Advertisement

Diet Quality and the Gut Microbiota in Women Living in Alabama

  • Rebecca B. Little
    Affiliations
    Division of Preventive Medicine, Department of Medicine, Heersink School of Medicine, The University of Alabama at Birmingham, Birmingham, Alabama
    Search for articles by this author
  • Anarina L. Murillo
    Correspondence
    Address correspondence to: Anarina L. Murillo, PhD, Department of Biostatistics, School of Global Public Health, New York University, 708 Broadway, New York NY 10003.
    Affiliations
    Department of Biostatistics, School of Public Health, The University of Alabama at Birmingham, Birmingham, Alabama

    Department of Pediatrics, Warren Alpert Medical School, Brown University, Providence, Rhode Island

    Center for Statistical Sciences, School of Public Health, Brown University, Providence, Rhode Island
    Search for articles by this author
  • William J. Van Der Pol
    Affiliations
    Biomedical Informatics, UAB Center for Clinical and Translational Science, The University of Alabama at Birmingham, Birmingham, Alabama
    Search for articles by this author
  • Elliot J. Lefkowitz
    Affiliations
    Biomedical Informatics, UAB Center for Clinical and Translational Science, The University of Alabama at Birmingham, Birmingham, Alabama

    Department of Microbiology, School of Medicine, Heersink School of Medicine, The University of Alabama at Birmingham, Birmingham, Alabama
    Search for articles by this author
  • Casey D. Morrow
    Affiliations
    Department of Cell, Developmental and Integrative Biology, Heersink School of Medicine, The University of Alabama at Birmingham, Birmingham, Alabama
    Search for articles by this author
  • Nengjun Yi
    Affiliations
    Department of Biostatistics, School of Public Health, The University of Alabama at Birmingham, Birmingham, Alabama
    Search for articles by this author
  • Tiffany L. Carson
    Affiliations
    Department of Health Outcomes and Behavior, H. Lee Moffitt Cancer Center & Research Institute, Tampa, Florida

    Department of Oncologic Sciences, Morsani College of Medicine, University of South Florida, Tampa, Florida
    Search for articles by this author

      Introduction

      The gut microbiota is associated with obesity and modulated by individual dietary components. However, the relationships between diet quality and the gut microbiota and their potential interactions with weight status in diverse populations are not well understood. This study examined the associations between overall diet quality, weight status, and the gut microbiota in a racially balanced sample of adult females.

      Methods

      Female participants (N=71) residing in Birmingham, Alabama provided demographics, anthropometrics, biospecimens, and dietary data in this observational study from March 2014 to August 2014, and data analysis was conducted from August 2017 to March 2019. Weight status was defined as a BMI (weight [kg]/height [m2]) <30 kg/m2 for non-obese participants and ≥30 kg/m2 for participants who were obese. Dietary data collected included an Automated Self-Administered 24-Hour recall and Healthy Eating Index-2010 (HEI-2010) score. Diet quality was defined as having a high HEI score (≥median) or a low HEI score (<median). The fecal microbiota was collected, and the 16S ribosomal RNA gene was amplified to profile the microbiota composition. Differences in diet quality based on weight status were assessed using 2-sample t-tests. The associations between diet quality, gut microbiota, and weight status were analyzed using negative binomial models.

      Results

      Participants (43 Black, 28 White) aged 40.39±13.86 years who were non-obese (56%) and obese (44%) were studied. Greater alpha diversity was observed among those with higher Healthy Eating Index scores (p=0.037) but did not differ by weight status. Higher abundances of Bacteroidetes (p=0.006) and Firmicutes (p=0.042) were associated with a higher HEI score. Higher Bacteriodetes levels were observed among non-obese (p=0.006).

      Conclusions

      Diet quality measured by the HEI was associated with alpha diversity of the gut microbiota among adult females. Abundances of phyla that have been linked with weight status (Bacteroidetes and Firmicutes) were positively associated with diet quality.

