A POTENTIAL ROLE FOR EARLY GUT MICROBIAL COLONIZATION IN INFANT BEHAVIOR AND CHILD SLEEP DISORDERS By Tengfei Ma A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Epidemiology—Doctor of Philosophy 2022 ABSTRACT A POTENTIAL ROLE FOR EARLY GUT MICROBIAL COLONIZATION IN INFANT BEHAVIOR AND CHILD SLEEP DISORDERS By Tengfei Ma There is increasing evidence from pre-clinical and human studies implicating the microbiota– gut–brain axis in behavior and sleep physiology. Infancy is a critical time period for brain development and is vulnerable to the harmful effects of gut dysbiosis. Thus, it is crucial to understand how gut microbial colonization during this period may influence behavior and sleep physiology in the later stages of life. We analyzed data from 194 mother-infant pairs from the Michigan Archive for Research on Child Health (MARCH) cohort Study. Clinical and demographic information was obtained from the birth certificate and interview during pregnancy and childhood. Fecal samples from infants at 3-9 months of age were sequenced at the V4 region of the 16S rRNA gene. In the first study, which examined whether feeding practices may affect early gut microbial colonization, we found that the gut microbiota of infants who were exclusively breastfed displayed a significantly lower Shannon diversity (p-adjust < 0.001) and a different gut microbiota composition than infants who were not breastfed (p-value = 0.001). Among the exclusively breastfed infants, recipients of supplemental vitamin D displayed a significantly lower Shannon diversity (p-adjust = 0.007) and different gut microbiota composition structure than non-supplemented, breastfed infants (p-value = 0.02). In addition, several individual taxa were identified to be associated with different feeding practices. In the second study, we examined whether gut microbiota in early infancy was associated with temperament in the nine-month-old infants. We identified that a microbial cluster characterized by a higher abundance of Bifidobacterium, Veillonella, and Escherichia-Shigella that was associated with lower emotionality scores (coefficient = -0.58, p-value = 0.02) compared to a cluster characterized by a higher abundance of Bacteroides. This association was especially prominent among infants who were not supplemented with vitamin D (coefficient = -1.01, p- value = 0.01), while no significant association was found among infants who were supplemented (coefficient = -0.43, p-value = 0.20). In the final aim, we assessed the association between gut microbiota in early infancy and the difficulty of initiating and maintaining sleep at age of two years. The gut microbiota of children who had difficulty maintaining sleep displayed significantly higher Shannon index (OR: 2.41, 95% CI= 1.23-4.93, p-adjust < 0.04) and Chao 1 index (OR: 1.01, 95% CI= 1.0-1.03, p-adjust < 0.008) after adjustment for covariates. We also observed that gut microbiota composition was significantly different between children with difficulty initiating (p-value= 0.043) and maintaining sleep (p-value= 0.004) by PERMANOVA based on the unweighted UniFrac distance metric. In conclusion, these results from analysis in a prospective cohort study suggest that early gut microbial colonization is shaped by breastfeeding status, vitamin D supplement, and maternal characteristics including gestational age, delivery mode and education level. Our findings suggested that the infant gut microbiome clusters may be associated with the temperament characteristic of negative emotionality in 9-month-old infants. We also demonstrated a significant association between infant gut microbiome composition and sleep problems in 2- year-old children. Thus, our results add to the evidence that early gut microbial colonization may be linked with brain outcomes with potential long-term effects. ACKNOWLEDGMENTS I would like to thank my advisor and dissertation committee chair, Dr. Nigel Paneth, for his invaluable suggestions from the beginning to end of PhD program. I am also very grateful to my dissertation committee members, Dr. Sarah Comstock and Dr. Jean Kerver for their guidance and inspiration for my research ideas. I would like to thank Dr. Lixin Zhang and Dr. Chenxi Li, for their help, and for being on my dissertation committee. I would like to thank the whole CHARM study team who have worked hard to on this large cohort and provide me with the data used in this dissertation and to provide me with an inspiration for my research ideas. iv TABLE OF CONTENTS LIST OF TABLES ........................................................................................................................ vii LIST OF FIGURES ....................................................................................................................... ix KEY TO ABBREVIATIONS ..........................................................................................................x CHAPTER 1. INTRODUCTION ....................................................................................................1 1.1 Background .............................................................................................................................1 1.2 Early microbial colonization...................................................................................................2 1.3 The microbiota–gut–brain axis ...............................................................................................4 1.4 Gut microbiome and sleep ......................................................................................................7 1.5 Gut microbiome research methods .........................................................................................8 1.6 Study population ...................................................................................................................10 REFERENCES ..............................................................................................................................12 CHAPTER 2. SPECIFIC AIMS ....................................................................................................23 CHAPTER 3. VITAMIN D SUPPLEMENTATION IN EXCLUSIVELY BREASTFED INFANTS IS ASSOCIATED WITH ALTERATIONS IN THE FECAL MICROBIOME .........25 3.1 Abstract .................................................................................................................................25 3.2 Introduction...........................................................................................................................25 3.3 Materials and Methods ..........................................................................................................27 3.3.1 Study participants............................................................................................................27 3.3.2 Data collection ................................................................................................................27 3.3.3 Fecal microbiota analysis................................................................................................28 3.3.4 Statistical analysis ...........................................................................................................28 3.4 Results ...................................................................................................................................29 3.4.1 Participants and feeding practices ...................................................................................29 3.4.2 Gut microbiota analysis ..................................................................................................33 3.5 Discussion .............................................................................................................................40 APPENDIX ....................................................................................................................................45 REFERENCES ..............................................................................................................................51 CHAPTER 4. ASSOCIATION OF THE INFANT GUT MICROBIOME WITH TEMPERAMENT..........................................................................................................................59 4.1 Abstract .................................................................................................................................59 4.2 Introduction ...........................................................................................................................60 4.3 Materials and Methods ..........................................................................................................62 4.3.1 Study participants............................................................................................................62 4.3.2 Data collection ................................................................................................................62 4.3.3 Fecal microbiota analysis................................................................................................63 4.3.4 Statistical analysis ...........................................................................................................63 v 4.4 Results ...................................................................................................................................65 4.4.1 Study Population Characteristics and temperament scales .............................................65 4.4.2 Alpha diversity and temperament scores ........................................................................69 4.4.3 Cluster analysis ...............................................................................................................69 4.4.4 Individual taxa analysis...................................................................................................71 4.5 Discussion .............................................................................................................................75 4.5.1 Alpha diversity and temperament scales.........................................................................76 4.5.2 Cluster analysis ...............................................................................................................76 4.5.3 Individual taxa ................................................................................................................77 4.5.4 Strength and limitation....................................................................................................78 APPENDIX ....................................................................................................................................80 REFERENCES ..............................................................................................................................83 CHAPTER 5. ASSOCIATION BETWEEN INFANT GUT MICROBIOME AND SLEEP PROBLEMS DURING CHILDHOOD .........................................................................................90 5.1 Abstract .................................................................................................................................90 5.2 Introduction...........................................................................................................................90 5.2.1 Sleep in childhood...........................................................................................................90 5.2.2 The gut microbiome and sleep ........................................................................................91 5.3 Materials and Methods ..........................................................................................................92 5.3.1 Study participants............................................................................................................92 5.3.2 Data collection ................................................................................................................92 5.3.3 Fecal microbiota analysis................................................................................................93 5.3.4 Statistical analysis ...........................................................................................................94 5.4 Results ...................................................................................................................................95 5.5 Discussion ...........................................................................................................................104 APPENDIX ..................................................................................................................................108 REFERENCES ............................................................................................................................111 CHAPTER 6. SUMMARY, LIMITATIONS AND FUTURE RESEARCH ..............................118 REFERENCES ............................................................................................................................122 vi LIST OF TABLES Table 3.1. Characteristics of the mothers and infants by breastfeeding status ..............................32 Table 3.2. Association between breastfeeding and perinatal characteristics .................................33 Table 3.3. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) for all infants.............................................................................................................................................36 Table 3.4. MaAsLin Analysis Results: Associations of infant feeding practices and gut microbiome taxa at genus level adjusted by covariates in all infants ............................................38 Table 3.5. MaAsLin Analysis Results: Associations of infant feeding practices and gut microbiome taxa at genus level within exclusively breastfed and no breastfed infants ................40 Table S3.1. Association between Shannon/Chao1 index and breastfeeding status in all infants ..47 Table S3.2. Association between Shannon/Chao1 index and infant vitamin D supplement intake ........................................................................................................................................................47 Table S3.3. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) on weighted UniFrac distances ...........................................................................................................49 Table S3.4. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) for exclusively breastfed infants on Bray-Curtis distances .................................................................49 Table S3.5. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) for not breastfed infants on Bray-Curtis distances ....................................................................................50 Table 4.1. Characteristics of mothers and infants by infant temperament scales ..........................66 Table 4.2. Association between alpha diversity and infant temperament......................................69 Table 4.3. Association between gut microbiota clusters and infant temperament scales ..............74 Table 4.4. MaAsLin Analysis Results: Associations of gut microbiome taxa at genus level and infant temperament scales adjusted by covariates .........................................................................75 Table S4.1. Association between gut microbiota clusters and infant temperament scales stratifying by infant vitamin D intake ............................................................................................81 Table S4.2. Association between gut microbiota clusters and infant temperament scales stratifying by infant sex .................................................................................................................82 Table 5.1. Characteristics of mothers and infants by difficulty initiating sleep and maintaining sleep ...............................................................................................................................................96 vii Table 5.2. Association between alpha diversity and difficulty initiating and maintaining sleep ...... ......................................................................................................................................................101 Table 5.3. DESeq2 Analysis Results: Associations of gut microbiome taxa at genus level and difficulty initiating sleep adjusted by covariates .........................................................................103 Table 5.4. DESeq2 Analysis Results: Associations of gut microbiome taxa at genus level and difficulty maintaining sleep adjusted by covariates .....................................................................103 Table S5.1 Association between maternal/infant characteristics and difficulty falling sleep in infants...........................................................................................................................................109 Table S5.2 Association between maternal/infant characteristics and difficulty maintaining sleep in infants.......................................................................................................................................110 viii LIST OF FIGURES Figure 3.1. Distribution of infant age (month) at stool sample collection .....................................31 Figure 3.2. Infant alpha and beta diversity by infant breastfeeding and Vitamin D supplement ..35 Figure S3.1. Infant alpha diversity by infant different feeding practices ......................................46 Figure S3.2. Association between breastfeeding status and relative abundance of 8 dominant genera .............................................................................................................................................48 Figure 4.1. Dirichlet multinomial mixture clustering identified three optimal clusters from 157 fecal samples ..................................................................................................................................72 Figure 4.2. Relative abundance of the top 4 genera that contribute by clusters ............................73 Figure 5.1. Boxplot of the alpha diversity (Shannon and Chao 1) by difficulty initiating and maintaining sleep .........................................................................................................................100 Figure 5.2. Principal Coordinates Analysis (PCoA) for difficulty initiating and maintaining sleep ......................................................................................................................................................102 ix KEY TO ABBREVIATIONS MARCH Michigan Archive for Research on Child Health SCFAs Short chain fatty acids GF Germ-free CS Caesarean section HMOs Human milk oligosaccharides MGBA Microbiota-gut-brain axis CNS Central nervous system ANS Autonomic nervous system ENS Enteric nervous system HPA Hypothalamic pituitary adrenal SPF Specific pathogen-free ASD Autism spectrum disorder MTT Microbiota transplantation therapy OUT Operational Taxonomic Unit ASV Amplicon Sequence Variant PERMANOVA Permutational multivariate analysis of variance ECHO Environmental Influences on Child Health Outcomes NEG Negative emotionality PAS Positive affect/surgency ORC Orienting and regulatory capacity BMI Body mass index x FDR False discovery rate MaAsLin Multivariate association with linear models ANOVA Analysis of variance PCoA Principal coordinate analysis VDR Vitamin D receptor IBQ-RVSF Rothbart Infant Behavior Questionnaire-Revised Very Short Form DMM Dirichlet multinomial mixture OSA Obstructive sleep apnea PSDS PROMIS Sleep Disturbance scale DSM-4 Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition xi CHAPTER1. INTRODUCTION 1.1 Background The human gut microbiome is composed of a wide variety of microorganisms, including bacteria, archaea, viruses, and eukaryotic microbes that all reside in our intestines. Several basic functions conferred by the gut microbiome on the human host suggest its importance in health and disease. These functions include regulation of the immune system, protection against pathogens, and fermentation of indigestible food components into metabolites[1,2]. However, these functions can be disrupted by an altered microbial composition, which is referred to as dysbiosis[1]. Several diseases are now thought to be influenced by dysbiosis, including some types of cancer[3,4], mental disorders[5–8], inflammatory bowel diseases[9,10], type 2 diabetes[11,12], and obesity[1,13]. The gut microbiota produces a set of through the breakdown of indigestible carbohydrates [14,15]. Among the metabolites, short chain fatty acids (SCFAs) perform complex but important roles in the human body[16]. The three most prominent SCFAs are acetate, propionate, and butyrate, which occur in a ratio of approximately 3:1:1 in the human intestinal lumen, respectively[16]. Butyrate is thought to be the most important SCFA for human health, as it forms the major energy source for human colonocytes[17,18]. Butyrate also potentially prevents cancer activity by inducing apoptosis of colon cancer cells and by regulating gene expression by inhibiting histone deacetylases[19]. Propionate is also an energy source for the epithelial cells but is largely taken up by the liver[18]. The transformation of propionate to glucose in intestinal gluconeogenesis has beneficial effects on energy homeostasis by decreasing hepatic activity, and in turn, reduces adiposity[20]. Acetate is the most abundant SCFA, and is important for gut environment stability and the growth of beneficial bacteria[21]. In addition, Acetate can cross the 1 blood–brain barrier and regulate hypothalamic neuronal activation patterning and neuropeptide release[22].The joint study of the microbiome and the metabolome is considered one of the best approaches to study host–microbiome interactions[23,24]. The gut microbiota is also critical for the development of both the intestinal mucosal and systemic immune system as demonstrated by the studies in animal models. A number of studies have shown that germ-free (GF) mice are more susceptible to infection than specific-pathogen free mice[25]. In addition, administration of antibiotics is associated with an increased risk of pathogen colonization in both mice and humans, suggesting an important function of the commensal microbiota in protecting the host from infection[26,27]. A major immune deficiency in germ-free animals is the absence of expansion of CD4+ T-cell populations, which can be completely reversed by polysaccharide A produced by B. fragili[28,29]. This reversion is mainly performed by the pattern recognition receptors (PRRs) of epithelial cells, such as Toll-like or Nod-like receptors, which can recognize the molecular effectors that are produced by gut microbiota[30]. Dysbiosis can alter the microbiol molecules sensed by the host, and in turn lead to a different activation state of the immune system[31]. A recent research which studied prenatal and early life bacterial colonization found that by transferring altered gut microbiome from pregnant patients with Crohn’s disease to germ-free mice, an imbalanced immune system lacking critical homeostatic elements was displayed in the GF mice, suggesting gut dysbiosis could trigger abnormal imprinting of the intestinal immune system[32]. 1.2 Early microbial colonization The colonization of gut bacteria begins at birth and remains highly dynamic until about 2-3 years of age when more stable microbial profiles begin to emerge[33,34]. Several reports highlighted that the early life development of the infant gut microbiota plays a critical role in the health of 2 later life, such as allergic diseases, obesity, and autism[33,35]. Mode of delivery, infant feeding practices, and antibiotic exposure are thought to be three key factors that influence early microbial colonization and establishment[36]. The relationship between these 3 factors and infant gut microbiome have been well established in the studies of recent years. Mode of delivery is one of the major contributors to influencing the infant's gut microbiome. During vaginal delivery, the infant’s microbial colonization starts at the contact with the maternal vaginal and intestinal flora, while infants delivered by CS acquire bacteria derived from the maternal skin, mouth, and from the hospital environment[37,38]. The gut microbiota patterns of infants delivered by CS differ from those who were vaginally delivered, including lower diversity and richness, lower abundance of Bacteroides, Lactobacillus, and Bifidobacterium, and higher abundance of Clostridium difficile and microbes associated with the human skin such as Staphylococcus, Streptococcus and Propionibacterium[39–41]. Previous studies found that from birth to 90 days of life, Bifidobacterium and Lactobacillus were significantly lower in the infants delivered by CS compared with those delivered vaginally[42]. These differences in newborn gut microbiome community may have an impact on health since the genera Bifidobacterium and Lactobacillus are considered to be the beneficial bacteria in the gut[43]. Infant feeding practices are also key factors in shaping early microbiota composition[44,45]. Recent studies have shown that gut microbial profiles in breastfed infants are significantly different from those in formula-fed infants and change rapidly after the transition from breastfeeding to formula or solid food[46,47]. Specifically, breastfed infants have higher abundance of bifidobacterial, and lower microbial diversity compared with formula-fed infants whose gut microbiota is more diverse and similar to older children[47,48]. The differences in gut microbiota composition observed between formula-fed and breastfed infants have, at least 3 partially, been attributed to the absence of human milk oligosaccharides (HMOs) in infant formula[49]. Human milk is enriched with HMOs, which have been linked to beneficial bacteria in the gut microbiota[50,51]. In recent years, some oligosaccharides have been added to some formula to help infant to establish a Bifidobacterium-rich microbiota[51]. However, formula-fed infants still have distinct features of their microbiotas compared with breastfed infants, with a higher abundance of C. difficile[52]. The introduction of solid food represents another key factor of feeding practices influencing the composition of infant gut microbiota, producing an adult-like complex microbiome dominated by the phyla Bacteroidetes and Firmicutes[53,54]. Infant gut bacterial abundance changed significantly with the introduction of solid foods between 9 and 18 months. Specifically, the abundance of Bacteroidetes increases and the abundance of Bifidobacterium and Lactobacillus decrease[55,56]. During birth and immediately thereafter, broad-spectrum antibiotics are commonly used in newborns and their mothers who are at high risk for infection. However, antibiotics can also contribute to gut dysbiosis[57]. Early antibiotic exposure can reduce the diversity of the gut microbiota of infants and change its composition, with a decreased abundance of Bifidobacterium and increased abundance of Proteobacteria[58]. Healthy infants whose mothers received ampicillin for group B Streptococcus before delivery displayed significantly decreased abundance of Bifidobacterium by one week of life, highlighting the modulatory influences of intrapartum antibiotic interventions[59]. Studies have also demonstrated that the prophylactic antibiotic treatment in preterm infants can reduce the diversity of gut flora and delay the colonization of commensal flora[60,61]. 