Background
Metabolic syndrome (MetS) has attracted widespread attention due to its association with cardiovascular disease (CVD), which is a major contributor to the burden of disability and the leading cause of death worldwide [
1‐
3]. MetS is a complex disorder with several risk factors including abdominal obesity, dyslipidemia (increased triglyceride [TG] and decreased high-density lipoprotein cholesterol [HDL-C]), elevated blood pressure (BP), and hyperglycemia [
4]. It has been reported that the prevalence of individual and clustering of MetS risk factors among children and adolescents has been alarmingly increasing in recent years [
5,
6]. In addition, the number of MetS risk factors was associated with gradually increasing odds of short-term and long-term cardiovascular damage [
7,
8]. Therefore, early detection of the accumulation of MetS risk factors is important for the prevention of CVD and related morbidity later in life.
Recently, microbiome-based interventions have been gaining popularity to treat and prevent metabolic disorders. Previous studies based on mice models and adults showed the association of alteration of gut microbiota with MetS and its risk factors [
9‐
12]. However, limited studies have paid attention to the association in children. The gut microbiota of children has been shown to be more susceptible to environmental factors than that of adults, and thus the gut microbiota associated with adult MetS cannot be generalized to children [
13]. Several studies have investigated the effect of the change in gut microbiota on individual MetS risk factor (e.g., obesity, elevated BP, hyperglycemia) among children [
14‐
18]. However, to the best of our knowledge, little is known about the gut microbiota in identifying the number of MetS risk factors among children. Clarifying this association may open avenues for convenient prevention, diagnosis, and treatment of the clustering of MetS risk factors in childhood and thus reduce the huge burden of CVD in adulthood.
Therefore, in this study, we aimed to identify the differential gut microbiota among children aged 10 − 11 years with none, one, and two or more numbers of MetS risk factors.
Methods
Participants and sample collection
This was a nested case–control study from the baseline of the “Huantai Childhood Cardiovascular Health Cohort Study” including 1515 children aged 6 − 11 years old, among whom we identified 24 children with two or more MetS risk factors. To control for confounding factors such as age and sex, we conducted a 1:1:1 propensity score matching to select 24 children without MetS risk factors and 24 children with one MetS risk factor (none MetS risk factor: 10.83 ± 0.35 years old; one MetS risk factor: 10.73 ± 0.32 years old; two-or-more MetS risk factors: 10.72 ± 0.33 years old; male: female = 15:9 in each group). Thus, a total of 72 children aged 10 − 11 years without the use of antibiotics and probiotics in the past three months were included in this study (24 children without MetS risk factors, 24 with one risk, and 24 with two or more risks). All included children were without a history of gastrointestinal disorders or diarrhea and were not taking medications at the time of the study. The informed consent was written by all participants and their guardians. This study was approved by the Ethics Committee of Shandong University.
Clinical data collection
Anthropometrics (e.g., weight, height, BP, and waist circumference [WC]), demographic characteristics (e.g., age and sex), and blood biochemistry indexes (e.g., fasting plasma glucose [FPG], TG, HDL-C, and low-density lipoprotein cholesterol [LDL-C]) were collected in this study. Specifically, height and weight were measured twice in light clothes without shoes using an ultrasonic height and weight scale (Shengyuan Co. Ltd, HGM-300, Henan, China). The mean values of two height and weight measurements were used for data analyses with an accuracy of 0.1 cm and 0.1 kg for height and weight, respectively [
19]. Body mass index (BMI) was calculated by weight (kg) divided by height (m) squared. WC was measured twice using a non-elastic measuring tape at 1 cm above the navel around a week horizontally, and the mean values of two WC measurements were used for data analyses with an accuracy of 0.1 cm [
19]. BP was measured three times continuously with the deviation of any two BP values controlled within 4 mmHg (OMRON-HEM 7012, Osaka, Japan), and the mean values of three BP measurements were used for data analyses [
20]. FPG, TG, HDL-C, and LDL-C were measured using a Beckman Coulter AU480 automatic analyzer (Mishima, Shizuoka, Japan) [
20].