      INTRODUCTION

      The abundance of certain bacterial groups is a hallmark of lower risk of several metabolic and gastrointestinal diseases.
      • So D
      • Whelan K
      • Rossi M
      • et al.
      Dietary fiber intervention on gut microbiota composition in healthy adults: a systematic review and meta-analysis.
      Diet contributes nutrients that the human gut microbiota needs to survive,
      • So D
      • Whelan K
      • Rossi M
      • et al.
      Dietary fiber intervention on gut microbiota composition in healthy adults: a systematic review and meta-analysis.
      and a growing number of studies have considered the impacts of diet quality on the gut microbiome.
      • Bowyer RCE.
      • Jackson MA
      • Pallister T
      • et al.
      Use of dietary indices to control for diet in human gut microbiota studies.
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      Few studies have evaluated the interaction between weight status and dietary quality with characteristics of the gut microbiota,
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      and even fewer focus on individuals living in the Southeastern region of the U.S.
      Undigested dietary substrates are metabolized by the human gut microbiota, thus dietary intake can affect the microbe abundances. Namely, fiber survives the human digestive system and is a major fuel source for the microbiota. Although dietary fiber may not increase microbial diversity, it does selectively increase the abundance of bacterial groups associated with health.
      • So D
      • Whelan K
      • Rossi M
      • et al.
      Dietary fiber intervention on gut microbiota composition in healthy adults: a systematic review and meta-analysis.
      Over the last 2 decades, the gut microbiota associated with distinct diets worldwide have been described.
      • Holscher HD.
      Dietary fiber and prebiotics and the gastrointestinal microbiota.
      ,
      • Xu Z
      • Knight R.
      Dietary effects on human gut microbiome diversity.
      Overall, plant-based rural and hunter-gatherer diets are characteristically higher in dietary fiber, leading to a higher abundance of phylum Bacteroidetes and genus Prevotella than urban and Westernized diets.
      • Xu Z
      • Knight R.
      Dietary effects on human gut microbiome diversity.
      Westernized diets are typically higher in fat, animal proteins, and refined carbohydrates while lower in fiber; however, variability exists within the diets of Westernized countries.
      • Holscher HD.
      Dietary fiber and prebiotics and the gastrointestinal microbiota.
      The amounts and types of carbohydrates, dietary fibers, fats, and proteins consumed can alter both harmful and beneficial bacteria in the gut.
      • So D
      • Whelan K
      • Rossi M
      • et al.
      Dietary fiber intervention on gut microbiota composition in healthy adults: a systematic review and meta-analysis.
      ,
      • Holscher HD.
      Dietary fiber and prebiotics and the gastrointestinal microbiota.
      ,
      • Yang Q
      • Liang Q
      • Balakrishnan B
      • Belobrajdic DP
      • Feng QJ
      • Zhang W.
      Role of dietary nutrients in the modulation of gut microbiota: a narrative review.
      The observed variance in both the diet and microbiota has sparked interest in the impact of the overall diet quality on the microbiota.
      • Bowyer RCE.
      • Jackson MA
      • Pallister T
      • et al.
      Use of dietary indices to control for diet in human gut microbiota studies.
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      ,
      • Liu Y
      • Ajami NJ
      • El-Serag HB
      • et al.
      Dietary quality and the colonic mucosa-associated gut microbiome in humans.
      Conformance to dietary patterns or guidelines can be quantified by diet indices. The Healthy Eating Index (HEI), based on the U.S. Department of Agriculture Dietary Guidelines for Americans (DGA),
      • Guenther PM
      • Casavale KO
      • Reedy J
      • et al.
      Update of the healthy eating index: HEI-2010.
      along with other dietary indices are beginning to be explored in relation to the microbiota. Diet quality is linked to the abundance of some bacterial groups and some studies
      • Bowyer RCE.
      • Jackson MA
      • Pallister T
      • et al.
      Use of dietary indices to control for diet in human gut microbiota studies.
      ,
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      ,
      • Liu Y
      • Ajami NJ
      • El-Serag HB
      • et al.
      Dietary quality and the colonic mucosa-associated gut microbiome in humans.
      but not all.
      • Garcia-Mantrana I
      • Selma-Royo M
      • Alcantara C
      • Collado MC.
      Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
      Furthermore, some studies show that alpha diversity, which is defined as a measure of microbiome diversity within a sample, increases with diet quality.
      • Bowyer RCE.
      • Jackson MA
      • Pallister T
      • et al.
      Use of dietary indices to control for diet in human gut microbiota studies.
      ,
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      ,
      • Liu Y
      • Ajami NJ
      • El-Serag HB
      • et al.
      Dietary quality and the colonic mucosa-associated gut microbiome in humans.
      In one of the largest ethnically diverse microbiota and diet studies, the Adiposity Phenotype Study cohort of the Multiethnic Cohort study, both alpha and beta diversity were positively associated with diet quality, with fruit and vegetable intake displaying the greatest impact on microbial diversity.
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      In this study, beta diversity is defined as a measure of similarity or dissimilarity of microbial composition between samples. However, there is a knowledge gap in the association between the gut microbiome and regional diets in the U.S.
      The association between gut microbiota and obesity
      • Rastelli M
      • Knauf C
      • Cani PD.
      Gut microbes and health: a focus on the mechanisms linking microbes, obesity, and related disorders.
      has been documented. Results of previous studies weaken the previous assumption that the microbiota of lean individuals and those who are obese is characterized by differences in phyla Firmicutes, Bacteroidetes, and F:B ratio. Instead, lifestyle factors may better explain the observed variation in Firmicutes and Bacteroidetes abundances.
      • Magne F
      • Gotteland M
      • Gauthier L
      • et al.
      The Firmicutes/Bacteroidetes ratio: a relevant marker of gut dysbiosis in obese patients?.
      In a study of Mediterranean diet adherence on the microbiota, Garcia-Mantrana et al.
      • Garcia-Mantrana I
      • Selma-Royo M
      • Alcantara C
      • Collado MC.
      Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
      observed phylum and genus level differences between the microbiota of normal weight and overweight individuals; however the interaction of diet and weight status was not considered. Given the complex interactions among the gut microbiota, metabolism, and lifestyle of its host,
      • Rastelli M
      • Knauf C
      • Cani PD.
      Gut microbes and health: a focus on the mechanisms linking microbes, obesity, and related disorders.
      a better understanding of the interplay between the microbiota and potentially obesogenic lifestyle of its host is needed.
      Adherence to the DGA is not optimal in the Southern region of the U.S.
      • Savoca MR
      • Arcury TA
      • Leng X
      • et al.
      The diet quality of rural older adults in the South as measured by Healthy Eating Index-2005 varies by ethnicity.
      • Yu D
      • Sonderman J
      • Buchowski MS
      • et al.
      Healthy eating and risks of total and cause-specific death among low-income populations of African-Americans and other adults in the Southeastern United States: a prospective cohort study.
      • McCabe-Sellers BJ
      • Bowman S
      • Stuff JE
      • Champagne CM
      • Simpson PM
      • Bogle ML.
      Assessment of the diet quality of U.S. adults in the Lower Mississippi Delta.
      • Hsiao PY
      • Mitchell DC
      • Coffman DL
      • et al.
      Dietary patterns and diet quality among diverse older adults: The University of Alabama at Birmingham Study of Aging.
      and may contribute to the disproportionate burden of obesity
      • Slack T
      • Myers CA
      • Martin CK
      • Heymsfield SB.
      The geographic concentration of U.S. adult obesity prevalence and associated social, economic, and environmental factors.
      and other chronic conditions observed in this region.
      • Barker LE
      • Kirtland KA
      • Gregg EW
      • Geiss LS
      • Thompson TJ.
      Geographic distribution of diagnosed diabetes in the U.S.: a diabetes belt.
      ,

      State Cancer Profiles. National Cancer Institute. https://statecancerprofiles.cancer.gov/index.html. Updated 2018. Accessed July 2017.

      The studies linking weight status, diet quality, and the gut microbiota are limited, particularly in the Southeastern region of the U.S. Previously, significant differences in overall diet quality and individual dietary components when comparing females who are obese and non-obese females
      • Little RB
      • Desmond R
      • Carson TL.
      Dietary intake and diet quality by weight category among a racially diverse sample of women in Birmingham, Alabama, USA.
      were reported from this observational study. The objectives of this study were (1) to test for the differences in the gut microbiota by weight status among a racially balanced sample of females, (2) to compare the gut microbiota by diet quality as measured by the HEI, and (3) to determine whether an interaction between weight status and HEI is associated with the characteristics of the gut microbiota. Although this study does not evaluate differences in the gut microbiota by race, the findings of this work give insight into the differences among a racially balanced sample of Black and White participants residing in the Southeastern region of the U.S.

