The human gut microbiome is considered a metabolic organ due to its myriad benefits (O’Hara and Shanahan, 2006), which include the degradation of indigestible dietary polysaccharides into short-chain fatty acids (SCFAs) by fermentative enzymes in the microbiome (Flint et al., 2008; den Besten et al., 2013). Furthermore, several types of bacteria in the distal intestine can synthesize vitamins and essential amino acids (Metges, 2000; Martens et al., 2002; Morowitz et al., 2011). The gut microbiota matures around 3 years after birth. At that time, it is characterized by the presence of anaerobic microorganisms and is dominated by two bacterial phyla, Firmicutes and Bacteroidetes, which account for more than 90% of the gut microbiota population (Matamoros et al., 2013; Meehan and Beiko, 2014). When the bacterial ecosystem is disrupted (dysbiosis), some of the benefits of this metabolic organ may be reduced, and the microbiome may even harm the host (Hawrelak and Myers, 2004; Larsen and Dai, 2015). Early-life exposures, including those known to impact the composition of the gut microbiome, have been associated with an increased risk of childhood obesity, which can impact the subsequent development of obesity, type 2 diabetes, hypertension, and dyslipidemia in adulthood (Munyaka et al., 2014; Wallace et al., 2016). The gut microbiome also regulates obesity by increasing the energy harvest from the diet and by regulating peripheral metabolism (Yamashiro et al., 2015). In fact, the gut microbiota of individuals who are obese seem to have less bacterial richness, an increased level of Firmicutes, and a reduced level of Bacteroidetes (Turnbaugh et al., 2006; Le Chatelier et al., 2013). Notably, elevated fat consumption stimulates a bloom of Proteobacteria. The abundance of these bacterial lineages has been suggested as a potential diagnostic criterion for dysbiosis (Hildebrandt et al., 2009). Children with moderate and severe undernutrition have gut microbiota that are less mature and less diverse than those of healthy controls (Subramanian et al., 2014). Consequently, dysbiosis of gut microbiota is linked to undernutrition (Ghosh et al., 2014). Notably, gnotobiotic mice who received fecal microbiota transplants from children with kwashiorkor showed marked weight loss and metabolic abnormalities, including disturbances in amino acid, carbohydrate, and intermediary metabolism. This finding supports the idea that the gut microbiome is implicated in this type of severe undernutrition (Smith et al., 2013; Kane et al., 2015). Stunting, or low height-for-age, is the most common form of undernutrition and is very prevalent in resource-limited areas of the world. Stunting is considered the main indicator of childhood undernutrition, and the etiology of stunting is poorly understood (Ahmed et al., 2014; Dinh et al., 2016). Several studies have reported associations between stunting and altered body composition and fat distribution that predispose individuals to excess adiposity and abdominal fat distribution, which is strongly associated with the development of obesity, insulin resistance, and diabetes (Lear et al., 2009; Wilson et al., 2012; Pomeroy et al., 2014; Hardikar et al., 2015).

Mexico is going through a nutritional transition that affects school-age children, who bear the double burden of malnutrition, which includes undernutrition and obesity (13.6% and 15.3%, respectively) (Shamah-Levy et al., 2017). An undernourished population often moves from undernutrition to overnutrition, resulting in weight gain and central adiposity (Barquera et al., 2007). There is a risk of obesity in populations who move from famine early in life to abundance or even to excessive nutrition in adulthood (Martorell et al., 2001; Tzioumis and Adair, 2014). The mechanisms underlying these alterations remain unknown. However, low-income families may have an obesogenic environment due to the enrichment of sugar and edible oils in inexpensive food, resulting in a diet that is energy dense but micronutrient-poor that could alter their gut microbiome (Sawaya et al., 2003; David et al., 2014; Doak et al., 2016).

Considering this context, describing the gut microbiota in nutrition-related diseases may provide insights into the roles of microbiota in the pathogenesis of undernutrition and in obesity risk over time (Gordon et al., 2012). Thus, the aim of the present study was to describe and compare the richness and diversity of the intestinal microbiota in undernourished and obese Mexican school-age children living in low-income (marginal) and vulnerable communities (Vite-Pérez and Pérez-Zamorano, 2010; Moreno-Sánchez and Espejel-Mena, 2013).