Definition
Children received one point for each MetS risk factor if they met the criteria outlined as follows: (1) elevated BP: systolic BP (SBP) and/or diastolic BP (DBP) ≥ the age- and sex-specific 90
th percentile [
21]; (2) hyperglycemia: FPG ≥ 5.6 mmol/L [
22]; (3) dyslipidemia: TG ≥ 1.47 mmol/L; (4) dyslipidemia: HDL-C ≤ 1.03 mmol/L [
23]; (5) abdominal obesity: WC ≥ the age- and sex-specific 90
th percentile [
24]. Thus, based on the number of MetS risk factors, children were classified into three groups (non-risk, one-risk, and two-or-more-risks). In this study, alterations of gut microbiota refer to the difference in the composition and relative abundance of dominant species, community richness (i.e., the total number of species in the community) and diversity (i.e., the richness and evenness of species in the community), and differential species among groups with an increasing number of MetS risk factors compared with the group without MetS risk factors.
Basic characteristics
The frequency of fruit and vegetable intake each day, the frequency of soft drink intake every week, and physical activity were classified into < 3 times/day vs. ≥ 3 times/day, < 3 times/week vs. ≥ 3 times/week, and < 1 h/day vs. ≥ 1 h/day, respectively. Parental education was divided into lower than high school and high school or higher (i.e., one of parents with high school or higher). Both paternal smoking and drinking were classified as yes and no.
Fecal samples collection and processing
Fresh fecal samples from children who had not received antibiotics within the past three months were collected in sterile fecal tubes and then frozen in − 80 °C refrigerators. Microbial genomic DNA (gDNA) was extracted from the fecal samples and detected by 1% agarose gel electrophoresis [
25]. 16S rRNA gene was selected as bacterial specific fragment using 338F (5'-ACTCCTACGGGAGGCAGCAG-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3') primers [
26]. Amplifications were performed using TransGen AP221-02: TransStart Fastpfu DNA Polymerase [
27]. A two-stage PCR was performed in the ABI GeneAmp® 9700 (Applied Biosystems Inc. USA) in triplicate [
28]. Then, the PCR products of the same sample were mixed and detected using 2% agarose gel electrophoresis, followed by recovering with AxyPrepDNA Gel Recovery Kit (Axygen; Corning, Inc., Corning, NY, USA) and eluting with Tris–HCl [
29].
According to the preliminary quantitative results of electrophoresis, the PCR products were detected and quantified with the QuantiFluor™-ST Blue Fluorescence Quantitative System (Promega Corp., Madison, WI, USA) [
30]. The purified amplicons were pooled in equimolar amounts and sequenced on an Illumina Hiseq3000 platform (Illumina, SanDiego, CA, USA) according to the standard protocols.
Sequence processing and analysis
Quality control of the raw sequencing reads was performed using the FastQC tool (
https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) [
31] to filter the bad reads, the low-quality bases, adapters, and N-bases [
32]. According to the overlap relation between Pair-end (PE) reads, we merged pairs of reads into a sequence with a minimum length of overlap of 10 bp [
33]. After detecting and filtering the chimera sequence, the data were analyzed with the Quantitative Insights Into Microbial Ecology (QIIME 1.9.1;
http://qiime.org/install/index.html) toolkit to obtain the optimization sequence [
34]. The raw sequencing reads have been submitted to the Sequence Read Archive (SRA) of the National Center for Biotechnology Information (NCBI) database (BioProject ID: PRJNA775883;
https://www.ncbi.nlm.nih.gov/bioproject/PRJNA775883), and this deposited data is available in the NCBI database.
According to a 97% similarity cut-off, the Operational Taxonomic Units (OTU) clustering was performed for non-repeating sequences (excluding single sequences) using Uparse software (version 7.0.1090;
http://drive5.com/uparse/). At the same time, chimeric sequences were identified and removed to obtain representative sequences of OTUs [
29]. According to the Silva database (Release 138;
http://www.arb-silva.de), taxonomic annotation was performed on the OTUs representative sequences of each sample based on the RDP classifier Bayesian algorithm Classifier (version 2.11;
http://sourceforge.net/projects/rdp-classifier/) with a 0.7 confidence threshold [
29].
Statistical analyses
The continuous variables were presented as mean and standard deviation (SD), and the categorical variables were presented as n (%). The Analysis of Variance was performed to compare the differences in age, WC, BMI, SBP, DBP, FPG, TG, LDL-C, and HDL-C among the three groups, and the Chi-square test was performed to compare the differences in sex, the frequency of fruit and vegetable intake each day, the frequency of soft drink intake every week, physical activity, parental education, paternal smoking, paternal drinking, parental BMI, parental history of hypertension, heart disease, stroke, and diabetes among the three groups. SPSS 25.0 software (IBM, Armonk, NY, USA) and R language (version 3.3.1) were used for analysis. Two-sided P values < 0.05 indicate a significant difference.
