EARLY - LIFE EXPOSURES AND THE IR IMPACT ON GUT MICROBIOTA ASSEMBLY: A LONGITUDINAL ANALYSIS OF THE INFANT GUT MICROBIOTA IN MICHIGAN COHORTS By Kameron Sugino A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Human Nutrition - Doctor of Philosophy 2020 ABSTRACT EARLY - LIFE EXPOSURES AND THE IR IMPACT ON GUT MICROBIOTA ASSEMBLY: A LONGITUDINAL ANALYSIS OF THE INFANT GUT MICROBIOTA IN MICHIGAN COHORTS By Kameron Sugino Background: Obesity is a global epidemic and is responsible for increasing burdens on public health. F actors like d iet a nd lifestyle have been investigated as modifiable factors that can help prevent the development of obesity, but another possible mediator of obesity development is the gut microbiome. The gut contains a diverse community of microbes that influences several physiological functions in humans and animals. There have been many studies investigating this association between obesity and the microbiota in adults, however, few studies have been done on the establishment of the infant gut microbiota in the context o f its association with maternal pre - pregnancy body mass index (BMI), breastfeeding and delivery mode. Hypothesis: We expect maternal pre - pregnancy BMI will be associated with altered microbiota composition of women and their children. However, in children , this association will have less of an effect on the gut microbiota than exposure to human milk in the diet. Methods: Fecal samples and participant information were collected from a subset of dyads enrolled in two related prospective cohorts (ARCH GUT and BABY GUT ) in Michigan. Sequencing the V4 region of the 16S gene was used to analyze fecal bacterial samples collected from mothers in their third trimester and infants at 1, 6, 12 and 24 months of age . The microbiota data was analyzed using alpha and beta diversity metrics, negative binomial regression to compare taxa abundances between groups and LonGP, a microbiota analysis tool for longitudinal data sets . Results: We found that fecal bacterial commun ities from overweight women had lower microbiota diversity than communities from normal weight or obese women and beta diversity of overweight women differed from those of normal and obese women at the genus and phylum levels. Infant alpha diversity at 1 m onth of age differed in membership (Sorensen index) by maternal pre - pregnancy BMI category and also differed by delivery mode and breastfeeding exclusivity. At 6 months of age, fewer infants born to women with pre - pregnancy obesity were breastfeeding compa red to infants born to non - obese women (35.7% and 81.8%, respectively). Maternal pre - pregnancy BMI and human milk exposure were both significantly associated with alpha and beta diversity of the infant microbiota in univariate analys e s . However, in multivariate analyses, human milk exposure accounted for 20% of variation in alpha diversity, but pre - pregnancy BMI was not significantly associated with microbiota diversity. The infant diet at six months was the major determinant of alpha a nd beta diversity of the infant. In our longitudinal analysis of the infant fecal microbiota from 1 month to 2 years of age, w e found that age and participant explained most of the changes in abundance in our dataset. I n the top 10 most abundant taxa, hum an milk exposure and antibiotic exposure at the time of sampling were the only variables important besides age and ID, specifically for Lachnospiraceae unclassified and Bacteroides abundances. Lachnospiraceae abundance was much higher as infants aged, in i nfants receiving <50% human milk in their diet , and if their mom had a pre - pregnancy BMI of > 3 0 . Conclusion: Here, we found several associations between the infant gut microbiota and infant exposures during the first two years of life. Exploring the relationship between gut health and early infant exposures could help develop potential preventative treatments for many gut - linked chronic diseases such as obesity. iv ACKNOWLEDGEMENTS I would like to start by thanking my mentor Dr. Sara h Comstock , for her amazing guidance and support through out my graduate studies. At times, I have refused to listen to your advice and knowledge (that thing about the flowering plants comes to mind) and failed to invite you to my impromptu marriage - documen t - signing ( I honestly figured you had better things to do) . Despite these things, you have been a constant source of encouragement both professionally and personally I promise to invite you to the real wedding in the future. Thank you to my committee members Jim Pestka, Won Song and Jim Tiejde for all the advice they gave and difficult questions they asked to help me improve my research and think more like a scientist. I thank our collaborators and the many individuals involved wit h the (M)ARCH project, without which this research would not have been possible. Specifically, Dr. Nigel Paneth, Dr. Jean Kerver and Tengfei Ma, who provided me with valuable insight and guidance. Thanks to my fellow graduate students for making graduate school so much more than just classes and research, specifically Amanda Feighner, Alyssa Beavers, Karl Se i wert , Ankita Bhattacharya, Lauren Gable, Ran Tao and Natasha Sloniker for all the shenanigans we got up to and the odd foods cons umed. And finally, I would like to thank my friend s , family , and wife for their support. To my friends , soon, I promise). To my family, who always believed I could finish my doctorate and have always supported my endeavors, including moving two thousand miles away (or further) to v Michigan . And to my wife, who has been immensely patient and understanding through this whole PhD, even as I am ignoring her to finish this disse rtation. Thank you all. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ..................... viii LIST OF FIGURES ................................ ................................ ................................ ...................... x KEY TO ABBREVIATIONS ................................ ................................ ................................ ..... xi CHAPTER 1: INTRODUCTION, LITERATURE REVIEW AND RESEARCH AIMS ..... 1 Introduction ................................ ................................ ................................ ............................... 1 Literature Review ................................ ................................ ................................ ...................... 2 Maternal obesity and infant health ................................ ................................ .......................... 2 Obesity and the gut microbiota ................................ ................................ ................................ 3 Preg nancy and the gut microbiota ................................ ................................ ........................... 4 Gut microbiota assembly: potential sources of infant bacterial exposure ............................... 5 The infant gut microbiota ................................ ................................ ................................ ........ 7 The infant gut microbiota and delivery mode ................................ ................................ ......... 8 The infant gut microbiota and human milk exposure ................................ .............................. 9 Infant gut microbiota and antibiotic exposure ................................ ................................ ....... 11 Infant gut microbiota and maternal pre - pregnancy BMI ................................ ....................... 12 The infant gut microbiota and future health ................................ ................................ .......... 13 Gaps in the literature ................................ ................................ ................................ .............. 14 Significance ................................ ................................ ................................ .............................. 14 Specific Aims ................................ ................................ ................................ ............................ 15 CHAPTER 2: METHODS OVERVIEW ................................ ................................ ................. 17 Study participants ................................ ................................ ................................ ................... 17 Sample Collection, DNA Extraction and Amplification ................................ ...................... 18 Processing and Analysis of Sequence Data ................................ ................................ ........... 20 Data Analysis ................................ ................................ ................................ ........................... 20 CHAPTER 3: MICHIGAN COHORTS TO DETERMINE ASSOCIATIONS OF MATERNAL PRE - PREGNANCY BODY MASS INDEX WITH PREGNANCY AND INFANT GASTROINTESTINAL MICROBIAL COMMUNITIES : LATE PREGNANCY AND EARLY INFANCY ................................ ................................ ................................ ........... 22 Abstract ................................ ................................ ................................ ................................ .... 22 Introduction ................................ ................................ ................................ ............................. 23 Materials and Methods ................................ ................................ ................................ ........... 26 Results ................................ ................................ ................................ ................................ ...... 27 Alpha diversity ................................ ................................ ................................ ...................... 30 Beta diversity ................................ ................................ ................................ ......................... 32 Sensitivity analys is of infant microbiota data ................................ ................................ ........ 39 Differential patterns of bacterial taxa ................................ ................................ .................... 43 vii Discussion ................................ ................................ ................................ ................................ . 48 Conclusion ................................ ................................ ................................ ................................ 53 CHAPTER 4: HUMAN MILK FEEDING PATTERNS AT SIX MONTHS OF AGE ARE THE MAJOR DETERMINANTS O F FECAL BACTERIAL DIVERSITY IN INFANTS 54 Abstract ................................ ................................ ................................ ................................ .... 54 Introduction ................................ ................................ ................................ ............................. 55 Materials and Methods ................................ ................................ ................................ ........... 56 Results ................................ ................................ ................................ ................................ ...... 57 Subject Characteristics ................................ ................................ ................................ .......... 57 Alpha Diversity ................................ ................................ ................................ ...................... 60 Beta Diversity ................................ ................................ ................................ ........................ 64 Individual Taxa ................................ ................................ ................................ ...................... 67 Discussion ................................ ................................ ................................ ................................ . 69 Limitations ................................ ................................ ................................ ............................... 72 Conclusion ................................ ................................ ................................ ................................ 72 CHAPTER 5: EFFECT OF INFANT EXPOSURES ON THE GUT MICROBIOTA FROM EARLY INFANCY TO 2 YEARS OF AGE ................................ ............................... 74 Abstract ................................ ................................ ................................ ................................ .... 74 Introduction ................................ ................................ ................................ ............................. 75 Materials and Method s ................................ ................................ ................................ ........... 76 Results ................................ ................................ ................................ ................................ ...... 78 Discussion ................................ ................................ ................................ ................................ . 93 SUMMARY DISCUSSION ................................ ................................ ................................ ........ 97 Implications and Future Directions ................................ ................................ ........................... 99 APPENDICES ................................ ................................ ................................ ........................... 102 APPENDIX A: CONSENT FORMS, SAMPLE COLLECTION INFORMATION AND QUESTIONNAIRES ................................ ................................ ................................ .............. 103 APPENDIX B: IRB APPROVAL ................................ ................................ ........................... 144 BIBLIOGRAPHY ................................ ................................ ................................ ..................... 145 viii LIST OF TABLES Table 1 . Participant Characteristics ................................ ................................ ......................... 28 Table 2 . Cohort Characteristics ................................ ................................ ................................ . 29 Ta ble 3 . Alpha Diversity of the Fecal Microbiota of Mothers and Infants by Maternal Pre - pregnancy BMI Category ................................ ................................ ................................ ........... 30 Table 4 . Infant alpha diversity by infant age, sex, mode of delivery and sample shipping time ................................ ................................ ................................ ................................ ............... 31 Table 5. P - values of infant alpha diversity by delivery mode and breastfeeding stratified by BMI category ................................ ................................ ................................ ............................... 32 Table 6. Sensitivity of infant fecal microbiota alpha diversity to infant age -- Days of Age ................................ ................................ ................................ ................................ .. 40 Table 7. Sensitivity of infant fecal microbiota alpha di versity to infant age -- Days of Age ................................ ................................ ................................ ................................ .. 41 Table 8 . Significantly Different Taxa in the Fecal Microbiota by Maternal Pre - pregnancy BMI Category ................................ ................................ ................................ .............................. 44 Table 9 . Significantly Different Taxa in Fecal Microbiota by Delivery Mode and Breastfeeding ................................ ................................ ................................ ............................... 46 Table 10. Significantly different taxa by delivery mode and breastfeeding in the gut microbiota of infants born to women who were overweight prior to becoming pregnant ... 47 Table 11. Significantly different taxa by delivery mode and breastfeeding in the gut microbiota of infants born to women who were obese prior to becoming pregnant ............ 48 Table 12 . Participant characteristics of the 6 - month - old infants by maternal pre - pregnancy BMI category ................................ ................................ ................................ ............................... 58 Table 13 . Participant characteristics of the 6 - month - old infants by cohort .......................... 59 Table 14 . Alpha diversity of the fecal bacterial community of 6 - month - old infants by maternal pre - pregnancy BMI category or human milk exposure ................................ ......... 61 Table 15 . Multivariate analysis of the alpha diversity of the 6 - month - old infant microbiota ................................ ................................ ................................ ................................ ....................... 62 ix Table 16 . Beta values of the 6 - month - old infant microbiota alpha diversity by human milk exposure levels in each model ................................ ................................ ................................ .... 63 Table 17 . Multivariate PERMANOVA analysis of the individual variable marginal effects in the beta diversity of the 6 - month - old inf ant microbiota ................................ ..................... 66 Table 18 . Genera/phyla abundances in the 6 - month - old infant microbiota classified by human milk exposure ................................ ................................ ................................ ................. 68 Table 19. Population characteristics of the infants at 1, 6, 12 and 24 months ...................... 79 Table 20. Population characteristics of infants with all four samples ................................ ... 81 Table 21 Model parameters significantly associated with changes in abundance of the top 50 most abundant taxa and the variance explained by each parameter ................................ 82 Table 22. The mean relative abundance predictions by each variable for the 10 most abundant taxa ................................ ................................ ................................ .............................. 88 x LIST OF FIGURES Figure 1. The fecal bacterial community structure of pregnant women differs significantly from that of infants ................................ ................................ ................................ ..................... 33 Figure 2. Pre - pregnancy overweight women have different fecal microbiota compositions than normal and obese women at phyl um/genus levels ................................ ........................... 34 Figure 3. Bacterial community membership does not differ in pregnant women by pre - pregnancy BMI category ................................ ................................ ................................ ............ 35 Figure 4. Relationship between infant fecal bacterial membership and maternal/infant variables at the genus level ................................ ................................ ................................ ......... 36 Figure 5. Relationship between infant fecal bacterial structure at the genus level and maternal/infant variables ................................ ................................ ................................ ........... 37 Figure 6. Infant fecal microbiota stratified by maternal overweight and obese categories . 38 Figure 7. Sensitivity analysis of the infant microbiota, including infants from 2 to 9 days old ................................ ................................ ................................ ................................ ....................... 42 Figure 8. Sensitivity analysis of the infant microbiota, including infants from 2 to 18 days old ................................ ................................ ................................ ................................ ................. 43 Figure 9. Maternal pre - pregnancy BMI and level of human milk in the diet are associated with differences in the infant microbiota. ................................ ................................ ................. 65 Figure 10. Predicted abundance and actual abundance over time for the top 10 most abundant taxa ................................ ................................ ................................ .............................. 89 Figure 11. Predicted change in abundance as the infant ages for the 40 least abundant taxa ................................ ................................ ................................ ................................ ....................... 90 Figure 12. Relative abundance of Lachnospiraceae unclassified predicted by current antibiotic exposure, human milk exposure and maternal pre - pregnancy BMI .................... 92 Figure 13. Relative abundance of Bacteroides predicted by human milk exposure ............. 93 xi KEY TO ABBREVIATIONS Alpha diversity - The bacterial diversity within a single community. Here we use Chao1 (presence/absence), Shannon (bacterial abundances, even weight for all bacteria) and inverse Simpson (bacterial abundances, weighted towards more abundant bacterial) ARCH - Archive for Research in Child Health ARCH GUT - Our study, which uses a subset of ARCH participants and collects fecal samples from the women and their children BABY GUT - A study designed to add participants to the ARCH GUT study cohort Beta diversity - The bacterial community similarity between two communities. Here, we use Sorensen (presence/absence) and Bray - Curtis (bacterial abundances) BMI - Body mass index calculated as C - section - Cesarean section deliveries. Here we do not differentiate between planned and emergency procedures Microbiome - The genetic material of all microbes in a community Microbiota - The microbes present i n a community; in this document, "microbiota" will refer to only the bacteria PCOA - Principal Coordinate Analysis, a type of ordination used to visualize multivariate data PERMANOVA - Permutational Analysis of Variance, a non - parametric multivariate sta tistical test used to test for differences between groups based on multivariate similarities/differences between said groups PERMDISP - Permutational Dispersion, a test for homogeneity of multivariate dispersions; is used as a companion test to PERMANOVA to test whether differences between groups found in a PERMANOVA test can be attributed to differences in the variance between groups SCFA Short chain fatty acids T1 and T3 - Trimesters 1 and 3 of pregnancy, respectively 1 CHAPTER 1 : INTRODUCTION, LITERATURE REVIEW AND RESEARCH AIMS Introduction Obesity is an epidemic which affects 38.1% of men and 41.2% of women in the US, with over two - thirds of adults categorized as either overweight or obese in 2016 (75.0% and 67.8% for men and women, respectively) (1) . Obesity is responsible for globally increasing burdens on the healthcare system due to higher rates of mo rbidities such as cardiovascular disease, type II diabetes and certain types of cancer (2) . High maternal BMI negatively affects maternal health and is linked to adverse pregnancy outcomes such as C - section delivery (3 5) as well as issues breastfeeding post - partum (6 8) . Each of these three closely related factors have been individually associated with a higher risk of childhood overweight/ obesity (3,5,8 10) . Obesity during childhood is likely to carry over into adulthood (11) , thus maternal obesity increases the generations. The BMI - heritability of obesity is estimated to be up to 80%, but only around 3% has been explained by genetic contributions (12) . Diet, lifestyle and environment are all known to contribute to obesity risk (12) , but another possible contributing factor is the gut microbiome. The gut contains a diverse community of microbes that influences several physiological functions in humans and animals (13,14) . It plays a role in immune system development, synthesizes vitamins that the host can utilize and could possibly have effects on cognitive development (15,16) . Which specific bacteria can affect weight gain and obesity is currently unknown in humans, but there is ample evidence that the microbiome does modulate weight gain in animal models. In germ - free (GF) mouse models, the absence of bacteria in the gut is 2 associated with significantly decreased weight gain (13,15) , but transplantation of the microbiota from wil d type mice into GF mice significantly increases the weight gain of the GF mice to levels similar to that of the wild type mice (13) . In another study using humanized mice (17) , fecal samples from human twin pairs were collected, where one twin was obese and one normal weight. The GF mice were then seeded with bacteria using either the obese or normal weight and fat mass independent of food intake (17) , demonstrating that the gut microbiome is capable of increasing weight gain and may contribute to the development of obesity. There have been many studies investigating this association between obesity and the microbiota in humans (13,14,18 20) , however, few studies have assessed how the infant gut microbiota assem bles in the context of maternal pre - pregnancy BMI, breastfeeding, delivery mode and other related exposures (21,22) . This research will use mother - infant dyads to determine how the maternal microbiota during pregnancy is affected by pre - pregnancy BMI and what environmental influences affect infant gut microbiot a development over the first 2 years post - partum. Literature Review Maternal obesity and infant health Among women giving birth in the US in 2014, over half were categorized as overweight or obese (25.6% overweight and 24.8% obese) (23) . Maternal obesity increases the risk of preeclampsia, gestational hypertension, gestational diabete s mellitus (24,25) and is especially concerning due to the implications for infant health and development. For