      METHODS

      This study is a secondary analysis of data derived from a cross-sectional study of healthy female volunteers from the Birmingham, Alabama Metropolitan area. For the original study, participants provided demographic, anthropometric, survey, and dietary data to examine for associations with the gut bacteria using collected fecal samples. Methods and primary outcomes of this study
      • Carson TL
      • Wang F
      • Cui X
      • et al.
      Associations between race, perceived psychological stress, and the gut microbiota in a sample of generally healthy Black and White women: a pilot study on the role of race and perceived psychological stress.
      and a more thorough description of the diet quality of the sample
      • Little RB
      • Desmond R
      • Carson TL.
      Dietary intake and diet quality by weight category among a racially diverse sample of women in Birmingham, Alabama, USA.
      are reported elsewhere. For this secondary analysis, comparisons of the characteristics of the gut microbiota by diet quality measured by HEI are the outcomes of interest.

      Study Population

      A total of 106 female participants enrolled in the original study between March 2014 and August 2014. Inclusion criteria included being either a self-reported non-Hispanic Black or non-Hispanic White female aged ≥19 years. Individuals were excluded if they were pregnant or smokers at the time of screening, were unable to read or write, had a previous cancer diagnosis, or had taken medications known to alter the gut (e.g., proton-pump inhibitors) or antibiotics in the previous 90 days. Two did not provide fecal samples, and 22 were excluded after reporting medications that are known to alter the microbiome. The University of Alabama at Birmingham IRB approved the study, and all participants provided written informed consent.

      Measures

      Collected demographic data included race, ethnicity, age, income, education, and marital status. Race and ethnicity were self-identified as non-Hispanic Black or non-Hispanic White. Demographics and anthropometrics were collected at the first visit, and diet and fecal samples were collected at the second visit, approximately 1 week later.
      The gut microbiota was analyzed by fecal samples. Details of collection and processing are published elsewhere.
      • Carson TL
      • Wang F
      • Cui X
      • et al.
      Associations between race, perceived psychological stress, and the gut microbiota in a sample of generally healthy Black and White women: a pilot study on the role of race and perceived psychological stress.
      In short, samples were collected with a wipe and frozen until delivered to the microbiome laboratory. A fecal DNA isolation kit from Zymo Research was used to isolate the microbial genomic DNA. Polymerase chain reaction was then used on the prepared DNA samples with unique bar-coded primers to amplify the Variable Region 4 region of the 16S ribosomal DNA gene to create an amplicon library from individual samples. The polymerase chain reaction product was ∼255 bases from the Variable Region 4 segment of the 16S ribosomal DNA gene, and 251 base paired-end reads were sequenced using Illumina MiSeq. Analysis of the sequence data utilized the QIIME-based QWRAP pipeline
      • Kumar R
      • Eipers P
      • Little RB
      • et al.
      Getting started with microbiome analysis: sample acquisition to bioinformatics.
      to produce sample operational taxonomic unit (OTU) tables. Analysis included quality control, merging of paired-end reads, OTU picking, and taxonomic assignment.
      Trained research personnel measured participants’ weight and height using a standardized protocol described in a previous article
      • Carson TL
      • Wang F
      • Cui X
      • et al.
      Associations between race, perceived psychological stress, and the gut microbiota in a sample of generally healthy Black and White women: a pilot study on the role of race and perceived psychological stress.
      published from this observational study. Weight and height were measured using a calibrated 2-in-1 measuring station (Seca 284 measuring station, Hanover, MD) in light-weight clothing without shoes. BMI was calculated as weight (kg)/height (m2).
      One 24-hour diet recall was collected by the National Cancer Institute Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24), version 2014, which was collected in person with the assistance of a trained data collector. ASA24 is administered as a multiple-pass standardized interview and provides a series of prompts with multi-level food probes to assess food types and amounts. The program computes total intake, macronutrient composition, and nutrient and food group estimates. Because this study included one 24-hour recall, a more conservative approach was used to avoid days of more extreme intake. Participants who did not report calories within 600‒4,400 kcal, outliers determined by National Health and Nutrition Examination Survey data,

      Reviewing & cleaning ASA24 data. National Cancer Institute Division of Cancer Control & Population Sciences. https://epi.grants.cancer.gov/asa24/resources/cleaning.html. Updated January 18, 2017. Accessed August 2017.

      were excluded from the current analysis (n=2). After reviewing triaged records that did not meet the National Health and Nutrition Examination Survey guidelines for portion and nutrient outliers (grams of protein and fat, vitamin C, and β-carotene), an additional n=9 participants with extreme diets or believed errors in reported food were excluded from the current analysis.
      The HEI-2010 scoring system was used to assess diet quality. HEI-2010 is an a priori scoring system based on the 2010 DGA, the most current recommendations at the time of the study. The DGA promote whole-grain consumption; fruit and vegetable intake; attaining protein through a variety of sources (i.e., plants and seafood); and limiting empty calories, sodium, and refined grains while allowing flexibility in eating patterns for individual preferences, cultural and ethnic influences, and vegetarianism.
      • Guenther PM
      • Casavale KO
      • Reedy J
      • et al.
      Update of the healthy eating index: HEI-2010.
      ,
      U.S. Department of Agriculture, HHS
      Dietary guidelines for Americans.
      The HEI-2010 separates food intake into 12 components: 9 adequacy components (total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and fatty acid ratio) score high by reaching recommended intake, and 3 moderation components (refined grains, sodium, and empty calories) receive high scores for maintaining moderation. Intake evaluation is density based (food component intake per 1,000 kcal or percentage of calories) and is scored proportionally when between minimum and maximum standards. Age and sex alter daily requirements; therefore, HEI uses the least restrictive recommendation to award the optimal score.
      • Guenther PM
      • Casavale KO
      • Reedy J
      • et al.
      Update of the healthy eating index: HEI-2010.
      The maximum total score is 100 points, with a higher score signifying closer compliance with 2010 DGA recommendations
      • Guenther PM
      • Casavale KO
      • Reedy J
      • et al.
      Update of the healthy eating index: HEI-2010.
      and a score ≥80 is ideal.
      • Savoca MR
      • Arcury TA
      • Leng X
      • et al.
      The diet quality of rural older adults in the South as measured by Healthy Eating Index-2005 varies by ethnicity.
      Base code available online