1.3 The microbiota–gut–brain axis The ‘microbiota-gut-brain axis’ (MGBA) refers to the biological network involving multiple 4 biological systems that allow bidirectional communication between gut bacteria and the brain[62]. This axis is becoming popular in fields investigating the biological and physiological basis of stroke, psychiatric, neurodevelopmental, age-related, and neurodegenerative disorders[62–64]. The communication network includes the central nervous system (CNS), both brain and spinal cord, the autonomic nervous system (ANS), the enteric nervous system (ENS) and the hypothalamic pituitary adrenal (HPA) axis[63]. The communication pathways in these biological networks include both direct and indirect signaling via neurotransmitters, metabolic pathways, and the immune system[62,63]. The most well-studied neuronal pathway for the MGBA is the vagus nerve signaling. The vagus nerve innervates the muscle and mucosa layers of the gut both in the lamina propria and muscularis externa, detects sensory signals and then relays these signals to the CNS[65,66]. The most important function of the vagus nerve is afferent activity which brings information of the inner organs, such as gut, liver, heart, and lungs to the brain[67]. A study reported that in vagotomized mice where the vagus nerve has been surgically severed, administration of Lactobacillus rhamnosus JB-1 does not affect GABA receptor expression[68]. whereas in normal mice, administration of Lactobacillus rhamnosus JB-1 alters the expression of GABA receptors in brain regions associated with fear and emotions.These receptors modulate anxiety- like behaviors, suggesting that the vagus as a major modulatory communication pathway between gut microbiota and the brain[68]. Inspired by how microbiota influence the brain through their ability to produce and modify many metabolic, immunological and neurochemical factors that impact the nervous system, flood of research is now connecting microbial communities, and their function to neuropsychiatric disorders associated with development (for example, autism spectrum disorder and 5 schizophrenia), with mood (for example, depression and anxiety) and with neurodegeneration (for example, Parkinson disease and Alzheimer disease)[69,70]. GF mice are one of the most widely used technical strategies in studying MGBA[71]. The main advantage of the GF mice model is that gut microbiome strains thought to be risk factors in gastrointestinal and systemic disease, such as candidate psychobiotics, can be studied in GF mice to learn more about their specific effects. In addition, fecal microbiota of human donors can be transplanted into GF mice to directly investigate the role of bacteria on disease pathogenesis. Several studies have shown that GF mice displayed increased stress response and decreased anxiety with augmented levels of adreno-corticotrophic hormone and cortisol compared with specific pathogen-free (SPF) mice with normal gut microbiota[72,73]. The abnormalities could be partially reversed when the gastrointestinal tracts of the germ-free mice were reconstituted with stool from normally raised mice[72]. The administration of microbiota from a patient with Alzheimer’s disease has been shown to trigger cognitive decline in recipient GF mice[74]. In addition, metabolites related to the nervous system, including γ-aminobutyrate, taurine, and valine, were significantly less abundant in the feces of mice transplanted with microbiota from the affected patient, reinforcing the idea of a microbiota-gut-brain axis[74]. Human studies have also demonstrated that the gut microbiota is associated with neuropsychiatric and neurodegenerative disorders[70,75,76]. Autism spectrum disorder (ASD) are a serious neurodevelopmental disorder in children and about 1 in 44 children has been identified with ASD according to the estimates from CDC in the United States[77]. Children with ASDs have been found to have lower abundances of Coprococcus, Prevotella, and unclassified Veillonellaceae, and an increased Firmicutes/Bacteroidetes ratio[78,79]. Recent research has also reported beneficial effects of fecal microbiota transplantation therapy (MTT) 6 for individuals with ASD[80]. In this open-label clinical trial of MTT, 18 children diagnosed with ASD received an antibiotic treatment for 2 weeks followed by an initial, high dose of fecal microbiota transplants and subsequent lower maintenance doses administered daily for 7– 8 weeks[80]. This study demonstrated that MTT appeared to reduce gastrointestinal symptoms (such as, constipation and abdominal pain) and improved ASD symptoms, such as social skills deficits and repetitive behavior. In a follow-up with the same participants after 2 years, most improvements in gastrointestinal symptoms were maintained, and autism-related symptoms improved even more after the end of treatment, suggesting a long-term impact[81]. The treatment also increased overall bacterial diversity and the abundance of Bifidobacteria and Prevotella, and these effects were remained over time as they were still observed in the 2-year follow-up[81]. The relationship between gut microbiota and cognitive development has been well studied in humans[82,83]. Carlson et al. demonstrated that higher alpha diversity of gut microbiota was associated with lower scores on the overall cognitive score, visual reception scale, and expressive language scale in 89 2-year-old children[82]. This study also showed that microbiome has minimal effects on regional brain volumes at 1 and 2 years of age by using MRI imaging[82]. 1.4 Gut microbiome and sleep Sleep quality and quantity are increasingly being considered as critical factors for child health and development. Epidemiology studies indicate that up to 50% of children experience a sleep disorder between 0 to 6 years old[84–86]. Difficulty initiating and maintaining sleep are the two most frequent sleep problems in childhood and often co-exist[87]. Spruyt et al. reported that 31% of children aged 6 to 13 years had disorder of initiating and maintaining sleep in a normal school-age population[88]. Liu et al. also showed that 16% of the parents in the United States 7 reported their children aged 4 to 11 years “sometimes” have difficulties falling asleep[89]. Multiple biological, psychosocial, and environmental factors are linked to sleep disorders in children[90]. Leone et al. found that gut microbiome dysbiosis could impair central and hepatic circadian clock gene expression, suggesting that gut microbiota may play a role in regulating or modifying circadian rhythm[91]. In a small group of breast cancer survivors (n=12), global sleep dysfunction was found to be associated with higher abundance of Paracoccus, Rikenellaceae, and Clostridium[92]. Another cross-sectional study of 37 healthy older adult found that global sleep dysfunction was associated with lower abundance of the phyla Verrucomicrobia and Lentisphaerae. Valentini et al. reported that children with Obstructive Sleep Apnea syndrome had a lower microbiota diversity and higher abundance of pro-inflammatory bacteria (Proteobacteria, Clostridiaceae, Oscillospiraceae, Klebsiella) compared to healthy participants[93]. 1.5 Gut microbiome research methods Nowadays, microbiome studies often rely on the analysis of 16S ribosomal RNA sequences for the taxonomic identification of bacteria. The 16S rRNA gene sequence is about 1,500 bp long and contains a highly conserved sequence that includes nine regions or windows of variable nucleotide sequence[94,95]. These nine regions constitute the most informative portions of the gene sequence for use in taxonomic classification. With the development of next-generation sequencing technology, 16S rRNA gene has become a powerful tool for pathogen detection and identification[94]. The pipeline for 16S amplicon analyses usually starts with using primers designed to amplify the hypervariable regions of the 16S rRNA gene (typically the V1–V3 region or the V3–V5 region). Sequences are clustered into Operational Taxonomic Units’ (OTUs) that contain similar 16S rRNA sequences with high sequence similarity[96]. A common 8 similarity threshold used is 97%, which was derived from an empirical study that showed most strains had 97% 16S rRNA sequence similarity[96]. For each OUT cluster, a single sequence is selected as a representative sequence, which is annotated using a 16S classification method[97]. The annotation is applied to all other sequences within the same OUT. Several pipelines have been developed to perform the entire 16S analysis from end to end, including QIIME and MOTHUR[98,99]. Recently, a new method called Amplicon Sequence Variant (ASV) approach has been developed. The ASV approach starts by determining which exact sequences are read and how many times each exact sequence is read from Illumina-scale amplicon data without imposing the arbitrary dissimilarity thresholds[100]. ASV methods infer the biological sequences in the sample before the introduction of amplification and sequencing errors. This allows ASV methods distinguish sequence variants differing from only one nucleotide[100]. Therefore, an ASV-based analysis is able to provide a higher-resolution taxonomic result allowing for more precise identification down to the species level and even potentially beyond[101]. Alpha diversity is the ecological diversity of a single sample and is commonly used as a measurable outcome in microbiome research with respect to its richness (number of different species present in an area), evenness (relative abundance of the different species in an area), or both[102]. In microbial ecology, analyzing the alpha diversity of gut microbiota data is a common first approach to assessing differences between environments. The Shannon index is one of the popular index for alpha diversity, which accounts for both richness and evenness in a single equation, while the Chao 1 index only accounts for the richness[103]. The diversity of gut microbiota within an individual has been linked to several human diseases. For example, low diversity in the gut has been associated with obesity and inflammatory bowel disease[104]. 9 While alpha diversity is a measure of microbiome diversity in a single sample, beta diversity is a measure of similarity or dissimilarity between two communities. It provides a measure of the degree to which samples differ from one another and it can reveal the structure difference between microbiota samples[105]. Some of the most popular beta diversity measures in microbiome research include the Bray-Curtis index (compositional dissimilarity), the Jaccard index (presence / absence measures, ignoring abundance information), and the UniFrac distances (which take into account the phylogenetic tree information)[106]. Many popular statistical methods, such as ordination-based methods, and permutational multivariate analysis of variance (PERMANOVA) are relied on the beta diversity[107,108]. Beta diversity is widely used for studying the association between environmental variables and microbial composition. 1.6 Study population The Michigan Archive for Research on Child Health (MARCH) cohort is an ongoing population- based pregnancy and birth cohort set in Michigan’s lower peninsula. Many important problems in child health and development may result from a mother's diet, her infections, and chemicals in her environment during her pregnancy. The MARCH study plans to examine 1,100 pregnancies in detail by interviewing women in pregnancy, acquiring abstracts of their medical records, and by saving biological specimens obtained in pregnancy (blood, urine, and placenta). The MARCH study assesses the child's health and development in relation to these factors, to learn what changes might be made during a woman's pregnancy that could prevent later problems in child health and development. The MARCH study contributes to a nation-wide study of child health called the Environmental Influences on Child Health Outcomes (ECHO). ECHO is a research program launched by the National Institutes of Health in 2016 to understand early environmental factors on child health and development from nearly 40 cohorts across the US. ECHO research 10 focuses on four key pediatric outcomes: pre-, peri-, and postnatal outcomes, upper and lower airway, obesity, and neurodevelopment. With the large sample size and generalizability of the study population, the ECHO program should allow for new insights into many prenatal factors and child health, that can lead to the development of interventions and prevention strategies to improve child health across the U.S[109]. 11 REFERENCES 12 REFERENCES 1. Thursby, E.; Juge, N. Introduction to the Human Gut Microbiota. Biochem J 2017, 474, 1823– 1836, doi:10.1042/BCJ20160510. 2. Shreiner, A.B.; Kao, J.Y.; Young, V.B. The Gut Microbiome in Health and in Disease. Curr Opin Gastroenterol 2015, 31, 69–75, doi:10.1097/MOG.0000000000000139. 3. Zitvogel, L.; Galluzzi, L.; Viaud, S.; Vétizou, M.; Daillère, R.; Merad, M.; Kroemer, G. Cancer and the Gut Microbiota: An Unexpected Link. Science Translational Medicine 2015, 7, 271ps1-271ps1, doi:10.1126/scitranslmed.3010473. 4. Louis, P.; Hold, G.L.; Flint, H.J. The Gut Microbiota, Bacterial Metabolites and Colorectal Cancer. Nat Rev Microbiol 2014, 12, 661–672, doi:10.1038/nrmicro3344. 5. Sani, G.; Manchia, M.; Simonetti, A.; Janiri, D.; Paribello, P.; Pinna, F.; Carpiniello, B. The Role of Gut Microbiota in the High-Risk Construct of Severe Mental Disorders: A Mini Review. Frontiers in Psychiatry 2021, 11, 1590, doi:10.3389/fpsyt.2020.585769. 6. Liang, S.; Wu, X.; Hu, X.; Wang, T.; Jin, F. Recognizing Depression from the Microbiota– Gut–Brain Axis. International Journal of Molecular Sciences 2018, 19, 1592, doi:10.3390/ijms19061592. 7. Liu, T.; Feenstra, K.A.; Heringa, J.; Huang, Z. Influence of Gut Microbiota on Mental Health via Neurotransmitters: A Review. Journal of Artificial Intelligence for Medical Sciences 2020, 1, 1–14, doi:10.2991/jaims.d.200420.001. 8. Kang, D.-W.; Adams, J.B.; Gregory, A.C.; Borody, T.; Chittick, L.; Fasano, A.; Khoruts, A.; Geis, E.; Maldonado, J.; McDonough-Means, S.; et al. Microbiota Transfer Therapy Alters Gut Ecosystem and Improves Gastrointestinal and Autism Symptoms: An Open-Label Study. Microbiome 2017, 5, 10, doi:10.1186/s40168-016-0225-7. 9. Miyoshi, J.; Chang, E.B. The Gut Microbiota and Inflammatory Bowel Diseases. Translational Research 2017, 179, 38–48, doi:10.1016/j.trsl.2016.06.002. 10. Pittayanon, R.; Lau, J.T.; Leontiadis, G.I.; Tse, F.; Yuan, Y.; Surette, M.; Moayyedi, P. Differences in Gut Microbiota in Patients With vs Without Inflammatory Bowel Diseases: A Systematic Review. Gastroenterology 2020, 158, 930-946.e1, doi:10.1053/j.gastro.2019.11.294. 11. Larsen, N.; Vogensen, F.K.; Berg, F.W.J. van den; Nielsen, D.S.; Andreasen, A.S.; Pedersen, B.K.; Al-Soud, W.A.; Sørensen, S.J.; Hansen, L.H.; Jakobsen, M. Gut Microbiota in Human Adults with Type 2 Diabetes Differs from Non-Diabetic Adults. PLOS ONE 2010, 5, e9085, doi:10.1371/journal.pone.0009085. 13 12. Gurung, M.; Li, Z.; You, H.; Rodrigues, R.; Jump, D.B.; Morgun, A.; Shulzhenko, N. Role of Gut Microbiota in Type 2 Diabetes Pathophysiology. EBioMedicine 2020, 51, 102590, doi:10.1016/j.ebiom.2019.11.051. 13. Zhao, L. The Gut Microbiota and Obesity: From Correlation to Causality. Nat Rev Microbiol 2013, 11, 639–647, doi:10.1038/nrmicro3089. 14. Rowland, I.; Gibson, G.; Heinken, A.; Scott, K.; Swann, J.; Thiele, I.; Tuohy, K. Gut Microbiota Functions: Metabolism of Nutrients and Other Food Components. Eur J Nutr 2018, 57, 1–24, doi:10.1007/s00394-017-1445-8. 15. Vernocchi, P.; Del Chierico, F.; Putignani, L. Gut Microbiota Metabolism and Interaction with Food Components. Int J Mol Sci 2020, 21, 3688, doi:10.3390/ijms21103688. 16. Besten, G. den; Eunen, K. van; Groen, A.K.; Venema, K.; Reijngoud, D.-J.; Bakker, B.M. The Role of Short-Chain Fatty Acids in the Interplay between Diet, Gut Microbiota, and Host Energy Metabolism. Journal of Lipid Research 2013, 54, 2325–2340, doi:10.1194/jlr.R036012. 17. Macfarlane, G.T.; Gibson, G.R.; Cummings, J.H. Comparison of Fermentation Reactions in Different Regions of the Human Colon. J Appl Bacteriol 1992, 72, 57–64, doi:10.1111/j.1365- 2672.1992.tb04882.x. 18. Rowland, I.; Gibson, G.; Heinken, A.; Scott, K.; Swann, J.; Thiele, I.; Tuohy, K. Gut Microbiota Functions: Metabolism of Nutrients and Other Food Components. Eur J Nutr 2018, 57, 1–24, doi:10.1007/s00394-017-1445-8. 19. Han, A.; Bennett, N.; Ahmed, B.; Whelan, J.; Donohoe, D.R. Butyrate Decreases Its Own Oxidation in Colorectal Cancer Cells through Inhibition of Histone Deacetylases. Oncotarget 2018, 9, 27280–27292, doi:10.18632/oncotarget.25546. 20. De Vadder, F.; Kovatcheva-Datchary, P.; Goncalves, D.; Vinera, J.; Zitoun, C.; Duchampt, A.; Bäckhed, F.; Mithieux, G. Microbiota-Generated Metabolites Promote Metabolic Benefits via Gut-Brain Neural Circuits. Cell 2014, 156, 84–96, doi:10.1016/j.cell.2013.12.016. 21. Duncan, S.H.; Holtrop, G.; Lobley, G.E.; Calder, A.G.; Stewart, C.S.; Flint, H.J. Contribution of Acetate to Butyrate Formation by Human Faecal Bacteria. Br J Nutr 2004, 91, 915–923, doi:10.1079/BJN20041150. 22. Frost, G.; Sleeth, M.L.; Sahuri-Arisoylu, M.; Lizarbe, B.; Cerdan, S.; Brody, L.; Anastasovska, J.; Ghourab, S.; Hankir, M.; Zhang, S.; et al. The Short-Chain Fatty Acid Acetate Reduces Appetite via a Central Homeostatic Mechanism. Nat Commun 2014, 5, 3611, doi:10.1038/ncomms4611. 23. Visconti, A.; Le Roy, C.I.; Rosa, F.; Rossi, N.; Martin, T.C.; Mohney, R.P.; Li, W.; de Rinaldis, E.; Bell, J.T.; Venter, J.C.; et al. Interplay between the Human Gut Microbiome and Host Metabolism. Nat Commun 2019, 10, 4505, doi:10.1038/s41467-019-12476-z. 14 24. Turnbaugh, P.J.; Gordon, J.I. An Invitation to the Marriage of Metagenomics and Metabolomics. Cell 2008, 134, 708–713, doi:10.1016/j.cell.2008.08.025. 25. Fukuda, S.; Toh, H.; Hase, K.; Oshima, K.; Nakanishi, Y.; Yoshimura, K.; Tobe, T.; Clarke, J.M.; Topping, D.L.; Suzuki, T.; et al. Bifidobacteria Can Protect from Enteropathogenic Infection through Production of Acetate. Nature 2011, 469, 543–547, doi:10.1038/nature09646. 26. Lawley, T.D.; Clare, S.; Walker, A.W.; Goulding, D.; Stabler, R.A.; Croucher, N.; Mastroeni, P.; Scott, P.; Raisen, C.; Mottram, L.; et al. Antibiotic Treatment of Clostridium Difficile Carrier Mice Triggers a Supershedder State, Spore-Mediated Transmission, and Severe Disease in Immunocompromised Hosts. Infect Immun 2009, 77, 3661–3669, doi:10.1128/IAI.00558-09. 27. Rupnik, M.; Wilcox, M.H.; Gerding, D.N. Clostridium Difficile Infection: New Developments in Epidemiology and Pathogenesis. Nat Rev Microbiol 2009, 7, 526–536, doi:10.1038/nrmicro2164. 28. Mazmanian, S.K.; Liu, C.H.; Tzianabos, A.O.; Kasper, D.L. An Immunomodulatory Molecule of Symbiotic Bacteria Directs Maturation of the Host Immune System. Cell 2005, 122, 107–118, doi:10.1016/j.cell.2005.05.007. 29. Round, J.L.; Mazmanian, S.K. The Gut Microbiome Shapes Intestinal Immune Responses during Health and Disease. Nat Rev Immunol 2009, 9, 313–323, doi:10.1038/nri2515. 30. Hevia, A.; Delgado, S.; Sánchez, B.; Margolles, A. Molecular Players Involved in the Interaction Between Beneficial Bacteria and the Immune System. Frontiers in Microbiology 2015, 6, 1285, doi:10.3389/fmicb.2015.01285. 31. Levy, M.; Kolodziejczyk, A.A.; Thaiss, C.A.; Elinav, E. Dysbiosis and the Immune System. Nat Rev Immunol 2017, 17, 219–232, doi:10.1038/nri.2017.7. 32. Torres, J.; Hu, J.; Seki, A.; Eisele, C.; Nair, N.; Huang, R.; Tarassishin, L.; Jharap, B.; Cote- Daigneault, J.; Mao, Q.; et al. Infants Born to Mothers with IBD Present with Altered Gut Microbiome That Transfers Abnormalities of the Adaptive Immune System to Germ-Free Mice. Gut 2020, 69, 42–51, doi:10.1136/gutjnl-2018-317855. 33. Tanaka, M.; Nakayama, J. Development of the Gut Microbiota in Infancy and Its Impact on Health in Later Life. Allergol Int 2017, 66, 515–522, doi:10.1016/j.alit.2017.07.010. 34. Rodríguez, J.M.; Murphy, K.; Stanton, C.; Ross, R.P.; Kober, O.I.; Juge, N.; Avershina, E.; Rudi, K.; Narbad, A.; Jenmalm, M.C.; et al. The Composition of the Gut Microbiota throughout Life, with an Emphasis on Early Life. Microb Ecol Health Dis 2015, 26, 10.3402/mehd.v26.26050, doi:10.3402/mehd.v26.26050. 35. Zhuang, L.; Chen, H.; Zhang, S.; Zhuang, J.; Li, Q.; Feng, Z. Intestinal Microbiota in Early Life and Its Implications on Childhood Health. Genomics Proteomics Bioinformatics 2019, 17, 13–25, doi:10.1016/j.gpb.2018.10.002. 15 36. Yang, I.; Corwin, E.J.; Brennan, P.A.; Jordan, S.; Murphy, J.R.; Dunlop, A. The Infant Microbiome: Implications for Infant Health and Neurocognitive Development. Nurs Res 2016, 65, 76–88, doi:10.1097/NNR.0000000000000133. 37. Mueller, N.T.; Bakacs, E.; Combellick, J.; Grigoryan, Z.; Dominguez-Bello, M.G. The Infant Microbiome Development: Mom Matters. Trends Mol Med 2015, 21, 109–117, doi:10.1016/j.molmed.2014.12.002. 38. Moore, R.E.; Townsend, S.D. Temporal Development of the Infant Gut Microbiome. Open Biology 9, 190128, doi:10.1098/rsob.190128. 39. Kim, H.; Sitarik, A.R.; Woodcroft, K.; Johnson, C.C.; Zoratti, E. Birth Mode, Breastfeeding, Pet Exposure, and Antibiotic Use: Associations With the Gut Microbiome and Sensitization in Children. Curr Allergy Asthma Rep 2019, 19, 22, doi:10.1007/s11882-019-0851-9. 40. Levin, A.M.; Sitarik, A.R.; Havstad, S.L.; Fujimura, K.E.; Wegienka, G.; Cassidy-Bushrow, A.E.; Kim, H.; Zoratti, E.M.; Lukacs, N.W.; Boushey, H.A.; et al. Joint Effects of Pregnancy, Sociocultural, and Environmental Factors on Early Life Gut Microbiome Structure and Diversity. Sci Rep 2016, 6, 31775, doi:10.1038/srep31775. 41. Ferretti, P.; Pasolli, E.; Tett, A.; Asnicar, F.; Gorfer, V.; Fedi, S.; Armanini, F.; Truong, D.T.; Manara, S.; Zolfo, M.; et al. Mother-to-Infant Microbial Transmission from Different Body Sites Shapes the Developing Infant Gut Microbiome. Cell Host Microbe 2018, 24, 133-145.e5, doi:10.1016/j.chom.2018.06.005. 42. Rutayisire, E.; Huang, K.; Liu, Y.; Tao, F. The Mode of Delivery Affects the Diversity and Colonization Pattern of the Gut Microbiota during the First Year of Infants’ Life: A Systematic Review. BMC Gastroenterology 2016, 16, 86, doi:10.1186/s12876-016-0498-0. 43. Azad, M.A.K.; Sarker, M.; Li, T.; Yin, J. Probiotic Species in the Modulation of Gut Microbiota: An Overview. Biomed Res Int 2018, 2018, 9478630, doi:10.1155/2018/9478630. 44. Hill, C.J.; Lynch, D.B.; Murphy, K.; Ulaszewska, M.; Jeffery, I.B.; O’Shea, C.A.; Watkins, C.; Dempsey, E.; Mattivi, F.; Tuohy, K.; et al. Evolution of Gut Microbiota Composition from Birth to 24 Weeks in the INFANTMET Cohort. Microbiome 2017, 5, 4, doi:10.1186/s40168- 016-0213-y. 45. Azad, M.B.; Konya, T.; Maughan, H.; Guttman, D.S.; Field, C.J.; Chari, R.S.; Sears, M.R.; Becker, A.B.; Scott, J.A.; Kozyrskyj, A.L. Gut Microbiota of Healthy Canadian Infants: Profiles by Mode of Delivery and Infant Diet at 4 Months. CMAJ 2013, 185, 385–394, doi:10.1503/cmaj.121189. 46. Ho, N.T.; Li, F.; Lee-Sarwar, K.A.; Tun, H.M.; Brown, B.P.; Pannaraj, P.S.; Bender, J.M.; Azad, M.B.; Thompson, A.L.; Weiss, S.T.; et al. Meta-Analysis of Effects of Exclusive Breastfeeding on Infant Gut Microbiota across Populations. Nat Commun 2018, 9, 4169, doi:10.1038/s41467-018-06473-x. 16 47. Fehr, K.; Moossavi, S.; Sbihi, H.; Boutin, R.C.T.; Bode, L.; Robertson, B.; Yonemitsu, C.; Field, C.J.; Becker, A.B.; Mandhane, P.J.; et al. Breastmilk Feeding Practices Are Associated with the Co-Occurrence of Bacteria in Mothers’ Milk and the Infant Gut: The CHILD Cohort Study. Cell Host Microbe 2020, 28, 285-297.e4, doi:10.1016/j.chom.2020.06.009. 48. Sugino, K.Y.; Ma, T.; Kerver, J.M.; Paneth, N.; Comstock, S.S. Human Milk Feeding Patterns at 6 Months of Age Are a Major Determinant of Fecal Bacterial Diversity in Infants. J Hum Lact 2020, 890334420957571, doi:10.1177/0890334420957571. 49. Moossavi, S.; Sepehri, S.; Robertson, B.; Bode, L.; Goruk, S.; Field, C.J.; Lix, L.M.; de Souza, R.J.; Becker, A.B.; Mandhane, P.J.; et al. Composition and Variation of the Human Milk Microbiota Are Influenced by Maternal and Early-Life Factors. Cell Host Microbe 2019, 25, 324-335.e4, doi:10.1016/j.chom.2019.01.011. 50. Walsh, C.; Lane, J.A.; van Sinderen, D.; Hickey, R.M. Human Milk Oligosaccharides: Shaping the Infant Gut Microbiota and Supporting Health. J Funct Foods 2020, 72, 104074, doi:10.1016/j.jff.2020.104074. 51. Lawson, M.A.E.; O’Neill, I.J.; Kujawska, M.; Gowrinadh Javvadi, S.; Wijeyesekera, A.; Flegg, Z.; Chalklen, L.; Hall, L.J. Breast Milk-Derived Human Milk Oligosaccharides Promote Bifidobacterium Interactions within a Single Ecosystem. ISME J 2020, 14, 635–648, doi:10.1038/s41396-019-0553-2. 52. Azad, M.B.; Konya, T.; Maughan, H.; Guttman, D.S.; Field, C.J.; Chari, R.S.; Sears, M.R.; Becker, A.B.; Scott, J.A.; Kozyrskyj, A.L.; et al. Gut Microbiota of Healthy Canadian Infants: Profiles by Mode of Delivery and Infant Diet at 4 Months. CMAJ 2013, 185, 385–394, doi:10.1503/cmaj.121189. 53. Stewart, C.J.; Ajami, N.J.; O’Brien, J.L.; Hutchinson, D.S.; Smith, D.P.; Wong, M.C.; Ross, M.C.; Lloyd, R.E.; Doddapaneni, H.; Metcalf, G.A.; et al. Temporal Development of the Gut Microbiome in Early Childhood from the TEDDY Study. Nature 2018, 562, 583–588, doi:10.1038/s41586-018-0617-x. 54. Moossavi, S.; Sepehri, S.; Robertson, B.; Bode, L.; Goruk, S.; Field, C.J.; Lix, L.M.; de Souza, R.J.; Becker, A.B.; Mandhane, P.J.; et al. Composition and Variation of the Human Milk Microbiota Are Influenced by Maternal and Early-Life Factors. Cell Host Microbe 2019, 25, 324-335.e4, doi:10.1016/j.chom.2019.01.011. 55. Voreades, N.; Kozil, A.; Weir, T.L. Diet and the Development of the Human Intestinal Microbiome. Frontiers in Microbiology 2014, 5, 494, doi:10.3389/fmicb.2014.00494. 56. Bergström, A.; Skov, T.H.; Bahl, M.I.; Roager, H.M.; Christensen, L.B.; Ejlerskov, K.T.; Mølgaard, C.; Michaelsen, K.F.; Licht, T.R. Establishment of Intestinal Microbiota during Early Life: A Longitudinal, Explorative Study of a Large Cohort of Danish Infants. Appl Environ Microbiol 2014, 80, 2889–2900, doi:10.1128/AEM.00342-14. 57. Vangay, P.; Ward, T.; Gerber, J.S.; Knights, D. Antibiotics, Pediatric Dysbiosis, and Disease. Cell Host Microbe 2015, 17, 553–564, doi:10.1016/j.chom.2015.04.006. 17 58. Tanaka, S.; Kobayashi, T.; Songjinda, P.; Tateyama, A.; Tsubouchi, M.; Kiyohara, C.; Shirakawa, T.; Sonomoto, K.; Nakayama, J. Influence of Antibiotic Exposure in the Early Postnatal Period on the Development of Intestinal Microbiota. FEMS Immunology & Medical Microbiology 2009, 56, 80–87, doi:10.1111/j.1574-695X.2009.00553.x. 59. Aloisio, I.; Mazzola, G.; Corvaglia, L.T.; Tonti, G.; Faldella, G.; Biavati, B.; Di Gioia, D. Influence of Intrapartum Antibiotic Prophylaxis against Group B Streptococcus on the Early Newborn Gut Composition and Evaluation of the Anti-Streptococcus Activity of Bifidobacterium Strains. Appl Microbiol Biotechnol 2014, 98, 6051–6060, doi:10.1007/s00253- 014-5712-9. 60. La Rosa, P.S.; Warner, B.B.; Zhou, Y.; Weinstock, G.M.; Sodergren, E.; Hall-Moore, C.M.; Stevens, H.J.; Bennett, W.E.; Shaikh, N.; Linneman, L.A.; et al. Patterned Progression of Bacterial Populations in the Premature Infant Gut. Proc Natl Acad Sci U S A 2014, 111, 12522– 12527, doi:10.1073/pnas.1409497111. 61. Fricke, W.F. The More the Merrier? Reduced Fecal Microbiota Diversity in Preterm Infants Treated with Antibiotics. J Pediatr 2014, 165, 8–10, doi:10.1016/j.jpeds.2014.03.022. 62. Cryan, J.F.; O’Riordan, K.J.; Cowan, C.S.M.; Sandhu, K.V.; Bastiaanssen, T.F.S.; Boehme, M.; Codagnone, M.G.; Cussotto, S.; Fulling, C.; Golubeva, A.V.; et al. The Microbiota-Gut- Brain Axis. Physiological Reviews 2019, 99, 1877–2013, doi:10.1152/physrev.00018.2018. 63. Carabotti, M.; Scirocco, A.; Maselli, M.A.; Severi, C. The Gut-Brain Axis: Interactions between Enteric Microbiota, Central and Enteric Nervous Systems. Ann Gastroenterol 2015, 28, 203–209. 64. Morais, L.H.; Schreiber, H.L.; Mazmanian, S.K. The Gut Microbiota–Brain Axis in Behaviour and Brain Disorders. Nat Rev Microbiol 2021, 19, 241–255, doi:10.1038/s41579-020- 00460-0. 65. Fülling, C.; Dinan, T.G.; Cryan, J.F. Gut Microbe to Brain Signaling: What Happens in Vagus…. Neuron 2019, 101, 998–1002, doi:10.1016/j.neuron.2019.02.008. 66. Berthoud, H.-R.; Neuhuber, W.L. Functional and Chemical Anatomy of the Afferent Vagal System. Autonomic Neuroscience 2000, 85, 1–17, doi:10.1016/S1566-0702(00)00215-0. 67. Breit, S.; Kupferberg, A.; Rogler, G.; Hasler, G. Vagus Nerve as Modulator of the Brain–Gut Axis in Psychiatric and Inflammatory Disorders. Front Psychiatry 2018, 9, 44, doi:10.3389/fpsyt.2018.00044. 68. Bravo, J.A.; Forsythe, P.; Chew, M.V.; Escaravage, E.; Savignac, H.M.; Dinan, T.G.; Bienenstock, J.; Cryan, J.F. Ingestion of Lactobacillus Strain Regulates Emotional Behavior and Central GABA Receptor Expression in a Mouse via the Vagus Nerve. PNAS 2011, 108, 16050– 16055, doi:10.1073/pnas.1102999108. 