example, mothers obese du ring the 1st trimester had children who were more than twice as likely to be obese when 3 measured at 2, 3 and 4 years of age (10) and other studies report maternal pre - pregnancy obesity also increas weight mothers (9) . Another maternal obesity - related risk factor for childhood obesity is C - section delivery. Women with obesity have 2 - 4 - fold increased odds of delivering via C - section compared to no rmal weight women (3 5) old through 15 years of age and is associated with almost twice the risk of developing overweight or obesity by 11 years of age (8) . Infant exposure to breast milk may reduce the risk of childhood overweight/obesity in children from obese women and those born via C - section (9) . Infants exposed to breast milk are generally at a lower risk of developing overweight/obesity regardless of maternal BMI (26) . However, obese mothers often have difficulties breastfeeding due to insufficie nt milk yield in the first few weeks of lactation (7) and are more likely to give their infant formula rather than breastmilk (6,7) . Obesity and the gut microbiota At around 2 - 3 years of age, the human gut microbiome stabilizes and the microbiota that dominate the gu t belong to the phyla Firmicutes or Bacteroidetes. After the gut matures to this state, the overall community structure remains unique and distinguishable from other communities (27) . Changes that occur after this point are generally within the relative abundances of extan t bacteria in the community, not shifts in composition due to colonization of new bacterial taxa (28) . One change that has previously been reported is an increased ratio of the phyla Firmicutes and Bac teroidetes in obese individuals (13,14) , though this association is not consistent between different studies (18,19) . Others have found many members of the Firmicutes 4 phylum are negatively associated with vi sceral fat mass, while only a few genera such as Blautia , Ruminococcus , Clostridiales and Acidaminococcus are positively associated (20) . The same authors found that the gut microbiota community explained more variance in participant visceral fat mass than did dietary nutrient estimates (R 2 =0.56 and R 2 =0.19, respectively) ca lculated from a food frequency questionnaire , following the EPIC - Norfolk guidelines, which is designed to estimate food intake over the past year (29) . Pregnanc y and the gut microbiota hormones, metabolism, increases in markers of inflammation as well as depressed immune function (30) . These physiological shifts create an altered environment for the microbes in the m trimester 1 (T1) to trimester 3 (T3) (31) . This shift was characterized by lower alpha and higher beta diversity and higher abundances of the phyla Proteobacteria and Actinobacteria (31) . When the fecal communities from T1 and T3 were transplanted into mice, the mice who received the T3 microbiota had higher levels of inflammatory markers such as lipocalin, gained mo re adiposity, and had reduced oral glucose tolerance compared to the mice who received the T1 microbiota (31) . Others observed that pregnancy - related hormone levels such as insulin were positively associated with increases in Collinsella abundance (32) , which has previously been asso ciated with type 2 diabetes (33) . These results suggest that the gut microbiota is associated with changes in metabolism that develop in the late stages of pregnancy (30) , and it is hypothesized that these change s to the gut microbiota are important for a healthy pregnancy (30) . However, 5 another study of 49 women who provided weekly fecal samples during gestation found no changes in the pregnancy gut microbiota (34) . Gut microbiota assembly: potential sources of infant bacterial exposure There are many early - life exposures that can provide colonizing bacteria in the infant gut microbiota. Possible human sources of bacteria include the vagina, s kin, oral, breast milk and the controversial placental microbiota, all of which have their own unique microbial communities. The vaginal microbiota is generally dominated by Lactobacillus crispatus, L. jensenii, L. gasseri and L. iners (35) . Like the gut microbiota, the vaginal microbiota has been found to have decreased overall diversity and richness related to increased abundances of Lactobacillus species during pregnancy (18 - 40 weeks gestation) compared to non - pregnant women (36) . Thi s increase in Lactobacillus in the vagina during pregnancy may help lower the pH (37) and prevent vaginosis by inhibiting bacteria such as Gardnerella (37 39) . G. vaginalis has also been associated with incr eased risk of preterm birth (34) and other adverse pregnancy outcomes (35) , suggesting that these changes in the vaginal microbiota during pregnancy are important for a healthy pregnancy. In comparison to other human microbiota environments such as the gut, our skin is relatively nutrient poor. As a result, the bacteria that reside on our skin are those that can utilize and tolerate sweat, lipids from secreted oils (sebum) and dead skin cel ls (40) . T he most abundant bacteria colonizing the skin be long to the genera Staphylococcus, Corynebacterium and Propionibacterium (41) . These genera can be found on all skin sites, but the halotolerant Staphylococcus and Corynebacterium tend to prefer humid environments such as the armpit, while oily sites such as the face are dominated by lipophilic Propionibacterium (40) . 6 Like the skin, the oral cavity hosts a wide range of bacterial species that have been reported to be site - dependent the communities on the tongue, cheek, teeth, saliva and other parts of the mouth differ from each other despite the close proximity and exch ange of microbes between these areas (42) . However, the genera found most commonly in a healthy oral cavity include Streptococcus, Bifidobacterium, Eubacterium, Lactobacillus, Propionibacterium and Veillonella among others (43) . Breast milk is thought to provide the ideal infant nutrition as well as important colonizing bacteria. The core milk microbiota is thought to be mainly composed of Staphylococcus and Streptococcus (44,45) as well as lower abundances of L actobacill us and Bifidobacterium (44) . There are many similarities between the infant oral, skin and milk microbiota, which is thought to be a result of reflux of bacteria into the breast during breastfeeding (46) . A recent study by Williams et al. (47) reported a high correlation between the milk microbiota and infant oral, infant fecal and maternal oral microbiota communities. The authors estimate that the milk microbiota contributes approximately 5% to the infant fecal microbiota at 2 days of age, but only 0.3% at 6 months (4 7) . Other bacteria that are commonly found in the gut such as Faecalibacterium and Bacteroides have also been reported in milk (48,49) and are theo rized to be a possible source of gut - colonizing bacteria. It is unknown how these anaerobic bacteria appear in the milk and whether they are viable colonizers, but it has been hypothesized that bacteria from the intestines may be translocated to the breast by dendritic cells (46) . Besides the probiotic microbiota that may help seed the infant gut, human milk also contains prebiotic human milk oligosaccharides that help stimulate proliferation of Bifidobacterium and Lactobacillus in the infant gut (50) , both of which are considered im portant components of a healthy infant gut community (46) . 7 The placenta was considered a sterile organ until the - based study found placental bacteria related to unhealthy pregnancy outcomes such as chorioamnionitis (51) . Since then, culture - independent studies have expanded on the microbes thought to compose the pla cental microbiota. These studies found that the placental microbiota is characterized by a low biomass with relatively higher abundances of Lactobacillus, Propionibacterium and members of the family Enterococcaceae (52) . However, other studies that compare qPCR results of placental samples to that of negative controls found that the two runs are indistinguishable from each other (53,54) . A recent study by de Goffau et al. suggests that the placenta does not have a microbiota community during a healthy pregnancy but may contain potentially pathogenic bacteria such as group B Streptococcus - causing S. agalactiae (55) . Because of these conflicting results between studies, the existence of the placental microbiota is controversial and further studies need to confirm the presence of a placental microbiota. The infant gut microbiota At birth, the infant gastrointestinal tract is minimally colonized and undergoes a period of instability and high diversity before maturing around two to three years of age (31,56 59) . - associated microbial communities are thought to provide most of the microbes that initially populate the infa (58,60) . However, (60,61) , antibiotic exposure (30) (31,61) are also important sources of bacteria as gut maturation progres ses. These early - life 8 affecting nutrition/metabolism (62,63) , allergy risk (62,64) and possibly the development of obesity (21,22) . The infant gut microbiota and delivery mode One of the first environmental influences on microbiome development is mode of delivery, which can shift both the abundance and type of bacteria found in the infa nt gut. vaginal community (enriched in Lactobacillus, Prevotella, Escherichia - Shigella and other bacterial communities found in the vagina) (60) , while C - section infants are most similar to bacteria found in the environment ( Enterobacter, Haemophilus, Staphylococcus, Streptococ cus, Veillonella and C. difficile ) (15,60,65) . Some studies have found that vaginal birth allows the infant to be inoculated by a mixture of aerobic and anaerobic bacteria, which may be an important in step towards creating an anaerobic gut environment for the colonization and development a healthy g ut microbiome (62,66) . As the infant ages, vaginally - born infants have been found to have higher abundances of Bacteroides (67 69) and Bifidobacterium (65,67) , while Enterococcus and Clostridium (67) as well as Staphylococcus, Rothia and Propionibacterium (68) have been reported to dominate the gut of C - section infants. Bacteria such as Staphylococcus and Propionibacterium likely come from t (70) . A study by Tun et al. found that differences in the infant gut associated with delivery mode may be dependent on maternal BMI (21) . At around 3 - 6 months, vaginally delivered infants continue to have significantly greater abundances of Bacteroides (15,69,71,72) and Bifidobacterium (67,73) , but lower levels of Enterococcus (67,72) , Klebsiella and Clostridium compared to C - section infants (67) . Others 9 have reported no association between delivery mode and Bifidobacterium abundance at 6 months but d id find that C - section delivery was associated with delayed colonization of Bifidobacterium (74,75) . After around 6 months, the differences in the gut microbiota between delivery modes generally decrease and the gut microbiota become much more similar between groups (15,72) . The infant gut microbiota and human milk exposure The American Academy of Pediatrics recommends that infants exclusively breastfeed for gastrointestinal infections, atopy, obesity and many chronic diseases (76) . Microbes that can utilize human milk oligosaccharides (HMOs) in breast milk as their main source of energy thrive in the infant gut. Exposure to breastmilk is o ne of the most important early - life contributors shaping the development of the infant gut microbiota and can act as a source of common gut bacteria such as Bifidobacterium and Lactobacillus (15,70,77 80) . Pannaraj et al. (81) , found that bacterial colonization in the gut by breastmilk bacteria may act in a dose - dependent manner. During the first month after birth, i nfants whose intake of breastmilk from the breast was >75%, less than that receive only about 17% of their bacteria from milk over time, this difference in bact eria acquisition decreases as the infants are exposed to other sources of bacteria (81) . At 6 weeks, weighted beta diversity measures of the gut microbiota have found that exclusively breastfed infants have distinct gut c ommunities compared to both mixed fed and exclusive formula fed infants, while mixed feeding and exclusive formula feeding were not significantly different (68) . Others looking at differences between exclusive and non - exclusive breastfeeding found that exclusive breastfeeding has been associated with lower bacterial diversity (82) , 10 increased Bifidobacterium abundance and decreased abundance of Lachnospiraceae (82,83) compared to partial or no breastmilk. Infants who are not exclusively breastfed have higher abundances of Bacteroides and Megasphaera and tend to have a microbiota profile with a predicted microbiot a age that is older than their actual age (84) . In infants younger than 6 months of age, non - exclusively breastfed infants have much higher gut diversity and more mature gut communities compared to infants that exclusively receive breas t milk in their diet, which is associated with increases in the relative abundances of Bacteroides, Eubacterium, Veillonella, and Megasphaera (84) . The same study found that early weaning can also affect the gut microbiota long - term. In infants who were weaned at or before 2 months, they showed lower abundances of Bifidobacteriaceae, Enterococcaceae and higher levels of Lactobacillacae, Coriobacteriaceae, Prevotellaceae, Clostridiaceae, Erysipelotrichaceae and Lachnospiraceae in their gut measured at 6 months and up to 2 years of age (84) . The large - scale TEDDY study reported that during 3 - 14 months of age, any breastfeeding has the largest impact on shaping the gut microbiome until infants s top receiving any milk (72) , which matches results observed from previous research (15) . Throughout this time, the gut is enriched in Bifidobacterium, Lactobacillus and Staphylococcus , but depleted in Escherichia, Eggerthella, Ruminococcus, and Roseburia (72) . Increases in these bacterial genera could be due more to exposure to formula than a lack of exposure to breastmilk, as the intestinal bacterial communities of infants fed any formula tend to be enriched in Bacteroides, E scherichia, Enterobacteriaceae, Clostridium and other bacteria compared to breastfed infants at 4 months of age (15,79) . HMOs in breast milk have several important roles in infant microbiota development by selectively enriching bacteria such as Bifidobacterium (15,72) and improving infant health by supporting epithelial barrier fun ction (85) . Work has been done to find appropriate non - 11 digestible carbohydra te supplements for formula that can mimic the health benefits of human milk. Compounds such as galacto - oligosaccharides, fructo - oligosaccharides and pectin - oligosaccharides have been investigated and beneficially affect Bifidobacterium abundance, epithelia l barrier function and markers of immune function (85) . While these results are promising, more work needs to be done to fully characterize the components in human milk which benefit infant health that are missing in formula. Infant gut microbiota and antibiotic exposure Antibiotic exposure is another factor that contributes to obesity risk and causes disruptions in the developing microbiota. Exposure to antibiotics before 6 months of age has been associated with significantly increased BMI at 2 - 3 years of age (86,87) . In terms of the microbiota, when antibiotics are introduced, the bacterial loa d in the gut decreases creating space for new bacteria to grow or existing taxa to proliferate. This disruption during infancy may have effects on later life health and risk for obesity by altering the microbiome when it is still at a developmental point, allowing for pathogenic bacteria to establish. For example, oral administration of amoxicillin in the first month of life has been shown to cause decreased abundances of Bifidobacterium , which are commonly found in healthy breastfed infants (65) . The routine use of antibiotics for intrapartum antibiotic prophylaxis (IA P) as a preventative measure against Group B Staphylococcus and for C - section deliveries ha s effects on the microbiota and can further compound other factors that alter the microbiome and affect obesity risk such as C - section delivery and maternal obesity. In a study by Azad et al. (88) , the gastrointestinal bacterial communities of IAP infants (for both vaginal and C - section delivery) showed significantly lower Bacteroidetes abundance compared to those from non - IAP infants. 12 They also found that lower amounts of Bacteroidetes and increased abundance of Firmicutes persisted at 1 year in the gastrointestinal bacteri al communities of IAP emergency C - section infants that were not breastfed exclusively at 3 months. Exposure to antibiotics has also been found to have long - lasting effects on the infant gut composition. A course of amoxicillin in children has been shown to decrease gut diversity for several days after treatment ends, but diversity increases dramatically 1 - 6 months after treatment ends compared to children who were never exposed to antibiotics (89) . Infant gut microbiota and maternal pre - pregnancy BMI High maternal pre - pregnancy BMI has been associated with both an increased risk of developing childhood overweight/obesity and an altered infant gut microbiota even after accounting for breastmilk in the infant diet (21,22) . Alterations to the gut during infancy may explain how maternal BMI and infant milk cons obesity. Additionally, there is some evidence of microbial transfer from mother to child (90,91) , which may act as a direct link between the increased risk of childhood obesity in infants born to obese women and the obesity - related gut microbiota. Though duration and level of human milk in the infant diet is known to be negatively associated with maternal obesity (92) , few studies have looked at the relationship between these two variables in the context of the developing infant microbiota (21,22,93) . Of these studies, one found that in 3 - 4 - month old infants, those born to overweight/obese women have higher richness and significantly different community membership (unweighted UNIFRAC) at 3 - 4 month s compared to infants born to normal weight women (21) . Another found that in 3 - 6 - month old infants, maternal BMI was not significant after adjusting for exclusive breastfeeding, delivery mode, age, race and other factors (21) . Finally, a 13 longitudinal study of infants using data collected at 4, 10, 30, 120, 365 and 730 days old found no difference in alpha diversity by maternal BMI at any timepoint (22) . The infant gut microbiota and futur e health There are many known relationships between bacteria and increased risk of infancy disease s . S epsis in infants is generally caused by the infiltration of Staphylococcus , while other diseases have more complex etiologies like necrotizing enterocolitis , which is associated with increases in Enterobacteriaceae , Clostridium perfringens as well as decreases in protective Bifidobacterium (94) . In recent years, many other adverse health effects have been investigated as potentially associated with the microbiota community and acquisition during infancy. It is though t that interactions between gut bacteria and the host epithelium are important for immune system programming and development (94) . An example of this i s atopic eczema, which has been associated with increased abundances of C lostridia and decreased B ifidobacteria in the gut microbiota of children who would later develop atopy (95,96) . Other chronic disorders that have been associated with infant dysbiosis include irritable bowel syndrome, type 1 diabetes and metabolic disorders (94) . Impaired growth has also been associated with the gut microbiota. In a study on malnutrition in Malawian children , the microbiota from malnourished and healthy children was transferred into GF mice . T he mice with the gut microbes from malnour ished children gained significantly less weight than the mice with the healthy microbiota despite no difference in food consumption (97) . From these studies, it is clear that microbiota acquisition and development during infancy is an important area of study because of the short and long - term health implications of dysbiosis at this early life stage. 14 Gaps in the literature The human microbiome has been extensively research ed in the past few decades, revealing many associations between microbiota communities and disease outcomes (9,16,96,98,99) . S tudies that have investigate d the gut microbiota in infancy have found long - term health outcomes associated with the gut microbiota of children (94,96,97) . High pre - pregnancy BMI is associated with higher risk of C - section delivery and lower breastfeeding rates (3,7,9,84) , however few studies on the infant microbiota have investigated how these factors interact with the infant gut micr obiota (21,22) . Therefore, it is possible t he infancy mi crobiota may act as a mediator between high maternal pre - pregnancy BMI, C - section delivery and low rates of breastfeeding and higher risk of childhood obesity, atopy and other disease states. This research aims to understand how the infant gut microbiota i s associated with these exposures while the gut community is still maturing . This research may reveal how exposures such as maternal pre - pregnancy BMI, delivery mode and breastfeeding are associated with gut microbiota assembly and at what ages these factors have the strongest association with the gut microbiota. Significance Few studies have investigated the role of pre - pregnancy BMI on the 3rd trimester microbiome or maternal pre - pregnancy infant gut microbiota infant exposures over time is poorly understood, but there is evidence to suggest the possibility of an obesogenic microbiome whose inheritanc e would increase the risk of developing obesity. This research will describe how the infant gut microbiota matures and which exposures help modulate these changes to the gut microbiota . Understanding wh ich influences the infant gut 15 microbiome and how those changes affect health will lead to novel solutions that help prevent chronic diseases like obesity. Specific Aims Hypothesis 1A: The microbiota of overweight and obese mothers will be characterized by lo wer alpha diversity and unique community composition (beta diversity) compared to normal weight mothers. Aim 1A : Determine if maternal pre - pregnancy BMI is associated with gut bacterial communities in pregnant women in their third trimester. Hypothesis 1B: The microbiota of infants born to overweight or obese mothers will be characterized by distinct patterns of community membership of the bacteria compared to infants born to normal weight mothers . Delivery mode and milk in the diet will also affect gut membership at this timepoint. Aim 1B : Determine the effect of maternal pre - pregnancy BMI on gut bacterial communities in infants within the first 4 months post - partum. Compare the microbiomes of infants ( around 1 month of age) based on maternal pre - pregnancy BMI and how other covariates such as delivery mode and breastfeeding are associated with the gut microbiota composition in the early life. Hypothesis 2: Breastfeeding will have the strongest association with gut microbiota community composit ion . D elivery mode and maternal pre - pregnancy BMI may also contribute to gut bacterial composition . Aim 2 : Determine whether there are differences in the gut microbiota of 6 - month - old infants based on maternal pre - pregnancy BMI, delivery mode and breastfee ding and what factors are most important in explaining the gut community composition at this timepoint. 16 Hypothesis 3A: G ut community composition over time will have the strongest association with infant age. Aim 3 A : Determine how the gut assembles over the first two years of life using gut microbiota data collected around 1 month of age , 6 months, 12 months and 24 months. Hypothesis 3B: Early life exposures such as delivery mode will have significant effects on the gut microbiota in the early life but will be less important once the infants reach 6 months of age. Breastfeeding rate and duration will have the overall strongest impact on shaping the infant gut microbiome until 12 months (when the infants are consuming solid food for most of their diet). M aternal pre - pregnancy BMI may have small effects on the gut microbiota assembly at all timepoints after adjusting for breastmilk exposure. Aim 3B: Determine which early life exposures can be used to predict microbio ta abundances . 