      Healthy Eating Index: Choosing a Method. National Cancer Institute Division of Cancer Control & Population Sciences. https://epi.grants.cancer.gov/hei/tools.html. Updated April 17, 2017. Accessed July 2017.

      was used to convert the data generated by ASA24 to individual HEI-2010 component scores and total score. HEI scores were used to categorize participants into 2 groups (low-diet quality and high-diet quality) to assess differences in gut microbiota on the basis of diet quality.

      Statistical Analysis

      Descriptive statistics were calculated for all participants and stratified by 2 weight status groups: non-obese (BMI <30 kg/m2) or obese (BMI ≥30 kg/m2). Means and SDs were calculated for continuous variables. Frequencies and percentages were calculated for all categorical variables. Mean differences in anthropometric and dietary measures were assessed on the basis of weight status using 2-sample t-tests. Associations between categorical variables were evaluated using chi-square tests. Statistical significance was accepted when p<0.05 (2-tailed).
      Analysis of alpha and beta diversities were performed using R Package phyloseq.
      • McMurdie PJ
      • Holmes S.
      phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data.
      Descriptive statistics for alpha diversity were assessed for 2 types of stratified groups: weight status (non-obese with BMI <30 versus obese with BMI ≥30) and HEI status (HEI greater than the median versus HEI lower or equal to the median). The median, first quartile, and third quartile were calculated for 5 alpha diversity indexes: observed, Chao1, Abundance-based Coverage Estimator, Shannon, and Simpson.
      • Carson TL
      • Wang F
      • Cui X
      • et al.
      Associations between race, perceived psychological stress, and the gut microbiota in a sample of generally healthy Black and White women: a pilot study on the role of race and perceived psychological stress.
      Shapiro‒Wilk tests were used to evaluate normality and appropriate statistical tests for evaluating between-group differences. Two sample t-tests were used when normality assumptions were met, and Mann‒Whitney U-tests were used as the nonparametric test. Statistical significance accepted with p<0.05. Beta diversity distance metrics were evaluated using nonmetric multidimensional scaling. Differences in microbiota based on the 2 types of stratified groups―weight status or HEI status―were analyzed. A total of 3 beta diversity metrics were used: Bray‒Curtis, Weighted UniFrac, and Unweighted UniFrac.
      • Carson TL
      • Wang F
      • Cui X
      • et al.
      Associations between race, perceived psychological stress, and the gut microbiota in a sample of generally healthy Black and White women: a pilot study on the role of race and perceived psychological stress.
      All analyses were performed using OTU count data and the R package phyloseq. Negative binomial models in R package BhGLM were used to assess the associations between diet variables and OTUs at the phylum and genus levels.
      • Pendegraft AH
      • Guo B
      • Yi N.
      Bayesian hierarchical negative binomial models for multivariable analyses with applications to human microbiome count data.
      ,
      • Yi N
      • Tang Z
      • Zhang X
      • Guo B.
      BhGLM: Bayesian hierarchical GLMs and survival models, with applications to genomics and epidemiology.
      Models were adjusted for age, race, weight status, BMI, and waist circumference. First, all OTUs available were analyzed at the phylum and genus levels. False discovery rate was used, and statistical significance was accepted with false discovery rate p<0.05 level.
      All plots and figures were created using R Packages ggplot2 and vegan. All statistical analyses were performed using R (version 3.4.4).

