Recipient Organization
CORNELL UNIVERSITY
(N/A)
ITHACA,NY 14853
Performing Department
Nutritional Science
Non Technical Summary
Micronutrient malnutrition remains a major public health problem, affecting more than 2 billion people, both domestically and internationally. In particular, women of reproductive age are at a high risk for micronutrient deficiencies and anemia. Fortification of foods with folic acid and other micronutrients is an efficacious and cost-effective public health intervention to address micronutrient deficiencies, and has been implemented in over 85 countries. The impact of micronutrient-fortified foods on other health outcomes beyond micronutrient deficiencies is needed to further inform fortification policies and guidelines in the United States and globally - as any additional health benefits from a public health intervention could have a substantial impact on population health.Dietary intake of key micronutrients provides substrates that can be used by the gut microbiota, comprised of bacteria, fungi, and viruses. Not only can these dietary substrates modify the composition of microorganisms residing in the human gut, but they can also influence their metabolic function. Previous studies have explored the impact of dietary interventions on the gut microbiome among adults in higher income settings. However, no studies to date have focused on the impact of consuming micronutrient-fortified foods and nutritional status on the gut microbial composition, diversity, and function, among women of reproductive age.Our randomized controlled trial among 1,000 women of reproductive age in South India, funded by the Centers for Disease Control and Prevention (CDC), represents an ideal opportunity to evaluate the impact of multiple micronutrient-fortified salt on the gut microbiome (estimated date of completion: March 2023). This randomized trial will examine the efficacy of quadruple fortified salt (QFS; iron, iodine, vitamin B12, folic acid, and iodine), compared to double-fortified salt (DFS; iron, iodine), using a 2x2 factorial design, on anemia and micronutrient status among women of reproductive age over a 12-month period. We propose to add an assessment of the gut microbiome to the parent trial design, which is a cost-efficient approach to answer the question: "Does multiple micronutrient-fortified salt-- with iron, iodine, folic acid, and vitamin B12--impact the composition, diversity, and function of the gut microbiome over 12 months of ad libitum consumption?'. We aim to increase our understanding of the links between agricultural, nutritional, and dietary inputs and the gut microbiome in women of reproductive age (preconception) and help identify critical windows for intervention to improve women's health in the United States and globally. We hypothesize that the multiple micronutrient-fortified intervention in this trial will improve the composition and function of the gut microbiome in women of reproductive age.This proposal, in response to the priority of "Food Safety, Nutrition, and Health", will address the challenges of a growing diverse population and persisting micronutrient deficiencies using fortified foods. This represents a sustainable solution for improving micronutrient status, including human food and micronutrient needs. In terms of sustainability, fortified food can be produced by local producers, sustaining the economic viability of industry operations. Further, fortification is cost-effective and fortified foods such as salt are affordable, widely consumed, and can be incorporated into existing production and distribution systems.
Animal Health Component
50%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
0%
Goals / Objectives
Within the context of an established randomized controlled trial, we aimto increase our understanding of the links between agricultural, nutritional, and dietary inputs and the gut microbiome in women of reproductive age (preconception) and help identify critical windows for intervention to improve women's health in the United States and globally.Specifically, in a subset ofn= 250 women of reproductive age (15-40y) participating in a randomized efficacy trial of multiple micronutrient-fortified salt in Chittoor, South India,our objectives are to:Objective I.Characterize the taxonomic composition and functional profile of the gut microbiome using whole genome shotgun sequencing at baseline, midline, and endpoint, as part of an established 12-month randomized trial.Objective II.Determine the impact of multiple micronutrient-fortified salt (iron, iodine, folic acid, vitamin B12) on the composition and functional profile of the gut microbiome at baseline, midline, and endpoint, as part of an established 12-month randomized trial.