17 CHAPTER 2 : METHODS OVERVIEW Study participants Since many of the methods used for the research presented in this dissertation are the same in each chapter, we have summarized the study design, sample collection, sample processing and analysis sections here. More details on study methods can be found in the methods section of each chapter, respectively. The participants for this study were enrolled as part of the Archive for Research in Child Health ( ARCH ) GUT or BABY GUT cohorts. ARCH GUT participants were enrolled as a sub - study of the ARCH study . ARCH is a pregnancy cohort designed to be low - cost and low participant burden participants, where participants provide written informed consent to obtain medical records, birth certificates, new born blood spots, brief interviews and clinically obtained blood and urine samples . ARCH GUT enrolled participants during 2015 - 2017 were from one clinic in Lansing and one in Traverse City, MI. Participants provided written informed consent to obtain an enr ollment questionnaire (pre - pregnancy height and weight, antibiotic use in the past year, parity, diagnosed or suspected food allergies/intolerances) and fecal samples from the women during their third trimester of pregnancy and fecal samples from their inf ant at 1 month , 6 months, 12 months and 24 months of age. A questionnaire was completed and submitted with each fecal sample. Copies of the enrollment forms and questionnaires can be found in Appendix A . past month, whether they were taking any prebiotics or probiotics, their water source (well, city or bottled), current weight and were asked to list the foods they h ad eaten in the past 24 hours before sampling. The questionnaires for the infant samples asked the moms about infant antibiotic use in the past month, how much human milk the infant consumed in the week before 18 fecal sample collection, whether the infant re ceived formula, a list of the foods the infant had eaten in the past 24 hours before sampling, whether the infant was taking prebiotics or probiotics, delivery mode, infant sex and weight/length at the sampling time point. BABY GUT was designed to supplemen t the ARCH GUT cohort. Enrollment and sampling protocols in BABY GUT were the same as those used for ARCH GUT , but the recruitment area was restricted to clinics within the G reater Lansing area, Michigan. The Michigan State University Human Research Protectio n Program approved these studies (IRB C07 - 1201, 15 - 1240 and 14 - 170M). Sample Collection, DNA Extraction and Amplification Fecal samples were collected from pregnant women in their third trimester, and from infants at 1 month , 6 months, 12 months and 24 months of age using Para P ak collection tubes (Meridian Biosciences, Cincinnati, OH) . Samples were sent to our lab by mail or retrieved from - 80°C upon reaching the lab. Ideally, fecal samples would be immediately processed or frozen at - 80 °C (100) , however our participants shipped the samples from home, exposing the microbiota to air and room temperature storage conditions for several days. These conditions may affect the abundances of the microbiota (100 102) , but community differences across samples are pres erved regardless of storage method (101,103,104) . Another study found that , compared to immediate fecal DNA extraction, biases are introduced to microbiota data regardless of sample storage conditions (105) . In other words, ensuring that samples are all processed using the same method is potentially more important than immediate sample processing since each storage condition results in unique biases in the data. 19 DNA extractions were performed using the MoBio Powersoil DNA Isolation kit (Qiagen (106) with a few alterations: after incubating in C3 solution, the samples were centrifuged for 2 minutes and the DNA was eluted from the spin columns with 50 µ L of low EDTA TE buffer (IDT, Coralville, IA) heated to 55°C. Barcoded primers were used to amplify the V4 region of the 16S rRNA gene following the mothur wet lab documentation (107) . Prim ers SB501 - SB508 and SA701 - SA712 were ordered from IDT (Coralville, IA). PCR amplification also followed the wet lab protocol outlined in the mothur documentation (107) . Briefly, Accuprime Pfx Supermix (ThermoFisher, Waltham, MA) was mixed with at least 10 ng of template DNA and the primer pair (final concentration of 500 nM for both the forward and reverse primers) in a final reaction volume of 20 µ L. Reactions were performed in triplicate and amplified using the following thermocycler settings: (1x) 2 min at 95°C; (30x) 20 s at 95°C, 15 s at 55°C, 5 min at 72°C; 10 min 72°C. Amplification succe ss was checked by electrophoresis on a 1% agarose gel run at 200 V for 30 min. Successful amplification triplicates were pooled and purified using Agencourt AMPure XP (Beckman Coulter, Brea, CA) with the following alterations to the protocol: 0.7 times the sample volume of AMPure was used for purification and 16S rRNA DNA was eluted using 25 µ L of low EDTA TE buffer (IDT, Coralville, IA). This lower volume of AMPure excludes fragments of DNA with a lower base count than the amplified DNA. After purification , the concentration of 16S rRNA gene amplicons was quantified using the Quant - IT dsDNA assay kit (Invitrogen, Carlsbad, CA). Purified 16S rRNA amplicons were pooled and quality checked using an Agilent 2100 Bioanalyzer with the High Sensitivity DNA Chip (A gilent, 5067 - 4626). For sequencing, equal amounts (in nanograms) of the purified 16S samples were pooled and submitted to the 20 Michigan State University Research Technology Support Facility Genomics Core for paired - end 250 base - pair sequencing on the Illumi na MiSeq platform using V2 chemistry. Processing and Analysis of Sequence Data Sequence reads were processed in mothur using the Illumina MiSeq SOP (107) . The mothur program is a bioinformatics tool used mainly to process and analyze 16S DNA sequences produ ced from high - throughput DNA sequencing platforms such as the Illumina MiSeq. O perational taxonomic unit (OTU) taxonomies were assigned by phylotype in mothur using the SILVA reference taxonomy ( release V 128 for chapters 3 and 4 ; release V 132 for chapter 5 ) (108) . SILVA, RDP and Greengenes are all reference databases that can be used to identify microbes from their DNA sequences. We use the SILVA database because the mothur SOP suggests that the SILVA database is the best reference aligner to use for 16S data. Rea d processing was done in mothur using the High - Performance Computing Cluster at Michigan State University. Sample reads were rarefied (15,000 reads in chapter 3 , 10,000 reads in chapter 4 and 10,000 reads in chapter 5 ) 999 times, averaged, and rounded to t he nearest integer before further analysis. Rarefaction curves were generated to confirm adequate community coverage. Data Analysis Mothers were classified, by maternal pre - - 80%, 50%, 20 - 50%, 2 0% or 0% following Bonuck et al. (109) . Depending on the number of individuals per breastfeeding strat a , we transformed this variable by combining groups together t his will be discussed in more 21 detail in the methods section of the following chapters. We will us e the following variables: human milk exposure (100%, 80%, 50 - 80%, 50%, 20 - 50%, 20% or 0%; groups will be combined if there is a small n), pre - pregnancy BMI (continuous or grouped into normal, overweight and obese), delivery mode (C - section or vaginal), se x (male or female), age (continuous), sample shipping time (continuous) and cohort (ARCH GUT or BABY GUT ). There were no differences between our cohorts at any timepoint more information can be found in Table 2 , Table 13 and Table 19 . Two common analysis methods used in microbiome research are alpha (within a sample) and beta (between samples) diversity metrics. Alpha diversity is comprised of richness and evenness, which measure how many unique taxa are present and at what abundance th e unique taxa are found, respectively. Beta diversity is similar to alpha diversity but compares the richness and evenness between samples . This produces a similarity score for every pair of samples, which can then be plotted to show how similar or dissimi lar each sample is from the others . Alpha diversity was calculated in R (110) for Chao1 (presence/absence) , Shannon (abundance equal weight for all taxa present ) and inverse Simpson (abundance more weight given to high abundance taxa) indices with the vegan package (111) . A lpha diversity normality was determined using the Shapiro - Wilk test . Beta diversity was visualized using the Sorensen (presence/absence) and Bray - Curtis (abundance) dissimilarity scores and plotted with principle coordinate analysis (PCoA). Differences between group means w ere tested using PERMANOVA (adonis function) and differences in group variances w ere tested with PERMDISP (betadisper function) using the vegan package. More details on data analysis will be discussed in the methods section of the following chapters. 22 C HAPTER 3: MICHIGAN COHORTS TO DETERMINE ASSOCIATIONS OF MATERNAL PRE - PREGNANCY BODY MASS INDEX WITH PREGNANCY AND INFANT GASTROINTESTINAL MICROBIAL COMMUNITIES: LATE PREGNANCY AND EARLY INFANCY This chapter has been published at: Sugino K, Paneth N, Comstock 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. PL OS ONE. 2019;14:e0213733. Abstract Background : About 25% of women in the United States are obese prior to becoming pregnant. Although there is some knowledge about the relationship between the gastrointestinal microbiota and obesity, little is known about the relationship between pre - pregnancy obesity and the gastrointestinal microbiota in pregnancy or its impact on infant gut microbiota. However, the composition of the gut microbiota early in life may influence childhood health. Thus, the objective of this research was to identify associations betwee n maternal pre - pregnancy obesity and the pregnancy or early infancy microbiotas. Results : Fecal bacterial communities from overweight women had lower microbiota diversity (Chao1: p=0.0 2 ; inverse Simpson: p=0.0 5 ; Shannon: p=0.0 2 ) than communities from norma l weight or obese women. The beta diversity of overweight women differed from those of normal and obese women at the genus and phylum levels (p=0.003 and p=0.0 2 , respectively). Pre - pregnancy overweight women had higher abundances of Bacteroides and a lower abundance of 23 Phascolarctobacterium than women who were normal weight or obese prior to becoming pregnant. Normal weight women had lower abundances of Acidaminococcus and Dialister than overweight and obese women. Infant community composition tended to differ in membership (Sorensen index) by maternal pre - pregnancy BMI category, and significantly differed by delivery mode and breastfeeding exclusivity (p=0.0 6 , p=0.001 , p=0.0 08 , respectively). Infants from normal weight women had lower abundances of Megasphaera than infants from overweight or obese women. Streptococcus was lowest in abundance in infants from overweight women , and Staphylococcus was lowest in abundance in infants from obese women. Conclusion: Maternal and infant microbiotas are associated with maternal pre - pregnancy BMI. Futur e work should determine functional differences in the infant microbiome , how they are related to maternal pre - pregnancy BMI, and whether differences in composition or traits persist over time. Introduction The gut hosts a diverse community of microbes that interact with biological functions in both humans and animals. For example, the gut microbiota impacts immune system development, digestion of food components and influences weight gain (13 15,112) . In mice, the presence of a gut microbiota promotes increased adiposity and weight gain, possibly by increasing the amount of energy extracted and absorbed from food (13,14,17,113) . Similarly, germ - free mice transplanted with the microbiota from an obese hum an twin have more body fat compared to mice that receive a microbiota transplant from the lean twin (14) . In healthy human adults, there is evidence that shifts in the Firmicutes and Bacteroidetes abundances are associated with weight gain and obesity, though these shifts in phyla abundances are not consistent between studies (14,18,19) . Thus, although there is strong evidence for a relationship between gut 24 microbiota and obesi ty in mice, the evidence in humans is weak (19) . To address this question more definitively in humans, analysis of thes e communities during pregnancy and early infancy (31,58) can inform how the gut microbiota is associated with maternal pre - pregnancy BMI. While it is known that maternal pre - pregnancy BMI is associated with infant development and child weight (114,115) s unknown why these relationships exist. However, these effects may be mediated through the pregnancy and infancy microbiotas (21) . In the U nited States, 50% of women giving birth in 2014 were overweight or obese (23) . In adults, high BMI is associated with differences in the gut microbiome, such as low alpha diversity, altered community structure and changes in the functional capacity of the metagenome (116) . Similar changes have been observed over the course of pregnancy (31) possibly due to alterations in hormone levels, the immune system and metabolism (117) . In the first trimester, the gut micr obiota displays a significantly lower alpha and higher beta diversity concurrent with higher abundances of the phyla Proteobacteria and Actinobacteria compared to the third trimester (118) . There is evidence of differences in the micr obiome during pregnancy based on body weight (98,119) and gestational weight gain (120) , but the evidence on pre - pregnancy weight and microbiota in late pregnancy is contradictory. Some studies fo und no association between pre - pregnancy BMI and the pregnancy microbiota (31,120) , while another found overweight/obese women had a lower abundance of Parabacteroides and Bifidobacterium compared to normal weight (22) . This suggests that pregnan cy may be an important driving factor of the maternal gut microbiota in some populations, which can affect the bacteria infants are exposed to during and after birth (91) . The infant microbiota increases in bacterial abundance and becomes more diverse over time as the infant is exposed to changes in diet and environment (56,121) . One of the earliest 25 environmental exposures is mode of delivery. In the first week of life, vaginally born infants communities such as Bifidobacterium , Parabacteroides and Escherichia - Shigella compared to infants born via cesarean delivery (15) . In contrast, the microbiota of infants born via C - section are enriched in skin, oral and environmental bacteria such as Enterobacter , Staphylococcus and Streptococcus (15,60,65) . At 4 and 12 months, the association between mode of delivery and gut microbiota is still present, but the differences decrease over time (15) . Breastfeeding also affects gut microbiota (122) . Lactobacilli and Bifidobacterium , both of which utilize human milk oligosaccharides tend to dominate the gut microbiota of breastfed infants (15,79,123) . High maternal pre - pregnancy BMI has been linked to an increased risk of high childhood BMI (124) . Obese children are more likely to be obese adults with chronic medical problems (125) . The infant microbiome may explain predisposition to weight gain and metabolic dysregulation and may also explain the effects that breastfeeding (126) and cesarean delivery (127) have on obesity risk during childhood. A study by Tun et al. found that children at 1 year of age were 3 times more likely to be overweight if they were born vaginally from pre - pregnancy overweight or obese mothers, which was associated with an increased abundance of Lachnospiraceae (21) . Other studies that have assessed maternal pre - pregnancy BMI and the infant microbiota found that beta div ersity was significantly different between infants from normal and overweight/obese groups (22) and another found that only vaginally born neonates had altered micr obiota based on pre - pregnancy BMI (128) . Here we use samples from women in their th ird trimester of pregnancy and their infants to investigate associations between maternal pre - pregnancy BMI and the pregnancy or infant microbiotas, controlling for effects of other factors such as breastfeeding and delivery mode. 26 Materials and Methods Participant enrollment, sample collection, sample processing and sequence read processing were performed as described in chapter 2. Here, we are analyzing the maternal T3 fecal sample and the infant 1month fecal sample. Breast milk in the diet was reported by mothers who completed a validated survey that separates the infant diet into seven levels of human milk exposure (100%, 80%, 50 - 80%, 50%, 20 - 50%, 20% and 0%). However, because of insufficient sample size for each of the groups, diet was recategorized a s exclusively or non - exclusively receiving human milk in the diet, either from the breast or a bottle. Infants whose mothers reported the infant had been given any antibiotic since birth were considered antibiotic exposed (n=4, 2 in overweight and 2 in obe se, Table 1). However, none of these infants were taking antibiotics at the time the sample was collected. Alpha (within - sample) diversity (Chao1, inverse Simpson and Shannon) were compared using a Kruskal - Wallis test or a Spearman correlation test for cat egorical and continuous variables, respectively. Pairwise tests between BMI categories was performed using a Dunn test. Sorensen and Bray - Curtis dissimilarities were calculated and ordinated using principle coordinate analysis (PCoA). PERMANOVA was perform ed to test for significant differences in beta - diversity using both univariate and multivariate models. In GUT or BABY GUT ) and shipping time were included as covariates with pr e - pregnancy BMI. For the infants, shipping time, breastfeeding, sex, cohort, infant age and delivery mode were covariates with maternal pre - pregnancy BMI. PERMDISP was used to test for differences in sample dispersion. Individual taxa were compared across BMI categories using a negative binomial model in the MASS package (1 29) on the taxa that composed >1% abundance on average. The 27 Benjamini - Hochberg method was used for false discovery rate correction. P - values less than 0.05 were considered significant, and p - values less than 0.10 were considered a trend. Results We collected fecal samples from 3 9 dyads . Of these, there was one twin birth, and only one twin was included in this analysis. In total, there were 1 2 normal weight, 1 1 overweight and 1 6 obese women. Among the women, none of the self - reported characteristics differed by BMI category ( Table 1 ). Fewer infants of obese women consumed breastmilk exclusively ( Table 1 ). A higher percent of infants born to normal weight mothers tended to be born vaginally compared to infants of overweight and obese mothers ( Table 1 ). Infant age at the time of sampling ranged from 2 to 111 days (median=8.5 days). Infants born to normal weight women tended to be younger at the time of sampling than those born to overweight or obese women ( Table 1 ). The characteristics of participants in ARCH GUT (n=24) and BABY GUT (n=15) were similar ( Table 2 ). 28 Table 1 . Participant Characteristics Pregnant Women All p - value Participants, n 39 12 11 16 Pre - pregnancy BMI (kg m^ - 2) 1 28.5 ± 5.5 22.7 ± 1.4 c 26.7 ± 1.5 b 34.1 ± 3.3 a <0.001 Maternal Age (years) 1 31.4 ± 4.5 33.3 ± 2.5 29.6 ± 5.1 4 31.0 ± 5.1 0.15 Currently on Antibiotics 3 1 (2.6) 1 (8.3) 0 (0.0) 0 (0.0) 0.32 Parity 2 2.0 (1 - 6) 2.0 (1 - 5) 2.0 (1 - 3) 2.0 (1 - 6) 0.96 Sample Shipping Time (days) 2 4.0 (0 - 11) 4.0 (1 - 7) 4.0 (0 - 11) 4.5 (2 - 7) 0.96 Infants All p - value Vaginal Delivery 3 26 (66.7) 11 (91.7) 5 (45.5) 10 (62.5) 0.06 Female 3 14 (35.9) 3 (25.0) 4 (36.4) 7 (43.8) 0.59 Exclusively Breastfed 3 24 (61.5) 11 (91.7) a 6 (54.5) ab 7 (43.8) b 0.03 Antibiotic Exposure Since Birth 3 4 (10.3) 0 (0.0) 2 (18.2) 2 (12.5) 0.33 Infant Sample Shipping Time (days) 2 4.0 (1 - 15) 3.0 (2 - 11) 4 5.0 (1 - 15) 5.0 (2 - 14) 4 0.85 Infant Age at Sampling Time (days) 2 8.5 (2 - 111) 5.5 (2 - 57) 10.0 (7 - 60) 17.0 (3 - 111) 4 0.05 Values in a row that do not contain the same superscript are significantly different, p<0.05 1 mean ± SD 2 median (range) 3 n (%) 4 Missing information for one sample 29 Table 2 . Cohort Characteristics Pregnant Women ARCH BABY p - value N 24 15 Pre - pregnancy BMI (kg m^ - 2) 1 29.0 ± 5.4 27.7 ± 5.8 0.47 Normal 3 7 (29.2) 5 (33.3) 0.29 Overweight 3 5 (20.8) 6 (40.0) Obese 3 12 (50.0) 4 (26.7) Maternal Age (years) 1 31.3 ± 3.8 31.4 ± 5.7 4 0.95 Currently on Antibiotics 3 0 (0.0) 1 (6.7) 0.81 Parity 2 2.0 (1 - 6) 2.0 (1 - 3) 0.99 Sample Shipping Time (days) 2 4.0 (1 - 7) 5.0 (0 - 11) 0.09 Infants ARCH BABY p - value Vaginal Delivery 3 17 (70.8) 9 (60.0) 0.73 Female 3 11 (45.8) 3 (20.0) 0.2 Exclusively Breastfed 3 14 (58.3) 10 (66.7) 0.86 Antibiotic Exposure Since Birth 3 1 (4.2) 3 (20.0) 0.3 Infant Sample Shipping Time (days) 2 4.0 (2 - 14) 4 6.0 (1 - 15) 4 0.2 Infant Age at Sampling Time (days) 2 20.7 ± 26.2 4 13.9 ± 14.4 0.66 1 mean ± SD 2 median (range) 3 n (%) 4 Missing information for one sample 30 Alpha diversity Pregnant women had significantly higher microbiota diversity at the genus level than their infants, as measured by either Chao1 (p<0.001), inverse Simpson (p<0.001) and Shannon indices (p<0.001). The fecal bacterial communities of women who were overweight prior to becoming pregnant were less rich (Chao1), more even (inverse Simpson) and had a lower diversity (Shannon) than those of women who were normal weight or obese prior to becoming pregnant ( Table 3 ). There was no difference in the alpha diversity of infant fecal bacterial communities by maternal pre - pregnancy BMI ( Table 3 ). Table 3 . Alpha Diversity of the Fecal Microbiota of Mothers and Infants by Maternal Pre - pregnancy BMI Category Pregnant Women 1 All Normal Overweight Obese p - value Chao1 122.7 ± 23.0 133.3 ± 24.9 a 107.7 ± 18.9 b 125.1 ± 19.7 ab 0.02 Inverse Simpson 9.6 ± 4.7 10.9 ± 5.4 6.7 ± 5.0 10.6 ± 3.1 0.05 Shannon 2.8 ± 0.5 3.0 ± 0.4 a 2.4 ± 0.7 b 2.9 ± 0.3 a 0.02 Infants 1 Chao1 46.3 ± 20.6 49.5 ± 22.0 44.2 ± 11.4 45.4 ± 25.0 0.58 Inverse Simpson 3.3 ± 1.2 3.7 ± 1.5 3.0 ± 0.9 3.2 ± 1.1 0.49 Shannon 1.4 ± 0.4 1.5 ± 0.4 1.4 ± 0.3 1.4 ± 0.4 0.51 Values in a row that do not contain the same superscript are significantly different, p<0.05 1 Values reported as mean ± SD Alpha diversity compared by infant age, sex , mode of delivery, breastfeeding exclusivity, antibiotic use since birth, and sample shipping time were not significant; however, the ARCH GUT infants had a significantly higher richness compared to the BABY GUT cohort ( Table 4 ). After stratifying the infants by maternal pre - pregnancy BMI and re moving infants exposed to antibiotics since birth (n=4), those in the obese category that were breastfed exclusively had 31 lower richness and Shannon diversity scores compared to non - exclusively breastfed infants ( Table 5 ). Alpha diversity of the infant fecal microbiota did not differ by mode of delivery for women who were overweight or obese prior to becoming pregnant. No comparisons between delivery mode and breastfeeding were possible in the normal weight group because of a lack of c - section deliveries and infants fed a mixed diet (n=1 for both). Table 4 . Infant alpha diversity by infant age, sex, mode of delivery and sample shipping time Age and Shipping 1 Infant Age Shipping Time Chao1 - 0.06, p=0.72 - 0.01, p=0.93 Inverse Simpson - 0.08, p=0.63 - 0.13, p=0.45 Shannon - 0.005, p=0.98 - 0.13, p=0.43 Delivery Mode 2 Vaginal C - Section p - value Chao1 47.0 ± 16.2 44.9 ± 28.2 0.16 Inverse Simpson 3.3 ± 1.4 3.2 ± 0.8 0.8 Shannon 1.4 ± 0.4 1.5 ± 0.3 0.65 Sex 2 Males Females Chao1 49.0 ± 24.7 41.5 ± 8.6 0.63 Inverse Simpson 3.2 ± 1.3 3.4 ± 1.2 0.65 Shannon 1.5 ± 0.4 1.4 ± 0.4 0.78 Cohort 2 Baby ARCH Chao1 51.0 ± 23.9 38.8 ± 10.9 0.03 Inverse Simpson 3.6 ± 1.4 2.9 ± 0.9 0.14 Shannon 1.5 ± 0.4 1.3 ± 0.3 0.05 Breastfeeding 2 Exclusive Mixed Chao1 48.4 ± 24.6 43.0 ± 12.0 0.98 Inverse Simpson 3.2 ± 1.4 3.3 ± 1.4 0.48 Shannon 1.4 ± 0.4 1.5 ± 0.3 0.45 Antibiotic Used Since Birth 2 Any None Chao1 56.8 ± 51.8 45.1 ± 14.9 0.39 Inverse Simpson 2.6 ± 0.5 3.4 ± 1.3 0.19 Shannon 1.3 ± 0.4 1.5 ± 0.4 0.46 1 rho and p - values reported 2 mean ± SD 32 Table 5 . P - values of infant alpha diversity by delivery mode and breastfeeding stratified by BMI category Normal ROUT BF Chao1 NA NA Inverse Simpson NA NA Shannon NA NA Overweight ROUT BF Chao1 0.19 0.73 Inverse Simpson 0.41 1 Shannon 0.73 0.73 Obese ROUT BF Chao1 0.78 0.04 Inverse Simpson 0.3 0.23 Shannon 0.45 0.03 p - values reported Beta diversity Pregnant women had different bacterial communities than infants at both the genus and phylum levels for both Sorensen and Bray - Curtis dissimilarities (p=0.001) ( Figure 1 A - C). The dispersion was also significantly different at the genus and phylum levels (p=0.001) due to the large variation in infant fecal bacteri al composition and the relatively more similar communities in the women. 33 Figure 1 . The fecal bacterial community structure of pregnant women differs significantly from that of infants PCoA of the microbiota for all samples at the (A) phylum - level using Bray - Curtis dissimilarity, and at the genus - level using (B) Sorensen index and (C) Bray - Curtis Dissimilarity. Axes percentages represent the amount of variation in the data explained by the a xis, calculated from the PCoA eigen values. Axes ranges represent the relative dissimilarity present between the samples. 34 We separately assessed changes in community composition (accounting for the presence and and community structure (accounting for the relative contributions of taxa based on Bray - Curtis dissimilarity) in the pregnant women and infants. During pregnancy, the fecal bacterial community structure of overweight women differed by BMI category at the phylum (F= 4 . 1 , p=0.003) and genus levels (F=1. 7 , p=0.0 2 ) ( Error! Reference source not found. A - B). Figure 2 . Pre - pregnancy overweight women have different fecal microbiota compositions than normal and obese women at phylum/genus levels PCoA of the by Bray - Curtis dissimilarity at the (A) phylum - level and (B) genus - level microbiota. Axes percentages represent the amount of variation in the data explained by the axis, calculated from the PCoA eigen values. Axes ranges represent the relative dis similarity present between the samples. 35 Adjusting for maternal age, cohort and shipping time, BMI category remained significant at the phylum (F=3.9, p=0.003) and genus level (F=1.6, p=0.03) for community structure. In addition, cohort (F=2.1, p=0.02) an d shipping time (F=1.8, p=0.04) were significant ly associated with the fecal bacterial communities at the genus level. The overall model for maternal age, cohort, shipping time and BMI category was significant at the phylum (F=2.32, p=0.01) and genus level s (F=1. 