69. Cenit, M.C.; Sanz, Y.; Codoñer-Franch, P. Influence of Gut Microbiota on Neuropsychiatric Disorders. World J Gastroenterol 2017, 23, 5486–5498, doi:10.3748/wjg.v23.i30.5486. 18 70. Morais, L.H.; Schreiber, H.L.; Mazmanian, S.K. The Gut Microbiota–Brain Axis in Behaviour and Brain Disorders. Nat Rev Microbiol 2021, 19, 241–255, doi:10.1038/s41579-020- 00460-0. 71. Luczynski, P.; McVey Neufeld, K.-A.; Oriach, C.S.; Clarke, G.; Dinan, T.G.; Cryan, J.F. Growing up in a Bubble: Using Germ-Free Animals to Assess the Influence of the Gut Microbiota on Brain and Behavior. International Journal of Neuropsychopharmacology 2016, 19, doi:10.1093/ijnp/pyw020. 72. Neufeld, K.-A.M.; Kang, N.; Bienenstock, J.; Foster, J.A. Effects of Intestinal Microbiota on Anxiety-like Behavior. Commun Integr Biol 2011, 4, 492–494, doi:10.4161/cib.4.4.15702. 73. Heijtz, R.D.; Wang, S.; Anuar, F.; Qian, Y.; Björkholm, B.; Samuelsson, A.; Hibberd, M.L.; Forssberg, H.; Pettersson, S. Normal Gut Microbiota Modulates Brain Development and Behavior. PNAS 2011, 108, 3047–3052, doi:10.1073/pnas.1010529108. 74. Fujii, Y.; Nguyen, T.T.T.; Fujimura, Y.; Kameya, N.; Nakamura, S.; Arakawa, K.; Morita, H. Fecal Metabolite of a Gnotobiotic Mouse Transplanted with Gut Microbiota from a Patient with Alzheimer’s Disease. Bioscience, Biotechnology, and Biochemistry 2019, 83, 2144–2152, doi:10.1080/09168451.2019.1644149. 75. Sasmita, A.O. Modification of the Gut Microbiome to Combat Neurodegeneration. Reviews in the Neurosciences 2019, 30, 795–805, doi:10.1515/revneuro-2019-0005. 76. Tomova, A.; Husarova, V.; Lakatosova, S.; Bakos, J.; Vlkova, B.; Babinska, K.; Ostatnikova, D. Gastrointestinal Microbiota in Children with Autism in Slovakia. Physiol Behav 2015, 138, 179–187, doi:10.1016/j.physbeh.2014.10.033. 77. Xu, G.; Strathearn, L.; Liu, B.; O’Brien, M.; Kopelman, T.G.; Zhu, J.; Snetselaar, L.G.; Bao, W. Prevalence and Treatment Patterns of Autism Spectrum Disorder in the United States, 2016. JAMA Pediatrics 2019, 173, 153–159, doi:10.1001/jamapediatrics.2018.4208. 78. Coretti, L.; Paparo, L.; Riccio, M.P.; Amato, F.; Cuomo, M.; Natale, A.; Borrelli, L.; Corrado, G.; De Caro, C.; Comegna, M.; et al. Gut Microbiota Features in Young Children With Autism Spectrum Disorders. Front Microbiol 2018, 9, 3146, doi:10.3389/fmicb.2018.03146. 79. Kang, D.-W.; Park, J.G.; Ilhan, Z.E.; Wallstrom, G.; LaBaer, J.; Adams, J.B.; Krajmalnik- Brown, R. Reduced Incidence of Prevotella and Other Fermenters in Intestinal Microflora of Autistic Children. PLOS ONE 2013, 8, e68322, doi:10.1371/journal.pone.0068322. 80. Kang, D.-W.; Adams, J.B.; Gregory, A.C.; Borody, T.; Chittick, L.; Fasano, A.; Khoruts, A.; Geis, E.; Maldonado, J.; McDonough-Means, S.; et al. Microbiota Transfer Therapy Alters Gut Ecosystem and Improves Gastrointestinal and Autism Symptoms: An Open-Label Study. Microbiome 2017, 5, 10, doi:10.1186/s40168-016-0225-7. 81. Kang, D.-W.; Adams, J.B.; Coleman, D.M.; Pollard, E.L.; Maldonado, J.; McDonough- Means, S.; Caporaso, J.G.; Krajmalnik-Brown, R. Long-Term Benefit of Microbiota Transfer 19 Therapy on Autism Symptoms and Gut Microbiota. Sci Rep 2019, 9, 5821, doi:10.1038/s41598- 019-42183-0. 82. Carlson, A.L.; Xia, K.; Azcarate-Peril, M.A.; Goldman, B.D.; Ahn, M.; Styner, M.A.; Thompson, A.L.; Geng, X.; Gilmore, J.H.; Knickmeyer, R.C. Infant Gut Microbiome Associated With Cognitive Development. Biological Psychiatry 2018, 83, 148–159, doi:10.1016/j.biopsych.2017.06.021. 83. Kelsey, C.; Dreisbach, C.; Alhusen, J.; Grossmann, T. A Primer on Investigating the Role of the Microbiome in Brain and Cognitive Development. Developmental Psychobiology 2019, 61, 341–349, doi:10.1002/dev.21778. 84. Owens, J.A.; Spirito, A.; McGuinn, M.; Nobile, C. Sleep Habits and Sleep Disturbance in Elementary School-Aged Children. J Dev Behav Pediatr 2000, 21, 27–36, doi:10.1097/00004703-200002000-00005. 85. Pagel, J.F.; Forister, N.; Kwiatkowki, C. Adolescent Sleep Disturbance and School Performance: The Confounding Variable of Socioeconomics. J Clin Sleep Med 2007, 3, 19–23. 86. Liu, X.; Liu, L.; Owens, J.A.; Kaplan, D.L. Sleep Patterns and Sleep Problems among Schoolchildren in the United States and China. Pediatrics 2005, 115, 241–249, doi:10.1542/peds.2004-0815F. 87. Kraenz, S.; Fricke, L.; Wiater, A.; Mitschke, A.; Breuer, U.; Lehmkuhl, G. [Prevalence and stress factors of sleep disorders in children starting school]. Prax Kinderpsychol Kinderpsychiatr 2004, 53, 3–18. 88. Spruyt, K.; O’Brien, L.M.; Cluydts, R.; Verleye, G.B.; Ferri, R. Odds, Prevalence and Predictors of Sleep Problems in School-Age Normal Children. J Sleep Res 2005, 14, 163–176, doi:10.1111/j.1365-2869.2005.00458.x. 89. Liu, X.; Liu, L.; Owens, J.A.; Kaplan, D.L. Sleep Patterns and Sleep Problems among Schoolchildren in the United States and China. Pediatrics 2005, 115, 241–249, doi:10.1542/peds.2004-0815F. 90. Owens, J.; Group, A.S.W.; Adolescence, C.O. Insufficient Sleep in Adolescents and Young Adults: An Update on Causes and Consequences. Pediatrics 2014, 134, e921–e932, doi:10.1542/peds.2014-1696. 91. Leone, V.; Gibbons, S.M.; Martinez, K.; Hutchison, A.L.; Huang, E.Y.; Cham, C.M.; Pierre, J.F.; Heneghan, A.F.; Nadimpalli, A.; Hubert, N.; et al. Effects of Diurnal Variation of Gut Microbes and High-Fat Feeding on Host Circadian Clock Function and Metabolism. Cell Host Microbe 2015, 17, 681–689, doi:10.1016/j.chom.2015.03.006. 92. Paulsen, J.A.; Ptacek, T.S.; Carter, S.J.; Liu, N.; Kumar, R.; Hyndman, L.; Lefkowitz, E.J.; Morrow, C.D.; Rogers, L.Q. Gut Microbiota Composition Associated with Alterations in Cardiorespiratory Fitness and Psychosocial Outcomes among Breast Cancer Survivors. Support Care Cancer 2017, 25, 1563–1570, doi:10.1007/s00520-016-3568-5. 20 93. Valentini, F.; Evangelisti, M.; Arpinelli, M.; Di Nardo, G.; Borro, M.; Simmaco, M.; Villa, M.P. Gut Microbiota Composition in Children with Obstructive Sleep Apnoea Syndrome: A Pilot Study. Sleep Medicine 2020, 76, 140–147, doi:10.1016/j.sleep.2020.10.017. 94. Clarridge, J.E. Impact of 16S RRNA Gene Sequence Analysis for Identification of Bacteria on Clinical Microbiology and Infectious Diseases. Clin Microbiol Rev 2004, 17, 840–862, doi:10.1128/CMR.17.4.840-862.2004. 95. Johnson, J.S.; Spakowicz, D.J.; Hong, B.-Y.; Petersen, L.M.; Demkowicz, P.; Chen, L.; Leopold, S.R.; Hanson, B.M.; Agresta, H.O.; Gerstein, M.; et al. Evaluation of 16S RRNA Gene Sequencing for Species and Strain-Level Microbiome Analysis. Nat Commun 2019, 10, 5029, doi:10.1038/s41467-019-13036-1. 96. Nguyen, N.-P.; Warnow, T.; Pop, M.; White, B. A Perspective on 16S RRNA Operational Taxonomic Unit Clustering Using Sequence Similarity. npj Biofilms Microbiomes 2016, 2, 1–8, doi:10.1038/npjbiofilms.2016.4. 97. Wang, Q.; Garrity, G.M.; Tiedje, J.M.; Cole, J.R. Naïve Bayesian Classifier for Rapid Assignment of RRNA Sequences into the New Bacterial Taxonomy. Applied and Environmental Microbiology 2007, 73, 5261–5267, doi:10.1128/AEM.00062-07. 98. Caporaso, J.G.; Kuczynski, J.; Stombaugh, J.; Bittinger, K.; Bushman, F.D.; Costello, E.K.; Fierer, N.; Peña, A.G.; Goodrich, J.K.; Gordon, J.I.; et al. QIIME Allows Analysis of High- Throughput Community Sequencing Data. Nat Methods 2010, 7, 335–336, doi:10.1038/nmeth.f.303. 99. Schloss, P.D.; Westcott, S.L.; Ryabin, T.; Hall, J.R.; Hartmann, M.; Hollister, E.B.; Lesniewski, R.A.; Oakley, B.B.; Parks, D.H.; Robinson, C.J.; et al. Introducing Mothur: Open- Source, Platform-Independent, Community-Supported Software for Describing and Comparing Microbial Communities. Applied and Environmental Microbiology 2009, 75, 7537–7541, doi:10.1128/AEM.01541-09. 100. Callahan, B.J.; McMurdie, P.J.; Holmes, S.P. Exact Sequence Variants Should Replace Operational Taxonomic Units in Marker-Gene Data Analysis. ISME J 2017, 11, 2639–2643, doi:10.1038/ismej.2017.119. 101. Gao, B.; Chi, L.; Zhu, Y.; Shi, X.; Tu, P.; Li, B.; Yin, J.; Gao, N.; Shen, W.; Schnabl, B. An Introduction to Next Generation Sequencing Bioinformatic Analysis in Gut Microbiome Studies. Biomolecules 2021, 11, 530, doi:10.3390/biom11040530. 102. Reese, A.T.; Dunn, R.R. Drivers of Microbiome Biodiversity: A Review of General Rules, Feces, and Ignorance. mBio 9, e01294-18, doi:10.1128/mBio.01294-18. 103. Beck, J.; Schwanghart, W. Comparing Measures of Species Diversity from Incomplete Inventories: An Update. Methods in Ecology and Evolution 2010, 1, 38–44, doi:10.1111/j.2041- 210X.2009.00003.x. 21 104. Matsuoka, K.; Kanai, T. The Gut Microbiota and Inflammatory Bowel Disease. Semin Immunopathol 2015, 37, 47–55, doi:10.1007/s00281-014-0454-4. 105. Socolar, J.B.; Gilroy, J.J.; Kunin, W.E.; Edwards, D.P. How Should Beta-Diversity Inform Biodiversity Conservation? Trends in Ecology & Evolution 2016, 31, 67–80, doi:10.1016/j.tree.2015.11.005. 106. Yang, Z.; Xu, F.; Li, H.; He, Y. Beyond Samples: A Metric Revealing More Connections of Gut Microbiota between Individuals. Comput Struct Biotechnol J 2021, 19, 3930–3937, doi:10.1016/j.csbj.2021.07.009. 107. Anderson, M.J. A New Method for Non-Parametric Multivariate Analysis of Variance. Austral Ecology 2001, 26, 32–46, doi:10.1111/j.1442-9993.2001.01070.pp.x. 108. Anderson, M.J.; Willis, T.J. Canonical Analysis of Principal Coordinates: A Useful Method of Constrained Ordination for Ecology. Ecology 2003, 84, 511–525, doi:10.1890/0012- 9658(2003)084[0511:CAOPCA]2.0.CO;2. 109. Gillman, M.W.; Blaisdell, C.J. Environmental Influences on Child Health Outcomes, a Research Program of the National Institutes of Health. Curr Opin Pediatr 2018, 30, 260–262, doi:10.1097/MOP.0000000000000600. 110. Luthold, R.V.; Fernandes, G.R.; Franco-de-Moraes, A.C.; Folchetti, L.G.D.; Ferreira, S.R.G. Gut Microbiota Interactions with the Immunomodulatory Role of Vitamin D in Normal Individuals. Metabolism 2017, 69, 76–86, doi:10.1016/j.metabol.2017.01.007. 22 CHAPTER 2. SPECIFIC AIMS The overall goal of this dissertation is to determine how feeding practices and other factors shape the early development of the infant gut microbiota, and to study the gut-microbiota-brain axis by assessing the relationship between gut microbiota and infant temperament at 9 months, and infant sleep problems at 2 years. Aim 1: To determine the association between infant feeding practices and gut bacterial composition in early infancy, accounting for other factors affecting the gut microbiota. Hypothesis1a: We hypothesize that the gut microbiota of infants who are exclusively breastfed will have higher abundance of Bifidobacterial, Streptococcus, and Lactobacillus, and a lower alpha diversity, compared with partially breastfed infants and exclusively formula infants at 3-9 months of infant life. Hypothesis1b: We hypothesize that the composition of gut microbiota from partially breastfed shows a greater resemblance to the gut microbiota of exclusively breastfed infants than to the gut microbiota from exclusively formula infants. Hypothesis1c: We hypothesize that among the exclusively breastfed, the gut microbiota of infants who had been given a vitamin D supplement will display a lower alpha diversity and a lower abundance of Haemophilus. Aim 2: To determine whether gut microbiota at the age of 3 months is associated with infant temperament at the age of 9 months. Hypothesis2a: We hypothesize that gut microbiome with higher alpha diversity at 3 months of age will be associated with higher scores of negative emotionality (NEG), and lower scores of positive affect/surgency (PAS) and orienting and regulatory capacity (ORC) of infant temperament at the age of 9 months. 23 Hypothesis2b: We hypothesize that a gut microbiome with higher abundance of beneficial bacteria, such as Bifidobacterial, Streptococcus, and Lactobacillus will be associated with lower scores of NEG, and higher scores of PAS and ORC of infant temperament at the age of 9 months. Hypothesis2c: We hypothesize that gut microbiota clusters characterized by higher abundance of beneficial bacteria are associated with lower scores of NEG, and higher scores of PAS and ORC of infant temperament at the age of 9 months. Aim 3: Determine whether gut microbiota in early infancy is associated with infant sleeping disorders at the age of 2 years. Hypothesis3a: We hypothesize that a gut microbiome with higher alpha diversity in early infancy will be associated with a higher risk of difficulty initiating and maintaining sleep at the age of 2 years. Hypothesis3b: We hypothesize that a gut microbiome with higher abundance of beneficial bacteria, such as Bifidobacteria, Streptococcus, and Lactobacillus, will be associated with a higher risk of difficulty initiating and maintaining sleep at the age of 2 years. 24 CHAPTER 3. VITAMIN D SUPPLEMENTATION IN EXCLUSIVELY BREASTFED INFANTS IS ASSOCIATED WITH ALTERATIONS IN THE FECAL MICROBIOME 3.1 Abstract Breastfeeding and introduction of solid food are the two major components of infant feeding practices that influence gut microbiota composition in early infancy. But it is unclear whether additional factors influence the microbiota of infants either exclusively breastfed or not breastfed. We obtained 194 fecal samples from infants at 3-9 months of age, extracted DNA, and sequenced the V4 region of the 16S rRNA gene. Feeding practices and clinical information were collected by questionnaire and abstraction of birth certificates. The gut microbiota of infants who were exclusively breastfed displayed significantly lower Shannon diversity (p-adjust < 0.001) and different gut microbiota composition compared to infants who were not breastfed (p-value = 0.001). Among the exclusively breastfed infants, recipients of vitamin D supplements displayed significantly lower Shannon diversity (p-adjust = 0.007), and different gut microbiota composition structure than non-supplemented, breastfed infants (p-value = 0.02). MaAslin analysis identified microbial taxa that associated with breastfeeding and vitamin D supplementation. Breastfeeding and infant vitamin D supplement intake play an important role in shaping infant gut microbiota. 3.2 Introduction The gut microbiota has been considered an "invisible organ" of the human body, playing important roles in modulating host functions, including metabolism, digestion, and gut mucosal immune responses and integrity[1,2]. Dysbiosis of the gut microbiota may be associated with various adverse health outcomes in infants such as asthma, Crohn's disease, inflammatory bowel disease, and type 1 diabetes (T1D)[3–7]. The colonization of gut bacteria begins at birth and remains remarkably dynamic until about 2-3 years of age when more stable microbial profiles 25 begin to emerge[8,9]. In addition to the mode of delivery and antibiotic exposure, infant feeding practices are key factors in shaping early microbiota composition[10,11]. Recent studies have shown that gut microbial profiles in breastfed infants are significantly different from those in formula-fed infants and change rapidly after the transition from breastfeeding to formula or solid food[12–15]. The differences in gut microbiota composition observed between formula-fed and breastfed infants have, at least partially, been attributed to the absence of human milk oligosaccharides (HMOs) in infant formula[16]. Human milk is enriched with HMOs, which have been linked to beneficial bacteria in the gut microbiota[17,18]. The introduction of solid food represents another key factor influencing the composition of infant gut microbiota, producing an adult-type complex microbiome dominated by the phyla Bacteroidetes and Firmicutes[19,20]. Most gut microbiota research on infant nutrition to date has focused on breastfeeding and the introduction of solid food. Little is known about the effect of other dietary features within the two different feeding practices. It is recommended that babies who are breastfed exclusively should take vitamin D supplements every day due to the variability of vitamin D content in human breastmilk[21]. Because all infant formula in the United States is fortified with vitamin D, infants who are fed exclusively with formula usually do not need a vitamin D supplementation. Vitamin D not only prevents rickets, but also plays an important role in immune responses and metabolic processes that maintain the integrity of the gut epithelium[22– 24]. Human milk can also be provided by bottle, from banking of milk by the mother or from human milk banks. This indirect form of breastfeeding can lead to enrichment by environmental bacteria, such as Stenotrophomonas and Pseudomonadacea[25]. The water used to reconstitute 26 powdered infant formulas may also be an important exposure for infant health outcomes. Reconstituting formula with tap water can lead to excessive fluoride and lead intake [26,27]. Different water types (e.g., city water, well water, filtration systems) can be a source of varied bacterial composition. We sought to determine the association between maternal and infant characteristics and infant feeding practices and the gut microbiota profiles in 3- to 9-month-old infants. In addition, we analyzed the exclusively breastfed and non-breastfed infants separately to assess the impact of unique feeding practices features on the microbiotas of the infants in each group. 3.3 Materials and Methods 3.3.1 Study participants The study population was drawn from the Michigan Archive for Research on Child Health (MARCH) cohort[28], an ongoing population-based pregnancy and birth cohort set in Michigan’s lower peninsula. The purpose of the MARCH study is to store biological specimens and other health information in pregnancy and early life that can be used to better understand the causes of problems in pregnancy and optimize the health of children. This cohort is a component of a nation-wide study of child health called the Environmental influences on Child Health Outcomes (ECHO)[29]. Our analysis included mothers who provided informed consent for providing infant stool samples. During the MARCH 3 month phone interview mothers confirmed their interest in participating in this sample collection. Fecal collection kits were sent by mail. 194 fecal samples have so far been collected from singleton infants aged from 3 to 9 months old. The infants in this analysis were 3-9 months of age between 2018 and 2021. 3.3.2 Data collection 27 Several questionnaires were administered to mothers from the first prenatal visit through 9 months postpartum. The questionnaire at the first prenatal visit included demographic information about the mothers, their breastfeeding plans and many health-related practices and conditions as well as their estimated due date. Infant dietary feeding patterns, including breast milk and/or formula, detailed feeding practices, and complementary food intake, were collected at the same time as the fecal samples. Detailed information, including the infant's sex, birth weight, complications of pregnancy, mode of delivery (vaginal vs C-section), pre-pregnancy BMI and gestational age, was abstracted from the birth certificate. 3.3.3 Fecal microbiota analysis Once received in the lab, the fecal samples were aliquoted into sterile tubes and stored at -80°C. DNA was extracted following a modified version of the Human Microbiome Project’s protocol as described previously [30]. Barcoded primers were used to amplify the V4 region of the 16S rRNA gene following the mothur wet lab documentation. PCR amplification also followed the wet lab protocol outlined in the mothur documentation. The resulting 16S rRNA libraries were sequenced using 250 base pair Illumina MiSeq with V2 chemistry at the Michigan State University genomics core. After trimming, clean sequences were analyzed using the QIIME2 (2021.2 version) pipeline[31]. Demultiplexed sequences were further quality filtered and clustered using QIIME2's DADA2 plugin to generate the ASV table[32]. Unique amplicon sequence variants (ASVs) were assigned a taxonomy by the QIIME2 feature-classifier plugin, using the Silva 132 database at the similarity threshold of 99% (for 16S data)[33,34]. Samples were rarefied to 6,000 sequencing reads per sample, leaving 191 stool samples with 6,905 unique ASVs, and findings were summarized at the genus taxonomic level. 3.3.4 Statistical analysis 28 We used multivariate ordinal logistic regression models to estimate the association between pre- pregnancy BMI and breastfeeding practices, with adjustment for demographic variables and delivery mode. Gut microbiota is analyzed by alpha diversity (Chao1 and Shannon) and beta diversity (Bray– Curtis dissimilarity and Weighted UniFrac) using the "vegan" package in R[35]. The difference of alpha diversity and relative abundance of taxa between feeding practices groups were tested by Wilcoxon rank test and Kruskal–Wallis test with false discovery rate (FDR) correction for multiple comparisons. We assessed the influence of factors significantly associated with gut bacterial community structure by multivariate models using a Permutational Multivariate Analysis of Variance (PERMANOVA) with 999 permutations based on Bray-Curtis dissimilarities (adonis, R vegan package)[35,36]. PERMANOVA is non-parametric multivariate statistical test, with p-values obtained using appropriate distribution‐free permutation techniques. We used the multivariate association with linear models (MaAsLin) to identify associated microbiological taxa with the feeding practices and other related factors[37,38]. MaAsLin is a multivariate statistical framework that identifies associations between clinical metadata and microbial community abundance and provides both nominal p-values and FDR adjusted p-values (q-values) by Benjamini–Hochberg procedure[39]. Associations are considered significant when the q-value is below the threshold of 0.1. 3.4 Results 3.4.1 Participants and feeding practices We analyzed gut microbiome samples from 191 infants. The distribution of the age at stool collection was shown in Figure 3.1. In Table 3.1, maternal and infant characteristics are compared by breastfeeding status (exclusive breastfeeding, partial breastfeeding, and not 29 breastfeeding). During the week immediately preceding stool sample collection, 88 (46.1%) infants were fed exclusively with breast milk, 43 (22.5%) were fed partially with breast milk, and 60 (31.4%) were not fed with breast milk. The median age at the time of specimen collection was 3.8 months (range: 3.0 months – 9.3 months). Partially breastfed infants were more likely to be fed with complementary foods than those who were not breastfed (44.2% vs 35.0%, P > 0.4). Infants who were exclusively breastfed were more likely to be given vitamin D supplementation than partially breastfed or not-breastfed infants (39.8% vs 18.6% vs 1.7%, P<0.01). A higher proportion of mothers who practiced exclusive breastfeeding were of normal BMI (18.5- 25.0) prior to pregnancy comparing to those who practiced partial breastfeeding or who were not breastfeeding (50.0% vs 39.5% vs 26.7%, P<0.01). Mothers who practiced exclusive breastfeeding were more likely to have a college degree than women who partially breastfed or did not breastfeed (72.4% vs 61.9% vs 32.2%, P<0.01). In a multivariate model adjusted for maternal age, maternal educational level, pre-pregnancy BMI (continuous), delivery mode, and infant age, mothers with higher pre-pregnancy BMI were less likely to practice breastfeeding (OR = 0.95, CI:0.91-0.99, p-value = 0.01, Table 3.2), and mothers with higher educational level were more likely to practice breastfeeding (OR= 2.66, CI: 1.72 - 4.21, p-value < 0.001, Table 3.2). 30 Figure 3.1. Distribution of infant age (month) at stool sample collection 31 Table 3.1. Characteristics of the mothers and infants by breastfeeding status1 Exclusive Partial No P- Characteristic breastfeeding breastfeeding breastfeeding value (N=88) (N=43) (N=60) Infant age at sample collection 115.5(17.3) 135.9 (39.8) 126.3 (31.9) <0.01 (day), mean (SD) Baby had any antibiotics since 14 (15.9) 5 (11.6) 8 (13.3) 0.40 birth, n (%) Consumption of complementary food during 0 (0.0) 19 (44.2) 21 (35.0) <0.013 past 24 hours, n (%) 2 Infant probiotic supplement 4 (4.5) 1 (2.4) 3 (3.3) 0.90 during past 24 hours, n (%) Infant Vitamin D supplement 35 (39.8) 8 (18.6) 1 (1.7) <0.01 during past 24 hours, n (%) Delivery mode, n (%) Vaginal delivery 66 (75) 30 (69.8) 36 (60) 0.10 C-section 22 (25) 13 (30.2) 24 (40) Baby weight at delivery 3461 (551) 3336 (529) 3269 (598) 0.10 (gram), mean (SD) Baby sex, n (%) Male 43 (48.9) 22 (51.2) 30 (50) 0.98 Female 45 (51.1) 21 (48.8) 30 (50) Maternal pre-pregnancy BMI, n (%) <18.5 1 (1.1) 0 (0.0) 3 (5.0) 18.5-25 44 (50) 17 (39.5) 16 (26.7) <0.01 >25-30 24 (27.3) 11 (25.6) 12 (20.0) >30 19 (21.6) 15 (34.9) 29 (48.3) Maternal education level, n (%) Did not finish high 0 (0.0) 0 (0.0) 6 (10.2) school High school graduate or 4 (4.6) 4 (9.5) 21 (35.6) <0.01 GED Some college 20 (23.0) 12 (28.6) 13 (22.0) College graduate or more 63 (72.4) 26 (61.9) 19 (32.2) 1 Breastfeeding status information was collected at the time of fecal sample collection. Values are mean (SD) for continuous variables or n (%) for categorical variables. Difference by breastfeeding status was calculated using an ANOVA or chi-squared test. 2 Including probiotic supplement, kefir and kimchi 3 Post hoc analysis with Bonferroni adjustment showed significant difference in consumption of complementary food between partial breastfeeding and no breastfeeding groups. 32 Table 3.2. Association between breastfeeding and perinatal characteristics1 Proportional odds Maternal characteristics 95% CI p-value ratio Maternal age(year) 1.02 0.96-1.09 0.48 Maternal educational level 2.66 1.72-4.21 <0.001 Pre-pregnancy BMI (continuous) 0.95 0.91-0.99 0.01 Delivery mode (C-section vs 0.56 0.29-1.05 0.07 Vaginal) Infant age (day) 0.99 0.98-1.0 0.07 1 A multivariate ordinal logistic regression analysis was performed to assess the association. Variables in the model include maternal age, maternal education level, pre-pregnancy BMI, delivery mode, and infant age. 3.4.2 Gut microbiota analysis Fecal samples from infants who were exclusively breastfed displayed lower Shannon diversity than samples from those who were not breastfed (FDR adjusted p-value <0.01, Figure 3.2A). Samples from infants who were partially breastfed displayed Shannon diversity intermediate between the other two groups, but not significantly different from either. (FDR adjusted p-value = 0.9). Chao 1 index was not significantly different across the three groups (FDR adjusted p- value =1.0, Figure 3.2B). Among the exclusively breastfed, infants who had been given a vitamin D supplement during the previous 24 hours displayed lower Shannon index (FDR adjusted p-value < 0.01, Figure 3.2C) and lower Chao 1 index (FDR adjusted p-value = 0.6, Figure 3.2D) than those infants who were not given a vitamin D supplement. The multivariate model confirmed that fecal samples from infants who were not breastfed displayed significantly higher Shannon diversity than samples from those who were exclusively breastfed (Beta=0.18, 33 95% CI: 0.01, 0.34, p-value= 0.03, Table S3.1). The multivariate model also confirmed that fecal samples from infants who had been given a vitamin D supplement during the previous 24 hours displayed significantly lower Shannon index than those who were not given a vitamin D supplement (Beta=-0.27, 95% CI=-0.48, -0.08, p-value= 0.006, Table S3.2) in exclusively breastfed infants. All the multivariate models were adjusted for maternal age, maternal educational level, pre-pregnancy BMI (continuous), delivery mode, and infant age. Whether breast milk was fed directly or was pumped and fed to the infant using a bottle, the Shannon and Chao 1 indices of the infant gut microbiota alpha diversity were similar (FDR adjusted p-value =1.0 and 0.88, respectively, Figure S3.1). Among the non-breastfed infants, neither the water type used to reconstitute the formula nor the consumption of complementary food during past 24 hours was associated with gut microbiota alpha diversity as measured by the Shannon or Chao 1 indices (Figure S3.1). When classified by breastfeeding status, the gut microbiota communities of the infants were well separated in principal coordinate analysis (PCoA) based on the Bray–Curtis distance matrix (univariate PERMANOVA: R2= 4.1%, p-value <0.01, Figure 3.2E). In addition to the feeding practices, gestational age (R2=4.0%, p-value = 0.001) and delivery mode (R2=2.0%, p-value = 0.003) were significantly associated with overall gut microbiome composition (Table 3.3; multivariate PERMANOVA model on Bray-Curtis distances). The PERMANOVA results were consistent with results from the Weighted UniFrac distance metric (Table S3.3). We then repeated the PERMANOVA analysis on Bray-Curtis distances within exclusively breastfed and not breastfed infants separately. These subgroup analyses also included additional variables. Accordingly, delivery mode (R2 = 3.5%, P = 0.01) and infant vitamin D supplement in the past 24 hours (R2 = 3.4%, P = 0.02) were significantly associated with gut microbiota composition in 34 exclusively breastfed infants (Table S3.4). Among the not breastfed infants, only maternal education level (R2 = 4.1%, p-value = 0.02) was significantly associated with gut microbiota composition (Table S3.5). The PERMANOVA results of these two subgroup analyses were consistent with results from the Weighted UniFrac distance metrics (results not shown). Figure 3.2. Infant alpha and beta diversity by infant breastfeeding and Vitamin D supplement. FDR adjusted p-value for alpha diversity was displayed in upper-left. (A) The Shannon diversity was used for alpha diversity. All the participants were included in the analysis (N=191). Group differences were tested by Kruskal-Wallis test. We then performed post hoc test for multiple comparisons. After FDR adjustment, no breastfeeding group has significant difference with exclusive breastfeeding (adjusted p-value <0.01). Partial breastfeeding group has no significant difference with exclusive breastfeeding group (p-value= 0.4, adjusted p-value=0.9) and no breastfeeding group (p-value= 0.03, adjusted p-value=0.09). (B) The Chao 1 index was used for alpha diversity. All the participants were included in the analysis (N=191). Post hoc test didn’t find any significant difference between groups. (C) The Shannon diversity was used for alpha diversity. Only breastfeeding participants were included in this subgroup (N=92). Group differences were tested by Wilcoxon rank test. (D) The Chao 1 index was used for alpha diversity. Only breastfeeding participants were included in this subgroup analysis (N=92). (E) Principal component analysis (PCoA) ordinations of variation based on the Bray–Curtis distance matrix for all infants. R2 and p-value were calculated by univariate PERMANOVA test. 35 Table 3.3. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) for all infants F p- Variable value R2 value Breastfeeding during past week 4 4.10% 0.001* Gestational age 2.3 1.20% 0.03* Infant sex 1.1 0.50% 0.36 Delivery mode (vaginal vs C-section) 3.6 1.80% 0.004* Baby weight at delivery 0.7 0.30% 0.72 Infant probiotic supplement during past 24 hours 1 0.50% 0.38 Infant had any antibiotics since birth 0.9 1.00% 0.47 Maternal educational level 2.7 1.30% 0.02* Maternal pre-pregnancy BMI (continuous) 0.37 0.20% 0.96 1 Bray-Curtis distance was used for the PERMANOVA * indicates the p-value <0.05 We further assessed the association between breastfeeding status and relative abundance of 8 dominant genera by univariate analysis (Figure S3.2). These 8 dominant genera were Bacteroides, Bifidobacterium, Veillonella, Escherichia-Shigella, Ruminococcus gnavus, Clostridium sensu stricto 1, Prevotella, and Lachnoclostridium. Exclusive breastfeeding was significantly associated with a higher relative abundance of Bifidobacterium (FDR adjusted p- value= 5 × 10-5) and a lower relative abundance of Lachnoclostridium (FDR adjusted p-value =5.6 × 10-7). 36 MaAsLin results revealed that exclusive breastfeeding was significantly associated with the relative abundance of a set of genera, including Intestinibacter, Flavonifractor, Lachnoclostridium, Clostridium innocuum group, Lactobacillus, Bifidobacterium etc. (Table 3.4). Infant age at sample collection and maternal pre-pregnancy BMI were associated with higher relative abundance of Lachnospira and Alistipes, respectively (Table 3.4). Among exclusively breastfed infants, infants who had taken a vitamin D supplement in the previous 24 hours had a lower relative abundance of Haemophilus (Table 3.5). Among the not breastfed infants, having taken a probiotic supplement in the past 24 hours was associated with higher relative abundance of uncultured Lachnospiraceae and Faecalitalea (Table 3.5). Maternal pre- pregnancy BMI was associated with a higher relative abundance of Alistipes (Table 3.5). 37 Table 3.4. MaAsLin Analysis Results: Associations of infant feeding practices and gut microbiome taxa at genus level adjusted by covariates in all infants1 Taxonomy at Meta data N/N not Coefficient p-value q-value2 genus level value 0 Exclusive 3.6 × 10- 3.0 × 10- Intestinibacter -0.567 191/64 10 7 breastfeeding Exclusive 4.0 × 10- 1.7 × 10- Flavonifractor -0.896 191/145 8 5 breastfeeding Exclusive 1.9 × 10- 5.2 × 10- Lachnoclostridium -0.998 191/146 7 5 breastfeeding Clostridium Exclusive 4.9 × 10- 6.8 × 10- -0.393 19144 6 4 innocuum group breastfeeding Exclusive 4.3 × 10- 6.8 × 10- Lactobacillus 0.68 191/115 6 4 breastfeeding Exclusive 1.7 × 10- 1.7 × 10- Lactococcus -0.287 191/29 5 3 breastfeeding Exclusive 2.4 × 10- Bifidobacterium 0.535 191/186 4 0.018 breastfeeding Exclusive 2.4 × 10- Eisenbergiella -0.322 191/24 4 0.018 breastfeeding Exclusive 3.2 × 10- Colidextribacter -0.398 191/40 4 0.022 breastfeeding Exclusive 1.3 × 10- Akkermansia -0.55 191/124 3 0.066 breastfeeding Uncultured Exclusive 1.4 × 10- -0.188 191/20 3 0.069 Lachnospiraceae breastfeeding 38 Table 3.4. (cont’d) Exclusive 1.7 × 10- Haemophilus 0.463 191/141 0.073 breastfeeding 3 Exclusive 1.8× 10- Staphylococcus 0.293 191/56 3 0.073 breastfeeding Exclusive 1.6 × 10- Incertae_Sedis -0.358 191/99 3 0.073 breastfeeding Partial 7.2 × 10- 1.5 Flavonifractor -0.909 191/145 7 breastfeeding × 10-4 Partial 1.3 × 10- Haemophilus 0.74 191/141 5 0.001 breastfeeding Partial 5.8 × 10- Lachnoclostridium -0.86 191/146 5 0.005 breastfeeding Partial 5.6 × 10- Lactococcus -0.258 191/29 4 0.031 breastfeeding Pre- 4.0 × 10- Alistipes pregnancy 0.206 191/99 4 0.025 BMI Age at 5.7 × 10- Lachnospira sample 0.171 191/84 4 0.032 collection 1 Model was adjusted for infant antibiotic use, sex, infant birth weight, delivery mode, age at fecal sample collection, infant probiotic supplement and pre-pregnancy BMI. Not breastfeeding is the reference for the breastfeeding status in the regression model. 2 q-value is the FDR (Benjamini-Hochberg) adjusted p-value. q-value < 0.1 for multiple comparisons was considered statistically significant and included in the table. 39 Table 3.5. MaAsLin Analysis Results: Associations of infant feeding practices and gut microbiome taxa at genus level within exclusively breastfed and no breastfed infants1 Breastfeeding Taxonomy at N/N status genus level Meta data value Coefficient not 0 p-value q-value2 Exclusively Vitamin D Haemophilus supplement (Yes) -0.683 88/74 6.7 × 10-5 0.058 Probiotic Faecalitalea supplement (Yes) 1.718 60/9 2.9 × 10-9 2.4 × 10-6 Not Uncultured Probiotic breastfeeding Lachnospiraceae supplement (Yes) 1.269 60/10 5.1 × 10-5 0.021 Pre-pregnancy Alistipes BMI 0.431 60/27 2.6 × 10-4 0.072 1 Both Models were adjusted for infant antibiotic use, sex, infant birth weight, delivery mode, age at fecal sample collection, infant probiotic supplement and pre-pregnancy BMI. 2 q-value is the FDR (Benjamini-Hochberg) adjusted p-value. q-value < 0.1 for multiple comparisons was considered statistically significant and results were included in the table. 3.5 Discussion Our study was conducted in a population with somewhat higher than average rates of exclusive breastfeeding, with nearly half (46.1%) of the infants exclusively breastfed at the median age of 3.8 months. This percentage is higher than found in the Infant Feeding Practices Study II in the US ( 34% at 3 months) in 2007 [40], while it is closed to the percentage in the CDC National Immunization Survey in 2018 (46.3% at 3 month)[41]. Although the American Academy of Pediatrics recommends vitamin D supplementation for all breast-fed babies[42], only 39.8% of the exclusively breastfed infants in our study followed this recommendation, a lower frequency than that has been found in Canadian and European cohorts in 2009 and 2014, respectively[43,44]. 40 We found that higher maternal education level and lower pre-pregnancy BMI were independently and significantly associated with an increased odds of being exclusively breastfed, consistent with previous studies in developed and developing countries[40,45,46]. Breastfeeding initiation and duration are also negatively correlated with high pre-pregnancy BMI[47,48]. These associations may be attributed to the physiological factors such as delayed onset of lactogenesis II and imbalances of hormones[49]. Previous studies have showed that maternal obesity can cause the delayed onset of lactogenesis II (DOL), a hormonal process that is associated with mother's confidence that her milk is sufficient for her child[50,51]. As a result, it can lead to lower rates of breastfeeding initiation and early termination of exclusive breastfeeding. The associations between maternal BMI and lactation success have been recently reviewed[52,53]. Our study demonstrated the importance of both breastfeeding and infant vitamin D supplements in shaping infant gut microbiota composition. Breastfeeding is significantly associated with both alpha and beta diversity of infant gut microbiota. We observed that the Shannon diversity of partially breastfed infants was between that of exclusively breastfed infants and not breastfed infants, though somewhat closer to the exclusively breast fed, suggesting a dose-response relationship between breastfeeding and Shannon diversity of infant gut microbiota. These results agree with a previous study, which reported that the composition of gut microbiota from partially breastfed infants are similar to that from exclusively breastfed infants[13]. Among the subgroup analysis of exclusively breastfed infants, the alpha and beta diversity results demonstrated that vitamin D supplementation is associated with infant gut microbiota overall composition. These results are in good agreement with the study of Lei et al. who investigated the role of vitamin D supplement on gut microbiome from 31 exclusively breastfed infants at 4-months-old[54]. Animal studies demonstrate that vitamin D plays a critical role in maintaining the integrity of the 41 intestinal mucosal barrier by preserving the integrity of junctions that control mucosal permeability and reduction of pro-inflammatory cytokines such as IL-8[55–57]. In addition, studies also found that VDR-mediated signaling inhibits inflammation-induced apoptosis of intestinal epithelial cells[56,58]. As a result of these effects on the intestinal mucosa, vitamin D acts as an important factor influencing the gut microbiota. Besides the infant feeding practices, the results herein confirm that gestational age, delivery mode and maternal educational level are also significantly associated with gut microbiome composition. These results are consistent with many previous studies[10,13,59]. However, delivery mode is only associated with gut microbiota composition among the exclusively breastfed infants, while no significant association was found in the not breastfed infants. This might be attributed to the fact that c-section delivery can delay lactation initiation[60] and shape the bacterial composition of breast milk[61,62]. Maternal educational level is the only factor that significantly associated with infant gut microbiota composition among the not breastfed infants, whereas it was not significant among the exclusively breastfed infants. This finding suggests that not breastfed infants are more susceptible to socio-economic factors, such as educational level, which is normally be connected to offspring diet and nutritional status[63]. Hence, this association among the not breastfed infants might be mediated by the types of solid food introduction and quality of formula purchased. However, our data set did not allow us to test these associations. Our study confirmed that Bifidobacterium was enriched in breastfed infants when compared with non-breastfed infants. Lower abundance of Bifidobacterium in infants due to early cessation of breastfeeding could potentially inhibit the interaction of bifidobacterial-mediated metabolites with the immune system, leading to higher levels of inflammation[64,65]. In contrast, the genus 42 Lachnoclostridium (Lachnospiraceae family) was found to be enriched in the non-breastfed infants when compared with exclusively breastfed or partially breastfed infants. In addition, the genera Eisenbergiella and Lachnospiraceae_uncultured, which also belong to Lachnospiraceae family were found to be enriched in the non-breastfed infants by MaAslin. These observations agree with previous studies that lower abundance of Lachnospiraceae is associated with breastfeeding at 3 months of age[66]. The evidence from many studies showed that Lachnospiraceae family or specific genera of Lachnospiraceae may be associated with several inflammatory conditions, such as metabolic syndrome, obesity, diabetes, and liver diseases[67– 70]. Notably, the genus Haemophilus was enriched in the breastfed infants. However, exclusively breastfed infants who had taken a vitamin D supplement in the past 24 hours had a lower relative abundance of Haemophilus compared to those exclusively breastfed infants who were not supplemented. Consistent with our study, Fehr et al. showed that breastmilk may specifically provide Haemophilus to the infant gut[13]. Luthold et al also demonstrated that Haemophilus was less abundant in the group of highest vitamin D intake[71], supporting the hypothesis that a reduced immune response in vitamin D deficiency could augment the competitive advantage of Haemophilus and influence the composition of the infant gut microbiome[72]. Our study did not detect any effect of feeding with expressed milk, infant antibiotic intake, and water type for formula on gut microbiome. However, sample sizes were small for many of these comparisons. Therefore, pooling data from multiple cohort studies or analysis in larger cohorts with similar data are necessary to confirm this lack of association. For instance, previous studies demonstrated that infants who had been exposed to antibiotics had decreased abundance of Bifidobacteria and Bacteroides in the infant gut microbiome[73]. In our study, we asked the 43 mothers if the infant had taken any antibiotics since birth, whereas the timing of antibiotics administration was unknown. Hence, it's possible that infant gut microbiome had recovered from the dysbiosis states caused by antibiotics at the time of stool sample collection. The inconsistent results may also be attributed to variations in the antibiotic type, dosage, duration[74]. An important limitation of this study is that only a single stool sample was available for analysis. Although the infant feeding practice information was collected at the same time as the stool sample collection and can demonstrate the impact of short-term exposures on the infant gut microbiota composition, we are unable to determine how these factors contribute to the temporal development of the infant gut microbiome. Also, we did not collect more detailed information, such as dose of vitamin D, maternal vitamin D status, timing of antibiotic administration. Future studies would benefit from a longitudinal stool sample collection during infancy and a more detailed infant feeding practices questionnaire that not only collects proximal but also long-term data about infant nutritional exposures. 44 APPENDIX 45 Figure S3.1. Infant alpha diversity by infant different feeding practices. FDR adjusted p- value for alpha diversity was displayed in upper-left. (A) The Shannon diversity was used for alpha diversity. Only exclusively breastfed infants were included in this subgroup analysis (N=92). Group differences were tested by Wilcoxon rank test. (B) The Chao 1 index was used for alpha diversity. Only exclusively breastfed infants were included in this subgroup analysis (N=92). (C) The Shannon diversity was used for alpha diversity. Only no breastfed infants were included in this subgroup (N=60). Group differences were tested by Kruskal-Wallis test. We then performed post hoc test for multiple comparisons and no significant associations was found. (D) The Chao 1 index was used for alpha diversity. Only no breastfed infants were included in this subgroup (N=60). Group differences were tested by Kruskal-Wallis test. We then performed post hoc test for multiple comparisons and no significant associations was found (N=92). (E) The Shannon diversity was used for alpha diversity. Only no breastfed infants were included in this subgroup (N=60). Group differences were tested by Kruskal-Wallis test. We then performed post hoc test for multiple comparisons and no significant associations was found. (F) The Chao 1 index was used for alpha diversity. Only no breastfed infants were included in this subgroup (N=60). Group differences were tested by Kruskal-Wallis test. We then performed post hoc test for multiple comparisons and no significant associations was found (N=60). 46 Table S3.1. Association between Shannon/Chao1 index and breastfeeding status in all infants1 Shannon index Chao 1 index Beta (95% CI) p-value Beta (95% CI) p- value Breastfeeding status Exclusive ref - ref - breastfeeding Partial breastfeeding 0.04 (-0.12, 0.20) 0.65 -0.32(-12.6, 6.3) 0.52 No breastfeeding 0.18 (0.01, 0.34) 0.03 0.02 (-9.6, 9.68) 0.98 1 A multivariate linear regression regression analysis was performed to assess the association. Covariates in the models include maternal age, maternal education level, pre-pregnancy BMI, delivery mode, and infant age. Table S3.2. Association between Shannon/Chao1 index and infant vitamin D supplement intake1 Shannon index Chao 1 index Beta (95% CI) p-value Beta (95% CI) p- value Vitamin D supplement for infant No ref - ref - Yes -0.27 (-0.48, - 0.006 -10.8 (-22.08, 0.54) 0.06 0.08) 1 A multivariate linear regression regression analysis was performed to assess the association. Covariates in the models include maternal age, maternal education level, pre- pregnancy BMI, delivery mode, and infant age. 47 Figure S3.2. Association between breastfeeding status and relative abundance of 8 dominant genera. Group differences were tested by Kruskal-Wallis test. Wilcoxon rank test was used for post hoc test. P-value was adjusted by Bonferroni correction. * indicates the adjusted p-value <0.05 NS indicates the adjusted p-value ≥ 0.05 48 Table S3.3. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) on weighted UniFrac distances F p- 2 Variable value R value Breastfeeding during past week 4.8 4.70% 0.001* Gestational age 4.1 2.10% 0.007* Infant sex 1.7 0.80% 0.14 Delivery mode (virginal vs C-section) 6.0 3.00% 0.001* Baby weight at delivery 0.4 0.20% 0.85 Infant probiotic supplement during past 24 hours 0.7 0.40% 0.57 Infant had any antibiotics since birth 0.9 0.90% 0.50 Maternal educational level 2.3 1.10% 0.06 Maternal pre-pregnancy BMI (continuous) 0.3 0.10% 0.95 * indicates the p-value <0.05 Table S3.4. Results of Permutational Multivariate Analysis of Variance (PERMANOVA) for exclusively breastfed infants on Bray-Curtis distances F p- Variable value R2 value Gestational age 1.1 1.20% 0.31 Infant sex 0.9 1.00% 0.46 Delivery mode (virginal vs C-section) 3.0 3.50% 0.01* Baby weight at delivery 0.3 0.40% 0.92 Infant probiotic supplement during past 24 hours 1.0 1.10% 0.41 Infant had any antibiotics since birth 0.8 0.90% 0.54 Maternal educational level 0.7 0.80% 0.67 Maternal pre-pregnancy BMI (continuous) 0.3 0.30% 0.95 Infant vitamin D supplement in the past 24 hours 2.9 3.40% 0.02* Feeding with expressed breast milk 0.3 0.30% 0.95 * indicates the p-value <0.05 49 Table S3.5. Results of Permutational Multivariate Analysis of Variance (PERMANOVA)1 for not breastfed infants on Bray-Curtis distances F p- 2 Variable value R value Gestational age 1.1 2.00% 0.31 Infant sex 1.2 2.10% 0.29 Delivery mode (virginal vs C-section) 1.3 2.30% 0.21 Baby weight at delivery 0.4 0.70% 0.94 Infant probiotic supplement during past 24 hours 0.7 1.20% 0.68 Infant had any antibiotics since birth 0.4 0.70% 0.95 Maternal educational level 2.4 4.10% 0.02* Maternal pre-pregnancy BMI (continuous) 0.6 1.10% 0.75 Water type for formula 1.1 7.40% 0.39 Introduction of complementary food 0.8 1.40% 0.66 * indicates the p-value <0.05 50 REFERENCES 51 REFERENCES 1. Guinane, C.M.; Cotter, P.D. Role of the Gut Microbiota in Health and Chronic Gastrointestinal Disease: Understanding a Hidden Metabolic Organ. Therap Adv Gastroenterol 2013, 6, 295–308, doi:10.1177/1756283X13482996. 2. Kau, A.L.; Ahern, P.P.; Griffin, N.W.; Goodman, A.L.; Gordon, J.I. Human Nutrition, the Gut Microbiome and the Immune System. Nature 2011, 474, 327–336, doi:10.1038/nature10213. 3. Arrieta, M.-C.; Stiemsma, L.T.; Dimitriu, P.A.; Thorson, L.; Russell, S.; Yurist-Doutsch, S.; Kuzeljevic, B.; Gold, M.J.; Britton, H.M.; Lefebvre, D.L.; et al. Early Infancy Microbial and Metabolic Alterations Affect Risk of Childhood Asthma. Science Translational Medicine 2015, 7, 307ra152-307ra152, doi:10.1126/scitranslmed.aab2271. 4. Gevers, D.; Kugathasan, S.; Denson, L.A.; Vázquez-Baeza, Y.; Van Treuren, W.; Ren, B.; Schwager, E.; Knights, D.; Song, S.J.; Yassour, M.; et al. The Treatment-Naive Microbiome in New-Onset Crohn’s Disease. Cell Host & Microbe 2014, 15, 382–392, doi:10.1016/j.chom.2014.02.005. 5. Vatanen, T.; Franzosa, E.A.; Schwager, R.; Tripathi, S.; Arthur, T.D.; Vehik, K.; Lernmark, Å.; Hagopian, W.A.; Rewers, M.J.; She, J.-X.; et al. The Human Gut Microbiome in Early-Onset Type 1 Diabetes from the TEDDY Study. Nature 2018, 562, 589–594, doi:10.1038/s41586-018- 0620-2. 6. Tilg, H.; Moschen, A.R. Microbiota and Diabetes: An Evolving Relationship. Gut 2014, 63, 1513–1521, doi:10.1136/gutjnl-2014-306928. 7. Manichanh, C.; Borruel, N.; Casellas, F.; Guarner, F. The Gut Microbiota in IBD. Nat Rev Gastroenterol Hepatol 2012, 9, 599–608, doi:10.1038/nrgastro.2012.152. 8. Tanaka, M.; Nakayama, J. Development of the Gut Microbiota in Infancy and Its Impact on Health in Later Life. Allergology International 2017, 66, 515–522, doi:10.1016/j.alit.2017.07.010. 9. Rodríguez, J.M.; Murphy, K.; Stanton, C.; Ross, R.P.; Kober, O.I.; Juge, N.; Avershina, E.; Rudi, K.; Narbad, A.; Jenmalm, M.C.; et al. The Composition of the Gut Microbiota throughout Life, with an Emphasis on Early Life. Microb Ecol Health Dis 2015, 26, 10.3402/mehd.v26.26050, doi:10.3402/mehd.v26.26050. 10. Bokulich, N.A.; Chung, J.; Battaglia, T.; Henderson, N.; Jay, M.; Li, H.; Lieber, A.; Wu, F.; Perez-Perez, G.I.; Chen, Y.; et al. Antibiotics, Birth Mode, and Diet Shape Microbiome Maturation during Early Life. Sci Transl Med 2016, 8, 343ra82, doi:10.1126/scitranslmed.aad7121. 11. Cortes-Macías, E.; Selma-Royo, M.; García-Mantrana, I.; Calatayud, M.; González, S.; Martínez-Costa, C.; Collado, M.C. Maternal Diet Shapes the Breast Milk Microbiota 52 Composition and Diversity: Impact of Mode of Delivery and Antibiotic Exposure. The Journal of Nutrition 2021, 151, 330–340, doi:10.1093/jn/nxaa310. 12. Ho, N.T.; Li, F.; Lee-Sarwar, K.A.; Tun, H.M.; Brown, B.P.; Pannaraj, P.S.; Bender, J.M.; Azad, M.B.; Thompson, A.L.; Weiss, S.T.; et al. Meta-Analysis of Effects of Exclusive Breastfeeding on Infant Gut Microbiota across Populations. Nat Commun 2018, 9, 4169, doi:10.1038/s41467-018-06473-x. 13. Fehr, K.; Moossavi, S.; Sbihi, H.; Boutin, R.C.T.; Bode, L.; Robertson, B.; Yonemitsu, C.; Field, C.J.; Becker, A.B.; Mandhane, P.J.; et al. Breastmilk Feeding Practices Are Associated with the Co-Occurrence of Bacteria in Mothers’ Milk and the Infant Gut: The CHILD Cohort Study. Cell Host & Microbe 2020, 28, 285-297.e4, doi:10.1016/j.chom.2020.06.009. 14. Haddad, E.N.; Sugino, K.Y.; Kerver, J.M.; Paneth, N.; Comstock, S.S. The Infant Gut Microbiota at 12 Months of Age Is Associated with Human Milk Exposure but Not with Maternal Pre-Pregnancy Body Mass Index or Infant BMI-for-Age z-Scores. Current Research in Physiology 2021, 4, 94–102, doi:10.1016/j.crphys.2021.03.004. 15. Sugino, K.Y.; Ma, T.; Kerver, J.M.; Paneth, N.; Comstock, S.S. Human Milk Feeding Patterns at 6 Months of Age Are a Major Determinant of Fecal Bacterial Diversity in Infants. J Hum Lact 2020, 0890334420957571, doi:10.1177/0890334420957571. 16. Moossavi, S.; Sepehri, S.; Robertson, B.; Bode, L.; Goruk, S.; Field, C.J.; Lix, L.M.; de Souza, R.J.; Becker, A.B.; Mandhane, P.J.; et al. Composition and Variation of the Human Milk Microbiota Are Influenced by Maternal and Early-Life Factors. Cell Host Microbe 2019, 25, 324-335.e4, doi:10.1016/j.chom.2019.01.011. 17. Walsh, C.; Lane, J.A.; van Sinderen, D.; Hickey, R.M. Human Milk Oligosaccharides: Shaping the Infant Gut Microbiota and Supporting Health. J Funct Foods 2020, 72, 104074, doi:10.1016/j.jff.2020.104074. 18. Lawson, M.A.E.; O’Neill, I.J.; Kujawska, M.; Gowrinadh Javvadi, S.; Wijeyesekera, A.; Flegg, Z.; Chalklen, L.; Hall, L.J. Breast Milk-Derived Human Milk Oligosaccharides Promote Bifidobacterium Interactions within a Single Ecosystem. ISME J 2020, 14, 635–648, doi:10.1038/s41396-019-0553-2. 19. Tanaka, M.; Nakayama, J. Development of the Gut Microbiota in Infancy and Its Impact on Health in Later Life. Allergology International 2017, 66, 515–522, doi:10.1016/j.alit.2017.07.010. 20. Stewart, C.J.; Ajami, N.J.; O’Brien, J.L.; Hutchinson, D.S.; Smith, D.P.; Wong, M.C.; Ross, M.C.; Lloyd, R.E.; Doddapaneni, H.; Metcalf, G.A.; et al. Temporal Development of the Gut Microbiome in Early Childhood from the TEDDY Study. Nature 2018, 562, 583–588, doi:10.1038/s41586-018-0617-x. 21. CDC Vitamin D Is Needed to Support Healthy Bone Development. Available online: https://www.cdc.gov/breastfeeding/breastfeeding-special-circumstances/diet-and- micronutrients/vitamin-d.html (accessed on 27 July 2021). 53 22. Aranow, C. Vitamin D and the Immune System. Journal of Investigative Medicine 2011, 59, 881–886, doi:10.2310/JIM.0b013e31821b8755. 23. Borges, M.C.; Martini, L.A.; Rogero, M.M. Current Perspectives on Vitamin D, Immune System, and Chronic Diseases. Nutrition 2011, 27, 399–404, doi:10.1016/j.nut.2010.07.022. 24. Kong, J.; Zhang, Z.; Musch, M.W.; Ning, G.; Sun, J.; Hart, J.; Bissonnette, M.; Li, Y.C. Novel Role of the Vitamin D Receptor in Maintaining the Integrity of the Intestinal Mucosal Barrier. American Journal of Physiology-Gastrointestinal and Liver Physiology 2008, 294, G208–G216, doi:10.1152/ajpgi.00398.2007. 25. Moossavi, S.; Sepehri, S.; Robertson, B.; Bode, L.; Goruk, S.; Field, C.J.; Lix, L.M.; de Souza, R.J.; Becker, A.B.; Mandhane, P.J.; et al. Composition and Variation of the Human Milk Microbiota Are Influenced by Maternal and Early-Life Factors. Cell Host & Microbe 2019, 25, 324-335.e4, doi:10.1016/j.chom.2019.01.011. 26. Till, C.; Green, R.; Flora, D.; Hornung, R.; Martinez-Mier, E.A.; Blazer, M.; Farmus, L.; Ayotte, P.; Muckle, G.; Lanphear, B. Fluoride Exposure from Infant Formula and Child IQ in a Canadian Birth Cohort. Environment International 2020, 134, 105315, doi:10.1016/j.envint.2019.105315. 27. Triantafyllidou, S.; Edwards, M. Lead (Pb) in Tap Water and in Blood: Implications for Lead Exposure in the United States. Critical Reviews in Environmental Science and Technology 2012, 42, 1297–1352, doi:10.1080/10643389.2011.556556. 28. About | CHARM Study Available online: https://www.epi.msu.edu/charmstudy/about (accessed on 30 September 2021). 29. Paneth, N.; Monk, C. THE IMPORTANCE OF COHORT RESEARCH STARTING EARLY IN LIFE TO UNDERSTANDING CHILD HEALTH. Curr Opin Pediatr 2018, 30, 292–296, doi:10.1097/MOP.0000000000000596. 30. Sugino, K.Y.; Paneth, N.; Comstock, S.S. Michigan Cohorts to Determine Associations of Maternal Pre-Pregnancy Body Mass Index with Pregnancy and Infant Gastrointestinal Microbial Communities: Late Pregnancy and Early Infancy. PLOS ONE 2019, 14, e0213733, doi:10.1371/journal.pone.0213733. 