Psychosocial adversity experienced in utero and early life is associated with variation in gut microbiota: a prospective case-control study

Barbara B. Warner, Bruce A. Rosa, I. Malick Ndao, Phillip I. Tarr, J. Phillip Miller, Sara K. England, Joan L. Luby, Cynthia E. Rogers, Carla Hall-Moore, Renay E. Bryant, Jacqueline D. Wang, Laura Linneman, Tara A. Smyser, Christopher D. Smyser, Deanna M. Barch, Gregory E. Miller, Edith Chen, John Martin, Makedonka Mitreva
doi: https://doi.org/10.1101/2022.11.17.22282482

Abstract

Social disadvantage (SD) and psychological stressors (PS) trap some populations in poverty, resulting in health inequities. How these two factors become biologically embedded and the pathways leading to adverse health outcomes is unclear, especially in infants exposed to psychosocial adversity in utero and during early life. Variation in gut microbiome structure and functions, and systemic elevations in circulating cytokines levels as indices of inflammation, offer two possible causative pathways. Here, we interrogate the gut microbiome of mother-child dyads and maternal inflammatory markers, and compare high-SD/high-PS dyads to pairs with low-SD/Low-PS, and demonstrate that the GM of high-SD and high-PS mothers may already be compromised, resulting in the lowest observed inter-individual similarity in that group. The strong predictors of maternal high-SD and high-PS based on mothers and children microbiomes were phylogenetically very distinct bacteria indicating different GM pathways associated with SD versus PS. We identified sets of SD- and PS-discriminatory metabolic pathways in the mothers and in the children, however their predictive power was lower compared to the discriminatory bacterial species. Prediction accuracy was consistently greater for IL-6 than for the other inflammatory markers, supporting an association between systemic inflammation and psychosocial adversity. The gut microbiome of the infants can be used to predict the psychosocial adversity of mothers, and are embedded in the gut microbiota of 4-month-old infants.

INTRODUCTION

Health inequities experienced by socially disadvantaged populations is an increasing and urgent societal issue [1, 2]. The psychosocial factors contributing to these disparities often begin early in life and include the prenatal environment which has profound effects on fetal and infant outcomes that can last a lifetime [3-5]. How social disadvatage (SD) and psychological stress (PS) becomes biologically embedded, and then leads to disparete health outcomes remains unclear.

The gut microbiome (GM) is one candidate driver of adverse outcomes. Many SD-related morbidities are associated with systemic chronic inflammation [6], and the GM, by shaping and modulating the immune system [7, 8], is associated with systemic inflammation and autoimmune disease, including many disorders associated with populations in which SD is widespread, and include diabetes [9], obesity [10], cardiovascular disease, and neurologic disorders [11].

The GM is itself shaped primarily by environment, with influences including environmental exposures of diet, antibiotics, exercise, stress and sleep deprivation [12]. Many of these exposures have distinct characteristics for individuals living with economic and psychosocial hardships. Despite increasing awareness of an intersection between the GM and psychosocial inequities [13, 14], few human studies [15-17] have examined the impact on gut microbial community structure and function, particularly in the perinatal period. Studies in pregnant women have focused on either maternal psychological state [18-22] or socioeconomic status [23-25] but not both, have used 16S RNA analysis, which limits, taxanomic and metabolic pathway identification, and have not linked mother-infants samples to examine maternal transfer within this context.

This study aims to fill these knowledge gaps. Using a prospective cohort assembled prenatally, we identify the distinct impact of exposure to SD and PS on GM structure and functions for mothers and their infants. We identify discriminatory taxa driving this association using whole metagenomic shotgun (WMS) sequencing, and attempt to relate systemic inflammation to the GM by examining maternal prenatal serum cytokines. Understanding the interaction of the gut microbiome with social determinants of health holds great promise for future preventions due to its potential alterability and plasticity over time. This is particularly true for the perinatal period, a critical developmental window in which perturbations in GM community structure and functions have long-lasting effects [26-28].

RESULTS AND DISCUSSION

A subset of 121 mother–child dyads were drawn from the larger parent study, Early Life Adversity Biological Embedding and Risk for Developmental Precursors of Mental Disorders (eLABE), which enrolled mothers prenatally and followed their infants after brith. These 121 mothers were drawn from across the psychosocial spectrum and representative of the parent cohort with average ages of 29.8 (± 5.1 years) vs. 29.2 (± 5.3 years), similar dietary Healthy Eating index profiles 59 (± 10.6%) vs 58 (± 9.9%), and gestational age at delivery 39.0 (± 1.1wks) vs. 38.3 (± 2 weeks). Mothers and their children were classified based on composite scores from two latent constructs, SD and PS [29]. The SD construct included 7 individual domains across income, neighborhood deprivation, insurance, and education. The PS construct included 9 individual domains across depression, stress, short term and lifetime, and racial discrimination (see Methods for details). To define an effect of SD from that of PS on the maternal and children microbiome, we examined the interrelationships among pre- and post-natal adversity and circulating cytokines as biomarkers of inflammation, with the GM. The mothers’ GM was profiled using third trimester stool samples before parturition (average -18.0 ± 16.9). Their children’s stools at were sampled 130.4 (± 13.1) days after birth. Circulating cytokines were analyzed in maternal blood samples obtained during the 3rd trimester of pregnancy. Participant characteristics are provided in Table 1 and Supplementary Table S1.

View this table:
Table 1.

Patients characteristics at study entry for each of the primary comparisons of interest. “Low” and “High” Disadvantage and Psych scores are separated according to the distribution of the metadata, as shown in Figure 1. Complete characteristics of all samples organized per individual sample is available in Supplementary Table S2.

Maternal SD and PS relation to maternal and child pre- and post-natal GM structure and function, based on high level taxonomic profiling

We performed targeted metagenomic sequencing (see Methods) for 121 mother-child dyads (Figure 1). Among the 121 mother–child dyads (Figure 1A), the SD scores were positively correlated with PS scores (r = 0.394, P < 10−5; Figure 1B), corroborating the literature [30] and demonstrating the importance of differentiating effects from these two overlapping variables. Several variables, including race, breast milk feeding frequency, healthy eating index and income-to-needs ratio, were significantly associated with high-vs-low SD and PS scores, also expected because these variables were used as inputs to calculate these scores [29] (Table 1; Supplementary Table S1).

Figure 1:
Figure 1:

Overview of cohorts. (A) Portrayal of the 16S rRNA and WMS sample sets. (B) Social Disadvantage (SD) scores are positively correlated with Psychological Stressor (PS) scores (across the 121 sample pairs). (C) The 89 WMS samples are divided into “low-SD”, “high-SD”, “low-PS” and “high-PS” groups based on the distribution of values (samples within 1 standard deviation of the average value for each variable are excluded).

Across all samples, 3,072 amplicon sequence variants (ASVs) representing unique 16S nucleotide sequences were detected. Relative abundance values per ASV per sample are provided for all 242 samples in Supplementary Table S2, and nucleotide sequences for each identified ASV are provided in Supplementary Table S3. In the mothers, SD and PS scores had an overall negative correlation with α-(within sample) diversity based on the bacterial taxonomic composition, although this correlation was not significant (Figure 2A), consistent with previous reports of decreased α-diversity associated with lower Socioeconomic Status (SES) [15-17]. In contrast, among the children, α-diversity was positively correlated with SD (r = 0.581, P < 10−5) and PS scores (r = 0.350, P = 8.3×10−5) (Figure 2B). This observation may be partly explained by the lower frequency of breast feeding in the high-SD mothers (Figure 2B), as it is reported that breast-fed infants had lower GM α-diversity compared to formula-fed infants at three months of age, but the GM diversity increased by six months [31].

Figure 2:
Figure 2:

GM sample diversity and composition comparisons with Social Disadvantage (SD) and Psychological Stressors (PS) scores. (A) SD scores and PD scores do not significantly correlate with GM α-diversity (Shannon diversity index) in the 121 stool samples from mothers. (B) α-diversity in the children is positively correlated with SD (P < 10−5) and PS (P = 8.3×10−5) scores. Relative proportions of human milk feeding are also shown, to provide additional context for potential sources of differential diversity. (C) β-diversity, measured by Bray-Curtis dissimilarity between the GM of each mother-child dyad, is positively correlated with SD (P < 10−5) and PS (P = 1.9×10−3) scores. (D) Hierarchical clustering (based on Bray-Curtis dissimilarity across all ASVs, complete linkage) identifies four major GM profile-based clusters in the mother. (E) No significant differences in SD or PS scores were identified between the four clusters based on maternal GM profiles (one-way ANOVA). (F) Hierarchical clustering identifies four major GM clusters in the children. (G) Among the GM clusters in children, cluster 4 has greater SD (one-way ANOVA P < 3.4×10−5) and PS (3.1×10−4) scores. Tukey post-hoc tests were used to identify significant differences between each cluster, *P ≤ 0.05, **** P ≤ 10−4, *****, P ≤ 10−5.

An examination of the similarity between mother and child GM (β-diversity) measured by the Bray-Curtis statistics, showed that mother–child dyads with low-SD scores had significantly less similarity (P < 10−5, Figure 2C). This may be partially due differences in α-diversity, where formula-fed/high-SD infants have GM that are more similar to their mother’s GM because of greater overall α-diversity. Bray-Curtis dissimilarity-based clustering identified significant associations between SD scores and PS scores in the microbiome profiles in the children, but not in the mothers (Figures 2D-2G). We identified four major GM clusters in the mothers, but none significantly differ in SD or PS (Figures 2D and 2E). We also identified four major sample clusters in children, with cluster 4 having significantly greater SD scores than the other three clusters (ANOVA P < 10−5), and also higher PS scores than clusters 1 and 3 (ANOVA P = 3.1×10−4) (Figures 2F and 2G). These results suggest that high-SD and high-PS scores are significantly associated with a distinct overall GM profile in the children, but not in the mothers.

We next compared Bray-Curtis similarity between each sample dyad, within and between high and low SD and PS groups (Figure 3). High-SD mothers had significantly more variable microbiome profiles than those of low-SD mothers, who presumably had more consistent healthy microbiome profiles. This effect was even stronger for the PS comparison, with low-PS mothers having the most similar GM to each other than in any of the other comparisons, and the high-PS mothers having as similar between-sample diversity as the low-SD mothers (Figure 3A). High-SD children had the most similar microbiome profiles (Figure 3B), which may relate to their increased α-diversity (Figure 2B). Low-SD children also have some overall similarity in their GM profiles, but the low-SD and high-SD children GMs have little similarity (consistent with clustering results in Figure 2F). The same is true for PS in the children, except the within-group similarity for low-PS and high-PS were not significantly different. Overall, these results suggest that high-SD and high-PS mothers have divergent, variable microbiomes compared to low-SD and low-PS mothers, who share some commonality in overall GM profiles. Taken together, these results suggest that high-SD and high-PS mother already have compromised GM resulting in low observed inter-individual similarity compared to the low-SD and low-PS group, which has a more similar and healthier mature GM.

Figure 3:
Figure 3:

Comparisons of β-diversity between sample sets based on high-vs-low SD and PS, quantified by Bray-Curtis similarity between sample pairs, and FDR-corrected two-tailed T-tests with unequal variance used to test significance (A) Comparisons of within- and between-group β-diversity of the GM for mothers with high-SD, low-SD, high-PS and low-PS. (B) Comparisons of within- and between-group GM β-diversity in children with high-SD, low-SD, high-PS and low-PS.

