Recipient Organization
CORNELL UNIVERSITY
(N/A)
ITHACA,NY 14853
Performing Department
(N/A)
Non Technical Summary
Micronutrient deficiencies are important threats to maternal and child health.In our population-based cohorts in Southern India, we have identified a high burden of anemia and micronutrient deficiencies(including iron, folate, andvitamin B12 )in women of reproductive age, pregnant women, children, and adolescents. Iron and folic acid supplementation is recommended during pregnancy, for prevention of anemia andadverse pregnancy outcomes.However, the impact of micronutrient fortificationinterventions on other aspects of health in children,remainsa critical gap for informing fortification guidelines. We are conducting arandomized trial to evaluate theeffects of micronutrient-fortified salt intervention onanemia and micronutrient status in women of reproductive age in1,000 households in Southern India over a 12-month period.We propose to examine the impact of micronutrient-fortified salt on on anemia and other health outcomes in children who are already receiving this household-based intervention.Additionally, we will use metabolomics to examine mechanisms that can explain the effects of the micronutrient-fortified salt intervention on health outcomesin children.Findings will establish the effects of long-term consumption of micronutrient-fortified foods on child health outcomes in a population that does not currently have mandatory fortification. This research will generate evidence to inform fortification programs - as low-cost strategies that reach the most at-risk populations to improve micronutrient status and health outcomes- both in the U.S. and globally.
Animal Health Component
50%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
0%
Goals / Objectives
GoalsWe propose to examine the effects of micronutrient-fortified salt on gut microbiome and health outcomes in childrenwho are already receiving micronutrient-fortified salt as part of a household-based intervention. Additionally, we will use metabolomics to examine mechanisms that can explain the effects of the micronutrient-fortified salton health outcomes in children. Findings will establish the effects of long-term consumption of micronutrient-fortified foods on child health outcomes in a population that does not currently have mandatory micronutrientfortification. This research will generate evidence to inform fortification programs - as low-cost strategies that reach the most at-risk populations to improve micronutrient status while potentially benefiting metabolic health - both in the United States and globally.ObjectivesWe aim toevaluate the impact of micronutrient-fortified salton gut microbiome and health outcomes in childrenresiding in households participating in a randomized trial in Southern India.Objective I. To evaluate the effects of themicronutrient-fortified saltintervention on anemia and other health outcomes in 500 childrenover a 12-month period. Objective II. To evaluate the effects of the micronutrient-fortified saltintervention on composition and functional profile of the gut microbiome prospectively as part of a 12-month randomized trial in a subset of 200 children. Objective III. To use stool metabolomics to examine potential mechanisms linking components of the microbiome affected by themicronutrient-fortified salt interventionand health outcomes in children.
Project Methods
We plan to evaluate the impact of a micronutrient-fortified salton gut microbiome and health outcomes in childrenwho are already receiving the household-based intervention as part of an established randomized trial in Southern India. Objective I. To evaluate the effects of themicronutrient-fortified saltintervention on anemia and other health outcomes in 500 childrenover a 12-month period.Statistical Analyses: The effect of the micronutrient-fortified salt intervention on health outcomes will be evaluated using intention-to-treat analyses. Continuous outcomes will be presented as geometric means (95% CI), based on the distributions in the population. Non-normally distributed variables will be transformed prior to analyses. Categorical outcomes will be presented as n (%). Generalized estimating equations with an exchangeable correlation matrix will be usedto evaluate the effects of the intervention onoutcomes. Statistical analyses will be performed using SAS v9.4.Objective II. To evaluate the effects of themicronutrient-fortified saltintervention on composition and functional profile of the gut microbiome prospectively as part of a 12-month trial in a subset of 200 children.Laboratory analyses:DNA and molecular extraction:After thawing, 300-mg aliquots of the stool sample will be spun down and separated for LC/MS/MS and DNA extraction. The pellet will be added to stool lysis with a custom bead preparation that improves efficiency of extraction across taxa and extracted on the Kingfisher platform,then quantitated using Quant-iT PicoGreen dsDNA assay kit as per the manufacturers' instructions.Library preparation and sequencing:Library preparation will be performed using miniaturized KAPA HyperPlus as in our Leaderboard Sequencing paper,using the epMotion, Mosquito and Echo liquid handling platforms. Multiplex pools will be prepared, and the fragments size distribution of the final sequencing library pools will be analyzed on Agilent 2200 TapeStation. Samples will be pooled with final, single-stranded normalized libraries and sequenced on Illumina iSeq to obtain each sample's cluster number and stoichiometric distribution, pooled again, and sequenced using NovaSeq to obtain a set of sequencing reads representing sample-specific metagenomes. Formetabolomics, we will use an UltiMate 3000 UHLPC system with a reverse phase C18 column prepended with a guard cartridge, at a column compartment temperature of 40°C. Samples will be chromatographically separated with a constant flow rate of 0.5 mL per minute. The UHPLC system is coupled to a Maxis Q-TOF Impact II mass spectrometer equipped with an electrospray ionization source. MS spectra will be acquired in positive ionization mode using Data Dependent Acquisition with a mass range of m/z 50-1500. MS/MS spectra will be acquired for the top 5 ions in each MS1 spectrum, with active exclusion after 2 spectra. Raw high resolution mass spectrometry data files will be converted to open source .mzXML format using Bruker DataAnalysis software after lock mass correction. Raw data files as well as converted .mzXML files will be uploaded to MassIVE and further analyzed on the GNPS server.Taxonomic abundance and diversity: Raw sequence data from WGSwill be filtered for quality and length using KneadData. Then, readswill be run through MetaPhlAn2 to assign taxonomic profiles to each sample, with a resolution to the species level, using the QIIME 2 plugin, q2-metaphlan2. StrainPhlAn3 and PanPhlAn3 will be used for within-species profiling.Functional profiling:Functional profiles of gut microbial communities are more informative for human health than microbial compositions.Functional capacity of the microbiome will be determined by metagenomic metabolic reconstruction using HUMAnN2. We will complement these established approaches with our more recently developed woltka pipeline.Diversity metrics: QIIME 2 will be used to calculateα-diversity (Chao1, Simpson, Shannon) of both taxa and metabolic pathways.β-diversity will be calculated using UniFrac andBray-Curtis distancesfor the relative abundance of bacterial taxa and metabolic pathways (nonphylogenetic), respectively. TheDEICODE plug-in in QIIME 2 will also be used for assessingβ-diversity. This method uses Robust Aitchison Principal Components Analysison log-ratio transformations of non-zero values (log of the ratio of each proportion relative to a reference taxon).We will use permutational multivariate analysis of varianceand test for homogeneity of multivariate dispersions to evaluate differences by the intervention group over time.Statistical analyses:We will useMaAsLin2 in R to conduct linear regression models and linear mixed regression modelsto assess the associations between exposure variables and gut microbiome outcomes (α-diversity and relative abundances of taxa and metabolic pathways) at baseline and over time. Linear mixed models will be used to evaluate the effects of the micronutrient-fortified saltintervention on gut microbiome and metabolomic outcomes over the 12-month period (Objective IIandIII). We will account for repeated measures in linear mixed models by including participant as a random effect. Benjamini-Hochberg false discovery rate will be used formultiple comparison testing. Associations will be visualized using heat maps.Songbirdin QIIME 2 and q2-songbird plugin will be used to determine the differential abundance for responders versus non-responders, and will visualize results using Qurro.In addition to the aforementioned methods,Bayesian Multivariate Sparse Functional Principal Components Analysis will be used to model the dynamic temporal changes in microbiome outcomes and to detect its dependence on intervention assignment and biological covariates. mSFPCA can overcome irregular sampling intervals, limited sample size, missing values, and dropouts, making it appropriate for this trial design. mSFPCA will be conducted using R, with code available on GitHub.Objective III. To use stool metabolomics to examine potential mechanisms linking components of the microbiome affected by the micronutrient-fortified salt intervention and health outcomesin the host.Intervention and exposure variables: We will evaluate the effects of the micronutrient-fortified salt interventionon the gut metabolome over the 12-month study period, including differences in the stool metabolome in response to the intervention. Of particular interest will be compounds in the biosynthesis and degradation pathways, and metabolites known to be produced by bacteria that depend on key micronutrients in the micronutrient fortified salt intervention (i.e., iron, folic acid,vitamin B12)which will be determined by our in-development MicrobeMASST tool, and assessment of diet-derived moleculesviaour Foodomics pipeline.Outcome variables:Fecal metabolome outcomes include relative abundance,α-diversity, andβ-diversity of metabolites and metabolic pathways. While all metabolites and pathways will be assessed, we will focus onspecific pathways involved in micronutrientbiosynthesis and degradation, pathways that involve enzymes that use these nutrients ascofactors, and products of microbially-mediated pathways related to human health, such as production of bile acid derivatives.Statistical analyseswill be conducted largely as inObjective II, as metabolomics and microbiome data share many common characteristics.Specific to the metabolome data, we will use partial least squares discriminant analysis, which has been widely used as a supervised ordination approach that complements the unsupervised approaches that are more commonly used with microbiome data. We will also integrate the microbiome and metabolome data using our recently introduced tool mmvec, which uses TensorFlow to learn representations of the relationships between specific microbial taxa and metabolites.