31. Bolyen, E.; Rideout, J.R.; Dillon, M.R.; Bokulich, N.A.; Abnet, C.C.; Al-Ghalith, G.A.; Alexander, H.; Alm, E.J.; Arumugam, M.; Asnicar, F.; et al. Reproducible, Interactive, Scalable and Extensible Microbiome Data Science Using QIIME 2. Nat Biotechnol 2019, 37, 852–857, doi:10.1038/s41587-019-0209-9. 32. Callahan, B.J.; McMurdie, P.J.; Rosen, M.J.; Han, A.W.; Johnson, A.J.A.; Holmes, S.P. DADA2: High-Resolution Sample Inference from Illumina Amplicon Data. Nat Methods 2016, 13, 581–583, doi:10.1038/nmeth.3869. 33. Bokulich, N.A.; Kaehler, B.D.; Rideout, J.R.; Dillon, M.; Bolyen, E.; Knight, R.; Huttley, G.A.; Gregory Caporaso, J. Optimizing Taxonomic Classification of Marker-Gene Amplicon 54 Sequences with QIIME 2’s Q2-Feature-Classifier Plugin. Microbiome 2018, 6, 90, doi:10.1186/s40168-018-0470-z. 34. Quast, C.; Pruesse, E.; Yilmaz, P.; Gerken, J.; Schweer, T.; Yarza, P.; Peplies, J.; Glöckner, F.O. The SILVA Ribosomal RNA Gene Database Project: Improved Data Processing and Web- Based Tools. Nucleic Acids Research 2013, 41, D590–D596, doi:10.1093/nar/gks1219. 35. Dixon, P. VEGAN, a Package of R Functions for Community Ecology. Journal of Vegetation Science 2003, 14, 927–930, doi:10.1111/j.1654-1103.2003.tb02228.x. 36. Anderson, M.J. A New Method for Non-Parametric Multivariate Analysis of Variance. Austral Ecology 2001, 26, 32–46, doi:10.1111/j.1442-9993.2001.01070.pp.x. 37. Morgan, X.C.; Tickle, T.L.; Sokol, H.; Gevers, D.; Devaney, K.L.; Ward, D.V.; Reyes, J.A.; Shah, S.A.; LeLeiko, N.; Snapper, S.B.; et al. Dysfunction of the Intestinal Microbiome in Inflammatory Bowel Disease and Treatment. Genome Biology 2012, 13, R79, doi:10.1186/gb- 2012-13-9-r79. 38. Mallick, H.; Rahnavard, A.; McIver, L.J.; Ma, S.; Zhang, Y.; Nguyen, L.H.; Tickle, T.L.; Weingart, G.; Ren, B.; Schwager, E.H.; et al. Multivariable Association Discovery in Population- Scale Meta-Omics Studies. bioRxiv 2021, 2021.01.20.427420, doi:10.1101/2021.01.20.427420. 39. Thissen, D.; Steinberg, L.; Kuang, D. Quick and Easy Implementation of the Benjamini- Hochberg Procedure for Controlling the False Positive Rate in Multiple Comparisons. Journal of Educational and Behavioral Statistics 2002, 27, 77–83, doi:10.3102/10769986027001077. 40. Nnebe-Agumadu, U.H.; Racine, E.F.; Laditka, S.B.; Coffman, M.J. Associations between Perceived Value of Exclusive Breastfeeding among Pregnant Women in the United States and Exclusive Breastfeeding to Three and Six Months Postpartum: A Prospective Study. Int Breastfeed J 2016, 11, 8, doi:10.1186/s13006-016-0065-x. 41. CDC Results: Breastfeeding Rates Available online: https://www.cdc.gov/breastfeeding/data/nis_data/results.html (accessed on 1 October 2021). 42. Gartner, L.M.; Greer, F.R.; Breastfeeding, S. on; Nutrition, C. on Prevention of Rickets and Vitamin D Deficiency: New Guidelines for Vitamin D Intake. Pediatrics 2003, 111, 908–910, doi:10.1542/peds.111.4.908. 43. Uday, S.; Kongjonaj, A.; Aguiar, M.; Tulchinsky, T.; Högler, W. Variations in Infant and Childhood Vitamin D Supplementation Programmes across Europe and Factors Influencing Adherence. Endocrine Connections 2017, 6, 667–675, doi:10.1530/EC-17-0193. 44. Aghajafari, F.; Field, C.J.; Weinberg, A.R.; Letourneau, N.; APrON Study Team Both Mother and Infant Require a Vitamin D Supplement to Ensure That Infants’ Vitamin D Status Meets Current Guidelines. Nutrients 2018, 10, 429, doi:10.3390/nu10040429. 55 45. Flores, T.R.; Mielke, G.I.; Wendt, A.; Nunes, B.P.; Bertoldi, A.D. Prepregnancy Weight Excess and Cessation of Exclusive Breastfeeding: A Systematic Review and Meta-Analysis. Eur J Clin Nutr 2018, 72, 480–488, doi:10.1038/s41430-017-0073-y. 46. Tao, X.-Y.; Huang, K.; Yan, S.-Q.; Zuo, A.-Z.; Tao, R.-W.; Cao, H.; Gu, C.-L.; Tao, F.-B. Pre-Pregnancy BMI, Gestational Weight Gain and Breast-Feeding: A Cohort Study in China. Public Health Nutrition 2017, 20, 1001–1008, doi:10.1017/S1368980016003165. 47. Thompson, L.A.; Zhang, S.; Black, E.; Das, R.; Ryngaert, M.; Sullivan, S.; Roth, J. The Association of Maternal Pre-Pregnancy Body Mass Index with Breastfeeding Initiation. Matern Child Health J 2013, 17, 1842–1851, doi:10.1007/s10995-012-1204-7. 48. Amir, L.H.; Donath, S. A Systematic Review of Maternal Obesity and Breastfeeding Intention, Initiation and Duration. BMC Pregnancy Childbirth 2007, 7, 9, doi:10.1186/1471- 2393-7-9. 49. Bever Babendure, J.; Reifsnider, E.; Mendias, E.; Moramarco, M.W.; Davila, Y.R. Reduced Breastfeeding Rates among Obese Mothers: A Review of Contributing Factors, Clinical Considerations and Future Directions. Int Breastfeed J 2015, 10, 21, doi:10.1186/s13006-015- 0046-5. 50. Chapman, D.J.; Pérez-escamilla, R. Identification of Risk Factors for Delayed Onset of Lactation. Journal of the American Dietetic Association 1999, 99, 450–454, doi:10.1016/S0002- 8223(99)00109-1. 51. Lovelady, C.A. Is Maternal Obesity a Cause of Poor Lactation Performance? Nutrition Reviews 2005, 63, 352–355, doi:10.1111/j.1753-4887.2005.tb00113.x. 52. Robinson, D.T.; Josefson, J.; Van Horn, L. Considerations for Preterm Human Milk Feedings When Caring for Mothers Who Are Overweight or Obese. Adv Neonatal Care 2019, 19, 361– 370, doi:10.1097/ANC.0000000000000650. 53. Knight, C.H. An Endocrine Hypothesis to Explain Obesity-Related Lactation Insufficiency in Breastfeeding Mothers. J Dairy Res 2020, 87, 78–81, doi:10.1017/S0022029920000047. 54. Lei, W.-T.; Huang, K.-Y.; Jhong, J.-H.; Chen, C.-H.; Weng, S.-L. Metagenomic Analysis of the Gut Microbiome Composition Associated with Vitamin D Supplementation in Taiwanese Infants. Sci Rep 2021, 11, 2856, doi:10.1038/s41598-021-82584-8. 55. Kong, J.; Zhang, Z.; Musch, M.W.; Ning, G.; Sun, J.; Hart, J.; Bissonnette, M.; Li, Y.C. Novel Role of the Vitamin D Receptor in Maintaining the Integrity of the Intestinal Mucosal Barrier. American Journal of Physiology-Gastrointestinal and Liver Physiology 2008, 294, G208–G216, doi:10.1152/ajpgi.00398.2007. 56. Li, Y.C.; Chen, Y.; Du, J. Critical Roles of Intestinal Epithelial Vitamin D Receptor Signaling in Controlling Gut Mucosal Inflammation. J Steroid Biochem Mol Biol 2015, 148, 179–183, doi:10.1016/j.jsbmb.2015.01.011. 56 57. Kanhere, M.; Chassaing, B.; Gewirtz, A.T.; Tangpricha, V. Role of Vitamin D on Gut Microbiota in Cystic Fibrosis. J Steroid Biochem Mol Biol 2018, 175, 82–87, doi:10.1016/j.jsbmb.2016.11.001. 58. Liu, W.; Chen, Y.; Golan, M.A.; Annunziata, M.L.; Du, J.; Dougherty, U.; Kong, J.; Musch, M.; Huang, Y.; Pekow, J.; et al. Intestinal Epithelial Vitamin D Receptor Signaling Inhibits Experimental Colitis. J Clin Invest 2013, 123, 3983–3996, doi:10.1172/JCI65842. 59. Moossavi, S.; Sepehri, S.; Robertson, B.; Bode, L.; Goruk, S.; Field, C.J.; Lix, L.M.; Souza, R.J. de; Becker, A.B.; Mandhane, P.J.; et al. Composition and Variation of the Human Milk Microbiota Are Influenced by Maternal and Early-Life Factors. Cell Host & Microbe 2019, 25, 324-335.e4, doi:10.1016/j.chom.2019.01.011. 60. Prior, E.; Santhakumaran, S.; Gale, C.; Philipps, L.H.; Modi, N.; Hyde, M.J. Breastfeeding after Cesarean Delivery: A Systematic Review and Meta-Analysis of World Literature. Am J Clin Nutr 2012, 95, 1113–1135, doi:10.3945/ajcn.111.030254. 61. Hermansson, H.; Kumar, H.; Collado, M.C.; Salminen, S.; Isolauri, E.; Rautava, S. Breast Milk Microbiota Is Shaped by Mode of Delivery and Intrapartum Antibiotic Exposure. Front. Nutr. 2019, 0, doi:10.3389/fnut.2019.00004. 62. Rautava, S.; Luoto, R.; Salminen, S.; Isolauri, E. Microbial Contact during Pregnancy, Intestinal Colonization and Human Disease. Nat Rev Gastroenterol Hepatol 2012, 9, 565–576, doi:10.1038/nrgastro.2012.144. 63. Wachs, T.D.; Creed-Kanashiro, H.; Cueto, S.; Jacoby, E. Maternal Education and Intelligence Predict Offspring Diet and Nutritional Status. The Journal of Nutrition 2005, 135, 2179–2186, doi:10.1093/jn/135.9.2179. 64. Turroni, F.; Milani, C.; Duranti, S.; Mancabelli, L.; Mangifesta, M.; Viappiani, A.; Lugli, G.A.; Ferrario, C.; Gioiosa, L.; Ferrarini, A.; et al. Deciphering Bifidobacterial-Mediated Metabolic Interactions and Their Impact on Gut Microbiota by a Multi-Omics Approach. ISME J 2016, 10, 1656–1668, doi:10.1038/ismej.2015.236. 65. Okada, Y.; Tsuzuki, Y.; Hokari, R.; Komoto, S.; Kurihara, C.; Kawaguchi, A.; Nagao, S.; Miura, S. Anti-Inflammatory Effects of the Genus Bifidobacterium on Macrophages by Modification of Phospho-IκB and SOCS Gene Expression. Int J Exp Pathol 2009, 90, 131–140, doi:10.1111/j.1365-2613.2008.00632.x. 66. Baumann-Dudenhoeffer, A.M.; D’Souza, A.W.; Tarr, P.I.; Warner, B.B.; Dantas, G. Infant Diet and Maternal Gestational Weight Gain Predict Early Metabolic Maturation of Gut Microbiomes. Nat Med 2018, 24, 1822–1829, doi:10.1038/s41591-018-0216-2. 67. Chávez-Carbajal, A.; Nirmalkar, K.; Pérez-Lizaur, A.; Hernández-Quiroz, F.; Ramírez-del- Alto, S.; García-Mena, J.; Hernández-Guerrero, C. Gut Microbiota and Predicted Metabolic Pathways in a Sample of Mexican Women Affected by Obesity and Obesity Plus Metabolic Syndrome. International Journal of Molecular Sciences 2019, 20, 438, doi:10.3390/ijms20020438. 57 68. Kameyama, K.; Itoh, K. Intestinal Colonization by a Lachnospiraceae Bacterium Contributes to the Development of Diabetes in Obese Mice. Microbes Environ 2014, 29, 427–430, doi:10.1264/jsme2.ME14054. 69. Shen, F.; Zheng, R.-D.; Sun, X.-Q.; Ding, W.-J.; Wang, X.-Y.; Fan, J.-G. Gut Microbiota Dysbiosis in Patients with Non-Alcoholic Fatty Liver Disease. Hepatobiliary & Pancreatic Diseases International 2017, 16, 375–381, doi:10.1016/S1499-3872(17)60019-5. 70. Vacca, M.; Celano, G.; Calabrese, F.M.; Portincasa, P.; Gobbetti, M.; De Angelis, M. The Controversial Role of Human Gut Lachnospiraceae. Microorganisms 2020, 8, 573, doi:10.3390/microorganisms8040573. 71. Luthold, R.V.; Fernandes, G.R.; Franco-de-Moraes, A.C.; Folchetti, L.G.D.; Ferreira, S.R.G. Gut Microbiota Interactions with the Immunomodulatory Role of Vitamin D in Normal Individuals. Metabolism 2017, 69, 76–86, doi:10.1016/j.metabol.2017.01.007. 72. Waterhouse, M.; Hope, B.; Krause, L.; Morrison, M.; Protani, M.M.; Zakrzewski, M.; Neale, R.E. Vitamin D and the Gut Microbiome: A Systematic Review of in Vivo Studies. Eur J Nutr 2019, 58, 2895–2910, doi:10.1007/s00394-018-1842-7. 73. Yassour, M.; Vatanen, T.; Siljander, H.; Hämäläinen, A.-M.; Härkönen, T.; Ryhänen, S.J.; Franzosa, E.A.; Vlamakis, H.; Huttenhower, C.; Gevers, D.; et al. Natural History of the Infant Gut Microbiome and Impact of Antibiotic Treatments on Strain-Level Diversity and Stability. Sci Transl Med 2016, 8, 343ra81, doi:10.1126/scitranslmed.aad0917. 74. Liu, Y.; Qin, S.; Song, Y.; Feng, Y.; Lv, N.; Xue, Y.; Liu, F.; Wang, S.; Zhu, B.; Ma, J.; et al. The Perturbation of Infant Gut Microbiota Caused by Cesarean Delivery Is Partially Restored by Exclusive Breastfeeding. Front. Microbiol. 2019, 0, doi:10.3389/fmicb.2019.00598. 58 CHAPTER 4. ASSOCIATION OF THE INFANT GUT MICROBIOME WITH TEMPERAMENT 4.1 Abstract Recent studies in both animals and humans have shown that gut microbiota is linked to brain development and control of behavior, but the association between gut microbiota and behavior in healthy infants is largely unknown. We undertook this prospective study to determine associations between gut microbiota at 3 months during infancy and infant temperament in a populational-based birth cohort. We analyzed data from 157 infants from the Michigan Archive for Research on Child Health Cohort Study. Infant temperament outcomes were reported by mothers using the Rothbart Infant Behavior Questionnaire-Revised Very Short Form at a mean age of 9.1 months. Microbiota profiling with 16S rRNA gene sequencing was conducted on fecal samples obtained at approximately 3 months of age. We identified three clusters of infants based on the relative abundance of gut microbiota: cluster A was characterized by a higher abundance of Bacteroides; cluster C was characterized by a higher abundance of Bifidobacterium, Veillonella, and Escherichia-Shigella; cluster B is intermediate between the other two clusters. Fully adjusted multivariate linear regression analysis showed a negative association between cluster C and negative emotionality score (coefficient = -0.58, p-value= 0.02) compared to cluster A, prominently among infants who were not given vitamin D supplement. However, no associations were evident between gut microbiota clusters and temperament scales after FDR correction. MaAslin analysis identified that individual microbial taxa were associated with three scales of temperament. This study found an association between infant gut microbiota composition and temperament by connecting temperament to microbiota clusters and individual taxa. 59 4.2 Introduction The human gastrointestinal tract is the habitat of trillions of microorganisms that are closely associated with many aspects of human health, including physiological functions, metabolism, and immune functions[1–3]. A critical time for the microorganism, typically referred to as the microbiota, is during the first year of life since the gastrointestinal tract moves from a sterile to a bacteria rich environment[4]. The gut microbiota does not only aid human health physically but also plays a critical role in neurodevelopment through the interaction between microbiota and the brain, also known as the gut-brain axis[5,6]. The gut-brain axis hypothesizes that the brain and the microbiota communicate with each other through various chemical processes, which impact psychological and mental health[7]. The bidirectional relationship between the brain and the microbiota, or gut-brain axis, has been established using animal models. Previous studies have demonstrated that altered gut microbiota can lead to higher stress responsiveness[8], anxiety-like behaviors, abnormal social behaviors[9], and autism spectrum disorder behaviors[10]. Several studies have shown that germ-free (GF) mice who have no commensal microbiota and an undeveloped immune system[11,12] displayed increased motor activity and reduced anxiety compared with specific pathogen-free (SPF) mice with normal gut microbiota[13]. The abnormalities could be partially reversed when the gastrointestinal tracts of the germ-free mice were reconstituted with stool from normally raised mice[14]. Besides the behavioral differences, the brains of GF mice displayed various molecular differences, including brain-region-specific changes in levels of brain-derived neurotrophic factor, oxytocin, and vasopressin expression[15]. There is mounting evidence that gut microbiota plays a critical role in brain development and the control of behavior in humans[16,17]. Cross-sectional studies have discovered an association 60 between gut microbiome composition and neuropsychiatric outcomes, including autism spectrum disorder (ASD) and depression[18–20]. The disturbance of the gut microbiota in early life can lead to adverse mental health outcomes later in life[21]. Previous studies reported that children with ASD had lower abundances of Coprococcus, Prevotella, and unclassified Veillonellaceae compared to neurotypical children[22]. A recent study also reported a significant increase in the Firmicutes/Bacteroidetes ratio in children with ASD[23]. However, there are also some studies that observed no difference regarding severity ASD in the bacterial composition among children[24,25]. In terms of infant physical development, Sordillo et al. reported that infant gut microbiome composition at 3–6 months was associated with fine motor skills assessed by the Ages and Stages Questionnaire[26]. In a longitudinal study of 201 children, Loughman et al. found that the decreased abundance of the genus Prevotella in fecal samples collected at 12 months of age was associated with increased behavioral problems at age 2[27]. These studies suggest that the effects of the microbiome in early infancy play a more permanent role over the course of human life, whether it is in brain development or future behavioral outcomes. The microbiome plays an essential role in behavior which determines temperament, a constitutionally based individual difference in emotion, motor behavior, attention, and self- regulation[28,29]. Longitudinal studies have reported that temperament characteristics in early life are associated with psychiatric problems, including anxiety, depression, ADHD, and autism in mid‐childhood[30–32]. Despite the importance of temperament characteristics, few prospective studies have examined the associations between gut microbiome composition and temperament in infancy. In this study, we aimed to investigate the association between the fecal microbiota composition at 3 months of age and infant temperament at 9 months of age. 61 4.3 Materials and Methods 4.3.1 Study participants The study population was drawn from the Michigan Archive for Research on Child Health (MARCH) cohort[33], an ongoing pregnancy and birth cohort set in Michigan’s lower peninsula. The purpose of the MARCH study is to store biological specimens, and other health information that can be used to better understand the causes of problems in pregnancy and the health of children and to contribute to a nation-wide study of child health called the Environmental Influences on Child Health Outcomes (ECHO)[34]. Our analysis included mothers who provided informed consent for providing infant stool samples. During the MARCH 3 month phone interview, mothers confirmed their interest in participating in this sample collection. Fecal collection kits were sent by mail. 157 samples collected from singleton infants were included in the analysis. 4.3.2 Data collection At approximately nine months of age, the Rothbart Infant Behavior Questionnaire-Revised Very Short Form (IBQ-RVSF) was administered to mothers by phone interview to assess the infant temperament[35]. The Infant Behavior Questionnaire is a widely used parent-report measure of infant temperament, first introduced by Rothbart in 1981[36,37]. IBQ-RVSF consists of 37 items measuring three dimensions of infant temperament, including positive affect/surgency (PAS, 13 items), negative emotionality (NEG, 12 items), and orienting/regulatory capacity (ORC, 12 items)[35]. Each item asked caregivers to report how often their babies engaged in a particular behavior during the last seven days. The items were rated on a scale ranging from 1 (never) to 7 (always). Higher scores of each scale indicate more of the measured temperament characteristic. PAS is characterized by positive loadings on approach, vocal reactivity, high intensity pleasure, 62 smiling and laughter, activity level, and perceptual sensitivity. NEG is analogous to the personality trait of neuroticism, and is characterized by positive loadings on sadness, distress to limitations, and fear, as well as negative loadings on falling reactivity. ORC is characterized by duration of orienting, low intensity pleasure, cuddliness, and soothability. Previous studies have demonstrated IBQR–VSF has adequate internal consistency, test-retest reliability, and interrater agreement between mothers and fathers[35]. Besides the IBQ-RVSF, multiple questionnaires were administered to mothers from the first prenatal visit through 9 months postpartum to collect a variety of socioeconomic and feeding practices factors. Detailed information on the neonate, including sex, birth weight, and gestational age, was abstracted from the birth certificate. 4.3.3 Fecal microbiota analysis The fecal samples were aliquoted into sterile tubes and stored at -80°C, once received in the lab. DNA was extracted following a modified version of the Human Microbiome Project’s protocol as described previously[38]. Barcoded primers were used to amplify the V4 region of the 16S rRNA gene region. The resulting 16S rRNA libraries were sequenced using 250 base pair Illumina MiSeq with V2 chemistry at the MSU genomics core. After trimming, clean sequences were analyzed using the QIIME2 (2021.2 version) pipeline[39]. QIIME2's DADA2 plugin was used to process the demultiplexed sequences and generate the amplicon sequence variants (ASV) table[40]. ASVs were assigned to taxonomy by the QIIME2 feature-classifier plugin, using the Silva 132 database at the similarity threshold of 99% (for 16S data)[41,42]. Samples were rarefied to 6,000 sequencing reads per sample, and taxa present in less than one sample were excluded, leaving 157 stool samples with 6,905 unique ASVs. ASVs were summarized at the genus taxonomic level. 4.3.4 Statistical analysis 63 All the analyses were performed using R software (version 4.0). The Shannon index, which represents microbial richness and evenness, and the Chao 1 index, which represents microbial richness, were calculated using the "vegan" package[43]. Associations between Shannon index or Chao1 index to temperament scales were assessed by multivariate linear regression models, adjusted by delivery mode (vaginal vs C-section), race, maternal education, maternal pre- pregnancy BMI, breastfeeding status, infant sex, and infant age at IBQ-RVSF collection. Dirichlet multinomial mixture (DMM) clustering is an unsupervised Bayesian clustering method to identify clusters or enterotypes of microbial community data[44,45], as performed previously[46,47]. The best fitting DMM model was determined using the Laplace approximation. The difference of alpha diversity and relative abundance of taxa between DMM clusters were tested by Kruskal–Wallis test with Dunn test for post hoc. Multiple comparisons were adjusted for false discovery rate (FDR) correction using the Benjamini-Hochberg procedure. We performed Principal Coordinates Analysis (PCOA) and Permutational Multivariate Analysis of Variance (PERMANOVA) based on Bray–Curtis dissimilarity using the "vegan" package to compare the difference in microbial community structure between DMM clusters. We used multivariate linear regression models to determine the association between DMM clusters and infant temperament scales, adjusting by race, maternal education level, pre- pregnancy weight, delivery mode (vaginal vs C-section), infant sex, infant age at IBQ-RVSF collection, breastfeeding status, and infant vitamin D supplement. Previous studies discovered that vitamin D and infant sex play roles as effect modifiers in the association between gut microbiota and neurodevelopment[48,49]. So, we conducted the sensitivity analyses to confirm the potential for infant vitamin D and sex to modify gut microbiome associations with temperament scales by stratification. we utilized MaAsLin, a 64 multivariate statistical framework that identifies associations between clinical metadata and individual microbiome abundance[50,51]. MaAslin provides both nominal p-values and FDR adjusted p-values (q-values) by Benjamini–Hochberg procedure. Associations were considered significant if the q-value was below the threshold of 0.2 in MaAslin results. 4.4 Results 4.4.1 Study Population Characteristics and temperament scales 157 participants were included in the final analysis (male= 52.2%, female= 47.8%). Descriptive statistics for maternal factors, infant factors by the three scales of infant temperament were displayed in Table 4.1. More than half of the mothers (58.6%) earned a college graduate degree, and 66.2% had a vaginal delivery. 57.3% of the mothers fed their children with exclusive breast milk at the time of fecal sample collection, and 92.4% of them had ever fed their children with breast milk. The mean (SD) age at temperament measurement was 9.1 (0.7) months, and the median was 9.0 (range: 8.0-13.0) months. Race is significantly associated with all three scales. Infants of white race were reported to have significantly lowest scores of all three scales, while infants of black race were reported to have significantly highest scores of NEG and ORC. Male has significantly higher PAS score than female (p-value <0.001). Higher maternal age is significantly associated with lower PAS score (coefficient = -0.03, p-value= 0.006) and lower NEG score (coefficient = -0.04, p-value= 0.009). 65 Table 4.1. Characteristics of mothers and infants by infant temperament scales Positive Negative Orienting/regulatory Scale affect/surgency emotionality capacity Mean Mean Mean n (%) (SD) (SD) (SD) p- Variable or mean or β or β p-value or β p-value value (SD) (95% (95% (95% CI) CI) CI) Categorical variable1 Delivery mode, n (%) 104 5.53 4.05 5.34 Vaginal 0.71 0.31 0.24 (66.2%) (0.65) (0.96) (0.68) delivery C- 53 5.58 4.22 5.48 section (33.8%) (0.70) (0.99) (0.70) Maternal education level, n (%) Did 4 5.92 4.92 4.85 not finish 0.18 0.22 0.004* (2.6%) (0.98) (1.05) (1.10) high school High school 22 5.75 4.30 5.81 graduate or (14.0%) (0.69) (1.0) (0.53) GED Some 39 5.59 4.14 5.46 college (24.8%) (0.67) (0.95) (0.70) 66 Table 4.1. (cont’d) College 92 5.46 4.02 5.28 graduate or (58.6%) (0.64) (0.96) (0.67) more Race, n (%) 126 5.46 4.02 5.29 White <0.001* 0.03* 0.002* (80.3%) (0.63) (0.94) (0.69) 20 5.82 4.53 5.85 Black (12.7%) (0.73) (0.88) (0.58) 11 6.04 4.42 5.66 Other (7.0%) (0.70) (1.25) (0.56) Breastfeeding status at fecal sample collection, n (%) 90 5.51 4.17 5.29 Exclusive 0.37 0.68 0.1 (57.3%) (0.67) (0.95) (0.70) breastfeeding Partial 41 5.72 4.03 5.60 breastfeeding (26.1%) (0.63) (0.96) (0.75) Not 26 5.53 4.03 5.46 breastfeeding (16.6) (0.68) (1.04) (0.62) Ever breastfeeding, n (%) 145 5.53 4.09 5.38 Yes 0.32 0.38 0.73 (92.4%) (0.66) (0.97) (0.71) 12 5.73 4.34 5.46 No (7.6%) (0.70) (1.02) (0.51) 67 Table 4.1. (cont’d) Baby sex, n (%) 82 5.72 4.14 5.40 Male <0.001* 0.69 0.77 (52.2%) (0.64) (0.96) (0.73) 57 5.36 4.08 5.37 Female (47.8%) (0.65) (0.98) (0.65) Continuous variable2 -0.03 -0.04 Maternal age -0.006 31.5 (-0.05, (- (year), mean 0.006* 0.009* (-0.03, 0.61 (5.0) - 0.07, - (SD) 0.02) 0.008) 0.01) Pre- 0.001 0.01 (- 0.01 (- pregnancy 27.8 (- 0.005, 0.18 0.96 0.002, 0.09 BMI, mean (7.2) 0.02, 0.02) 0.03) (SD) 0.02) Gestational 0.02 age at 0.03 (- -0.05 38.8 (- delivery 0.04, 0.38 0.67 (-0.13, 0.14 (1.5) 0.08, (week), 0.10) 0.02) 0.13) mean (SD) Infant age at 0.004 0.04 0.002 IBQ (- (- (- measurement 9 (0.7) 0.14 0.32 0.55 0.001, 0.003, 0.004, (month), 0.008) 0.01) 0.007) mean (SD) 1 Summary of categorical variables were displayed as n (%). The difference in temperament scores between categorical variables was analyzed by One-Way Analysis of Variance (ANOVA). 2 Summary of continuous variables were displayed as mean (SD). The association between continuous variables and temperament scales was analyzed by linear regression. * p-value < 0.05 68 4.4.2 Alpha diversity and temperament scores Table 4.2 presents the association between alpha diversity of fecal samples and infant temperament scales. Higher Shannon index is associated with higher positive affect/surgency scores (coefficient=0.16, p-value= 0.12) and lower negative emotionality score (coefficient= - 0.16, p-value= 0.27), but the associations were not statistically significant. We also did not observe any significant associations between Chao 1 index and temperament scales (Table 4.2). Table 4.2. Association between alpha diversity and infant temperament1 Shannon Chao 1 Scale Beta p-value Beta p-value Positive affect/surgency 0.13 0.18 0.001 0.44 Negative emotionality -0.20 0.20 -0.004 0.08 Orienting/regulatory capacity 0.003 0.98 -0.001 0.66 1 Linear regression Models were adjusted for delivery mode, race, maternal education, maternal pre-pregnancy BMI, breastfeeding status, infant sex, and infant age at IBQ-RVSF collection. 