      RESULTS

      This study included 71 female participants with a mean age of 40.4 years (SD=13.9 years) and a BMI of 30.6 kg/m2 (SD=8.9 kg/m2) (Table 1). The average BMI of non-obese participants and participants who were obese was 24.7 kg/m2 (SD=2.7 kg/m2) and 38.3 kg/m2 (SD=8.3 kg/m2), respectively. More than half reported some college education and a total household income ≥$40,000. A greater proportion of participants with obesity were Black than White (75% vs 25%; p=0.01).
      Table 1Participant Characteristics With Statistical Significance (p<0.05)
      CharacteristicsAll (N=71)Non-obese (n=40)Obese (n=31)p-value
      Age, mean (SD)40.4 (13.9)39.7 (14.8)41.3 (12.8)0.65
      BMI, mean (SD)30.6 (8.9)24.7 (2.7)38.3 (8.3)<0.01
      Education, n (%)0.24
       ≤High school7 (9.9)2 (28.6)5 (71.4)
       Some college or degree47 (66.2)29 (61.7)18 (38.3)
       Post graduate17 (23.9)9 (52.9)8 (47.1)
      Household income, n (%)0.26
       <$10,00014 (19.7)7 (50)7 (50)
       $10,000–$19,9995 (7.0)3 (60)2 (40)
       $20,000–$29,9993 (4.2)1 (33.3)2 (66.7)
       $30,000–$39,99912 (16.9)5 (41.7)7 (58.3)
       $40,000–$49,99915 (21.1)7 (46.7)8 (53.3)
       ≥$50,00022 (31.0)17 (77.3)5 (22.7)
      Black, n (%)43 (60.6)19 (44.2)24 (55.8)0.02
      Calories, kcal (SD)1,998.8 (1,032.7)2,041.3 (1,046.9)1,943.9 (1,028.7)0.70
      Carbohydrate, g (SD)241.9 (160.5)251.4 (170.5)229.7 (148.5)0.58
      Protein, g (SD)76.4 (34.6)80.5 (33.9)71.1 (35.3)0.26
      Fat, g (SD)83.8 (45.4)83.3 (45.5)84.3 (46.1)0.93
      Dietary fiber, g (SD)17.6 (9.9)17.9 (10.5)17.2 (9.1)0.77
      Total grains, oz eq (SD)5.5 (3.8)5.0 (2.7)6.2 (4.9)0.22
      Non-whole grain, oz eq (SD)4.8 (3.8)4.3 (2.7)5.5 (4.9)0.22
      Whole grains, oz eq (SD)0.7 (1.1)0.7 (1.0)0.7 (1.2)0.96
      Total vegetables, cup eq (SD)1.9 (1.4)1.9 (1.6)1.9 (1.1)0.92
      Total fruit, cup eq (SD)1.1 (1.2)1.5 (1.4)0.7 (0.8)0.01
      Saturated fatty acids, g (SD)27.2 (16.8)27.9 (17.7)26.2 (15.9)0.67
      Monounsaturated fatty acids, g (SD)30.9 (18.7)29.8 (18.4)32.3 (19.4)0.58
      Polyunsaturated fatty acids, g (SD)18.7 (11.4)18.7 (11.6)18.7 (11.3)0.99
      Total unsaturated fat, g (SD)49.6 (28.6)48.5 (28.2)51.0 (29.5)0.71
      HEI-2010 scores (maximum points), mean (SD)
       Vegetable intake (5)3.7 (1.5)3.5 (1.7)3.9 (1.1)0.19
       Dark green vegetable and bean intake (5)2.3 (2.3)2.2 (2.3)2.4 (2.4)0.68
       Fruit intake (5)2.6 (2.2)3.1 (2.2)2.0 (2.0)0.04
       Whole-fruit intake (5)2.4 (2.4)2.9 (2.4)1.7 (2.3)0.03
       Whole grain intake (10)2.4 (3.2)2.4 (3.2)2.3 (3.3)0.85
       Total Dairy intake (10)5.0 (3.5)5.2 (3.4)4.8 (3.7)0.62
       Total protein intake (5)4.1 (1.4)4.3 (1.2)3.9 (1.6)0.19
       Seafood and plant protein intake (5)1.9 (2.3)2.0 (2.4)1.6 (2.1)0.46
       Type of fat intake (10)5.0 (3.9)4.4 (3.9)5.7 (3.7)0.18
       Sodium intake (10)3.1 (3.4)3.5 (3.6)2.5 (3.1)0.22
       Refined grain intake (10)6.5 (3.6)6.8 (3.6)6.1 (3.6)0.40
       Empty calories (20)12.2 (6.1)12.8 (6.5)11.4 (5.4)0.34
       HEI-2010 total score (0–100)51.0 (16.2)53.2 (17.4)48.3 (14.3)0.21
      Note: Boldface indicates statistical significance (p<0.05).
      cup eq, cup equivalent; HEI, Healthy Eating Index; oz eq, ounce equivalent.
      Mean intake was 1,998.8 (SD=1,032.7) kcal, 76.4 (SD=34.6) g protein, 83.8 (SD=45.4) g fat, 241.9 (SD=160.5) g carbohydrate, and 17.6 (SD=9.9) g fiber. There were no significant differences in calorie, macronutrient, or fiber intake on the basis of weight status (Table 1). The mean HEI total score was 51.0 (SD=16.2) and did not differ by weight status. Fruit intake was greater in non-obese individuals than in individuals who were obese, both by the number of servings (1.5 cup equivalent vs 0.7 cup equivalent; p=0.01) and HEI scoring (3.1 vs 2.0; p=0.04). Whole-fruit intake was higher in non-obese females than in females who were obese (HEI score: 2.9 vs 1.7; p=0.03).
      The low HEI group had a higher BMI (33.4 vs 28.0; p=0.02) and higher caloric intake (2,257.8 vs 1,746.9; p=0.04). Fat, protein, and fiber intake did not differ, but the low HEI group reported a higher intake of carbohydrates (p=0.02). The high HEI group reported greater intake of vegetables (p=0.02), green vegetables and beans (p<0.001), total fruit (p<0.001), whole fruit (p<0.001), whole grains (p=0.04), total protein (p=0.02), and protein from seafood and plant sources (p<0.001); a better monounsaturated to saturated fat ratio (p<0.01); and greater intake of fewer refined grains (p<0.001) and empty calories (p<0.001).
      From fecal samples, 13 phyla, 123 genera, and 718 OTUs were identified. At the phylum level, the abundance of Bacteroidetes (65.06%) was greatest, followed by those of Firmicutes (24.02%) and Actinobacteria (10.92%) (Figure 1A). The most dominant genera were Lactobacillus (28.49%), Bacteroides (24.02%), Blautia (19.85%), Bifidobacterium (10.92%), Faecalibacterium (5.17%), Roseburia (4.46%), and other unclassified genera (7.09%) (Figure 1D).
      Figure 1
      Figure 1Microbial abundance comparisons of the most dominant groups. Microbial abundance at the phylum level is captured for (A) the general profile, (B) according to HEI status, and (C) according to weight status. At the genus level, the microbial abundance is shown for (D) the general profile, (E) according to HEI status, and (F) according to weight status.
      HEI, Healthy Eating Index.