The rarefaction curve was used to explore the sequencing depth as well as the abundance of sample species with different sequencing quantities based on the Sobs index (community richness) and Shannon index (community diversity). The Venn diagram analysis was performed to count the number of common and unique OTUs among the three groups. Among genera with a relative abundance greater than 0.01%, the bar plot and the non-parametric Kruskal–Wallis H test were used to compare the changes in composition and relative abundance of genera among the three groups with the adjustment of the false discovery rate (FDR).
The
α-diversity indexes including the Ace, Chao 1, Shannon, and Inverse Simpson were calculated to evaluate the community diversity and richness of gut microbiota at the OTU level using the Mothur software platform (version 1.30.2;
https://www.mothur.org/wiki/Download_mothur) [
35]. A trend test was used to estimate trends in
P values for the
α-diversity indexes of gut microbiota and the number of MetS risk factors. The difference in
β-diversity among the three groups was calculated according to the Bray–Curtis distance matrix using permutational ANOSIM in the Principal Coordinate Analysis (PCoA) and Non-metric multidimensional scaling analysis (NMDS) [
36].
At the genus level, the Linear Discriminant Analysis Effect Size (LEfSe) and the linear discriminant analyses (LDA) were performed to evaluate the extent of the contribution of differential gut microbiota to the different numbers of MetS risk factors. The random forest model analysis was conducted to screen out the top ten genera biomarkers to distinguish the groups with at least one MetS risk group from the non-risk group based on the randomForest package of the R language. We further evaluate the differences in the relative abundance of these top ten genera biomarkers among the three groups based on the non-parametric Kruskal–Wallis H test with the FDR. Finally, the significant genera in the LEfSe analysis, the random forest analyses, and the non-parametric Kruskal–Wallis H test were selected as potential diagnostic biomarkers for the different numbers of MetS risk factors in children. The receiver operating characteristic curve (ROC) was performed to evaluate the ability of these selected genera in identifying the number of MetS risk factors (i.e., area under the ROC curve [AUC]) based on the pROC package of the R language.
Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 (PICRUSt2) software was performed to predict the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways of gut microbiota based on the Greengene database [
37]. The non-parametric Kruskal–Wallis H test was performed to compare the differences in KEGG level 3 pathways among the three groups. Then, the Post-hoc test with Tukey Kramer was used for further pairwise comparisons on the differential pathways based on the STAMP software [
38] (Additional file
1).
Discussion
To the best of our knowledge, we initially found that the dysbiosis of gut microbiota was associated with the different numbers of MetS risk factors among children. There was a downward trend in the community diversity and richness of gut microbiota with the increased number of MetS risk factors. Among genera with a relative abundance greater than 0.01%, the genus Lachnoclostridium increased in the MetS risk groups, whereas the genera Alistipes and Lachnospiraceae_NK4A136_group decreased in the MetS risk groups compared with the non-risk group. In addition, the genera Christensenellaceae_R-7_group, Family_XIII_AD3011_group, and Lachnoclostridium performed moderately well in identifying the number of MetS risk factors. Our KEGG pathway analyses showed that D-Glutamine and D-glutamate metabolism and cysteine and methionine metabolism pathways might participate in cardiovascular homeostasis through the regulation of gut microbiota. These findings confirmed that gut microbiota played a pivotal part in identifying the number of MetS risk factors among Chinese children.
It has been shown that specific MetS risk factors are inversely associated with the community richness and diversity of gut microbiota among adults [
39]. For example, individuals with obesity, dyslipidemia, or hypertension were more likely to obtain a lower community richness of gut microbiota compared with normal controls [
40‐
42]. In addition, previous studies found that increased BMI and blood lipid levels (e.g., TG) were associated with reduced bacterial diversity among adults [
42,
43]. Alterations in the composition of gut microbiota have also been reported to be linked to atherosclerosis, hypertension, obesity, and type 2 diabetes mellitus [
39]. However, these previous findings only focused on one MetS risk factor or were limited to adults, ignoring the association between the clustering of MetS risk factors and the imbalance of gut microbiota. In this study, we not only found that the richness estimators of gut microbiota decreased among children with one MetS risk factor, but also further found that there was a decreasing trend of community diversity and richness with the accumulation of MetS risk factors among children.
