The Role of Maternal Secretor Status and Human Milk Oligosaccharides on Early Childhood Development: A Systematic Review and Meta-Analysis

Abstract

Background:

Breast milk is the gold standard of infant nutrition, delivering nutrients and bioactive molecules as needed to support optimal infant growth and cognitive development. Increasing evidence links human milk oligosaccharides (HMOs) to these early childhood development milestones.

Aims:

To summarize and synthesize the evidence relating to HMOs and infant brain development, physical growth, and cognitive development. In addition, HMO concentrations in secretor and nonsecretor mothers were compared via a meta-analysis.

Study Design:

A systematic review and meta-analysis were carried out in accordance with the PRISMA statement. This review used three databases (PubMed, Scopus, and Web of Science) and was limited to English-language articles published between 2000 and June 30, 2023.

Results:

The initial searches yielded 245 articles, 27 of which were included in the systematic review and 12 in the meta-analysis. The meta-analysis revealed a substantial between-study heterogeneity, I² = 97.3%. The pooled effect was 0.21 (95% CI: −0.41 to 0.83; p = 0.484), indicating that secretors had higher HMO concentrations, although this difference was not statistically significant. At one month of age, 2′FL, 3FL, and 3′SL play an important role in brain maturation and thus play a critical role in cognitive development. Secretors produce higher concentrations of 2′FL and 3′SL, explaining the benefits to infants of secretor mothers. Growth velocity was correlated to fucosylated and sialylated HMO concentrations, with lower concentrations linked to stunting.

Conclusions:

According to evidence from the systematically reviewed articles, HMOs are essential for a child’s early development, but the extent to which they have an impact depends on maternal secretor status.

Introduction

Breastfeeding is the gold standard of infant nutrition and has undisputed benefits for infants, including protection against childhood infections, increased intelligence, and a possible reduction in obesity and diabetes later in life.¹ It is also known that breastfed infants of secretor mothers benefit more than those breastfed by nonsecretor mothers. This is due, in part, to Bifidobacterium spp. depletion caused by the absence of 2′FL in human milk of nonsecretors, as well as colonization by Clostridium spp. and Enterobacteriaceae in the absence of Bifidobacterium spp.² Suboptimal breastfeeding, defined as lack of exclusive and continued breastfeeding, can disrupt physical growth and neurodevelopment, potentially leading to lifelong consequences.³ The benefit of human milk is the result of millions of years of evolution, which resulted in a perfect multifunctional fluid composed of macronutrients, micronutrients, bioactive molecules, and microbiota.⁴ Bioactive molecules are non-nutritive factors that influence biological processes.⁵ One example of such factors is human milk oligosaccharides (HMOs), nondigestible glycans that have been demonstrated to be important in promoting microbiota growth.⁶

Human milk oligosaccharides are now known to be composed of the following monosaccharides: galactose, glucose, fucose, N-acetylglucosamine, and the sialic acid derivative N-acetyl-neuraminic acid, with 12 possible linkages.⁷ They are synthesized by lactocytes or imported into milk from the blood supply of the breasts based on infant needs, with colostrum having the highest concentration (>20 g/L) and mature milk having a lower concentration (5–14 g/L).^8,9 It is important to note that HMO composition varies between mothers and is influenced by secretor status, geographic region, season, maternal diet and BMI, gestational age, and duration of lactation.² The genetic variation that causes changes in HMO patterns and concentrations has been linked to chromosome 19, with further mutations discovered on chromosomes 4, 8, 10, and 11.¹⁰ Maternal genetics is important because secretor mothers express the FUT2 gene and thus produce milk with an abundance of α1–2 linked fucose moieties such as 2′FL and LNFP I, whereas nonsecretor mothers produce little or no such HMOs.¹¹ Mothers who are Lewis-positive express the FUT3 gene and produce milk containing oligosaccharides with α1–3 and α1–4 linkages to fucose such as 3FL and LNFP II.^12,13

Differences in HMO production based on secretor status have been linked to stimulation of body growth and brain development in the first 6 months of life.^14–16 Specific HMOs have been linked to infant growth and anthropometry in both high- and low-income populations.^17,18 Another aspect of child development is brain development, which begins a few weeks after conception and continues until the early postnatal years. HMOs are thought to promote early brain development by acting as prebiotics, which are metabolized by gut microbiota and transported from the gut to the brain through the vagus nerve, where they affect brain functions.^19,20 These brain development functions have been linked to sialic acid, which is abundant in HMOs. Notably, sialic acid is an essential nutrient for brain development because it crosses the blood–brain barrier and reaches the frontal cortex gray matter, where it is involved in ganglioside formation and myelination, both of which are key components of developing cortical gray matter and white matter.²¹ As a result, it is vital to identify HMOs that are important for newborn growth and prioritize synthetic HMOs that can be added in formulas, while keeping in mind that some HMOs may increase the risk of excessive weight gain and obesity in childhood, which may persist into adulthood. As a result, it is critical to identify HMOs that are important in infant growth and cognitive development, given that technology exists to synthesize HMOs, which can be used as a supplement on infants who receive inadequate breast milk.

A review of published systematic reviews up to June 2023 yielded 5 articles relevant to this topic. Two studies explored the effect of breastfeeding on infant body composition,^22,23 while three articles focused on the role of HMOs on neurodevelopment and cognitive functions in children.^24–26 All of these systematic reviews address the effect of human milk components on either infant growth or neurodevelopment, but none has systematically evaluated and synthesized evidence on the effect of maternal secretor status on HMOs on both infant growth and cognitive development, which are two intertwined concepts of early human development. The primary goal of this systematic review is to summarize evidence on the influence of maternal secretor status on HMOs and on brain development, infant growth velocity, and cognitive development using Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. The secondary goal was to perform meta-analyses on HMO concentration differences between secretors and nonsecretors, as well as to determine between-study heterogeneity.

Methods

The PRISMA guidelines were followed when conducting this systematic review and meta-analysis.²⁷ An ethical review was not necessary because this was a review of studies that had already been published.

Data sources and search strategy

The literature search was conducted by HK. Articles were published in English from three databases (PubMed, Scopus, and Web of Science) using identified keywords and index terms. The search terms were divided into three concepts, combined with the Boolean phrases “AND” between concepts and “OR” within concepts: (1) breastfeeding AND (2) human milk oligosaccharides OR human milk sugars AND (3) infant growth OR brain development OR neurodevelopment OR cognitive development. The search was limited to articles published between 2000 and June 30, 2023, based on an extensive review published by Kunz and colleagues focusing on the structural, functional, and metabolic aspects of HMOs.⁸ We conducted the search and limited the selection to original articles published in English, excluding gray literature, using the syntax shown in Supplementary Table S1. The search terms used on PubMed were modified for Scopus and Web of Science database searches.

Eligibility criteria

We developed a Population, Intervention, Comparators, Outcome, and study design approach as the eligibility criteria, determined by the following research question: What effect do HMOs in human milk [Intervention] have on brain development, infant growth velocity, and cognitive development [outcomes] in breastfed infants [population]? Population: Studies that included nursing mothers and their infants were eligible. Intervention: exposure to HMOs via breastfeeding. Comparators: Secretor versus nonsecretor mothers. Outcome: brain development, infant growth velocity, and cognitive development. Study design: Observational and intervention studies. Exclusion: Inappropriate study types (animal studies, in vitro studies, abstracts from conferences, reviews, and meta-analyses), as well as articles, that did not link HMOs to brain development, infant growth rate, or cognitive development. Two authors (MM and HA) conducted the literature selection for eligibility by removing all undoubtedly irrelevant articles before using Rayyan (https://www.rayyan.ai/), a web tool that speeds up the screening process by multiple authors concurrently.²⁸

Data extraction

The data extracted from each article included the following: author, year of publication, country, study design, population size (n), milk samples (n), HMO profiled (n), HMO profiling technique, proportion of secretors (%) and nonsecretor mothers (%), and effect of HMOs on brain development, growth velocity, and cognitive development. M.M. and H.A. independently extracted data into Microsoft excel template. When there were disagreements, a consensus was reached after discussing with L.O.

Quality assessment

The study’s quality was independently assessed by M.M., H.A., and H.K. A majority of the studies were observational, and therefore, Newcastle–Ottawa Scale (NOS) was used to assess the risk of bias.²⁹ The NOS is a three-domain scale that assesses participant selection, comparability, and outcome. Each study may receive up to four points in the selection section (if the study is truly representative of the community and prospective) and three points in the outcome section (if outcome measure data are collected as accurately as possible, and the cohort follow-up rate is >80%). Depending on the adjustment for confounders, the comparability section can receive up to two points. As a result, studies can be scored on a scale of 0–9. A high-quality study will be awarded 8–9 points, a medium-quality study will earn 6–7 points, and a low-quality study will receive 6 points. Cross-sectional studies receive up to 6 stars, whereas longitudinal studies receive up to 9 stars.