Significant differences (P < 0.05) in the groups of undernourished, obese, and normal-weight Mexican school-age children in anthropometrical characteristics, such as BMI, waist circumference, tricipital and subscapular skinfolds, and body fat percentage. Regarding biochemical parameters, there were significant differences (P < 0.05) in glucose, triglycerides, insulin, ALT, and leptin levels (Supplementary Table S1). Division and discrimination achieved between the groups are shown in the Principal Component Analysis (Figure 1). The first principal component (PC1) accounted for 53.1% of the variance of the performance measures data, while the second principal component (PC2) accounted for 18.1%. The correlations between variables and the contribution of each variable to PC1 and PC2 are shown in Supplementary Figure S1.

Previous studies have shown that obesity is associated with changes in gut microbiota (Ridaura et al., 2013), but few studies have investigated the microbiota composition in undernourished children, including those affected by poverty in developing countries. The undernourished population tends to shift from undernutrition to overnutrition, but the mechanisms underlying the associated alterations remain unknown (Barquera et al., 2007). However, they may involve dysbiosis in the gut microbiota.

The present study categorized children into normal-weight, obese, and undernourished groups. Independent of nutrition status, the gut microbiota of the children were dominated by Bacteroidetes (including Bacteroides and Prevotella genera) and Firmicutes (including Clostridium, Enterococcus, Lactobacillus, and Ruminococcus), which account for more than 90% of the phylogenetic lineages (Sagheddu et al., 2016).

Microbiota diversity and richness was lower in the undernourished and obese groups than in the normal-weight group. Interestingly, this pattern has also been reported in undernourished mice (Preidis et al., 2015). Studies indicate that individuals with low intestinal bacterial richness have more overall adiposity and insulin resistance and gain more weight over time (Le Chatelier et al., 2013). Indeed, significant decreases in diversity and phylum-level changes have been associated with obesity (Turnbaugh et al., 2009a). Ley et al. (2006) found a higher F/B ratio in obese people than in thin people. In the present study, the F/B ratio was higher in the undernutrition group than in the other groups; however, this difference was not significant.

Bacteroidetes have a very large repertoire of genes that are involved in the acquisition and metabolism of polysaccharides (Mahowald et al., 2009). Bacteroidetes can easily adapt to any environmental niche, owing in part to the plasticity of their genomes, which undergo continuous genetic rearrangements, duplications, and lateral gene transfers between species (Thomas et al., 2011). The relative abundance of Bacteroidetes was lowest in the obese group and was higher in the normal-weight group compared to the undernourished group. This is in accordance with data from undernourished neonatal mice, who have a relatively lower proportion of Bacteroidetes compared to controls and are deficient in multiple microbial pathways, including the N-glycan pathway. This deficiency may result in less efficient energy extraction from non-digestible polysaccharides or diet fiber (Preidis et al., 2015).

In addition, we found that the F/B ratio in the gut microbiota of the undernourished group was higher than in the obese and normal-weight groups. This might be because the undernourished group had a diet high in sugar and low in fiber, which could predispose them to future obesity. There is evidence that high-glucose or -sucrose consumption induces changes in gut microbiota, increasing the F/B ratio (Magnusson et al., 2015; Do et al., 2018). On the other hand, a so-called Western diet, which is high in sugar and fat, has been associated with an overgrowth of Firmicutes and a parallel decrease in Bacteroidetes (Turnbaugh et al., 2009b). The Western lifestyle, including the diet, is associated with high incidences of chronic diseases, such as cardiovascular disease and type II diabetes, which individually and collectively have a hefty socioeconomic burden (Turnbaugh et al., 2009b). Most Western populations have omnivorous diets that are rich in refined food and of poor nutritional quality (Conlon and Bird, 2015).