Species-level gut microbiome profiles and metabolic pathway reconstructions of 89 mother–child dyads

We performed WMS sequencing on a subset of stools from 89 of the 121 mother–child dyads, selected based on the distribution of SD and PS scores (Figure 1C) (greater and less than the average value + 0.5 and – 0.5 standard deviations). The profiles of the 178 GMs (89 mothers and their 89 children), were generated from an average of 6Gb reads per sample, characterized using the Unified Human Gastrointestinal Genome (UHGG) [32]. Relative abundance values per species per sample were calculated using normalized depths [33] for downstream analysis. A total of 2,219 bacterial genomes were detected across all samples (Figure 4A). Genome taxonomic annotations and relative genome abundance values per sample are provided for all 178 samples (Supplementary Table S2). Twelve genomes were detected frequently across all mothers and all children samples (≥ 40% of both), including five Bifidobacterium genus members (B. infantis, B. breve, B. bifidum, B. catenulatum and B. adolescentis), with Bifidobacterium infantis being the most frequently detected genome across all samples (69.7% of mothers and 89.9% of children). These 12 also included two Bacteroides species (B. dorei and B. xylanisolvens), two Faecalicatena species (F. gnavus and F. unclassified), Flavonifractor plautii and Eggerthella lenta (Figure 4B). Four of these species were also identified as “core mother-infant shared species” in a previous WGS study [34], however the limited overlap may be in part a result of using different genome reference databases (identifying clade-specific marker genes from MetaPhlAn2 [34, 35] vs mapping to the Unified Human Gastrointestinal Genome (UHGG) [32]) in these 2 studies.

Figure 4:
Figure 4:

Genome and pathway detection in the 89 mother and 89 child WMS samples. (A) The total number of bacterial genomes and genomes with species-level taxonomic annotation detected in at least 3 maternal samples and/or at least 3 child samples. (B) For each of the 2,219 bacterial genomes detected in any sample, the proportion of mother and child samples that were detected in the dataset. (C) The total number of metabolic pathways detected in the GM from at least 3 mother and at least 3 child stools. (D) For each the 468 metabolic pathways detected in any sample, the proportion of mother and child samples with detection in the dataset.

We next reconstructed the metabolic pathways to compare the functional potential of the GM communities (based on the relative abundance of metabolic pathways [36] in read counts per million, CPM) using HUMAnN3 [37-39]. A total of 468 pathways were detected across all samples, 438 of which were detected in at least 3 mother or 3 children GMs (Figure 4C). Relevant pathway annotations and relative pathway abundance values per sample are provided for all 178 samples (Supplementary Table S2). In contrast to the genomes with relatively sparse identification across samples, almost half (46.3%) of all detected pathways were identified in ≥ 90% of samples in both the mothers and the children (top right of the plot, Figure 4D), including 130 pathways (27.8%) detected in all 178 samples. Despite the taxonomic differences between the GM of mothers and children (31.71% shared genomes; Figure 4A), the core set of microbial functions is conserved across the groups with 87.2% of the pathways being encoded by both mothers and children’s GMs (Figure 4C). This greater similarity between mothers and children in their microbial metabolic pathways compared to their microbial taxonomic profiles has been observed previously, and may be attributed to “core” microbial community functions essential for all species, despite the distinct populations of species adapted to different diets at different stages of life [34].

Bacterial species discriminating between mothers SD and PS

To dissociate the impact of the highly inter-related SD from PS on the maternal GM, and to identify discriminatory bacterial taxa that are strong predictors of mother’s SD and PS scores, we analyzed taxonomic and pathway GM profiles using two statistical approaches. First, we used supervised Random Forest (RF; [40]) machine-learning was used with a two-round approach [41] to (i) quantify the ability to predict SD and PS classification based on the mothers’ and the childrens’ microbiome profiles (Figure 5), and (ii) identify the specific genomes and pathways that most strongly differentiate between the high and low SD and PS scores (Figures 6 and 7; see Methods). Second, we used linear discriminant analysis effect size (LEfSe; [42]) to test differential genome and pathway abundance testing, to assign P values and “effect size” values for significance, as additional validation of RF results. Differential abundance statistics for all genomes and all comparisons are provided in Supplementary Table S4.

Figure 5:
Figure 5:

The Random Forest (RF) classification accuracy (low-SD vs high-SD and low-PS vs high-PS, out-of-bag error) based on relative UHGG genome abundance in the mothers and children. (A) Overall classification accuracy, with P values indicating significance based on FDR-corrected binomial distribution tests (compared to random sample assignment). (B) For each of the four classification tests, receiver operating characteristic (ROC) curves are shown based on RF models, with the area under the curve (AUC) scores and associated Wilcoxon rank sum test results for each ROC curve indicated.

Figure 6:
Figure 6:

WMS genome differential abundance in the mothers, based on comparisons of high-vs-low SD scores and high-vs-low Psych scores. (A) Based on WMS genome profiles, RF can successfully classify mothers as having low and high SD 80.5% of the time (P = 1.4×10−6 compared to random assignment; FDR-corrected binomial distribution test). Taxonomy and relative abundance per sample for the genomes with the highest predictive value in the top-25 RF model are shown (ranked by mean decrease in accuracy of the RF model; MDA). Also displayed are -Log of the Kruskal-Wallis test P values from LEfSe (no value shown if the effect size was <2) and the overall abundance of the taxa when present. (B) Based on WMS genome profiles, Random Forest (RF) can successfully classify mothers into high and low Psych groups 79.4% of the time (P = 1.5×10−6). *Associated with low-SD in the children, **associated with high Psych in the children.

We identified a set of SD and PS-discriminatory bacteria (Figure 6) whose relative abundances can classify mothers into low- or high-SD 80.5% of the time (P = 1.4×10−6 compared to random assignment; FDR-corrected binomial distribution test; AUC = 0.862, P = 7.1×10−8, Wilcoxon rank sum test), and as low- or high-PS 79.4% of the time (P =1.5×10−6; AUC = 0.778, P = 8.2×10−7; Figure 5). The best predictors of SD (Figure 6A) and PS (Figure 6B) scores in the mothers were identified according to mean decrease of accuracy (MDA) scores (see Methods) for each genome across the low-SD group and the high-SD groups. The microbiome signature of low-SD and low-PS samples consisted of bacterial species present in high abundance in most of the samples, but which were detected with zero or very low abundance in the high-SD or high-PS samples (Figure 6).

Low-SD mothers are identified by increased abundance of many Firmicutes A genomes (Figure 6A). Lawsonibacter asaccharolyticus, a recently-identified butyrate-producing species [43, 44], was the strongest predictor of low-SD scores in the mothers (MDA = 10.2%, LEfSe effect size = 2.7, P = 3.7×10−6). GM-derived butyrate has a wide range of beneficial effects on health including regulating fluid transport, reducing inflammation, reinforcing the epithelial defense barrier, and modulating intestinal motility via mechanisms that include potent regulation of gene expression [45], but L. asaccharolyticus has not been previously and independently associated with these beneficial effects. Streptococcus thermophilus was also among the top 5 predictors (MDA = 5.0%, LEfSe effect size = 3.2, P = 1.7×10−4) and is used to produce yogurts and cheeses, has properties beneficial to health including the prevention of chronic gastritis and diarrhea, and immunomodulatory properties with possible benefits in inflammation [46]. S. thermophilus was also among the predictors of low-PS, and this species was detected with zero or low abundance in high-PS and high-SD individuals.

The predictors of high-SD and high-PS were phylogenetically quite distinct, enriched for Actinobacteria and Firmicutes C, vs. mainly Firmicutes A, respectively. Genomes from four Bifidobacterium species (B. catenulatum, B. bifidum, B. breve and B. infantis) are among the six most strongly associated with high-SD, and B. catenulatum is associated with high-PS as well. It is recognized that Bifidobacteria in the human gut vary with age, and while quantitively some are particularly important in infant GM its presence with aging is stable but abundance changes over time. In general high abundance of Bifidobacteria is related to gut homeostasis and health maintenance and protection, in part by producing a number of potentially health promoting metabolites including short chain fatty acids, conjugated linoleic acid and bacteriocins, and Bifidobacteria is postulated to improve health [47]. However, qualitative and quantitative (increasing abundance) in Bifidobacteria are associated with inflammatory disorders such as diverticulitis, inflammatory bowel disease, and colorectal cancer [48]. Additionally, a recent review of GM variations associated with major depressive disorder (MDD) identified Bifidobacterium as one of three genera most consistently associated with MDD across studies [49]. While the specific functional role of these high-SD associated Bifidobacteria species is unclear, there is a striking increase in overall Bifidobacterial abundance in the high-SD mothers (Supplementary Figure S1).

Bacteroides A mediterraneensis was most strongly associated with low-PS in mothers (MDA = 10.9%, effect size = 2.5, P = 1.5×10−4; Figure 6B). In mice, stress exposure reduces abundance of Bacteroides in the GM [50], and in humans, Bacteroides is one of five genera associated with healthy status vs. MDD patients [51], the but this is the first report of Bacteroides A mediterraneensis specifically being associated with PS in a human cohort. The same human MDD study [51] also identified Faecalibacterium and Prevotella as being negatively associated with MDD, and in this study Faecalibacterium sp. and Prevotella sp001275135 were the 5th and 7th strongest predictors to low-PS (respectively).

Three species of Blautia sp. were among the top four most strong predictors of high-PS in the mothers, and none of these four species were associated with SD, suggesting a specific link with psychological stressors. In human studies, Blautia was one of ten genera associated with MDD [51], and Blautia and Eggerthella (represented in the high-PS mothers by E. lenta) were significantly correlated with PSS scores [52]. The latter study also identified Blautia and Bifidobacteria (represented in the high-PS mothers by B. catenulatum) as being significantly associated with MDD [52]. However, we wish to note that the overall abundance of Blautia in the microbiome remained fairly consistent across the mothers (Supplementary Figure S2), and only the species identified in Figure 6B predict PS.

Bacterial species in the infant GM discriminating prenatal maternal SD and PS

A set of SD and PS-discriminatory bacterial genomes correctly classify children into low- or high-SD in 84.6% of comparisons (P = 6.7×10−9 compared to random assignment; FDR-corrected binomial distribution test; AUC = 0.920, P = 7.0×10−10,, Wilcoxon rank sum test), and as low- or high-PS in 82.1% of comparisons (P =7.7×10−8; AUC = 0.889, P = 9.8×10−8) Figure 5). The top predictors of SD score (Figure 7A) and PS score (Figure 7B) in the children were identified according to MDA scores.

Figure 7:
Figure 7:

WMS genome differential abundance in the children, based on comparisons of high-vs-low Social Disadvantage (SD) scores and high-vs-low Psychological Stressors (PS) scores. (A) Based on WMS genome profiles, Random Forest (RF) can successfully classify children as low- and high-SD 84.6% of the time (P = 6.7×10−9 compared to random assignment; FDR-corrected binomial distribution test). Taxonomy and relative abundance per sample for the genomes with the highest predictive value in the top-25 RF model are shown (ranked by mean decrease in accuracy of the RF model; MDA). Also displayed are -Log of the Kruskal-Wallis test P values from LEfSe (no value shown if the effect size was <2) and the overall abundance of the taxa when present. (B) Based on WMS genome profiles, RF can successfully classify children into high and low-PS groups 82.1% of the time (P = 7.7×10−8). *Associated with high-SD in the mothers, **associated with low Psych in the mothers, ***also associated with high-PS in the mothers.