7 , p=0.001). There were no differences by BMI category or any other variables for community composition by the Sorensen index ( Figure 3 ). Figure 3 . Bacterial community membership does not differ in pregnant women by pre - pregnancy BMI category PCoA of the genus - level microbiota using Sorensen index. Axes percentages represent the amount of variation in the data explain ed by the axis, calculated from the PCoA eigen values. Axes ranges represent the relative dissimilarity present between the samples. 36 In univariate analyses, infant community composition (Sorensen index) at the genus level tended to differ by maternal pre - pregnancy BMI Category (F=1. 4 , p=0.0 6 ), and differed by delivery mode (F= 2 . 9 , p=0.001) and breastmilk in the infant diet (F=2. 2 , p=0.0 08 ) ( Figure 4 ). Figure 4 . Relationship between infant fecal bacterial membership and maternal/infant variables at the genus level PCoA of the genus - level microbiota using the Sorensen index comparing (A) maternal pre - pregnancy BMI category, (B) delivery mode, (C) breast milk in diet and (D) a ntibiotic exposure . Axes percentages represent the amount of variation in the data explained by t he axis, calculated from the PCoA eigen values. Axes ranges represent the relative dissimilarity present between the samples. In a multivariate model including BMI, breastfeeding, delivery mode, sex and antibiotic exposure, delivery mode was significant using the Sorensen index (F=2.1, p=0.01), and exposure to antibiotics was significant for Bray - Curtis (F=2.1, p=0.04). After additionally adjusting for 37 sample shipping time, sex and cohort, none of the variables were significant for Sorensen. In univariate analyses, antibiotic exposure was significant for Bray - Curtis (F=2.2, p=0.03), but maternal pre - pregnancy BMI category was not associated with infant gut community structure (Bray - Curtis dissimilarity matrix, Figure 5 ). Similarly, the overall multivariate model for community structure (Bray - Curtis dissimilarity matrix) of the in fant microbiota was not significant. At the phylum level, there were no significant differences. Figure 5 . Relationship between infant fecal bacterial structure at the genus level and maternal/infant variables PCoA of the genus - level microbiota by Bray - Curits dissimilarity comparing (A) maternal pre - pregnancy BMI category, (B) delivery mode, (C) breast milk in diet and (D) a ntibiotic exposure . Axes percentages represent the amount of variation in the data explained by the axis, calculated from the PCoA eigen values. Axes ranges represent the relative dissimilarity present between the samples. 38 When stratifying the infant samples by maternal pr e - pregnancy BMI category and removing infants exposed to antibiotics (n=4), only breastfeeding was associated with the Sorensen index at the genus level in the subset of infants born to women who were obese prior to becoming pregnant (F=2.0, p=0.02) ( Figure 6 ). At the phylum level, breastfeeding was significantly associated with the gut microbiota community in infants born to either overweight (F=2.7, p=0 .048) or obese (F=3.7, p=0.009) women ( Figure 6 ). Figure 6 . Infant fecal microbiota stratified by maternal overweight and obese categories PCoA of the infant gut microbiota by breastfeeding status of the (A) overweight group using the Sorensen index at the genus level (B) obese group using the Sorensen index at the genus level (C) overweight group using Bray - Cu rtis dissimilarity at the phylum level (D) obese group using Bray - Curtis dissimilarity at the phylum level. Axes percentages represent the amount of variation in the data explained by the axis, calculated from the PCoA eigen values. Axes ranges represent t he relative dissimilarity present between the samples. 39 Sensitivity analysis of infant microbiota data Since age is known to be associated with changes in the infant microbiota (15,31,130) , we conducted a sensitivity analysis on two subs ets of the infant data based on the median age (8.5 days) and the mean age (18 days) of the infants. Alpha diversity in either age subset was not associated with maternal pre - pregnancy BMI, delivery mode, age, sex, sample shipping time or cohort ( Table 6 & Table 7 ). 40 Table 6 . Sensitivity of infant fecal microbiota alpha diversity to infant age -- Days of Age Age and Shipping 1 Infant Age Shipping Time Chao1 - 0.26, p=0.29 - 0.06, p=0.81 Inverse Simpson - 0.18, p=0.46 Shannon - 0.26, p=0.29 Normal Overweight Obese p - value n 9 5 5 Chao1 50.8 ± 24.6 51.8 ± 12.0 37.2 ± 6.9 0.2 Inverse Simpson 3.8 ± 1.7 3.1 ± 1.1 2.5 ± 0.6 0.41 Shannon 1.5 ± 0.5 1.6 ± 0.4 1.2 ± 0.2 0.53 Delivery Mode 2 Vaginal C - Section Chao1 48.7 ± 20.5 42.9 ± 10.7 0.66 Inverse Simpson 3.3 ± 1.5 3.4 ± 1.0 0.47 Shannon 1.4 ± 0.4 1.5 ± 0.4 0.74 Sex 2 Males Females Chao1 49.3 ± 21.0 42.5 ± 10.5 0.89 Inverse Simpson 3.3 ± 1.5 3.3 ± 1.3 0.69 Shannon 1.4 ± 0.4 1.4 ± 0.4 0.89 Cohort 2 Baby ARCH Chao1 38.7 ± 10.7 53.9 ± 21.2 0.06 Inverse Simpson 2.9 ± 1.0 3.6 ± 1.6 0.54 Shannon 1.3 ± 0.3 1.5 ± 0.4 0.18 Breastfeeding 2 Exclusive Mixed Chao1 47.9 ± 21.1 46.7 ± 14.3 1 Inverse Simpson 3.5 ± 1.6 2.8 ± 0.7 0.77 Shannon 1.4 ± 0.5 1.4 ± 0.25 0.97 1 rho and p - values reported 2 mean ± SD 41 Table 7 . Sensitivity of infant fecal microbiota alpha diversity to infant age -- Age and Shipping 1 Infant Age Shipping Time Chao1 - 0.25, p=0.22 - 0.04, p=0.86 Inverse Simpson - 0.10, p=0.63 Shannon - 0.03, p=0.88 - 0.19, p=0.35 Normal Overweight Obese p - value n 10 7 8 Chao1 50.4 ± 23.3 48.4 ± 11.8 37.8 ± 6.6 0.2 Inverse Simpson 3.8 ± 1.6 3.1 ± 1.0 2.9 ± 1.1 0.59 Shannon 1.5 ± 0.5 1.4 ± 0.3 1.3 ± 0.4 0.57 Delivery Mode 2 Vaginal C - Section Chao1 46.9 ± 18.6 42.3 ± 9.2 0.73 Inverse Simpson 3.2 ± 1.4 3.7 ± 1.0 0.22 Shannon 1.4 ± 0.4 1.6 ± 0.3 0.33 Sex 2 Males Females Chao1 48.1 ± 19.2 40.9 ± 9.1 0.51 Inverse Simpson 3.4 ± 1.5 3.3 ± 1.1 0.84 Shannon 1.5 ± 0.4 1.4 ± 0.4 0.67 Cohort 2 Baby ARCH Chao1 39.4 ± 10.2 49.4 ± 18.9 0.11 Inverse Simpson 2.8 ± 1.0 3.7 ± 1.4 0.21 Shannon 1.3 ± 0.3 1.5 ± 0.4 0.1 Breastfeeding 2 Exclusive Mixed Chao1 46.0 ± 18.8 45.3 ± 12.8 0.93 Inverse Simpson 3.4 ± 1.5 3.2 ± 0.9 0.98 Shannon 1.4 ± 0.4 1.5 ± 0.3 0.8 1 rho and p - values reported 2 mean ± SD 42 days of age, only infant sex significantly associated with gut microbiota membership as measured by the Sorensen metric (F=1.8, 0.03), however, the dispersion was also significantly different (F=5.6, p=0.03) ( Figure 7 metric was significant by maternal BMI (F=1.5, p=0.04) and delivery mode (F=1.92, p=0.03) ( Figure 8 ). Figure 7 . Sensitivity analysis of the infant microbiota, including infants from 2 to 9 days old PCoA of the genus - level microbiota by Sorensen dissimilarity comparing (A) maternal pre - pregnancy BMI category , (B) delivery mode , (C) breastfeeding status and (D) sex . Axes percentages represent the amount of variation in the data explained by the axis, calcu lated from the PCoA eigen values. Axes ranges represent the relative dissimilarity present between the samples. 43 Figure 8 . Sensitivity analysis of the infant microbiota, including infants from 2 to 18 days old PCoA of the genus - level microbiota by Sorensen dissimilarity comparing (A) maternal pre - pregnancy BMI category , (B) delivery mode , (C) breastfeeding status and (D) sex . Axes percentages represent the amount of variation in the data explained by the axis, calcu lated from the PCoA eigen values. Axes ranges represent the relative dissimilarity present between the samples. Differential patterns of bacterial taxa Women who were overweight prior to becoming pregnant had higher abundances of Bacteroides and lower abundances of Phascolarctobacterium than women who were normal weight or obese prior to becoming pregnant. Women who were normal weight prior to becoming pr egnant had lower abundances of Acidaminococcus and Dialister but higher abundances of Phascolarctobacterium than overweight and obese women ( Table 8 ). At the phylum level, the 44 overweight women had higher Bacteroidete s than both normal weight and obese women, but lower abundances of Firmicutes than obese women ( Table 8 ). Table 8 . Significantly Different Taxa in the Fecal Microbiota by Maternal Pre - pregnancy BMI Category Pregnant Women - Genus Normal Overweight Obese Bacteroides 20.0 ± 7.2 b 38.1 ± 21.7 a 19.8 ± 8.2 b Phascolarctobacterium 3.0 ± 3.1 a 1.7 ± 2.3 c 2.3 ± 3.1 b Acidaminococcus 0.0006 ± 0.002 b 2.4 ± 6.7 a 1.9 ± 4.8 a Dialister 0.7 ± 1.0 c 1.1 ± 1.6 b 1.7 ± 2.6 a Pregnant Women - Phylum Bacteroidetes 29.5 ± 8.8 b 49.3 ± 17.6 a 30.2 ± 12.1 b Firmicutes 49.5 ± 11.1 ab 37.6 ± 15.0 b 52.7 ± 14.0 a Infants - Genus Megasphaera 0.09 ± 0.2 b 6.0 ± 19.0 a 9.5 ± 21.5 a Streptococcus 7.0 ± 9.2 a 0.9 ± 1.2 b 2.6 ± 4.1 ab Staphylococcus 5.3 ± 9.4 a 2.1 ± 2.5 ab 0.7 ± 1.4 b Acidaminococcus 0.0006 ± 0.002 b 0.0006 ± 0.002 b 3.4 ± 13.6 a Escherichia - Shigella 28.1 ± 28.3 a 19.9 ± 20.1 b 17.7 ± 24.1 c Akkermansia 0.004 ± 0.004 b 0.006 ± 0.009 b 2.6 ± 9.7 a Infants - Phylum Verrucomicrobia 0.005 ± 0.004 b 0.005 ± 0.008 b 2.6 ± 9.7 a Values reported as mean (%) ± SD Values in a row that do not contain the same superscript are significantly different, p<0.05 P - values are Benjamini - Hochberg corrected Infants born to women who were normal weight prior to becoming pregnant had lower abundances of Megasphaera , but higher abundances of Escherichia - Shigella than infants born to overweight or obese women. Streptococcus was less abundant in infants from women who were overweight prior to pregnancy and Staphylococcus was lower in infants from obese women ( Table 8 ). Infants born to obese women had significantly more Acidaminococcus and 45 Akkermansia in their fecal bacterial communities than infants born to normal or overweight women. Vaginally born infant s had significantly higher abundances of Megasphaera , Parabacteroides , and Escherichia - Shigella , but lower Acidaminococcus and Akkermansia than C - section born infants ( Table 9 ). Infants that exclusively consumed human milk had higher abundances of Staphylococcus and Escherichia - Shigella , but lower abundances of Acidaminococcus and Akkermansia than infants that consumed a mixed diet ( Table 9 ). 46 Table 9 . Significantly Different Taxa in Fecal Microbiota by Delivery Mode and Breastfeeding Infant - Genus Vaginal C - section Megasphaera 8.4 ± 20.5 0.009 ± 0.02 Parabacteroides 4.8 ± 10.8 0.3 ± 0.9 Escherichia - Shigella 22.4 ± 26.4 19.9 ± 20.0 Akkermansia 0.01 ± 0.03 3.1 ± 10.8 Acidaminococcus 0.0005 ± 0.002 4.2 ± 15.1 Infant - Phylum Verrucomicrobia 0.01 ± 0.03 3.1 ± 10.8 Infant - Genus Exclusively Breastfed Mixed Feeding Escherichia - Shigella 24.1 ± 24.1 17.4 ± 18.2 Akkermansia 0.07 ± 0.3 2.6 ± 10.1 Acidaminococcus 0.002 ± 0.006 3.6 ± 14.1 Staphylococcus 3.8 ± 6.9 0.4 ± 1.0 Infant - Phylum Verrucomicrobia 0.08 ± 0.3 2.6 ± 10.1 Infant - Genus Antibiotics Since Birth No Antibiotics Megasphaera 0.04 ± 0.04 6.2 ± 18.0 Parabacteroides 0.01 ± 0.02 3.6 ± 9.5 Escherichia - Shigella 0.15 ± 0.3 24.0 ± 24.4 Acidaminococcus 0.007 ± 0.01 1.6 ± 9.2 Infant - Phylum Bacteroidetes 0.9 ± 1.0 13.7 ± 20.4 Values reported as mean (%) ± SD All comparisons, p<0.05 ( Benjamini - Hochberg corrected) After stratifying the infant data by BMI category, infants born vaginally to women who were overweight prior to becoming pregnant had higher abundances of Bifidobacterium, Megasphaera and Parabacteroides but lower abundances of unclassified Enterobacteriaceae, Clostridium sensu stricto , Escherichia - Shig ella , and Enterococcus ( Table 10 ). 47 Table 10 . Significantly different taxa by delivery mode and breastfeeding in the gut microbiota of infants born to women who were overweight prior to becoming pregnant Infant - Genus Vaginal C - section Uncl. Enterobacteriaceae 3.6 ± 1.7 17.8 ± 26.0 Bifidobacterium 27.7 ± 26.8 11.4 ± 11.4 Megasphaera 16.6 ± 31.1 0.004 ± 0.006 Parabacteroides 2.9 ± 3.9 0.6 ± 1.5 Clostridium Sensu Stricto 0.01 ± 0.02 16.9 ± 24.0 Escherichia - Shigella 17.6 ± 23.0 30.0 ± 17.0 Enterococcus 0.04 ± 0.04 1.7 ± 2.3 Infant - Genus Exclusively Breastfed Mixed Feeding Uncl. Enterobacteriaceae 3.3 ± 1.4 18.7 ± 25.8 Bifidobacterium 30.1 ± 25.3 9.5 ± 10.0 Megasphaera 16.6 ± 31.1 0.005 ± 0.006 Parabacteroides 2.9 ± 3.9 0.6 ± 1.4 Veillonella 0.2 ± 0.4 1.7 ± 1.9 Clostridium Sensu Stricto 0.01 ± 0.02 16.9 ± 23.9 Escherichia - Shigella 13.9 ± 16.8 32.6 ± 19.1 Staphylococcus 2.8 ± 3.1 1.2 ± 1.5 Enterococcus 0.1 ± 0.2 1.6 ± 2.4 Because 4 of the 5 infants born vaginally to overweight women were exclusively breastfed, the same associations apply to exclusively breastfed and mixed diet infants. In the obese group, Megasphaera and Clostridium sensu stricto were more abundant, and Akk ermansia and Klebsiella were less abundant in the vaginally delivered infants compared to the c - section delivered infants ( Table 11 ). In contrast to bab ies born vaginally to women who were overweight prior to becoming pregnant, the infants born vaginally to women who were obese prior to becoming pregnant were just as likely to be mixed fed as they were to be exclusively breastfed. Infants born to women wh o were obese prior to becoming pregnant and fed a diet of exclusive breastmilk had higher abundances of Staphylococcus , but lower abundances of Megasphaera, 48 Akkermansia and Klebsiella than infants fed a mixed diet. At the phylum level, the gut microbiota o f infants born to women that were obese prior to becoming pregnant had a greater abundance of Verrucomicrobia when delivered by c - section or fed a mixed diet ( Table 11 ). Table 11 . Significantly different taxa by delivery mode and breastfeeding in the gut microbiota of infants born to women who were obese prior to becoming pregnant Infant - Genus Vaginal C - section Megasphaera 15.1 ± 26.0 0.002 ± 0.003 Clostridium sensu stricto 13.1 ± 17.2 0.7 ± 0.2 Akkermansia 0.02 ± 0.05 9.8 ± 19.5 Klebsiella 0.5 ± 0.8 9.1 ± 10.6 Infant - Phylum Verrucomicrobia 0.02 ± 0.05 9.8 ± 19.5 Infant - Genus Exclusively Breastfed Mixed Feeding Megasphaera 0.3 ± 0.6 18.7 ± 28.2 Staphylococcus 10.9 ± 16.8 8.5 ± 15.5 Akkermansia 0.004 ± 0.003 4.9 ± 13.8 Klebsiella 0.5 ± 1.0 4.8 ± 8.3 Infant - Phylum Verrucomicrobia 0.006 ± 0.005 4.9 ± 13.8 Discussion In this population of pregnant women and infants, we investigated the relationship between maternal pre - pregnancy BMI and the gut microbiota of women during late pregnancy and of their children during early infancy. The fecal bacterial communities of women who were overweight prior to becoming pregnant differed from those of women who were normal weight and obese prior to becoming pregnant. The overweight women had a lower alpha diversity and a high abundance of Bacteroides , which was likely the main driver of their overall community differences measured by beta diversity, compared to the other BMI categories. The alpha diversity of the infant microbiota did not differ by any of variables tested, but the beta diversity 49 of the infant microbiota was associated with delivery mode, human milk in the diet and antibiotic exposure when analyzed independently. However, only exclusive breastfeeding was significantly associated with the infant gut microbiota after stratifying by maternal pre - pregnancy BMI category. Wo men who were overweight prior to becoming pregnant tended to have a different microbiota than normal weight and obese women. Although some studies have found no differences in the gut communities of normal, overweight or obese women during pregnancy (31,32) , our population of pre - pregnancy overweight women had a different community structure at the genus and phylum lev els than normal and obese women. This difference was driven by the higher abundance of Bacteroides and Bacteroidetes in the feces of overweight women compared to that of women in the other BMI categories. That the third trimester microbiota of overweight w omen was different from the third trimester microbiota of normal and obese women was surprising. A higher abundance of Bacteroides in obese women compared to normal weight has been observed previously using fluorescent in - situ hybridization (119) , however, there were no participants in the overweight category as reference. Our population of overweight women also had microbiota with lower richness, lower diversity and higher evenness than that of normal weight or obese women as well as a significantly different community composition. Our results in overweight pregnant women may be due to other variables related to BMI , such as diet or lifestyle (131) , which have also been shown to alter the microbiome. Another possible reason for this difference could be gestational weight gain. Excessive gestational weight gain has been associated with a gut microbiot a dominated by Bacteroidetes rather than Firmicutes and a lower alpha diversity ( 120) . Thus, the overweight women herein may have had a different microbiota because of higher gestational weight gain compared to the normal or obese women. 50 Unfortunately, this information was not collected. Both gestational weight gain and maternal BMI have b years of age (132) . Shifts in gut microbiota composition over the cou rse of pregnancy due to hormonal, metabolic and immunological changes (117) may affect what taxa establish in the strain level (133,134) suggesting there is a partial transfer of gut bacteria from mother to child (91) . Since the microbiota influences weight gain and there is evidence of microbial transmittance from mother to child, it is possible an obesogenic microbiota can be transferred from mother to child. This, in turn, may increase the infant population, we found that microbiota membership was affected by both maternal characteristics and environmental exposures. Delivery mode and amount of breast milk in the diet were associated with the infa nt fecal microbiota before adjustment, while maternal pre - pregnancy BMI tended to be associated with the infant fecal microbiota. Many of the bacterial abundances that differed by maternal pre - pregnancy BMI, delivery mode and breast milk in the diet were s hared, such as Escherichia - Shigella and Acidaminococcus . In fact, Acidaminococcus has been shown to negatively affect growth in children that are at risk for malnourishment (135) . These associations suggest that maternal pre - pregnancy BMI category, delivery mode and amount of breast milk in the diet all interact in si milar ways with the infant microbiota. This may be due to associations between the pre - pregnancy BMI, delivery mode and breastfeeding variables themselves. For example, infants born to overweight and obese women have a higher risk of developing childhood o verweight/obesity and a 2 - 4 fold increased odds of being delivered via C - section compared to normal weight women (3 5) . Moreover, obese mothers often have 51 difficulties breastfeeding due to insufficient milk yields in the first few weeks of lactation (7) and are more likely to give their infant formula rather than breastmilk as a result (6,7) . After adjusting for maternal pre - pregnancy BMI, shipping time, infant age, breast milk and mode of delivery, delivery mode remained significantly associated with the beta - div ersity (Sorenson index) of the infant gut microbiota. Additionally, vaginally born infants had higher abundance of Parabacteroides and Escherichia - Shigella . Others have found that vaginally born infants have a higher proportion of their community in common such as Bifidobacterium and Parabacteroides and Escherichia - Shigella (15) . Bifidobacteriu m was significantly higher in our vaginally born infants only in the overweight, but not obese, category. This was related to the disparities in breastfeeding exclusivity between infants born to overweight women versus those born to obese women with each m ode of delivery category. Regardless of maternal pre - pregnancy BMI status, 33.3% of C - section born infants were exclusively fed human milk. In contrast, 80% of vaginally born infants of overweight women and 50% of vaginally born infants of obese women were exclusively fed human milk. Infants born to obese women, even if they are vaginally born, may be at a greater risk for allergies, overweight/obesity and other chronic diseases due to their lack of Bifidobacteria (136) . Breastfeeding impacts the abundances of several taxa, especially the genera Lactobacillus and Bifidobacteria , which can utilize the oligosaccharides found in breast milk (15,79,88,123) . Breastfeeding was no t associated with the alpha - or beta - diversity of the infant microbiota after adjustment for covariates, but breastfeeding was significantly associated with community membership at the genus - level and composition at the phylum - level when stratifying by BMI . Compared to the relationship we found between delivery mode and the microbiota, we saw similar associations between BMI category and Bifidobacterium abundance when comparing 52 exclusively breastfed infants to mixed - fed infants. However, this is likely due to collinearity between delivery mode and breastfeeding within the overweight group (15) . Others have shown that exclusiv e breastfeeding alters the microbiota compared to mixed feeding (15,137) and reduces the risk of overweight/obesity in both childhood and adulthood (138) . Infant age is associated with changes to the microbiota for several reasons: alterations to the gut environment, allowing for a shift from a high abundance of facultative anaerobes to obligate anaerobes over time; alterations in dietary intake; and alterations in gut transit time (130) . The full analysis and the sensitivity analyses of infants 9 days old or less and infants 18 days old or less led to similar conclusions for alpha diversity in infant fecal bacterial communities. Although, beta - diversity results were similar for the full population and the subset of infants 2 - 1 8 days of age, these results were not observed in the subset of infants 2 - 9 days of age likely due to decreasing sample size which reduced the infants in the obese category to 31%, those in the overweight category to 45% and those in the normal category to 75% of their original size. There are several limitations to this study. One limitation is the sample collection method. Women collected the samples in their home and shipped the samples to the lab allowing the sample to remain at non - ideal temperatures for an average of 4 days before storage at - 80°C. In general, it is known that the methods used to collect, store, extract and amplify samples have effects on microbiome data (100 102) . These conditions may affect the relative abundances of the microbiota, but community differences across sample s have been shown to be preserved regardless of storage method (101,103,104) . Thus, comparisons across groups within our study population are valid. The BMI and breastfeeding variables relied on self - reported data from the women. However, both the pre - pregna ncy self - report of height and weight (139) as well as the 53 breastmilk in the infant diet (109) instruments are validated, so self - report of these measures are not expected to affect the results. Furthermore, because of the relatively limited sample size and collinearity among variables, we were not able to test some associations, such as breastfeeding exclusivity and mode of delivery within the normal weight participants, in this data set. For instance, if an infant was born vaginally, s/he was also likely to be exclusively fed human milk in this study population. Antibiotic use was included in the analysis because antibiotic use has been shown to alter the g ut microbiota (140) . Strengths of this study include the enrollment of participants of low socioeconomic status , a factor which has been shown to reduce gut microbial diversity in adults (141) . Conclusion Maternal pre - pregnancy BMI is associated with the pregnancy fecal bacterial community and tends to be associated with the early infancy fecal bacterial community. Other maternal characteristics and environmental exposures were also associated with the micr obiota during early infancy. Factors such as pre - pregnancy BMI, C - section delivery and formula feeding, affect the infant microbiota and have also been shown to increase the risk of developing adverse health outcomes such as childhood overweight/obesity (8) and allergies/asthma (142 144) . The health effects associated with these factors may partially be explained by their effects on the microbiota. Other characteristics of the infant microbiome, such as species, strain, or functional differences, are important aspects of microbiota - host interactions and may give further insight on the roles the microbiota plays in ch ildhood development. Future work will determine if these bacterial differences persist as the child ages as well as describe associations between the microbiota and later health outcomes. 54 CHAPTER 4: HUMAN MILK FEEDING PATTERNS AT SIX MONTHS OF AGE ARE THE MAJOR DETERMINANTS OF FECAL BACTERIAL DIVERSITY IN INFANTS Abstract Background: Maternal pre - pregnancy obesity and breast - feeding have been associated with altered infant gut microbiota. We hypothesized that the gut microbiota of infants at 6 months o f age is influenced by pre - pregnancy obesity and by infant diet. Research Aim: The aim of this study was to examine the correlates of infant gut microbiota composition at 6 months of age . Methods: Fecal samples and participant information were collected from a subset of dyads enrolled in two related prospective cohorts (ARCH GUT and BABY GUT ) in Michigan. Sequencing the V4 region of the 16S gene was used to analyze fecal bacterial samples collected from six - month - old infants (n=36). Infants were grouped into four categories by extent of human milk exposure (100%, 80%, 50 - in the infant diet) and by maternal pre - pregnancy BMI category (normal weight, overweight, obese). Results: Fewer women with pre - pregnancy obesity were breastfeeding at 6 months post - partum compared to non - obese women (35.7% and 81.8%, respectively). In univariate analyses, maternal pre - pregnancy BMI and human milk exposure were both significantly associated with alpha and beta diversity of the infant microbiota. However, in multivariate analyses, human milk exposure accounted for some 20% of variation in alpha diversity, but pre - pregnancy BMI was not significantly associated with any form of microbiota diversity. Conclusions: The proportion of the infant diet that was human milk at six months was the major determinant of alpha and beta diversity of the in fant. Maternal obesity contributes to the gut microbiota by its association with extent of breast feeding. 