Infant Gut Microbiota Development Is Driven by Transition to Family Foods Independent of Maternal Obesity

Abstract

The first years of life are paramount in establishing our endogenous gut microbiota, which is strongly affected by diet and has repeatedly been linked with obesity. However, very few studies have addressed the influence of maternal obesity on infant gut microbiota, which may occur either through vertically transmitted microbes or through the dietary habits of the family. Additionally, very little is known about the effect of diet during the complementary feeding period, which is potentially important for gut microbiota development. Here, the gut microbiotas of two different cohorts of infants, born either of a random sample of healthy mothers (n = 114), or of obese mothers (n = 113), were profiled by 16S rRNA amplicon sequencing. Gut microbiota data were compared to breastfeeding patterns and detailed individual dietary recordings to assess effects of the complementary diet. We found that maternal obesity did not influence microbial diversity or specific taxon abundances during the complementary feeding period. Across cohorts, breastfeeding duration and composition of the complementary diet were found to be the major determinants of gut microbiota development. In both cohorts, gut microbial composition and alpha diversity were thus strongly affected by introduction of family foods with high protein and fiber contents. Specifically, intake of meats, cheeses, and Danish rye bread, rich in protein and fiber, were associated with increased alpha diversity. Our results reveal that the transition from early infant feeding to family foods is a major determinant for gut microbiota development. IMPORTANCE The potential influence of maternal obesity on infant gut microbiota may occur either through vertically transmitted microbes or through the dietary habits of the family. Recent studies have suggested that the heritability of obesity may partly be caused by the transmission of "obesogenic" gut microbes. However, the findings presented here suggest that maternal obesity per se does not affect the overall composition of the gut microbiota and its development after introduction of complementary foods. Rather, progression in complementary feeding is found to be the major determinant for gut microbiota establishment. Expanding our understanding of the influence of complementary diet on the development and establishment of the gut microbiota will provide us with the knowledge to tailor a beneficial progression of our intestinal microbial community.

Keywords: 16S rRNA sequencing; breastfeeding; complementary diet; family foods; infant gut microbiota; maternal obesity.

FIG 1

Gut microbial beta and alpha diversity is independent of maternal obesity but changes over time. (A) PCoA plot based on Bray-Curtis dissimilarity, with the centroid for each group shown with a black boarder. The distance to the group centroid for each point provides a measure of homogeneity of variance, used to estimate beta diversity. PC1 and PC2, principal coordinates 1 and 2, respectively. (B) Alpha diversity measures as estimated by the Shannon index, observed genera, and Pielou’s evenness index. Boxes indicate 25th to 75th percentiles, with mean values marked as a line and whiskers indicating minimum and maximum values. ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001 (according to Tukey’s honestly significant difference test for beta diversity and paired [within cohorts, across time points] or unpaired [across cohorts at the same time point] t tests for alpha diversity measures).

FIG 2

Composition of gut microbiota in SKOT I and SKOT II. Relative abundances of bacterial phyla (small panels) and families (large panels) in the SKOT I cohort (n = 114) at the ages of 9 months (A) and 18 months (B) and in the SKOT II cohort (n = 113) at the ages of 9 months (C) and 18 months (D). Boxes indicate 25th to 75th percentiles, with mean relative abundances marked as lines and whiskers indicating the range (minimum/maximum) multiplied by the interquartile range (25th to 75th percentiles) from the boxes. Bacterial families are ranked by average relative abundances at the age of 9 months. Detailed information can be found in Table S1 in the supplemental material.

FIG 3

Gut microbiota composition is independent of maternal obesity but changes over time. (A) PCA biplot of the relative abundances of bacterial families at 9 and 18 months of age in SKOT I and SKOT II. Ellipses indicate 95% confidence intervals for each group, while arrows show loadings. var., variance. (B) Log₂-transformed fold changes of relative abundances of bacterial families between the ages of 9 and 18 months within SKOT I and SKOT II. Error bars indicate the standard error of the mean. ns, not significant; *, q < 0.05; **, q < 0.01; ***, q < 0.001 (according to false-discovery-rate-corrected [5%] paired Wilcoxon signed-rank tests of relative abundances at 9 months versus 18 months). No significant differences were found between the fold changes of bacterial families occurring in the two cohorts after we performed false-discovery-rate-corrected (5%) Mann-Whitney tests.

FIG 4

Duration of exclusive breastfeeding is reflected in late-infancy gut microbiota. Hierarchical clustering of Spearman’s rank correlations of duration of exclusive breastfeeding with gut microbial composition at 9 months of age at the family (A) and genus (B) levels in SKOT I and II. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (C) Spearman’s rank correlations of alpha diversity measures (Shannon index, observed genera, and Pielou’s evenness index) to duration of exclusive breastfeeding (0 to 6 months) and age at introduction of complementary feeding (3 to 6 months) for compiled data from SKOT I and II. Boxes indicate 25th to 75th percentiles, with mean values marked as a line and whiskers indicating minimum and maximum values.

FIG 5

Complementary diet affects late-infancy gut microbiota composition. (A) Macronutrient intake in SKOT I and II is expressed as a mean percentage of total energy intake (E%) from carbohydrates, protein, and fat, whereas fiber intake is expressed in milligrams per kilojoule, with error bars indicating standard deviations. (B and C) PCA plot of consumption of the 23 food groups (grams per day per kilogram of body weight) for each individual in SKOT I and II (n = 217), resulting in principal component 1 (PC1), designated family food, and PC2, designated health-conscious food. (D) Heatmap illustrating hierarchical clustering of Spearman’s rank correlations between relative abundances of family-level gut microbes and macronutrient intake. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (E and F) Spearman’s rank correlations between family food and relative abundances of Bifidobacteriaceae and Lachnospiraceae. q values are false-discovery-rate-corrected P values.

FIG 6

The transition to family foods with higher protein and fiber content correlates with increased gut microbiota diversity. Pearson correlations of gut microbial alpha diversity (Shannon index) with macronutrient intake (A to D) and the dietary patterns family food and health-conscious food (E and F) at 9 months of age (SKOT I, green; SKOT II, purple).