Among the predictors that are most important for classification to high-SD in children and with the largest effect size are Enterobacter nimipressuralis, nearly absent in low-SD children (detected in only one sample), and Klebsiella pneumoniae, both proinflammatory lipopolysaccharide expressing Proteobacteria [53]. The strongest predictor of children with low-SD was B. infantis, a species frequently used as a probiotic to diminish inflammation and associated with breastfeeding [54]. B. infantis represented an average of 28.3% of the GM of the low-SD infants, but only 5.2% of the GM in the high-SD infants (MDA = 10.14%, LEfSe effect size = 5.1, P = 0.0083; Supplementary Figure S1). The top taxa based on the comparison of the SD scores in the mothers were not proinflammatory and belonged to genomes from four Bifidobacterium species previously associated with several putatively beneficial metabolites [47].

There was an overlap of the predictors for low-SD and low-PS, including Veillonella parvula A, and several Collinsella spp. Veillonella is a signature taxa of the 4-month microbiome, and with Collinsella have been found in breast-fed microbiome indicating reduced concentration of oxygen, increased production and utilization of lactic acid which is specific for milk dominated diet [55]. The high-SD and high-PS-discriminating bacteria were enriched for a broad range of evolutionarily distinct Firmicutes species (Figure 7). The best high-PS predictors included F. gnavus reported as pathobiont associated with inflammatory bowel disease [56] and Sutterella sp. Sutterella species are prevalent commensals in the human GM with mild-proinflammatory properties [57].

Metabolic pathways associated with SD and PS in mothers and children

The relative genomic abundance data from the WMS dataset has more features and more taxonomic specificity than metabolic pathway abundance data, because pathways are much more conserved across samples (Figure 4). Genomic data therefore enables better estimation of overall GM associations with SD and PS scores (Figure 5). Nevertheless, the same statistical approaches (see Methods; RF and LEfSe) were used to identify sets of SD- and PS-discriminatory metabolic pathways in the mothers and in the children (defined here using MetaCyc, a curated database of experimentally elucidated metabolic pathways from all domains of life [36]). Differential abundance statistics for all pathways and all comparisons are provided in Supplementary Table S5.

Based on metabolic pathway profiles, machine learning correctly classifies mothers as high- and low-SD in 71.8% of comparisons (P = 8.5×10−4) and as high- and low-PS in 69.1% of comparisons (P = 0.0023). The top predictive pathways for both comparisons are shown in Table 2. The pathways with greatest association with high-SD in mothers include three related to carbohydrate degradation (sucrose degradation IV, glycogen degradation I, starch degradation III), which might relate to the accompanying abundance of Bifidobacterium species, which are rich in carbohydrate metabolism pathways [58]. The “myo-, chiro- and scyllo-inositol degradation” pathway (PWY-7237) was most strongly associated with high-PS in mothers (MDA = 5.95, P = 2.1×10−3). Myo-inositol and chiro-inositol degradation by the gut microbiome contributes to inositol deficiency [59], which includes metabolic disorders involved with insulin function [59] and MDD when myo-inositol is deficient in the prefrontal cortex [60].

View this table:
Table 2.

Metabolic pathways differential abundance in the mothers, based on comparisons of high-vs-low SD scores and high-vs-low PS scores. Machine learning algorithm can classify mothers as low and high SD 71.8% of the time (P = 8.6×10−4 compared to random assignment; FDR-corrected binomial distribution test). Average abundance and association (high or low) are shown for the pathways predictive value in the top-25 RF model (ranked by mean decrease in accuracy of the RF model; MDA), and with LEfSe effect size ≥ 1. The -log of the Kruskal-Wallis test P values from LEfSe is also shown.

Children are classified correctly as high- and low-SD in 85.9% of comparisons (P = 1.3×10−9) and as high- and low-PS in 72.1% of comparisons (P = 4.1×10−4) based on pathway abundance profiles. The top predictive pathways for both of these comparisons are shown in Table 3. Synthesis of L-glutamate and L-glutamine (PWY-5505) was the pathway most strongly associated with high-SD in the children (MDA = 7.9%, P = 7.9×10−7), and has been previously associated with obesity and visceral fat accumulation [61]. Here, L-glutamine biosynthesis was also associated with high-SD in both mothers and children. In supplementation studies of the gut microbiome, glutamine reduces the ratio of Firmicutes to Bacteroidetes and bacterial overgrowth or bacterial translocation, and increases the density of secretory immunoglobulin A (IgA) and IgA+ cells in the intestinal lumen [62].

View this table:
Table 3.

Metabolic pathways differential abundance in the children, based on comparisons of high-vs-low SD scores and high-vs-low Machine learning algorithm can correctly classify children as low and high mother SD 71.8% of the time (P = 8.6×10−4 compared to random assignment; FDR-corrected binomial distribution test). Average abundance and association (high or low) are shown for the pathways predictive value in the top-25 RF model (ranked by mean decrease in accuracy of the RF model; MDA), and with LEfSe effect size ≥ 1. The -log of the Kruskal-Wallis test P values from LEfSe is also shown.

Bacterial species discrimination between SD and PS groups based on maternal circulating cytokines

The relationship between inflammation and the GM is examined through maternal systemic cytokines IL-6, IL-8, IL-10 and TNF alpha, measured across trimesters. In mothers, prediction accuracy of the GM was greatest for third trimester cytokine samples, consistent with timing of maternal stool samples. Prediction accuracy was consistently greater for IL-6 than for the other inflammatory markers (Table 4). Discriminatory bacterial genomes demonstrate the most consistent and greatest predictive accuracy for IL-6, which in preclinical models is centrally important in altering fetal brain development in maternal immune activation models, where placental inflammatory signals are relayed to the fetal brain [63-65]. We identified a set of discriminatory taxa from maternal third trimester stools, whose relative abundance successfully classify mothers into low- or high-IL-6 75.6% of comparisons (P = 2.2×10−4; Supplementary Figure S3) vs. random assignment (FDR-corrected binomial distribution test), and in 72.1% of comparisons of metabolic pathways (P = 2.2×10−4). In this case, high prenatal IL-6 concentrations are associated with the lower abundance of anti-inflammatory Bacteroides species, as evidenced by the RF and LEfSe significance (effect size 2.3, P = 0.017 for Bacteroides A and effect size 2.9, P = 0.02 for Bacteroides faecis), as reported in other inflammatory states [47, 66]. These data illustrate the importance of community balance, including presence and absence, on target outcomes, and the importance in the ability to identify these at species/strain levels. Interestingly, we also identified discriminatory taxa in the infant 4-month GM profile (Figure 8). Of all inflammatory marker comparisons, microbial genomes in children had the best overall accuracy for maternal third trimester IL-6 concentrations (high/low) based on relative abundance (77.8% accuracy, P = 6.0×10−5) and metabolic pathways (75.6%, P = 6.9×10−4). However, maternal IL-6 concentration was not significantly correlated with SD or PS in this subsample (Supplementary Figure S4), although higher SD is associated with higher maternal IL-6 in the full sample set. Thus, the relationship of the maternal GM to IL-6 is driven by taxa distinct from those identified in composite SD or PS values, but still have proinflammatory characteristics.

View this table:
Table 4.

Random forest (RF) machine learning predictive accuracy for high-vs-low inflammatory markers based on GM taxonomic and pathway profiles. Prediction accuracy was consistently greater for IL-6 than for the other inflammatory markers. Bolded values correspond to comparisons for which P < 0.005 (after FDR correction).

View this table:
Table 5.

MetaCyc pathways differential abundance in the mothers and children, based on comparisons of high-vs-low IL-6 abundance. Based on MGS HUMAnN3 MetaCyc pathway abundance, RF can successfully classify mothers as low and high IL-6 73.3% of the time (P = 6.9×10−4 compared to random assignment; FDR-corrected binomial distribution test). Average abundance and association (high or low) are shown for the pathways predictive value in the top-25 RF model (ranked by mean decrease in accuracy of the RF model; MDA), and with LEfSe effect size ≥ 1. The -log of the Kruskal-Wallis test P values from LEfSe is also shown.

Figure 8:
Figure 8:

MGS genome differential abundance in children, based on comparisons of high-vs-low IL-6 abundance. Based on MGS genome profiles, Random Forest (RF) can successfully classify children into high and low IL-6 abundance 77.8% of the time (P = 6.0×10−5) compared to random assignment; FDR-corrected binomial distribution test). Taxonomy and relative abundance per sample for the genomes with the highest predictive value in the top-25 RF model are shown (ranked by mean decrease in accuracy of the RF model; MDA). Also displayed are -Log of the Kruskal-Wallis test P values from LEfSe (no value shown if the effect size was <2) and the overall abundance of the taxa when present.

CONCLUSIONS

Our prospectively assembled cohort of mother-infant dyads, enabled us to quantify for the first time the impact of exposure to both, SD and PS on GM structure and function for mothers and their infants. Mothers and infants classified as “high’ (case) or “low” (control) SD/PS, show distinct discriminatory taxonomic, metabolic and inflammatory features that accurately ‘predict’ maternal prenatal SD and PS status over 80% of the time, with SD having greater predictive accuracy than PS. Mothers with high-SD/high-PS have highly variable microbiomes compared to low-SD/low PS mothers, reflecting greater permutations from environmental influences. The distinct nature of the taxonomic and functional GM predictors for SD compared to PS, indicate different underlying mechanisms driving the relationship.

The human GM modulates inflammatory cytokine production [8, 67] and has been linked to chronic inflammatory disorders [6]. We identified a significant relationship between the maternal prenatal and infant GM, and prenatal circulating cytokine concentrations in mothers. Prediction accuracy is highest and most consistent for IL-6 suggesting a contributing role to subtle chronic inflammation, although the discriminating taxa are distinct from SD. Elevated IL-6 concentrations have been linked to specific GM profiles in disease states among adults [68, 69] and recently to neuropathology when elevated during pregnancy in animal models [64, 65].

Our findings should be viewed within the strengths and limitations of the study. The finding are associations, and do not necessarily imply causality. Identification of strain level taxa and metabolic pathways through WMG, however, form the basis for testing mechanistic causation in preclinical models. Use of mother-infant dyads allowed interrogation of SD and PS between mother and infant GM, and examination of genomes and metabolic pathways. The population we studied is from a circumscribed region but contains a broad range of socio-economic backgrounds. We did not examine the role of race, because of the collinearity of race and SD, with no additional contribution of race in the model beyond that found with SD alone.

While longitudinal studies are needed to determine the stability of these GM signatures from early life to early childhood in children, the results identify unique features of the maternal and infant GMs and host response. Such findings should be pursued to better understand the effect of socioeconomic status and mental health determinants on GM health and stability. More critically, information on causal pathways triggered or sustained by the GM that affect child health and development could lead to new biomarkers and interventions. The potential malleability of the GM leaves room for optimism that unfavorable neurodevelopment outcomes might not be inevitable in children with living with high SD and PS values.