      When comparing non-obese individuals with those who were obese, measures of alpha diversity indicated similar within-sample gut microbial diversity (Table 2). There was no evidence of differences in the overall microbial distribution across groups or beta diversity, as shown by the lack of clustering by group in Figure 2A and Figure 2B and the Bray‒Curtis (p=0.577), Unweighted Unifrac (p=0.461), and Weighted Unifrac (p=0.788) methods (Table 3 and Figure 2B).
      Table 2Measures of Alpha Diversity by Weight Status With p<0.05 Considered Statistically Significant
      Weight status
      MeasuresNon-obese (n=40)Median (Q1, Q3)Obese (n=31)Median (Q1, Q3)p-value
      Observed275 (259.5, 307.5)266 (247.5, 290.2)0.248
      Chao1335 (298.2, 361.4)315 (280.6, 346.6)0.143
      ACE327 (311.4, 358.9)307 (286.0, 343.0)0.077
      Shannon4 (3.2, 3.8)4 (3.3, 3.8)0.471
      Simpson0.916 (0.903, 0.957)0.944 (0.898, 0.956)0.553
      ACE, Abundance-based Coverage Estimator; Q, quartile.
      Figure 2
      Figure 2Illustration of an NMDS plot. Results are shown for (A) HEI status and (B) weight status.
      Notes: The NMDS analysis provides a visualization of the level of similarity of each observation in the data set on the basis of the distance matrix. Then, each observation is assigned a location in a low-dimensional space. Ellipses represent 95% CI around the centroid.
      HEI, Healthy Eating Index; NMDS, nonmetric multidimensional scaling.
      Table 3Measures of Beta Diversity by Weight Status and HEI Status With p<0.05 Considered Statistically Significant
      MeasuresWeight status p-valueHEI status p-value
      Bray‒Curtis0.5770.297
      Unweighted UniFrac0.4610.280
      Weighted UniFrac0.7880.071
      HEI, Healthy Eating Index.
      Alpha diversity assessed by the Chao1 (p=0.030) and Abundance-based Coverage Estimator (p=0.037) method displayed significant differences by HEI groups (Table 4). Beta diversity did not differ when comparing HEI groups using the Bray‒Curtis (p=0.297), Unweighted Unifrac (p=0.280), and Weighted Unifrac (p=0.071) methods (Table 3).
      Table 4Measures of Alpha Diversity by HEI Status With p<0.05 Considered Statistically Significant
      HEI status
      MeasuresLow (n=35)Median (Q1, Q3)High (n=36)Median (Q1, Q3)p-value
      Observed263 (246.0, 291.0)274 (260.0, 306.0)0.136
      Chao1300 (280.9, 341.9)333 (309.8, 361.4)0.030
      ACE306 (282.8, 341.4)331 (307.5, 364.2)0.037
      Shannon3.59 (3.24, 3.95)3.66 (3.29, 3.80)0.818
      Simpson0.930 (0.900, 0.959)0.939 (0.902, 0.953)0.868
      Note: Boldface indicates statistical signifiance (p<0.05).
      ACE, Abundance-based Coverage Estimator; HEI, Healthy Eating Index; Q, quartile.
      At the phylum level, better diet quality was inversely associated with Proteobacteria and TM7 (Figure 3A). No dietary components were significantly associated with the 2 most abundant phyla, Firmicutes and Bacteroidetes. Whole-grain intake was inversely associated with Proteobacteria. Whole-fruit intake was inversely associated with Verrucomicrobia. A higher ratio of unsaturated fats to saturated fats was inversely associated with Proteobacteria. HEI for fruit and empty calories was inversely associated with Proteobacteria and TM7. Intake of dark green vegetables and beans was inversely associated with Proteobacteria and Thermi. Diet interactions by weight status were significantly associated with the Proteobacteria phyla (Figure 3B). The relationship between Proteobacteria and whole-fruit intake was positively associated with weight status. Finally, the total HEI score was significantly and positively associated with weight status and Firmicutes.
      Figure 3
      Figure 3Associations between diet and all OTUs at the phylum level: (A) main effect of diet and (B) interaction between diet and weight status.
      Note: A + sign indicates a (A) positive main effect or (B) positive interaction.
      OTU, Operational Taxonomic Unit.
      The associations between OTUs, diet, and weight status were explored at the genus level (Figure 4) . Significant associations between diet quality and microbiota groups were found after assessing fruit and fiber intake. HEI fruit score was inversely associated with Akkermansia (p<0.05) and Gernella (p<0.01). Bifidobacterium was positively associated with HEI fruit score (p<0.01) and showed a negative interaction between weight status and fruit intake (p<0.01). Total and HEI scores for whole grain were positively associated with Dorea (p<0.01), but an interaction between weight status and intake did not significantly relate to Dorea abundance. However, the genus Prevotella was negatively associated with whole-grain intake that differed on the basis of weight status group. HEI vegetable score was inversely associated with Lactococcus (p<0.05). Dietary fat intake was strongly associated with several species in the gut microbiota. Fat intake was inversely associated with Prevotella, Escherichia, and Lachnospira (p<0.01). HEI fat ratio, total unsaturated fats (monosaturated and polyunsaturated fat total), and polyunsaturated fats were inversely associated with Lactobacillus (p<0.01). Furthermore, total fat, mono and poly fats, and monounsaturated fats were positively associated with both Haemophilus (p<0.001) and Odoribacter (p<0.01). Finally, Devosia was negatively associated with vegetable intake that differed on the basis of weight status group (p<0.01).
      Figure 4
      Figure 4Associations between OTUs and diet with (A) the main effect of diet and interaction between (B) diet and weight status at the genus level.
      Note: + signs indicate a (A) positive main effect or (B) positive interaction.
      OTU, Operational Taxonomic Unit.