There were complex interactive effects of genetic background, gut microbiota, and diet on the development of obesity and MetS features [
44]. In this study, we found that there are no differences in diet across these three groups, suggesting that the differential gut microbiota across groups might be due to genetic influences. However, the diet information we collected was self-reported, which might bias the true results. Besides, we additionally performed analyses on the difference in parental BMI, and parental history of hypertension, heart disease, stroke, and diabetes across three groups and found that there were no differences across three groups. Our findings suggest that other unmeasured variables might affect the association between gut microbiota and the number of MetS risk factors. Future studies with larger sample sizes, more accurate dietary information, and detailed lifestyle information were called for verifying our findings.
Consistent with our findings in the association of
Lachnoclostridium,
Alistipes, and
Lachnospiraceae_NK4A136_group with the different number of MetS risk factors, previous animal studies showed that the relative abundance of
Lachnoclostridium was positively related to TG and negatively related to HDL-C in rats [
45]. However, the relative abundance of
Lachnospiraceae_NK4A136_group was negatively associated with weight gain and serum lipid levels in mice [
46‐
48]. Additionally, the relative abundance of
Alistipes was reduced in obese adults with metabolic diseases from China [
49], obese adults from the Netherlands [
50], and obese individuals with type 2 diabetes mellitus from Germany [
51]. These findings suggest that the higher abundance of the genus
Lachnoclostridium and the lower abundance of the genera
Alistipes and
Lachnospiraceae_NK4A136_group might contribute to the accumulation of MetS risk factors among children.
In this study, we also found that the genera
Christensenellaceae_R-7_group and
Family_XIII_AD3011_group performed a high ability in differentiating the two-or-more MetS risk factor group from the non-risk group.
Christensenellaceae_R-7_group was found to be negatively related to body weight, visceral fat percentage, and FPG levels in animal experiment studies [
52‐
54]. Moreover, it has been reported that the
Family_XIII_AD3011_group was inversely related to glycated hemoglobin, 2 h glucose level and insulin, BMI, and secretion index in patients with type 2 diabetes mellitus, suggesting that it could be a novel predictive microbial biomarker for type 2 diabetes mellitus [
55,
56]. These findings imply that the
Christensenellaceae_R-7_group and
Family_XIII_AD3011_group may provide a non-invasive, practical, and clinical diagnosis of the accumulation of MetS risk factors in children.
Previous studies supported our findings that the D-Glutamine and D-glutamate metabolism, as well as cysteine and methionine metabolism, play an important role in the accumulation of MetS risk factors [
57‐
64]. It has been reported that the glutamate concentration was positively related to TG, glucose, BMI, and the increased risks of type 2 diabetes mellitus, whereas the ratio of glutamine/glutamate was negatively related to TG, glucose, BMI, and the increased risks of type 2 diabetes mellitus in Mediterranean and Spanish populations [
57‐
59]. Furthermore, plasma total cysteine and methionine were strongly associated with MetS risk factors such as higher total cholesterol concentration, elevated BP, obesity, and type 2 diabetes mellitus [
60‐
64]. These findings suggest that gut microbiota may be associated with the number of MetS risk factors through D-Glutamine and D-glutamate metabolism, as well as cysteine and methionine metabolism pathways.
This study has some limitations. First, the sample size in this study is smaller compared to studies in adults, so further validation with a larger sample size is warranted. However, the small sample size could still provide statistical confidence for our results because of the measurement of substantial changes in gut microbiota [
65]. Second, due to the case–control study, a causal association between gut microbiota and the number of MetS risk factors cannot be concluded. Third, limited information on CRP or other cytokines and parental risk factors for MetS prohibited us from performing the effects of these clinical indicators on the association between alteration of gut microbiota and the number of MetS risk factors. Fourth, the differential gut microbiota in this study should be carefully extrapolated to children of other ethnicities, and future studies with children from other regions or countries are needed to validate our findings based on qPCR experiments. Fifth, 16S rRNA gene sequencing technology provides limited information on bacterial genes and their functions, and further research is needed to explicate the pathogenesis and mechanism of gut microbiota in children with an accumulation of MetS risk factors. Sixth, because we focused on MetS risk factors, LDL cholesterol as an important cardiovascular risk factor was not included in this study. However, we compared LDL cholesterol across the three groups and found no significant difference, partly suggesting the effect of LDL cholesterol on associations between gut microbiota and the number of MetS risk factors might be attenuated.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.