METHODS

Study Design and Cohort

The study population Consists of 121 mother-child dyads drawn from a larger parent study of 399 dyads, Early Life Adversity Biological Embedding and Risk for Developmental Precursors of Mental Disorders (eLABE; details provided in [29]). The eLABE study used a prospective observational design to examine the impact of pre- and post-natal psychological and social factors on infant neurodevelopment. Preganant women (N=395) and their offspring were recruited from the March of Dimes Prematurity Research Center at Washington University in St. Louis between 2017-2020, with delivery of singleton births at the Barnes Jewish Hospital in St. Louis [70]. This sub-cohort was chosen from the extremes of maternal social disadvantage and psychosocial stress, among women with relatively healthy offspring, who donated third trimester stools. Exclusions included multiple gestation, congenital malformations and infections, premature birth (< 37 weeks gestational age), maternal alcohol or drug use during pregnancy (excluding tobacco, marijuana, and maternal enteral steroid use). Race and ethnicity were based on maternal self reporting extracted from the medical record. Options included American Indian/Alaskan Native, Asian, Black or African American, Native Hawaiian/Pacific Islander, White, unknown, or other (free text) for race, and the following options for ethnicity: Hispanic/Latina, non-Hispanic/Latina, or unknown/not applicable for ethnicity. All procedures were approved by the Human Research Protection Office, informed consent was obtained from the mother for each dyad. The study was performed in accordance with Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [71].

Maternal measures

At each trimester of pregnancy, measures of maternal depression, experiences of stress, as well as demographic information including insurance, education, address, and household composition were obtained from participants or extracted from the medical record by trained staff. Components of the two latent constructs, maternal social disadvantage and maternal psychosocial stress are previously described [29]. Briefly, the components of each latent variable included:

Maternal Social Disadvantage

Insurance status obtained was verified at the third trimseter from the medical record and maternal self reporting, Income to Needs ratio in each trimseter based on family income and houshold size (1.0 being the poverty line for the U.S.) was self reported, highest maternal educational level was self reported, the national Area Deprivation Index is a national multidimensional geotracking method based on census block data, providing percentile rankings of neighborhood disadvantage status [72], and maternal nutrition over the past year was categorized using the validated Healthy Eating Index (using National Cancer Institute. The Healthy Eating Index – Population Ratio Method. Updated December 14, 2021; [73]) obtained using the Diet History Questionnaire (DHQII).

Maternal Psychosocial Stress

In each trimester mothers completed the Edinburgh Postnatal Depression Scale (EPDS) [74], Perceived Stress scale (PSS) [75] at each trimester, averaged over trimesters a one-time lifetime STRAIN survey [76], a comprehensive measure of lifetime stressful and traumatic life events. Experiences of discrimination based on race were assessed using the Everyday Discrimination Scale [77].

Maternal medical risks

were defined by the Maternal Medical Risk score in pregnancy, a validated measure of maternal co-morbidities weighted by severity in pregnancy, and maternal pre-pregnancy Body Mass Index was extracted from the medical record.

Infant measures

Gestational age was determined by the best obstetric estimate using last menstrual period or earliest ultrasound dating. Birthweight and route of delivery were extracted from the electronic medical record delivery note. Breastfeeding data was collected by parental reporting at the time of home stool sample collection and based on the Center for Disease Control Infant Feeding Practices II study food frequency checklist data [78, 79].

Biological specimen collection and processing

Maternal blood samples for serum were collected from venous draws during routine clinic visits across each trimester, processed within 12 hours of collection, and stored at -80°C in 1 mL aliquots (details in [70]). Stools from mothers and infants were collected from home and processed as previously described [79]. Briefly maternal samples were collected during the third trimester, and infant stools were retrieved directly from the diaper and scooped into barcoded tubes. Stool samples were placed in insulated packaging (VWR) with U-tex gel packs (Fischer Scientific) and placed in home freezers. A community-based courier system available 24 hours per day was used to retrieve samples within 90 minutes. Weekend or overnight samples were taken to the courier office freezers (−20°C) and then delivered to the laboratory freezer (−80°C) during working hours; weekday stools were couriered directly to the laboratory. Previous testing showed no quality difference between temperatures used in delivery methods [80]. DNA was extracted from stool using the Qiagen (Hilden, Germany) QIAamp Power Fecal Pro DNA Mini Kit (50) and the automated QIAcube platform (Qiagen) as previously described [79, 81].

Targeted GM profiling using V4-16S rRNA sequencing, data processing and analyses

DNA extracted from stool was sequenced on an Illumina MiSeq, producing 2×250bp paired-end reads spanning the V4 hypervariable region, for 242 samples (121 samples from mothers and 121 from their matched children). Data were imported into QIIME2 [82], using standard methods and the developer’s docker container (qiime2/core:2018.8). V4 region amplicons were assembled and denoised using the Qiime2 method ‘DADA2 denoise-paired’. Processed V4 amplicons were grouped into amplicon sequence variants (ASVs) with 100% sequence similarity. ASVs were classified using a pre-trained classifier based on SILVA (release 132) [83], a comprehensive database that provides accurate annotations [84]. ASV counts per sample were exported as biom files from a qiime2 artifact and converted into a human readable tsv file using “biom convert”. Read counts per sample were rarefied to 11,929 reads per sample (the lowest count among the 242 samples) using the “rrarefy” command in the R package “vegan” (version 2.5-7, https://CRAN.R-project.org/package=vegan), and normalized read counts were calculated per sample by dividing the number of reads associated with each ASV by the total number of reads assigned across ASVs. Taxonomic identifications used are directly provided by SILVA [83]. Raw 16S rRNA can be downloaded from public database (submission in progress).

Whole Metagenome Shotgun (WMS) sequencing and GM profiling

Whole Metagenome Shotgun (WMS) datasets for 178 samples (89 samples from mothers and 89 from their respective children) were generated on the Illumina NovaSeq S4. For each sample ∼ 6Gbp was generated. The reads for all 178 samples were cleaned of barcodes, adapters and low-quality ends using Trimmomatic [85] (version 0.36). The BMTagger program (installed using conda, July 30th, 2020) was used to identify human contaminant reads using the human reference genome (GRCh38.98 [86]). Reads identified as human were removed to produce final paired-end fastq per each of the 178 samples. Raw WMS read data can be downloaded from public database (submission in progress).

The 178 WMS samples were mapped against the Unified Human Gastrointestinal Genome (UHGG) collection, comprising 204,938 nonredundant genomes from 4,644 gut prokaryotes, each theoretically representing an individual bacterial or archaeal species (95% average nucleotide identity [32]) using bowtie2 [39] (v2.3.5.1). The profile module of the inStrain [33] program was then run to generate sequencing breadth and depth of coverage statistics for every genome, in addition to nucleotide diversity measures per genome per sample. The depth of coverage values were normalized within every sample by dividing each genome’s depth by the sum of the depths across all genomes.

The 178 WMS samples were also used as input for HUMAnN [37] (version 3), which was ran from the biobakery/humann docker container (latest version as of October 2020) using the Chocophlan nucleotide database and Uniref90 [87] protein database. HUMAnN3 runs the MetaPhlAn [38] program as an intermediate step to assign organism-specific functional profiling, and we used the developer-provided Metaphlan3 [38] bowtie2 [39] database for this intermediate step. The HUMAnN3 pipeline was used to generate MetaCyc [36] pathway abundance per sample. The “humann_renorm_table” script (included in the HUMAnN3 distribution) was used to convert Reads Per Kilobase (RPK) values in the MetaCyc abundance table to a normalized value, Copies Per Million (CPM), which can be compared across samples.

Statistical analysis

Significant differences between Low and High SD and PS sample sets based on metadata classifications (Table 1, Supplementary Table S1) were calculated using two-tailed T-tests with unequal variance for continuous variables and using Fischer exact tests for categorical data. Correction for multiple testing was not performed across tests, because significant differences indicate potential bias in the data. Significant differences in components of SD and PS were expected since samples were chosen from the extremes of phenotype for SD and PS to improve detection of GM differences between groups [29].

For the 16S/ASV sample analysis, Shannon index diversity values were calculated for each sample using the normalized read counts across all taxa using the “diversity” function in the “vegan” library in R, and Bray-Curtis distance diversity values were calculated using the “vegdist” function [88]. Correlation R2 values and Pearson r values were calculated using MS Excel, and the significance of the correlation was tested using the two-tailed t statistic with degrees of freedom N -1. ASV-based sample clustering was performed using the relative abundance profiles of all ASVs across all samples as input for Bray-Curtis dissimilarity-based clustering (complete linkage) using the “hclust” function in R, with additional metadata visualized using MS Excel. Significant differences in SD and PS between clusters were identified using ANOVA with a Post Hoc Tukey HSD test.

For the WMG sequence analysis, samples were divided into “high” and “low” SD and PS based on the distribution of these values across the sample set. Samples above the average value + 0.5 standard deviations were considered “high” and samples below the average value – 0.5 standard deviations were considered “low” (35 “low-SD”, 43 “high-SD, 36 “low-Psych” and 32 “high-Psych”; Figure 1C). The same approach was used to separate samples into “high” and “low” sample sets based on inflammatory marker data (IL-6, IL-8, IL-10 and TNFα).

To identify bacterial taxa that strongly predict mothers’ SD and PS scores, we analyzed taxonomic and pathway GM profiles using two approaches. First, a supervised machine-learning approach (Random Forest [40]) that identifies non-linear relationships from high dimensional and dependent data [40] and two-round approach using only the best 25 predictors (as previously described for gut microbiome associations with mental health measures [41]) was used to (i) quantify the ability to predict metadata classification based on the microbiome profiles, indicative of the overall association between the microbiome and the composite scores, and (ii) for each comparison, identify the specific genomes and pathways that most strongly differentiate between the high and low SD and PS scores. The generalization error of the model was evaluated by out of bag (OOB) error. The association of the metadata with the microbiome was quantified using the RF classification accuracy, and the significance of the accuracy was measured using FDR-corrected binomial distribution tests. RF model accuracy was also examined using receiver operating characteristic (ROC) curves, quantified using the area under the curve (AUC). Significance values for the ROC curves were assigned by Mann-Whitney U statistics [89], using the “roc.area” function in the R library “ROCR”.

Second, linear discriminant analysis effect size (LEfSe [42]), the most frequently used statistical tool to determine significant differences in microbiome member abundance [90], was used for differential genome abundance testing (default settings at an adjusted P ≤ 0.05 for significance) for the non-parametric factorial Kruskal-Wallis (KW) sum-rank test, and requiring a linear discriminant analysis (LDA) “effect size” of at least 2 in order to identify differentially abundant taxa). The same approach was used for the pathway analysis, but the “effect size” test cutoff applied was reduced to a value of 1 instead of 2, since the effect size is designed for the more sparse nature of metagenomic abundance data [42]. However, the same MDA and KW cutoffs were applied for pathway analysis.

Data Availability

All data produced in the present work are contained in the manuscript.

Author Contributions

Conceived and designed the study: BBW, MM JL, DB, CER CDS

Assembled the cohort, collected specimens: IMN, LL, SKE

Biological and clinical data database maintenance: IMN, REB, JDW, CH-M

Sample preparation and extraction of DNA: IMN, REB, JDW, CH-M

Performed data analysis: BAR, MM, JM, PJM, GEM,EC

Interpreted the data: BAR, MM, BBW, JM PPT, IMN

Wrote the paper: BAR, MM, BBW

All authors approve the manuscript for publication.

Competing Interest

The authors have no competing interests to disclose.

Additional information

We thank the families involved in the study, the MARCH of Dime Premature Research Center at Washington University, and the GTAC@MGI for the sequence data generation.

Supplementary Figures

Supplementary Figure S1:
Supplementary Figure S1:

The relative abundance of Bifidobacterium species identified in all mothers and children in the high-SD vs low-SD comparisons.