We found that the Lachnospiraceae family within the phylum Firmicutes was significantly overrepresented in undernourished children and correlated negatively with energy intake. Early blooms of this family have been associated with adiposity, weight gain, and diabetes (Cho et al., 2012; Kameyama and Itoh, 2014). According to Turnbaugh et al. (2006), the gut microbiomes of obese people with the lowest levels of microbiota diversity and richness had higher energy harvesting. The mechanisms involved in this phenomenon are not yet clear, but a recent study showed that Lachnospiraceae, which was detected in the microbiomes of animals with more efficient energy harvesting, could be a contributing factor (Shabat et al., 2016). Considering these data, we hypothesize that having high Lachnospiraceae levels contributes to a process of adaptation that protects against poor nutrition. In this context, although the microbial alpha diversity was low in undernourished children, this type of bacteria helps improve the energy balance, suggesting that the Lachnospiraceae family might serve as a metabolic regulator in individuals with an undernourished phenotype. As mentioned above, we found that Firmicutes, particularly Lachnospiraceae, was enriched in the undernourished group. Notably, the presence of Lachnospiraceae is closely related to obesity. Hence, we hypothesize that an expansion of Firmicutes could improve energy homeostasis and protect against undernutrition, but it could also have negative cumulative effects on the children’s health in the future. Interestingly, Lachnospiraceae abundance correlated with leptin levels; moreover, malnourished children had higher concentrations of leptin than normal-weight children. Studies indicate that independent of body fat, high serum leptin levels may be an indicator of increased leptin resistance, which predisposes children who are at high risk of adult obesity to weigh more and to have more body fat during childhood (Fleisch et al., 2007). According to the nutrition transition theory, populations tend to shift from undernutrition to overnutrition as they experience the dietary and demographic changes associated with socioeconomic development (Barquera et al., 2007). Undernutrition in early life results in an increased risk of hyperinsulinemia, high blood pressure, obesity, diabetes, and cardiovascular diseases in adulthood (Guerrant et al., 2013). In Mexico, child stunting is one of the biggest nutritional public health problems, particularly in underserved groups, including those in rural areas (Kroker-Lobos et al., 2014). Interestingly the prevalence of stunting in Mexico has decreased in recent decades, but this has been coupled to a dramatic change in the prevalence of overweight and obesity in children (Shamah-Levy et al., 2017). Thus, our data suggest that Lachnospiraceae may be linked to increased energy harvest by gut microbiota. Additional studies are needed to determine whether the abundance of Lachnospiraceae early in life predisposes undernourished children to obesity later in life.

Our results showed that the obese group had higher levels of Proteobacteria, and we also found a positive correlation between Proteobacteria and fat intake. Preliminary investigation showed that an increase in Proteobacteria represents a risk factor for human health, including dysbiosis, and therefore the abnormal growth of Proteobacteria may represent an imbalance in the gut microbial community and be a potential marker of disease risk (Shin et al., 2015). Moreover, the combination of low alpha diversity and Proteobacteria expansion may reflect intestinal dysbiosis in obese children. A recent study in mice found an association between increased abundance of Bilophila and fat feeding, inflammation, and colitis (Devkota et al., 2012). In the present study, the Bilophila level was increased in the obese group of children.

To determine whether the enterotypes of the human gut microbiome are linked to the Western diet, we analyzed possible correlations between Bacteroides and Prevotella and energy and macronutrient intake. The Bacteroides enterotype was positively associated with energy intake and fat intake, including the intake of monounsaturated, polyunsaturated, and saturated fat. Bacteroides are associated with consumption of a long-term fat-enriched diet (Lim et al., 2014). These results indicate that, despite the alterations of the gut microbiota in undernourished and obese children, enterotype 1 (Bacteroides) can be considered a potential bacterial marker of a Western diet that predisposes them to chronic diseases. A recent study showed that Prevotella was associated with long-term consumption of dietary fiber. Likewise, we found a strong, positive correlation between Prevotella and bean consumption. Beans are one of the most widely consumed foods in the Mexican population (Mojica and de Mejia, 2015).

This is the first study to examine the gut microbial community structure in undernourished and obese Mexican children living in low-income neighborhoods. Our analysis revealed distinct taxonomic profiles for undernourished and obese children.

EM-S, MG-S, BP-G, and MM conceived and designed the experiments, analyzed the data, and wrote the paper. EM-S, MO-L, MG-S, and BP-G performed the experiments.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.