4.4.3 Cluster analysis We employed Dirichlet multinomial mixture (DMM) modeling to assign microbiota composition into clusters. Using the minimum Laplace approximation, we identified three optimal clusters, which we also referred to as enterotypes (Figure 4.1A). Cluster A, B, and C accounted for 25%, 56%, and 19% of the total samples, respectively. A significantly lower Shannon index and Chao 1 index were observed in cluster B, compared to cluster A and C (Figure 4.1B-4.1C). In addition, the gut microbiota communities were well-separated by the three clusters in principal coordinate analysis (PCoA) based on the Bray–Curtis distance matrix (univariate PERMANOVA: R2= 11.9%, p-value =0.001, Figure 4.1D). Thus, these results revealed broad community differences 69 across the three clusters. Figure 4.1E displayed the heatmap, which shows the relative abundance of the top ten genera within each cluster. Samples in each cluster were ordered by the relative abundance of Bacteroides. Bacteroides, Bifidobacterium, Veillonella, and Escherichia-Shigella were the top 4 genera that drove the clustering (Figure 4.2). Samples in cluster A exhibited a higher relative abundance of Bacteroides than the cluster C (adjusted p-value = 1.7 × 10-10), but similar abundance of Bacteroides to cluster B (adjusted p-value= 0.33). Cluster C was characterized by the highest relative abundance of Bifidobacterium, Veillonella, and Escherichia- Shigella, compared to the cluster A and B. To investigate the potential association between clusters and infant temperament, we applied linear regression models, adjusted by covariates including delivery mode (vaginal vs C-section), race, maternal education, maternal pre-pregnancy BMI, breastfeeding status, infant vitamin D intake, infant sex, and infant age at IBQ-RVSF collection (Table 4.3). The results of univariate models demonstrated that cluster C is significantly associated with lower PAS score (coefficient = -0.43, p-value=0.009), and lower NEG score (coefficient = -0.54, p-value= 0.02), comparing to the cluster A. In multivariate linear regression models, cluster C is significantly associated with a lower NEG score (coefficient = -0.58, p-value=0.02), comparing to cluster A. However, after FDR correction, none of these clusters remained significantly associated with the temperament scales. In the sensitivity analysis, among the infants who had not taken any vitamin D supplements, infants of cluster C showed a significantly lower NEG scores than cluster A (coefficient = -1.01, p-value = 0.01) before FDR correction (Table S4.1). While in the vitamin D group, cluster C was no longer significant with NEG score, and displayed less pronounced coefficient (coefficient = - 0.43, p-value = 0.20), which suggested the associations were potentially attenuated by the 70 vitamin D intake. We did not find any significant association between gut microbiota clusters and temperament scales when stratifying by infant sex (Table S4.2). 4.4.4 Individual taxa analysis We next sought to assess the association between individual taxa at the genus level and infant temperament scales using MaAslin (Table 4.4). Infants with higher positive affect/surgency scores had significantly higher relative abundance of a set of Firmicutes, including genera Oscillospiraceae UCG-003, Oscillospiraceae UCG-002, Christensenellaceae R-7 group. Infants with higher negative emotionality score had significantly lower relative abundance of Lachnospiraceae FCS020 group (MaAslin coefficient = -0.07, q-value= 0.012), and higher relative abundance of Eggerthella (MaAslin coefficient = 0.10, q-value= 0.13). Infants with higher orienting/regulatory capacity scores had a significantly lower relative abundance of Clostridioides (MaAslin coefficient = -0.14, q-value= 0.19). 71 Figure 4.1. Dirichlet multinomial mixture clustering identified three optimal clusters from 157 fecal samples. (a) The number of clusters (k=3) was chosen by selecting the minimal Laplace approximation to the negative log model evidence. (b-c) Boxplot of the alpha diversity (Shannon and Chao 1) distributed between the 3 clusters. Group differences were tested by Wilcoxon signed-rank test, and p-values were adjusted for multiple testing using Bonferroni. Adjusted p-value <0.05 was labeled as *, and adjusted p-value ≥ 0.05 was labeled as NS. (d) Principal component analysis (PCoA) ordinations of variation based on the Bray–Curtis distance matrix. R2 and p-value were calculated by the univariate PERMANOVA test. (e) Heatmap of relative abundance of top 10 genera by the three clusters. Within the clusters, samples were ordered by the relative abundance of the genus Bacteroides. 72 Figure 4.2. Relative abundance of the top 4 genera that contribute by clusters. Group differences were tested by Wilcoxon signed-rank test. P-values were adjusted for multiple testing using Bonferroni. Adjusted p-value <0.05 was labeled as *, and adjusted p-value ≥ 0.05 was labeled as NS. 73 Table 4.3. Association between gut microbiota clusters and infant temperament scales1 Univariate model Multivariate model Scale Cluster p- q- p- q- Beta value value Beta value value2 Cluster A ref - - ref - - -0.20 (-0.46, -0.12 (-0.38, Positive Cluster B 0.05) 0.12 0.18 0.13) 0.34 0.51 affect/surgency -0.43 (-0.74, - -0.29 (-0.61, Cluster C 0.11) 0.009* 0.05 0.02) 0.07 0.21 Cluster A Ref - - ref - - -0.04 (-0.41, -0.04 (-0.43, Negative Cluster B 0.33) 0.84 0.84 0.34) 0.82 0.82 emotionality -0.54 (-1.00, - -0.58 (-1.05, - Cluster C 0.08) 0.02* 0.06 0.10) 0.02* 0.12 Cluster A ref - ref - - -0.09 (-0.36, -0.06 (-0.33, Orienting/regulatory Cluster B 0.18) 0.53 0.64 0.22) 0.69 0.82 capacity -0.30 (-0.63, -0.21 (-0.56, Cluster C 0.04) 0.08 0.16 0.13) 0.22 0.44 1 Models were adjusted for delivery mode, race, maternal education, maternal pre-pregnancy BMI, breastfeeding status, infant vitamin D intake, infant sex, and infant age at IBQ-RVSF collection. 2 P-values were adjusted by false discovery rate (FDR) correction for multiple comparisons using the Benjamini-Hochberg procedure. *p-value < 0.05 74 Table 4.4. MaAsLin Analysis Results: Associations of gut microbiome taxa at genus level and infant temperament scales adjusted by covariates1 q- Coeffici N/ N p- Scale Phylum Family Genus valu ent not 0 value e2 Lachnospiraceae_FCS0 157/ 0.000 0.01 Firmicutes Lachnospiraceae -0.07 Negative 20_group 16 03 3 emotionality Actinobacte 157/ 0.13 Eggerthellaceae Eggerthella 0.10 0.002 riota 44 1 Christensenellac Christensenellaceae R- 157/ 0.11 Firmicutes 0.16 0.001 eae 7 group 57 2 Positive 157/ 0.11 Firmicutes Oscillospiraceae UCG-003 0.12 0.001 affect/surgency 44 3 157/ Firmicutes Oscillospiraceae UCG-002 0.14 0.003 0.15 73 Orienting/regul Peptostreptococ 157/ Firmicutes Clostridioides -0.14 0.009 0.20 atory capacity caceae 79 1 Model was adjusted for delivery mode, race, maternal education, maternal pre-pregnancy BMI, breastfeeding status, infant vitamin D intake, infant sex, and infant age at IBQ-RVSF collection. 2 Q-value is the FDR (Benjamini-Hochberg) adjusted p-value. Q-value < 0.2 for multiple comparisons was considered statistically significant and included in the table. 4.5 Discussion Accumulating evidence from animal and human studies report that the gut microbiota plays a role in neurodevelopment during the early critical time window. In the current study, we focus on the three subscales of the infant temperament: positive affect/surgency, negative emotionality, 75 and orienting/regulatory capacity. Our results show an association between infant gut microbiota composition and temperament by connecting temperament to both DMM clusters and individual taxa. 4.5.1 Alpha diversity and temperament scales No significant associations were found between alpha diversity and temperament scales. This finding is consistent with some results of studies from longitudinal and cross-section studies. Fox et al. investigated the relationship between gut microbiota at each age group (1–3 weeks, 2, 6, and 12 months) and IBQ scores at 12-months of age[52]. No temperament scales at age 12 months demonstrated a significant association with the alpha diversity measures at each age. Similarly, Kelsey et al. demonstrated that neither alpha diversity nor richness was associated with any of the temperament scales in a cross-sectional study[53]. 4.5.2 Cluster analysis We identified three clusters based on the DMM method. Cluster A was characterized by a higher abundance of Bacteroides; cluster C is characterized by a higher abundance of Bifidobacterium, Veillonella, and Escherichia-Shigella; while cluster B is intermediate between the other two clusters. Our result shows that cluster C had a significant inverse association with negative emotionality. Negative emotionality, which is defined by the disposition to experience negative emotions such as anger and fear[54], has often been linked with internalizing and externalizing problems[55,56]. Previous work by Aatsinki et al. also showed the same trend toward the relationship between Bifidobacterium-dominated/ Bacteroides-dominated cluster and negative emotionality, though it was not statistically significant[57]. In addition, two studies of early infants discovered the association between Bacteroides-dominant community and poor fine 76 motor skills, which also supports the adverse effect of Bacteroides-dominated microbiome composition on infant neurodevelopment[47,58]. In the sensitivity analysis, we discovered the potential role of infant vitamin D intake in modifying the association between clusters and negative emotionality. The association was only identified among the infants who were not recipients of vitamin D supplements, suggesting infant vitamin D intake potentially protects against the adverse effect of the gut microbiome– associated increments on infant negative emotionality. Similarly, a prior study reported that in participants who received prenatal vitamin D treatment, the Veillonella-dominated gut microbiota community was associated with improved communication scores in infants, whereas no association between the Veillonella-dominated community and communication scores was observed in the control group[58]. A possible explanation is that, vitamin D and vitamin D receptor (VDR) levels modify gut microbiota in neurodevelopment through the gut–microbiota– brain axis, such as cytokines, neurotransmitters, and SCFAs[59]. 4.5.3 Individual taxa Our results showed that several taxa were significantly associated with infant temperament outcomes, suggesting a mechanism of influence on infant temperament involving specific taxa of the gut microbiome. Of note, our results agree with the previous literature, which suggests that the genus Clostridioides (Clostridium) in the infant's gut is associated with adverse neurodevelopment outcomes[57,60]. Clostridioides is demonstrated to be an important predictor of infant temperament, as a similar study also reports a significant negative association between Clostridioides at age 2.5 months and regulation scale at age 6 months[57]. Clostridioides difficile (C difficile), which can cause severe diarrhea and colitis, is the most common Clostridioides species in the infant. Animal studies showed that C. difficile can produce Propionic Acid, a short- 77 chain fatty acid that can introduce ASD-related symptoms[61,62]. Future research should further examine the specific strain of Clostridioides in the infant gut to confirm the association between C difficile and infant temperament. Our results did not identify any significant associations between Bifidobacterium and positive affect/surgency, which is in contrast to other studies. Both Fox et al. and Aatsinki et al. reported that Bifidobacterium abundance at early infancy was positively associated with surgency/extraversion[52,57]. Bifidobacterium is typically considered as a beneficial bacteria for infant neurodevelopment, demonstrating benefits for gut epithelium integrity and function as well as gastrointestinal motility[63]. Larger sample size may need for our study to replicate the results from previous studies. 4.5.4 Strength and limitation Our current study may have a number of strengths. We demonstrated prospective associations between infant gut microbiome and temperament in a population-based cohort study. In addition to adjustment for prenatal and infant variables, our models also accounted for the mode of delivery, gestational age, and nutrition. The associations between microbiome clusters and negative emotionality are robust to adjustment for these variables. Our analysis not only demonstrated the effect of gut microbiota composition on infant temperament but also highlighted the importance of individual taxa. Of note, our study is not free of limitations. First, infant gut microbiota is rapidly maturated over the first year of life. However, only one fecal sample was included in our analysis and cannot fully represent the temporal development of the infant gut microbiome, which may be critical to neurodevelopment in early life. Second, 16S rRNA sequencing can only provide very limited strain-level information. Previous studies demonstrate the biological importance and different metabolic capabilities of specific bacterial 78 strains on human health. Thus, Shotgun metagenomic sequencing is needed for our future studies. 79 APPENDIX 80 Table S4.1. Association between gut microbiota clusters and infant temperament scales stratifying by infant vitamin D intake1 No Vitamin D Vitamin D Scale Cluster p- q- p- q- 2 Beta value value Beta value value Cluster ref - ref - A Cluster -0.34 (-0.86, -0.05 (-0.35, Positive affect/surgency 0.20 0.46 0.76 0.91 B 0.18) 0.25) Cluster -0.33 (-0.87, -0.35 (-0.77, 0.23 0.46 0.11 0.46 C 0.20) 0.08) Cluster ref - ref - A Cluster -0.50 (-1.21, 0.03 (-0.43, Negative emotionality 0.18 0.46 0.89 0.91 B 0.22) 0.49) Cluster -1.01 (-1.75, - -0.43 (-1.08, 0.01* 0.12 0.20 0.46 C 0.26) 0.23) Cluster ref - ref - A Cluster -0.03 (-0.54, -0.15 (-0.44, Orienting/regulatory capacity 0.91 0.91 0.38 0.57 B 0.48) 0.39) Cluster -0.13 (-0.66, -0.23 (-0.90, 0.63 0.84 0.34 0.57 C 0.40) 0.25) 1 Models were adjusted for delivery mode, race, maternal education, maternal pre-pregnancy BMI, breastfeeding status, infant age at IBQ-RVSF collection and infant sex. 2 P-values were adjusted by false discovery rate (FDR) correction for multiple comparisons using the Benjamini-Hochberg procedure. *p-value < 0.05 81 Table S4.2. Association between gut microbiota clusters and infant temperament scales stratifying by infant sex1 Male Female Scale Cluster q- p- q- 2 Beta p-value value Beta value value Cluster ref - ref - A Cluster -0.08 (-0.42, -0.24 (-0.67, Positive affect/surgency 0.63 0.83 0.54 B 0.26) 0.19) 0.27 Cluster -0.30 (-0.74, -0.41 (-0.91, 0.18 0.48 0.48 C 0.14) 0.10) 0.12 Cluster ref - ref - A Cluster -0.13 (-0.74, Negative emotionality 0.04 () 0.86 0.86 0.69 0.83 B 0.49) Cluster -0.51 (-1.23, -0.59 () 0.08 0.48 0.18 0.48 C 0.22) Cluster ref - ref - A Orienting/regulatory Cluster 0.06 (-0.32, -0.20 (-0.64, 0.76 0.83 0.38 0.60 capacity B 0.44) 0.24) Cluster -0.33 (-0.82, -0.22 (-0.74, 0.20 0.48 0.40 0.60 C 0.16) 0.30) 1 Models were adjusted for delivery mode, race, maternal education, maternal pre-pregnancy BMI, breastfeeding status, infant age at IBQ-RVSF collection, infant vitamin D. 2 P-values were adjusted by false discovery rate (FDR) correction for multiple comparisons using the Benjamini-Hochberg procedure. *p-value < 0.05 82 REFERENCES 83 REFERENCES 1. Hakansson, A.; Molin, G. Gut Microbiota and Inflammation. Nutrients 2011, 3, 637–682, doi:10.3390/nu3060637. 2. Levy, M.; Kolodziejczyk, A.A.; Thaiss, C.A.; Elinav, E. Dysbiosis and the Immune System. Nat Rev Immunol 2017, 17, 219–232, doi:10.1038/nri.2017.7. 3. Sekirov, I.; Russell, S.L.; Antunes, L.C.M.; Finlay, B.B. Gut Microbiota in Health and Disease. Physiol Rev 2010, 90, 859–904, doi:10.1152/physrev.00045.2009. 4. Quigley, E.M.M. Gut Bacteria in Health and Disease. Gastroenterol Hepatol (N Y) 2013, 9, 560–569. 5. Foster, J.A.; McVey Neufeld, K.-A. Gut-Brain Axis: How the Microbiome Influences Anxiety and Depression. Trends Neurosci 2013, 36, 305–312, doi:10.1016/j.tins.2013.01.005. 6. Burokas, A.; Moloney, R.D.; Dinan, T.G.; Cryan, J.F. Microbiota Regulation of the Mammalian Gut-Brain Axis. Adv Appl Microbiol 2015, 91, 1–62, doi:10.1016/bs.aambs.2015.02.001. 7. Morais, L.H.; Schreiber, H.L.; Mazmanian, S.K. The Gut Microbiota–Brain Axis in Behaviour and Brain Disorders. Nat Rev Microbiol 2021, 19, 241–255, doi:10.1038/s41579-020-00460-0. 8. Sudo, N.; Chida, Y.; Aiba, Y.; Sonoda, J.; Oyama, N.; Yu, X.-N.; Kubo, C.; Koga, Y. Postnatal Microbial Colonization Programs the Hypothalamic–Pituitary–Adrenal System for Stress Response in Mice. The Journal of Physiology 2004, 558, 263–275, doi:10.1113/jphysiol.2004.063388. 9. Neufeld, K.-A.M.; Kang, N.; Bienenstock, J.; Foster, J.A. Effects of Intestinal Microbiota on Anxiety-like Behavior. Communicative & Integrative Biology 2011, 4, 492–494, doi:10.4161/cib.15702. 10. Sharon, G.; Cruz, N.J.; Kang, D.-W.; Gandal, M.J.; Wang, B.; Kim, Y.-M.; Zink, E.M.; Casey, C.P.; Taylor, B.C.; Lane, C.J.; et al. Human Gut Microbiota from Autism Spectrum Disorder Promote Behavioral Symptoms in Mice. Cell 2019, 177, 1600-1618.e17, doi:10.1016/j.cell.2019.05.004. 11. Kennedy, E.A.; King, K.Y.; Baldridge, M.T. Mouse Microbiota Models: Comparing Germ- Free Mice and Antibiotics Treatment as Tools for Modifying Gut Bacteria. Frontiers in Physiology 2018, 9, 1534, doi:10.3389/fphys.2018.01534. 12. Fiebiger, U.; Bereswill, S.; Heimesaat, M.M. Dissecting the Interplay Between Intestinal Microbiota and Host Immunity in Health and Disease: Lessons Learned from Germfree and Gnotobiotic Animal Models. Eur J Microbiol Immunol (Bp) 2016, 6, 253–271, doi:10.1556/1886.2016.00036. 84 13. Heijtz, R.D.; Wang, S.; Anuar, F.; Qian, Y.; Björkholm, B.; Samuelsson, A.; Hibberd, M.L.; Forssberg, H.; Pettersson, S. Normal Gut Microbiota Modulates Brain Development and Behavior. PNAS 2011, 108, 3047–3052, doi:10.1073/pnas.1010529108. 14. Bercik, P.; Denou, E.; Collins, J.; Jackson, W.; Lu, J.; Jury, J.; Deng, Y.; Blennerhassett, P.; Macri, J.; McCoy, K.D.; et al. The Intestinal Microbiota Affect Central Levels of Brain-Derived Neurotropic Factor and Behavior in Mice. Gastroenterology 2011, 141, 599–609, 609.e1-3, doi:10.1053/j.gastro.2011.04.052. 15. Desbonnet, L.; Clarke, G.; Traplin, A.; O’Sullivan, O.; Crispie, F.; Moloney, R.D.; Cotter, P.D.; Dinan, T.G.; Cryan, J.F. Gut Microbiota Depletion from Early Adolescence in Mice: Implications for Brain and Behaviour. Brain Behav Immun 2015, 48, 165–173, doi:10.1016/j.bbi.2015.04.004. 16. Rogers, G.B.; Keating, D.J.; Young, R.L.; Wong, M.-L.; Licinio, J.; Wesselingh, S. From Gut Dysbiosis to Altered Brain Function and Mental Illness: Mechanisms and Pathways. Mol Psychiatry 2016, 21, 738–748, doi:10.1038/mp.2016.50. 17. Sampson, T.R.; Mazmanian, S.K. Control of Brain Development, Function, and Behavior by the Microbiome. Cell Host Microbe 2015, 17, 565–576, doi:10.1016/j.chom.2015.04.011. 18. Cenit, M.C.; Sanz, Y.; Codoñer-Franch, P. Influence of Gut Microbiota on Neuropsychiatric Disorders. World J Gastroenterol 2017, 23, 5486–5498, doi:10.3748/wjg.v23.i30.5486. 19. Kim, H.-N.; Yun, Y.; Ryu, S.; Chang, Y.; Kwon, M.-J.; Cho, J.; Shin, H.; Kim, H.-L. Correlation between Gut Microbiota and Personality in Adults: A Cross-Sectional Study. Brain, Behavior, and Immunity 2018, 69, 374–385, doi:10.1016/j.bbi.2017.12.012. 20. Białecka-Dębek, A.; Granda, D.; Szmidt, M.K.; Zielińska, D. Gut Microbiota, Probiotic Interventions, and Cognitive Function in the Elderly: A Review of Current Knowledge. Nutrients 2021, 13, 2514, doi:10.3390/nu13082514. 21. Borre, Y.E.; O’Keeffe, G.W.; Clarke, G.; Stanton, C.; Dinan, T.G.; Cryan, J.F. Microbiota and Neurodevelopmental Windows: Implications for Brain Disorders. Trends Mol Med 2014, 20, 509–518, doi:10.1016/j.molmed.2014.05.002. 22. Kang, D.-W.; Park, J.G.; Ilhan, Z.E.; Wallstrom, G.; LaBaer, J.; Adams, J.B.; Krajmalnik- Brown, R. Reduced Incidence of Prevotella and Other Fermenters in Intestinal Microflora of Autistic Children. PLOS ONE 2013, 8, e68322, doi:10.1371/journal.pone.0068322. 23. Strati, F.; Cavalieri, D.; Albanese, D.; De Felice, C.; Donati, C.; Hayek, J.; Jousson, O.; Leoncini, S.; Renzi, D.; Calabrò, A.; et al. New Evidences on the Altered Gut Microbiota in Autism Spectrum Disorders. Microbiome 2017, 5, 24, doi:10.1186/s40168-017-0242-1. 24. Gondalia, S.V.; Palombo, E.A.; Knowles, S.R.; Cox, S.B.; Meyer, D.; Austin, D.W. Molecular Characterisation of Gastrointestinal Microbiota of Children With Autism (With and Without Gastrointestinal Dysfunction) and Their Neurotypical Siblings. Autism Research 2012, 5, 419–427, doi:10.1002/aur.1253. 85 25. Son, J.S.; Zheng, L.J.; Rowehl, L.M.; Tian, X.; Zhang, Y.; Zhu, W.; Litcher-Kelly, L.; Gadow, K.D.; Gathungu, G.; Robertson, C.E.; et al. Comparison of Fecal Microbiota in Children with Autism Spectrum Disorders and Neurotypical Siblings in the Simons Simplex Collection. PLOS ONE 2015, 10, e0137725, doi:10.1371/journal.pone.0137725. 26. Sordillo, J.E.; Korrick, S.; Laranjo, N.; Carey, V.; Weinstock, G.M.; Gold, D.R.; O’Connor, G.; Sandel, M.; Bacharier, L.B.; Beigelman, A.; et al. Association of the Infant Gut Microbiome With Early Childhood Neurodevelopmental Outcomes: An Ancillary Study to the VDAART Randomized Clinical Trial. JAMA Netw Open 2019, 2, e190905, doi:10.1001/jamanetworkopen.2019.0905. 27. Loughman, A.; Ponsonby, A.-L.; O’Hely, M.; Symeonides, C.; Collier, F.; Tang, M.L.K.; Carlin, J.; Ranganathan, S.; Allen, K.; Pezic, A.; et al. Gut Microbiota Composition during Infancy and Subsequent Behavioural Outcomes. EBioMedicine 2020, 52, 102640, doi:10.1016/j.ebiom.2020.102640. 28. Rothbart, M.K.; Sheese, B.E.; Rueda, M.R.; Posner, M.I. Developing Mechanisms of Self- Regulation in Early Life. Emot Rev 2011, 3, 207–213, doi:10.1177/1754073910387943. 29. Rothbart, M.K.; Ahadi, S.A.; Evans, D.E. Temperament and Personality: Origins and Outcomes. J Pers Soc Psychol 2000, 78, 122–135, doi:10.1037//0022-3514.78.1.122. 30. Rettew, D.C.; McKee, L. Temperament and Its Role in Developmental Psychopathology. Harv Rev Psychiatry 2005, 13, 14–27, doi:10.1080/10673220590923146. 31. Tang, A.; Crawford, H.; Morales, S.; Degnan, K.A.; Pine, D.S.; Fox, N.A. Infant Behavioral Inhibition Predicts Personality and Social Outcomes Three Decades Later. PNAS 2020, 117, 9800–9807, doi:10.1073/pnas.1917376117. 32. Lavigne, J.V.; Gibbons, R.D.; Christoffel, K.K.; Arend, R.; Rosenbaum, D.; Binns, H.; Dawson, N.; Sobel, H.; Isaacs, C. Prevalence Rates and Correlates of Psychiatric Disorders among Preschool Children. Journal of the American Academy of Child & Adolescent Psychiatry 1996, 35, 204–214, doi:10.1097/00004583-199602000-00014. 33. About | CHARM Study Available online: https://www.epi.msu.edu/charmstudy/about (accessed on 30 September 2021). 34. About ECHO Available online: https://www.nih.gov/echo/about-echo (accessed on 30 September 2021). 35. Putnam, S.P.; Helbig, A.L.; Gartstein, M.A.; Rothbart, M.K.; Leerkes, E. Development and Assessment of Short and Very Short Forms of the Infant Behavior Questionnaire-Revised. J Pers Assess 2014, 96, 445–458, doi:10.1080/00223891.2013.841171. 36. Rothbart, M.K. Measurement of Temperament in Infancy. Child Development 1981, 52, 569–578, doi:10.2307/1129176. 86 37. Rothbart, M.K.; Ahadi, S.A.; Hershey, K.L.; Fisher, P. Investigations of Temperament at Three to Seven Years: The Children’s Behavior Questionnaire. Child Development 2001, 72, 1394–1408, doi:10.1111/1467-8624.00355. 38. Sugino, K.Y.; Paneth, N.; Comstock, S.S. Michigan Cohorts to Determine Associations of Maternal Pre-Pregnancy Body Mass Index with Pregnancy and Infant Gastrointestinal Microbial Communities: Late Pregnancy and Early Infancy. PLoS One 2019, 14, e0213733, doi:10.1371/journal.pone.0213733. 39. Bolyen, E.; Rideout, J.R.; Dillon, M.R.; Bokulich, N.A.; Abnet, C.C.; Al-Ghalith, G.A.; Alexander, H.; Alm, E.J.; Arumugam, M.; Asnicar, F.; et al. Reproducible, Interactive, Scalable and Extensible Microbiome Data Science Using QIIME 2. Nat Biotechnol 2019, 37, 852–857, doi:10.1038/s41587-019-0209-9. 40. Callahan, B.J.; McMurdie, P.J.; Rosen, M.J.; Han, A.W.; Johnson, A.J.A.; Holmes, S.P. DADA2: High-Resolution Sample Inference from Illumina Amplicon Data. Nat Methods 2016, 13, 581–583, doi:10.1038/nmeth.3869. 41. Bokulich, N.A.; Kaehler, B.D.; Rideout, J.R.; Dillon, M.; Bolyen, E.; Knight, R.; Huttley, G.A.; Gregory Caporaso, J. Optimizing Taxonomic Classification of Marker-Gene Amplicon Sequences with QIIME 2’s Q2-Feature-Classifier Plugin. Microbiome 2018, 6, 90, doi:10.1186/s40168-018-0470-z. 42. Quast, C.; Pruesse, E.; Yilmaz, P.; Gerken, J.; Schweer, T.; Yarza, P.; Peplies, J.; Glöckner, F.O. The SILVA Ribosomal RNA Gene Database Project: Improved Data Processing and Web- Based Tools. Nucleic Acids Res 2013, 41, D590–D596, doi:10.1093/nar/gks1219. 43. Dixon, P. VEGAN, a Package of R Functions for Community Ecology. Journal of Vegetation Science 2003, 14, 927–930, doi:10.1111/j.1654-1103.2003.tb02228.x. 44. Nigam, K.; Mccallum, A.K.; Thrun, S.; Mitchell, T. Text Classification from Labeled and Unlabeled Documents Using EM. Machine Learning 2000, 39, 103–134, doi:10.1023/A:1007692713085. 45. Holmes, I.; Harris, K.; Quince, C. Dirichlet Multinomial Mixtures: Generative Models for Microbial Metagenomics. PLOS ONE 2012, 7, e30126, doi:10.1371/journal.pone.0030126. 46. Stewart, C.J.; Ajami, N.J.; O’Brien, J.L.; Hutchinson, D.S.; Smith, D.P.; Wong, M.C.; Ross, M.C.; Lloyd, R.E.; Doddapaneni, H.; Metcalf, G.A.; et al. Temporal Development of the Gut Microbiome in Early Childhood from the TEDDY Study. Nature 2018, 562, 583–588, doi:10.1038/s41586-018-0617-x. 47. Acuña, I.; Cerdó, T.; Ruiz, A.; Torres-Espínola, F.J.; López-Moreno, A.; Aguilera, M.; Suárez, A.; Campoy, C. Infant Gut Microbiota Associated with Fine Motor Skills. Nutrients 2021, 13, 1673, doi:10.3390/nu13051673. 48. Tamana, S.K.; Tun, H.M.; Konya, T.; Chari, R.S.; Field, C.J.; Guttman, D.S.; Becker, A.B.; Moraes, T.J.; Turvey, S.E.; Subbarao, P.; et al. Bacteroides-Dominant Gut Microbiome of Late 87 Infancy Is Associated with Enhanced Neurodevelopment. Gut Microbes 2021, 13, 1930875, doi:10.1080/19490976.2021.1930875. 49. Sordillo, J.E.; Korrick, S.; Laranjo, N.; Carey, V.; Weinstock, G.M.; Gold, D.R.; O’Connor, G.; Sandel, M.; Bacharier, L.B.; Beigelman, A.; et al. Association of the Infant Gut Microbiome With Early Childhood Neurodevelopmental Outcomes: An Ancillary Study to the VDAART Randomized Clinical Trial. JAMA Network Open 2019, 2, e190905–e190905, doi:10.1001/jamanetworkopen.2019.0905. 50. Morgan, X.C.; Tickle, T.L.; Sokol, H.; Gevers, D.; Devaney, K.L.; Ward, D.V.; Reyes, J.A.; Shah, S.A.; LeLeiko, N.; Snapper, S.B.; et al. Dysfunction of the Intestinal Microbiome in Inflammatory Bowel Disease and Treatment. Genome Biology 2012, 13, R79, doi:10.1186/gb- 2012-13-9-r79. 51. Mallick, H.; Rahnavard, A.; McIver, L.J.; Ma, S.; Zhang, Y.; Nguyen, L.H.; Tickle, T.L.; Weingart, G.; Ren, B.; Schwager, E.H.; et al. Multivariable Association Discovery in Population- Scale Meta-Omics Studies; 2021; p. 2021.01.20.427420; 52. Fox, M.; Lee, S.M.; Wiley, K.S.; Lagishetty, V.; Sandman, C.A.; Jacobs, J.P.; Glynn, L.M. Development of the Infant Gut Microbiome Predicts Temperament across the First Year of Life. Development and Psychopathology undefined/ed, 1–12, doi:10.1017/S0954579421000456. 53. Kelsey, C.M.; Prescott, S.; McCulloch, J.A.; Trinchieri, G.; Valladares, T.L.; Dreisbach, C.; Alhusen, J.; Grossmann, T. Gut Microbiota Composition Is Associated with Newborn Functional Brain Connectivity and Behavioral Temperament. Brain, Behavior, and Immunity 2021, 91, 472–486, doi:10.1016/j.bbi.2020.11.003. 54. Rothbart, M.K. Becoming Who We Are: Temperament and Personality in Development; Guilford Press, 2011; ISBN 978-1-60918-071-3. 55. Lemery, K.S.; Essex, M.J.; Smider, N.A. Revealing the Relation between Temperament and Behavior Problem Symptoms by Eliminating Measurement Confounding: Expert Ratings and Factor Analyses. Child Development 2002, 73, 867–882, doi:10.1111/1467-8624.00444. 56. Lengua, L.J.; Wolchik, S.A.; Sandler, I.N.; West, S.G. The Additive and Interactive Effects of Parenting and Temperament in Predicting Adjustment Problems of Children of Divorce. Journal of Clinical Child Psychology 2000, 29, 232–244, doi:10.1207/S15374424jccp2902_9. 57. Aatsinki, A.-K.; Lahti, L.; Uusitupa, H.-M.; Munukka, E.; Keskitalo, A.; Nolvi, S.; O’Mahony, S.; Pietilä, S.; Elo, L.L.; Eerola, E.; et al. Gut Microbiota Composition Is Associated with Temperament Traits in Infants. Brain Behav Immun 2019, 80, 849–858, doi:10.1016/j.bbi.2019.05.035. 