55 Introduction The infant gut microbiome experiences rapid change and development for the first two to three years of age, before eventually stabilizin g at two to three years of age (31,56 59) . There are many potential routes of microbial exposure such as the environment (60,61) and infant diet (31,61) . However, maternal sources of bacteria, such as those found in breastmilk and vaginal communities, may convey most of the bacteria found in the infant gut during this early development al stage (58,60) . From birth to 2 - 3 years of age, the gut changes rapidly in bacterial abundance, composition and gene function (76) . One of the strongest influences on the gut during this time is breastfeed until six months of age due to the protective effects of human milk on the development of respiratory and gastrointestinal infections, atopy, obesity and many chronic diseases (145) . Exclusive breastfeeding has also been associated with lower bacterial diversity (82) , increased Bifidobacterium abundance and decreased abundance of Lachnospiraceae (82,83) compared to partial or no breastfeeding. Infants who are not exclusively breastfed have higher abundances of Bacteroides and Megasphaera and tend to have a microbiota profile with a predict ed age that is older than their actual age (84) . However, the amount of breastmilk an infant consumes in infancy is closely related to maternal BMI. In the US, women with obesity typically have lower rates of breastfeeding initiation a nd a shorter duration of any/exclusive breastfeeding compared to women with lower (92,146) , possibly due to higher rates of delayed lactogenesis in obese women (147) . High maternal pre - pregnancy BMI has also been associated with both an increased risk of developing childhood overweight/obesity and an altered infant gut microbiota even after 56 accounting for breastmilk in the infant diet (21,22) . Alterations to the gut microbiota during risk of obesity. Some studies have suggested that perturbations to the microbiota early in life (3 - 4 months of age) is associated with increased risk of infant overweight later in life (12 months) (21,82) . Additionally, there is some evidence of microbial transfer from mother to child (90,91) , which may act as a direct link between the increased risk of childhood obesity in infants born to obese women and the obesity - related gut microbiota. Though obese women have been shown to have lower rates of breastfeeding duration and intensity (percent of breast milk vs. in fant formula) (92) , few studies have assessed the relationship of both of these variables (i.e. maternal obesity and exposure to breast milk) in t he context of the developing infant microbiota (21,22,93) . Thus, there is a need for more studies to analyze the relationships between maternal pre - pregnancy BMI , human milk exposure and the ir effects on the infant microbiota simultaneously . Here we examine associations between the gut microbiota of 6 - month - old infants and maternal pre - pregnancy BMI, breastfeeding in the past week, delivery mode and other factors. We hypothesized that the gut microbiota of infants at 6 months of age is influenced by maternal characteristics such as pre - pregnancy obesity as well as by infant diet. Materials and Methods Participant enrollment, sample collection, sample processing and sequence read processing we re performed as described in chapter 2. Here, we analyze the infant 6 - month fecal data. Comparison of population characteristics was done using a chi - square test for categorical variables or Kruskal - Wallis for continuous variables. Alpha diversity was calculated for Chao1, Shannon and inverse Simpson indices . Normality of the alpha diversity was confirmed using the 57 Shapiro - Wilk test , and differences across groups were tested by ANOVA . Post - hoc comparisons of pairs of BMI categories w ere visualized using the Sorensen and Bray - Curtis dissimilarity scores and plotted with principle coordinate analysis (PCoA). Differences between group means w ere tested using PERMANOVA and differences in group variances w ere tested with PERMDISP . For both alpha and beta diversity, multivariable linear regression models were used to test the associations between maternal BMI category, human milk exposure, delivery mode, sex, shipping ti me, infant age and cohort. The unadjusted model, model 1, included only human milk exposure, model 2 added BMI category, and model 3 added shipping time, age and cohort to the model . We included the variables delivery mode, sex, shipping time, infant age a nd cohort in our models because they have previously been reported to affect the gut microbiota. Thus, we felt it important to account for the influence those variables in our analyses. Alpha diversity models were tested using a type II ANOVA. Analyses of beta diversity used PERMANOVA (adonis2 function in the vegan package) to fit the models. Individual taxa were compared within human milk exposure groups using a negative binomial model in the MASS package (129) . Rarefied count data of taxa that composed 1% abundance (on average) were analyzed. Benjamini - Hochberg was used to correct for false discovery rate. P - values less than 0.05 were considered significant. Results Subject Characteristics Delivery mode, sex, infant age and sample shipping time did not differ by maternal pre - pregnancy BMI category ( Table 12 ). Human milk exposure differed significantly across pre - pregnancy BMI categories. All infants whose mothers were in the normal weight category had at 58 least 50% human milk in their diet, while more than one - third of the infants whose mothers were overweight and two - thirds of infants whose mothers were obese consumed less than 20% human milk in their diet ( Table 12 ). There were no significant differences in characteristics between participants in ARCH GUT and BABY GUT ( Table 13 ). Table 12 . Participant characteristics of the 6 - month - old infants by maternal pre - pregnancy BMI category 1 Overall (n=36) Normal (n=11) Overweight (n=11) Obese (n=14) P - value Infant feeding, % breast milk in the past week, n (%) 2 0.02 100% 7 (19.4) 3 (27.3) 3 (27.3) 1 (7.1) 80% 9 (25) 4 (36.4) 1 (9.1) 4 (28.6) 50 - 80% 7 (19.4) 4 (36.4) 3 (27.3) 0 (0) 13 (36.1) 0 (0) 4 (36.4) 9 (64.3) Vaginal delivery, n (%) 2 23 (63.9) 9 (81.8) 5 (45.5) 9 (64.3) 0.21 Girls, n (%) 2 12 (33.3) 3 (27.3) 3 (27.3) 6 (42.9) 0.63 Age at stool sample collection, days (mean ± SD) 3 199.8 ±23.3 202 ±25.2 202.9 ±30.4 195.6 ±15.1 0.68 Shipping, days (mean ± SD) 3 3.8 ±1.9 3.7 ±1.7 3.9 ±1.8 3.9 ±2.2 0.96 1 BMI categories were 2 Proportions tested using Chi - squared test 3 Medians tested using Kruskal - Wallis 59 Table 13 . Participant characteristics of the 6 - month - old infants by cohort ARCHGUT (n=21) BABYGUT (n=15) P - value Infant feeding, % human milk in the past week, n (%) 2 0.35 100% 3 (14.3) 4 (26.7) 80% 4 (19.0) 5 (33.3) 50 - 80% 4 (19.0) 3 (20.0) 10 (47.6) 3 (20.0) Maternal BMI category 1,2 0.41 Normal 5 (23.8) 6 (40.0) Overweight 6 (28.6) 5 (33.3) Obese 10 (47.6) 4 (26.7) Vaginal delivery, n (%) 2 15 (71.4) 8 (53.3) 0.44 Girls, n (%) 2 9 (42.9) 3 (20.0) 0.28 Age at stool sample collection, days (mean ± SD) 3 197.4 ± 14.6 203.1 ± 32.1 0.94 Shipping, days (mean ± SD) 3 4.1 ± 2.0 3.4 ± 1.8 0.33 1 2 Proportions tested using Chi - squared test 3 Medians tested using Kruskal - Wallis 60 Alpha Diversity Infants whose mothers were obese prior to becoming pregnant had greater microbiota with normal weight mothers ( Table 14 ). There were no differences in richness according to the Chao1 index ( Table 14 ). Alpha dive significantly higher than those of infants who received human milk exclusively, while the other exposure groups were similar to the infants receiving exclusively human milk ( Table 14 ). Sex, delivery mode, age and shipping were not associated with gut microbiota alpha diversity of bacterial communities in fecal samples from 6 - month - old infants. 61 Table 14 . Alpha diversity of the fecal bacterial community of 6 - month - old infants by maternal pre - pregnancy BMI category or human milk exposure 1,2 BMI category 3 Human milk 4 Normal Overweight Obese P - value 100% 80% 50 - 80% P - value Chao1 36.4 ± 13.9 42.8 ± 12.8 45.2 ± 14.2 0.28 33.2 ± 9.2 b 39.4 ± 16.3 ab 36.3 ± 8.4 ab 50.9 ± 12.1 a 0.01 Shannon 1.4 ± 0.4 b 1.7 ± 0.7 ab 2.1 ± 0.4 a 0.003 1.3 ± 0.7 b 1.7 ± 0.3 b 1.5 ± 0.2 b 2.3 ± 0.3 a <0.001 Inverse Simpson 3.0 ± 1.1 b 4.8 ± 2.7 ab 6.4 ± 2 a 0.001 3.2 ± 2.1 b 4.2 ± 1.2 b 3.0 ± 0.6 b 7.2 ± 2.1 a <0.001 1 Values reported as mean ± SD; differences in means tested with ANOVA and a Tukey's HSD for pairwise comparisons 2 Values in a row that do not contain the same superscript are significantly different, P <0.05 3 4 % breast milk in the past week 62 Since human milk exposure was associated with BMI, we tested these variables together along with the variables: delivery mode, sex, age, sample shipping time and cohort using the models described in the methods. Human milk exposure explained the most var iance within the Chao1, Shannon and inverse Simpson scores across all models ( Table 15 ). Effect sizes are presented in Table 16 human milk had higher Chao1, Shannon and inverse Simpson scores compared to infants who received exclusively human milk, while the infants in the other milk categories had similar alpha diversity scores ( Table 14 & Table 15 ). Delivery mode was the only other variable significantly associated with any of the alpha diversity scores ( Table 15 ) infants who were born via C - section had a lower inverse Simpson index than infants born vaginally ( = - 1.5). Table 15 . Multivariate analysis of the alpha diversity of the 6 - month - old infant microbiota Chao1 Shannon Inverse Simpson R 2 P - value R 2 P - value R 2 P - value Model 1 Human milk 0.28 0.01 0.54 <0.001 0.57 <0.001 Model 2 Human milk 0.23 0.04 0.38 0.001 0.38 0.001 BMI category 0.01 0.77 0.03 0.42 0.04 0.36 Overall 0.29 0.04 0.57 <0.001 0.6 <0.001 Model 3 Human milk 0.19 0.11 0.31 0.006 0.34 0.002 BMI category 0.02 0.75 0.05 0.3 0.07 0.16 Delivery mode 0.02 0.42 0.07 0.07 0.11 0.02 Sex <0.001 0.92 <0.001 0.89 <0.001 0.84 Age 0.02 0.39 0.02 0.33 0.02 0.34 Shipping 0.005 0.67 0.004 0.65 0.01 0.39 Cohort <0.001 0.99 0.01 0.42 <0.001 0.85 Overall 0.33 0.27 0.65 0.001 0.69 <0.001 63 Table 16 . Beta values of the 6 - month - old infant microbiota alpha diversity by human milk exposure levels 1 in each model 2,3 Chao1 Model 1 Model 2 Model 3 100% Ref Ref Ref 80% 6.1 7.2 4.7 50 - 80% 3.1 2.9 0.7 17.7** 18.9** 16.9* Shannon 100% Ref Ref Ref 80% 0.4* 0.4 0.3 50 - 80% 0.2 0.23 0.1 1.0** 0.9** 0.8** Inverse Simpson 100% Ref Ref Ref 80% 1 0.9 0.7 50 - 80% - 0.2 - 0.1 - 0.6 4.0** 3.4** 3.1** 1 % breast milk in the past week 2 Values reported are the effect size of human milk exposure relative to 100% exposure 3 Multivariable linear regression was used to test the significance of each of the following models: Model 1: Human milk Model 2: Human milk and BMI category Model 3 : Human milk, BMI category, delivery mode, sex, infant age, sample shipping time and cohort * P - value <0.05 ** P - value <0.01 64 Beta Diversity In the univariate analyses, BMI category was not associated with microbiota richness ( Figure 9 A; Sorensen: p=0.20), but was significantly associated with gut community structure at the genus ( Figure 9 B; Bray - Curtis: p=0.01) and phylum levels ( Figure 9 C; Bray - Curtis: p=0.005). Human milk exposure was associated with the bacteria present in the infant gut ( Figure 9 ; Sorensen: p=0.001) and the bacterial community structure at the genus ( Figure 9 E; Bray - Curtis: p=0.03) and phylum levels ( Figure 9 F; Bray - Curtis: p=0.03). Delivery mode, sex, age, sample shipping time and study cohort were not significantly associated with the beta diversity measures in the univariate analysis (data not shown). 65 Figure 9 . Maternal pre - pregnancy BMI and level of human milk in the diet are associated with differences in the infant microbiota. PCoA of the genus - level Sorensen (A/D) and Bray - Curtis (B/E) dissimilarities as well as the phylum Bray - Curtis (C/F) dissimilarities. PC1 and PC2 are plotted on the x - and y - axis of each pl ot. The percentage of variation explained by each axis is indicated on each axis. Ellipses are around the mean location of each maternal pre - pregnancy BMI category (A/B/C) or human milk group (D/E/F). Each label is placed at the center of its corresponding ellipse. 66 At the genus level, human milk exposure alone was sufficient to explain differences in the types of distinct bacteria present in the infant gut at 6 months ( Table 17 ). Similar results were seen in the community structure using Bray - Curtis dissimilarity, human milk exposure explained the most variance (~10%) at the genus level across all models ( Table 17 ). In model 2, though the overall model was significant, no single variable included in the model reached statistical significance. Similar results were ob served at the phylum level model 2 had significant overall P - values, but none of the individual variables reached statistical significance ( Table 17 ). Table 17 . Multivariate PERMANOVA analysis of the individual variable marginal effects in the beta diversity of the 6 - month - old infant microbiota Sorensen (Genus) Bray - Curtis (Genus) Bray - Curtis (Phylum) R 2 P - value R 2 P - value R 2 P - value Model 1 Human milk 0.19 <0.001 0.13 0.03 0.16 0.03 Model 2 Human milk 0.16 <0.001 0.1 0.17 0.09 0.35 BMI category 0.04 0.72 0.07 0.17 0.09 0.11 Overall 0.23 <0.001 0.2 0.02 0.23 0.03 Model 3 Human milk 0.14 0.002 0.1 0.2 0.08 0.42 BMI category 0.05 0.46 0.08 0.11 0.09 0.1 Delivery mode 0.04 0.14 0.04 0.18 0.03 0.33 Sex 0.03 0.36 0.03 0.26 0.06 0.07 Age 0.01 0.91 0.01 0.94 0 1 Shipping 0.04 0.08 0.02 0.77 0.01 0.71 Cohort 0.02 0.48 0.02 0.58 0.05 0.15 Overall 0.38 <0.001 0.31 0.18 0.35 0.1 67 Individual Taxa Human Milk Exposure at 6 months of Age ( Tabl e 18 ) T he infants who were exclusively breastfed had lower Megasphaera , unclassified Lachnospiraceae and Blautia. For Akkermansia, Veillonella and Parabacteroides , there were significant difference s in bacterial abundances across breastfeeding groups, but no clear dose - response relationship, suggesting that the differences may be due to another variable or due to chance variation. At the family level, Ruminococcaceae, Porphyromonadaceae and Acidamin ococcaceae abundances were significantly higher in the low milk group compared to all other groups. Verrucomicrobaceae displayed no dose - response relationship, and Veillonellaceae abundance was significantly lower in the exclusively breastfed infants compa red to all other groups. At the phylum level, Firmicutes abundance was lower in exclusively breastfed infants compared to all other milk groups. Verrucomicrobia abundances were significantly higher in the 50 - 80% group compared to all other group and second highest in the - response relationship with decreasing amount of human milk exposure. 68 Tabl e 18 . Genera/phyla abundances in the 6 - month - old infant microbiota classified by human milk exposure 1,2,3,4 Genus 100% 80% 50 - 80% P - value Megasphaera 0.04 ± 0.1 d 4.9 ± 9.6 a 2.8 ± 7.2 c 4.5 ± 6.8 b <0.001 unclassified Lachnospiraceae 0.03 ± 0.1 c 0.2 ± 0.3 b 0.3 ± 0.4 b 3.9 ± 4.4 a <0.001 Blautia 0.006 ± 0.01 c 0.9 ± 2.2 ab 0.1 ± 0.3 b 3.8 ± 4.5 a <0.001 Akkermansia 1.5 ± 3.8 c 0.009 ± 0.02 d 9.0 ± 22.7 a 3.3 ± 6.9 b <0.001 Bifidobacterium 35.3 ± 24.6 26.4 ± 19.1 25.2 ± 21.8 13.3 ± 10.8 0.41 Veillonella 1.9 ± 2.7 c 16.7 ± 14.0 a 13.7 ± 18.3 a 3.8 ± 3.9 b 0.03 Parabacteroides 2.3 ± 4.4 b 1.2 ± 2.4 c 0.4 ± 0.8 d 4.1 ± 6.0 a <0.001 unclassified Enterobacteriaceae 14.6 ± 36.8 5.9 ± 9.0 8.3 ± 21.1 2.5 ± 3.3 0.34 Bacteroides 21.4 ± 23.3 14.4 ± 14.6 7.5 ± 16.7 13.4 ± 10.6 0.89 Escherichia/Shigella 12.8 ± 8.5 15.8 ± 14.1 15.6 ± 15.5 10.9 ± 11.7 0.91 Lachnospiraceae_ge 3.1 ± 8.1 4.2 ± 5.6 2.6 ± 5.7 5.7 ± 3.0 0.89 Enterococcus 1.5 ± 2.4 1.1 ± 1.7 1.3 ± 2.7 0.5 ± 0.9 0.67 Klebsiella 0.5 ± 0.8 0.1 ± 0.3 6.3 ± 11.9 3.7 ± 7.5 0.13 Family 100% 80% 50 - 80% P - value Ruminococcaceae 0.1 ± 0.2 b 0.6 ± 0.7 b 0.5 ± 0.5 b 4.2 ± 5.8 a <0.001 Porphyromonadaceae 2.3 ± 4.4 b 1.2 ± 2.4 c 0.4 ± 0.8 d 4.2 ± 6.0 a <0.001 Acidaminococcaceae 0.1 ± 0.3 b 0.1 ± 0.3 b 0.0 ± 0.0 c 2.7 ± 6.2 a <0.001 Verrucomicrobiaceae 1.5 ± 3.8 c 0.0 ± 0.0 d 9.0 ± 22.7 a 3.3 ± 6.9 b <0.001 Veillonellaceae 2.2 ± 2.8 b 22.0 ± 13.7 a 16.7 ± 20.7 a 12.0 ± 13.3 a 0.04 Lachnospiraceae 3.5 ± 8.0 6.7 ± 9.3 3.8 ± 6.2 18.2 ± 8.5 0.05 Streptococcaceae 0.5 ± 0.9 2.0 ± 3.0 0.4 ± 0.2 0.9 ± 1.2 0.08 Enterobacteriaceae 27.9 ± 31.5 22.7 ± 16.5 30.2 ± 25.5 17.7 ± 18.6 0.73 Bacteroidaceae 21.4 ± 23.3 14.4 ± 14.6 7.5 ± 16.7 13.4 ± 10.6 0.79 Bifidobacteriaceae 35.4 ± 24.6 26.5 ± 19.2 25.3 ± 21.8 13.3 ± 10.8 0.42 Coriobacteriaceae 0.5 ± 1.1 0.6 ± 1.2 2.1 ± 5.2 1.3 ± 2.3 0.46 Enterococcaceae 1.5 ± 2.4 1.1 ± 1.8 1.4 ± 2.8 0.5 ± 0.9 0.61 Phylum 100% 80% 50 - 80% P - value Firmicutes 8.5 ± 9.4 b 33.9 ± 17.9 a 25.1 ± 19.8 a 41.7 ± 14.6 a <0.001 Proteobacteria 29.1 ± 30.9 23.2 ± 16.7 30.3 ± 25.5 21.5 ± 19.9 0.81 Bacteroidetes 24.3 ± 23.6 15.6 ± 16.0 8.2 ± 17.1 18.9 ± 13.6 0.81 Actinobacteria 36.2 ± 24.4 27.3 ± 19.1 27.4 ± 26.6 14.6 ± 11.1 0.35 Verrucomicrobia 1.5 ± 3.8 c 0.01 ± 0.02 d 9.0 ± 22.7 a 3.3 ± 6.9 b <0.001 1 Values reported as mean % abundance ± SD; negative binomial regression on the bacterial count data 2 Values in a row that do not contain the same superscript are significantly different, P <0.05 3 P - values were false discovery rate corrected using the Benjamini - Hochberg method 4 % breast milk in the past week 69 Discussion In this sample of six - month - old infants from two pregnancy cohorts in Michigan, univariate analyses revealed maternal pre - pregnancy BMI and human milk exposure were significantly associated with alpha and beta diversity measures of the infant microbiota, but multivariate analyses showed that extent of human milk exposure (assessed in four categories in the week before sam pling), and not pre - pregnancy BMI, was the key correlate of microbial diversity. Additional exposures such as delivery mode, sex, infant age, sample shipping time and potential differences in the two study cohorts were also examined. In these adjusted mode ls, alpha diversity (Chao1, Shannon) was strongly influenced by human milk exposure, and delivery mode was also associated with differences in the inverse Simpson index, but maternal BMI no longer contributed. Comparing the beta diversity of the gut bacter ial community structures of the infants, human milk exposure alone explained differences in community membership (Sorensen) and community structure (Bray - Curtis) at the genus and phylum levels. When comparing the abundances of different taxa by human milk exposure, genera such as unclassified Lachnospiraceae and Blautia increased as human milk exposure decreased, while other taxa such as Akkermansia and Veillonella differed between milk exposure groups, but not linearly. At the phylum level, Firmicutes abundance was significantly lower in infants in the exclusive human milk exposure group. Generally, differences in the gut microbiotas of infants born by different de livery modes decrease by 6 months of age, and gut microbiotas are similar between these groups (15,72) . We found that some differences persisted to 6 months in our infants, with C - section infants having a 70 significantly lower inverse Simpson index than vaginally born infants, but only in the multivariate mode ls. High maternal pre - pregnancy BMI is associated with an increased risk of developing childhood overweight/obesity and an altered infant gut microbiota (21,22) and is directly related to other microbiota - influencing factors such as human milk exposure (92,146) . A handful of studies have examined the relationship between maternal pre - pregnancy BMI and infant microbiota (21,22,128,148,149) , only a few of which addressed infants aged 6 months (21,22) , and the findings of these studies are contradic tory. One study found that 3 - 4 - month old infants born to overweight/obese women had microbiota of higher richness and different community membership (unweighted UNIFRAC) than infants of normal weight women, controlling for breast feeding (exclusive/partial /none) (21) . But another study found no difference in alpha diversity at several time points in the first two years (4, 10, 30, 120, 365 and 730 days old) by maternal BMI (22) , adjusting for exclusive breastfeeding at the time of sample collection. We found that the diversity in the infant gut measured by inverse Simpson and Shannon indices increased with increasing maternal BMI, but this relationship disappeared after adjusting for the effects of human milk exposure. High gut diversity at 3 - 4 months of age has been associated with increased risk of chi ldhood overweight by 12 months of age, while the 12 - month microbiota was not associated with overweight at 12 months (82) . This suggests that the timing of infant exposure to a more diverse and mature microbial community is important for long - term heal th despite short - term microbiota alterations not persisting into later childhood. Human milk exposure was the main variable associated with gut microbiota composition in our sample of 6 - month - old infants. Differences in human milk exposure appear to be res ponsible for the significant findings we observed in the univariate analyses of maternal pre - pregnancy BMI, since 71 maternal BMI was no longer significant in multivariate models that included maternal BMI and human milk exposure. Alpha diversity effect size for human milk exposure was consistent across all models tested and the magnitude of the effect for the lowest human milk exposure group compared to exclusive breastfeeding was much higher than those found between maternal pre - pregnancy BMI category compar isons. The effect breastmilk has on the infant gut microbiota is strong whereas the effects of maternal pre - pregnancy BMI on the infant gut microbiota at 6 months, while potentially still important, are much weaker, which could lead to infant milk exposur e masking the effects that pre - pregnancy BMI have on the gut microbiota. Infants receiving less than 100% human milk had a higher abundance of unclassified Lachnospiraceae compared to all other milk exposure groups. Enrichment in Lachnospiraceae in the ea rly gut microbiota has been associated with an increased risk of overweight/obesity later in life (21,82) . This relationship could be due to increased production of short chain fatty acids (SCFA) by bacteria like Lachnospiraceae or other bacteria that are also in the phylum Firmicutes (14,150,1 51) . Though the abundance of Lachnospiraceae was not significantly different between milk groups after false discovery - rate correction, the infants in the low human milk exposure group had a numerically higher abundance of this family compared to the ot her milk groups. Our population of infants who were not in the 100% human milk exposure group had higher levels of Firmicutes. High relative abundances of this family in the gut of infants could contribute to an increased risk of developing overweight/obes ity during childhood due to increased butyrate and other SCFA production in the gut (152) . 72 Limitations There are several strengths and limitations to this study. One limitation of this study is sample collection method. Infant stool was collected by the mother at home and shipped to the lab without using a preservation method for an average of four days before storage at - 80 ° C. These conditions may affect the abundances of the microbiota (100 102) , but community differences across samples are preserved regardless of storage method (101,103,104) . Another limitation is the lack of infant complimentary feeding data in this analys is. Others have reported an association between a reduction in Bifidobacteriaceae and enrichment in Lachnospiraceae in infants who had diets high in meat, cheese and animal fats (153) . Components of complimentary foods may explain some of the associations between breastfeeding intensity and these bacterial taxa. However, there are a limited number of studies that have investigated this (154) . The small sample size also limits the conclusions that can be drawn from this study . Variables with a large effect on the infant gut microbiota such as human milk exposure could mask effects that other variables may have. A larger sample size would allow for a more nuanced description of the infant gut assembly. Conclusion This study included categorization of recent human milk exposure. Using a previously published survey method (26) , we collected a more nuanced estimate of human milk exposure to elucidate how differing levels of exposure might change the infant microbiota in the context of maternal BMI. However, our metric of human milk exposure lacked a sufficient sample size to include all seven levels in this analysis, and there were significant associations between maternal pre - pregnancy BMI and human milk in the infant diet, making detailed comparisons by 73 stratification on BMI groups impossible. Our population of 6 - month - old infant fecal bacterial communities was strongly influenced by recent human milk exposure, suggesting that the extent of recent human milk exposure must be controlled thoroughly in order to investigate the nuanced impact that other covariates may have on the infant gut microbiome. 