Supplementary Figure S2:
Supplementary Figure S2:

The relative abundance of Blautia species identified in all mothers and children in the high-SD vs low-SD comparisons.

Supplementary Figure S3:
Supplementary Figure S3:

Based on WMS genome profiles, RF can successfully classify mothers as low and high IL-6 abundance 75.6% of the time (P = 2.2×10−4 compared to random assignment; FDR-corrected binomial distribution test). Taxonomy and relative abundance per sample for the genomes with the greatest predictive value in the top-25 RF model are shown (ranked by mean decrease in accuracy of the RF model; MDA). Also displayed are -Log of the Kruskal-Wallis test P values from LEfSe (no value shown if the effect size was <2) and the overall abundance of the taxa when present.

Supplementary Figure S4:
Supplementary Figure S4:

Comparison of Social Disadvantage scores and Psychological Stressors scores to maternal third trimester inflammatory marker serum concentrations. Only IL-8 and PS scores correlated significantly.

Supplementary Tables

Supplementary Table S1: Complete patient characteristics at study entry for each of the primary comparisons of interest. “Low” and “High” Disadvantage and Psych scores are separated according to the distribution of the metadata, as shown in Figure 1.

Supplementary Table S2: Database of sample data, including metadata, indications of sample groups for each comparison, relative 16S ASV abundance, relative MGS genome abundance and relative MGS pathway abundance values.

Supplementary Table S3: Full ASV sequences for each unique ASV identifier from Table S2.

Supplementary Table S4: Differential abundance statistics for each UHGG genome in each comparison.

Supplementary Table S5: Differential abundance statistics for each MetaCyc pathway in each comparison.

Acknowledgments

Funding, Washington University in St. Louis’ GTAC@MGI for library construction and sequencing, NIMH grant RO1 MH113883, (BBW, BAR, IMN, PPT, PJM, SKE, JLL, CER, TAS, CDS, DMB, GEM, EC, JM, MM) March of Dimes (SKE), Children’s Discovery Institute II MD-II-2015-489 (BBW, IMN), Biobank Core P30DK052574 (PIT).

Footnotes

  • Email addresses Barbara B Warner: warnerbb{at}wustl.edu, Bruce A. Rosa: barosa{at}wustl.edu, I. Malick Ndao: ndao_m{at}wustl.edu, Phillip I. Tarr: tarr{at}wustl.edu, J. Philip Miller: jphilipmiller{at}wustl.edu, Sara K. England: englands{at}wustl.edu, Joan Luby: lubyj{at}wustl.edu, Cynthia Rogers: rogersc{at}wustl.edu, Carla Hall-Moore: cmhall{at}wustl.edu, Renay E. Bryant: renay{at}wustl.edu, Jacqueline D. Wang: jackie.wang{at}wustl.edu, Laura Linneman: llinneman{at}wustl.edu, Tara A. Smyser: smysert{at}wustl.edu, Christopher D. Smyser: smyserc{at}wustl.edu, Deanna M. Barch: dbarch{at}wustl.edu, Gregory E. Miller: greg.miller{at}northwestern.edu, Edith Chen: edith.chen{at}northwestern.edu, John Martin: jmartin{at}wustl.edu, Makedonka Mitreva: mmitreva{at}wustl.edu