Data analysis

The primary goal of all of the studies was to determine the effect of breast milk HMOs on brain development, infant growth velocity, or cognitive development of breastfed infants. If means and standard deviations (SD) for HMO concentrations were not reported, they were calculated from the median and (IQR) using an online (https://www.math.hkbu.edu.hk/∼tongt/papers/median2mean.html). A weighted mean was computed using the Hmisc R package to determine the proportion of secretor versus nonsecretor mothers. To compare concentrations between secretor and nonsecretor mothers, meta-analyses were performed using the meta R package. The majority of studies were cross-sectional, but if longitudinal, the earliest timepoint was used. The standardized mean difference (SMD) was used as the outcome measure to assess statistical heterogeneity in the studies that were included in the meta-analysis. Since we expected significant between-study heterogeneity, we used a random-effects model to pool effect sizes. The heterogeneity variance τ² was estimated using the restricted maximum likelihood (REML) estimator. Because percentages are easier to report, the I² statistic is reported in addition to the τ² statistic. The confidence interval around the pooled effect was computed using Knapp–Hartung adjustments. Studentized residuals and Cook’s distances were used to determine if studies were outliers or influential in the context of the model. Publication bias was assessed through the Egger weighted regression and visualized using funnel plots.

Results

Eligible studies

A PRISMA flowchart of the study selection process is shown in (Fig. 1). H.K. conducted an initial search in the three databases, and the results were 245 articles. After the duplicates were removed, 176 articles were left. After evaluating titles and abstracts for study type, or irrelevance, a further 61 articles were eliminated. M.M. and H.K. downloaded the full texts of the remaining 69 articles and reviewed them for eligibility. At this point, 42 articles were removed from consideration because they did not meet the inclusion criteria, leaving only 27 articles in the systematic review.^{17,18,30–54} In addition, 12 studies that compared HMO concentrations in secretors and nonsecretor mothers were included in the meta-analysis.^{30,31,33–36,38–43,48}

FIG. 1.

PRISMA diagram of study selection. The relevant number of articles at each step are indicated.

Quality assessment

According to the NOS findings (Supplementary Table S2), 17 studies (71%) had medium quality, which was attributed to higher scores in comparability and outcome aspects. The main reason for the lower quality studies was a lack of detail about potential baseline differences between groups or confounding variables. The aspect of outcome (HMO profile) within the NOS was adequately conducted in most studies in accordance with the study’s research question.

Study characteristics

Table 1 summarizes the sociodemographic and HMO data from the 27 studies that were included in this review. The selected studies were conducted in 21 countries: Australia,^31,43 Bangladesh,⁵³ Brazil,^36,50 China,^32,41 Denmark,⁵⁴ Finland,³⁰ France,⁴⁸ Gambia,¹⁸ Germany,³⁹ Kenya,⁴² Malawi,^47,51 Netherlands,⁴⁴ Philippines,⁵² Singapore,³³ Spain,⁴⁰ Sweden,⁴⁹ and United States.^{17,34,35,37,38,45} In addition, there was one European Union study involving milk samples collected from France, Italy, Norway, Portugal, Romania, Spain, and Sweden.⁴⁶ A total of 7,648 milk samples were examined in the 27 studies, which involved 4,066 mothers and 4,103 infants. The differences between mothers and infants are due to twin pregnancies. The weighted mean of secretor mothers was 77.8% (95% CI: 60.3–95.3).

Table 1.

Summary of Selected Studies (n = 27)

Author	Country	Study design	Population (n)	Sample (n)	Secretor (%)	HMO (n)	Technique	Outcome measure	Key findings
Cheema et al.,2022⁴³	Australia	Cross-sectional	Mothers = 60 Infants = 60	60	81.7	19	HPLC	Growth velocity	Infant body composition had a significant association with HMO intakes and concentrations; however, the associations were influenced by secretor status.
Mansell et al., 2023³¹	Australia	Cross-sectional	Mothers = 292 Infants = 292	292	75.7	19	HPLC	Growth velocity	HMO composition is associated with several anthropometry measures up to 6 months of age in a potential secretor status-specific manner.
Nuzhat et al., 2022⁵³	Bangladesh	Cross-sectional	Mothers = 45 Infants = 45	45	55.6	73	HPLC-qTOF-MS	Growth velocity	A higher relative abundance of sialylated HMOs in mothers’ breast milk adversely affected infant nutritional status. There is no association between severe acute malnutrition and fucosylated or undecorated HMOs.
Tonon et al., 2019³⁶	Brazil	Cross-sectional	Mothers =78 Infants = 78	78	87.2	16	LC-MS	Growth velocity	Differences in HMO composition due to maternal Se/Le phenotype have no effect on infant growth, but higher concentrations of specific HMOs may protect against excessive weight gain.
Ferreira et al., 2021⁵⁰	Brazil	Longitudinal	Mothers =74 Infants =74	173	91.9	19	HPLC	Cognitive development	Higher concentrations of LNT HMOs in Brazilian women are associated with their infants being less likely to be at risk of inadequate development for personal-social skills or for ≥2 developmental domains during the first year of life.
Wang et al., 2020⁴¹	China	Longitudinal	Mothers = 116 Infants =116	464	81.0	23	HPLC-ESI-MS/MS	Growth velocity	There was no significant difference in infant weight and length gain between the Se+/Se-groups.
Dou et al., 2023³²	China	Longitudinal	Mothers = 22 Infants = 22	66	68.2	14	IC-LC	Growth velocity	LNnT was significantly higher in GDM mothers in all time points, and its concentrations in colostrum and transitional milk were correlated positively with the infant’s weight-for-age Z-score at 6 months postnatal in the GDM group.
Larsson et al., 2019⁵⁴	Denmark	Longitudinal	Mothers = 30 Infants = 30	60	76.7	19	HPLC	Growth velocity	2’FL, an additive to infant formula, was associated with excessive weight gain. Total HMO concentrations, as well as total HMO-bound fucose at 5 months, were positively associated with both fat mass index and weight velocity from 0 to 5 months.
Binia et al., 2021⁴⁶	Europe (7 countries)	Longitudinal	Mothers = 357 Infants = 359	2220	63.0	21	HPLC & NMR	Growth velocity	No significant associations between growth and body composition and Lewis or secretor-dependent HMOs in a cohort of health mother–infant dyads.
Lagström et al., 2020³⁰	Finland	Cross-sectional	Mothers = 802 Infants = 802	802	87.2	19	HPLC	Growth velocity	The concentration of LNnT was inversely associated with that of 2′FL and was directly associated with infant growth velocity.
Rozé et al., 2022⁴⁸	France	Longitudinal	Mothers = 117 Infants = 137	570	80.3	24	HPLC	Growth velocity & Cognitive development	In very preterm infants, total HMOs were positively correlated with neonatal length growth. LNFP-III was significantly associated with total ASQ score among infants born to Secretor(+) Lewis(+) mothers at 2 years.
Davis et al., 2017¹⁸	Gambia	Longitudinal	Mothers = 33 Infants = 33	99	63.6	39	Nano-HPLC-Chip/TOF MS	Growth velocity	Mothers nursing in the wet season (July-October) produced significantly less HMOs compared to those nursing in the dry season (November-June). 3′SL was found to be a good indicator of infant growth.
Menzel et al., 2021³⁹	Germany	Cross-sectional	Mothers = 143 Infants = 143	143	85.3	9	HPLC	Growth velocity	The growth velocity of nonsecretors’ children was inversely associated with LNnT up to 6 years of age. 2’FL was negatively associated with BMI 3 months and 4 and 5 years. Children of nonsecretors showed higher BMI at 3 months, 6 months, and 3, 6, and 7 years of age.
Paganini et al., 2019⁴²	Kenya	Clinical trial	Mothers = 75 Infants = 75	75	72.0	13	HPAEC-PAD	Growth velocity	There were no significant differences in anthropometrics between secretor mothers’ infants and nonsecretor mothers’ infants.
Charbonneau et al., 2016⁴⁷	Malawi	Cross-sectional	Mothers = 303 Infants = 303	303	73.9	50	Nano-HPLC-Chip/TOF MS	Growth velocity	Mothers with undernourished infants have decreased HMOs. Nonsecretor mothers are unable to compensate for deficiencies in fucosylated HMOs through increased output of other (e.g., sialylated) HMOs, resulting in milk that is less supportive of infant growth.
Jorgensen et al., 2021⁵¹	Malawi	Clinical trial	Mothers = 647 Infants = 647	659	75.0	51	Nano-LC chip-TOF MS	Growth velocity & Cognitive development	There is an association between the relative abundance of fucosylated and sialylated HMOs and language at 18 months. In addition, the absolute abundance of HMOs was associated with infant growth among secretors.
Willemsen et al., 2023⁴⁴	Netherlands	Longitudinal	Mothers = 63 Infants = 63	189	n.s	24	PGC-UPLC-MS /HPAEC-PAD	Cognitive development	Concentrations of 2′FL and grouped fucosylated HMOs were associated with better executive functions. In contrast, grouped sialylated HMOs were associated with worse executive functions at three years of age.
Samuel et al., 2022⁵²	Philippines	Longitudinal	Mothers = 41 Infants = 41	205	73.2	24	HPLC	Growth velocity	Specific HMOs, partly as proxy for milk groups (defined by secretor and Lewis status), may be associated with infant growth. There was no association between HMOs and infant appetitive traits.
Sprenger et al., 2017³³	Singapore	Longitudinal	Mothers = 50 Infants = 50	147	68.0	5	HPAEC	Growth velocity	HMO variation between the high and low 2′FL clusters do not impact infant growth of either sex up to 4 months of age.
Oliveros et al., 2021⁴⁰	Spain	Clinical trial	Mothers = 82 Infants = 82	82	57.3	2	UPLC-MS/MS	Cognitive development	2′FL and 6′SL levels in human milk were associated with infant cognitive development.
Wejryd et al., 2018⁴⁹	Sweden	Clinical trial	Mothers = 91 Infants = 106	200	94.5	15	HPAEC-PAD	Growth velocity	There is an association between 3′SL and 6′SL and culture-proven sepsis in very low birth weight (<1000 g) infants who were exclusively breastfed. LST was correlated positively with weight, length, and head growth at 14 days of life, whereas 6′SL, LST b, and LST c correlated negatively with weight and head growth at 14 and 28 days of life.
Alderete et al., 2015¹⁷	USA	Longitudinal	Mothers = 25 Infants = 25	50	72.0	16	HPLC	Growth velocity	Maternal HMO composition is associated with infant growth and body composition.
Berger et al., 2020a³⁴	USA	Longitudinal	Mothers = 50 Infants = 50	100	86.0	19	HPLC	Cognitive development	Early exposure to 2′FL (1 month) contributed infant cognitive development, but not at 6 months. This suggests that early exposure to 2′FL may be a critical temporal window for positively influencing infant cognitive development.
Berger et al., 2020 b³⁵	USA	Cross-sectional	Mothers = 157 Infants = 157	157	86.0	19	HPLC	Growth velocity	Higher LNFPII in human milk may decrease obesity risk across all infants, whereas higher LNnT and DSLNT may increase obesity risk in infants of nonsecretor mothers.
Saben et al., 2021³⁸	USA	Cross-sectional	Mothers = 194 Infants = 194	194	73.2	19	HPLC	Growth velocity	Infant intakes of 3FL, 3′SL, DSLNT, and total acidic HMOs were positively associated with infant growth over the first 6 months of life. Maternal obesity is associated with changes in HMO concentrations that are associated with infant adiposity.
Cho et al., 2021⁴⁵	USA	Longitudinal	Mothers = 99 Infants = 99	183	n.s	8	HPLC	Cognitive development	There is an association between 3’SL and cognitive development, particularly language functions, in children who received human milk containing detectable A-tetra during infancy.
Berger et al., 2022³⁷	USA	Cross-sectional	Mothers = 20 Infants = 20	20	85.0	19	HPLC	Brain development	Fucosylated and sialylated HMOs differentially associate with indices of tissue microstructure and rCBF, suggesting specific roles for 2′FL, 3′FL, and 3’SL in early brain maturation.