74 CHAPTER 5 : EFFECT OF INFANT EXPOSURES ON THE GUT MICROBIOTA FROM EARLY INFANCY TO 2 YEARS OF AGE Abstract Background : There is much unknown about how the infant gut develops and how various early exposures a ffect the normal course of gut maturation over the first two years of life. Here we investigate how the infant gut matures over time and which exposures such as human milk and antibiotic exposure alter the development of the infant gut microbiota. Methods : We used the LonGP program to create predictive models using our longitudinal data to determine the contribution of various exposures on the abundances of the top ten most abundant infant gut bacteria . Results : As expected, age and participant explained most of the variations in abundance in our dataset. Infant antibiotic exposure at the time of sampling was important for 13 genera, antibiotic exposure ever in 6, breastfeeding in 7, maternal pre - pregnancy BMI in 7, sex in 1 and sample shipping time was important in 1 genus. La chnospiraceae unclassified and Bacteroides abundances , were partially predicted by human milk exposure and antibiotic exposure at the time of sampling . Independent of age, h igh levels of human milk exposure (>50% in the infant diet) w ere associated with much lower predicted levels of Lachnospiraceae unclassified, and no antibiotic exposure at the sampling time was associated with higher abundances of this taxa. Conclusion : nique microbiota composition and their age predicted most of the bacterial abundances during this critical developmental period , however human milk and antibiotic exposure helped predict Lachnospiraceae and Bacteroides abundances. High Lachnospiraceae abundance in the early infant gut microbiota has been associated with increased risk of overweight later in childhood. 75 Identifying the factors that alter the gut microbiota and lead to adverse health outcomes may help select potential bacte rial targets for microbiota modification in the developing infant gut. Introduction There is much unknown about how the infant gut develops and how various early exposures a ffect the normal course of gut community maturation , which occurs during the first 2 - 3 years of life (15,72 ) . G ut microbiota development is associated with high early abundances of Bifidobacterium, Bacteroides and Escherichia , which are gradually replaced by other bacteria, notably members of the Firmicutes phylum such as Clostridiaceae and Lachnospiraceae (155) . During this period of microbiota maturation, there is high interindividual variability in gut composition before the microbiota compo sition converges to an adult - like community (156) . During this period of high microbiota instability, changes in gut composition are strongly linked to early - life exposures such as C - section delivery, formula feeding and antibiotic exposure (66) . These exposures have also been associated with early gut maturation (i.e., higher abundance of Firmicutes) and adverse e ffect s on immune system development (157) and increased obesity risk in children (21) . Of these exposures, infant feeding, specifically human milk in the infant diet, has strong, long - term associations with the gut microbiota (15) and lowers infant risk of developing obesity (82) . Human milk exposure is associated with inc reased Bifidobacterium abundance and decreased abundance of the Firmicute Lachnospiraceae (82,83) . F ormula - fed infants tend to be enriched in Bacteroides, Escherichia, Enterobacteriaceae, Clostridium (15,79) and other bacteria associated with a more mature microbiota (84) . Understanding what these bacterial differences are and how they are altered in response to changes in diet and other gut - perturbing events are key in understanding which taxa are most 76 affected during th is developmental window and may lead to a better understanding of what a health y gut community should look like as it matures. Here, we use longitudinal fecal microbiota data collected from children from about 1 month to 2 years of ag e to describe how the infant gut changes over time and how infant exposures such as antibiotics or hum an milk affect changes in taxa abundance. Materials and Methods This analysis include d fecal microbiota samples from infants at 1 month (n=43), 6 months (n=41), 12 months (n=4 0 ) and 24 months (n= 39 ) of age . In total, there were n=3 4 participants with a sample at each timepoint. For this analysis, we used the methods and a MatLab ( release R2019b) package (LonGP) by Cheng et al. to build additive Gaussian process regression models and determine which variables were important for predicting changes to the infant gut microbiota (158) . Briefly, an additi ve Gaussian process model assumes that our data follows a multivariate normal distribution composed of D - dimensional variables that can be described by a D - dimensional function (159) . This high dimensional function can then be broken into lower dimensional functions that describe the contribution each variable has when fitting a model to the data and has t he same mathematical form as a regular multivariate regression (159) ( Equation 1 ) . For this analysis, we h ad 10 variables of interest: participant ( ID ), infant age ( age ), sample shipping time ( ship ), breastfeeding status ( bf ), maternal pre - pregnancy BMI ( BMI ), antibiotic exposure at the time of sampling ( abx1 ), any antibiotic exposure since birth ( abx2 ), sex, delivery mode ( rout ) and cohort. Since we expect ed age and participant to have the largest effect on the microbiota, the variables age and participant w ere included in e ach model. Additionally, ag e, participant , maternal pre - pregnancy BMI and breastfeeding status were 77 allowed to form interaction terms with any other variable in our models . T he fecal microbiota data of the infants at 1 month, 6 months, 12 months and 24 months was used for this analysis. The bacterial counts were log transformed for the analysis and then the scale was back transformed for interpretation of the data . Only the top 50 most abundant taxa were included. Continuous covariates and target variables were centered to the mean to help generate appropriate priors for each function. Continuous variables were then back transformed to their original scale for inter pretation of the results. The most complicated model tested with this method was : Equation 1 . Gaussian process regression model that includes all variables and possible interactions ca= categorical, co=continuous, bi=binary Where each variable or interaction was described by its own function. In this analysis, breastfeeding was transformed from our seven - points scale to 100%, >50 - 80%, 20 - 50% and 0% ( no breastmilk ) to accommodate decreasing consumption of human milk as the infant grows older. Model inference and prediction was done in LonGP by the leave - one - out cross - validation method and the stratified cross - validation method. The first method simulates the predictive power of each model by leaving out a single timepoint of one individual and using the rest of the data as training data for the models, while the latter does the same process but leaves out all 78 timepoint of an ind ividual (158) . Once the predictive densities of each model are calculated, the difference of the average prediction accuracy (calculated using a Bayesian bootstrap) is used to determine whether one model is significantly better than another, with a difference of 0 indicating that the models being compared have equal predictive accuracy. Results Population characteristics for all infant samples are shown in Table 19 . There were no significant differences between the two cohorts. C haracteristics for the participants with a sample at e very time point are shown in Table 20 . 79 Table 19 . Population characteristics of the infants at 1, 6, 12 and 24 months 1 Month 6 Months 12 Months 24 Months ARCH GUT BABY GUT p - val ARCH GUT BABY GUT p - val ARCH GUT BABY GUT p - val ARCH GUT BABY GUT p - val n 27 16 25 16 24 16 22 17 Vaginal Delivery 1 18 (41.9) 10 (23.3) 1 17 (41.5) 9 (22) 0.67 16 (40.0) 10 (25.0) 1 15 (38.5) 10 (25.6) 0.79 Girls 1 12 (27.9) 3 (7.0) 0.17 10 (24.4) 3 (7.3) 0.28 9 (22.5) 3 (7.5) 0.36 7 (17.9) 3 (7.7) 0.53 Infant breastmil k 1 0.28 0.47 0.12 0.35 100 15 (34.9) 13 (30.2) 3 (7.3) 4 (9.8) 0 (0) 0 (0) 0 (0) 0 (0) >50 - 80 4 (9.3) 2 (4.7) 11 (26.8) 8 (19.5) 0 (0) 2 (5) 0 (0) 0 (0) 20 - 50 6 (14) 1 (2.3) 1 (2.4) 1 (2.4) 9 (22.5) 3 (7.5) 1 (2.6) 2 (5.1) 0 2 (4.7) 0 (0) 10 (24.4) 3 (7.3) 15 (37.5) 11 (27.5) 21 (53.8) 14 (35.9) Missing 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1 (2.6) Antibiotic exposure currently 1 0 (0) 2 (4.7) 0.26 0 (0) 0 (0) 1 2 (5.0) 1 (2.5) 1 0 (0) 0 (0) 1 Antibiotic exposure ever 1 1 (2.3) 1 (2.3) 1 1 (2.4) 2 (4.9) 0.69 8 (20.0) 4 (10.0) 0.83 9 (23.1) 4 (10.3) 0.42 Maternal pre - pregnanc y BMI 2 29.4 ± 5.4 27.2 ± 5.9 0.24 29.6 ± 5.5 26.9 ± 6.0 0.15 29.5 ± 5.5 27.2 ± 5.9 0.21 29.8 ± 5.7 26.9 ± 5.8 0.14 Infant age (days) 2 38.5 ± 32.0 36.7 ± 33.7 0.86 204.9 ± 15.7 208.6 ± 32.4 0.68 385.5 ± 17.4 386.2 ± 22.0 0.9 749.6 ± 21.7 767.1 ± 49.0 0.19 Infant age (days) 3 26 (6, 116) 21 (9, 125) 0.83 204.0 (179, 239) 198.5 (163, 295) 0.7 383 (356, 432) 379 (364, 438) 0.55 746 (712, 800) 746 (723, 916) 0.51 80 Sample shipping time (days) 2 4.7 ± 2.8 5.5 ± 3.9 0.48 5.6 ± 4.4 3.6 ± 1.9 0.05 4.4 ± 2.6 5.2 ± 7.1 0.66 4.4 ± 3.6 4.6 ± 3.8 0.88 Sample shipping time (days) 3 4 (1, 14) 4 (1, 15) 0.69 5 (1, 22) 3 (0, 7) 0.14 4 (1, 11) 3.5 (2, 31) 0.55 3 (1, 18) 4 (0, 14) 0.95 1 n (%) 2 mean ± SD 3 median (min, max) 81 Table 20 . Population characteristics of infants with all four samples 1 Month 6 Months 12 Months 24 Months n 34 34 34 34 Vaginal Delivery 1 22 (64.7) Girls 10 (29.4) Infant breastmilk 100 22 (64.7) 6 (17.6) 0 (0) 0 (0) >50 - 80 6 (17.6) 16 (47.1) 2 (5.9) 0 (0) 20 - 50 6 (17.6) 2 (5.9) 11 (32.4) 2 (5.9) 0 0 (0) 10 (29.4) 21 (61.8) 31 (91.2) Missing 0 (0) 0 (0) 0 (0) 1 (2.9) Antibiotic exposure currently 1 1 (2.9) 0 (0) 3 (8.8) 0 (0) Antibiotic exposure ever 1 2 (5.9) 3 (8.8) 11 (32.4) 11 (32.4) Maternal pre - pregnancy BMI 2 28.9 ± 5.9 Infant age 2 39.0 ± 32.3 204.1 ± 23.1 386.0 ± 20.5 750.2 ± 23.0 Sample shipping time 2 5.4 ± 3.4 4.8 ± 3.8 4.7 ± 5.2 4.6 ± 3.8 1 n (%) 2 mean ± SD 82 Table 21 shows the model parameters that best explain changes in genera abundance. For most of the models, the variables age, participant and the interaction between these variables explained most of the variation found in this longitudinal dataset. Of the 50 genera included in this analysis, the mode ls for 30 of these included at least one of the other eight variables besides age and participant . Infant antibiotic exposure at the time of sampling was selected to be in the final model for 13 genera, antibiotic exposure ever was selected in 6, breastfee ding in 7 , maternal pre - pregnancy BMI in 7, s ex in 1 and sample shipping time in 1 ( Table 21 ) . The variance explained ranged from <0.01% to up to 21.1 % ( Table 21 ). Table 21 Model parameters significantly associated with changes in abundance of the top 50 most abundant taxa and the variance explained by each parameter Genera Model Parameters and Variance Explained Pyramidobac ter age abxcurrently abxever I d age*id error var_Pyramid obacter 0.00% 0.60% 21.50% 2.00% 60.70% 15.30 % Ruminococc us_2 age abxcurrently id age*id error var_Ruminoc occus_2 75.80% 1.50% 1.20% 6.30% 15.20% Prevotella_9 age abxcurrently id age*id error var_Prevotell a_9 32.40% 0.70% 37.80% 17.20% 11.90% Parasutterella age abxcurrently id age*id error var_Parasutte rella 55.40% 2.00% 1.00% 38.30% 3.30% Ruminococc aceae_ unclassified age abxcurrently id age*id error var_Ruminoc occaceae_ unclassified 60.00% 0.80% 2.00% 26.20% 11.00% Fusobacteriu m age abxcurrently id age*id error var_Fusobact erium 3.70% 0.00% 3.30% 88.10% 4.90% 83 Odoribacter age abxcurrently id age*id error var_Odoriba cter 2.90% 0.00% 3.30% 93.50% 0.30% Ruminococc aceae_ unclassified. 1 age abxcurrently id age*id error var_Ruminoc occaceae_ unclassified. 1 18.10% 0.20% 2.90% 77.00% 1.80% uncultured_g e age abxcurrently id age*id error var_uncultur ed_ge 11.00% 0.70% 2.00% 86.20% 0.20% Stenotropho monas age abxcurrently id age*id error var_Stenotro phomonas 0.10% 0.00% 15.60% 44.70% 39.50% Prevotella_7 age abxcurrently id age*id error var_Prevotell a_7 0.20% 0.00% 0.90% 95.30% 3.60% Bilophila age abxever id age*id error var_Bilophil a 9.90% 3.80% 15.30% 65.50% 5.40% Megamonas age abxever id age*id error var_Megamo nas 2.10% 0.30% 32.60% 31.60% 33.20% Ruminococc aceae_ge age abxever id age*id error var_Ruminoc occaceae_ge 37.00% 1.20% 3.60% 15.30% 42.90% Cloacibacillu s age abxever id age*id error var_Cloaciba cillus 0.80% 2.10% 0.80% 91.20% 5.00% Bacteroides age bf id age*id error var_Bacteroi des 37.10% 7.20% 9.50% 19.70% 26.60% Parabacteroi des age bf id age*id error 84 var_Parabact eroides 1.10% 2.60% 33.30% 56.50% 6.60% Staphylococc us age bf id age*id error var_Staphylo coccus 44.00% 10.10% 0.90% 42.90% 2.10% Peptostreptoc occaceae_ unclassified age bf id age*id error var_Peptostr eptococcacea e_unclassifie d 29.80% 16.60% 0.90% 51.20% 1.60% Acinetobacte r age bf id age*id error var_Acinetob acter 10.10% 21.10% 4.70% 40.00% 24.20% Erysipelotric haceae_ge age bf id age*id error var_Erysipel otrichaceae_ ge 15.70% 14.90% 1.90% 61.60% 5.90% Turicibacter age bf id age*id error var_Turiciba cter 31.10% 0.60% 3.50% 64.60% 0.20% Escherichia.S higella age id age*id error var_Escheric hia.Shigella 26.40% 1.10% 67.90% 4.70% Bifidobacteri um age id age*id error var_Bifidoba cterium 18.00% 3.90% 72.10% 6.00% Veillonella age id age*id error var_Veillone lla 15.30% 12.90% 45.70% 26.10% Faecalibacter ium age id age*id error var_Faecalib acterium 77.00% 0.30% 7.10% 15.60% 85 Clostridium_ sensu_stricto _1 age id age*id error var_Clostridi um_ sensu_stricto _1 1.00% 2.40% 88.50% 8.10% Streptococcu s age id age*id error var_Streptoc occus 10.00% 1.30% 87.60% 1.00% Akkermansia age id age*id error var_Akkerm ansia 32.90% 2.50% 41.80% 22.80% Alistipes age id age*id error var_Alistipes 48.60% 11.00% 35.70% 4.80% Collinsella age id age*id error var_Collinsel la 13.20% 43.20% 34.50% 9.10% Haemophilus age id age*id error var_Haemop hilus 13.80% 3.70% 68.00% 14.50% Pseudomona s age id age*id error var_Pseudom onas 0.20% 1.70% 93.80% 4.30% Erysipelatocl ostridium age id age*id error var_Erysipel atoclostridiu m 51.30% 0.40% 30.40% 17.80% Paraprevotell a age id age*id error var_Paraprev otella 12.00% 3.80% 75.60% 8.70% Ruminococc aceae_ UCG.014 age id age*id error var_Ruminoc occaceae_ UCG.014 23.80% 0.50% 74.00% 1.70% 86 Eggerthella age id age*id error var_Eggerthe lla 26.80% 2.00% 61.70% 9.50% Gastranaerop hilales_ge age id age*id error var_Gastrana erophilales_g e 5.00% 6.00% 84.70% 4.30% Actinomyces age id age*id error var_Actinom yces 25.10% 2.90% 55.70% 16.30% Aeromonas age id age*id error var_Aeromo nas 0.10% 1.00% 93.30% 5.70% Butyricimon as age id age*id error var_Butyrici monas 1.00% 11.50% 86.90% 0.60% Desulfovibri o age id age*id error var_Desulfov ibrio 0.00% 0.90% 95.40% 3.60% Lachnospirac eae_ unclassified age mombmi abxcurrent ly bf id age*id error var_Lachnos piraceae_ unclassified 36.90% 0.80% 2.20% 25.20% 0.30% 33.40 % 1.40% Barnesiella age mombmi id age*id error var_Barnesie lla 7.30% 1.20% 3.90% 87.20% 0.30% Christensenel laceae_ R.7_group age mombmi id age*id error var_Christen senellaceae_ R.7_group 24.30% 0.10% 2.20% 73.40% 0.00% Muribaculac eae_ge age mombmi id age*id error var_Muribac ulaceae_ge 0.20% 0.20% 4.00% 57.20% 38.30% 87 Table 21 Lactobacillus age mombmi id age*id error var_Lactobac illus 0.50% 0.00% 26.10% 38.60% 34.80% Epulopisciu m age mombmi id age*id error var_Epulopis cium 0.10% 1.80% 1.00% 79.10% 18.00% Phascolarcto bacterium age mombmi sex id age*id error var_Phascola rctobacteriu m 10.80% 0.60% 4.30% 8.10% 44.90% 31.20 % Muribaculac eae_ge.1 age ship abxcurrent ly abxever id age*id error var_Muribac ulaceae_ge.1 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% Of the top 10 most abundant taxa , Bacteroides, Escheri ch a - Shigella, Lachnospiraceae unclassified , Bifidobacterium and Veillonella had an average abundance of >5% in our pooled dataset , while the other taxa were much lower in abundance on average ( Table 22 ). Table 22 also shows that age, participant (ID) and the interaction between the two variables contribut e the most to predicting the abundances of these bacteria . In the top 10 most abundant bacteria, th e variables antibiotic exposure during fecal sampling , breastfeeding and maternal pre - pregnancy BMI were selected in the final models for some of these bacteria . Current antibiotic exposure helped explain changes in relative abundance of Lachnospiraceae unclassified of around 4%, maternal pre - pregnancy BMI explained a 2% change in relative abundance, while breastfeeding explained around 5% relative abundance of b oth Bacteroides and Lachnospiraceae unclassified. 88 Table 22 . The mean r elative abundance predict ions by each variable for the 10 most abundant taxa age id ageid bf abxcurrently mombmi Bacteroides 10.4 ± 6.2 5.6 ± 5.9 6.6 ± 7.8 4.5 ± 6.2 NA NA Escherichia.Shigella 9.1 ± 6.5 6.3 ± 6.9 13.0 ± 9.7 NA NA NA Lachnospiraceae_ unclassified 4.2 ± 6.3 2.9 ± 3.0 5.5 ± 3.5 5.1 ± 3.8 4.0 ± 5.6 2.0 ± 4.5 Bifidobacterium 4.1 ± 3.2 3.2 ± 3.6 6.8 ± 5.0 NA NA NA Veillonella 1.9 ± 1.6 1.8 ± 1.9 2.4 ± 2.8 NA NA NA Faecalibacterium 1.9 ± 0.3 0.3 ± 0.3 0.4 ± 0.4 NA NA NA Clostridium_sensu_ stricto_1 0.3 ± 0.3 0.3 ± 0.3 1.0 ± 0.5 NA NA NA Streptococcus 1.0 ± 0.9 0.9 ± 1.0 1.7 ± 1.1 NA NA NA Akkermansia 0.2 ± 0.1 0.1 ± 0.1 0.3 ± 0.2 NA NA NA Phascolarctobacterium 0.1 ± 0 0 ± 0 0.1 ± 0.1 NA NA NA Values shown as the mean ± SD predicted abundance for each variable included in the model Means calculated by averaging the predicted abundances within each variable NA values indicate the variable was not included in the final model As the infants aged, Bacteroides, Lachnospiraceae unclassified, Faecalibacterium, Akkermansia and Phascolarctobacterium abundance increased rapidly after 6 months, while Eschericia - Shigella, Bifidobacterium, Veillonella and Streptococcus decreased in abundance over time ( Figure 10 ). Clostridium _sensu_stricto_1 was predicted to have a relative abundance of less than 0.5% at all timepoints and followed a bimodal distribution with high abundance at the 1 month timepoint, a sharp decrease in abundance at 6 and 12 months, follo wed by an abundance comparable to the 1 month timepoint at 24 months. The age association for the other taxa can be found in Figure 11 . 89 Figure 10 . Predicted abundance and actual abundance over time for the top 10 most abundant taxa Each taxon is shown as the predicted relative abundance as a function of infant age. Note that t he scale of the y - axis is different for each bacterium . 90 Figure 11 . Predicted change in abundance as the infant ages for the 40 least abundant taxa 91 Figure 12 show s the predicted abundance due to antibiotic exposure at the time of sampling , human milk in the infant diet and maternal pre - pregnancy BMI have on Lachnospiraceae unclassified abundance . Lachnospiraceae unclassified abundance was higher in infants who were not exposed to antibiotics at the time of sampling ( Figure 12 A). Lachnospiraceae u nclassified was lower in infants who were receiving >50% human milk in their diet compared to infants receiving 50% or less in their diet ( Figure 12 B ) . The abundances in the 100% and >50 - 80% groups were predicted to be at around 1% abundance, while infants receiving 20 - 50% or 0% human milk were predicted to have at least 6 times the abunda nce of Lachnospiraceae unclassified. BMI increased ( Figure 12 increased past 30. Interestingly, the Bacteroides abundance was predicted to be highest in the 100% milk group, and these predicted levels were similar to those predicted levels for the 0% milk gro up ( Figure 13 ) the >50 - 80% had lower predicted abundance compared to the 100% and 0% groups, and 20 - 50% group had the lowest predicted abundance , creati ng a bimodal distribution . 92 Figure 12 . Relative abundance of Lachnospiraceae unclassified predicted by current antibiotic exposure , human milk exposure and maternal pre - pregnancy BMI 93 Figure 13 . Relative abundance of Bacteroides predicted by human milk exposure Discussion In this population of infants from 1 to 24 months of age, age, participant and the interaction between these variables explain ed most of the bacterial abundances during this critical developmental period. Infant antibiotic exposure at the time of sampling , antibiotic exposure ever, human milk exposure maternal pre - pregnancy BMI, sex and sample shipping time helped predict the abund ance s of several taxa, however most of these taxa were at very low abundance in the infant gut microbiota. Only human milk exposure , antibiotic exposure at the time of sampling , and maternal pre - pregnancy BMI were predictive of some of the more abundant ba cteria, specifically Lachnospiraceae unclassified and Bacteroides . 94 Lachnospiraceae abundance generally increases with infant age (83,160) , suggesting it is a common ta xon associated with gut maturation (161) . In this cohort, infants receiving <50% milk in their diet were predicted to have much higher abundances of Lachnospiraceae , giving these infants a more mature gut c ommunity at an earlier timepoint than infant s who are receiving >50% human milk. Another study found similar results, where infants fed >50% human milk in their diets had much lower abundances of Lachnospiraceae compared to infants receiving >50% - or soy - based formula (83) . Lachnospiraceae abundance in 3 month - old infants has also been associated with maternal BMI as well as increased risk for childhood overweight at 1 year of age (21) . We found a similar association between a high predicted abundance of Lachnospiraceae and high maternal pre - pregnancy BMI. Additionally, s tudies in mice have associated Lachnospiraceae and the promotion of increased adiposity and markers of adipose inflammation (162) . The mechanism driving th ese association s are not known , but it is hypothesized that short chain fatty acid production (mainly acetate, propionate and butyrate) by members of the Lachnospiraceae family contribute to obesity risk by upregulat i ng lipogenesis (163) and causing alterations to immune system responses (152,164) . Together, this suggests that high maternal BMI and an infant diet low in human milk influences the infant gut composition, which in turn may increase infant risk for obesity development. Lachnospiraceae abundance was lower in infants on antibiotics compar ed to infants not on antibiotics at the time of sampling . There is some evidence that only certain members of the Lachnospiraceae family decrease when the host is exposed to antibiotics, while others are not affected (165) . Lachnospiraceae recovers quickly after cessation of antibiotic treatment in infants, reaching higher abundances 1 month after treatment compared to infants of the same age who received no antibiotics (89 ) . Additionally, other studies have found Lachnospiraceae 95 blooms when exposed to a subtherapeutic antibiotic treatment in mice (166) and has a higher abundance in the microbiota of 3 - 4 month - old infants who live in households that frequently use cleaning products (167) , supporting the idea that Lachnospiraceae thrives in the post - antibiotic gut environment. The same study reported that the increase in Lachno spiraceae abundance at 2 - 3 months of age was also associated with higher BMI - for - age z - scores at 3 years of age (167) . Bacteroides are a common genus of bacteria found in the human gut from infancy through ad ulthood (168) and in the gut of many other mammals (169) . This suggests that Bacteroides is well - adapted to living in the mammalian gut and can survive on a range of different dietary substrates in cluding human milk oligosaccharides during infancy (170) and many plant - and host - derived glycans (171) . We found that Bacteroides abundance increased over time and the association with human milk ex posure had a bimodal distribution, where exclusive breastfeeding and no breastfeeding had similarly high levels of predicted Bacteroides abundance, and the 20 - 80% milk groups were both lower in predicted abundance. This could be due to decreased energy har vesting capability when exposed to mixed feeding, forcing the Bacteroides community to quickly switch between different dietary substrates and causing an overall decrease in Bacteroides abundance. There are several strengths and limitations to this study. First, the sample collection method was not ideal, since immediate freezing is the gold - standard followed by refrigeration or storage in a preservative during transit (100) . This could have affected the bacterial abundances in our samples and may explain some of our findings such as a lack of si gnificance between Bifidobacterium abundance and human milk exposure. Strengths include sample collection during a critical developmental window and a more granular look at the impact of different levels of human milk exposure. 