by 16S ribosomal RNA sequencing. Infants with a weight for length exceeding the 85th percentile were considered to be at risk for overweight.

Results: There were 1087 infants in the study (507 girls and 580 boys); at 3 months, 579 of 1077 (53.8%) were exclusively breastfed according to maternal report. Infants who were exclusively formula fed at 3 months had an increased risk of overweight in covariate-adjusted models (53 of 159 [33.3%] vs 74 of 386 [19.2%]; adjusted odds ratio, 2.04; 95% CI, 1.25-3.32). This association was attenuated (adjusted odds ratio, 1.33; 95% CI, 0.79-2.24) after further adjustment for microbiota features characteristic of formula feeding at 3 to 4 months, including higher overall richness and enrichment of Lachnospiraceae. A total of 179 of 579 infants who were exclusively breastfed (30.9%) received formula as neonates; this brief supplementation was associated with lower relative abundance of Bifidobacteriaceae and higher relative abundance of Enterobacteriaceae at 3 to 4 months but did not influence the risk of overweight. At 12 months, microbiota profiles differed significantly according to feeding practices at 6 months; among partially breastfed infants, formula supplementation was associated with a profile similar to that of nonbreastfed infants (higher diversity and enrichment of Bacteroidaceae), whereas the introduction of complementary foods without formula was associated with a profile more similar to that of exclusively breastfed infants (lower diversity and enrichment of Bifidobacteriaceae and Veillonellaceae). Microbiota profiles at 3 months were more strongly associated with risk of overweight than were microbiota profiles at 12 months.

Conclusions and relevance: Breastfeeding may be protective against overweight, and gut microbiota may contribute to this effect. Formula feeding appears to stimulate changes in microbiota that are associated with overweight, whereas other complementary foods do not. Subtle microbiota differences emerge after brief exposure to formula in the hospital. These results identify important areas for future research and distinguish early infancy as a critical period when transient gut dysbiosis may lead to increased risk of overweight.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Azad reported holding a Canada Research Chair in the Developmental Origins of Chronic Disease. Dr Sears reported holding the AstraZeneca endowed chair in respiratory epidemiology. No other disclosures were reported.

Figures

Figure 1.
Figure 1.. Infant Gut Microbiota at 3 to 4 Months According to Breastfeeding (BF) Status
A, Alpha diversity evaluated by richness (abundance-based coverage estimator [ACE] and Chao1) and diversity (Simpson and Shannon). Median estimates are compared across feeding groups using the Kruskal-Wallis test (nonparametric analysis of variance) and Dunn post hoc tests for multiple comparisons. Boxes indicate interquartile range, lines indicate medians, diamonds indicate means, and whiskers represent range. B, Mean phylum-level composition. C, Relative abundance of dominant taxa across feeding groups. Breastfeeding (BF) status is assessed at the time of sample collection. Breastfeeding refers to breast milk feeding regardless of feeding mode (at the breast or from a bottle). D, Principal coordinate analysis (PC1 and PC2) based on unweighted UniFrac distances, with community structure differences tested by permutational analysis of variance with 999 permuations. P values represent false discovery rate–corrected P values testing for overall differences across the 4 feeding groups. Significant pairwise comparisons: aNo BF/partial BF; bNo BF/exclusive BF after hospital; cNo BF/exclusive BF; dPartial BF/exclusive BF after hospital; ePartial BF/exclusive BF; fExclusive BF after hospital/exclusive BF. gP < .001. hP < .05.
Figure 2.
Figure 2.. Infant Gut Microbiota at 12 Months According to Diet at 6 Months
A, Alpha diversity evaluated by richness (abundance-based coverage estimator [ACE] and Chao1) and diversity (Simpson and Shannon). Median estimates are compared across feeding groups using the Kruskal-Wallis test and Dunn post hoc tests for multiple comparisons. Boxes indicate interquartile range, lines indicate medians, diamonds indicate means, and whiskers represent range. B, Mean phylum-level composition. C, Relative abundance of dominant taxa across feeding groups. Breastfeeding (BF) refers to breast milk feeding regardless of feeding mode (at the breast or from a bottle). D, Principal coordinate analysis (PC1 and PC2) based on unweighted UniFrac distances, with community structure differences tested by permutational analysis of variance with 999 permuations. P values represent false discovery rate–corrected P values testing for overall differences across the 4 feeding groups. Significant pairwise comparisons: aNo BF/exclusive BF; bPartial BF with formula/partial BF without formula; cPartial BF with formula/exclusive BF; no significant differences observed between partial BF without formula and exclusive BF; dNo BF/partial BF without formula; eNo BF/partial BF with formula. fP < .01. gP < .05.
Figure 3.
Figure 3.. Infant Gut Microbiota Characterization at 3 Months According to Infant Weight Status at 12 Months
A, Alpha diversity evaluated by richness (abundance-based coverage estimator [ACE] and Chao1) and diversity (Simpson and Shannon). Median estimates are compared across weight status using the Kruskal-Wallis test and Dunn post hoc tests for multiple comparisons. Boxes indicate interquartile range, lines indicate medians, diamonds indicate means, and whiskers represent range. B, Mean phylum-level composition. C, Relative abundance of dominant taxa across weight status groups. Breastfeeding refers to breast milk feeding regardless of feeding mode (at the breast or from a bottle). Significant pairwise comparisons: aNormal/overweight; bNormal/at risk.

Comment in

Similar articles

See all similar articles

Cited by 21 articles

See all "Cited by" articles

Publication types

Grant support

Feedback