The bold numbers under population (n) depict twin pregnancies where the number of infants is greater than the number of mothers. The studies are organized by country.

ASQ, age and stages questionnaire; GDM, gestational diabetes mellitus; HPLC, high performance liquid chromatography; HPLC-ESI-MS, high performance liquid chromatography coupled with negative mode electrospray ionization tandem mass spectrometry; HPAEC-PAD, high-performance anion exchange chromatography with pulsed amperometric detection; IC-LC, ion chromatograph liquid chromatography; LC-MS, liquid chromatography-mass; n.s, not stated; NMR, nuclear magnetic resonance; Nano-HPLC-Chip/TOF MS, nano-high performance liquid chromatography-chip/time-of-flight (TOF) mass spectrometry; PGC-UPLC-MS, porous graphitized carbon-ultra high-performance liquid chromatography–mass spectrometry; rCBF, regional cerebral blood flow; SAM, severe acute malnutrition.

Pooled effect and heterogeneity of the reported results

Twelve articles comparing HMO concentrations in secretor and nonsecretor mothers were included in the meta-analysis. Overall, the pooled Hedges’ g (bias corrected standardized mean difference) was 0.21 (95% CI: −0.41 to 0.83), with the majority of estimates being positive (Fig. 2). In the context of our meta-analysis, this indicates a small effect, with secretors having higher HMO concentrations than nonsecretors, but the difference between the two groups is not statistically significant (p = 0.48). Moreover, the 95% prediction interval ranged from g = −2.47 to 2.89, indicating that, although the average outcome is estimated to be positive, in some studies the true outcome may in fact be negative. The REML method estimated a substantial between-study heterogeneity variance (I² value of 97.3%) (95% CI: 96.6–97.9; p < 0.001). Looking at τ² = 1.54 (95% CI: 0.85–3.62; p < 0.001), since the confidence interval does not include zero, it indicates that the variance of true effect sizes is significantly greater than zero (Fig. 2). According to Higgins and Thompson’s I² statistic, 15/19 HMOs profiled were generated by studies with moderate to substantial heterogeneity, whereas the remaining 4/19 HMOs were generated by studies which had low heterogeneity. The τ² statistic included a zero in the remaining HMOs, indicating homogeneity (Table 2). The g of 2′FL produced exclusively by secretors was 2.8 (95% CI: 2.40–3.20; p < 0.001). Other key HMOs with significantly higher bias corrected SMDs in secretors were LNFP-1, 1.45 (95% CI: 1.08–1.82; p < 0.001), and DFlac, 2.78 (95% CI: 0.63–4.97; p = 0.024). Nonsecretor mothers had significantly higher concentrations of the following HMOs compared to secretor mothers: LNFP II, −1.97 (95% CI: –2.92 to –1.02; p < 0.001) and FDSLNH, −2.07 (95% CI: −2.58 to −1.57; p < 0.001). The concentrations of three key HMOs involved in early childhood development (3-sialyllactose [3’SL], 3FL, and 6-sialyllactose [6’SL]) are highlighted although the corrected SMDs did not reach statistical significance (Table 2). The concentration of 3′SL was higher in secretor mothers, but the averaged bias corrected SMD did not differ significantly from zero, 0.21 (95% CI: −0.17 to 0.59; p = 0.23). The corrected SMD of 3FL was higher in secretors, 0.44 (95% CI: −2.57, 3.45) p = 0.746, while the corrected SMD for 6’SL was higher in nonsecretors 0.34 (95% CI: −0.96, 0.26) p = 0.231. An examination of the studentized residuals revealed that one study³⁸ was a significant outlier that influenced the pooled statistic. Although the mean was not particularly extreme, the narrow standard deviations influenced its weight on the overall pooled effect. When it was removed from the analysis of studies profiling 2′FL and 3′SL, the value of τ² decreased from 13.8 to 0.3 for 2′FL and from 10.6 to 0.19 for 3′SL. The pooled effect of g = 2.8 was slightly lower than our initial estimate of g = 3.8 for 2′FL and 0.21 from 1.21 for 3′SL, but it is still within the same order of magnitude (Supplementary Table S3). There was no evidence of outlier HMOs in the final fitted model. Egger’s test did not indicate the presence of funnel plot asymmetry: intercept 2.29 (95% CI: −3.49 to 8.07; t = 0.78, p = 0.44), which is consistent with funnel plot (Fig. 3).

FIG. 2.

Forest plot of the standardized mean difference (g) of HMO concentration between secretors and nonsecretors. A positive effect (g) indicates a higher concentration in secretors, whereas a negative effect (−g) indicates higher HMO concentration in nonsecretors. Note that the Confidence Intervals in this figure differ from those in Table 2 due to Hartung–Knapp (HK) adjustment for random effects model (df = 18). Prediction interval is based on t-distribution (df = 17). g = Hedges’ g (bias corrected standardized mean difference). HMO, human milk oligosaccharides; SMD, standardized mean difference; SE, standard error; 2’FL, 2-fucosyllactose; 3FL, 3-fucosyllactose; DFLac, difucosyllactose; DFLNH, difucosyllacto-N-hexaose; DFLNT, difucosyllacto-N-tetraose; FLNH, fucosyllacto-N-hexaose; LNFP I, lacto-N-fucopentaose I; LNFP II, lacto-N-fucopentaose II; LNFP III, lacto-N-fucopentaose III; 3′SL, 3-sialyllactose; 6′SL, 6-sialyllactose; DSLNT, disialyllacto-N-tetraose; DSLNH, disialyllacto-N-hexaose; LST b, sialyllacto-N-tetraose b; LST c, sialyllacto-N-tetraose c; FDSLNH, fucodisialyllacto-N-hexaose; LNH, lacto-N-hexaose; LNnT, lacto-N-neotetraose; LNT, lacto-N-tetraose.

FIG. 3.

