Background: To compare and characterize the gut microbiota in women of childbearing age from sub-Saharan Africa (the Democratic Republic of the Congo, DRC) and South Asia (India), in relation to dietary intakes. Methods: Women of childbearing age were recruited from rural DRC and India as part of the Women First (WF) preconception maternal nutrition trial. Findings presented include fecal 16S rRNA gene-based profiling of women in the WF trial from samples obtained at the time of randomization, prior to initiation of nutrition intervention and to conception. Results: Stool samples were collected from 217 women (DRC n = 117; India n = 100). Alpha diversity of the gut microbiota was higher in DRC than in India (Chao1: 91 ± 11 vs. 82 ± 12, P = 6.58E-07). The gut microbial community structure was not significantly affected by any demographical or environmental variables, such as maternal BMI, education, and water source. Prevotella, Succinivibrio, and Roseburia were at relatively high abundance without differences between sites. Bifidobacterium was higher in India (4.95 ± 1.0%) than DRC (0.3 ± 0.1%; P = 2.71E-27), as was Lactobacillus (DRC: 0.2 ± 0.0%; India: 1.2 ± 0.1%; P = 2.39E-13) and Faecalibacterium (DRC: 6.0 ± 1.7%; India: 8.4 ± 2.9%; P = 6.51E-7). Ruminococcus was higher in DRC (2.3 ± 0.7%) than in India (1.8 ± 0.4%; P = 3.24E-5) and was positively associated with consumption of flesh foods. Succinivibrio was positively associated with dairy intake in India and fish/insects in DRC. Faecalibacterium was positively associated with vitamin A-rich fruits and vegetables. Overall, these observations were consistent with India being primarily vegetarian with regular fermented dairy consumption and DRC regularly consuming animal-flesh foods. Conclusion: Consumption of animal-flesh foods and fermented dairy foods were independently associated with the gut microbiota while demographic variables were not, suggesting that diet may have a stronger association with microbiota than demographic characteristics.
Women of childbearing age were recruited from rural DRC (Equateur Province) and India (Belagavi, Karnataka) as part of the WF preconception maternal nutrition trial (Hambidge et al., 2014, 2019). Inclusion criteria were 16–35 years of age; parity 0–5; expectation to have first or additional pregnancy within next 2 years and without intent to utilize contraception. Enrollment occurred after screening, and informed consent was obtained by the home visitor research assistant if the potential participant was eligible. Findings presented in the current report include the gut microbiota of women at the time of randomization prior to the initiation of the nutrition intervention and at least 3 months prior to conception in the WF trial and represented participants from two of the WF sites, with distinctive ethnicity, diet, culture and geographical locations. Women were recruited from 12 villages in rural DRC and 9 villages from rural India. The project was approved by the Colorado Multiple Institutional Review Board, University of Colorado, the local and/or national ethics committees for each site (registered with the US Office of Human Research Protection and with Federal-wide Assurance in place). Written informed consent was obtained from all participants and the study was registered at ClinicalTrials.gov ({“type”:”clinical-trial”,”attrs”:{“text”:”NCT01883193″,”term_id”:”NCT01883193″}}NCT01883193). Questionnaires of demographic information were administered: cell phone (Yes/No), education (None vs. At least secondary), electricity (Yes/No), man-made floor (Yes/No), flush toilet (Yes/No), improved water (Yes/No), landline (Yes/No), motorcycle (Yes/No), fridge (Yes/No), worry no food (Yes/No), and open sewage near house (Yes/No). Improved water means the participant had access to filtered or treated water. Questionnaires were administered in the home by the local home visitor research assistant and were completed within 1 week of enrollment. A mobile assessment team obtained past medical history and height and weight, from which body mass index (BMI) was calculated. A stool sample was collected from each participant (DRC n = 117; India n = 100). A pre-labeled fecal bag, Ziploc bag, a black cryogenic pen, and a Styrofoam storage box containing ice or ice packs were provided to each participant. Stool was collected into the fecal bag using a sterile scoop and then placed into a second Ziploc bag. Participants then placed the bag into the Styrofoam storage container until picked up by the research team on the day the stool was passed. When receiving samples, the research team labeled the sample date and time of stool passage. The research team scooped about a teaspoon of stool and transferred the sample to a sterile stool storage tube with 3 ml RNAlaterTM (ThermoFisher Scientific Inc., Waltham, MA, United States), ensuring that the specimen was coated with RNAlaterTM and that the label was complete. The stool samples were then frozen at −20°C or