      DISCUSSION

      This study assessed the differences in the gut microbiota by diet quality and weight status among a racially balanced sample of 71 females living in Birmingham, Alabama. Greater alpha diversity was associated with a higher diet quality score but was not related to weight status. Beta diversity did not differ on the basis of weight status nor on the basis of diet quality. Several gut bacterial groups were associated with individual dietary components. At the phylum level, Bacteriodetes and Firmicutes were abundant but were not found to be significantly associated with any specific dietary intake in this sample. However, the phylum Proteobacteria was found to be inversely related to whole grain, fat, dark green vegetables and beans, fruit, and better adherence to limited empty calorie intake. Higher HEI scores were associated with lower abundances of Proteobacteria and TM7. At the genus level, Bifidobacteria was positively related to fruit intake and was inversely linked to weight status. This finding was consistent with another study that found that Bifidobacteria was associated with a plant-based diet.
      • Garcia-Mantrana I
      • Selma-Royo M
      • Alcantara C
      • Collado MC.
      Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
      Several fats such as total fat, unsaturated fats (mono and poly fat), and monounsaturated fats were positively associated with Lactobacillus, Haemophilus, and Odoribacter. Notably, Lactobacillus has been studied in other contexts to examine irritable bowel syndrome, and thus this bacteria has been documented in nutrition and health literature. The HEI score for fat intake was inversely related to Prevotella, Escherichia, and Lachnospira. This study did not find differences between beta diversity on the basis of diet quality or weight status. However, these findings are consistent with other reports that higher diet quality assessed by HEI is associated with greater alpha diversity.
      • Bowyer RCE.
      • Jackson MA
      • Pallister T
      • et al.
      Use of dietary indices to control for diet in human gut microbiota studies.
      ,
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      Similar to the study by Maskarinec and colleagues,
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      this study did not observe any significant diet associations with Firmicutes. In contrast to the study by Maskarinec et al.
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      who observed an inverse association between Actinobacteria and diet quality, this study only noted a nonsignificant trend.
      Previously, fruit and vegetable intake has been found to be most beneficial to alpha diversity.
      • Maskarinec G
      • Hullar MAJ
      • Monroe KR
      • et al.
      Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
      In this sample, fruit or vegetable intake was associated with increased Bifidobacterium and inversely associated with Proteobacteria, Akkermansia, and Verrucomicrobia. Polysaccharides such as starches and fiber found in fruit are associated with higher Bifidobacterium.
      • Garcia-Mantrana I
      • Selma-Royo M
      • Alcantara C
      • Collado MC.
      Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
      This finding that Proteobacteria was associated with consumption of dietary fats is consistent with another study that found a higher abundance of Proteobacteria in children with obesity or a high fat intake.
      • Méndez-Salazar EO
      • Ortiz-López MG
      • MLÁ Granados-Silvestre
      • Palacios-González B
      • Menjivar M
      Altered gut microbiota and compositional changes in Firmicutes and Proteobacteria in Mexican undernourished and obese children.
      A positive association was observed between total fat and healthy unsaturated fats and Haemophilus. In contrast, Garcia-Mantrana and colleagues
      • Garcia-Mantrana I
      • Selma-Royo M
      • Alcantara C
      • Collado MC.
      Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
      observed that less healthy saturated fats were associated with Haemophilus. Reasons for this contradictory finding may be a result of using different metrics for assessing diet quality given that Garcia-Mantrana et al.
      • Garcia-Mantrana I
      • Selma-Royo M
      • Alcantara C
      • Collado MC.
      Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
      examined the Mediterranean diet, and this study used the HEI-2010.

      Limitations

      This study has a few limitations. The small sample limits the strength to detect potentially significant relationships between diet, weight status, and the gut microbiota. Hence, a larger sample would strengthen the ability to detect potential differences in gut microbiota, diet, and weight status. Self-reported dietary intake may also allow for recall bias or inaccurate reporting by participants, thus the decision to exclude participants with implausible dietary recalls. This study also has significant strengths that outweigh the limitations. This sample has an equivalent proportion of females who are obese and non-obese females and non-Hispanic White and Black females representing the Southeastern region of the U.S., which allows us to explore the gut microbial differences in a diverse, understudied population with a regional-specific diet. With this sample, these findings provide valuable insight into chronic disease risk in a potentially high-risk, underrepresented, and vulnerable population because 20% of the participants reported <$10,000 household income.

      CONCLUSIONS

      These findings suggest strong correlations between diet quality and the gut microbiota in a racially balanced sample population. Specifically, whole grain, fiber, fruit, vegetable, and fat intake were strongly linked to bacterial groups related to health outcomes. As more information is discovered about signatures for optimal health‒promoting gut microbiota, these findings will help to inform strategies to target for modification of the gut microbiota. This research will offer insight into the behavioral and potential physiologic pathways for chronic disease prevention and progression.

      ACKNOWLEDGMENTS

      The authors wish to acknowledge funding that supported this work. TLC and RBL thank the NIH Center for Clinical and Translational Science/National Center for Advancing Translational Sciences grant UL1TR003096, The University of Alabama at Birmingham Cancer Research Experiences for Students National Cancer Institute/NIH R25CA076023-17, and The University of Alabama at Birmingham Comprehensive Cancer Center Young Supporters Board Young Investigator Award for supporting this research. ALM thanks the NIH/National Heart, Lung, and Blood Institute grant number T32HL072757, which supported her postdoctoral fellowship during this research.
      No financial disclosures were reported by the authors of this paper.

      CRediT AUTHOR STATEMENT

      Tiffany L. Carson: conceptualization, investigation, methodology, writing - original draft. Rebecca B. Little: conceptualization, data curation, writing - original draft. Anarina L. Murillo: formal analysis, visualization, writing - original draft. Nengjun Yi: formal analysis, writing - review and editing. Casey D. Morrow: data curation, validation, writing - review and editing. Elliot J. Lefkowitz: data curation, validation, writing - review and editing. William J. Van Der Pol: data curation, validation, writing - review and editing.

      SUPPLEMENT NOTE

      This article is part of a supplement entitled Obesity-Related Health Disparities: Addressing the Complex Contributors, which is sponsored by the National Institute on Minority Health and Health Disparities (NIMHD), National Institutes of Health (NIH), U.S. Department of Health and Human Services (HHS). The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of NIMHD, NIH, or HHS.