Contour enhanced funnel plot for 19 HMO concentrations between secretors and nonsecretors. It shows a scatter plot of the studies’ observed effect sizes on the x-axis against a measure of their standard error on the y-axis. The higher values on the y-axis represent HMOs with the lower standard errors. In this diagram there is no indication of asymmetry. Nine HMOs are found in regions of high significance (p < 0.01) in [gray zone], three in regions of significance (p < 0.05) in [turquoise zone], and six HMOs in regions of no significance (p < 0.1) in [white zone]. HMO, human milk oligosaccharides.

Table 2.

The Heterogeneity and Estimated Average Standardized Mean Differences of 19 HMOs Computed by Random-Effects Model

Heterogeneity statistics					Individual HMO random effects
HMO	N	I² (95% CI)	τ² (95% CI)	p value	g (95% CI)	p value
2’FL	12	73.0% (51.9%, 84.8%)	0.28 (0.09, 1.07)	<0.001	2.80 (2.40, 3.20)	<.001
3FL	9	98.5% (98.0%, 98.9%)	14.6 (6.53, 57.7)	<0.001	0.44 (−2.57, 3.45)	0.746
LNFP I	8	73.6% (46.2%, 87.0%)	0.15 (0.03, 0.69)	<0.001	1.45 (1.08, 1.82)	<.001
LNFP II	7	94.1% (89.8%, 96.6%)	0.77 (0.26, 4.67)	<0.001	–1.97 (−2.92, −1.02)	0.003
LNFP III	6	94.2% (90.0%, 96.7%)	0.48 (0.15, 3.04)	<0.001	–0.85 (−1.61, −0.09)	0.035
FDSLNH	5	52.7% (0.0%, 82.6%)	0.06 (0.00, 1.88)	0.076	–2.07 (−2.58, −1.57)	<.001
FLNH	5	82.9% (61.1%, 92.5%)	0.19 (0.04, 1.82)	<0.001	0.28 (−0.33, 0.89)	0.269
DFLNH	5	90.1% (79.8%, 95.2%)	0.52 (0.14, 4.92)	<0.001	0.81 (−0.15, 1.77)	0.078
DFLNT	5	94.4% (89.8%, 97.0%)	0.89 (0.28, 7.72)	<0.001	1.59 (0.37, 2.80)	0.022
DFlac	5	92.2% (84.7%, 96.0%)	2.76 (0.85, 27.0)	<0.001	2.79 (0.60, 4.97)	0.024
3’SL	9	81.2% (65.4%, 89.8%)	0.19 (0.05, 0.77)	<0.001	0.21 (−0.17, 0.59)	0.230
6’SL	8	95.7% (93.3%, 97.2%)	0.51 (0.19, 2.07)	<0.001	–0.34 (−0.96, 0.26)	0.231
LST b	6	35.5% (0.0%, 74.2%)	0.03 (0.01, 0.15)	0.171	–0.83 (−1.05, −0.60)	<.001
LST c	6	62.4% (8.5%, 84.5%)	0.06 (0.00, 0.44)	0.021	0.65 (0.33, 0.97)	0.004
LNT	8	90.0% (82.7%, 94.2%)	0.25 (0.08, 1.06)	<0.001	–0.47 (−0.92, −0.02)	0.043
LNnT	8	87.5% (77.5%, 93.0%)	0.32 (0.10, 1.45)	<0.001	0.58 (0.07, 1.09)	0.032
LNH	6	41.9% (0.0%, 77.0%)	0.03 (0.00, 0.48)	0.126	0.05 (−0.24, 0.33)	0.695
DSLNH	5	65.2% (8.8%, 86.7%)	0.09 (0.00, 1.49)	0.021	–0.38 (−0.87, 0.11)	0.099
DSLNT	5	16.1% (0.0%, 82.6%)	0.02 (0.01, 0.18)	0.311	–0.07 (−0.23, 0.16)	0.435

n, Number of studies; I², Higgins & Thompson’s Statistic is the percentage of variability in the effect sizes that is not caused by sampling error, I², 25%: low heterogeneity, I², 50%: moderate heterogeneity I², 75%: substantial heterogeneity; τ², heterogeneity variance (tau) statistic which quantifies the variance of the true effect sizes underlying our data, if it does not include zero then there is heterogeneity in the data, τ² was estimated using the restricted maximum likelihood (REML) estimator. Hedges’ g, bias corrected standardized mean difference. The confidence interval (CI) around the pooled effect was computed using Knapp–Hartung adjustments. The computations were done on meta R package.

Analytical methods for characterizing HMOs

Various analytical techniques were used to identify HMOs, depending on the researchers’ goals. The majority of the studies (14/27) used high-performance liquid chromatography (HPLC). Four studies (4/27) used high-pH anion exchange chromatography (HPAEC) as a detector, whereas seven (7/27) used mass spectrometry (MS). The following 19 HMOs were studied in multiple studies and were included in the meta-analysis: fucosylated HMOs [2-fucosyllactose (2’FL), 3-fucosyllactose (3FL), difucosyllactose (DFLac), difucosyllacto-N-hexaose (DFLNH), difucosyllacto-N-tetraose (DFLNT), fucosyllacto-N-hexaose (FLNH), lacto-N-fucopentaose I (LNFP I), lacto-N-fucopentaose II (LNFP II), lacto-N-fucopentaose III (LNFP III)]; sialylated HMOs [3-sialyllactose (3’SL), 6-sialyllactose (6’SL), disialyllacto-N-tetraose (DSLNT), disialyllacto-N-hexaose (DSLNH), sialyllacto-N-tetraose b (LST b), sialyllacto-N-tetraose c (LST c)]; fucosylated and sialylated HMOs [fucodisialyllacto-N-hexaose (FDSLNH)]; and undecorated HMOs [lacto-N-hexaose (LNH), lacto-N-neotetraose (LNnT), lacto-N-tetraose (LNT)]. Given that different authors used different abbreviations in their published articles, we used Lars Bode’s (2012) abbreviations for consistency in our article.⁶

HMOs involved in brain tissue organization

In one study, a 3.0-Tesla MRI scanner was used to investigate the relationship between HMO concentrations and brain development.³⁷ In this groundbreaking study, Berger and colleagues compared MRI indices of tissue microstructure and regional cerebral blood flow (rCBF) in infants with HMO levels. Most notably in the cortical mantle, 2′FL was connected to decreased fractional anisotropy (FA), increased mean diffusivity (MD), and decreased rCBF values, suggesting that this HMO is crucial for the formation of dendritic arbors and synapses for circuit formation. In the developing white matter, 3FL and 3′SL were linked to higher FA, lower MD, and higher rCBF values, indicating that these HMOs may improve structural connectivity in the brain. While 2′FL, 3FL, and 3′SL were linked to early brain maturation, 6′SL was not associated to these MRI indices.

HMOs involved in cognitive development

This section summarizes the involvement of HMOs in cognitive development in six studies using various techniques such as: Ratings of Everyday Executive Functioning (REEF); Behavior Rating Inventory of Executive Function (BRIEF); Age and Stages Questionnaire (ASQ); Bayler III Cognitive Scale; and Mullen Scales of Early Learning (MSEL). The REEF and BRIEF scores were used to determine executive functions at different time points. 2′FL and total fucosylated HMOs were associated with better executive functions at three years, and 2′FL was associated with 4-behavioral tasks at 2 and 12 weeks, regardless of whether the infants were exclusively breastfed or partially breastfed; however, 3′SL, 6′SL, and total sialylated HMOs were associated with worse executive functions at 3 years.⁴⁴ The ASQ score was used in 2/6 studies to assess the development of communication skills, fine motor skills, gross motor skills, and personal social skills. Positive associations were determined for: 3FL for communication skills; DSLNT, FLNH for fine motor skills; and negative associations for: 2′FL, DFLac, LNH, and 3′SL for personal social skills, LNnT for problem solving, and LNFP III for gross motor skills.⁵⁰ In a study by Roze and colleagues, LNFP III was associated with ASQ score in secretor mothers but not in the entire cohort.⁴⁸ The Bayler III cognitive scale, which measures cognitive development at 24 months using five domains: cognitive, language, motor, adaptive, behavior, and social-emotional development, was used in 2/6 studies. 2′FL, LNH, and FLNH were positively associated with cognitive development, whereas LST b was negatively associated with cognitive development.³⁴ In addition, 2′FL and 6′SL were linked to better cognitive development in infants born to secretor mothers.⁴⁰ The MSEL is a tool for assessing infant cognitive development that is divided into five subdomains as follows: fine motor, gross motor, visual reception, receptive language, and expressive language. In A-tetra+ HMO mothers, there was a positive relationship between 3′SL and early learning composite, receptive language t score.⁴⁵

