Prenatal vitamin C and fish oil supplement use are associated with human milk microbiota composition in the Canadian CHILD Cohort Study

Abstract

Maternal diet may modulate human milk microbiota, but the effects of nutritional supplements are unknown. We examined the associations of prenatal diet and supplement use with milk microbiota composition. Mothers reported prenatal diet intake and supplement use using self-administered food frequency and standardised questionnaires, respectively. The milk microbiota was profiled using 16S rRNA gene sequencing. Associations of prenatal diet quality, dietary patterns, and supplement use with milk microbiota diversity and taxonomic structure were examined using Wilcoxon signed-rank tests and multivariable models adjusting for relevant confounders. A subset of 645 mothers participating in the CHILD Cohort Study (originally known as the Canadian Healthy Infant Longitudinal Development Study) provided one milk sample between 2 and 6 months postpartum and used prenatal multivitamin supplements ≥4 times a week. After adjusting for confounders, vitamin C supplement use was positively associated with milk bacterial Shannon diversity (β = 0.18, 95% CI = 0.05, 0.31) and Veillonella and Granulicatella relative abundance (β = 0.54; 95% CI = 0.05, 1.03 and β = 0.44; 95% CI = 0.04, 0.84, respectively), and negatively associated with Finegoldia relative abundance (β = -0.31; 95% CI = -0.63, -0.01). Fish oil supplement use was positively associated with Streptococcus relative abundance (β = 0.26; 95% CI = 0.03, 0.50). Prenatal diet quality and dietary patterns were not associated with milk microbiota composition. Prenatal vitamin C and fish oil supplement use were associated with differences in the milk microbiota composition. Future studies are needed to confirm our findings and elucidate mechanisms linking maternal supplement use to milk microbiota and child health.

Keywords: Breastmilk; CHILD Cohort Study; Diet; Microbiome; Pregnancy; Supplements.

Fig. 1.

Flowchart for selecting mother-infant dyads in the CHILD Cohort Study included in the current analysis.

Fig. 2.

Human milk bacterial Shannon diversity by prenatal diet quality, dietary pattern, and supplement use among mothers in the CHILD Cohort Study. (a) Unadjusted associations examined using Wilcoxon-sign rank test. (b) Adjusted associations examined using multivariable linear regression permutation models adjusted for parity, pre-pregnancy BMI, breastfeeding exclusivity, mode of breast milk feeding at the time of milk sample collection, and fish oil and vitamin C supplement use; models were additionally adjusted for batch of analysis. Diet quality was examined using HEI-2010 scores categorised as low (<50^th percentile) vs. high (≥50^th percentile). Dietary patterns derived using PCA were examined as adherence (positive PCA scores) vs. avoidance (negative PCA scores). PCA scores reflected how closely the mother’s dietary intake was similar to the components of the dietary pattern, which were as follows: plant-based (dairy, legumes, vegetables, whole grains, and an aversion to meats), Western (fats, meats, processed foods, and starchy vegetables) and balanced (diverse sources of animal proteins (especially fish), vegetables, fruits, nuts and seeds). Supplement ever use was defined as use at least once a month during pregnancy. BMI, body mass index; CI, confidence interval; HEI, healthy eating index; PCA, principal component analysis. NS, not significant; *: P-value < 0.05; **: P-value < 0.01.

Fig. 3.

Human milk microbiota β diversity by prenatal diet quality, dietary patterns, and supplement use among mothers in the CHILD Cohort Study. β diversity assessed on Bray–Curtis dissimilarity matrix using permutational analysis of variance (PERMANOVA). Diet quality was examined using HEI-2010 scores categorised as low (<50^th percentile) vs. high (≥50^th percentile). Dietary patterns derived using PCA were examined as adherence (positive PCA scores) vs. avoidance (negative PCA scores). PCA scores reflected how closely the mother’s dietary intake was similar to the components of the dietary pattern, which were as follows: plant-based (dairy, legumes, vegetables, whole grains, and an aversion to meats), Western (fats, meats, processed foods, and starchy vegetables) and balanced (diverse sources of animal proteins (especially fish), vegetables, fruits, nuts and seeds). Supplement ever use was defined as use at least once a month during pregnancy. PCA, principal component analysis.

Fig. 4.

Genus relative abundance in the human milk of mothers in the CHILD Cohort Study by prenatal supplement use. Associations were examined using multivariable linear regression permutation models adjusted for parity, pre-pregnancy BMI, breastfeeding exclusivity, mode of breast milk feeding at the time of milk sample collection, and fish oil and vitamin C supplement use; models were additionally adjusted for batch of analysis. Supplement ever use was defined as use at least once a month during pregnancy. Genera relative abundance was centred log-ratio transformed. BMI, body mass index. *: Confidence interval not including 0.