References

  1. 1.↵
    The L. Taking urgent action on health inequities. Lancet. 2020;395(10225):659. doi: 10.1016/S0140-6736(20)30455-4. PubMed PMID: 32113488.
    OpenUrlCrossRefPubMed
  2. 2.↵
    Woolf SH, Schoomaker H. Life Expectancy and Mortality Rates in the United States, 1959-2017. JAMA. 2019;322(20):1996–2016. doi: 10.1001/jama.2019.16932. PubMed PMID: 31769830; PMCID: PMC7146991.
    OpenUrlCrossRefPubMed
  3. 3.↵
    Barker DJ, Osmond C. Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet. 1986;1(8489):1077–81. doi: 10.1016/s0140-6736(86)91340-1. PubMed PMID: 2871345.
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.
    Nobile S, Di Sipio Morgia C, Vento G. Perinatal Origins of Adult Disease and Opportunities for Health Promotion: A Narrative Review. J Pers Med. 2022;12(2). Epub 20220125. doi: 10.3390/jpm12020157. PubMed PMID: 35207646; PMCID: PMC8877993.
    OpenUrlCrossRefPubMed
  5. 5.↵
    Simeoni U, Armengaud JB, Siddeek B, Tolsa JF. Perinatal Origins of Adult Disease. Neonatology. 2018;113(4):393–9. Epub 20180531. doi: 10.1159/000487618. PubMed PMID: 29852488.
    OpenUrlCrossRefPubMed
  6. 6.↵
    Furman D, Campisi J, Verdin E, Carrera-Bastos P, Targ S, Franceschi C, Ferrucci L, Gilroy DW, Fasano A, Miller GW, Miller AH, Mantovani A, Weyand CM, Barzilai N, Goronzy JJ, Rando TA, Effros RB, Lucia A, Kleinstreuer N, Slavich GM. Chronic inflammation in the etiology of disease across the life span. Nat Med. 2019;25(12):1822–32. Epub 20191205. doi: 10.1038/s41591-019-0675-0. PubMed PMID: 31806905; PMCID: PMC7147972.
    OpenUrlCrossRefPubMed
  7. 7.↵
    Fan Y, Pedersen O. Gut microbiota in human metabolic health and disease. Nat Rev Microbiol. 2021;19(1):55–71. Epub 20200904. doi: 10.1038/s41579-020-0433-9. PubMed PMID: 32887946.
    OpenUrlCrossRefPubMed
  8. 8.↵
    Olin A, Henckel E, Chen Y, Lakshmikanth T, Pou C, Mikes J, Gustafsson A, Bernhardsson AK, Zhang C, Bohlin K, Brodin P. Stereotypic Immune System Development in Newborn Children. Cell. 2018;174(5):1277–92 e14. doi: 10.1016/j.cell.2018.06.045. PubMed PMID: 30142345; PMCID: PMC6108833.
    OpenUrlCrossRefPubMed
  9. 9.↵
    Penders J, Thijs C, van den Brandt PA, Kummeling I, Snijders B, Stelma F, Adams H, van Ree R, Stobberingh EE. Gut microbiota composition and development of atopic manifestations in infancy: the KOALA Birth Cohort Study. Gut. 2007;56(5):661–7. Epub 20061017. doi: 10.1136/gut.2006.100164. PubMed PMID: 17047098; PMCID: PMC1942165.
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    Mbakwa CA, Hermes GDA, Penders J, Savelkoul PHM, Thijs C, Dagnelie PC, Mommers M, Zoetendal EG, Smidt H, Arts ICW. Gut Microbiota and Body Weight in School-Aged Children: The KOALA Birth Cohort Study. Obesity (Silver Spring). 2018;26(11):1767–76. Epub 20181008. doi: 10.1002/oby.22320. PubMed PMID: 30296366; PMCID: PMC6646907.
    OpenUrlCrossRefPubMed
  11. 11.↵
    Hou K, Wu ZX, Chen XY, Wang JQ, Zhang D, Xiao C, Zhu D, Koya JB, Wei L, Li J, Chen ZS. Microbiota in health and diseases. Signal Transduct Target Ther. 2022;7(1):135. Epub 20220423. doi: 10.1038/s41392-022-00974-4. PubMed PMID: 35461318; PMCID: PMC9034083.
    OpenUrlCrossRefPubMed
  12. 12.↵
    Karl JP, Hatch AM, Arcidiacono SM, Pearce SC, Pantoja-Feliciano IG, Doherty LA, Soares JW. Effects of Psychological, Environmental and Physical Stressors on the Gut Microbiota. Front Microbiol. 2018;9:2013. Epub 20180911. doi: 10.3389/fmicb.2018.02013. PubMed PMID: 30258412; PMCID: PMC6143810.
    OpenUrlCrossRefPubMed
  13. 13.↵
    Amato KR, Arrieta MC, Azad MB, Bailey MT, Broussard JL, Bruggeling CE, Claud EC, Costello EK, Davenport ER, Dutilh BE, Swain Ewald HA, Ewald P, Hanlon EC, Julion W, Keshavarzian A, Maurice CF, Miller GE, Preidis GA, Segurel L, Singer B, Subramanian S, Zhao L, Kuzawa CW. The human gut microbiome and health inequities. Proc Natl Acad Sci U S A. 2021;118(25). doi: 10.1073/pnas.2017947118. PubMed PMID: 34161260; PMCID: PMC8237592.
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    Herd P, Palloni A, Rey F, Dowd JB. Social and population health science approaches to understand the human microbiome. Nat Hum Behav. 2018;2(11):808–15. Epub 20181022. doi: 10.1038/s41562-018-0452-y. PubMed PMID: 31457107; PMCID: PMC6711373.
    OpenUrlCrossRefPubMed
  15. 15.↵
    Bowyer RCE, Jackson MA, Le Roy CI, Ni Lochlainn M, Spector TD, Dowd JB, Steves CJ. Socioeconomic Status and the Gut Microbiome: A TwinsUK Cohort Study. Microorganisms. 2019;7(1). Epub 20190111. doi: 10.3390/microorganisms7010017. PubMed PMID: 30641975; PMCID: PMC6351927.
    OpenUrlCrossRefPubMed
  16. 16.
    Miller GE, Engen PA, Gillevet PM, Shaikh M, Sikaroodi M, Forsyth CB, Mutlu E, Keshavarzian A. Lower Neighborhood Socioeconomic Status Associated with Reduced Diversity of the Colonic Microbiota in Healthy Adults. PLoS One. 2016;11(2):e0148952. Epub 20160209. doi: 10.1371/journal.pone.0148952. PubMed PMID: 26859894; PMCID: PMC4747579.
    OpenUrlCrossRefPubMed
  17. 17.↵
    He Y, Wu W, Wu S, Zheng HM, Li P, Sheng HF, Chen MX, Chen ZH, Ji GY, Zheng ZD, Mujagond P, Chen XJ, Rong ZH, Chen P, Lyu LY, Wang X, Xu JB, Wu CB, Yu N, Xu YJ, Yin J, Raes J, Ma WJ, Zhou HW. Linking gut microbiota, metabolic syndrome and economic status based on a population-level analysis. Microbiome. 2018;6(1):172. Epub 20180924. doi: 10.1186/s40168-018-0557-6. PubMed PMID: 30249275; PMCID: PMC6154942.
    OpenUrlCrossRefPubMed
  18. 18.↵
    Zijlmans MA, Korpela K, Riksen-Walraven JM, de Vos WM, de Weerth C. Maternal prenatal stress is associated with the infant intestinal microbiota. Psychoneuroendocrinology. 2015;53:233–45. Epub 20150119. doi: 10.1016/j.psyneuen.2015.01.006. PubMed PMID: 25638481.
    OpenUrlCrossRefPubMed
  19. 19.
    Aatsinki AK, Keskitalo A, Laitinen V, Munukka E, Uusitupa HM, Lahti L, Kortesluoma S, Mustonen P, Rodrigues AJ, Coimbra B, Huovinen P, Karlsson H, Karlsson L. Maternal prenatal psychological distress and hair cortisol levels associate with infant fecal microbiota composition at 2.5 months of age. Psychoneuroendocrinology. 2020;119:104754. Epub 20200605. doi: 10.1016/j.psyneuen.2020.104754. PubMed PMID: 32531627.
    OpenUrlCrossRefPubMed
  20. 20.
    Kang LJ, Vu KN, Koleva PT, Field CJ, Chow A, Azad MB, Becker AB, Mandhane PJ, Moraes TJ, Sears MR, Lefebvre DL, Turvey SE, Subbarao P, Lou WYW, Scott JA, Kozyrskyj AL. Maternal psychological distress before birth influences gut immunity in mid-infancy. Clin Exp Allergy. 2020;50(2):178–88. Epub 20200115. doi: 10.1111/cea.13551. PubMed PMID: 31845414.
    OpenUrlCrossRefPubMed
  21. 21.
    Hantsoo L, Jasarevic E, Criniti S, McGeehan B, Tanes C, Sammel MD, Elovitz MA, Compher C, Wu G, Epperson CN. Childhood adversity impact on gut microbiota and inflammatory response to stress during pregnancy. Brain Behav Immun. 2019;75:240–50. Epub 20181103. doi: 10.1016/j.bbi.2018.11.005. PubMed PMID: 30399404; PMCID: PMC6349044.
    OpenUrlCrossRefPubMed
  22. 22.↵
    Hechler C, Borewicz K, Beijers R, Saccenti E, Riksen-Walraven M, Smidt H, de Weerth C. Association between Psychosocial Stress and Fecal Microbiota in Pregnant Women. Sci Rep. 2019;9(1):4463. Epub 20190314. doi: 10.1038/s41598-019-40434-8. PubMed PMID: 30872645; PMCID: PMC6418257.
    OpenUrlCrossRefPubMed
  23. 23.↵
    Levin AM, Sitarik AR, Havstad SL, Fujimura KE, Wegienka G, Cassidy-Bushrow AE, Kim H, Zoratti EM, Lukacs NW, Boushey HA, Ownby DR, Lynch SV, Johnson CC. Joint effects of pregnancy, sociocultural, and environmental factors on early life gut microbiome structure and diversity. Sci Rep. 2016;6:31775. Epub 20160825. doi: 10.1038/srep31775. PubMed PMID: 27558272; PMCID: PMC4997337.
    OpenUrlCrossRefPubMed
  24. 24.
    Jahnke JR, Roach J, Azcarate-Peril MA, Thompson AL. Maternal precarity and HPA axis functioning shape infant gut microbiota and HPA axis development in humans. PLoS One. 2021;16(5):e0251782. Epub 20210520. doi: 10.1371/journal.pone.0251782. PubMed PMID: 34015045; PMCID: PMC8136730.
    OpenUrlCrossRefPubMed
  25. 25.↵
    Lewis CR, Bonham KS, McCann SH, Volpe AR, D’Sa V, Naymik M, De Both MD, Huentelman MJ, Lemery-Chalfant K, Highlander SK, Deoni SCL, Klepac-Ceraj V. Family SES Is Associated with the Gut Microbiome in Infants and Children. Microorganisms. 2021;9(8). Epub 20210728. doi: 10.3390/microorganisms9081608. PubMed PMID: 34442687; PMCID: PMC8398307.
    OpenUrlCrossRefPubMed
  26. 26.↵
    Jain N. The early life education of the immune system: Moms, microbes and (missed) opportunities. Gut Microbes. 2020;12(1):1824564. Epub 20201012. doi: 10.1080/19490976.2020.1824564. PubMed PMID: 33043833; PMCID: PMC7781677.
    OpenUrlCrossRefPubMed
  27. 27.
    Cowan CSM, Dinan TG, Cryan JF. Annual Research Review: Critical windows -the microbiota-gut-brain axis in neurocognitive development. J Child Psychol Psychiatry. 2020;61(3):353–71. Epub 20191126. doi: 10.1111/jcpp.13156. PubMed PMID: 31773737.
    OpenUrlCrossRefPubMed
  28. 28.↵
    Laursen MF, Bahl MI, Licht TR. Settlers of our inner surface - factors shaping the gut microbiota from birth to toddlerhood. FEMS Microbiol Rev. 2021;45(4). doi: 10.1093/femsre/fuab001. PubMed PMID: 33428723; PMCID: PMC8371275.
    OpenUrlCrossRefPubMed
  29. 29.↵
    Luby JL, Barch DM, Warner B, Rogers C, Smyser CD, Triplett R, Arora J, Smyser T, England SK, Slavich G, Stout M, Miller P. Modeling Prenatal Adversity/Advantage: Effects on Birth Weight. MedRxiv. 2021. doi: doi: https://doi.org/10.1101/2021.12.16.21267938.
  30. 30.↵
    Ridley M, Rao G, Schilbach F, Patel V. Poverty, depression, and anxiety: Causal evidence and mechanisms. Science. 2020;370(6522). Epub 2020/12/12. doi: 10.1126/science.aay0214. PubMed PMID: 33303583.
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    Ma J, Li Z, Zhang W, Zhang C, Zhang Y, Mei H, Zhuo N, Wang H, Wang L, Wu D. Comparison of gut microbiota in exclusively breast-fed and formula-fed babies: a study of 91 term infants. Sci Rep. 2020;10(1):15792. Epub 2020/09/27. doi: 10.1038/s41598-020-72635-x. PubMed PMID: 32978424; PMCID: PMC7519658.
    OpenUrlCrossRefPubMed
  32. 32.↵
    Almeida A, Nayfach S, Boland M, Strozzi F, Beracochea M, Shi ZJ, Pollard KS, Sakharova E, Parks DH, Hugenholtz P, Segata N, Kyrpides NC, Finn RD. A unified catalog of 204,938 reference genomes from the human gut microbiome. Nat Biotechnol. 2021;39(1):105–14. Epub 2020/07/22. doi: 10.1038/s41587-020-0603-3. PubMed PMID: 32690973; PMCID: PMC7801254.
    OpenUrlCrossRefPubMed
  33. 33.↵