HMOs involved in infant growth velocity

Various anthropometric indexes were used to report infant growth in 19 studies. Weight gain was reported as weight-for-age z-score in 10/19 studies. The following studies found positive associations between HMOs and weight gain: 2’FL,^30,43,54 3FL,^30,38,43 3’SL,^18,30,38,43 LNnT,³⁵ LNFP II,³⁸ DFLac,^30,43,54 LST b,³⁸ LNH,⁴¹ DFLNH,⁴³ DSLNH,³⁸ DFLNT,⁴³ and negative associations for 3’SL,^31,32 6’SL,⁴³ LNnT,^30,32,54 LNFP II,³⁵ LNH,³¹ LNFP I,¹⁷ LST b,³⁰ LST c,¹⁸ DFLNH,⁵⁴ DSLNT,^32,35 and FDSLNH.³¹

The second metric, height gain, was reported in 12/19 studies as either length-for-age z-score or height-for-age z-score. The following HMOs had a positive correlation with height gain: 2’FL,^30,31,47,48 3FL,⁴³ 3’SL,^39,43 LNnT,³⁹ LST a,^49,51 DFLNH,^41,43,51 and DFLNT,⁴³ whereas a negative correlation was linked with: 3FL,⁵² 3’SL,⁴⁶ 6’SL,⁴⁹ LNnT,^30,39,43,46 LNT,^31,39 LNFP I,^39,47 LNFP II,³² LST b,^30,31,49 and FLNH.³⁹

The third infant growth metric was head growth which was reported in 7/19 studies as head circumference-for-age z-score. This metric which also serves as an index of cognitive development and was positively correlated with the following HMOs: 2’FL,³¹ 3’SL,^51,54 6’SL,⁴⁸ LNFP III,⁴⁶ LST a,⁴⁹ DFLac,⁵⁴ and negatively correlated with 6’SL,^49,51 LNFP III,¹⁸ LNFP I,¹⁸ DFLNH,⁵⁴ LST b,⁴⁹ and DSLNH.³¹

Body–mass index, a gauge of body composition, was reported as body–mass index-for-age z-score in 7/19 studies. The following were reported to have positive associations with this metric: DFLac,⁴³ LST b,⁴³ and DFLNT,⁴³ whereas a negative association was reported for 2’FL,³⁹ 6’SL,⁵⁴ LNnT,³⁰ and LNnT.³⁹ The percentage of fat mass change was independently reported in 4/19 studies and is closely related to body–mass index and weight gain. Positive correlations were noted for: 2’FL,⁴³ 3’SL,³⁸ 6’SL,³⁸ LNFP III,³⁸ LNFP II,^17,38 DSLNT,^17,38 LST b,³⁸ FDSLNH,¹⁷ DSLNH,³⁸ DFLNT,⁴³ whereas negative association was reported for: 6’SL,⁴³ LNnT,^17,54 LNFP I,¹⁷ LNFP III,⁴³ and DFLNH.⁵⁴ In 4/19 studies, the change in fat-free (lean) mass was reported, and it was positively related to 3FL,⁴³ 3’SL,^43,48 DFLac,⁴³ DFLNH,⁴³ and DFLNT,⁴³ whereas a negative correlation was found for: LNT,⁴⁸ LNFP I,¹⁷ LST c.^48,49 Skinfold thickness-for-age z-score, a novel metric, was positively correlated with 3FL and 3′SL and negatively correlated with 6′SL, LNnT, and LST c.³¹

HMOs predominant in stunting or severe acute malnutrition

Two studies focused on the effects of undernutrition on HMO composition and consequently on stunting⁴⁷ and severe acute malnutrition (SAM)⁵³ in Malawi and Bangladeshi children, respectively. In these studies, ultrasensitive qTOF-MS was used to conduct robust HMO profiling, yielding 50 and 73 HMOs, respectively. Lower levels of fucosylated and sialylated HMOs were linked to stunting in Malawian infants.⁴⁷ In contrast, total sialylated HMOs were linked to a higher risk of SAM in Bangladeshi children, while there was no evidence of a significant association between SAM and fucosylated or undecorated HMOs.⁵³

Discussion

Our study provides a comprehensive analysis of 19 HMOs on early childhood milestones based on published studies in order to better understand their roles, which have remained understudied compared to other human milk components. We collected and pooled data that suggest that 2′FL, 3FL, and 3′SL are associated with early brain maturation, which is important in understanding the optimal growth and cognitive development benefits observed in breastfed infants and superior outcomes of infants nursed by secretor mothers.^37,55 Suboptimal breastfeeding or poor maternal nutrition resulted in lower total fucosylated and sialylated HMOs, resulting in stunting.^47,53 In a Danish study, Larsson et al. found that adding 2′FL to formula milk resulted in overnutrition and excessive weight gain in children born to secretor mothers, whose milk contains 2′FL.⁵⁴ However, in two studies that used 2′FL supplementation, there were no differences between formula-fed and age-matched exclusively breastfed infants.^56,57 Interestingly, LNFP II was found to be protective against pediatric obesity, and early supplementation with this HMO may help to prevent childhood obesity, which can persist into adolescence and adulthood.³⁵ The pooled studies were highly heterogeneous, with secretors having higher HMO concentrations despite the fact that the difference was not statistically significant. Nevertheless, key HMOs involved in early childhood development, such as 2′FL and 3′SL, were predominant in secretors explaining the benefits to their nursing infants, including brain development, infant growth velocity, and cognitive development, all of which are covered in more detail in the sections that follow.

Among the 224 identified HMOs,⁵⁸ this review focused on 19 well-profiled HMOs and their effects on brain development, infant growth velocity, and cognitive development. As expected, there was evidence of between-study heterogeneity, which ranged from low to substantial in accordance with Higgins and Thompson’s I² statistic.⁵⁹ It is reasonable to anticipate between-study heterogeneity given that the pooled data come from various studies with varying study designs, encompassing variations in study populations, HMO characterization analytical techniques, and covariate adjustments. In this systematic review, the pooled concentrations of 2′FL, DFlac, DFLNT, LNFP I, LST c, LNnT, FDSLNH, LNFP II, LNFP III, LST b, and LNT were statistically different between secretors and nonsecretors, consistent with previous studies that found HMO concentrations varied by maternal secretor status.^60–63 The selected studies also confirmed the previously reported observation that secretor status varies by geographical location, with the highest proportions of secretor mothers in South America and Northern Europe and the lowest proportions in Africa and Asia.^64–69 The expression of the FUT2 gene, whose locus is mapped to chromosome 19, is now known to be the primary driver of this genetic variation.¹⁰ A compensatory mechanism occurs through expression of FUT3 gene (Lewis positive) where mothers produce milk containing oligosaccharides with α1–3 and α1–4 linked fucoses, including 3FL and LNFP II.^12,13 This finding confirms that evolutionary dynamics ensured that nonsecretors retained the ability to produce key fucosylated HMOs required for early childhood development although not as efficiently as secretors. Seasonality was identified as a key environmental factor regulating HMO synthesis in Gambian mothers,¹⁸ and similar patterns were observed in Israeli mothers, with HMO concentrations significantly lower in the summer.⁶⁸ Given that seasons are a common phenomenon all over the world, this is a critical factor that should be considered in study designs. Although not frequently reported as a probable causal factor of interest, infants’ biological sex influenced HMOs in Brazilian mothers,³⁶ which is consistent with the findings of Asher and colleagues in a study in which seasonality was adjusted for.⁶⁸ The milk collection technique, which was dependent on geographical location, could also have contributed to the observed differences. Moossavi and colleagues discovered that using breast pumps was associated with lower relative abundance of Bifidobacterium spp. and higher levels of DSLNH, emphasizing the indirect role of the microbiome, which will be discussed in detail under infant growth.⁷⁰ Maternal weight and BMI adjusted for in 13 studies may have also played a role in HMO concentrations, consistent with previous findings.^66,71

Breastfeeding that starts as soon as the baby is born and continues exclusively for 6 months and extended breastfeeding for up to 2 years have been linked to higher cognitive functions in children studied using REEF, BRIEF, ASQ, and MSEL tools since the early ages.^24,72 This has been attributed to the abundance of sialic acid in sialylated HMOs, as well as its well-known function in brain development, neuronal transmission, and synaptogenesis.^73,74 This link was clarified by Berger and colleagues’ groundbreaking study,³⁷ which demonstrated the link between key fucosylated and sialylated HMOs on brain maturation, explaining why breastfed infants of secretor mothers have more health benefits than those of nonsecretor mothers. It also suggests that in the absence of 2′FL supplementation, infants of nonsecretor mothers who express no 2′FL and lower amounts of 3′SL benefit less or that compensation occurs through unknown mechanisms. With the link between HMOs and brain development and thus higher cognitive functions established, we now have a solid foundation to advocate for the importance of exclusive breastfeeding after birth to reap the greatest benefits, as the benefits identified at 1 month diminish after 6 months, indicating that early exposure is temporal and the most critical time which has an influence on infant cognitive development.³⁴ This knowledge may also benefit preterm infants born when their mothers’ entero-mammary pathway is not fully developed,⁷⁵ by exposing them to mature donor milk from confirmed secretor mothers which are loaded with sufficient 2’FL and microbiome for optimal development. In a study of preterm infants, LNFP III was found to be significantly associated with higher cognitive development in infants born to Secretor(+) Lewis(+) mothers. This suggests that LNFP III might play a significant role in this vulnerable group, but only in the context of secretor mothers’ infants.⁴⁸ The microbiota is also involved in cognitive development through the gut-brain axis. For instance, the relative abundances of Veillonella spp., a critical microbiota regulating γ-aminobutyric acid (GABA), an important neurotransmitter for early brain development,⁷⁶ were inversely correlated with concentrations of sialylated HMOs.⁶⁵ Some HMOs, including LNT and DSLNT, were linked to poor cognitive development and were notably higher in nonsecretor mothers in our pooled analyses.^34,50