      REFERENCES

        • So D
        • Whelan K
        • Rossi M
        • et al.
        Dietary fiber intervention on gut microbiota composition in healthy adults: a systematic review and meta-analysis.
        Am J Clin Nutr. 2018; 107: 965-983https://doi.org/10.1093/ajcn/nqy041
        • Bowyer RCE.
        • Jackson MA
        • Pallister T
        • et al.
        Use of dietary indices to control for diet in human gut microbiota studies.
        Microbiome. 2018; 6: 77https://doi.org/10.1186/s40168-018-0455-y
        • Garcia-Mantrana I
        • Selma-Royo M
        • Alcantara C
        • Collado MC.
        Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population.
        Front Microbiol. 2018; 9: 890https://doi.org/10.3389/fmicb.2018.00890
        • Maskarinec G
        • Hullar MAJ
        • Monroe KR
        • et al.
        Fecal microbial diversity and structure are associated with diet quality in the Multiethnic Cohort Adiposity Phenotype Study.
        J Nutr. 2019; 149: 1575-1584https://doi.org/10.1093/jn/nxz065
        • Holscher HD.
        Dietary fiber and prebiotics and the gastrointestinal microbiota.
        Gut Microbes. 2017; 8: 172-184https://doi.org/10.1080/19490976.2017.1290756
        • Xu Z
        • Knight R.
        Dietary effects on human gut microbiome diversity.
        Br J Nutr. 2015; 113: S1-S5https://doi.org/10.1017/S0007114514004127
        • Yang Q
        • Liang Q
        • Balakrishnan B
        • Belobrajdic DP
        • Feng QJ
        • Zhang W.
        Role of dietary nutrients in the modulation of gut microbiota: a narrative review.
        Nutrients. 2020; 12: 381https://doi.org/10.3390/nu12020381
        • Liu Y
        • Ajami NJ
        • El-Serag HB
        • et al.
        Dietary quality and the colonic mucosa-associated gut microbiome in humans.
        Am J Clin Nutr. 2019; 110: 701-712https://doi.org/10.1093/ajcn/nqz139
        • Guenther PM
        • Casavale KO
        • Reedy J
        • et al.
        Update of the healthy eating index: HEI-2010.
        J Acad Nutr Diet. 2013; 113 ([published correction appears in J Acad Nutr Diet. 2016;116(1):170]): 569-580https://doi.org/10.1016/j.jand.2012.12.016
        • Rastelli M
        • Knauf C
        • Cani PD.
        Gut microbes and health: a focus on the mechanisms linking microbes, obesity, and related disorders.
        Obesity (Silver Spring). 2018; 26: 792-800https://doi.org/10.1002/oby.22175
        • Magne F
        • Gotteland M
        • Gauthier L
        • et al.
        The Firmicutes/Bacteroidetes ratio: a relevant marker of gut dysbiosis in obese patients?.
        Nutrients. 2020; 12: 1474https://doi.org/10.3390/nu12051474
        • Savoca MR
        • Arcury TA
        • Leng X
        • et al.
        The diet quality of rural older adults in the South as measured by Healthy Eating Index-2005 varies by ethnicity.
        J Am Diet Assoc. 2009; 109: 2063-2067https://doi.org/10.1016/j.jada.2009.09.005
        • Yu D
        • Sonderman J
        • Buchowski MS
        • et al.
        Healthy eating and risks of total and cause-specific death among low-income populations of African-Americans and other adults in the Southeastern United States: a prospective cohort study.
        PLoS Med. 2015; 12e1001830https://doi.org/10.1371/journal.pmed.1001830
        • McCabe-Sellers BJ
        • Bowman S
        • Stuff JE
        • Champagne CM
        • Simpson PM
        • Bogle ML.
        Assessment of the diet quality of U.S. adults in the Lower Mississippi Delta.
        Am J Clin Nutr. 2007; 86: 697-706https://doi.org/10.1093/ajcn/86.3.697
        • Hsiao PY
        • Mitchell DC
        • Coffman DL
        • et al.
        Dietary patterns and diet quality among diverse older adults: The University of Alabama at Birmingham Study of Aging.
        J Nutr Health Aging. 2013; 17: 19-25https://doi.org/10.1007/s12603-012-0082-4
        • Slack T
        • Myers CA
        • Martin CK
        • Heymsfield SB.
        The geographic concentration of U.S. adult obesity prevalence and associated social, economic, and environmental factors.
        Obesity (Silver Spring). 2014; 22: 868-874https://doi.org/10.1002/oby.20502
        • Barker LE
        • Kirtland KA
        • Gregg EW
        • Geiss LS
        • Thompson TJ.
        Geographic distribution of diagnosed diabetes in the U.S.: a diabetes belt.
        Am J Prev Med. 2011; 40: 434-439https://doi.org/10.1016/j.amepre.2010.12.019
      1. State Cancer Profiles. National Cancer Institute. https://statecancerprofiles.cancer.gov/index.html. Updated 2018. Accessed July 2017.

        • Little RB
        • Desmond R
        • Carson TL.
        Dietary intake and diet quality by weight category among a racially diverse sample of women in Birmingham, Alabama, USA.
        J Nutr Sci. 2020; 9: e58https://doi.org/10.1017/jns.2020.51
        • Carson TL
        • Wang F
        • Cui X
        • et al.
        Associations between race, perceived psychological stress, and the gut microbiota in a sample of generally healthy Black and White women: a pilot study on the role of race and perceived psychological stress.
        Psychosom Med. 2018; 80: 640-648https://doi.org/10.1097/PSY.0000000000000614
        • Kumar R
        • Eipers P
        • Little RB
        • et al.
        Getting started with microbiome analysis: sample acquisition to bioinformatics.
        Curr Protoc Hum Genet. 2014; 82 (18.8.1–18.8.29)https://doi.org/10.1002/0471142905.hg1808s82
      2. Reviewing & cleaning ASA24 data. National Cancer Institute Division of Cancer Control & Population Sciences. https://epi.grants.cancer.gov/asa24/resources/cleaning.html. Updated January 18, 2017. Accessed August 2017.

        • U.S. Department of Agriculture, HHS
        Dietary guidelines for Americans.
        U.S. Department of Agriculture, HHS, Washington, DC2010 (Published December 2010. Accessed XXX)
      3. Healthy Eating Index: Choosing a Method. National Cancer Institute Division of Cancer Control & Population Sciences. https://epi.grants.cancer.gov/hei/tools.html. Updated April 17, 2017. Accessed July 2017.

        • McMurdie PJ
        • Holmes S.
        phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data.
        PLoS One. 2013; 8: e61217https://doi.org/10.1371/journal.pone.0061217
        • Pendegraft AH
        • Guo B
        • Yi N.
        Bayesian hierarchical negative binomial models for multivariable analyses with applications to human microbiome count data.
        PLoS One. 2019; 14e0220961https://doi.org/10.1371/journal.pone.0220961
        • Yi N
        • Tang Z
        • Zhang X
        • Guo B.
        BhGLM: Bayesian hierarchical GLMs and survival models, with applications to genomics and epidemiology.
        Bioinformatics. 2019; 35: 1419-1421https://doi.org/10.1093/bioinformatics/bty803
        • Méndez-Salazar EO
        • Ortiz-López MG
        • MLÁ Granados-Silvestre
        • Palacios-González B
        • Menjivar M
        Altered gut microbiota and compositional changes in Firmicutes and Proteobacteria in Mexican undernourished and obese children.
        Front Microbiol. 2018; 9 ([published correction appears in Front Microbiol. 2018;9:2693]): 2494https://doi.org/10.3389/fmicb.2018.02494