    Olm MR, Crits-Christoph A, Bouma-Gregson K, Firek BA, Morowitz MJ, Banfield JF. inStrain profiles population microdiversity from metagenomic data and sensitively detects shared microbial strains. Nat Biotechnol. 2021;39(6):727–36. Epub 2021/01/20. doi: 10.1038/s41587-020-00797-0. PubMed PMID: 33462508.
    OpenUrlCrossRefPubMed
  34. 34.↵
    Wang S, Zeng S, Egan M, Cherry P, Strain C, Morais E, Boyaval P, Ryan CA E MD, Ross RP, Stanton C. Metagenomic analysis of mother-infant gut microbiome reveals global distinct and shared microbial signatures. Gut Microbes. 2021;13(1):1–24. doi: 10.1080/19490976.2021.1911571. PubMed PMID: 33960282; PMCID: PMC8115609.
    OpenUrlCrossRefPubMed
  35. 35.↵
    Truong DT, Franzosa EA, Tickle TL, Scholz M, Weingart G, Pasolli E, Tett A, Huttenhower C, Segata N. MetaPhlAn2 for enhanced metagenomic taxonomic profiling. Nat Methods. 2015;12(10):902–3. doi: 10.1038/nmeth.3589. PubMed PMID: 26418763.
    OpenUrlCrossRefPubMed
  36. 36.↵
    Caspi R, Billington R, Keseler IM, Kothari A, Krummenacker M, Midford PE, Ong WK, Paley S, Subhraveti P, Karp PD. The MetaCyc database of metabolic pathways and enzymes -a 2019 update. Nucleic Acids Res. 2020;48(D1):D445–D53. Epub 2019/10/06. doi: 10.1093/nar/gkz862. PubMed PMID: 31586394; PMCID: PMC6943030.
    OpenUrlCrossRefPubMed
  37. 37.↵
    Franzosa EA, McIver LJ, Rahnavard G, Thompson LR, Schirmer M, Weingart G, Lipson KS, Knight R, Caporaso JG, Segata N, Huttenhower C. Species-level functional profiling of metagenomes and metatranscriptomes. Nat Methods. 2018;15(11):962–8. Epub 2018/11/01. doi: 10.1038/s41592-018-0176-y. PubMed PMID: 30377376; PMCID: PMC6235447.
    OpenUrlCrossRefPubMed
  38. 38.↵
    Beghini F, McIver LJ, Blanco-Miguez A, Dubois L, Asnicar F, Maharjan S, Mailyan A, Manghi P, Scholz M, Thomas AM, Valles-Colomer M, Weingart G, Zhang Y, Zolfo M, Huttenhower C, Franzosa EA, Segata N. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. Elife. 2021;10. Epub 2021/05/05. doi: 10.7554/eLife.65088. PubMed PMID: 33944776; PMCID: PMC8096432.
    OpenUrlCrossRefPubMed
  39. 39.↵
    Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9. Epub 2012/03/06. doi: 10.1038/nmeth.1923. PubMed PMID: 22388286; PMCID: PMC3322381.
    OpenUrlCrossRefPubMedWeb of Science
  40. 40.↵
    Knights D, Costello EK, Knight R. Supervised classification of human microbiota. FEMS Microbiol Rev. 2011;35(2):343–59. Epub 2010/11/03. doi: 10.1111/j.1574-6976.2010.00251.x. PubMed PMID: 21039646.
    OpenUrlCrossRefPubMedWeb of Science
  41. 41.↵
    Malan-Muller S, Valles-Colomer M, Foxx CL, Vieira-Silva S, van den Heuvel LL, Raes J, Seedat S, Lowry CA, Hemmings SMJ. Exploring the relationship between the gut microbiome and mental health outcomes in a posttraumatic stress disorder cohort relative to trauma-exposed controls. Eur Neuropsychopharmacol. 2022;56:24–38. Epub 20211216. doi: 10.1016/j.euroneuro.2021.11.009. PubMed PMID: 34923209.
    OpenUrlCrossRefPubMed
  42. 42.↵
    Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, Huttenhower C. Metagenomic biomarker discovery and explanation. Genome Biol. 2011;12(6):2011–12.
    OpenUrl
  43. 43.↵
    Sakamoto M, Iino T, Yuki M, Ohkuma M. Lawsonibacter asaccharolyticus gen. nov., sp. nov., a butyrate-producing bacterium isolated from human faeces. Int J Syst Evol Microbiol. 2018;68(6):2074–81. Epub 20180510. doi: 10.1099/ijsem.0.002800. PubMed PMID: 29745868.
    OpenUrlCrossRefPubMed
  44. 44.↵
    Sakamoto M, Ikeyama N, Yuki M, Ohkuma M. Draft Genome Sequence of Lawsonibacter asaccharolyticus JCM 32166(T), a Butyrate-Producing Bacterium, Isolated from Human Feces. Genome Announc. 2018;6(25). Epub 20180621. doi: 10.1128/genomeA.00563-18. PubMed PMID: 29930067; PMCID: PMC6013597.
    OpenUrlAbstract/FREE Full Text
  45. 45.↵
    Canani RB, Costanzo MD, Leone L, Pedata M, Meli R, Calignano A. Potential beneficial effects of butyrate in intestinal and extraintestinal diseases. World J Gastroenterol. 2011;17(12):1519–28. doi: 10.3748/wjg.v17.i12.151910.3748/wjg.v17.i12.1519. PubMed PMID: 21472114; PMCID: PMC3070119.
    OpenUrlCrossRefPubMedWeb of Science
  46. 46.↵
    Uriot O, Denis S, Junjua M, Roussel Y, Dary-Mourot A, Blanquet-Diot S. Streptococcus thermophilus: From yogurt starter to a new promising probiotic candidate? Journal of Functional Foods. 2017;37:74–89. doi: https://doi.org/10.1016/j.jff.2017.07.038.
    OpenUrl
  47. 47.↵
    Arboleya S, Watkins C, Stanton C, Ross RP. Gut Bifidobacteria Populations in Human Health and Aging. Front Microbiol. 2016;7:1204. Epub 2016/09/07. doi: 10.3389/fmicb.2016.01204. PubMed PMID: 27594848; PMCID: PMC4990546.
    OpenUrlCrossRefPubMed
  48. 48.↵
    Gueimonde M, Ouwehand A, Huhtinen H, Salminen E, Salminen S. Qualitative and quantitative analyses of the bifidobacterial microbiota in the colonic mucosa of patients with colorectal cancer, diverticulitis and inflammatory bowel disease. World J Gastroenterol. 2007;13(29):3985–9. Epub 2007/08/01. doi: 10.3748/wjg.v13.i29.3985. PubMed PMID: 17663515; PMCID: PMC4171173.
    OpenUrlCrossRefPubMed
  49. 49.↵
    Knudsen JK, Bundgaard-Nielsen C, Hjerrild S, Nielsen RE, Leutscher P, Sorensen S. Gut microbiota variations in patients diagnosed with major depressive disorder-A systematic review. Brain Behav. 2021;11(7):e02177. Epub 20210528. doi: 10.1002/brb3.2177. PubMed PMID: 34047485; PMCID: PMC8323045.
    OpenUrlCrossRefPubMed
  50. 50.↵
    Bailey MT, Dowd SE, Galley JD, Hufnagle AR, Allen RG, Lyte M. Exposure to a social stressor alters the structure of the intestinal microbiota: implications for stressor-induced immunomodulation. Brain Behav Immun. 2011;25(3):397–407. Epub 20101030. doi: 10.1016/j.bbi.2010.10.023. PubMed PMID: 21040780; PMCID: PMC3039072.
    OpenUrlCrossRefPubMed
  51. 51.↵
    Jiang H, Ling Z, Zhang Y, Mao H, Ma Z, Yin Y, Wang W, Tang W, Tan Z, Shi J, Li L, Ruan B. Altered fecal microbiota composition in patients with major depressive disorder. Brain Behav Immun. 2015;48:186–94. Epub 20150413. doi: 10.1016/j.bbi.2015.03.016. PubMed PMID: 25882912.
    OpenUrlCrossRefPubMed
  52. 52.↵
    Chung YE, Chen HC, Chou HL, Chen IM, Lee MS, Chuang LC, Liu YW, Lu ML, Chen CH, Wu CS, Huang MC, Liao SC, Ni YH, Lai MS, Shih WL, Kuo PH. Exploration of microbiota targets for major depressive disorder and mood related traits. J Psychiatr Res. 2019;111:74–82. Epub 20190119. doi: 10.1016/j.jpsychires.2019.01.016. PubMed PMID: 30685565.
    OpenUrlCrossRefPubMed
  53. 53.↵
    Wang J, Chen WD, Wang YD. The Relationship Between Gut Microbiota and Inflammatory Diseases: The Role of Macrophages. Front Microbiol. 2020;11:1065. Epub 2020/06/26. doi: 10.3389/fmicb.2020.01065. PubMed PMID: 32582063; PMCID: PMC7296120.
    OpenUrlCrossRefPubMed
  54. 54.↵
    Groeger D, O’Mahony L, Murphy EF, Bourke JF, Dinan TG, Kiely B, Shanahan F, Quigley EM. Bifidobacterium infantis 35624 modulates host inflammatory processes beyond the gut. Gut Microbes. 2013;4(4):325–39. Epub 2013/07/12. doi: 10.4161/gmic.25487. PubMed PMID: 23842110; PMCID: PMC3744517.
    OpenUrlCrossRefPubMed
  55. 55.↵
    Backhed F, Roswall J, Peng Y, Feng Q, Jia H, Kovatcheva-Datchary P, Li Y, Xia Y, Xie H, Zhong H, Khan MT, Zhang J, Li J, Xiao L, Al-Aama J, Zhang D, Lee YS, Kotowska D, Colding C, Tremaroli V, Yin Y, Bergman S, Xu X, Madsen L, Kristiansen K, Dahlgren J, Wang J. Dynamics and Stabilization of the Human Gut Microbiome during the First Year of Life. Cell Host Microbe. 2015;17(5):690–703. Epub 2015/05/15. doi: 10.1016/j.chom.2015.04.004. PubMed PMID: 25974306.
    OpenUrlCrossRefPubMed
  56. 56.↵
    Hall AB, Yassour M, Sauk J, Garner A, Jiang X, Arthur T, Lagoudas GK, Vatanen T, Fornelos N, Wilson R, Bertha M, Cohen M, Garber J, Khalili H, Gevers D, Ananthakrishnan AN, Kugathasan S, Lander ES, Blainey P, Vlamakis H, Xavier RJ, Huttenhower C. A novel Ruminococcus gnavus clade enriched in inflammatory bowel disease patients. Genome Med. 2017;9(1):103. Epub 2017/12/01. doi: 10.1186/s13073-017-0490-5. PubMed PMID: 29183332; PMCID: PMC5704459.
    OpenUrlCrossRefPubMed
  57. 57.↵
    Hiippala K, Kainulainen V, Kalliomaki M, Arkkila P, Satokari R. Mucosal Prevalence and Interactions with the Epithelium Indicate Commensalism of Sutterella spp. Front Microbiol. 2016;7:1706. Epub 2016/11/12. doi: 10.3389/fmicb.2016.01706. PubMed PMID: 27833600; PMCID: PMC5080374.
    OpenUrlCrossRefPubMed
  58. 58.↵
    Pokusaeva K, Fitzgerald GF, van Sinderen D. Carbohydrate metabolism in Bifidobacteria. Genes Nutr. 2011;6(3):285–306. Epub 20110216. doi: 10.1007/s12263-010-0206-6. PubMed PMID: 21484167; PMCID: PMC3145055.
    OpenUrlCrossRefPubMedWeb of Science
  59. 59.↵
    Caputo M, Bona E, Leone I, Sama MT, Nuzzo A, Ferrero A, Aimaretti G, Marzullo P, Prodam F. Inositols and metabolic disorders: From farm to bedside. J Tradit Complement Med. 2020;10(3):252–9. Epub 20200324. doi: 10.1016/j.jtcme.2020.03.005. PubMed PMID: pmcid:32670820; PMCID: PMC7340869.
    OpenUrlCrossRef
  60. 60.↵
    Coupland NJ, Ogilvie CJ, Hegadoren KM, Seres P, Hanstock CC, Allen PS. Decreased prefrontal Myo-inositol in major depressive disorder. Biol Psychiatry. 2005;57(12):1526–34. doi: 10.1016/j.biopsych.2005.02.027. PubMed PMID: 15953489.
    OpenUrlCrossRefPubMedWeb of Science
  61. 61.↵
    Nie X, Chen J, Ma X, Ni Y, Shen Y, Yu H, Panagiotou G, Bao Y. A metagenome-wide association study of gut microbiome and visceral fat accumulation. Comput Struct Biotechnol J. 2020;18:2596–609. Epub 20200920. doi: 10.1016/j.csbj.2020.09.026. PubMed PMID: 33033580; PMCID: PMC7528071.
    OpenUrlCrossRefPubMed
  62. 62.↵
    Perna S, Alalwan TA, Alaali Z, Alnashaba T, Gasparri C, Infantino V, Hammad L, Riva A, Petrangolini G, Allegrini P, Rondanelli M. The Role of Glutamine in the Complex Interaction between Gut Microbiota and Health: A Narrative Review. Int J Mol Sci. 2019;20(20). Epub 2019/10/28. doi: 10.3390/ijms20205232. PubMed PMID: 31652531; PMCID: PMC68341721.
    OpenUrlCrossRefPubMed
  63. 63.↵
    Smith SE, Li J, Garbett K, Mirnics K, Patterson PH. Maternal immune activation alters fetal brain development through interleukin-6. J Neurosci. 2007;27(40):10695–702. Epub 2007/10/05. doi: 10.1523/JNEUROSCI.2178-07.2007. PubMed PMID: 17913903; PMCID: PMC2387067.
    OpenUrlAbstract/FREE Full Text
  64. 64.↵
    Choi GB, Yim YS, Wong H, Kim S, Kim H, Kim SV, Hoeffer CA, Littman DR, Huh JR. The maternal interleukin-17a pathway in mice promotes autism-like phenotypes in offspring. Science. 2016;351(6276):933–9. Epub 20160128. doi: 10.1126/science.aad0314. PubMed PMID: 26822608; PMCID: PMC4782964.
    OpenUrlAbstract/FREE Full Text
  65. 65.↵