58. Sordillo, J.E.; Korrick, S.; Laranjo, N.; Carey, V.; Weinstock, G.M.; Gold, D.R.; O’Connor, G.; Sandel, M.; Bacharier, L.B.; Beigelman, A.; et al. Association of the Infant Gut Microbiome With Early Childhood Neurodevelopmental Outcomes: An Ancillary Study to the VDAART Randomized Clinical Trial. JAMA Network Open 2019, 2, e190905, doi:10.1001/jamanetworkopen.2019.0905. 88 59. Ogbu, D.; Xia, E.; Sun, J. Gut Instincts: Vitamin D/Vitamin D Receptor and Microbiome in Neurodevelopment Disorders. Open Biology 10, 200063, doi:10.1098/rsob.200063. 60. Loughman, A.; Quinn, T.; Nation, M.L.; Reichelt, A.; Moore, R.J.; Van, T.T.H.; Sung, V.; Tang, M.L.K. Infant Microbiota in Colic: Predictive Associations with Problem Crying and Subsequent Child Behavior. Journal of Developmental Origins of Health and Disease 2021, 12, 260–270, doi:10.1017/S2040174420000227. 61. Diseases, C. on I. Clostridium Difficile Infection in Infants and Children. Pediatrics 2013, 131, 196–200, doi:10.1542/peds.2012-2992. 62. Al-Owain, M.; Kaya, N.; Al-Shamrani, H.; Al-Bakheet, A.; Qari, A.; Al-Muaigl, S.; Ghaziuddin, M. Autism Spectrum Disorder in a Child with Propionic Acidemia. JIMD Rep 2012, 7, 63–66, doi:10.1007/8904_2012_143. 63. Cong, X.; Henderson, W.A.; Graf, J.; McGrath, J.M. Early Life Experience and Gut Microbiome: The Brain-Gut-Microbiota Signaling System. Adv Neonatal Care 2015, 15, 314– 323, doi:10.1097/ANC.0000000000000191. 89 CHAPTER 5. ASSOCIATION BETWEEN INFANT GUT MICROBIOME AND SLEEP PROBLEMS DURING CHILDHOOD 5.1 Abstract The difficulty of initiating and maintaining sleep are the two most predominant complaints of sleep disorders in childhood. To the best of our knowledge, no study has yet investigated the association between gut microbiota collected at early infancy and these two disorders in early childhood. We undertook this prospective study to determine the associations between gut microbiota collected at early infancy (3-9 months) and difficulty initiating and maintaining sleep at the age of 2 years in a populational-based birth cohort. We analyzed data from 195 infants from the Michigan Archive for Research on Child Health Cohort Study. Sleep disorders were abstracted from the items in the PROMIS Sleep Disturbance scale, which was reported by mothers. Feeding practices and clinical information were collected by questionnaire and abstraction of birth certificates. Microbiota profiling with 16S rRNA gene sequencing at the V4 region was conducted on fecal samples. Gut microbiota of children who had difficulty of maintaining sleep displayed significantly higher Shannon index (OR: 2.41, 95% CI= 1.23- 4.93, p-adjust < 0.04) and Chao 1 index (OR: 1.01, 95% CI= 1.0-1.03, p-adjust < 0.008) after adjustment for covariates. We also demonstrated that gut microbiota composition was significantly associated with difficulty initiating (p-value= 0.043) and maintaining sleep (p-value= 0.004) based on the unweighted UniFrac distance metric in 2-year-old children. In the DESeq2 analysis for individual taxa, we identified several taxa associated with each of two sleep disorders at the genus level. This study demonstrated a clear association between infant gut microbiota and sleep disorders in early childhood. 5.2 Introduction 5.2.1 Sleep in childhood Sleep quality and quantity are increasingly being considered critical factors for child health and development. Epidemiology studies indicate that up to 50% of children experience a sleep disorder between 0 to 6 years old[1–3]. The frequency of childhood sleep disorders from 2 90 general pediatric clinics indicated that 41% of parents reported insomnia in their children[4]. Difficulty initiating and maintaining sleep are the two most frequent sleep problems in childhood and often co-exist[5]. Multiple biological, psychosocial, and environmental factors are linked to sleep disorders in children[6]. Obesity is one of the critical factors associated with sleep disorders, and the relationship is bidirectional[7]. Cross-sectional studies from multiple countries suggested that increased levels of obesity are associated with decreased sleep duration[8–11]. Previous studies also have reported a higher prevalence and severity of obstructive sleep apnea (OSA) in children with obesity[12]. Environmental exposure also plays a vital role in sleep disorders among children. For example, elevated blood lead levels in early childhood are associated with increased risk for sleep disorders and excessive daytime sleepiness in later childhood[13]. Sleep disorders are associated with an increased risk of poor school performance, anxiety, depression, aggressive behaviors, and attention problems that continue in adulthood [14–17]. Sleep disorders interrupt the duration and depth of sleep, which are essential for child development. It is critical to identify the underlying factors involved in sleep disorders. 5.2.2 The gut microbiome and sleep Disruption of the gut microbiota had been linked to psychiatric disorders and behavior problems in children. Growing evidence points toward the gut-microbiota-brain axis, which refers to the network involving multiple biological systems that allow bidirectional communication between gut bacteria and the brain[18]. The communication pathways in these biological networks include both direct and indirect signaling via chemical transmitters, neuronal pathways, and the immune system[18]. Sleep disorders, which are closely related to psychiatric disorders, have been linked to gut microbiota in recent years. Animal studies showed that during sleep 91 fragmentation, the relative abundance of Actinobacteria, Lactobacillaceae, and Bifidobacteriaceae decreased, while the relative abundance of Ruminococcaceae increased in mice[19,20]. A cohort study showed that children with Obstructive Sleep Apnoea syndrome were associated with a lower microbiota diversity in respect to healthy subjects[21]. A study among old adults also suggested a possible relationship between the composition of the gut microbiome and sleep quality[22]. Although difficulty initiating and maintaining sleep has a high prevalence among children, the mechanisms underlying the development of these two sleep problems has not yet been understood[23]. To our knowledge, no study has investigated their association with gut microbiota in children. Therefore, the present study aimed to investigate the association between the fecal microbiota collected in early infancy and difficulty initiating and maintaining sleep at the age of two years. 5.3 Materials and Methods 5.3.1 Study participants The study population was drawn from the Michigan Archive for Research on Child Health (MARCH) cohort, an ongoing population-based pregnancy and birth cohort set in Michigan’s lower peninsula. During the MARCH 3 month phone interview, mothers confirmed their interest in participating in this sample collection. Our analysis used the subset of the cohort whose mothers provided informed consent for providing infant stool samples. Fecal collection kits were sent by mail. The infants in this analysis were 3-9 months of age between 2018 and 2021 at time of fecal sample collection. 195 fecal samples collected from singleton infants were included in the analysis. 5.3.2 Data collection 92 Several questionnaires were administered to mothers from the first prenatal visit through 2 years postpartum. The questionnaire during the prenatal visit included demographic information about the mothers, their breastfeeding plans, health-related practices, and their estimated due date. Detailed information, including the infant's sex, birth weight, pregnancy complications, mode of delivery (vaginal vs C-section), pre-pregnancy BMI, and gestational age, were abstracted from the birth certificate. To assess sleep problems of the children, we administered the PROMIS Sleep Disturbance scale (PSDS, four items)[24] to mothers when their children were age 2. PSDS has demonstrated high internal consistency and strong construct validity among children in previous studies[24,25]. For our outcomes, we used two items from the PSDS to represent difficulty initiating sleep (my child had difficulty falling asleep) and difficulty maintaining sleep (my child could not sleep through the night). The items followed a 5-point format (1= never, 2= almost never, 3= sometimes, 4= almost always, 5= always). We considered the children to be free of difficulty initiating or maintaining sleep if the scores reported by mothers were less or equal than 2. 5.3.3 Fecal microbiota analysis Once received in the lab, the fecal samples were aliquoted into sterile tubes and stored at -80°C. DNA was extracted following a modified version of the Human Microbiome Project’s protocol described previously[26]. Barcoded primers were used to amplify the V4 region of the 16S rRNA gene following the mothur wet lab documentation. PCR amplification also followed the wet lab protocol outlined in the mothur documentation. The resulting 16S rRNA libraries were sequenced using 250 base pair Illumina MiSeq with V2 chemistry at the Michigan State University genomics core. After trimming, clean sequences were analyzed using the QIIME2 (2021.2 version) pipeline[27]. Demultiplexed sequences were further quality filtered and 93 clustered using QIIME2's DADA2 plugin to generate the ASV table[28]. Unique amplicon sequence variants (ASVs) were assigned a taxonomy by the QIIME2 feature-classifier plugin, using the Silva 132 database at the similarity thresh-old of 99% (for 16S data)[29,30]. Samples were rarefied to 6,000 sequencing reads per sample, leaving 191 stool samples with 6,905 unique ASVs, and findings were summarized at the genus taxonomic level. 5.3.4 Statistical analysis We used multivariate logistic regression to assess the association between characteristics of the study population and sleeping disorders at age of two years, with adjustment for breastfeeding status, delivery mode, birth weight, gestational age, age at fecal sample collection. Characteristic variables included maternal education level, pre-pregnancy BMI, race, sex, and maternal age. Gut microbiota was analyzed in terms of alpha diversity (Chao1 and Shannon) and beta diversity (Unweighted UniFrac and Weighted UniFrac) using the "vegan" package in R[31]. The difference in alpha diversity and relative abundance of taxa between sleep outcomes were tested by the Wilcoxon rank test with false discovery rate (FDR) correction for multiple comparisons. We also performed multivariate logistic regression models to estimate the association between alpha diversity and sleep disorders, with adjustment for breastfeeding status, delivery mode, birth weight, gestational age, age at fecal sample collection, maternal education level, pre-pregnancy BMI, race, baby sex, and maternal age. We performed Principal Coordinates Analysis (PCoA) and Permutational Multivariate Analysis of Variance (PERMANOVA) based on Unweighted UniFrac and Weighted UniFrac distance metric to compare the difference in microbial community structure between sleep disorders[32,33]. PERMANOVA is a non-parametric multivariate statistical test, with p-values obtained using appropriate distribution‐free permutation techniques. Differential abundance of microbial genera was determined using 94 generalized linear models with a negative binomial family and a log link function implemented in DESeq2[34], with adjustment by delivery mode, maternal education level, maternal age, gestational age, pre-pregnancy BMI, race, breastfeeding status, age at fecal sample collection, child sex, and birth weight. DESeq2 assumes that counts can be modeled as a negative binomial distribution with a mean parameter, allowing for size factors and a dispersion parameter[34]. 5.4 Results In total, 95 boys and 99 girls with data on sleep disturbance at age two years were included in this analysis. The mean age of the children when the PROMIS Sleeping Disturbance questionnaire was implemented to mothers was 24.2 (SD= 1.2) months. The mean age of the children at fecal sample collection was 4.0 (SD= 1.1) months. Mean birth weight was 3,352 (SD= 583) grams, and mean gestational age was 38.5 weeks (SD= 1.9). Mean (SD) maternal age was 30.9 (5.4) years and mean pre-pregnancy BMI was 27.9 (SD= 7.6). 100 children were exclusively breastfed, 33 were partially breastfed, and 61 were fed with exclusive formula at the time of fecal sample collection. At one year, 169 children were ever breastfed or fed with pumped breast milk and 25 were never fed with breast milk. Descriptive summary of study participants by difficulty initiating sleep and maintaining sleep are shown in Table 5.1. Over the 194 children, 61 (31%) were reported to have difficulty initiating sleep, 63 (32.5%) were reported to have difficulty maintaining sleep, 32 (16.4%) were reported to have both disorders, and 92 (47.2%) were reported to have at least one disorder. Univariate analysis showed that a lower proportion of mothers whose children had difficulty initiating sleep were of normal BMI (< 25.0) prior to pregnancy compared to those whose children did not (36.1% vs 47.8%, p-value= 0.004). Children with difficulty maintaining sleep were more likely to be never fed with breast milk, compared to those who were ever fed with breast milk (22.2% vs 8.5%, p- 95 value= 0.01). In a multivariate model, children whose mothers were overweight were more likely to have difficulty initiating sleep (OR = 3.7, CI:1.6-8.7, p-adjust = 0.004) and more likely to have difficulty maintaining sleep (OR = 2.7, CI:1.2-6.4, p-adjust = 0.04) at age 2, compared to those whose mothers with normal BMI (Table S5.1-S5.2). Increased birth weight tended to be associated with a lower risk of having difficulty initiating sleep in infants (OR = 0.4, CI:0.2-0.9, p-adjust = 0.08, Table S5.2). Table 5.1. Characteristics of mothers and infants by difficulty initiating sleep and maintaining sleep P- P- Difficulty initiating sleep Difficulty maintaining sleep value valu 1 Yes (61) No (134) Yes (63) No (131) e N/mea %/S N/mea %/S N/mea N/mea n D n D n %/SD n %/SD Delivery mode, n (%) Vaginal 62.90 72.50 delivery 37 61% 98 74% 39 % 95 % C- 37.10 27.50 section 24 39% 35 26% 0.1 23 % 36 % 0.24 Materna l educatio n level, n (%) 96 Table 5.1. (cont’d) Did not finish high school 5 8.50% 8 6% 7 11.70% 6 4.60% High school graduate or GED 6 10.20% 18 13.50% 9 15.00% 15 11.50% Some college 13 22.00% 31 23.30% 15 25.00% 29 22.10% College graduate or more 35 59.30% 76 57.20% 0.85 29 48.30% 81 61.80% 0.17 Maternal pre- pregnancy BMI Normal (<25) 22 36.10% 64 47.80% 21 33.30% 64 Overweight (25-30) 24 39.30% 23 17.20% 19 30.20% 28 Obesity (≥ 30) 15 24.60% 47 35.00% 0.004 23 36.50% 39 0.12 Race, n (%) White 44 74.60% 105 78.90% 42 70.00% 106 80.90% Black 11 18.60% 20 15.00% 13 21.70% 18 13.70% Other 4 6.80% 8 6.10% 0.8 5 8.30% 7 5.30% 0.21 97 Table 5.1. (cont’d) Breastfeeding status at fecal sample collection, n (%) Exclusive breastfeeding 31 51.70% 69 51.50% 27 43.50% 72 55.00% Partial breastfeeding 11 18.30% 22 16.40% 10 16.10% 23 17.60% Not breastfeeding 18 30% 43 32.10% 0.92 25 40.40% 36 27.40% 0.19 Ever exposed to breast milk, n (%) Yes 53 86.90% 116 87.20% 49 77.80% 119 91.50% No 8 13.10% 17 12.80% 0.99 14 22.20% 11 8.50% 0.01 Baby sex, n (%) Male 31 50.80% 64 47.80% 27 43.50% 68 51.90% Female 30 49.20% 69 52.20% 0.85 35 56.50% 63 48.10% 0.35 98 Table 5.1. (cont’d) Maternal age (year), mean (SD) 30.7 5.26 31 5.42 0.7 29.9 5.9 30.1 5 0.1 Birth weight (gram), mean (SD) 3251 620 3398 562 0.12 3244 587 3402 579 0.08 Gestational age (week), mean (SD) 38.3 2.1 38.6 1.8 0.28 38.2 2.2 38.7 1.7 0.14 Age at sleeping disturbance assessment 24.3 1.57 24.1 0.91 0.16 24.1 1.27 24.2 1.11 0.72 1 The difference between sleep disorders and characteristics was analyzed by Chi-square test or One-Way Analysis of Variance (ANOVA). Infancy fecal samples from children who have difficulty initiating sleep displayed a higher Shannon index (p-adjusted= 0.07) and a higher Chao 1 index (p-adjusted= 0.09) than samples from those who did not have difficulty initiating sleep (Figure 5.1A-5.1B). Infancy fecal samples from children who have difficulty maintaining sleep displayed a higher Shannon index (p- adjust= 0.004) and a higher Chao 1 index (p-adjust= 0.0004) than samples from those who did not have difficulty maintaining sleep (Figure 5.1C-5.1D). The multivariate model showed that children who had a higher Shannon index (OR = 2.4, CI:1.2-4.9, p-adjust = 0.04) and a higher Chao 1 index (OR = 1.01, CI:1.0-1.03, p-adjust = 0.008) at early infancy were more likely to have difficulty maintaining sleep (Table 5.2). There was also a non-significant trend towards an increased Shannon index (OR = 1.8, CI:0.95-3.6, p-adjust = 0.32) and an increased Chao 1 index 99 (OR = 1.1, CI:0.99-1.01, p-adjust = 0.64) being associated with a higher risk of having difficulty initiating sleep (Table 5.2). Figure 5.1. Boxplot of the alpha diversity (Shannon and Chao 1) by difficulty initiating and maintaining sleep. (A) Association between difficulty initiating sleep and Shannon index. P- adjust = 0.07 after Bonferroni correction. (B) Association between difficulty initiating sleep and Chao 1 index. P-adjust = 0.09 after Bonferroni correction. (C) Association between difficulty maintaining sleep and Shannon index. P-adjust = 0.008 after Bonferroni correction. (D) Association between difficulty maintaining sleep and Chao 1 index. P-adjust = 0.0008 after Bonferroni correction. 100 Table 5.2. Association between alpha diversity and difficulty initiating and maintaining sleep Shannon index Chao1 index Odds ratio p- adjusted p- Odds ratio (95% p- adjusted p- (95% CI) value value1 CI) value value Difficulty initiating sleep 1.83 (0.95, 1.05 (0.99, 3.64) 0.08 0.32 1.01) 0.16 0.64 Difficulty maintaining sleep 2.41 (1.23, 1.01 (1.00, 4.93) 0.01 0.04 1.03) 0.002 0.008 P-value was adjusted by Bonferroni correction. All models were adjusted for breastfeeding status, delivery mode, birth weight, gestational age, age at fecal sample collection, maternal education level, pre-pregnancy BMI, race, baby sex, and maternal age When classified by difficulty initiating sleep, the gut microbiota communities of the children were significantly differed based on unweighted UniFrac distance metric (PERMANOVA: R2= 0.9%, p-value =0.04, Figure 5.2B), but were similar when compared based on Weighted UniFrac distance metric (PERMANOVA: R2= 0.4%, p-value =0.57, Figure 5.2A). When classified by difficulty maintaining sleep, the gut microbiota communities were significantly differed based on the unweighted UniFrac distance metric (PERMANOVA: 1.8%, p-value =0.004, Figure 5.2D), but were not distinct based on the weighted UniFrac distance metric (PERMANOVA: R2= 0.9%, p-value =0.1, Figure 5.2C). We then performed differential abundance testing with DESeq2 to see which taxa were associated with the sleep problems. At the genus level, we detected a higher abundance of Desulfovibrio, Butyricimonas, Roseburia, and a lower abundance of the Lachnospiraceae NK4A136 group, Collinsella in the children with difficulty initiating sleep (q-value < 0.001, Table 5.3). We also identified a higher abundance of Eubacterium coprostanoligenes group, Epulopiscium, Colidextribacter, Lactobacillus, Megamonas, Faecalibacterium, Roseburia, and a lower abundance of Lachnospiraceae uncultured in the children with difficulty maintaining sleep (q-value < 0.001, Table 5.4). 101 Figure 5.2. Principal Coordinates Analysis (PCoA) for difficulty initiating and maintaining sleep. R2 and p-value were calculated by the PERMANOVA test. (A) PCoA for difficulty initiating sleep based on weighted UniFrac distance metric. (B) PCoA for difficulty initiating sleep based on unweighted UniFrac distance metric. (C) PCoA for difficulty maintaining sleep based on weighted UniFrac distance metric. (D) PCoA for difficulty maintaining sleep based on unweighted UniFrac distance metric. 102 Table 5.3. DESeq2 Analysis Results: Associations of gut microbiome taxa at genus level and difficulty initiating sleep adjusted by covariates1 Log 2 folder Family Genus change p-value p-adjusted -8 -6 Desulfovibrionaceae Desulfovibrio 1.32 7.48×10 8.37×10 -6 -5 Marinifilaceae Butyricimonas 1.30 1.52×10 5.67×10 -6 -5 Lachnospiraceae NK4A136 group -1.94 1.20×10 5.67×10 -6 -4 Coriobacteriaceae Collinsella -1.58 3.62×10 1.02×10 -5 -4 Lachnospiraceae Roseburia 1.45 3.97×10 8.90×10 1 All models were adjusted for breastfeeding status, delivery mode, birth weight, gestational age, age at fecal sample collection, maternal education level, pre-pregnancy BMI, race, baby sex, and maternal age. P-adjusted is the FDR (Benjamini-Hochberg) adjusted p-value. p-adjusted < 0.001 for multiple comparisons was considered statistically significant and included in the table. Table 5.4. DESeq2 Analysis Results: Associations of gut microbiome taxa at genus level and difficulty maintaining sleep adjusted by covariates1 Log 2 folder Adjuste chang d p- Family Genus e p-value value Eubacterium coprostanoligenes gro Eubacterium coprostanoligenes gro 3.32×10 3.56×10 -9 -7 up up 1.81 3.09×10 3.56×10 -7 -7 Lachnospiraceae Epulopiscium 1.98 3.48×10 1.24×10 -7 -5 Oscillospiraceae Colidextribacter 1.98 6.12×10 1.64×10 -7 -5 Lachnospiraceae Lachnospiraceae uncultured -2.29 1.11×10 2.38×10 -6 -5 Lactobacillaceae Lactobacillus 1.85 1.32×10 -5 -4 Selenomonadaceae Megamonas 1.50 2.2×10 1.44×10 -5 -4 Ruminococcaceae Faecalibacterium 1.41 2.2×10 7.14×10 9.55×10 -5 -4 Lachnospiraceae Roseburia 1.39 1 All models were adjusted for breastfeeding status, delivery mode, birth weight, gestational age, age at fecal sample collection, maternal education level, pre-pregnancy BMI, race, baby sex, and maternal age. P-adjusted is the FDR (Benjamini-Hochberg) adjusted p-value. p-adjusted < 0.001 for multiple comparisons was considered statistically significant and included in the table. 103 5.5 Discussion Our prospective study demonstrated that the composition of the gut microbiome at early infancy was significantly associated with difficulty initiating sleep and maintaining sleep at the age of 2 years. We also identified those microbial candidates that might contribute to both sleep disorders. Our results suggested a clear association between infant gut microbiota composition and difficulty initiating sleep and maintaining sleep. Due to different inclusion criteria, prevalence estimates of insomnia symptoms in childhood have varied from 4–41%[35–39]. In the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-4), a patient with difficulty initiating or maintaining sleep for at least one month was considered having a risk for insomnia. In our study, nearly half of the population had difficulty initiating sleep or maintaining sleep, but our survey did not ask for the of these symptoms. The univariate analysis showed that difficulty initiating sleep was significantly associated with maternal pre-pregnancy BMI. The multivariate analysis confirmed this finding that children whose mothers were overweight were more likely to have difficulty initiating and maintaining sleep. This association could possibly be mediated by child obesity, which was shown to be significantly associated with maternal pre-pregnancy BMI in many studies. Obesity, which is a risk factor for sleep disorders of all ages, has been linked to short sleep duration and poor sleep quality during childhood[40]. However, child BMI was not recorded in our study. The univariate analysis also showed that children with difficulty maintaining sleep were more likely to be never fed with breast milk, although this association disappeared in the multivariate analysis with the odds ratio decreasing from 3.1 to 1.7. Breastfeeding was found in other studies to be associated with increased night waking and sleep fragmentation at infancy due to parenting practices of nursing at night[41–43]. However, this 104 association disappeared by the 9-month follow-up[41,44], suggesting that sleep disruptions associated with breastfeeding resolve when the breastfeeding ends. Turning to the microbiome findings, our results demonstrated that alpha diversity at 3 months of age predicted sleep disorder at 2 years of age. Higher Shannon index and Chao 1 index were significantly associated with a higher risk of difficulty maintaining sleep after adjustment of covariates. There was also a trend of increased Shannon index and Chao 1 index in the children with difficulty initiating sleep. The alpha diversity of gut microbiota is dynamic during infancy and is correlated with complementary feeding and delivery mode. High alpha diversity in infancy indicates a more mature, adult-like community[65]. Carlson et al. demonstrated that high alpha diversity at infancy was associated with poor cognitive development at 2 years of age[45]. However, numerous studies have shown that low alpha diversity in infancy is associated with asthma and type 1 diabetes[66-67]. A highly diverse microbiome can introduce functional redundancy[46], which allows individuals to adapt to environmental fluctuations, maintain intestinal homeostasis and support human health. Our results, together with the work done by Carlson et al.[45] suggest that increased alpha diversity at early infancy could contribute to sleep problems in childhood. We noted that the structure of the infant gut microbiota affected difficulty initiating and maintaining sleep when the unweighted UniFrac distance metric was used. An unweighted UniFrac distance considers only species presence and absence information and counts the fraction of branch length unique to either community, while a weighted UniFrac distance uses species abundance information and weights the branch length with abundance difference[47,48]. Thus, unweighted UniFrac distance is most efficient in detecting abundance change in rare taxa. 105 Our results suggested that the association between gut microbiota composition and sleep disorders was potentially driven by rare and less abundant taxa[49,50]. Our results indicated that several taxa were positively or negatively associated with infant sleep problems, suggesting different taxa were involved in the mechanisms that influence the sleep disorders in children. Our results agree with the previous literature, which suggests that the genera Desulfovibrio and Lactobacillus were significantly enriched in the gut microbiome of insomniac adult participants[51,52]. An animal study also confirmed that Desulfovibrio could increase the risk of obstructive sleep apnea (OSA) by causing intermittent hypoxia during sleep[53]. Many studies have also confirmed that a higher abundance of Desulfovibrio is associated with the incidence and the severity of autism[54,55]. Desulfovibrio is a sulfate- reducing bacterium in the human gut that can generate hydrogen sulfide, which is an effective inhibitor of the oxidation of short-chain fatty acids (SCFAs) in cells[56,57]. Several species in the Lactobacillus genus were found to have beneficial effects on sleep rhythms in animal models, which is in contradiction to our results[58,59]. However, some Lactobacillus species can produce GABA, and abnormal expression of GABA mRNA is linked to depression and insomnia[60,61]. Thus, more studies are needed at the species level of Lactobacillus to understand the mechanism of the influence of gut Lactobacillus on sleep disorders. An important limitation of this study is that only one stool sample collected. Gut microbial ecology and function are dynamic across the infancy lifestage by time and influenced by multiple factors, including feeding practice, delivery mode, and environmental exposures[62–64]. Future studies would benefit from a longitudinal stool sample collection during infancy that helps us better understand the association between infant gut microbiome and sleep disorders. Another limitation is that 16S rRNA sequencing can provide limited strain-level information. Previous studies 106 demonstrate a complex mechanism for several species of Lactobacillus in sleep disorders[60]. Thus, Shotgun metagenomic sequencing is needed for future studies. 