Differences in immune status and fecal SCFA between Indonesian stunted children and children with normal nutritional status

We recently showed that the gut microbiota composition of stunted children was different from that of children with normal nutritional status. Here, we compared immune status and fecal microbial metabolite concentrations between stunted and normal children, and we correlated macronutrient intake (including energy), metabolites and immune status to microbiota composition. The results show that macronutrient intake was lower in stunted children for all components, but after correction for multiple comparison significant only for energy and fat. Only TGF-β was significantly different between stunted children and children of normal nutritional status after correction for multiple comparisons. TNF-alpha, IL-10, lipopolysaccharide binding protein in serum and secretory IgA in feces were not significantly different. Strikingly, all the individual short-chain and branched-chain fatty acids were higher in fecal samples of stunted children (significant for acetate, valerate and total SCFA). These metabolites correlated with a number of different microbial taxa, but due to extensive cross-feeding between microbes, did not show a specific pattern. However, the energy-loss due to higher excretion in stunted children of these metabolites, which can be used as substrate for the host, is striking. Several microbial taxa also correlated to the intake of macronutrients (including dietary fibre) and energy. Eisenbergiella positively correlated with all macronutrients, while an uncharacterized genus within the Succinivibrionaceae family negatively correlated with all macronutrients. These, and the other correlations observed, may provide indication on how to modulate the gut microbiota of stunted children such that their growth lag can be corrected. Trail registered at https://clinicaltrials.gov/ct2/show/NCT04698759.


Population and Research Sample
The population is childrens between three and five years old in the village of stunting locus.
The sampling method to be used in this research was quota sampling, namely 50 stunting and 50 healthy childrens from each Regency.

Research Variables
The variables studied will be the nutritional status of children between three and five years old, the characteristics of these children (age, sex, LBW, birth length, history of immunization, history of diarrhea, history of upper respiratory tract infection, food intake), mother and family characteristics (mother's age, mother's education, mother's occupation, father's occupation), home environment (aspects of the components of the house, aspects of sanitation facilities, aspects of occupant behavior), composition of the intestinal microbiota.
Data will be obtained through measurements, interviews using a questionnaire and direct observation using a check list. The nutritional status of each children included in this study will be quantified using the WHO recommended three nutritional Z-scores namely, height for age (referred to in this study as Zscore 1); weight for age (referred as Z-score 2) and weight for height (referred as Z-score 3).
A structured questionnaire was used for face-to-face interviews with the respective child's mother to collect sociodemographic information. In addition, age and anthropometric measurements (height, weight) based on Department of Health Ministry of Indonesia Regulation will be recorded. For stunting, the thresholds for height-for-age are: 'severely stunted' (<-3 SD); 'stunted' (-3 SD to < -2 SD); 'normal' (-2 SD to +3 SD); 'tall' (> +3 SD).

Characteristics of Children
a. "Balita age" is the children age from births until the last birthday.
b. Gender is a physically differentiated identity based on external genital organs. e. Immunization history is the immunization scheduled, due to the children age.
f. History of diarrhea is a history of diarrhea in the last two weeks.
g. History of upper respiratory infection (cough, cold, and fever) in the last two weeks.
h. Food intake is the amount and type of food under five years of age within the last 7 days, measured by food records and calculated by nutritional surveys.

Intestinal microbiota profile a. Stool sampling method
Stool samples will be taken from all childrens that are screened, with a total of 200 from 2 locations, Pandeglang and Sumedang. Stools will be collected using pot stool containers, and not include urine and latrines. The stools are immediately transferred to a stool container using a spoon, and stored in a cold temperature box using an ice pack. Stools will then immediately be transported to the laboratory. Stool samples will be buffered to keep DNA from being damaged.

b. Analysis
The microbiota profile in stool samples will be analyzed using Next Generation Sequencing (NGS) at Maastricht University, by sequencing the V3-V4 region of the 16S rRNA gene. Stools samples will also be analyzed for microbiota composition using total plate count, isolation and identification according to standard microbiological methods, while the identification of worm eggs in feces will be carried out by the Kato Kantz method in the YARSI and UKI University Parasitology Microbiology Laboratory.

Data Analysis Methods
a. The quantitative Insights Into Microbial Ecology (QIIME) pipeline will be used to study microbiota diversity based on grouping of the Taxonomic Operational Unit (OTU) reads at 99% similarity with comparison against the Greengenes database, Q30 = 75.9%, cluster density 808K / mm2. α-Diversity (diversity within the sample) will be analyzed using the diversity metric: Phylogenetic diversity (PD-Whole-Tree), Shannon index, Chao and Observed UTO.
For β-diversity, UniFrac (weigthed and unweighted) will visualized using 3D-graph principal coordinate analysis (PCoA), using the Emperor tool. β-diversity of UniFrac can be divided into weighted (which directly account for differences in relative abundance) and unweighted (presence or absence of OTU) is the distance between two communities.
b. Apart from analysis using the QIIME system, statistical analysis will also be performed using SPSS (IBM SPSS Statistics for windows version 22.0 Armonk, NY; IBM Corp). The Shapiro-Wilk test will be conducted to test the normality of the distribution of all quantitative variables studied. If the variables do not follow a normal distribution, a non-parametric method is chosen for further data analysis. Nonparametric methods Mann Whitney, Kruskall wallis and Spearman in R (function cor.test) will be used to study the correlation between several parameters in the microbiota profile.