    Wu WL, Hsiao EY, Yan Z, Mazmanian SK, Patterson PH. The placental interleukin-6 signaling controls fetal brain development and behavior. Brain Behav Immun. 2017;62:11–23. Epub 20161109. doi: 10.1016/j.bbi.2016.11.007. PubMed PMID: 27838335; PMCID: PMC5373986.
    OpenUrlCrossRefPubMed
  66. 66.↵
    Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, Pace NR. Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proceedings of the National Academy of Sciences of the United States of America. 2007;104(34):13780–5. Epub 20070815. doi: 10.1073/pnas.0706625104. PubMed PMID: 17699621; PMCID: PMC1959459.
    OpenUrlAbstract/FREE Full Text
  67. 67.↵
    Schirmer M, Smeekens SP, Vlamakis H, Jaeger M, Oosting M, Franzosa EA, Ter Horst R, Jansen T, Jacobs L, Bonder MJ, Kurilshikov A, Fu J, Joosten LAB, Zhernakova A, Huttenhower C, Wijmenga C, Netea MG, Xavier RJ. Linking the Human Gut Microbiome to Inflammatory Cytokine Production Capacity. Cell. 2016;167(4):1125–36 e8. doi: 10.1016/j.cell.2016.10.020. PubMed PMID: 27814509; PMCID: PMC5131922.
    OpenUrlCrossRefPubMed
  68. 68.↵
    Shahini A, Shahini A. Role of interleukin-6-mediated inflammation in the pathogenesis of inflammatory bowel disease: focus on the available therapeutic approaches and gut microbiome. J Cell Commun Signal. 2022. Epub 20220916. doi: 10.1007/s12079-022-00695-x. PubMed PMID: 36112307.
    OpenUrlCrossRefPubMed
  69. 69.↵
    Kanazawa A, Aida M, Yoshida Y, Kaga H, Katahira T, Suzuki L, Tamaki S, Sato J, Goto H, Azuma K, Shimizu T, Takahashi T, Yamashiro Y, Watada H. Effects of Synbiotic Supplementation on Chronic Inflammation and the Gut Microbiota in Obese Patients with Type 2 Diabetes Mellitus: A Randomized Controlled Study. Nutrients. 2021;13(2). Epub 20210208. doi: 10.3390/nu13020558. PubMed PMID: 33567701; PMCID: PMC7914668.
    OpenUrlCrossRefPubMed
  70. 70.↵
    Stout MJ, Chubiz J, Raghuraman N, Zhao P, Tuuli MG, Wang LV, Cahill AG, Cuculich PS, Wang Y, Jungheim ES, Herzog ED, Fay J, Schwartz AL, Macones GA, England SK. A multidisciplinary Prematurity Research Cohort Study. PLoS One. 2022;17(8):e0272155. Epub 20220825. doi: 10.1371/journal.pone.0272155. PubMed PMID: 36006907; PMCID: PMC9409532.
    OpenUrlCrossRefPubMed
  71. 71.↵
    Cuschieri S. The STROBE guidelines. Saudi J Anaesth. 2019;13(Suppl 1):S31–S4. doi: 10.4103/sja.SJA_543_18. PubMed PMID: 30930717; PMCID: PMC6398292.
    OpenUrlCrossRefPubMed
  72. 72.↵
    Kind AJH, Buckingham WR. Making Neighborhood-Disadvantage Metrics Accessible - The Neighborhood Atlas. N Engl J Med. 2018;378(26):2456–8. Epub 2018/06/28. doi: 10.1056/NEJMp1802313. PubMed PMID: 29949490; PMCID: PMC6051533.
    OpenUrlCrossRefPubMed
  73. 73.↵
    Institute NC. The Healthy Eating Index – Population Ratio Method Updated December 14, 2021. Available from: https://epi.grants.cancer.gov/hei/population-ratio-method.html.
  74. 74.↵
    Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150:782–6. doi: 10.1192/bjp.150.6.782. PubMed PMID: 3651732.
    OpenUrlAbstract/FREE Full Text
  75. 75.↵
    1. Cohen S,
    2. Kessler RC,
    3. Gordon LU, editors
    Cohen S, Kessler RC, Gordon LU. Perceived stress scale: Measuring stress: A guide for health and social scientists. Cohen S, Kessler RC, Gordon LU, editors. New York, NY, US: Oxford University Press; 1997. xii, 236-xii, p.
  76. 76.↵
    Slavich GM, Shields GS. Assessing Lifetime Stress Exposure Using the Stress and Adversity Inventory for Adults (Adult STRAIN): An Overview and Initial Validation. Psychosom Med. 2018;80(1):17–27. doi: 10.1097/PSY.0000000000000534. PubMed PMID: 29016550; PMCID: PMC5757659.
    OpenUrlCrossRefPubMed
  77. 77.↵
    Lewis TT, Yang FM, Jacobs EA, Fitchett G. Racial/ethnic differences in responses to the everyday discrimination scale: a differential item functioning analysis. Am J Epidemiol. 2012;175(5):391–401. Epub 20120203. doi: 10.1093/aje/kwr287. PubMed PMID: 22306556; PMCID: PMC3282874.
    OpenUrlCrossRefPubMed
  78. 78.↵
    Fein SB, Labiner-Wolfe J, Shealy KR, Li R, Chen J, Grummer-Strawn LM. Infant Feeding Practices Study II: study methods. Pediatrics. 2008;122 Suppl 2:S28–35. doi: 10.1542/peds.2008-1315c. PubMed PMID: 18829828.
    OpenUrlAbstract/FREE Full Text
  79. 79.↵
    Planer JD, Peng Y, Kau AL, Blanton LV, Ndao IM, Tarr PI, Warner BB, Gordon JI. Development of the gut microbiota and mucosal IgA responses in twins and gnotobiotic mice. Nature. 2016;534(7606):263–6. Epub 20160525. doi: 10.1038/nature17940. PubMed PMID: 27279225; PMCID: PMC4902178.
    OpenUrlCrossRefPubMed
  80. 80.↵
    Bennett WE, Jr., Gonzalez-Rivera R, Shaikh N, Magrini V, Boykin M, Warner BB, Hamvas A, Tarr PI. A method for isolating and analyzing human mRNA from newborn stool. J Immunol Methods. 2009;349(1-2):56–60. Epub 20090803. doi: 10.1016/j.jim.2009.07.013. PubMed PMID: 19660464; PMCID: PMC2850193.
    OpenUrlCrossRefPubMed
  81. 81.↵
    Warner BB, Deych E, Zhou Y, Hall-Moore C, Weinstock GM, Sodergren E, Shaikh N, Hoffmann JA, Linneman LA, Hamvas A, Khanna G, Rouggly-Nickless LC, Ndao IM, Shands BA, Escobedo M, Sullivan JE, Radmacher PG, Shannon WD, Tarr PI. Gut bacteria dysbiosis and necrotising enterocolitis in very low birthweight infants: a prospective case-control study. Lancet. 2016;387(10031):1928–36. Epub 20160309. doi: 10.1016/S0140-6736(16)00081-7. PubMed PMID: 26969089; PMCID: PMC5553277.
    OpenUrlCrossRefPubMed
  82. 82.↵
    Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, Alexander H, Alm EJ, Arumugam M, Asnicar F, Bai Y, Bisanz JE, Bittinger K, Brejnrod A, Brislawn CJ, Brown CT, Callahan BJ, Caraballo-Rodriguez AM, Chase J, Cope EK, Da Silva R, Diener C, Dorrestein PC, Douglas GM, Durall DM, Duvallet C, Edwardson CF, Ernst M, Estaki M, Fouquier J, Gauglitz JM, Gibbons SM, Gibson DL, Gonzalez A, Gorlick K, Guo J, Hillmann B, Holmes S, Holste H, Huttenhower C, Huttley GA, Janssen S, Jarmusch AK, Jiang L, Kaehler BD, Kang KB, Keefe CR, Keim P, Kelley ST, Knights D, Koester I, Kosciolek T, Kreps J, Langille MGI, Lee J, Ley R, Liu YX, Loftfield E, Lozupone C, Maher M, Marotz C, Martin BD, McDonald D, McIver LJ, Melnik AV, Metcalf JL, Morgan SC, Morton JT, Naimey AT, Navas-Molina JA, Nothias LF, Orchanian SB, Pearson T, Peoples SL, Petras D, Preuss ML, Pruesse E, Rasmussen LB, Rivers A, Robeson MS, 2nd, Rosenthal P, Segata N, Shaffer M, Shiffer A, Sinha R, Song SJ, Spear JR, Swafford AD, Thompson LR, Torres PJ, Trinh P, Tripathi A, Turnbaugh PJ, Ul-Hasan S, van der Hooft JJJ, Vargas F, Vazquez-Baeza Y, Vogtmann E, von Hippel M, Walters W, Wan Y, Wang M, Warren J, Weber KC, Williamson CHD, Willis AD, Xu ZZ, Zaneveld JR, Zhang Y, Zhu Q, Knight R, Caporaso JG. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019;37(8):852–7. Epub 2019/07/26. doi: 10.1038/s41587-019-0209-9. PubMed PMID: 31341288; PMCID: PMC7015180.
    OpenUrlCrossRefPubMed
  83. 83.↵
    Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, Peplies J, Glöckner FO. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41. doi: 10.1093/nar/gks1219.
    OpenUrlCrossRefPubMedWeb of Science
  84. 84.↵
    Balvociute M, Huson DH. SILVA, RDP, Greengenes, NCBI and OTT - how do these taxonomies compare? BMC Genomics. 2017;18(Suppl 2):114. doi: 10.1186/s12864-017-3501-4. PubMed PMID: 28361695; PMCID: PMC5374703.
    OpenUrlCrossRefPubMed
  85. 85.↵
    Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20. doi: 10.1093/bioinformatics/btu170. PubMed PMID: 24695404; PMCID: PMC4103590.
    OpenUrlCrossRefPubMedWeb of Science
  86. 86.↵
    Martin FJ, Amode MR, Aneja A, Austine-Orimoloye O, Azov AG, Barnes I, Becker A, Bennett R, Berry A, Bhai J, Bhurji SK, Bignell A, Boddu S, Branco Lins PR, Brooks L, Ramaraju SB, Charkhchi M, Cockburn A, Da Rin Fiorretto L, Davidson C, Dodiya K, Donaldson S, El Houdaigui B, El Naboulsi T, Fatima R, Giron CG, Genez T, Ghattaoraya GS, Martinez JG, Guijarro C, Hardy M, Hollis Z, Hourlier T, Hunt T, Kay M, Kaykala V, Le T, Lemos D, Marques-Coelho D, Marugan JC, Merino GA, Mirabueno LP, Mushtaq A, Hossain SN, Ogeh DN, Sakthivel MP, Parker A, Perry M, Pilizota I, Prosovetskaia I, Perez-Silva JG, Salam AIA, Saraiva-Agostinho N, Schuilenburg H, Sheppard D, Sinha S, Sipos B, Stark W, Steed E, Sukumaran R, Sumathipala D, Suner MM, Surapaneni L, Sutinen K, Szpak M, Tricomi FF, Urbina-Gomez D, Veidenberg A, Walsh TA, Walts B, Wass E, Willhoft N, Allen J, Alvarez-Jarreta J, Chakiachvili M, Flint B, Giorgetti S, Haggerty L, Ilsley GR, Loveland JE, Moore B, Mudge JM, Tate J, Thybert D, Trevanion SJ, Winterbottom A, Frankish A, Hunt SE, Ruffier M, Cunningham F, Dyer S, Finn RD, Howe KL, Harrison PW, Yates AD, Flicek P. Ensembl 2023. Nucleic Acids Res. 2022. Epub 20221101. doi: 10.1093/nar/gkac958. PubMed PMID: 36318249.
    OpenUrlCrossRefPubMed
  87. 87.↵
    Suzek BE, Wang Y, Huang H, McGarvey PB, Wu CH, UniProt C. UniRef clusters: a comprehensive and scalable alternative for improving sequence similarity searches. Bioinformatics. 2015;31(6):926–32. Epub 2014/11/16. doi: 10.1093/bioinformatics/btu739. PubMed PMID: 25398609; PMCID: PMC4375400.
    OpenUrlCrossRefPubMed
  88. 88.↵
    Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara R, Simpson GL, Solymos P, Stevens M, Wagner H. vegan: Community Ecology Package. R package version 2.0–10. 2013. There is no corresponding record for this reference. 2015.
    OpenUrl
  89. 89.↵
    Mason SJ, Graham NE. Areas beneath the relative operating characteristics (ROC) and relative operating levels (ROL) curves: Statistical significance and interpretation. Quarterly Journal of the Royal Meteorological Society. 2002;128(584):2145–66. doi: https://doi.org/10.1256/003590002320603584.
    OpenUrl
  90. 90.↵
    Kleine Bardenhorst S, Berger T, Klawonn F, Vital M, Karch A, Rubsamen N. Data Analysis Strategies for Microbiome Studies in Human Populations-a Systematic Review of Current Practice. mSystems. 2021;6(1). Epub 20210223. doi: 10.1128/mSystems.01154-20. PubMed PMID: 33622856; PMCID: PMC8573962.
    OpenUrlAbstract/FREE Full Text
Back to top
Posted November 18, 2022.
Download PDF

Supplementary Material

Data/Code
Email
Psychosocial adversity experienced in utero and early life is associated with variation in gut microbiota: a prospective case-control study
Barbara B. Warner, Bruce A. Rosa, I. Malick Ndao, Phillip I. Tarr, J. Phillip Miller, Sara K. England, Joan L. Luby, Cynthia E. Rogers, Carla Hall-Moore, Renay E. Bryant, Jacqueline D. Wang, Laura Linneman, Tara A. Smyser, Christopher D. Smyser, Deanna M. Barch, Gregory E. Miller, Edith Chen, John Martin, Makedonka Mitreva
medRxiv 2022.11.17.22282482; doi: https://doi.org/10.1101/2022.11.17.22282482
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area