107 APPENDIX 108 Table S5.1. Association between maternal/infant characteristics and difficulty falling sleep in infants1 Odds ratio (95% CI) P-value Delivery mode, n (%) Vaginal delivery ref C-section 1.94 (0.91,4.15) 0.08 Maternal education level, n (%) Did not finish high school ref High school graduate or GED 0.55 (0.10, 3.18) 0.49 Some college 0.87 (0.18, 4.73) 0.87 College graduate or more 0.85 (0.17, 4.64) 0.84 Maternal pre-pregnancy BMI Normal ref Overweight 3.73 (1.64, 8.72) 0.002* Obesity 0.85 (0.35, 2.03) 0.72 Race, n (%) White ref Black 1.08(0.34, 3.4) 0.89 Other 1.22(0.29, 4.48) 0.77 Breastfeeding status at fecal sample collection, n (%) Exclusive breastfeeding ref Partial breastfeeding 0.87(0.33, 2.22) 0.78 Not breastfeeding 0.89(0.33, 2.33) 0.82 Ever breastfeeding, n (%) Yes ref No 0.85(0.22, 3.15) 0.81 Baby sex, n (%) Male ref Female 1.11(0.56, 2.23) 0.77 Maternal age (year) 0.98(0.91, 1.05) 0.62 Birth weight (kilogram) 0.41(0.17, 0.94) 0.04* Gestational age (week) 1.13(0.89, 1.46) 0.32 1 All variables listed in table were included in the multivariable model. *<0.05 109 Table S5.2. Association between maternal/infant characteristics and difficulty maintaining sleep in infants1 Odds ratio (95% CI) P-value Delivery mode, n (%) Vaginal delivery ref C-section 1.5(0.71, 3.15) 0.29 Maternal education level, n (%) Did not finish high school ref High school graduate or GED 0.52(0.11, 2.39) 0.40 Some college 0.69(0.16, 3.0) 0.62 College graduate or more 0.51(0.11, 2.28) 0.37 Maternal pre-pregnancy BMI Normal ref Overweight 2.69(1.15, 6.39) 0.02* Obesity 1.49(0.64, 3.48) 0.35 Race, n (%) White ref Black 0.87(0.29, 2.57) 0.81 Other 2.1(0.54, 7.74) 0.26 Breastfeeding status at fecal sample collection, n (%) Exclusive breastfeeding ref Partial breastfeeding 0.92(0.35, 2.33) 0.86 Not breastfeeding 0.99(0.38, 2.49) 0.98 Ever breastfeeding, n (%) Yes ref No 1.72(0.51, 5.81) 0.38 Baby sex, n (%) Male ref Female 1.64(0.83, 3.29) 0.16 Maternal age (year) 0.97(0.9, 1.04) 0.35 Birth weight (kilogram) 0.63(0.27, 1.38) 0.26 Gestational age (week) 1.01(0.80, 1.28) 0.93 1 All variables listed in table were included in the multivariable model. *<0.05 110 REFERENCES 111 REFERENCES 1. Owens, J.A.; Spirito, A.; McGuinn, M.; Nobile, C. Sleep Habits and Sleep Disturbance in Elementary School-Aged Children. J Dev Behav Pediatr 2000, 21, 27–36, doi:10.1097/00004703-200002000-00005. 2. Richdale, A.L.; Schreck, K.A. Sleep Problems in Autism Spectrum Disorders: Prevalence, Nature, & Possible Biopsychosocial Aetiologies. Sleep Medicine Reviews 2009, 13, 403–411, doi:10.1016/j.smrv.2009.02.003. 3. Hilliard, T. Principles and Practice of Pediatric Sleep Medicine. Arch Dis Child 2006, 91, 546–547, doi:10.1136/adc.2006.093955. 4. Archbold, K.H.; Pituch, K.J.; Panahi, P.; Chervin, R.D. Symptoms of Sleep Disturbances among Children at Two General Pediatric Clinics. J Pediatr 2002, 140, 97–102, doi:10.1067/mpd.2002.119990. 5. Kraenz, S.; Fricke, L.; Wiater, A.; Mitschke, A.; Breuer, U.; Lehmkuhl, G. [Prevalence and stress factors of sleep disorders in children starting school]. Prax Kinderpsychol Kinderpsychiatr 2004, 53, 3–18. 6. Owens, J.; Group, A.S.W.; Adolescence, C.O. Insufficient Sleep in Adolescents and Young Adults: An Update on Causes and Consequences. Pediatrics 2014, 134, e921–e932, doi:10.1542/peds.2014-1696. 7. Vgontzas, A.N.; Bixler, E.O.; Basta, M. Obesity and Sleep: A Bidirectional Association? Sleep 2010, 33, 573–574. 8. Vgontzas, A.N.; Lin, H.-M.; Papaliaga, M.; Calhoun, S.; Vela-Bueno, A.; Chrousos, G.P.; Bixler, E.O. Short Sleep Duration and Obesity: The Role of Emotional Stress and Sleep Disturbances. Int J Obes (Lond) 2008, 32, 801–809, doi:10.1038/ijo.2008.4. 9. Liu, J.; Liu, X.; Pak, V.; Wang, Y.; Yan, C.; Pinto-Martin, J.; Dinges, D. Early Blood Lead Levels and Sleep Disturbance in Preadolescence. Sleep 2015, 38, 1869–1874, doi:10.5665/sleep.5230. 10. Taheri, S.; Lin, L.; Austin, D.; Young, T.; Mignot, E. Short Sleep Duration Is Associated with Reduced Leptin, Elevated Ghrelin, and Increased Body Mass Index. PLOS Medicine 2004, 1, e62, doi:10.1371/journal.pmed.0010062. 11. Kohatsu, N.D.; Tsai, R.; Young, T.; VanGilder, R.; Burmeister, L.F.; Stromquist, A.M.; Merchant, J.A. Sleep Duration and Body Mass Index in a Rural Population. Archives of Internal Medicine 2006, 166, 1701–1705, doi:10.1001/archinte.166.16.1701. 112 12. Su, M.-S.; Zhang, H.-L.; Cai, X.-H.; Lin, Y.; Liu, P.-N.; Zhang, Y.-B.; Hu, W.-Z.; Li, C.-C.; Xiao, Y.-F. Obesity in Children with Different Risk Factors for Obstructive Sleep Apnea: A Community-Based Study. Eur J Pediatr 2016, 175, 211–220, doi:10.1007/s00431-015-2613-6. 13. Liu, J.; Liu, X.; Pak, V.; Wang, Y.; Yan, C.; Pinto-Martin, J.; Dinges, D. Early Blood Lead Levels and Sleep Disturbance in Preadolescence. Sleep 2015, 38, 1869–1874, doi:10.5665/sleep.5230. 14. Krystal, A.D. PSYCHIATRIC DISORDERS AND SLEEP. Neurol Clin 2012, 30, 1389– 1413, doi:10.1016/j.ncl.2012.08.018. 15. Chervin, R.D.; Archbold, K.H.; Dillon, J.E.; Pituch, K.J.; Panahi, P.; Dahl, R.E.; Guilleminault, C. Associations between Symptoms of Inattention, Hyperactivity, Restless Legs, and Periodic Leg Movements. Sleep 2002, 25, 213–218. 16. Johnson, E.O.; Chilcoat, H.D.; Breslau, N. Trouble Sleeping and Anxiety/Depression in Childhood. Psychiatry Res 2000, 94, 93–102, doi:10.1016/s0165-1781(00)00145-1. 17. Carvalho Bos, S.; Gomes, A.; Clemente, V.; Marques, M.; Pereira, A.T.; Maia, B.; Soares, M.J.; Cabral, A.S.; Macedo, A.; Gozal, D.; et al. Sleep and Behavioral/Emotional Problems in Children: A Population-Based Study. Sleep Med 2009, 10, 66–74, doi:10.1016/j.sleep.2007.10.020. 18. Morais, L.H.; Schreiber, H.L.; Mazmanian, S.K. The Gut Microbiota–Brain Axis in Behaviour and Brain Disorders. Nat Rev Microbiol 2021, 19, 241–255, doi:10.1038/s41579-020- 00460-0. 19. Wagner-Skacel, J.; Dalkner, N.; Moerkl, S.; Kreuzer, K.; Farzi, A.; Lackner, S.; Painold, A.; Reininghaus, E.Z.; Butler, M.I.; Bengesser, S. Sleep and Microbiome in Psychiatric Diseases. Nutrients 2020, 12, 2198, doi:10.3390/nu12082198. 20. Matenchuk, B.A.; Mandhane, P.J.; Kozyrskyj, A.L. Sleep, Circadian Rhythm, and Gut Microbiota. Sleep Medicine Reviews 2020, 53, 101340, doi:10.1016/j.smrv.2020.101340. 21. Valentini, F.; Evangelisti, M.; Arpinelli, M.; Di Nardo, G.; Borro, M.; Simmaco, M.; Villa, M.P. Gut Microbiota Composition in Children with Obstructive Sleep Apnoea Syndrome: A Pilot Study. Sleep Medicine 2020, 76, 140–147, doi:10.1016/j.sleep.2020.10.017. 22. Anderson, J.R.; Carroll, I.; Azcarate-Peril, M.A.; Rochette, A.D.; Heinberg, L.J.; Peat, C.; Steffen, K.; Manderino, L.M.; Mitchell, J.; Gunstad, J. A Preliminary Examination of Gut Microbiota, Sleep, and Cognitive Flexibility in Healthy Older Adults. Sleep Medicine 2017, 38, 104–107, doi:10.1016/j.sleep.2017.07.018. 23. Morin, C.M.; Drake, C.L.; Harvey, A.G.; Krystal, A.D.; Manber, R.; Riemann, D.; Spiegelhalder, K. Insomnia Disorder. Nat Rev Dis Primers 2015, 1, 1–18, doi:10.1038/nrdp.2015.26. 113 24. Yu, L.; Buysse, D.J.; Germain, A.; Moul, D.E.; Stover, A.; Dodds, N.E.; Johnston, K.L.; Pilkonis, P.A. Development of Short Forms From the PROMISTM Sleep Disturbance and Sleep- Related Impairment Item Banks. Behavioral Sleep Medicine 2012, 10, 6–24, doi:10.1080/15402002.2012.636266. 25. Hanish, A.E.; Lin-Dyken, D.C.; Han, J.C. PROMIS Sleep Disturbance and Sleep-Related Impairment in Adolescents: Examining Psychometrics Using Self-Report and Actigraphy. Nurs Res 2017, 66, 246–251, doi:10.1097/NNR.0000000000000217. 26. Sugino, K.Y.; Paneth, N.; Comstock, S.S. Michigan Cohorts to Determine Associations of Maternal Pre-Pregnancy Body Mass Index with Pregnancy and Infant Gastrointestinal Microbial Communities: Late Pregnancy and Early Infancy. PLoS One 2019, 14, e0213733, doi:10.1371/journal.pone.0213733. 27. Bolyen, E.; Rideout, J.R.; Dillon, M.R.; Bokulich, N.A.; Abnet, C.C.; Al-Ghalith, G.A.; Alexander, H.; Alm, E.J.; Arumugam, M.; Asnicar, F.; et al. Reproducible, Interactive, Scalable and Extensible Microbiome Data Science Using QIIME 2. Nature Biotechnology 2019, 37, 852– 857, doi:10.1038/s41587-019-0209-9. 28. Callahan, B.J.; McMurdie, P.J.; Rosen, M.J.; Han, A.W.; Johnson, A.J.A.; Holmes, S.P. DADA2: High-Resolution Sample Inference from Illumina Amplicon Data. Nature Methods 2016, 13, 581–583, doi:10.1038/nmeth.3869. 29. Bokulich, N.A.; Kaehler, B.D.; Rideout, J.R.; Dillon, M.; Bolyen, E.; Knight, R.; Huttley, G.A.; Gregory Caporaso, J. Optimizing Taxonomic Classification of Marker-Gene Amplicon Sequences with QIIME 2’s Q2-Feature-Classifier Plugin. Microbiome 2018, 6, 90, doi:10.1186/s40168-018-0470-z. 30. Quast, C.; Pruesse, E.; Yilmaz, P.; Gerken, J.; Schweer, T.; Yarza, P.; Peplies, J.; Glöckner, F.O. The SILVA Ribosomal RNA Gene Database Project: Improved Data Processing and Web- Based Tools. Nucleic Acids Research 2013, 41, D590–D596, doi:10.1093/nar/gks1219. 31. Dixon, P. VEGAN, a Package of R Functions for Community Ecology. Journal of Vegetation Science 2003, 14, 927–930, doi:10.1111/j.1654-1103.2003.tb02228.x. 32. Anderson, M.J. Permutational Multivariate Analysis of Variance (PERMANOVA). In Wiley StatsRef: Statistics Reference Online; American Cancer Society, 2017; pp. 1–15 ISBN 978-1- 118-44511-2. 33. Lozupone, C.; Lladser, M.E.; Knights, D.; Stombaugh, J.; Knight, R. UniFrac: An Effective Distance Metric for Microbial Community Comparison. ISME J 2011, 5, 169–172, doi:10.1038/ismej.2010.133. 34. Love, M.I.; Huber, W.; Anders, S. Moderated Estimation of Fold Change and Dispersion for RNA-Seq Data with DESeq2. Genome Biology 2014, 15, 550, doi:10.1186/s13059-014-0550-8. 114 35. Johnson, E.O.; Roth, T.; Schultz, L.; Breslau, N. Epidemiology of DSM-IV Insomnia in Adolescence: Lifetime Prevalence, Chronicity, and an Emergent Gender Difference. Pediatrics 2006, 117, e247-256, doi:10.1542/peds.2004-2629. 36. Camhi, null; Morgan, null; Pernisco, null; Quan, null Factors Affecting Sleep Disturbances in Children and Adolescents. Sleep Med 2000, 1, 117–123, doi:10.1016/s1389-9457(99)00005-2. 37. Zhang, J.; Li, A.M.; Kong, A.P.S.; Lai, K.Y.C.; Tang, N.L.S.; Wing, Y.K. A Community- Based Study of Insomnia in Hong Kong Chinese Children: Prevalence, Risk Factors and Familial Aggregation. Sleep Med 2009, 10, 1040–1046, doi:10.1016/j.sleep.2009.01.008. 38. Singareddy, R.; Moole, S.; Calhoun, S.; Vocalan, P.; Tsaoussoglou, M.; Vgontzas, A.N.; Bixler, E.O. Medical Complaints Are More Common in Young School-Aged Children with Parent Reported Insomnia Symptoms. J Clin Sleep Med 2009, 5, 549–553. 39. Archbold, K.H.; Pituch, K.J.; Panahi, P.; Chervin, R.D. Symptoms of Sleep Disturbances among Children at Two General Pediatric Clinics. J Pediatr 2002, 140, 97–102, doi:10.1067/mpd.2002.119990. 40. Morrissey, B.; Taveras, E.; Allender, S.; Strugnell, C. Sleep and Obesity among Children: A Systematic Review of Multiple Sleep Dimensions. Pediatric Obesity 2020, 15, e12619, doi:10.1111/ijpo.12619. 41. Mindell, J.A.; Du Mond, C.; Tanenbaum, J.B.; Gunn, E. Long-Term Relationship Between Breastfeeding and Sleep. Children’s Health Care 2012, 41, 190–203, doi:10.1080/02739615.2012.685038. 42. Touchette, É.; Petit, D.; Paquet, J.; Boivin, M.; Japel, C.; Tremblay, R.E.; Montplaisir, J.Y. Factors Associated With Fragmented Sleep at Night Across Early Childhood. Archives of Pediatrics & Adolescent Medicine 2005, 159, 242–249, doi:10.1001/archpedi.159.3.242. 43. Ramamurthy, M.B.; Sekartini, R.; Ruangdaraganon, N.; Huynh, D.H.T.; Sadeh, A.; Mindell, J.A. Effect of Current Breastfeeding on Sleep Patterns in Infants from Asia-Pacific Region. Journal of Paediatrics and Child Health 2012, 48, 669–674, doi:10.1111/j.1440- 1754.2012.02453.x. 44. Wolke, D.; Söhne, B.; Riegel, K.; Ohrt, B.; Osterlund, K. An Epidemiologic Longitudinal Study of Sleeping Problems and Feeding Experience of Preterm and Term Children in Southern Finland: Comparison with a Southern German Population Sample. J Pediatr 1998, 133, 224–231, doi:10.1016/s0022-3476(98)70224-0. 45. Carlson, A.L.; Xia, K.; Azcarate-Peril, M.A.; Goldman, B.D.; Ahn, M.; Styner, M.A.; Thompson, A.L.; Geng, X.; Gilmore, J.H.; Knickmeyer, R.C. Infant Gut Microbiome Associated With Cognitive Development. Biological Psychiatry 2018, 83, 148–159, doi:10.1016/j.biopsych.2017.06.021. 46. Moya, A.; Ferrer, M. Functional Redundancy-Induced Stability of Gut Microbiota Subjected to Disturbance. Trends in Microbiology 2016, 24, 402–413, doi:10.1016/j.tim.2016.02.002. 115 47. Chen, J.; Bittinger, K.; Charlson, E.S.; Hoffmann, C.; Lewis, J.; Wu, G.D.; Collman, R.G.; Bushman, F.D.; Li, H. Associating Microbiome Composition with Environmental Covariates Using Generalized UniFrac Distances. Bioinformatics 2012, 28, 2106–2113, doi:10.1093/bioinformatics/bts342. 48. Lozupone, C.; Lladser, M.E.; Knights, D.; Stombaugh, J.; Knight, R. UniFrac: An Effective Distance Metric for Microbial Community Comparison. ISME J 2011, 5, 169–172, doi:10.1038/ismej.2010.133. 49. Chen, J.; Wright, K.; Davis, J.M.; Jeraldo, P.; Marietta, E.V.; Murray, J.; Nelson, H.; Matteson, E.L.; Taneja, V. An Expansion of Rare Lineage Intestinal Microbes Characterizes Rheumatoid Arthritis. Genome Medicine 2016, 8, 43, doi:10.1186/s13073-016-0299-7. 50. Chen, J.; Bittinger, K.; Charlson, E.S.; Hoffmann, C.; Lewis, J.; Wu, G.D.; Collman, R.G.; Bushman, F.D.; Li, H. Associating Microbiome Composition with Environmental Covariates Using Generalized UniFrac Distances. Bioinformatics 2012, 28, 2106–2113, doi:10.1093/bioinformatics/bts342. 51. Man, H.; Chen, B.; Wang, Q.; Fu, S.; Xie, G.; Wang, J.; Zhao, C.; Gai, Z.; Zhang, C.; Heng, X.; et al. Correlations of Gut Microbiome, Serum Metabolome and Immune Factors in Insomnia 2021. 52. Mashaqi, S.; Gozal, D. Obstructive Sleep Apnea and Systemic Hypertension: Gut Dysbiosis as the Mediator? Journal of Clinical Sleep Medicine 15, 1517–1527, doi:10.5664/jcsm.7990. 53. Moreno-Indias, I.; Torres, M.; Montserrat, J.M.; Sanchez-Alcoholado, L.; Cardona, F.; Tinahones, F.J.; Gozal, D.; Poroyko, V.A.; Navajas, D.; Queipo-Ortuño, M.I.; et al. Intermittent Hypoxia Alters Gut Microbiota Diversity in a Mouse Model of Sleep Apnoea. Eur Respir J 2015, 45, 1055–1065, doi:10.1183/09031936.00184314. 54. Tomova, A.; Husarova, V.; Lakatosova, S.; Bakos, J.; Vlkova, B.; Babinska, K.; Ostatnikova, D. Gastrointestinal Microbiota in Children with Autism in Slovakia. Physiol Behav 2015, 138, 179–187, doi:10.1016/j.physbeh.2014.10.033. 55. Finegold, S.M. Desulfovibrio Species Are Potentially Important in Regressive Autism. Med Hypotheses 2011, 77, 270–274, doi:10.1016/j.mehy.2011.04.032. 56. Dordević, D.; Jančíková, S.; Vítězová, M.; Kushkevych, I. Hydrogen Sulfide Toxicity in the Gut Environment: Meta-Analysis of Sulfate-Reducing and Lactic Acid Bacteria in Inflammatory Processes. Journal of Advanced Research 2021, 27, 55–69, doi:10.1016/j.jare.2020.03.003. 57. Roediger, W.E.; Duncan, A.; Kapaniris, O.; Millard, S. Reducing Sulfur Compounds of the Colon Impair Colonocyte Nutrition: Implications for Ulcerative Colitis. Gastroenterology 1993, 104, 802–809, doi:10.1016/0016-5085(93)91016-b. 58. Wagner-Skacel, J.; Dalkner, N.; Moerkl, S.; Kreuzer, K.; Farzi, A.; Lackner, S.; Painold, A.; Reininghaus, E.Z.; Butler, M.I.; Bengesser, S. Sleep and Microbiome in Psychiatric Diseases. Nutrients 2020, 12, 2198, doi:10.3390/nu12082198. 116 59. Lin, A.; Shih, C.-T.; Chu, H.-F.; Chen, C.-W.; Cheng, Y.-T.; Wu, C.-C.; Yang, C.C.H.; Tsai, Y.-C. Lactobacillus Fermentum PS150 Promotes Non-Rapid Eye Movement Sleep in the First Night Effect of Mice. Sci Rep 2021, 11, 16313, doi:10.1038/s41598-021-95659-3. 60. Li, Y.; Hao, Y.; Fan, F.; Zhang, B. The Role of Microbiome in Insomnia, Circadian Disturbance and Depression. Front Psychiatry 2018, 9, 669, doi:10.3389/fpsyt.2018.00669. 61. Barrett, E.; Ross, R.P.; O’Toole, P.W.; Fitzgerald, G.F.; Stanton, C. γ-Aminobutyric Acid Production by Culturable Bacteria from the Human Intestine. J Appl Microbiol 2012, 113, 411– 417, doi:10.1111/j.1365-2672.2012.05344.x. 62. Kostic, A.D.; Gevers, D.; Siljander, H.; Vatanen, T.; Hyötyläinen, T.; Hämäläinen, A.-M.; Peet, A.; Tillmann, V.; Pöhö, P.; Mattila, I.; et al. The Dynamics of the Human Infant Gut Microbiome in Development and in Progression toward Type 1 Diabetes. Cell Host & Microbe 2015, 17, 260–273, doi:10.1016/j.chom.2015.01.001. 63. Gibson, M.K.; Wang, B.; Ahmadi, S.; Burnham, C.-A.D.; Tarr, P.I.; Warner, B.B.; Dantas, G. Developmental Dynamics of the Preterm Infant Gut Microbiota and Antibiotic Resistome. Nat Microbiol 2016, 1, 1–10, doi:10.1038/nmicrobiol.2016.24. 64. Bäckhed, F.; Roswall, J.; Peng, Y.; Feng, Q.; Jia, H.; Kovatcheva-Datchary, P.; Li, Y.; Xia, Y.; Xie, H.; Zhong, H.; et al. Dynamics and Stabilization of the Human Gut Microbiome during the First Year of Life. Cell Host & Microbe 2015, 17, 690–703, doi:10.1016/j.chom.2015.04.004. 65. Roswall, J.; Olsson, L.M.; Kovatcheva-Datchary, P.; Nilsson, S.; Tremaroli, V.; Simon, M.- C.; Kiilerich, P.; Akrami, R.; Krämer, M.; Uhlén, M.; et al. Developmental Trajectory of the Healthy Human Gut Microbiota during the First 5 Years of Life. Cell Host Microbe 2021, 29, 765-776.e3, doi:10.1016/j.chom.2021.02.021. 66. Kostic, A.D.; Gevers, D.; Siljander, H.; Vatanen, T.; Hyötyläinen, T.; Hämäläinen, A.-M.; Peet, A.; Tillmann, V.; Pöhö, P.; Mattila, I.; et al. The Dynamics of the Human Infant Gut Microbiome in Development and in Progression toward Type 1 Diabetes. Cell Host Microbe 2015, 17, 260–273, doi:10.1016/j.chom.2015.01.001. 67. Abrahamsson, T.R.; Jakobsson, H.E.; Andersson, A.F.; Björkstén, B.; Engstrand, L.; Jenmalm, M.C. Low Gut Microbiota Diversity in Early Infancy Precedes Asthma at School Age. Clin. Exp. Allergy J. Br. Soc. Allergy Clin. Immunol. 2014, 44, 842–850, doi:10.1111/cea.12253. 117 CHAPTER 6. SUMMARY, LIMITATIONS AND FUTURE RESEARCH Our studies examined whether feeding practices affect early gut microbial colonization and whether gut microbial colonization influence temperament and sleep disorders in children using a sample of singleton births from an ongoing population-based pregnancy and birth cohort set in Michigan’s lower peninsula. In this prospective study, fecal samples were collected between 3-9 months, and the bacterial 16S rRNA gene (V4 hypervariable regions) was amplified. We then collected temperament and sleep outcomes of infants at 9 months and 2 years, respectively. In the first analysis, we confirmed the importance of breastfeeding and vitamin D supplements in shaping gut microbiota composition in early infancy. By performing differential abundance analysis on individual taxa, our results demonstrated that vitamin D supplement potentially reduced the likelihood of Haemophilus colonization in exclusively breastfed infants. The Vitamin D family is a group of fat-soluble secosteroids absorbed from sunlight, food or supplements. Vitamin D exists in several forms, including ergocalciferol (D2), cholecalciferol (D3), 1-hydroxylated and 1,25-hydroxylated forms[12-13]. In addition to the food, humans acquire vitamin D synthesized in the skin upon exposure to sunlight. vitamin D synthesis is affected by season, time of day, skin pigmentation, cloth, aging etc.[13]. The liver and kidneys convert vitamin D into 25-hydroxyvitamin D which is a major circulating form and 1,25- dihydroxyvitamin D which is the biologically active form, respectively[13-14]. Vitamin D is essential for bone development and calcium homeostasis[15]. In addition, vitamin D may play a critical role in immune regulation, partially via modulating the gut microbial composition[16]. Vitamin D deficiency is associated with dysbiosis that promotes inflammation[16]. Our results confirm and extend existing findings and point out that vitamin D potentially protects against Haemophilus, which can cause a wide variety of infections in infants. However, only 39.8% of the exclusively breastfed infants received vitamin D supplements in our study population. Our 118 results highlight the need for actions to support the promotion of vitamin D among exclusive infants. In the analysis of the next two aims, we focused on temperament and sleep disorders which are two critical outcomes closely related to brain development during childhood. Uncovering the relationship between gut microbiota and these two outcomes will help us better understand the operation of the microbiota-gut-brain axis. Our results demonstrated an association between gut microbiota composition and the temperament characteristic described as negative emotionality. Negative emotionality, defined as a disposition to experience negative emotions such as anger and fear[1], is one of the most important dimensions of infant temperament. An excess of negative emotionality may contribute to a high risk of later childhood psychopathology and behavior problems[2]. Thus, the present study provides further evidence of the critical functional roles of the gut microbiota in behavior problems. In a sensitivity analysis, we discovered a potential role of infant vitamin D intake in modifying the association between microbiota clusters and negative emotionality. The association described above was identified only among the infants who were not recipients of vitamin D supplements, suggesting that infant vitamin D intake potentially protects against the adverse effect of the gut microbiome–associated on infant emotionality. Finally, we demonstrated that the composition of the gut microbiome in early infancy was significantly associated with difficulty initiating sleep and maintaining sleep at the age of 2 years. Epidemiologic studies indicate that up to 50% of children have experienced a sleep problem by age six[3]. A growing body of evidence shows that sleep disorders and insomnia in early childhood may be linked obesity, diabetes, and inflammatory diseases in later life[4–6]. 119 Previous research has focused on understanding the psychological, social, and physiological factors that regulate sleep. Our study, together with recent studies, provides a new basis for understanding the mechanisms of sleep problems. Circadian Rhythm, which is an important process in the regulation of sleep, has been linked to gut microbiota in recent studies[7]. Studies have demonstrated that the gut microbiota exhibit compositional and functional rhythmic fluctuations[8]. Consequently, the intestinal epithelium is exposed to different bacterial species and their metabolites throughout the day[8]. In turn, the circadian rhythm of the microbiota regulates the transcription of host circadian clock genes and affects epigenetic modifications and oscillations in metabolite levels[8,9]. These results add to the mounting evidence connecting the gut microbiota with the gut-brain axis, where early gut microbial colonization may be linked with neurodevelopmental outcomes with potential long-term effects. Within this axis, gut microbiota affects brain function through various pathways, including the immunoregulatory pathway, neuroendocrine pathway, vagus nerve pathway, and tryptophan metabolism[10]. The term psychobiotic was brought up by Dinan et al.[11], as a novel class of psychotropic medication, defined as a “live organism that, when ingested in adequate amounts, produces a health benefit in patients suffering from psychiatric illness.” Research into the microbiota-gut-brain axis could help us develop potential new therapeutic targets for the effective treatment or prevention of brain disorders. One limitation of this dissertation is that only one fecal sample was included in our analysis, and infant gut microbiota matures rapidly maturated the first year of life. Thus, our results may not fully represent the temporal development of the infant gut microbiome, which may be critical to studying neurodevelopment in early life. A second limitation is that 16S rRNA sequencing can only provide limited strain-level information. Previous studies have demonstrated the biological 120 importance and different metabolic capabilities of specific bacterial strains on human health. A third limitation is that due to the small sample size and lack of power, we are unable to assess the potential role of gut microbiota in the causal pathway linking infant feeding practices and brain development. Future studies using this same cohort may be able to extend our study in four ways. First, we may use shotgun metagenomic sequencing, which provides higher resolution and sensitivity in microbiome analysis. Metagenomics also allows us to access the species-level functional profiling of the gut microbiota, which is important for understanding gene and metabolic pathway content. Second, vitamin D, which is an important factor for gut microbiota, has only been studied in vitamin D supplements for infants in our study. However, previous studies highlight the influence of prenatal vitamin D and vitamin D concentration in breast milk on infant health. We may be able to estimate prenatal vitamin D intake by interview and to measure vitamin D concentration in mother’s breast milk and in archived prenatal serum. Third, we can use Mendelian randomization to make casual inferences for our existing findings. Mendelian randomization is a method of using measured variation in genes of known function to examine the causal effect of a modifiable exposure on disease in observational studies. Fourth, as the sample size of our cohort increases, we may be able to assess the potential role of gut microbiota in the causal pathway linking infant feeding practices and brain development. 121 REFERENCES 122 REFERENCES 1. Widiger, T.A.; Oltmanns, J.R. Neuroticism Is a Fundamental Domain of Personality with Enormous Public Health Implications. World Psychiatry 2017, 16, 144–145, doi:10.1002/wps.20411. 2. Kostyrka-Allchorne, K.; Wass, S.V.; Sonuga-Barke, E.J.S. Research Review: Do Parent Ratings of Infant Negative Emotionality and Self-Regulation Predict Psychopathology in Childhood and Adolescence? A Systematic Review and Meta-Analysis of Prospective Longitudinal Studies. Journal of Child Psychology and Psychiatry 2020, 61, 401–416, doi:10.1111/jcpp.13144. 3. Carter, K.A.; Hathaway, N.E.; Lettieri, C.F. Common Sleep Disorders in Children. AFP 2014, 89, 368–377. 4. Chan, W.S.; Levsen, M.P.; McCrae, C.S. A Meta-Analysis of Associations between Obesity and Insomnia Diagnosis and Symptoms. Sleep Medicine Reviews 2018, 40, 170–182, doi:10.1016/j.smrv.2017.12.004. 5. Vgontzas, A.N.; Liao, D.; Pejovic, S.; Calhoun, S.; Karataraki, M.; Bixler, E.O. Insomnia With Objective Short Sleep Duration Is Associated With Type 2 Diabetes: A Population-Based Study. Diabetes Care 2009, 32, 1980–1985, doi:10.2337/dc09-0284. 6. Basta, M.; Chrousos, G.P.; Vela-Bueno, A.; Vgontzas, A.N. Chronic Insomnia and the Stress System. Sleep Medicine Clinics 2007, 2, 279–291, doi:10.1016/j.jsmc.2007.04.002. 7. Voigt, R.M.; Forsyth, C.B.; Green, S.J.; Engen, P.A.; Keshavarzian, A. Chapter Nine - Circadian Rhythm and the Gut Microbiome. In International Review of Neurobiology; Cryan, J.F., Clarke, G., Eds.; Gut Microbiome and Behavior; Academic Press, 2016; Vol. 131, pp. 193– 205. 8. Thaiss, C.A.; Levy, M.; Korem, T.; Dohnalová, L.; Shapiro, H.; Jaitin, D.A.; David, E.; Winter, D.R.; Gury-BenAri, M.; Tatirovsky, E.; et al. Microbiota Diurnal Rhythmicity Programs Host Transcriptome Oscillations. Cell 2016, 167, 1495-1510.e12, doi:10.1016/j.cell.2016.11.003. 9. Li, Y.; Hao, Y.; Fan, F.; Zhang, B. The Role of Microbiome in Insomnia, Circadian Disturbance and Depression. Frontiers in Psychiatry 2018, 9, 669, doi:10.3389/fpsyt.2018.00669. 10. Kim, Y.-K.; Shin, C. The Microbiota-Gut-Brain Axis in Neuropsychiatric Disorders: Patho- Physiological Mechanisms and Novel Treatments. Curr Neuropharmacol 2018, 16, 559–573, doi:10.2174/1570159X15666170915141036. 11. Dinan, T.G.; Stanton, C.; Cryan, J.F. Psychobiotics: A Novel Class of Psychotropic. Biol Psychiatry 2013, 74, 720–726, doi:10.1016/j.biopsych.2013.05.001. 123 12. Wacker, M.; Holick, M.F. Sunlight and Vitamin D. Dermatoendocrinol 2013, 5, 51–108, doi:10.4161/derm.24494. 13. Wimalawansa, S. Biology of Vitamin D. 2019, 10, 1–8, doi:10.4172/2157-7536.1000198. 14. Adams, J.S.; Hewison, M. Update in Vitamin D. The Journal of Clinical Endocrinology & Metabolism 2010, 95, 471–478, doi:10.1210/jc.2009-1773. 15. Bikle, D. Nonclassic Actions of Vitamin D. The Journal of Clinical Endocrinology & Metabolism 2009, 94, 26–34, doi:10.1210/jc.2008-1454. 16. Yamamoto, E.A.; Jørgensen, T.N. Relationships Between Vitamin D, Gut Microbiome, and Systemic Autoimmunity. Front Immunol 2020, 10, 3141, doi:10.3389/fimmu.2019.03141. 124