While the link between cognitive development and sialylated HMOs is well established,³⁷ the connection between HMOs and growth velocity is blurred and indirect given that growth functions are largely influenced by lipids, lactose, and proteins.^5,77 In contrast to nonhuman mammals that gain weight considerably more quickly, human milk has relatively low protein and fat concentrations but higher HMO concentrations. This is probably an adaptation to the lower requirements of baby growth while allowing for the optimal development of higher order cognitive skills at a younger age.^58,78 Nonetheless, the importance of HMOs for physical growth can be seen in the context of malnutrition, where the amount of total HMOs in breast milk from mothers whose children were stunted was lower than the amount of total HMOs in breast milk from mothers whose children were healthy. This has an adverse effect on infants born to nonsecretor mothers because they are unable to compensate for deficiencies in fucosylated HMOs by secreting higher quantities of other HMOs, such as sialylated HMOs, resulting in milk that does not promote infant growth.⁴⁷ Addressing undernutrition is a global issue because, as of 2020, 144 million children under the age of five were stunted, which disadvantages them mainly because they have difficulty learning in school, earn less as adults, and face barriers to participation in their communities.⁷⁹ 2′FL, a critical HMO in brain development and cognitive ability, was linked to weight gain, height gain, head circumference, and body composition, which highlight the significance of this HMO, which we posit confers physical growth in a secretor-specific manner alongside other milk components.^{30,31,43,47,48,54} The second HMO of interest, 3′SL, had contradictory findings on weight gain, with four articles showing positive associations^18,30,38,43 and two articles showing negative associations.^31,32 3′SL was additionally linked to height gain, head growth, and body composition.^{31,38,39,43,48,51,54} The reason for this inconsistency in weight gain could be due to microbiota or macronutrients such as lipids and proteins found in milk, which have a direct relationship with growth. This can be further understood in the context of high-income overnutrition, which manifests as pediatric obesity, notably when breastfeeding is supplemented with formula containing 2′FL in mixed feeding settings occasioning over dosage on a critical HMO.^35,39,54

In our pooled analyses, LNFP II, a fucosylated HMO that was predominant in nonsecretor mothers, was protective against excessive weight gain regardless of maternal secretor status, implying that early exposure to LNFP-II could offset early obesity risk, which is important to policymakers as a mitigation strategy in high income societies with pediatric obesity challenges.³⁵ With the microbiome featuring as an important indirect regulator of infant growth, we explore its role in this early critical event, fully aware of the consequences of dysbiosis induced by antibiotics.⁷⁶ Breastfed infants appear to have better outcomes because breast milk provides both HMOs and commensal bacteria (microbiota) in early life, with the former acting as prebiotics that are metabolized by the latter and thus playing a critical role in immune maturation and the latter enhancing infant health by preventing pathogen adhesion and promoting gut colonization by beneficial microbes.^6,80 Because milk oligosaccharides are structurally similar to intestinal mucin O-glycans, bacterial glycosidases digest carbohydrates from the intestine’s protective mucin layer, allowing pathobionts like Enterobacteriaceae to cross-feed.⁸¹ Any factor can lead to microbiota dysbiosis early in life—such as inadequate breastfeeding, prolonged diarrhea, or antibiotics which eliminates crucial taxa for preserving homeostasis, leaving vacancies to be filled by blooms of pathobionts that may be more effective at extracting energy, predisposing some infants to obesity later in life.^76,82 Furthermore, HMOs function as receptor decoys, preventing bacterial adhesion to mucosal surfaces and thus preventing diarrhea.¹³ They also reduce the pH of the infant’s digestive tract, increasing the proportion of beneficial bacteria Bifidobacterium longum, while decreasing Escherichia coli and Clostridium perfringens.⁸³ In this context, we argue that children who are less sick will grow faster.

Strengths and limitations

The strength of this study was its ability to summarize and quantify evidence for the role of HMOs on three interconnected early growth metrics, namely brain development, physical growth, and cognitive development, using clearly defined research question. We were cognizant of the fact that evidence synthesis requires adequate representation for validity of the findings. We also included a wide spectrum of systematically selected studies from high- and low-income countries, which helps to understand the influence of HMOs in regions of the world with highest undernutrition burden and associated highest neonatal and infant mortality rates globally which is critical for decision-making.

The main limitation of the studies was inconsistent reporting of HMO concentration; some authors used mean and standard deviation, whereas others used median and IQR, which had to be converted, potentially losing some accuracy. Second, the random-effects model we used to calculate the pooled effect reports four heterogeneity statistics by default, which makes deciding what to interpret and what to leave out challenging. In this case, we reported both the τ² statistic, which is insensitive to changes in the number of studies and their precision but is often difficult to interpret, and the I² statistic, which is insensitive to changes in the number of studies and heavily depends on the precision of the included studies but is easier to interpret the percentage. Finally, we had only 12 studies comparing HMO concentrations between secretors and nonsecretors.

Conclusion and future opportunities

This study revealed that early exposure to 2′FL and 3′SL, which are abundant in secretor mothers, gives nursing infants an early childhood development advantage through optimal brain and cognitive development. The proportion of secretor mothers in our study was 78%, implying that 22% of infants of nonsecretors have no access to 2′FL and only have suboptimal access to critical 3′SL through breastfeeding, both of which play critical roles in early childhood growth and development. Key emerging issues warrant further investigation in future studies. There is a need for well-designed, sufficiently powered studies that distinguish between secretor and nonsecretor women. Since suboptimal breastfeeding or poor maternal nutrition is linked to poor developmental outcomes, it is important that these infants are supplemented with key HMOs that are linked with optimal growth. Modern technology has enabled large-scale production of essential HMOs such as 2′FL, which is approved as a safe ingredient in baby formula in both Europe and the United States. This is in contrast to the past, when large-scale production seemed unfeasible. In contrast, there is a risk of pediatric obesity, which may persist into adulthood. The finding that LNFP II is protective against pediatric obesity is of great interest, and more research is needed to confirm the findings. If successful, supplementation with this HMO holds the promise of new therapeutic applications to address obesity, a global concern.

Footnotes

Authors’ Contributions

Conceptualization: M.M., R.N., and D.W. Methodology: M.M. and H.K. Investigation: M.M., H.K., L.O., and H.A. Statistical analysis: M.M. and H.A. Writing—original draft: M.M. and H.A. Writing—review & editing: M.M., R.N., and D.W.

Disclosure Statement

The authors have declared that no competing interests exist.

Funding Information

The authors received no specific funding for this work.

Data Availability Statement

All the data included were extracted from the selected original articles. The code used for the meta-analysis can be requested from the corresponding author.

Supplementary Material

References

Victora

, Bahl

, Barros

AJD

, et al. Breastfeeding in the 21st century: Epidemiology, mechanisms, and lifelong effect. Lancet, 2016; 387(10017):475–490; doi: 10.1016/S0140-6736(15)01024-7

Dinleyici

, Barbieur

, Dinleyici

, et al. Functional effects of human milk oligosaccharides (HMOs). Gut Microbes, 2023; 15(1):2186115; doi: 10.1080/19490976.2023.2186115

Anderson

, Johnstone

, Remley

. Breast feeding and cognitive development: A meta-analysis. Am J Clin Nutr, 1999; 70(4):525–535.

Hennet

, Borsig

. Breastfed at Tiffany’s. Trends Biochem Sci, 2016; 41(6):508–518; doi: 10.1016/j.tibs.2016.02.008

Ballard

, Morrow

. Human milk composition. nutrients and bioactive factors. Pediatr Clin North Am, 2013; 60(1):49–74; doi: 10.1016/j.pcl.2012.10.002

Bode

. Human milk oligosaccharides: Every baby needs a sugar mama. Glycobiology, 2012; 22(9):1147–1162; doi: 10.1093/glycob/cws074

Chen

. Human milk oligosaccharides (HMOs): Structure, function, and enzyme-catalyzed synthesis. 1st ed. Elsevier Inc.; 2015; doi: 10.1016/bs.accb.2015.08.002

Kunz

, Rudloff

, Baier

, et al. Oligosaccharides in human milk: Structural, functional, and metabolic aspects. Annu Rev Nutr, 2000; 20:699–722; doi: 10.1146/annurev.nutr.20.1.699

Kunz

, Rudloff

, Hintelmann

, et al. High-pH anion-exchange chromatography with pulsed amperometric detection and molar response factors of human milk oligosaccharides. J Chromatogr B Biomed Appl, 1996; 685(2):211–221; doi: 10.1016/S0378-4347(96)00181-8

10.