colder. Samples were shipped to the University of Colorado on ice packs or at ambient temperature. Bacterial profiles were determined by broad-range amplification and sequence analysis of 16S rRNA genes following our previously described methods (Hara et al., 2012; Markle et al., 2013). In brief, DNA was extracted from 25 to 50 mg of stool using the QIAamp PowerFecal DNA kit (Qiagen Inc., Carlsbad, CA, United States), which employs chemical and mechanical disruption of biomass. PCR amplicons were generated using barcoded (Frank, 2009) primers that target approximately 450 basepairs of the V3V4 variable region of the 16S rRNA gene (338F: 5′ACTCCTACGGGAGGCAGCAG and 806R: 5′ GGACTACHVGGGTWTCTAAT) (Lane et al., 1985; Weisburg et al., 1991). PCR products were normalized using a SequalPrepTM kit (Invitrogen, Carlsbad, CA, United States) and then pooled. The amplicon pool was partially lyophilized to reduce its volume then purified and concentrated using a DNA Clean and Concentrator Kit (Zymo, Irvine, CA, United States). Pooled amplicons was quantified using a Qubit Fluorometer 2.0 (Invitrogen, Carlsbad, CA, United States). Illumina paired-end sequencing was performed following the manufacturer’s protocol on the MiSeq platform using a 600 cycle version 3 reagent kit and versions v2.4 of the MiSeq Control Software. Illumina Miseq paired-end reads were aligned to human reference genome hg19 with bowtie2 and matching sequences discarded (Homo Sapiens Ucsc Hg19 Human Genome Sequence from iGenome: Illumina, 2009; Langmead and Salzberg, 2012). As previously described, the remaining non-human paired-end sequences were sorted by sample via barcodes in the paired reads with a python script (Markle et al., 2013). The sorted paired reads were assembled using phrap (Ewing and Green, 1998; Ewing et al., 1998). Pairs that did not assemble were discarded. Assembled sequence ends were trimmed over a moving window of five nucleotides until average quality met or exceeded 20. Trimmed sequences with more than 1 ambiguity or shorter than 350 nt were discarded. Potential chimeras identified with Uchime (usearch6.0.203_i86linux32) (Edgar et al., 2011) using the Schloss (Schloss and Westcott, 2011) Silva reference sequences were removed from subsequent analyses. Assembled sequences were aligned and classified with SINA (1.3.0-r23838) (Pruesse et al., 2012) using the 418,497 bacterial sequences in Silva 115NR99 (Quast et al., 2013) as reference configured to yield the Silva taxonomy. Operational taxonomic units (OTUs) were produced by clustering sequences with identical taxonomic assignments. The software package Explicet (v2.10.5) (Robertson et al., 2013) was used for microbial diversity analysis. For those participants who conceived after at least 3 months after randomization and entered the pregnancy phase of the trial, repeat 24-h dietary recalls were conducted in first trimester on a randomly selected subgroup of the study participants. For the current analysis, dietary data were obtained for 50 women at each site (DRC n = 50; India n = 50). In brief, two 24-h dietary recalls (Lander et al., 2017) were conducted 2–4 weeks apart once pregnancy was confirmed prior to 12-week gestation. The analysis reflected only the participants’ food intakes and did not include any contribution from the trial intervention nutrient supplement. No counseling regarding diet choices or quality was provided over the course of the trial. Dietary assessment training was provided for each of the site nutritionists by the lead study nutritionist. A unique food nutrient composition database was constructed at each site based on the food intake data collected from the dietary recalls to quantify intakes as nutrients and food groups (Lander et al., 2017). Values are presented as mean ± SD for continuous variables. Alpha diversity indices measured for the gut microbiota were tested using the Mann–Whitney statistic. Beta diversity was calculated using weighted UniFrac distances. To compare between DRC and India for differences of the gut microbial profiles, non-parametric Mann–Whitney tests were used. P < 0.05 was considered significant between DRC and India for the gut microbial profiles comparison. Within each site (e.g., DRC and India), permutation-based multivariate analysis of variance (PERMANOVA) tests were used to assess the associations between gut microbial community composition and demographical variables. Associations between the relative abundances of bacterial taxa and nutrients/food-groups were assessed by Spearman rank-order correlation tests. Results were visualized by plotting heatmaps of Spearman’s rho correlation coefficient using the heatmap.2 R function; hierarchical clustering and dendrograms were generated using the default parameters of heatmap.2/dist/hclust functions, using Euclidean distances. Nominal p-values not accounting for multiple testing are reported. Data were analyzed using R version 2.7.2 (R Foundation for Statistical Computing; Vienna, Austria).
N/A