96 Using a longitudinal cohort of infants, we determine several associations between microbi al abundances and infant exposure s . Of particular interest is the Lachnospiraceae family, which has been implicated to increase childhood BMI and obesity risk (21) and is potentially a microbial target to alter in the early life by increasing human milk in the infant diet or providing an HMO - analogue as a supplement for infants receiving <50% human milk in their diet. 97 SUMMARY DISCUSSION This research suggests that the gut microbiota of infants is influenced heavily by infant exposure to human milk, but also by maternal pre - pregnancy BMI and other variables. Some exposures, like mode of delivery, were important early in life, but the effect disappeared as the infant gut became more complex. However, for each aim , we found that human milk was the most important infant exposure associated with the inf ant gut microbial composition. In study one , we report that the microbiota communit y of women who were overweight prior to becoming pregnant differed from those of women who were normal weight and obese prior to becoming pregnant. The overweight women had a lower alpha diversity and a high abundance of Bacteroides , which was the main driv er for their overall community differences compared to the other BMI categories . B eta diversity of the infant microbiota at 1 months of age was associated with delivery mode, human milk (exclusive vs non - exclusive) and antibiotic exposure (exposed during f ecal sampling time) in univariate analyses. However, only human milk in the diet was significantly associated with the infant gut microbiota after stratifying by maternal pre - pregnancy BMI category . In study 2 , data were presented whereupon maternal pre - pregnancy BMI and human milk exposure (100%, 80%, 50 - 80% and <20%) were significantly associated with alpha and beta diversity measures of the infant microbiota, but multivariate analyses showed that the extent of human milk exposure was the main driver of microbial diversity. In models that considered delivery mode, sex, infant age, sample shipping time and cohort differences , alpha diversity (Chao1, Shannon) remained strongly influenced by human milk exposure, and delivery mode was also associated with differences in inverse Simpson index. For beta diversity, human milk was strongly associated with differences in the Sorensen metric only. 98 In study 3 , we report that age had the most influence on t he bacterial abundances over time . H uman mi lk (100%, >50 - 80%, 20 - 50% and 0%) , antibiotic exposure at the time of sampling and maternal pre - pregnancy BMI helped explain Lachnospiraceae unclassified and Bacteroides . Lachnospiraceae was much higher as infants aged, in infants receiving <50% human milk in their diet and if their mom had a pre - pregnancy BMI > 30. M aternal BMI is associated with differences in the infant gut (21) , which m ay be due to intergenerational transmittance of certain bacteria from mother to child (90) , suggesting it is possible to transfer an obesity - associated microbiota from mother to child and increase childhood risk of obesity (172) . However, if there is a link between the gut microbiota of infants and obesity risk, it is likely due to the strong association between human milk exposure and the gut microbiota . Other exposures, like mode of delivery , were important early in life, but the effect disappeared as the infant gut became more complex. Studies have also found that effects on the gut microbiota due to delivery mode tend to disappear by 6 months of age, if not sooner (15,72) . In both studies 2 and 3, unclassified Lachnospiraceae was higher in abundance in the lower milk exposure groups at 6 months and in our longitudinal data. Members of the Lachnospiraceae family may contribute to obesity risk by upregulating lipogenesis (163) , thereby increas ing infant risk for obesity development . This could be due to increased production of SCFAs by bacteria like Lachnospiraceae or other s in the phylum Firmicutes (14,150,151) . SCFAs are produced from the fermentation of non - digestible carbohydrates (16) . Non - digestible carbohydrates are generally consumed as fiber from the diet but are also in human milk as HMOs. High fiber consumption and supplementation of SCFAs in the diet have been associated with improved health in adults, but the exact role of SCFAs and their producers in the infant gut are not known . A high load of fecal acetate, butyrate and propionate is associated with higher 99 BMI in adults (173) . However, o thers have found that SCFAs may attenuate obesity development and have several other beneficial effects on metabolism and on epithelial function (174,175) . In infants, it is possible that earl y exposure to these SCFA - producers in the Lachnospiraceae family could influence immune system training and epigenetic signaling via SCFA production (176) leading to increased risk for obesity later in life. I mplications and Future Directions There ha s been a massive amount of research to describe changes in the gut microbiome in response to changes in diet, weight , exercise and many other exposures (177) and how alte rations to gut composition affect obesity, allergies and even neurodevelopment (70) . In particular, high risk of obesity later in childhood is associated with increases in Firmicutes abundance early in life (21,167) , which is directly impacted by human milk exposure and maternal pre - pregnancy BMI. Additionall y, high abundance of Firmicutes is associated with early gut maturity, which negatively impacts immune system development (70) and can impact host metabolism (178) . Exploring th is relationship between gut health and long - term health effects in infants would help develop potential preventative treatments for many common chronic diseases such as obesity, diabetes, allergies and others (179) . The ARCH study collects questionnaire information on these children at 5 and 7 years of age. The fecal microbiota dataset presented here could be used in conjunction with this additional health and lifestyle information to confirm wh ether the early alterations on the gut microbiota we report are associated with childhood health outcomes such as obesity . Future research should explore how c hanges to the gut microbiota influences SCFA - production and whether SCF As a ffect infant development. Transferring the human gut 100 microbiota into germ - free mice causes obesity if the mice are colonized using the microbiota (14) and transferring the microbiota from infants born to obese mothers into germ - free mice has shown similar increase in obesity - related symptoms (172) . Other studies have reported that a malnourishment phenotype can also be transferred to m ice via colonization of gut microbes from children with kwashiorkor (180) , suggesting that the gut microbiota can modulate host metabolism. The causal mechanism behi nd the association between metabolism and gut microbes is not known, but SCFAs and SCFA - producers like Lachnospiraceae likely play a role. Early epithelial exposure to SCFA s could cause changes to the epigen ome in the developing infant gut (181) , thereby affecting infant metabolic programming (182) and immune system function (183) . One avenue to investigate this is to measure the SCFA fecal content, gene content and map metagenomic/transcriptomic SCFA pathways early in life to describe the abundance of SCFAs made in the gut and whether certain fermentation pathways are important pred ictors of gut microbiota change and childhood health months or years later. Another research topic of interest is how to alter formula such that it can mimic the beneficial effects human milk has on infant health. Fortifying formula with HMO - analogues co uld reduce gut changes and attenuate increased abundances of the Lachnospiraceae family , potentially reducing obesity risk and helping improve immune system function. One compound that has shown promising benefits is the addition of 2 - fucosyllactose to for mula. This compound leads to microbiota compositions similar to that of breastfed infants, has no adverse effects on infant growth and is well - tolerated by infants (184) . Kameyama et al. colonized germ - free mice with a member of the Lachnospiraceae family and a non - pathogenic strain of E. coli , leading to increased adiposity and symptoms of obesity in t hese mice (185) . The authors suggest that the 101 Lachnospiraceae helped translocate E. coli LPS across the epithelial barrier, causing adverse changes to health. Similar research could be done using Lachnospiraceae, E. coli and other health - associated bacteria to measure how the host responds to human milk exposure and whether the response to formula enriched with HMO - analogues is significantly different. In sum, the early gut microbiota shapes itself in response to different dietary and environmental exposures . Understanding what influences the infant gut microbiome a nd how those changes affect health will lead to novel solutions that help prevent chronic diseases like obesity. 102 APPENDICES 103 APPENDIX A: CONSENT FORMS, SAMPLE COLLECTION INFORMATION AND QUESTIONNAIRES 104 105 106 107 BABYGUT consent form 108 109 110 111 ARCHGUT sample collection forms 112 113 114 BABYGUT sample collection forms 115 116 117 ARCHGUT enrollment, adult sampling time and infant 1 - month questionnaires 118 119 120 121 122 ARCHGUT 6/12/24 month sample questionnaire and dietary checklist 123 124 ARCHGUT 3 - year questionnaire and dietary checklist 125 126 127 128 129 130 BABYGUT enrollment, adult sampling time and infant 1 - month questionnaires 131 132 133 134 BABYGUT 6/12/24 month sample questionnaire and dietary checklist 135 136 137 138 BABYGUT 3 - year questionnaire and dietary checklist 139 140 141 142 143 BABYGUT demographic survey 144 APPENDIX B: IRB APPROVAL 145 BIBLIOGRAPHY 146 1. Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence of Obesity Among Adults and Youth: United States, 2015 2016 Key findings Data from the National Health and Nutrition Examination Survey. Natl Cent Heal Stat [Internet]. 2017;1:1 8. Available from: https://www.cdc.gov/nchs/data/databriefs/db288_table.pdf#1. 2. Tremmel M, Gerdtham UG, Nilsson PM, Saha S. Economic burden of obesity: A systematic literature review. International Journal of Environmental Research and Public Health. 2017. 3. Lynch CM, Sexton DJ, Hession M, Morrison JJ. Obesity and mode of delivery in primigravid and mult igravid women. Am J Perinatol. 2008;25:163 7. 4. Weiss JL, Malone FD, Emig D, Ball RH, Nyberg DA, Comstock CH, Saade G, Eddleman K, Carter SM, Craigo SD, et al. Obesity, obstetric complications and cesarean delivery rate - A population - based screening st udy. Am J Obstet Gynecol. Mosby Inc.; 2004;190:1091 7. 5. Ajslev TA, Andersen CS, Gamborg M, Sørensen TIA, Jess T. Childhood overweight after establishment of the gut microbiota: The role of delivery mode, pre - pregnancy weight and early administration of antibiotics. Int J Obes. 2011;35:522 9. 6. González - Jiménez E, Montero - Alonso MA, Schmidt - RioValle J, García - García CJ, Padez C. Metabolic syndrome in Spanish adolescents and its association with birth weight, breastfeeding duration, maternal smoking, a nd maternal obesity: a cross - sectional study. Eur J Nutr. Dr. Dietrich Steinkopff Verlag GmbH and Co. KG; 2015;54:589 97. 7. Negative Association between Maternal Obesity a nd Exclusive Breastfeeding at 1 and 2 Months Postpartum. J Nutr [Internet]. 2015 [cited 2019 Sep 12];145:2369 78. Available from: https://academic.oup.com/jn/article/145/10/2369/4616766 8. Blustein J, Attina T, Liu M, Ryan AM, Cox LM, Blaser MJ, Trasande L. Association of caesarean delivery with child adiposity from age 6 weeks to 15 years. Int J Obes. 2013;37:900 6. 9. Portela DS, Vieira TO, Matos SMA, de Oliveira NF, Vieira GO. Maternal obesity, environmental factors, cesarean delivery and breastfeeding as determinants of overweight and obesity in children: Results from a cohort. BMC Pregnancy Childbirth. BioMed Central Ltd.; 2015;15. 10. Whitaker RC. Predicting Preschooler Obesity at Birth: The Role of Maternal Obesity in Early Pregnancy. Pediatrics [Internet]. 2004 [cited 2019 Sep 12];114:e29 36. Available from: http://pediatrics.aappublications.org/cgi/doi/10.1542/p eds.114.1.e29 11. Freedman DS, Khan LK, Dietz WH, Srinivasan SR, Berenson GS. Relationship of childhood obesity to coronary heart disease risk factors in adulthood: The Bogalusa heart study. Pediatrics. 2001;108:712 8. 147 12. Albuquerque D, Nóbrega C, Manc o L, Padez C. The contribution of genetics and environment to obesity. British Medical Bulletin. 2017. p. 159 73. 13. Ussar S, Griffin NW, Bezy O, Fujisaka S, Vienberg S, Softic S, Deng L, Bry L, Gordon JI, Kahn CR. Interactions between gut microbiota, h ost genetics and diet modulate the predisposition to obesity and metabolic syndrome. Cell Metab. Cell Press; 2015;22:516 30. 14. Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI. An obesity - associated gut microbiome with increased capac ity for energy harvest. Nature. 2006;444:1027 31. 15. 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. Cell Press; 2015;17:690 703. 16. Yang JY, Kweon MN. The gut microbiota: A key regulator of metabolic diseases. BMB Reports. The Biochemical Society of the Republic of Korea; 2016. p. 536 41. 17. Ridaura VK, Faith JJ, Rey FE, Cheng J, Duncan AE , Kau AL, Griffin NW, Lombard V, Henrissat B, Bain JR, et al. Gut microbiota from twins discordant for obesity modulate metabolism in mice. Science (80 - ). American Association for the Advancement of Science; 2013;341. 18. Schwiertz A, Taras D, Schäfer K , Beijer S, Bos NA, Donus C, Hardt PD. Microbiota and SCFA in lean and overweight healthy subjects. Obesity. 2010;18:190 5. 19. Sze MA, Schloss PD. Looking for a signal in the noise: Revisiting obesity and the microbiome. MBio. American Society for Micro biology; 2016;7. 20. Le Roy CI, Bowyer RCE, Castillo - Fernandez JE, Pallister T, Menni C, Steves CJ, Berry SE, Spector TD, Bell JT. Dissecting the role of the gut microbiota and diet on visceral fat mass accumulation. Sci Rep [Internet]. 2019 [cited 2019 Jul 19];9:9758. Available from: http://www.nature.com/articles/s41598 - 019 - 46193 - w 21. Tun HM, Bridgman SL, Chari R, Field CJ, Guttman DS, Becker AB, Mandhane PJ, Turvey SE, Subbarao P, Sears MR, et al. Roles of birth mode and infant gut microbiota in inte rgenerational transmission of overweight and obesity from mother to offspring. JAMA Pediatr. 2018;172:368 77. 22. Stanislawski MA, Dabelea D, Wagner BD, Sontag MK, Lozupone CA, Eggesbø M. Pre - pregnancy weight, gestational weight gain, and the gut microbi ota of mothers and their infants. Microbiome [Internet]. 2017 [cited 2019 Jul 9];5:113. Available from: http://www.ncbi.nlm.nih.gov/pubmed/28870230 23. Branum AM, Kirmeyer SE, Gregory ECW. Prepregnancy Body Mass Index by Maternal Characteristics and State : Data From the Birth Certificate, 2014. Natl Vital Stat Rep [Internet]. 2016 [cited 2019 Sep 12];65:1 11. Available from: 148 http://www.ncbi.nlm.nih.gov/pubmed/27508894 24. Sebire NJ, Jolly M, Harris JP, Wadsworth J, Joffe M, Beard RW, Regan L, Robinson S. Maternal obesity and pregnancy outcome: A study of 287 213 pregnancies in London. Int J Obes. 2001;25:1175 82. 25. Wallace JM, Horgan GW, Bhattacharya S. Placental weight and efficiency in relation to maternal body mass index and the risk of pregnancy co mplications in women delivering singleton babies. Placenta. 2012;33:611 8. 26. Binns C, Lee M, Low WY. The Long - Term Public Health Benefits of Breastfeeding. Asia - Pacific J Public Heal. 2016;28:7 14. 27. Franzosa EA, Huang K, Meadow JF, Gevers D, Lemon KP, Bohannan BJM, Huttenhower C. Identifying personal microbiomes using metagenomic codes. Proc Natl Acad Sci U S A. National Academy of Sciences; 2015;112:E2930 8. 28. Faith JJ, Guruge JL, Charbonneau M, Subramanian S, Seedorf H, Goodman AL, Clemente J C, Knight R, Heath AC, Leibel RL, et al. The long - term stability of the human gut microbiota. Science (80 - ). American Association for the Advancement of Science; 2013;341. 29. EPIC - Norfolk nutritional methods: food frequency questionnaire [Internet]. 20 19. Available from: http://www.srl.cam.ac.uk/epic/nutmethod/FFQ.shtml 30. Nuriel - Ohayon M, Neuman H, Koren O. Microbial changes during pregnancy, birth, and infancy. Frontiers in Microbiology. Frontiers Media S.A.; 2016. 31. Koren O, Goodrich JK, Cullen der TC, Spor A, Laitinen K, Kling Bäckhed H, Gonzalez A, Werner JJ, Angenent LT, Knight R, et al. Host remodeling of the gut microbiome and metabolic changes during pregnancy. Cell. 2012;150:470 80. 32. Gomez - Arango LF, Barrett HL, McIntyre HD, Callaway LK, Morrison M, Nitert MD, Tremellen A, Tobin J, Wilkinson S, McSweeney C, et al. Connections between the gut microbiome and metabolic hormones in early pregnancy in overweight and obese women. Diabetes [Internet]. 2016 [cited 2019 Sep 4];65:2214 23. Avail able from: http://www.ncbi.nlm.nih.gov/pubmed/27217482 33. Shah V, Lambeth SM, Carson T, Lowe J, Ramaraj T, Leff JW, Luo L, Bell CJ. Composition Diversity and Abundance of Gut Microbiome in Prediabetes and Type 2 Diabetes. J Diabetes Obes. Ommega Online P ublishers; 2015;2:108 14. 34. DiGiulio DB, Callahan BJ, McMurdie PJ, Costello EK, Lyell DJ, Robaczewska A, Sun CL, Goltsman DSA, Wong RJ, Shawa G, et al. Temporal and spatial variation of the human microbiota during pregnancy. Proc Natl Acad Sci U S A. N ational Academy of Sciences; 2015;112:11060 5. 35. Diop K, Dufour JC, Levasseur A, Fenollar F. Exhaustive repertoire of human vaginal 149 microbiota. Human Microbiome Journal. Elsevier Ltd; 2019. 36. Aagaard K, Riehle K, Ma J, Segata N, Mistretta TA, Coarfa C, Raza S, Rosenbaum S, van den Veyver I, Milosavljevic A, et al. A metagenomic approach to characterization of the vaginal microbiome signature in pregnancy. PLoS One. 2012;7. 37. Matu MN, Orinda GO, Njagi ENM, Cohen CR, Bukusi EA. In vitro inhibitory activity of human vaginal lactobacilli against pathogenic bacteria associated with bacterial vaginosis in Kenyan women. Anaerobe. 2010;16:210 5. 38. Atassi F, Brassart D, Grob P, Graf F, Servin AL. Lactobacillus strains isolated from the vaginal microbiota of healthy women inhibit Prevotella bivia and Gardnerella vaginalis in coculture and cell culture. FEMS Immunol Med Microbiol. 2006;48:424 32. 39. Ravel J, Gajer P, Abdo Z, Schneider GM, Koenig SSK, McCulle SL, Karlebach S, Gorle R, Russell J, Tacket CO, et al. Vaginal microbiome of reproductive - age women. Proc Natl Acad Sci U S A. National Academy of Sciences; 2011;108:4680 7. 40. Byrd AL, Belkaid Y, Segre JA. The human skin microbiome. Nature Reviews Microbiology. Nature Publishing Group; 2018. p. 143 55. 41. Findley K, Oh J, Yang J, Conlan S, Deming C, Meyer JA, Schoenfeld D, Nomicos E, Park M, Becker J, et al. Topographic diver sity of fungal and bacterial communities in human skin. Nature. 2013;498:367 70. 42. Benn AML, Heng NCK, Broadbent JM, Thomson WM. Studying the human oral microbiome: challenges and the evolution of solutions. Australian Dental Journal. Blackwell Publish ing; 2018. p. 14 24. 43. Deo PN, Deshmukh R. Oral microbiome: Unveiling the fundamentals. Journal of Oral and Maxillofacial Pathology. Wolters Kluwer Medknow Publications; 2019. p. 122 8. 44. Ruiz L, García - Carral C, Rodriguez JM. Unfolding the human m ilk microbiome landscape in the omicsera. Front Microbiol. Frontiers Media S.A.; 2019;10. 45. Hermansson H, Kumar H, Collado MC, Salminen S, Isolauri E, Rautava S. Breast milk microbiota is shaped by mode of delivery and intrapartum antibiotic exposure. Front Nutr. Frontiers Media S.A.; 2019;6. 46. Toscano M, Grandi R De, Grossi E, Drago L. Role of the human breast milk - associated Media S.A.; 2017;8. 47. Williams JE, Carrothers JM, Lackey KA, Beatty NF, Brooker SL, Peterson HK, Steinkamp KM, York MA, Shafii B, Price WJ, et al. Strong Multivariate Relations Exist Among Milk, Oral, and Fecal Microbiomes in Mother - Infant Dyads During the First Six Months Postpartum. J Nut r. NLM (Medline); 2019;149:902 14. 150 48. Jiménez E, De Andrés J, Manrique M, Pareja - Tobes P, Tobes R, Martínez - Blanch JF, Codoñer FM, Ramón D, Fernández L, Rodríguez JM. Metagenomic analysis of milk of healthy and mastitis - suffering women. J Hum Lact [Inte rnet]. 2015 [cited 2019 Sep 5];31:406 15. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25948578 49. Jost T, Lacroix C, Braegger CP, Rochat F, Chassard C. Vertical mother - neonate transfer of maternal gut bacteria via breastfeeding. Environ Microbiol [Internet]. 2014 [cited 2019 Sep 5];16:2891 904. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24033881 50. Walker WA, Iyengar RS. Breast milk, microbiota, and intestinal immune homeostasis. Pediatric Research. Nature Publishing Group; 2015. p. 220 8 . 51. Kovalovszki L, Villányi Z, Pataki I, Veszelowvsky I, Nagy ZB. Isolation of aerobic bacteria from the placenta. Acta Paediatr Acad Sci Hung [Internet]. 1982 [cited 2019 Sep 6];23:357 60. Available from: http://www.ncbi.nlm.nih.gov/pubmed/6817588 52. Pelzer E, Gomez - Arango LF, Barrett HL, Nitert MD. Review: Maternal health and the placental microbiome. Placenta. W.B. Saunders Ltd; 2017. p. 30 7. 53. Lauder AP, Roche AM, Sherrill - Mix S, Bailey A, Laughlin AL, Bittinger K, Leite R, Elovitz MA, Parry S, Bushman FD. Comparison of placenta samples with contamination controls does not provide evidence for a distinct placenta microbiota. Microbiome. BioMed Central Ltd.; 2016;4. 54. Leiby JS, McCormick K, Sherrill - Mix S, Clarke EL, Kessler LR, Taylor LJ, Hofstaedter CE, Roche AM, Mattei LM, Bittinger K, et al. Lack of detection of a human placenta microbiome in samples from preterm and term deliveries. Microbiome. NLM (Medline); 2018;6:196. 55. de Goffau MC, Lager S, Sovio U, Gaccioli F, Cook E, Peacock SJ, Parkhill J, Charnock - Jones DS, Smith GCS. Human placenta has no microbiome but can contain potential pathogens. Nature. Springer Science and Business Media LLC; 2019; 56. Palmer C, Bik EM, DiGiulio DB, Relman DA, Brown PO. Development of the human in fant intestinal microbiota. PLoS Biol [Internet]. 2007 [cited 2019 Sep 4];5:1556 73. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17594176 57. De Filippo C, Cavalieri D, Di Paola M, Ramazzotti M, Poullet JB, Massart S, Collini S, Pieraccini G, Lione tti P. Impact of diet in shaping gut microbiota revealed by a comparative study in children from Europe and rural Africa. Proc Natl Acad Sci U S A. 2010;107:14691 6. 58. Koenig JE, Spor A, Scalfone N, Fricker AD, Stombaugh J, Knight R, Angenent LT, Ley R E. Succession of microbial consortia in the developing infant gut microbiome. Proc Natl Acad Sci U S A. 2011;108:4578 85. 59. Wu GD, Chen J, Hoffmann C, Bittinger K, Chen YY, Keilbaugh SA, Bewtra M, Knights D, Walters WA, Knight R, et al. Linking long - te rm dietary patterns with gut microbial 151 enterotypes. Science (80 - ). American Association for the Advancement of Science; 2011;334:105 8. 60. Dominguez - Bello MG, Costello EK, Contreras M, Magris M, Hidalgo G, Fierer N, Knight R. Delivery mode shapes the a cquisition and structure of the initial microbiota across multiple body habitats in newborns. Proc Natl Acad Sci U S A. 2010;107:11971 5. 61. Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, Versalovic J. The placenta harbors a unique microbiome. Sci Transl Med. 2014;6. 62. Houghteling PD, Walker WA. Why is initial bacterial colonization of the intestine alth? Journal of Pediatric Gastroenterology and Nutrition. Lippincott Williams and Wilkins; 2015. p. 294 307. 63. Gehrig JL, Venkatesh S, Chang H - W, Hibberd MC, Kung VL, Cheng J, Chen RY, Subramanian S, Cowardin CA, Meier MF, et al. Effects of microbiota - directed foods in gnotobiotic animals and undernourished children. Science [Internet]. 2019 [cited 2019 Jul 22];365. Available from: http://www.ncbi.nlm.nih.gov/pubmed/31296738 64. van den Elsen LWJ, Garssen J, Burcelin R, Verhasselt V. Shaping the gut m icrobiota by breastfeeding: The gateway to allergy prevention? Frontiers in Pediatrics. 2019. 65. Penders J, Thijs C, Vink C, Stelma FF, Snijders B, Kummeling I, Van Den Brandt PA, Stobberingh EE. Factors influencing the composition of the intestinal mic robiota in early infancy. Pediatrics. 2006;118:511 21. 66. Tanaka M, Nakayama J. Development of the gut microbiota in infancy and its impact on health in later life. Allergology International. Japanese Society of Allergology; 2017. p. 515 22. 67. Yang R, Gao R, Cui S, Zhong H, Zhang X, Chen Y, Wang J, Qin H. Dynamic signatures of gut microbiota and influences of delivery and feeding modes during the first 6 months of life. Physiol Genomics. 2019;51:368 78. 68. Madan JC, Hoen AG, Lundgren SN, Farzan SF , Cottingham KL, Morrison HG, Sogin ML, Li H, Moore JH, Karagas MR. Association of cesarean delivery and formula supplementation with the intestinal microbiome of 6 - week - old infants. JAMA Pediatr. 2016;170:212 9. 69. Jakobsson HE, Abrahamsson TR, Jenmalm MC, Harris K, Quince C, Jernberg C, Björkstén B, Engstrand L, Andersson AF. Decreased gut microbiota diversity, delayed Bacteroidetes colonisation and reduced Th1 responses in infants delivered by Caesarean section. Gut. BMJ Publishing Group; 2014;63:559 66. 70. 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 and Bioinformatics. 2019. p. 13 25. 152 71. Azad MB, Konya T, Maughan H, Guttman DS, Field CJ, Cha ri RS, Sears MR, Becker AB, Scott JA, Kozyrskyj AL. Gut microbiota of healthy Canadian infants: Profiles by mode of delivery and infant diet at 4 months. CMAJ. Canadian Medical Association; 2013;185:385 94. 72. n DS, Smith DP, Wong MC, Ross MC, Lloyd RE, Doddapaneni HV, Metcalf GA, et al. Temporal development of the gut microbiome in early childhood from the TEDDY study. Nature. 2018. p. 583 8. 73. Kabeerdoss J, Ferdous S, Balamurugan R, Mechenro J, Vidya R, Sa nthanam S, Jana AK, Ramakrishna BS. Development of the gut microbiota in southern indian infants from birth to 6 months: A molecular analysis. J Nutr Sci [Internet]. 2013 [cited 2019 Jul 23];2:1 7. Available from: http://www.ncbi.nlm.nih.gov/pubmed/2519156 6 74. Hesla HM, Stenius F, Jäderlund L, Nelson R, Engstrand L, Alm J, Dicksved J. Impact of lifestyle on the gut microbiota of healthy infants and their mothers - the ALADDIN birth cohort. FEMS Microbiol Ecol [Internet]. 2014 [cited 2019 Jul 23];90:791 80 1. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25290507 75. Dogra S, Sakwinska O, Soh SE, Ngom - Bru C, Brück WM, Berger B, Brüssow H, Lee YS, Yap F, Chong YS, et al. Dynamics of infant gut microbiota are influenced by delivery mode and gestational d uration and are associated with subsequent adiposity. MBio. 2015;6. 76. Guittar J, Shade A, Litchman E. Trait - based community assembly and succession of the infant gut microbiome. Nat Commun. 