Williams

, McGuire

, Meehan

, et al. Key genetic variants associated with variation of milk oligosaccharides from diverse human populations. Genomics, 2021; 113(4):1867–1875; doi: 10.1016/j.ygeno.2021.04.004

11.

Totten

, Zivkovic

, Wu

, et al. Comprehensive profiles of human milk oligosaccharides yield highly sensitive and specific markers for determining secretor status in lactating mothers. J Proteome Res, 2012; 11(12):6124–6133; doi: 10.1021/pr300769g

12.

Thurl

, Munzert

, Henker

, et al. Variation of human milk oligosaccharides in relation to milk groups and lactational periods. Br J Nutr, 2010; 104(9):1261–1271; doi: 10.1017/S0007114510002072

13.

Bode

. The functional biology of human milk oligosaccharides. Early Hum Dev, 2015; 91(11):619–622; doi: 10.1016/j.earlhumdev.2015.09.001

14.

Hamosh

. Bioactive factors human milk. Pediatr Clin North Am, 2001; 48(1):69–86; doi: 10.1016/s0031-3955(05)70286-8

15.

Newburg

. Neonatal protection by an innate immune system of human milk consisting of oligosaccharides and glycans. J Anim Sci, 2009; 87(13 Suppl):26–34; doi: 10.2527/jas.2008-1347

16.

Triantis

, Bode

, van Neerven

JRJ

. Immunological effects of human milk oligosaccharides. Front Pediatr, 2018; 6(July):190; doi: 10.3389/fped.2018.00190

17.

Alderete

, Autran

, Brekke

, et al. Associations between human milk oligosaccharides and infant body composition in the first 6 mo of life. Am J Clin Nutr, 2015; 102(6):1381–1388; doi: 10.3945/ajcn.115.115451

18.

Davis

JCC

, Lewis

, Krishnan

, et al. Growth and morbidity of Gambian infants are influenced by maternal milk oligosaccharides and infant gut microbiota. Sci Rep, 2017; 7:40466; doi: 10.1038/srep40466

19.

, Chen

, Newburg

. Utilization of major fucosylated and sialylated human milk oligosaccharides by isolated human gut microbes. Glycobiology, 2013; 23(11):1281–1292; doi: 10.1093/glycob/cwt065

20.

Tarr

, Galley

, Fisher

, et al. The prebiotics 3’sialyllactose and 6’sialyllactose diminish stressor-induced anxiety-like behavior and colonic microbiota alterations: Evidence for effects on the gut-brain axis. Brain Behav Immun, 2015; 50:166–177; doi: 10.1016/j.bbi.2015.06.025

21.

Jacobi

, Yatsunenko

, Li

, et al. Dietary isomers of sialyllactose increase ganglioside sialic acid concentrations in the corpus callosum and cerebellum and modulate the colonic microbiota of formula-fed piglets. J Nutr, 2016; 146(2):200–208; doi: 10.3945/jn.115.220152

22.

Gale

, Logan

, Santhakumaran

, et al. Effect of breastfeeding compared with formula feeding on infant body composition: A systematic review and meta-analysis. Am J Clin Nutr, 2012; 95(3):656–669; doi: 10.3945/ajcn.111.027284

23.

Norrish

, Sindi

, Sakalidis

, et al. Relationships between the Intakes of Human Milk Components and Body Composition of Breastfed Infants: A systematic review. Nutrients, 2023; 15(10); doi: 10.3390/nu15102370

24.

Hou

, Li

, Yan

, et al. Impact of the duration of breastfeeding on the intelligence of children: A systematic review with network meta-analysis. Breastfeed Med, 2021; 16(9):687–696; doi: 10.1089/bfm.2020.0364

25.

Lechner

, Vohr

. Neurodevelopmental outcomes of preterm infants fed human milk: A systematic review. Clin Perinatol, 2017; 44(1):69–83; doi: 10.1016/j.clp.2016.11.004

26.

Lockyer

, McCann

, Moore

. Breast milk micronutrients and infant neurodevelopmental outcomes: A systematic review. Nutrients, 2021; 13(11):1–16; doi: 10.3390/nu13113848

27.

Page

, McKenzie

, Bossuyt

, et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. Bmj, 2021; 372:n71; doi: 10.1136/bmj.n71

28.

Ouzzani

, Hammady

, Fedorowicz

, et al. Rayyan-a web and mobile app for systematic reviews. Syst Rev, 2016; 5(1):210–210; doi: 10.1186/s13643-016-0384-4

29.

Wells

, Shea

, O’Connell

, et al. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses. 2017. Available from: https://www.ohri.ca//programs/clinical_epidemiology/oxford.asp [Last accessed: September 29, 2022].

30.

Lagström

, Rautava

, Ollila

, et al. Associations between human milk oligosaccharides and growth in infancy and early childhood. Am J Clin Nutr, 2020; 111(4):769–778; doi: 10.1093/ajcn/nqaa010

31.

Mansell

, Furst

, O’Hely

, et al. Age-dependent associations of human milk oligosaccharides with body size and composition up to 4 years of age. Am J Clin Nutr, 2023; 117(5):930–945; doi: 10.1016/j.ajcnut.2023.02.016

32.

Dou

, Luo

, Xing

, et al. Human milk oligosaccharides variation in gestational diabetes mellitus mothers. Nutrients, 2023; 15(6):1441; doi: 10.3390/nu15061441

33.

Sprenger

, Lee

, De Castro

, et al. Longitudinal change of selected human milk oligosaccharides and association to infants’ growth, an observatory, single center, longitudinal cohort study. PLoS One, 2017; 12(2):e0171814–15; doi: 10.1371/journal.pone.0171814

34.

Berger

, Plows

, Jones

, et al. Human milk oligosaccharide 2′-fucosyllactose links feedings at 1 month to cognitive development at 24 months in infants of normal and overweight mothers. PLoS One, 2020; 15(2):e0228323–12; doi: 10.1371/journal.pone.0228323

35.

Berger

, Plows

, Jones

, et al. Human milk oligosaccharides and Hispanic infant weight gain in the first 6 months. Obesity (Silver Spring), 2020; 28(8):1519–1525; doi: 10.1002/oby.22884

36.

Tonon

, de Morais

, Abrão

ACFV

, et al. Maternal and infant factors associated with human milk oligosaccharides concentrations according to secretor and Lewis phenotypes. Nutrients, 2019; 11(6):1–23; doi: 10.3390/nu11061358

37.

Berger

, Bansal

, Sawardekar

, et al. Associations of human milk oligosaccharides with infant brain tissue organization and regional blood flow at 1 month of age. Nutrients, 2022; 14(18):1–14; doi: 10.3390/nu14183820

38.

Saben

, Sims

, Abraham

, et al. Human milk oligosaccharide concentrations and infant intakes are associated with maternal overweight and obesity and predict infant growth. Nutrients, 2021; 13(2):1–16; doi: 10.3390/nu13020446

39.

Menzel

, Vogel

, Austin

, et al. Concentrations of oligosaccharides in human milk and child growth. BMC Pediatr, 2021; 21(1):481–411; doi: 10.1186/s12887-021-02953-0

40.

Oliveros

, Martín

, Torres-Espínola

, et al. Human milk levels of 2′-fucosyllactose and 6’-sialyllactose are positively associated with infant neurodevelopment and are not impacted by maternal BMI or diabetic status. J Nutr Food Sci, 2021; 4(1):100024.

41.

Wang

, Zhao

, et al. Neutral human milk oligosaccharides are associated with multiple fixed and modifiable maternal and infant characteristics. Nutrients, 2020; 12(3):10–3390; doi: 10.3390/nu12030826

42.

Paganini

, Uyoga

, Kortman

GAM

, et al. Maternal human milk oligosaccharide profile modulates the impact of an intervention with iron and galacto-oligosaccharides in Kenyan infants. Nutrients, 2019; 11(11); doi: 10.3390/nu11112596

43.

Cheema

, Gridneva

, Furst

, et al. Human milk oligosaccharides and bacterial profile modulate infant body composition during exclusive breastfeeding. Int J Mol Sci, 2022; 23(5):1–28; doi: 10.3390/ijms23052865

44.

Willemsen

, Beijers

, Gu

, et al. Fucosylated human milk oligosaccharides during the first 12 postnatal weeks are associated with better executive functions in toddlers. Nutrients, 2023; 15(6); doi: 10.3390/nu15061463

45.

Cho

, Zhu

, Li

, et al. Human milk 3′-Sialyllactose is positively associated with language development during infancy. Am J Clin Nutr, 2021; 114(2):588–597; doi: 10.1093/ajcn/nqab103

46.