Project Methods
Our overall hypothesis is that the multiple micronutrient-fortified salt intervention will improve the composition and function of the gut microbiome, compared to standard fortified salt. Specifically, among 250 women of reproductive age (15 to 40 y) participating in a randomized efficacy trial of multiple micronutrient fortified salt in Southern India, we propose to:Objective I. Characterize the taxonomic composition and functional profile of the gut microbiome using whole genome shotgun sequencing at baseline, midline, and endpoint, as part of an established 12-month randomized trial.Objective II. Determine the impact of multiple micronutrient-fortified salt (iron, iodine, folic acid, vitamin B12) on the composition and functional profile of the gut microbiome at baseline, midline, and endpoint, as part of an established 12-month randomized trial.DNA extraction:After thawing, 300-mg aliquots of stool samples will be mixed with stool lysis buffer (Qiagen), transferred to a SK38 stool-grinding tube (Precellys), and homogenized and centrifuged. DNA will be extracted from a 1.2 mL sample of each resulting supernatant (QIAmp stool kit, Qiagen) according to the manufacturer's instructions. Samples will be quantified (Quant-iT PicoGreen dsDNA assay kit [Life Technologies]); the total amount of resulting DNA is predicted to be about 3 µg (158).Library preparation and sequencing:Barcoded DNA fragment libraries will be generated with 0.25 ng of input DNA using a KAPA HyperPlus (Roche) sample preparation kit per manufacturer's instructions, as the Knight Lab's internal benchmarks have shown this kit to perform better than Nextera for assembly. A BioAnalyzer (Agilent) will determine the distribution of fragment sizes (predicted to be between 430 to 990 bp). Samples will be pooled with final, single-stranded normalized libraries and sequenced on Illumina MiSeq to obtain the cluster number and stoichiometric distribution of each sample, pooled again and sequenced using NextSeq 500 to obtain a set of sequencing reads representing sample-specific metagenomes.Bioinformatics:Microbial composition and function:Raw data from whole genome sequencing will be run through the bioinformatics tool MetaPhlAn3 (metagenomic phylogenetic analysis) to assign taxonomic profiles to each sample, with resolution to the species level. MetaPhlAn relies on unique clade-specific marker genes identified from ~17,000 reference genomes (~13,500 bacterial and archaeal, ~3,500 viral, and ~110 eukaryotic), allowing up to 25,000 reads-per-second (on one CPU) analysis speed (orders of magnitude faster compared to existing methods), unambiguous taxonomic assignments as the MetaPhlAn markers are clade-specific, accurate estimation of organismal relative abundance (in terms of number of cells rather than fraction of reads), species-level resolution for bacteria, archaea, eukaryotes and viruses, and extensive validation of the profiling accuracy on several synthetic datasets and on thousands of real metagenomes. StrainPhlAn3 and PanPhlAn3 will be used for strain-level profiling. QIIME2 will be used to calculate and β diversity measures, including the Shannon Diversity Index and Chao1, and principal coordinates analysis will be used to visualize differences in β diversityviathe Emperor framework within QIIME2 (111). Functional capacity of the microbiota will be determined by metagenomic metabolic reconstruction using HUMAnN2 (159, 160), a bioinformatics tool that performs quality filtering, removes human DNA, normalizes the abundances of genes, assigns genes to pathways using MinPath (161), and accounts for noise and gaps in the data (159).Statistical Analyses:Descriptive statistics: Demographic characteristics at baseline and at each timepoint will be compared usingt-tests or ANOVA for continuous outcomes and χ2tests for categorical outcomes, or their nonparametric equivalents. Differences in bacterial taxa and host characteristics at each time point and changes across timepoints will be compared by ANOVA and ANOCVA, respectively, with repeated measures and adjusting for the intervention arm, in combination with Tukey'spost-hoctests. A p-value of <0.05 will be considered statistically significant. To assess the effect of consuming multiple micronutrient-fortified salt, compared to standard double-fortified salt, on gut microbial community composition between study arms at each time-point, as well as changes within each arm over the 12-month feeding period, we will investigate relative abundance, ? diversity, β diversity, and functional capacity of the sequence data. Linear regression: MaAsLin2 will also be applied to determine multivariate associations between the trial's intervention, clinical, and dietary metadata and microbial features. We will account for repeated measures in linear mixed models by including participants as a random effect. For objective 1, intervention assignment will be included as covariates in regression models, and dietary intake of micronutrients and macronutrients will also be adjusted for total energy. We will adjust for multiple comparisons using Benjamini-Hochberg False discovery rate (FDR). Associations will be visualized using heat maps. β-diversity: For β-diversity (i.e., UniFrac, Bray-Curtis, and DEICODE) of taxa and metabolic pathways, we will use PERMANOVA to assess the percentage contribution of sociodemographic and health characteristics to β-diversity variation. Longitudinal inferential analysis: ABayesian Sparse Functional Principal Components Analysis (SFPCA) methodologywill be used to model dynamic temporal change and to detect its dependence on biological covariates in longitudinal microbiome analysis, which is a new method developed by the Knight Lab and Thompson Lab at UCSD which was presented at JSM (170) and will be available as a preprint in bioRxiv (171). In simulation and in real data applications, Bayesian SFPCA is able to overcome irregular sampling intervals, limited sample size, missing values and dropouts, making it appropriate for our randomized trial study design. Currently, the SFPCA method is appropriate for univariate associations, but extending this to multivariate models is currently in development. SFPCA will be carried out using R, code available on GitHub (172). An example of re-analyzed data from the early childhood and microbiome (ECAM) study (173) is shown inFigure 3 (A, B)modeling 0-2-year-old infant maturation of the gut microbiome based on Shannon Diversity Index (171). Comparing infant diet (breastmilk vs. formula), Bayesian SFPCA detected a significant difference in all three principal components (PCs) with p-values all less than 0.001 [Figure 3 (C-E)], suggesting that the difference between formula-fed and breastfed infants can belong to any of the temporal pattern shown in three PCs (Figure 8b-d) but with PC 3 the least distinguished (171). In contrast, there were no significant difference between breastfed and formula-fed infants using another method, splinectomeR estimated mean curves (174), despite there being significant differences in the growth rate of Shannon Diversity using loess estimated mean curves. Age will be explicitly accounted for as a variable using random forests regression vs. the whole dataset, a method that was introduced to the microbiome field by the Knight Lab with application to estimating time since death of corpse (176, 177) and most recently incorporated covariation among taxa (178). We will use these techniques to detrend for age, which we have used successfully in older adults (179).