2019;10. 77. Martin FPJ, Wang Y, Sprenger N, Holmes E, Lindo n JC, Kochhar S, Nicholson JK. Effects of probiotic Lactobacillus Paracasei treatment on the host gut tissue metabolic profiles probed via magic - angle - spinning NMR spectroscopy. J Proteome Res. 2007;6:1471 81. 78. Hunt KM, Foster JA, Forney LJ, Schütte U ME, Beck DL, Abdo Z, Fox LK, Williams JE, McGuire MK, McGuire MA. Characterization of the diversity and temporal stability of bacterial communities in human milk. PLoS One. 2011;6. 79. De Leoz MLA, Kalanetra KM, Bokulich NA, Strum JS, Underwood MA, Germa n JB, Mills DA, Lebrilla CB. Human milk glycomics and gut microbial genomics in infant feces show a correlation between human milk oligosaccharides and gut microbiota: A proof - of - concept study. J Proteome Res [Internet]. 2015 [cited 2019 Jul 9];14:491 502. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25300177 80. Simeoni M, Citraro ML, Deodato F, Provenzano M, Capria M, Comi A, Libri E, Andreucci M, Puja A, Abenavoli L, et al. An open - label, randomized, placebo - controlled study on the effectiveness o f a novel probiotics administration protocol (ProbiotiCKD) in patients with mild renal insufficiency (stage 3a of CKD). European Journal of Nutrition. Dr. Dietrich Steinkopff Verlag GmbH and Co. KG; 2018; 153 81. Pannaraj PS, Li F, Cerini C, Bender JM, Yang S, Rollie A, Adisetiyo H, Zabih S, Lincez PJ, Bittinger K, et al. Association between breast milk bacterial communities and establishment and development of the infant gut microbiome. JAMA Pediatr. 2017;171:647 54. 82. Forbes JD, Azad MB, Vehling L. Erratum: Association of exposure to formula in the hospital and subsequent infant feeding practices with gut microbiota and risk of overweight in the first year of life (JAMA Pediatr (2018) DOI: 10.1001/jamapediatrics.2018.1161) [Internet]. JAMA Pediatrics . 2018 [cited 2019 Jul 9]. p. 704. Available from: http://archpedi.jamanetwork.com/article.aspx?doi=10.1001/jamapediatrics.2018.1161 83. Baumann - and maternal gestational weight gain pr edict early metabolic maturation of gut microbiomes [Internet]. Nature Medicine. 2018 [cited 2019 Jul 9]. p. 1822 9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30374198 84. Ho NT, Li F, Lee - Sarwar KA, Tun HM, Brown BP, Pannaraj PS, Bender JM, Azad MB, Thompson AL, Weiss ST, et al. Meta - analysis of effects of exclusive breastfeeding on infant gut microbiota across populations. Nat Commun. 2018;9. 85. Akkerman R, Faas MM, de Vos P. Non - digestible carbohydrates in infant formula as substitution for human milk oligosaccharide functions: Effects on microbiota and gut maturation. Critical Reviews in Food Science and Nutrition. 2019. 86. Trasande L, Blustein J, Liu M, Corwin E, Cox LM, Blaser MJ. Infant antibiotic exposures and early - life body mass. In t J Obes. 2013;37:16 23. 87. Saari A, Virta LJ, Sankilampi U, Dunkel L, Saxen H. Antibiotic exposure in infancy and risk of being overweight in the first 24 months of life. Pediatrics. American Academy of Pediatrics; 2015;135:617 26. 88. Azad MB, Konya T, Persaud RR, Guttman DS, Cha ri RS, Field CJ, Sears MR, Mandhane PJ, Turvey SE, Subbarao P, et al. Impact of maternal intrapartum antibiotics, method of birth and breastfeeding on gut microbiota during the first year of life: A prospective cohort study. BJOG: An International Journal of Obstetrics and Gynaecology. Blackwell Publishing Ltd; 2016. p. 983 93. 89. Korpela K, Salonen A, Saxen H, Nikkonen A, Peltola V, Jaakkola T, de Vos W, Kolho KL. Antibiotics in early life associate with specific gut microbiota signatures in a prospecti ve longitudinal infant cohort. Pediatr Res. Springer Nature; 2020; 90. Yassour M, Jason E, Hogstrom LJ, Arthur TD, Tripathi S, Siljander H, Selvenius J, Oikarinen S, Hyöty H, Virtanen SM, et al. Strain - Level Analysis of Mother - to - Child Bacterial Transmis sion during the First Few Months of Life. Cell Host Microbe. Cell Press; 2018;24:146 - 154.e4. 91. Ferretti P, Pasolli E, Tett A, Asnicar F, Gorfer V, Fedi S, Armanini F, Truong DT, Manara 154 S, Zolfo M, et al. Mother - to - Infant Microbial Transmission from Dif ferent Body Sites Shapes the Developing Infant Gut Microbiome. Cell Host Microbe [Internet]. 2018 [cited 2019 Jul 9];24:133 - 145.e5. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30001516 92. Campbell T, Shackleton N. Pre - pregnancy body mass index and breastfeeding initiation, early cessation and longevity: Evidence from the first wave of the UK Millennium Cohort Study. Journal of Epidemiology and Community Health. 2018; 93. Sordillo JE, Zhou Y, McGeachie MJ, Ziniti J, Lange N, Laranjo N, Savage JR, Carey V, - 6 months: Findings from the ethnically diverse Vitamin D Antenatal Asthma Reduction Trial (VDAART). J Allergy Clin Immunol [Internet]. 2017 [cited 2019 Jul 9];139: 482 - 491.e14. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27746239 94. Milani C, Duranti S, Bottacini F, Casey E, Turroni F, Mahony J, Belzer C, Delgado Palacio S, Arboleya Montes S, Mancabelli L, et al. The First Microbial Colonizers of the Human G ut: Composition, Activities, and Health Implications of the Infant Gut Microbiota. Microbiol Mol Biol Rev [Internet]. 2017;81. Available from: https://doi.org/10.1128/MMBR 95. Björkstén B, Sepp E, Julge K, Voor T, Mikelsaar M. Allergy development and the intestinal microflora during the first year of life. J Allergy Clin Immunol. 2001;108:516 20. 96. Kalliomäki M, Kirjavainen P, Eerola E, Kero P, Salminen S, Isolauri E. Distinct patterns of neonatal gut microflora in infants in whom atopy was and was not developing. J Allergy Clin Immunol. Mosby Inc.; 2001;107:129 34. 97. Blanton L V., Charbonneau MR, Salih T, Barratt MJ, Venkatesh S, Ilkaveya O, Subramanian S, Manary MJ, Trehan I, Jorgensen JM, et al. Gut bacteria that prevent growth impairments transm itted by microbiota from malnourished children. Science (80 - ). American Association for the Advancement of Science; 2016;351. 98. Santacruz A, Collado MC, García - Valdés L, Segura MT, Marítn - Lagos JA, Anjos T, Martí - Romero M, Lopez RM, Florido J, Campoy C, et al. Gut microbiota composition is associated with body weight, weight gain and biochemical parameters in pregnant women. Br J Nutr. 2010;104:83 92. 99. Ata B, Yildiz S, Turkgeldi E, Brocal VP, Dinleyici EC, Moya A, Urman B. The Endobiota Study: Com parison of Vaginal, Cervical and Gut Microbiota Between Women with Stage 3/4 Endometriosis and Healthy Controls. Sci Rep. Nature Publishing Group; 2019;9. 100. Goodrich JK, Di Rienzi SC, Poole AC, Koren O, Walters WA, Caporaso JG, Knight R, Ley RE. Condu cting a microbiome study. Cell. 2014. p. 250 62. 101. Song SJ, Amir A, Metcalf JL, Amato KR, Xu ZZ, Humphrey G, Knight R. Preservation Methods Differ in Fecal Microbiome Stability, Affecting Suitability for Field Studies. 155 mSystems. 2016;1. 102. Choo JM , Leong LEX, Rogers GB. Sample storage conditions significantly influence faecal microbiome profiles. Sci Rep [Internet]. 2015 [cited 2019 Jul 9];5:16350. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26572876 103. Lauber CL, Zhou N, Gordon JI, Knigh t R, Fierer N. Effect of storage conditions on the assessment of bacterial community structure in soil and human - associated samples. FEMS Microbiol Lett [Internet]. 2010 [cited 2019 Jul 9];307:80 6. Available from: http://www.ncbi.nlm.nih.gov/pubmed/204123 03 104. Tedjo DI, Jonkers DMAE, Savelkoul PH, Masclee AA, Best N Van, Pierik MJ, Penders J. The effect of sampling and storage on the fecal microbiota composition in healthy and diseased subjects. PLoS One. 2015;10. 105. Jenkins S V., Vang KB, Gies A, G riffin RJ, Jun SR, Nookaew I, Dings RPM. Sample storage conditions induce post - collection biases in microbiome profiles. BMC Microbiol. 2018; 106. Methé BA, Nelson KE, Pop M, Creasy HH, Giglio MG, Huttenhower C, Gevers D, Petrosino JF, Abubucker S, Badge r JH, et al. A framework for human microbiome research. Nature. 2012;486:215 21. 107. Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of a dual - index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the miseq illumina sequencing platform. Appl Environ Microbiol. 2013;79:5112 20. 108. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, Peplies J, Glöckner FO. The SILVA ribosomal RNA gene database project: Improved data processing and web - based tools. Nucleic Acids Res. 2013;41. 109. Bonuck KA, Trombley M, Freeman K, McKee D. Randomized, controlled trial of a prenatal and postnatal lactation consultant intervention on duration and intensity of breastfeeding up to 12 months. Pediatrics. 20 05;116:1413 26. 110. Team RC. R: A language and environment for statistical computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2019. Available from: https://www.r - project.org/ 111. Jari Oksanen and F. Guillaume Blanchet and M ichael Friendly and Roeland Kindt and Simpson and Peter Solymos and M. Henry H. Stevens and Eduard Szoecs and Helene Wagner. vegan: Community Ecology Package [Internet]. 201 9. Available from: https://cran.r - project.org/package=vegan 112. Yang I, Corwin EJ, Brennan PA, Jordan S, Murphy JR, Dunlop A. The infant microbiome: Implications for infant health and neurocognitive development. Nurs Res. Lippincott Williams and Wilkins; 2016;65:76 88. 156 113. Bäckhed F, Ding H, Wang T, Hooper L V., Gou YK, Nagy A, Semenkovich CF, Gordon JI. The gut microbiota as an environmental factor that regulates fat storage. Proc Natl Acad Sci U S A. 2004;101:15718 23. 114. Stephenson J, Heslehurst N, Hall J, Schoenaker DAJM, Hutchinson J, Cade JE, Poston L, Barrett G, Crozier SR, Barker M, et al. Before the beginning: nutrition and lifestyle in the preconception period and its importance for future health. The Lancet. Lancet Publishing Group; 2018. p. 1830 41. 115. Pos ton L, Caleyachetty R, Cnattingius S, Corvalán C, Uauy R, Herring S, Gillman MW. Preconceptional and maternal obesity: epidemiology and health consequences. The Lancet Diabetes and Endocrinology. Lancet Publishing Group; 2016. p. 1025 36. 116. Yun Y, Kim H - N, Kim SE, Heo SG, Chang Y, Ryu S, Shin H, Kim H - L. Comparative analysis of gut microbiota associated with body mass index in a large Korean cohort. BMC Microbiol [Internet]. 2017 [cited 2020 Jan 25];17:151. Available from: http://bmcmicrobiol.biomedcen tral.com/articles/10.1186/s12866 - 017 - 1052 - 0 117. Barbour LA, McCurdy CE, Hernandez TL, Kirwan JP, Catalano PM, Friedman JE. Cellular mechanisms for insulin resistance in normal pregnancy and gestational diabetes. Diabetes Care. 2007;30. 118. El Aila NA, Tency I, Claeys G, Verstraelen H, Saerens B, Lopes dos Santos Santiago G, De Backer E, Cools P, Temmerman M, Verhelst R, et al. Identification and genotyping of bacteria from paired vaginal and rectal samples from pregnant women indicates similarity betwe en vaginal and rectal microflora. BMC Infect Dis. 2009;9:167. 119. Collado MC, Isolauri E, Laitinen K, Salminen S. Distinct composition of gut microbiota during pregnancy in. Am Soc Nutr. 2008;88:894 9. 120. Aatsinki AK, Uusitupa HM, Munukka E, Pesonen H, Rintala A, Pietilä S, Lahti L, Eerola E, Karlsson L, Karlsson H. Gut Microbiota Composition in Mid - Pregnancy Is Associated Mary Ann Liebert Inc.; 2018;27:1293 301. 121. 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. 122. Fitzstevens JL, Smith KC , Hagadorn JI, Caimano MJ, Matson AP, Brownell EA. Systematic Review of the Human Milk Microbiota. Nutr Clin Pract [Internet]. 2017 [cited 2020 Jan 25];32:354 64. Available from: http://doi.wiley.com/10.1177/0884533616670150 123. Simeoni U, Berger B, Juni ck J, Blaut M, Pecquet S, Rezzonico E, Grathwohl D, Sprenger N, Brüssow H, Szajewska H, et al. Gut microbiota analysis reveals a marked shift to bifidobacteria by a starter infant formula containing a synbiotic of bovine milk - derived oligosaccharides and B ifidobacterium animalis subsp. lactis CNCM I - 3446. Environ 157 Microbiol [Internet]. 2016 [cited 2020 Jan 25];18:2185 95. Available from: http://doi.wiley.com/10.1111/1462 - 2920.13144 124. Bider - Canfield Z, Martinez MP, Wang X, Yu W, Bautista MP, Brookey J, Page KA, Buchanan TA, Xiang AH. Maternal obesity, gestational diabetes, breastfeeding and childhood overweight at age 2 years. Pediatr Obes [Internet]. 2017 [cited 2020 Jan 25];12:171 8. Available from: http://doi.wiley.com/10.1111/ijpo.12125 125. Lakshma n R, Elks CE, Ong KK. Obesity Childhood Obesity. 2012 [cited 2020 Jan 25]; Available from: http://circ.ahajournals.org 126. Lobstein T, Jackson - Leach R, Moodie ML, Hall KD, Gortmaker SL, Swinburn BA, James WPT, Wang Y, McPherson K. Child and adolescent ob esity: Part of a bigger picture. The Lancet. Lancet Publishing Group; 2015. p. 2510 20. 127. Mueller NT, Bakacs E, Combellick J, Grigoryan Z, Dominguez - Bello MG. The infant microbiome development: Mom matters. Trends in Molecular Medicine. Elsevier Ltd; 2015. p. 109 17. 128. Mueller NT, Shin H, Pizoni A, Werlang IC, Matte U, Goldani MZ, Goldani HAS, Dominguez - Bello MG. Birth mode - dependent association between pre - pregnancy maternal weight status and the neonatal intestinal microbiome. Sci Rep. 2016;6. 129. Ripley WNV and BD. Modern Applied Statistics with S [Internet]. New York: Springer; 2002. Available from: http://www.stats.ox.ac.uk/pub/MASS4 130. Nagpal R, Tsuji H, Takahashi T, Nomoto K, Kawashima K, Nagata S, Yamashiro Y. Ontogenesis of the gut m icrobiota composition in healthy, full - term, vaginally born and breast - - eye view. Front Microbiol. Frontiers Media S.A.; 2017;8. 131. Gohir W, Ratcliffe EM, Sloboda DM. Of the bugs that sha pe us: Maternal obesity, the gut microbiome, and long - term disease risk. Pediatric Research. Nature Publishing Group; 2015. p. 196 204. 132. Stanislawski MA, Dabelea D, Wagner BD, Iszatt N, Dahl C, Sontag MK, Knight R, Lozupone CA, Eggesbø M. Gut Microbi ota in the First 2 Years of Life and the Association with Body Mass Index at Age 12 in a Norwegian Birth Cohort. MBio. NLM (Medline); 2018;9. 133. Asnicar F, Manara S, Zolfo M, Truong DT, Scholz M, Armanini F, Ferretti P, Gorfer V, Pedrotti A, Tett A, et al. Studying Vertical Microbiome Transmission from Mothers to Infants by Strain - Level Metagenomic Profiling. mSystems. American Society for Microbiology; 2017;2. 134. Duranti S, Andrea Lugli G, Mancabelli L, Armanini F, Turroni F, James K, Ferretti P, G orfer V, Ferrario C, Milani C, et al. Maternal inheritance of bifidobacterial communities and bifidophages in infants through vertical transmission. [cited 2020 Jan 25]; Available 158 from: http://www.hmpdacc.org/doc/HMP_MOP_Version12_ 135. Gough EK, Stephens DA, Moodie EEM, Prendergast AJ, Stoltzfus RJ, Humphrey JH, Manges AR. Linear growth faltering in infants is associated with Acidaminococcus sp. and community - level changes in the gut microbiota. Microbiome. Springer Science and Business Media LLC; 2015;3. 136. Koleva PT, Bridgman SL, Kozyrskyj AL. The infant gut microbiome: Evidence for obesity risk and dietary intervention. Nutrients. MDPI AG; 2015. p. 2237 60. 137. Thompson AL, Monteagudo - Mera A, Cadenas MB, Lampl ML, Azcarate - Peril MA. Milk - and sol id - feeding practices and daycare attendance are associated with differences in bacterial diversity, predominant communities, and metabolic and immune function of the infant gut microbiome. Front Cell Infect Microbiol. Frontiers Research Foundation; 2015;5. 138. Horta BL, Loret De Mola C, Victora CG. Long - term consequences of breastfeeding on cholesterol, obesity, systolic blood pressure and type 2 diabetes: A systematic review and meta - analysis. Acta Paediatrica, International Journal of Paediatrics. Blac kwell Publishing Ltd; 2015. p. 30 7. 139. Shin D, Chung H, Weatherspoon L, Song WO. Validity of Prepregnancy Weight Status Estimated from Self - reported Height and Weight. Matern Child Health J. Kluwer Academic/Plenum Press New York; 2014;18:1667 74. 140. Modi SR, Collins JJ, Relman DA. Antibioti cs and the gut microbiota. Journal of Clinical Investigation. American Society for Clinical Investigation; 2014. p. 4212 8. 141. Miller GE, Engen PA, Gillevet PM, Shaikh M, Sikaroodi M, Forsyth CB, Mutlu E, Keshavarzian A. Lower Neighborhood Socioeconomi c Status Associated with Reduced Diversity of the Colonic Microbiota in Healthy Adults. Driks A, editor. PLoS One [Internet]. 2016 [cited 2020 Jan 25];11:e0148952. Available from: https://dx.plos.org/10.1371/journal.pone.0148952 142. Papathoma E, Triga M, Fouzas S, Dimitriou G. Cesarean section delivery and development of food allergy and atopic dermatitis in early childhood. Pediatr Allergy Immunol. Blackwell Publishing Ltd; 2016;27:419 24. 143. Van Nimwegen FA, Penders J, Stobberingh EE, Postma DS, Kop pelman GH, Kerkhof M, Reijmerink NE, Dompeling E, Van Den Brandt PA, Ferreira I, et al. Mode and place of delivery, gastrointestinal microbiota, and their influence on asthma and atopy. J Allergy Clin Immunol. Mosby Inc.; 2011;128. 144. Johansson MA, Sjö gren YM, Persson JO, Nilsson C, Sverremark - Ekström E. Early colonization with a group of Lactobacilli decreases the risk for allergy at five years of age despite allergic heredity. PLoS One. 2011;6. 145. Section on Breastfeeding. Breastfeeding and the Us e of Human Milk. Pediatrics. 159 2012;129:e827 41. 146. Wojcicki JM. Maternal prepregnancy body mass index and initiation and duration of [cited 2019 Jul 9]. p. 341 7. Avai lable from: http://www.ncbi.nlm.nih.gov/pubmed/21434834 147. Turcksin R, Bel S, Galjaard S, Devlieger R. Maternal obesity and breastfeeding intention, initiation, intensity and duration: A systematic review. Maternal and Child Nutrition. 2014. p. 166 83. 148. microbiota acquisition, composition, and activity during early infancy: a prospective follow - up study initiated in early pregnancy. Am J Clin Nutr [Internet]. 2010 [cited 2020 Jan 24];92:1023 30. Available from: https://academic.oup.com/ajcn/article/92/5/1023/4597484 149. Galley JD, Bailey M, Dush CK, Schoppe - Sullivan S, Christian LM. Maternal obesity is associated with alterations in the gut microbiome in toddlers . PLoS One [Internet]. 2014 [cited 2019 Jul 9];9:e113026. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25409177 150. Fernandes J, Su W, Rahat - Rozenbloom S, Wolever TMS, Comelli EM. Adiposity, gut microbiota and faecal short chain fatty acids are lin ked in adult humans. Nutr Diabetes. 2014;4. 151. Riva A, Borgo F, Lassandro C, Verduci E, Morace G, Borghi E, Berry D. Pediatric obesity is associated with an altered gut microbiota and discordant shifts in Firmicutes populations. Environ Microbiol [Inte rnet]. 2017 [cited 2019 Jul 9];19:95 105. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27450202 152. Kumari M, Kozyrskyj AL. Gut microbial metabolism defines host metabolism: an emerging perspective in obesity and allergic inflammation [Internet]. O besity Reviews. 2017 [cited 2019 Jul 9]. p. 18 31. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27862824 153. Laursen MF, Andersen LBB, Michaelsen KF, Mølgaard C, Trolle E, Bahl MI, Licht TR. Infant Gut Microbiota Development Is Driven by Transition to Family Foods Independent of Maternal Obesity. mSphere. American Society for Microbiology; 2016;1. 154. Iozzo P, Sanguinetti E. Early Dietary Patterns and Microbiota Development: Still a Way to Go from Descriptive Interactions to Health - Relevant Solut ions. Front Nutr. Frontiers Media SA; 2018;5. 155. Mesa MD, Loureiro B, Iglesia I, Fernandez Gonzalez S, Llurba Olivé E, García Algar O, Solana MJ, Cabero Perez MJ, Sainz T, Martinez L, et al. The Evolving Microbiome from Pregnancy to Early Infancy: A Co mprehensive Review. Nutrients [Internet]. 2020 [cited 2020 Jan 30];12:133. Available from: https://www.mdpi.com/2072 - 6643/12/1/133 160 156. De Muinck EJ, Trosvik P. Individuality and convergence of the infant gut microbiota during the first year of life. Nat Commun. Nature Publishing Group; 2018;9. 157. Lazar V, Ditu LM, Pircalabioru GG, Gheorghe I, Curutiu C, Holban AM, Picu A, Petcu L, Chifiriuc MC. Aspects of gut microbiota and immune system interactions in infectious diseases, immunopathology, and cancer . Frontiers in Immunology. Frontiers Media S.A.; 2018. p. 1830. 158. Cheng L, Ramchandran S, Vatanen T, Lietzén N, Lahesmaa R, Vehtari A, Lähdesmäki H. An additive Gaussian process regression model for interpretable non - parametric analysis of longitudina l data. Nat Commun. 2019;10. 159. Duvenaud D, Nickisch H, Rasmussen CE. Additive Gaussian processes. Advances in Neural Information Processing Systems 24: 25th Annual Conference on Neural Information Processing Systems 2011, NIPS 2011 [Internet]. 2011. Available from: http://arxiv.org/abs/1112.4394 160. Matsuyama M, Morrison M, Cao KAL, Pruilh S, Davies PSW, Wall C, Lovell A, Hill RJ. Dietary intake influences gut microbiota development of healthy Australian children from the age of on e to two years. Sci Rep. Nature Publishing Group; 2019;9:1 11. 161. Rinninella E, Raoul P, Cintoni M, Franceschi F, Miggiano GAD, Gasbarrini A, Mele MC. What is the healthy gut microbiota composition? A changing ecosystem across age, environment, diet, a nd diseases. Microorganisms. MDPI AG; 2019;7. 162. Ravussin Y, Koren O, Spor A, Leduc C, Gutman R, Stombaugh J, Knight R, Ley RE, Leibel RL. Responses of gut microbiota to diet composition and weight loss in lean and obese mice. Obesity. NIH Public Acces s; 2012;20:738 47. 163. Chambers ES, Preston T, Frost G, Morrison DJ. Role of Gut Microbiota - Generated Short - Chain Fatty Acids in Metabolic and Cardiovascular Health. Current Nutrition Reports. Current Science Inc.; 2018. p. 198 206. 164. Morrison DJ, Preston T. Formation of short chain fatty acids by the gut microbiota and their impact on human metabolism. Gut Microbes. Taylor and Francis Inc.; 2016. p. 189 200. 165. Pérez - Cobas AE, Artacho A, Knecht H, Ferrús ML, Friedrichs A, Ott SJ, Moya A, Latorre A, Gosalbes MJ. Differential effects of antibiotic therapy on the structure and function of human gut microbiota. PLoS One. 2013; 166. Cho I, Yamanishi S, Cox L, Methé BA, Zavadil J, Li K, Gao Z, Mahana D, Raju K, Teitler I, et al. Antibiotics in early life alter the murine colonic microbiome and adiposity. Nature. 2012; 167. Tun MH, Tun HM, Mahoney JJ, Konya TB, Guttman DS, Becker AB, Mandhane PJ, Turvey SE, Subbarao P, Sears MR, et al. Postnatal exposure to household disinfectants, infant gut microbiota and subsequent risk of overweight in children. CMAJ. 2018; 161 168. Ottman N, Smidt H, de Vos WM, Belzer C. The function of our microbiota: who is out there and what do they do? Frontiers in cellular and infection microbiology. 2012. 169. Ley RE, Hamady M, Lozupone C, Turnbaugh PJ, Ramey RR, Bircher JS, Schlegel ML, Tucker TA, Schrenzel MD, Knight R, et al. Evolution of mammals and their gut microbes. Science (80 - ). 20 08; 170. Marcobal A, Barboza M, Sonnenburg ED, Pudlo N, Martens EC, Desai P, Lebrilla CB, Weimer BC, Mills DA, German JB, et al. Bacteroides in the infant gut consume milk oligosaccharides via mucus - utilization pathways. Cell Host Microbe. 2011; 171. W microbiome. Nature Microbiology. 2017. 172. Soderborg TK, Clark SE, Mulligan CE, Janssen RC, Babcock L, Ir D, Lemas DJ, Johnson LK, Weir T, Lenz LL, et al. The gut microbiot a in infants of obese mothers increases inflammation and susceptibility to NAFLD. Nat Commun. Nature Publishing Group; 2018;9. 173. Kim KN, Yao Y, Ju SY. Short chain fatty acids and fecal microbiota abundance in humans with obesity: A systematic review a nd meta - analysis. Nutrients. 2019. 174. Den Besten G, Bleeker A, Gerding A, Van Eunen K, Havinga R, Van Dijk TH, Oosterveer MH, Jonker JW, Groen AK, Reijngoud DJ, et al. Short - chain fatty acids protect against high - fat diet - induced obesity via a pparg - de pendent switch from lipogenesis to fat oxidation. Diabetes. 2015; 175. Paul HA, Bomhof MR, Vogel HJ, Reimer RA. Diet - induced changes in maternal gut microbiota and metabolomic profiles influence programming of offspring obesity risk in rats. Sci Rep. 201 6; 176. Li M, van Esch BCAM, Wagenaar GTM, Garssen J, Folkerts G, Henricks PAJ. Pro - and anti - inflammatory effects of short chain fatty acids on immune and endothelial cells. Eur J Pharmacol. 2018; 177. Monda V, Villano I, Messina A, Valenzano A, Espos ito T, Moscatelli F, Viggiano A, Cibelli G, Chieffi S, Monda M, et al. Exercise modifies the gut microbiota with positive health effects. Oxidative Medicine and Cellular Longevity. 2017. 178. Mohammadkhah AI, Simpson EB, Patterson SG, Ferguson JF. Develo pment of the Gut Microbiome in Children, and Lifetime Implications for Obesity and Cardiometabolic Disease. Children. 2018; 179. Baothman OA, Zamzami MA, Taher I, Abubaker J, Abu - Farha M. The role of Gut Microbiota in the development of obesity and Diabe tes. Lipids in Health and Disease. 2016. 180. Kane A V., Dinh DM, Ward HD. Childhood malnutrition and the intestinal microbiome. 162 Pediatric Research. 2015. 181. Miro - Blanch J, Yanes O. Epigenetic regulation at the interplay between gut microbiota and ho st metabolism. Frontiers in Genetics. 2019. 182. Krautkramer KA, Kreznar JH, Romano KA, Vivas EI, Barrett - Wilt GA, Rabaglia ME, Keller MP, Attie AD, Rey FE, Denu JM. Diet - Microbiota Interactions Mediate Global Epigenetic Programming in Multiple Host Tiss ues. Mol Cell. 2016; 183. Gonçalves P, Araújo JR, Di Santo JP. A cross - talk between microbiota - derived short - chain fatty acids and the host mucosal immune system regulates intestinal homeostasis and inflammatory bowel disease. Inflamm Bowel Dis. 2018; 1 84. Hegar B, Wibowo Y, Basrowi RW, Ranuh RG, Sudarmo SM, Munasir Z, Atthiyah AF, Widodo AD, Supriatmo, Kadim M, et al. The role of two human milk oligosaccharides, - fucosyllactose and lacto - N - neotetraose, in infant nutrition. Pediatric Gastroenterology, Hepatology and Nutrition. 2019. 185. Kameyama K, Itoh K. Intestinal colonization by a lachnospiraceae bacterium contributes to the development of diabetes in obese mice. Microbes Environ. Japanese Society of Microbial Ecology; 2014;29:427 30.