Binia

, Lavalle

, Chen

, et al. Human milk oligosaccharides, infant growth, and adiposity over the first 4 months of lactation. Pediatr Res, 2021; 90(3):684–693; doi: 10.1038/s41390-020-01328-y

47.

Charbonneau

, O’Donnell

, Blanton

, et al. Sialylated milk oligosaccharides promote microbiota-dependent growth in models of infant undernutrition. Cell, 2016; 164(5):859–871; doi: 10.1016/j.cell.2016.01.024

48.

Rozé

, Hartweg

, Simon

, et al. Human milk oligosaccharides in breast milk and 2-year outcome in preterm infants: An exploratory analysis. Clin Nutr, 2022; 41(9):1896–1905; doi: 10.1016/j.clnu.2022.07.024

49.

Wejryd

, Martí

, Marchini

, et al. Low diversity of human milk oligosaccharides is associated with necrotising enterocolitis in extremely low birth weight infants. Nutrients, 2018; 10(10); doi: 10.3390/nu10101556

50.

Ferreira

ALL

, Alves-Santos

, Freitas-Costa

, et al. Associations between human milk oligosaccharides at 1 month and infant development throughout the first Year of Life in a Brazilian Cohort. J Nutr, 2021; 151(11):3543–3554; doi: 10.1093/jn/nxab271

51.

Jorgensen

, Young

, Ashorn

, et al. Associations of human milk oligosaccharides and bioactive proteins with infant growth and development among Malawian mother-infant dyads. Am J Clin Nutr, 2021; 113(1):209–220; doi: 10.1093/ajcn/nqaa272

52.

Samuel

, Hartweg

, Lebumfacil

, et al. Dynamics of human milk oligosaccharides in early lactation and relation with growth and appetitive traits of Filipino breastfed infants. Sci Rep, 2022; 12(1):17304–17313; doi: 10.1038/s41598-022-22244-7

53.

Nuzhat

, Palit

, Mahfuz

, et al. Association of human milk oligosaccharides and nutritional status of young infants among Bangladeshi mother–infant dyads. Sci Rep, 2022; 12(1):9456–9459; doi: 10.1038/s41598-022-13296-w

54.

Larsson

, Lind

, Laursen

, et al. Human milk oligosaccharide composition is associated with excessive weight gain during exclusive breastfeeding—an explorative study. Front Pediatr, 2019; 7(July):297–214; doi: 10.3389/fped.2019.00297

55.

Robertson

, Manges

, Finlay

, et al. The human microbiome and child growth – first 1000 days and beyond. Trends Microbiol, 2019; 27(2):131–147; doi: 10.1016/j.tim.2018.09.008

56.

Marriage

, Buck

, Goehring

, et al. Infants fed a lower calorie formula with 2′FL show growth and 2′FL uptake like breast-fed infants. J Pediatr Gastroenterol Nutr, 2015; 61(6):649–658; doi: 10.1097/MPG.0000000000000889

57.

Puccio

, Alliet

, Cajozzo

, et al. Effects of infant formula with human milk oligosaccharides on growth and morbidity: A randomized multicenter trial. J Pediatr Gastroenterol Nutr, 2017; 64(4):624–631; doi: 10.1097/MPG.0000000000001520

58.

Durham

, Wei

, Lemay

, et al. Creation of a milk oligosaccharide database, MilkOligoDB, reveals common structural motifs and extensive diversity across mammals. Sci Rep, 2023; 13(1):10345–10326; doi: 10.1038/s41598-023-36866-y

59.

Higgins

JPT

, Thompson

. Quantifying heterogeneity in a meta-analysis. Stat Med, 2002; 21(11):1539–1558; doi: 10.1002/sim.1186

60.

Kunz

, Rudloff

. Compositional analysis and metabolism of human milk oligosaccharides in infants. Nestle Nutr Inst Workshop Ser, 2017; 88:137–147; doi: 10.1159/000455398

61.

Newburg

. Human milk oligosaccharides vary among populations. Am J Clin Nutr, 2017; 105(5):1027–1028; doi: 10.3945/ajcn.117.155721

62.

Thurl

, Munzert

, Boehm

, et al. Systematic review of the concentrations of oligosaccharides in human milk. Nutr Rev, 2017; 75(11):920–933; doi: 10.1093/nutrit/nux044

63.

Saben

, Abraham

, Bode

, et al. Third-trimester glucose homeostasis in healthy women is differentially associated with human milk oligosaccharide composition at 2 months postpartum by secretor phenotype. Nutrients, 2020; 12(8):1–15; doi: 10.3390/nu12082209

64.

Samuel

, Binia

, de Castro

, et al. Impact of maternal characteristics on human milk oligosaccharide composition over the first 4 months of lactation in a cohort of healthy European mothers. Sci Rep, 2019; 9(1):11767–11710; doi: 10.1038/s41598-019-48337-4

65.

Pace

, Williams

, Robertson

, et al. Variation in human milk composition is related to differences in milk and infant fecal microbial communities. Microorganisms, 2021; 9(6):1–17; doi: 10.3390/microorganisms9061153

66.

McGuire

, Meehan

, McGuire

, et al. What’s normal? Oligosaccharide concentrations and profiles in milk produced by healthy women vary geographically. Am J Clin Nutr, 2017; 105(5):1086–1100; doi: 10.3945/ajcn.116.139980.Am

67.

Vinjamuri

, Davis

JCCC

, Totten

, et al. Human milk oligosaccharide compositions illustrate global variations in early nutrition. J Nutr, 2022; 152(5):1239–1253; doi: 10.1093/jn/nxac027

68.

Asher

, Mangel

, Ari

, et al. Human milk oligosaccharide profile across lactation stages in Israeli women—A prospective observational study. Nutrients, 2023; 15(11):1–16; doi: 10.3390/nu15112548

69.

Pell

, Ohuma

, Yonemitsu

, et al. The human-milk oligosaccharide profile of lactating women in Dhaka, Bangladesh. Curr Dev Nutr, 2022; 5(12):nzab137; doi: 10.1201/9780429293016

70.

Moossavi

, Atakora

, Miliku

, et al. Integrated analysis of human milk microbiota with oligosaccharides and fatty acids in the child cohort. Front Nutr, 2019; 6(May):58–16; doi: 10.3389/fnut.2019.00058

71.

Lemay-Nedjelski

, Yonemitsu

, Asbury

, et al. Oligosaccharides and microbiota in human milk are interrelated at 3 months postpartum in a cohort of women with a high prevalence of gestational impaired glucose tolerance. J Nutr, 2021; 151(11):3431–3441; doi: 10.1093/jn/nxab270

72.

Dangat

, Joshi

. Breast Milk and Cognitive Performance in Children. Elsevier Inc.; 2023.; doi: 10.1016/b978-0-323-89834-8.00014-3

73.

Hauser

, Pisa

, Arias Vásquez

, et al. Sialylated human milk oligosaccharides program cognitive development through a non-genomic transmission mode. Mol Psychiatry, 2021; 26(7):2854–2871; doi: 10.1038/s41380-021-01054-9

74.

Wang

. Sialic acid is an essential nutrient for brain development and cognition. Annu Rev Nutr, 2009; 29:177–222; doi: 10.1146/annurev.nutr.28.061807.155515

75.

Khodayar-Pardo

, Mira-Pascual

, Collado

, et al. Impact of lactation stage, gestational age and mode of delivery on breast milk microbiota. J Perinatol, 2014; 34(8):599–605; doi: 10.1038/jp.2014.47

76.

Mulinge

, Mwanza

, Kabahweza

, et al. The impact of neonatal intensive care unit antibiotics on gut bacterial microbiota of preterm infants;: A systematic review. Front Microbiomes, 2023; 2:1180565; doi: 10.3389/frmbi.2023.1180565

77.

Dror

, Allen

. Overview of nutrients in human milk. Adv Nutr, 2018; 9(suppl_1):278S–294S; doi: 10.1093/advances/nmy022

78.

Capuco

, Akers

. The origin and evolution of lactation. J Biol, 2009; 8(4):37–11; doi: 10.1186/jbiol139

79.

UNICEF/WHO/World Bank. Levels and trends in child malnutrition. 2020.

80.

Lyons

, Ryan

, Dempsey

, et al. Breast milk, a source of beneficial microbes and associated benefits for infant health. Nutrients, 2020; 12(4); doi: 10.3390/nu12041039

81.

, Ferreyra

, Higginbottom

, et al. Microbiota-liberated host sugars facilitate post-antibiotic expansion of enteric pathogens. Nature, 2013; 502(7469):96–99; doi: 10.1038/nature12503

82.

Rautava

. Early-Life Antibiotic Exposure, the Gut Microbiome, and Disease in Later Life. INC; 2021; doi: 10.1016/b978-0-12-818097-6.00006-7

83.

, Chen

, Kling

, et al. The principal fucosylated oligosaccharides of human milk exhibit prebiotic properties on cultured infant microbiota. Glycobiology, 2013; 23(2):169–177; doi: 10.1093/glycob/cws138

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB