Abstract
Background
Left ventricular hypertrophy (LVH) is most common when driven by
hypertension, and it is a strong independent risk factor for adverse
cardiovascular events and death. Some animal models support a role for
gut microbiota and metabolites in the development of LVH, but cohort
studies confirming these findings in populations are lacking.
Methods and Results
We investigated the alterations of gut microbiota and metabolites in 30
patients with hypertension, 30 patients with hypertensive LVH, and 30
matched controls on the basis of 16S rDNA and metabolomic analyses.
Thirty stool and 90 serum samples were collected in fasting conditions.
ANOVA/Kruskal–Wallis/Pearson's χ^2/Fisher's exact test and Bonferroni's
correction were used (P<0.0167) for comparison among the 3 groups. A
regression analysis and subgroup analysis were performed between gut
microbiota and left ventricular mass index (LVMI) and metabolites and
LVMI, respectively. Spearman correlation analysis was performed between
metabolites and flora and metabolites and LVMI. We observed
LVH‐enriched Faecalitalea (β=6758.55 [95% CI, 2080.92–11436.18];
P=0.009), Turicibacter (β=8424.76 [95% CI, 2494.05–14355.47]; P=0.01),
Ruminococcus torques group (β=840.88 [95% CI, 223.1–1458.67]; P=0.013),
and Erysipelotrichaceae UCG‐003 (β=856.37 [95% CI, 182.76–1529.98];
P=0.019) were positively correlated with LVMI. A total of 1141 (in
sera) and 2657 (in feces) metabolites were identified. There was a
sex‐specific association between metabolites and LVMI. Significant
changes in metabolic pathways in LVH were also observed, especially
bile acid and lipid metabolism pathways.
Conclusions
Our study demonstrated the disordered gut microbiota and microbial
metabolite profiles in LVH. This highlights the roles of gut bacteria
and metabolite in this disease and could lead to new intervention,
diagnostic, or management paradigms for LVH.
Registration
URL: [32]https://www.chictr.org.cn; Unique Identifier:
ChiCTR2200055603.
Keywords: gut microbiota, hypertension, left ventricular hypertrophy,
metabolomics
Subject Categories: Hypertension, Biomarkers, Metabolism, Remodeling
__________________________________________________________________
Nonstandard Abbreviations and Acronyms
BA
bile acid
CDCA
chenodeoxycholic acid
DA
differential abundance
DCA
deoxycholic acid
F/B
Firmicutes/Bacteroidetes
KEGG
Kyoto Encyclopedia of Genes and Genomes
LA
linolenic acid
LDA
linear discriminant analysis
LVMI
left ventricular mass index
PLS‐DA
partial least squares discrimination analysis
TMAO
trimethylamine N‐oxide
Clinical Perspective.
What Is New?
* Microbial profiles of patients with left ventricular hypertrophy in
the real world, the metabolite signature profiles, and microbial
biomarkers for early detection of left ventricular hypertrophy.
What Are the Clinical Implications?
* Our study can provide ideas for early intervention targets to
prevent patients with hypertension from developing left ventricular
hypertrophy and to provide possible means of reversal for patients
who have already developed left ventricular hypertrophy.
Hypertension is a leading risk factor for death and strongly related to
cerebrovascular and cardiovascular disease.[33] ^1 , [34]^2 The
pathophysiological mechanisms of essential hypertension remain largely
unknown, which hinders the development of targeted therapies.
Pathological left ventricular hypertrophy (LVH) is most common when
driven by hypertension, and it is a strong independent risk factor for
adverse cardiovascular events and death, even when left ventricular
systolic function is normal and there is no preceding myocardial
infarction.[35] ^3 Therefore, timely diagnosis and early treatment are
the keys. It is well documented that treatment of hypertension is a
means of delaying the development of LVH.[36] ^4 , [37]^5 LVH does not
seem to be entirely caused by blood pressure (BP). Clinical treatment
has confirmed that although some patients with hypertension have good
long‐term BP control, the hypertrophic left ventricle is difficult to
return to normal.[38] ^6 Animal studies have demonstrated that some
spontaneously hypertensive rats develop ventricular hypertrophy before
the increase of BP.[39] ^7 There is mounting evidence to suggest that
the gut microbiome plays an important role in the development and
pathogenesis of hypertension.[40] ^8 Therefore, we believe that
hypertensive LVH is not only a simple process of cell volume increase
and corresponding subcellular structural changes caused by hemodynamic
overload but also a reasonable remodeling process of myocardial
structure, function, and metabolism under the regulation of
neurohumoral factors.[41] ^9 , [42]^10 , [43]^11 We sought to identify
new intervention targets for the treatment and reversal of LVH through
a novel approach to reduce the adverse outcomes of hypertensive LVH,
the risk of heart failure, and the hospitalization rate of heart
failure.
The gut microbiota plays an important role in maintaining human health;
it interacts with the host, and each influences the other.[44] ^12
Imbalances in the composition and function of these gut microbes are
linked to many diseases, such as obesity,[45] ^13 diabetes,[46] ^14
metabolic syndrome,[47] ^15 cardiovascular disease,[48] ^16 and other
chronic diseases. The gut microbiota mediates microbiome–host genomic
cross‐talk by sending biological signals to the host, including
microbial metabolites and proinflammatory molecules.[49] ^17 Metabolic
remodeling is a precursor to most other pathological changes and may
play an important role in cardiac hypertrophy and heart failure.[50] ^7
, [51]^18 , [52]^19 During hypertensive LVH, metabolic changes have
been observed in the cardiac muscle, which prefers to convert from
fatty acids to glucose to produce ATP.[53] ^7 Short‐chain fatty acids
are reported to affect host physiology and the cardiovascular system
and are thought to play a part in regulating blood pressure.[54] ^20
Organ et al[55] ^21 found that acetate supplementation significantly
reduced systolic and diastolic BP, cardiac fibrosis, and LVH. There are
limited studies indicating a direct association between gut microbe and
metabolites and LVH, especially in clinical trials. There are some
important gaps in knowledge about the gut and LVH that remain
unexplored, and critical issues that should be addressed, such as the
microbial profiles of patients with LVH in the real world, the
metabolite signature profiles, and microbial biomarkers for early
detection of LVH.
Our study aimed to address the following issues: (1) to find
alterations in microbial profiles and metabolite profiles of LVH; (2)
to identify potential biomarkers for early detection of LVH populations
in clinical trials; (3) to determine the relationship between gut
microbe and metabolites; and (4) to identify the metabolic pathways
involved in LVH and hypertension. Our study can provide ideas for early
intervention targets to prevent patients with hypertension from
developing LVH and to provide possible means of reversal for patients
who have already developed LVH.
METHODS
Data Availability Statement
To minimize the possibility of unintentionally sharing information that
can be used to reidentify private information, a subset of the data
generated for this study is available at the National Center for
Biotechnology Information under the BioProject and can be accessed at
[56]https://www.ncbi.nlm.nih.gov/bioproject/1065087. All untargeted
metabolomic data used in this publication have been deposited to the
European Molecular Biology Laboratory–European Bioinformatics Institute
MetaboLights database with the identifier MTBLS9381 (fecal
metabolomics) and MTBLS9370 (serum metabolomics). The complete data set
can be accessed at [57]https://www.ebi.ac.uk/metabolights/MTBLS9370 and
[58]https://www.ebi.ac.uk/metabolights/MTBLS9381.
Participants
The study was conducted in the Second Affiliated Hospital of Baotou
Medical College, Inner Mongolia, China, between November 8, 2021, and
July 1, 2022. Our study recruited 90 participants, including 30
patients with hypertension, 30 patients with hypertensive LVH, and 30
self‐reported healthy subjects (controls). All 3 groups were matched
for sex and age. Clinical information was collected, including age,
sex, height, weight, body mass index, systolic BP, diastolic (BP),
heart rate, fasting blood glucose, triglycerides, total cholesterol,
high‐density lipoprotein, low‐density lipoprotein, uric acid,
creatinine, blood routine, history of antihypertensive medication,
history of smoking, and alcohol status. Individuals were excluded if
they had secondary hypertension; were hospitalized for acute myocardial
infarction or unstable angina pectoris with coronary angiography
stenosis >50%; had previous revascularization surgery (percutaneous
coronary intervention or coronary artery bypass grafting); had acute or
chronic heart failure; had valvular heart disease, dilated or
hypertrophic cardiomyopathy, rheumatic heart disease, or congenital
heart disease; had diabetes; had severe liver or kidney disease or
dialysis; had serum creatinine >221 mol/L (>2.5 mg/dL); had a malignant
tumor; had infections, diarrhea, and anti‐inflammatory use in the past
week; or had other inflammatory diseases. Fecal and plasma samples were
collected for comparison of gut microbial communities using 16S rDNA
and of metabolites using nontargeted metabolomics (liquid
chromatography–mass spectrometry). Feces and plasma samples were
collected at the hospital and stored immediately at −80 °C until
processing. Only 30 stool specimens were collected, 10 from each group;
90 blood samples were collected. All subjects signed informed consent
forms. The present study was approved by the local ethics committee
(LW‐018). This study followed the Strengthening the Reporting of
Observational Studies in Epidemiology cohort reporting guidelines.[59]
^22
Measurement of Blood Pressure and Cardiac Structure
Resting BP was measured by experienced physicians, and the average of 3
readings was taken as our measure of BP. Hypertension was defined as
systolic BP ≥140 mm Hg or diastolic BP ≥90 mm Hg or a previous
diagnosis of hypertension.[60] ^23
All participants underwent full M‐type, 2‐dimensional, and Doppler
echocardiographic examinations, which were performed by the same
well‐trained physician who was blind to the participants' grouping. The
physician used a VIVID S70n ultrasound system and M5Sc probes, and the
patient was breathing quietly in the left‐lying position. The following
echocardiographic parameters were collected: left ventricular
end‐diastolic diameter, left ventricular posterior wall thickness, and
interventricular septal thickness. The left ventricular mass index
(LVMI) was calculated as left ventricular mass divided by body surface
area. LVH was defined according to the 2013 European Society of
Hypertension/European Society of Cardiology guidelines for the
management of arterial hypertension[61] ^24 : LVMI >95 g/m^2 in women
and >115 g/m^2 in men. The formulations are as follows:
graphic file with name JAH3-13-e034230-e004.jpg
[MATH: Body surface area=0.007184×heightcm×0.125×weightkg×0.425m2 :MATH]
[MATH: LVMI(g/m2)=left ventricularend‐diastolic
diameterg/body surface
area(m2) :MATH]
16S rDNA
Total genome DNA from samples were extracted using a Magnetic Soil and
Stool DNA Kit (Tiangen Biotech, Beijing, China). DNA concentration and
purity were monitored on 1% agarose gels. According to the
concentration, DNA was diluted to 1 ng/μL using sterile water. 16S rRNA
genes were amplified using the specific primer with the barcode. All
polymerase chain reaction (PCRs) were carried out in 30‐μL reactions
with 15 μL of Phusion High‐Fidelity PCR Master Mix (New England
Biolabs, Ipswich, MA); 0.2 μmol/L of forward and reverse primers, and
about 10 ng template DNA. Thermal cycling consisted of initial
denaturation at 98 °C for 1 minute, followed by 30 cycles of
denaturation at 98 °C for 10 seconds, annealing at 50 °C for
30 seconds, and elongation at 72 °C for 30 seconds and finally 72 °C
for 5 minutes. The same volume of 1X loading buffer (containing SYBR
green) was mixed with PCR products and electrophoresis on 2% agarose
gel operated for detection. Samples with bright main strips between 400
and 450 bp were chosen for further experiments. PCR products were mixed
in equidensity ratios. Then, a mixture of PCR products was purified
with Qiagen Gel Extraction Kit (Qiagen, Venlo, The Netherlands).
Sequencing libraries were generated using TruSeq DNA PCR‐Free Sample
Preparation Kit (Illumina, San Diego, CA) following manufacturer's
recommendations and index codes were added. The library quality was
assessed on the Qubit@ 2.0 Fluorometer (Thermo Scientific, Waltham, MA)
and Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara,
CA). At last, the library was sequenced on an Illumina NovaSeq6000
platform and 250‐bp paired‐end reads were generated. Paired‐end reads
from the original DNA fragments were merged using FLASH, which was
designed to merge paired‐end reads when at least some of the reads
overlap the read generated from the opposite end of the same DNA
fragment. Paired‐end reads were assigned to each sample according to
the unique barcodes.
Metabonomics
The plasma and fecal samples were thawed at 4 °C, and 100‐μL aliquots
were mixed with 400 μL of cold methanol/acetonitrile (1:1, v/v) to
remove the protein. The mixture was centrifuged for 20 minutes
(14 000g, 4 °C). The supernatant was dried in a vacuum centrifuge. For
liquid chromatography–mass spectrometry analysis, the samples were
redissolved in 100 μL acetonitrile/water (1:1, v/v) solvent and
centrifuged at 14 000g at 4 °C for 15 minutes, and then the supernatant
was injected. Analysis was performed using an ultra‐high‐performance
liquid chromatography (Vanquish UHPLC, Thermo) coupled to a quadrupole
time‐of‐flight (TripleTOF 6600; AB Sciex, Concord, ON, Canada) in
Shanghai Applied Protein Technology Co., Ltd. For hydrophilic
interaction liquid chromatography separation, samples were analyzed
using a 2.1 mm×100 mm Acquity UPLC BEH Amide 1.7‐μm column (Waters
Corp., Wexford, Ireland). In both electrospray ionization–positive and
–negative modes, the mobile phase contained A=25 mmol/L ammonium
acetate and 25 mmol/L ammonium hydroxide in water and B=acetonitrile.
The gradient was 98% B for 1.5 minutes and was linearly reduced to 2%
in 10.5 minutes, and then kept for 2 minutes, and then increased to 98%
in 0.1 minute, with a 3‐minute reequilibration period employed. The
electrospray ionization source conditions were set as follows: Ion
Source Gas1 as 60, Ion Source Gas2 as 60, and curtain gas as 30; source
temperature: 600 °C, IonSpray Voltage Floating±5500 V. In mass
spectrometry–only acquisition, the instrument was set to acquire over
the m/z range 80 to 1200 Da, the resolution was set at 60 000, and the
accumulation time was set at 100 milliseconds. In auto tandem mass
spectrometry acquisition, the instrument was set to acquire over the
m/z range 70 to 1200 Da, the resolution was set at 30 000, and the
accumulation time was set at 50 milliseconds and excluded time within
4 seconds. The raw mass spectrometry data were converted to MzXML files
using ProteoWizard MSConvert before importing into freely available
XCMS software. For peak picking, the following parameters were used:
centWave m/z=10 ppm, peakwidth=c (10, 60), prefilter=c (10, 100). For
peak grouping, bw=5, mzwid=0.025, minfrac=0.5 were used. CAMERA
(Collection of Algorithms of MEtabolite pRofile Annotation) was used
for annotation of isotopes and adducts. In the extracted ion features,
only the variables having >50% of the nonzero measurement values in at
least 1 group were kept, and then the nonzero measurement values were
filled by knn algorithm. Compound identification of metabolites was
performed by comparing of accuracy m/z value (<10 ppm) and tandem mass
spectrometry spectra with an in‐house database established with
available authentic standards.
Statistical Analysis
Statistical analyses were performed using R software (R Foundation for
Statistical Computing, Vienna, Austria). Normally distributed variables
were expressed as mean±SD, and the ANOVA test was used; abnormally
distributed variables were shown as the median (quartile), and
nonparametric (Kruskal–Wallis) tests were used. Categorical data were
expressed as the numbers and percentages, using Pearson's χ^2 test or
Fisher's exact test, as appropriate. Multiple testing correction was
performed using Bonferroni correction, and the significance of the P
value was assessed at 0.0167 (0.05/3), as pairwise comparisons were
performed 3 times among the 3 groups. A multivariable linear regression
analysis was performed between specific gut microbiota and LVMI.
Multivariate‐adjusted models were used: model 1: adjusted age and sex;
model 2=model 1+body mass index+low‐density lipoprotein+total
cholesterol+fasting blood glucose+history of antihypertensive
medication. Spearman correlation analysis was performed to determine
the correlation between differential metabolites and differential flora
and the correlation between metabolites and LVMI. Subgroup analyses
were also used to investigate the sex differences in associations
between metabolites and LVMI.
Sequences analyses were performed with UPARSE software package using
the UPARSE‐OTU and UPARSE‐OTUref algorithms. In‐house Perl scripts were
used to analyze α (within samples) and β (among samples) diversity.
Sequences with ≥97% similarity were assigned to the same operational
taxonomic units (OTUs). We pick a representative sequence for each OTU
and use the RDP classifier to annotate taxonomic information for each
representative sequence (Silva 132; [62]https://www.arb‐silva.de/).
Rarefaction curves were generated on the basis of these 3 metrics.
Graphical representation of the relative abundance of bacterial
diversity from phylum to species can be visualized using Krona chart.
Principal component analysis was applied to reduce the dimension of the
original variables using the QIIME software package. QIIME calculates
both weighted and unweighted unifrac distance, which are phylogenetic
measures of β diversity. We used weighted unifrac distance for
principal coordinate analysis. Linear discriminant analysis (LDA)
effective size was used for the quantitative analysis of biomarkers
within different groups. LDA effective size >2 and P value <0.05 were
used to screen significantly changed bacteria. Analysis of similarities
and permutational multivariate ANOVA were performed on the basis of the
Bray–Curtis dissimilarity distance matrices to identify differences of
microbial communities between the 2 groups.
After sum normalization, the processed metabolomics data were analyzed
by the R package ropls, where it was subjected to multivariate data
analysis, including Pareto‐scaled principal component analysis, partial
least‐squares discrimination analysis (PLS‐DA), and orthogonal PLS‐DA.
The 7‐fold cross‐validation and response permutation testing were used
to evaluate the robustness of the model. The variable importance in the
projection value of each variable in the orthogonal PLS‐DA model was
calculated to indicate its contribution to the classification. The
Kruskal–Wallis test with post hoc multiple comparison was applied to
determine the significance of differences among 3 groups of independent
samples. Variable importance in the projection >1 and P value <0.05
were used to screen significantly changed metabolites. All identified
metabolites were mapped to the Kyoto Encyclopedia of Genes and Genomes
(KEGG) database to find general biochemical pathways
([63]http://geneontology.org/). The KEGG pathway enrichment analysis
was performed using MBROLE version 2.0
([64]http://csbg.cnb.csic.es/mbrole2/). KEGG differential abundance
(DA) score plots were performed by ggplot2 package (R software). DA
score is calculated by first applying a nonparametric DA test to all
metabolites in a pathway. Then, after determining which metabolites are
significantly increased/decreased in abundance, the DA score is defined
as[65] ^25 :
[MATH: No.of metabolites
increased−No.of metabolites
decreasedNo.of measured metabolites in
pathway :MATH]
A score of 1 indicates all measured metabolites in the pathway
increase, and − 1 indicates all measured metabolites in a pathway
decrease.
RESULTS
Clinical Features of the Cohort
Our cohort recruited 30 patients with hypertension, 30 patients with
LVH, and 30 self‐reported healthy subjects, aged 63.9±8.3 years, 37 of
whom were women (41.1%). The clinical characteristics of the cohort are
shown in Table [66]1. We must acknowledge that power is low with so few
people. No differences were observed among groups concerning the age,
sex, height, body mass index, total cholesterol, triglycerides,
high‐density lipoprotein, low‐density lipoprotein, fasting blood
glucose (FBG), heart rate, and alcohol status. systolic BP, diastolic
BP, smoking status, left ventricular end‐diastolic diameter,
interventricular septal thickness, left ventricular posterior wall
thickness, left ventricular end‐diastolic diameter and LVMI were
significantly different among the 3 groups.
Table 1.
Demographics and Clinical Features of the Cohort
Variables Total (n=90) LVH (n=30) Hypertension (n=30) Control (n=30) P
value
Sex, n (%) 0.725
Male 53 (58.9) 18 (60) 19 (63.3) 16 (53.3)
Female 37 (41.1) 12 (40) 11 (36.7) 14 (46.7)
Age, y 63.9±8.3 64.1±9.1 64.0±8.0 63.5±8.0 0.963
Smoking, n (%) 0.001
No 65 (72.2) 17 (56.7) 19 (63.3) 29 (96.7)
Yes 25 (27.8) 13 (43.3) 11 (36.7) 1 (3.3)
Alcohol status, n (%) 0.067
No 72 (80.0) 21 (70) 23 (76.7) 28 (93.3)
Yes 18 (20.0) 9 (30) 7 (23.3) 2 (6.7)
History of hypertension medication, n (%) <0.001
No 31 (34.4) 1 (3.3) 0 (0) 30 (100)
Yes 59 (65.6) 29 (96.7) 30 (100) 0 (0)
Height, cm 166.4±7.2 168.0±7.5 166.6±7.2 164.6±6.6 0.195
Weight, kg 68.8±12.1 72.4±13.7 69.6±12.1 64.5±9.2 0.037
Body mass index, kg/m^2 24.7±3.2 25.6±3.9 24.9±3.1 23.7±2.4 0.076
Total cholesterol, mmol/L 4.6±0.9 4.7±0.9 4.4±1.0 4.8±0.7 0.175
Triglycerides, mmol/L 1.4 (1.0–1.9) 1.7 (1.1–2.1) 1.4 (1.1–1.9) 1.1
(0.8–1.6) 0.125
HDL, mmol/L 1.3 (1.0–1.6) 1.1 (1.0–1.5) 1.2 (1.0–1.5) 1.4 (1.2–1.8)
0.061
LDL, mmol/L 2.9±0.8 3.0±0.8 2.8±0.9 2.9±0.6 0.589
Fasting blood glucose, mmol/L 5.4±0.7 5.3±0.7 5.2±0.6 5.7±0.8 0.016
Creatinine, mmol/L 73.6±17.1 75.9±20.5 75.3±12.6 69.3±17.5 0.285
Uric acid, mmol/L 317.0±85.5 282.7±73.4 319.3±77.3 322.1±97.0 0.589
Systolic BP, mm Hg 135.0 (125.0–143.8) 141.5 (136.0–150.0) 140.0
(127.0–144.0) 122.5 (114.0–130.8) <0.001
Diastolic BP, mm Hg 82.8±12.6 88.6±11.6 85.0±13.4 74.7±8.1 <0.001
Heart rate, beats 76.4±10.3 78.3±10.3 74.8±10.2 76.2±10.5 0.409
LVDD, mm 4.7 (4.4–4.9) 4.8 (4.6–5.3) 4.6 (4.3–4.8) 4.5 (4.3–4.8) <0.001
Interventricular septal thickness, mm 1.0 (0.9–1.1) 1.2 (1.1–1.4) 0.9
(0.8–1.0) 0.9 (0.8–0.9) <0.001
LVPWT, mm 0.9 (0.8–1.0) 1.0 (1.0–1.2) 0.9 (0.8–1.0) 0.8 (0.8–0.9)
<0.001
LVM, g 146.7 (123.1–181.8) 219.4 (177.5–254.6) 137.7 (123.0–151.4)
125.5 (113.6–142.5)
<0.001
0.008
LVMI, g/m^2 91.8±25.3 121.4±18.8 78.7±10.8 75.3±11.3 <0.001
[67]Open in a new tab
Data are presented as mean±SD, n (%), or median (quartile). Normally
distributed variables were used for ANOVA test, abnormally distributed
variables were used for nonparametric tests (Kruskal–Wallis test); χ^2
or Fisher test for categorical variables. Normality was examined using
the Shapiro–Wilk test. Bonferroni's correction was applied
(0.05/3=0.0167). BP indicates blood pressure; HDL, high‐density
lipoprotein; LDL, low‐density lipoprotein; LVDD, left ventricular
end‐diastolic diameter; LVH, left ventricular hypertrophy; LVM, left
ventricular mass; LVMI, left ventricular mass index; and LVPWT, left
ventricular posterior wall thickness.
Gut Microbial
Rarefaction of samples to 55 000 reads showed consistent and plateauing
diversity metrics across samples (Figure [68]S1). The top 10
differentially enriched bacterial taxa of each group are presented in
our study. At the phylum level, hypertension, LVH, and control group,
all presented Firmicutes and Bacteroidetes as the major. Compared with
the control group, the proportion of Firmicutes and Actinobacteria
increased the most in the LVH group, followed by the hypertension
group. However, Bacteroidetes was significantly reduced in the LVH
group, followed by the hypertension group. In addition, Proteobacteria
and Tenericutes were significantly decreased in the LVH group and
significantly increased in the hypertension group, while
Verrucomicrobia was significantly increased in the LVH group and
significantly decreased in the hypertension group. An increased ratio
of Firmicutes/Bacteroidetes (F/B) in the gut microbiota of both
individuals with LVH and individuals with hypertensive compared with
controls was found, resulting from a reduction of Bacteroidetes and an
enrichment of Firmicutes (Figure [69]1). More detailed taxonomic levels
are shown in Figures [70]S2 through [71]S7. The above results indicate
that the imbalanced structure of the intestinal flora in patients with
LVH and hypertension and the alterations of some intestinal flora of
the 2 groups were different, suggesting that these changes of gut
microbiata increase the susceptibility to LVH.
Figure 1. The top 10 differentially enriched bacterial taxa of each group at
phylum level.
Figure 1
[72]Open in a new tab
HP indicates hypertension; and LVH, left ventricular hypertrophy.
The microbial β diversity (between samples) and α diversity (within
samples) metrics were quantified in our study. We observed an elevation
in bacterial Chao and Abundance‐Based Coverage Estimator richness in
the LVH group, although the difference was not statistically
significant (P>0.05; Table [73]S1, Figure [74]S8), which may suggest
the overgrowth of a variety of harmful bacteria in patients with LVH.
There was no association between α diversity and LVMI and BP (data not
shown). The microbial β diversity showed no significant differences in
gut microbiome composition among the 3 groups (Figure [75]2,
Figure [76]S9). Some specific taxa in control, hypertension, or LVH
groups were prevalent, detailed by LDA effective size (Figure [77]3A
through [78]3C). We detected a total of 44, 16, and 13 biomarkers with
LDA scores >2 that were significantly different between the LVH group
and controls, the hypertension group and controls, and the LVH and
hypertension groups, respectively. Of these, the specific taxa with the
highest LDA score in LVH belongs to Firmicutes, and in controls and
hypertension belong to Bacteroidetes. At the genus level, we identified
that individuals with LVH had higher levels of genera such as
Streptococcus, Romboutsia, Lactobacillus, Rothia, Granulicella,
Marmoricola, Enorma, Leuconostoc, Peptoniphilus, Faecalitalea,
Turicibacter, Desulfovibrio, Uruburuella, Ruminococcus torques group,
and Erysipelotrichaceae UCG‐003, while subjects with hypertension had
higher levels of Lactobacillus, Granulicatella, and Eubacterium
coprostanoligenes group, healthy participants had higher levels of
Roseburia, Acetitomaculum, Fournierella, Odoribacter, Prevotella 2,
Christensenellaceae R7 group, Eubacterium ruminantium group,
GCA_900 066 575, Lachnospira, Lachnospiraceae NK4A136 group, and
Lachnospiraceae UCG‐010.
Figure 2. Comparison of the microbial β diversity (as accessed by principal
coordinate analysis used weighted unifrac distance) in the 3 groups.
Figure 2
[79]Open in a new tab
ANOSIM (analysis of similarities) is a nonparametric test based on the
Bray–Curtis algorithm. R>0 indicates significant differences between
groups, R<0 indicates significant difference within the group is
greater than the difference between the groups. ADONIS is known as
permutational multivariate ANOVA or nonparametric multivariate ANOVA.
The F score ranges from 0 to 1, with the higher the value, the better
the model performance. Top left: LVH vs HP vs control; bottom left: LVH
vs HP; top right: LVH vs control; bottom right: HP vs control. HP
indicates hypertension; and LVH, left ventricular hypertrophy.
Figure 3. Linear discriminant analysis effect size (LEfSe) of each cohort to
visualize differences in bacterial species.
Figure 3
[80]Open in a new tab
A, LVH vs control; B, LVH vs HP; and C, HP vs control. LEfSe LDA>2 and
P value<0.05 were used to screen significant changed bacteria. HP
indicates hypertension; and LVH, left ventricular hypertrophy.
We performed correlations between the microbiome and LVMI (Table [81]2,
Table [82]S2). LVH‐enriched Faecalitalea (β=6758.55 [95% CI,
2080.92–11436.18]; P=0.009), Turicibacter (β=8424.76 [95% CI,
2494.05–14355.47]; P=0.01), Ruminococcus torques group (β=840.88 [95%
CI, 223.1–1458.67]; P=0.013), and Erysipelotrichaceae UCG‐003 (β=856.37
[95% CI, 182.76–1529.98]; P=0.019) were positively correlated with
LVMI, but control‐enriched GCA.900066575 (β=−46478.71 [95% CI=−90661.93
to −2295.49]; P=0.049) was negatively linked to LVMI. After adjusting
for 2 models, only Faecalitalea (β=6077.34 [95% CI, 1633.86–10520.82];
P=0.014), Ruminococcus torques group (β=853.73 [95%CI, 283.14–1424.32];
P=0.008), and Erysipelotrichaceae UCG‐003 (β=880.05 [95% CI,
282.6–1477.5]; P=0.009) remained statistically significantly associated
with LVMI. It is worth noting that the bacteria genera that distinguish
LVH from the hypertension group were Faecalitalea, Erysipelotrichaceae
UCG‐003, Ruminococcus torques group, and Turicibacter, which were
highly expressed in LVH.
Table 2.
Relationship Between Gut Microbiota and LVMI
Model 1, β (95% CI); P value Model 2, β (95% CI); P value
Turicibacter 8424.76 (2494.05 to 14355.47); 0.01 5770.46 (−1347.22 to
12888.13); 0.127
Faecalitalea 6758.55 (2080.92 to 11436.18); 0.009 6077.34 (1633.86 to
10520.82); 0.014
Enorma 124337.08 (4889.52 to 243784.63); 0.052 94315.07 (−26841.66 to
215471.8); 0.142
Ruminococcus torques group 840.88 (223.1 to 1458.67); 0.013 853.73
(283.14 to 1424.32); 0.008
Erysipelotrichaceae UCG‐003 856.37 (182.76 to 1529.98); 0.019 880.05
(282.6 to 1477.5); 0.009
GCA.900066575 −46478.71 (−90661.93 to −2295.49); 0.049 −31365.99
(−80514.88 to –17782.9); 0.225
[83]Open in a new tab
Model 1: adjusted for age and sex; model 2=Model 1+body mass index;
low‐density lipoprotein; total cholesterol; fasting blood glucose;
history of antihypertensive medication. LVMI indicates left ventricular
mass index.
Metabolomics (Feces and Plasma)
Many metabolites of the gut microbiota serve as microbe–host bridges
and can influence host physiology either in the gut or through entry
into the bloodstream. Therefore, the levels of the intestinal metabolic
profiling from feces and plasma were both measured in our study. A
subset of 30 participants (10 from each group) from the present study
were enrolled in the feces study (Tables [84]S3). PLS‐DA and orthogonal
PLS‐DA under both positive ion (ES+) and negative ion (ES−) modes
succeed in distinguishing samples of plasma and fecal metabolomics
between the LVH and control groups, the LVH and hypertension groups,
and the hypertension and control groups (Figures [85]4A and [86]4B;
Figures [87]5A and [88]5B). Variable importance in the projection plots
were also provided (Figures [89]S10 and [90]S11). The metabolites of 3
groups were identified in ES+ and ES− mode, and the metabolites of the
3 groups were significantly different (Figure [91]6). These results
indicated that LVH and hypertension significantly altered the
intestinal flora metabolic profiles.
Figure 4. PLS‐DA score plots based on the metabolic profiles from the
control, LVH, and HP groups in ES+ and ES−.
Figure 4
[92]Open in a new tab
A, In plasma samples; and B, in feces samples. The 7‐fold
cross‐validation and response permutation testing were used to evaluate
the robustness of the model. ES+ indicates positive ion; ES–, negative
ion; HP, hypertension; LVH, left ventricular hypertrophy; and PLS‐DA,
partial least‐squares discriminant analysis.
Figure 5. OPLS‐DA score plots based on the metabolic profiles from the
control, LVH, and HP groups in ES+ and ES−.
Figure 5
[93]Open in a new tab
A, In plasma samples; and (B) in feces samples. The 7‐fold
cross‐validation and response permutation testing were used to evaluate
the robustness of the model. ES+ indicates positive ion; ES–, negative
ion; HP, hypertension; LVH, left ventricular hypertrophy; and OPLS‐DA,
orthogonal partial least‐squares discriminant analysis.
Figure 6. Venn diagram shows the number of altered metabolites in plasma and
feces samples from the control, LVH, and HP groups in ES+ and ES−.
Figure 6
[94]Open in a new tab
ES+ indicates positive ion; ES–, negative ion; HP, hypertension; and
LVH, left ventricular hypertrophy.
In plasma tests, a total of 1141 metabolites were identified. A total
of 132 (ES+ mode, 72; ES− mode, 60) and 49 (ES+ mode, 20; ES− mode, 29)
metabolites were significantly different between the LVH and control
groups and the LVH and hypertension groups, respectively. In addition,
116 (ES+ mode, 66; ES− mode, 50) serum metabolites were significantly
different between hypertension and controls. Of these, the majority
were lipids and lipid‐like molecules. The expression of the lipids and
lipid‐like molecules was downregulated in both LVH and hypertension.
Inversely, the majority of the organic acids and derivatives and
organoheterocyclic compounds were upregulated in LVH and hypertension
(Figures [95]7A and [96]7C, Tables [97]S4 and [98]S5). Notably,
metabolites such as bile acids (BAs), trimethylamine N‐oxide (TMAO),
linolenic acid (LA), and choline were significantly changed in patients
with LVH. The free BAs of deoxycholic acid (DCA) was significantly
upregulated, while the contents of conjuncted BAs such as glycocholic
acid was significantly downregulated. In addition, the content of TMAO
and choline in LVH was significantly upregulated, while LA in LVH was
significantly downregulated. TMAO is derived mainly from the oxidation
of trimethylamine by the gut flora. Intestinal microorganisms
metabolize the ingested lecithin and choline and produce
trimethylamine, which is then oxidized by flavin monooxygenase 3 or
other flavin monooxygenase to produce TMAO.[99] ^26 Both BAs and TMAO
have been found to be involved in the occurrence and development of
hypertension, cardiovascular disease, diabetes, and other diseases, and
may be related to inflammatory response.[100] ^27 , [101]^28 , [102]^29
, [103]^30 , [104]^31 , [105]^32 Inflammation and oxidative stress are
the main factors causing BP increase and organ damage, such as LVH, in
hypertension. This is a possible reason for the changes in BAs and TMAO
levels in patients with LVH. LA and α‐LA, are essential fatty acids
that have been shown to have a variety of beneficial effects, including
antimetabolic syndrome effects, antimetabolic syndrome, anticancer,
antiinflammatory, antioxidant, antiobesity, neuroprotection, and
regulation of the intestinal flora properties.[106] ^33
Figure 7. The relative amount of plasma metabolites varied in the LVH, HP,
and control groups.
Figure 7
[107]Open in a new tab
A, LVH vs HP; B, LVH vs control; and C, HP vs control. Variable
importance in the projection >1 and P value <0.05 were used to screen
significant changed metabolites. HP indicates hypertension; and LVH,
left ventricular hypertrophy.
In fecal metabolite tests, a total of 2657 metabolites were identified.
Seventy‐four (ES+ mode, 53; ES− mode, 21) and 59 (ES+ mode, 31; ES−
mode, 28) fecal metabolites were significantly different between LVH
and controls and LVH and hypertension, respectively. Additionally, 95
(ES+ mode, 69; ES− mode, 26) fecal metabolites were significantly
different between the hypertension and controls. Different from the
differential metabolites in blood, organic acids and derivatives were
mostly found in feces. The expression of the lipids and lipid‐like
molecules were downregulated in the LVH group compared with the
hypertension group, but its expression was upregulated in the LVH group
compared with the controls. Lipids and lipid‐like molecules were
down‐expressed in the LVH group compared with the controls and
up‐expressed when compared with the hypertension group. When compared
with controls, organic acids and derivatives were upregulated and
organoheterocyclic compounds were downregulated in the hypertension
group (Figure [108]8A through [109]8C; Tables [110]S6 and [111]S7). The
free BAs, including DCA, chenodeoxycholic acid (CDCA), ursodeoxycholic
acid, lithocholic acid, and hyodeoxycholic acid, as well as
l‐homocitrulline, were significantly upregulated in patients with LVH.
Homocitrulline is produced by carbamylation. Carbamylated proteins are
associated with atherosclerosis.[112] ^34 A recent study found that
l‐homocitrulline was associated with nonalcoholic fatty liver
disease.[113] ^35 Nicotinuric acid, the main catabolite of niacin, was
considered a good indicator to evaluate the biotransformation of niacin
in the liver.[114] ^36 A study proved that urinary nicotinuric acid
level is positively correlated with metabolic syndrome.[115] ^37
Figure 8. The relative amount of fecal metabolites varied in the LVH, HP, and
control groups.
Figure 8
[116]Open in a new tab
A, LVH vs HP; B, LVH vs control; and C, HP vs control. Variable
importance in the projection >1 and P value <0.05 were used to screen
significant changed metabolites. HP indicates hypertension; and LVH,
left ventricular hypertrophy.
It is worth mentioning that there were metabolites identified in both
serum and feces. We found 2 (zerumbone and DCA) and 1 (pregnenolone
sulfate) common metabolites (with KEGG ID) both in plasma and feces
between the LVH control groups and the hypertension and control groups,
respectively. No common metabolites in plasma and feces were found in
the LVH and hypertension groups (Figure [117]9; Tables [118]S8 through
[119]S10). Zerumbone and DCA showed the same tendency both in serum and
feces. However, pregnenolone sulfate showed an increased tendency in
feces but decreased in serum (Figures [120]7B and [121]7C;
Figures [122]8B and [123]8C). This pathogenic substance might originate
from a pathway other than gut microbes.[124] ^38
Figure 9. Venn diagram of the altered metabolites shared between plasma
(blue) and feces (red).
Figure 9
[125]Open in a new tab
A, common metabolites (with KEGG ID) both in plasma and feces between
LVH and the control group; and B, common metabolites (with KEGG ID)
both in plasma and feces between HP and control group. HP indicates
hypertension; KEGG, Kyoto Encyclopedia of Genes and Genomes; and LVH,
left ventricular hypertrophy.
We examined the relationship between differential metabolites in the
LVH group and LVMI and found that 131 (ES+ mode, 74; ES− mode, 57)
metabolites in plasma and 56 (ES+ mode, 46; ES− mode, 10) metabolites
in feces were associated with LVMI. The correlation between metabolites
and LVMI differed across the sex stratification (Figures [126]S12
through [127]S15). This suggests that there was a sex‐specific
association between metabolites and LVMI. Spearman correlation analysis
showed that glycocholic acid and LA was significantly negatively
correlated with LVMI, while hyodeoxycholic acid, CDCA, lithocholic
acid, DCA, TMAO, choline, l‐homocitrulline, and nicotinuric acid were
significantly positively correlated with LVMI. These metabolic
variations might aggravate or even promote the pathological processes
of LVH.
Functional Alteration in Metabolic Pathways of Hypertension and LVH
We performed an analysis based on the KEGG pathway using metabolites of
DA between the LVH and control and hypertension groups to identify the
metabolic pathways associated with the pathogenesis of LVH.
Differential abundance scores were calculated, which captured the
tendency for metabolites in a pathway to be increased/decreased
relative to the control and hypertension groups.
In plasma tests, compared with the control group, there were 6 and 11
significantly different metabolic pathways in the LVH and hypertension
groups, respectively. Among them, 3 pathways were overrepresented in
the LVH group, mainly involved in amino acid metabolism (arginine and
proline metabolism), central carbon metabolism in cancer, and digestive
system (protein digestion and absorption pathways); and 6 pathways were
overrepresented in the hypertension group, amino acid metabolism
(arginine and proline metabolism; glycine, serine, and threonine
metabolism), digestive system (protein digestion and absorption,
mineral absorption), central carbon metabolism in cancer, and choline
metabolism in cancer (Figure [128]10). Upregulation of amino acid
metabolic and protein digestion and absorption pathways has also been
reported in diabetes[129] ^39 and gastric adenocarcinoma,[130] ^40
respectively.
Figure 10. Functional alteration in metabolic pathways of HP and LVH.
Figure 10
[131]Open in a new tab
The DA score captures the average, gross changes for all metabolites in
a pathway. The length of the line segment represents the absolute value
of DA score, and the size of the dot at the end of the line segment
represents the number of metabolites in the pathway. The larger the
dot, the greater the number of metabolites. The darker the red of the
line segment and the dot, the more upregulated the overall expression
of the pathway, and the darker the blue, the more downregulated the
overall expression of the pathway. A score of 1 indicates all measured
metabolites in the pathway increase, and −1 indicates all measured
metabolites in a pathway decrease. DA indicates differential abundance;
HP, hypertension, and LVH left ventricular hypertrophy.
It should be noted that the metabolic pathways with significant
differences between LVH and hypertension in plasma tests included lipid
metabolism (LA metabolism, α‐LA metabolism, biosynthesis of unsaturated
fatty acids) and nervous system pathways (retrograde endocannabinoid
signaling), which were significantly downregulated in the LVH group
(Figure [132]9). Downregulation of the linoleic acid metabolic pathway
has been reported to be associated with metabolic syndrome.[133] ^41 In
addition, there were 4 significantly different metabolic pathways in
fecal tests in the LVH group compared with the hypertension group, of
which 2 pathways (alanine, aspartate, and glutamate metabolism;
β‐alanine metabolism) were downregulated, and 1 pathway was upregulated
(bile secretion) (Figure [134]10). No significantly different pathways
were found in any of the other group comparisons in fecal tests (LVH
versus controls, hypertension versus controls). Consistently, BA
synthesis and excretion is the main pathway of cholesterol and lipid
metabolism and is associated with a variety of metabolic diseases,
including obesity, diabetes, and nonalcoholic fatty liver disease.[135]
^42
Correlations Between Gut Microbiome and Metabolites
The significantly different metabolites in the 3 groups may be
biomarkers for the development of LVH and might be derived from
intestinal flora or their fermented products. To explore the
association between aberrant metabolites and disordered gut microflora,
we carried out a correlation analysis between differential flora and
differential metabolites in plasma or feces (Figures [136]11, [137]12,
[138]13, [139]14). Interestingly, we found a significant positive
correlation between free BAs including DCA, hyodeoxycholic acid, CDCA,
and lithocholic acid and flora including Desulfovibrio, Streptococcus,
Leuconostoc, Marmoricola, and Faecalitalea, which were enriched in LVH.
However, there was a significant negative correlation between these
free BAs and control‐enriched Lachnospira and Lachnospiraceae NK4A136
group. On the contrary, glycocholic acid was negatively correlated with
LVH‐enriched Enorma and positively correlated with control‐enriched
Lachnospira and Lachnospiraceae NK4A136 group. Similarly, metabolites
such as nicotinuric acid, choline, and urobilin were positively
correlated with the bacteria enriched in the LVH group but negatively
correlated with the bacteria in the control group.
Figure 11. Correlations between metabolites in plasma and gut microbiome from
the LVH and control groups.
Figure 11
[140]Open in a new tab
Spearman correlation analysis was used. LVH indicates left ventricular
hypertrophy.
Figure 12. Correlations between metabolites in feces and gut microbiome from
the LVH and control groups.
Figure 12
[141]Open in a new tab
Spearman correlation analysis was used. LVH indicates left ventricular
hypertrophy.
Figure 13. Correlations between metabolites in plasma and gut microbiome from
the LVH and HP groups.
Figure 13
[142]Open in a new tab
Spearman correlation analysis was used. HP indicates hypertension; and
LVH, left ventricular hypertrophy.
Figure 14. Correlations between metabolites in feces and gut microbiome from
the LVH and HP groups.
Figure 14
[143]Open in a new tab
Spearman correlation analysis was used. HP indicates hypertension; and
LVH, left ventricular hypertrophy.
We examined the relationship between metabolites existing in both serum
and feces, including zerumbone, DCA, and pregnenolone sulfate and
microbiome. DCA in plasma and feces is positively correlated with
Marmoricola, Faecalitalea, and Desulfovibrio, and inversely correlated
with Lachnospira. Zerumbone in feces is positively correlated with
Lachnospiraceae UCG‐010 but inversely related to Faecalitalea,
Desulfovibrio, and Uruburuella. Pregnenolone sulfate in plasma was
positively correlated with Odoribacter but inversely related to
Eubacterium coprostanoligenes group. Pregnenolone sulfate in feces was
inversely related to Fournierella (Figures [144]11 and [145]12). The
close relationship between microbes and metabolites suggests that
specific metabolites may be produced directly or indirectly by the
corresponding gut microbes[146] ^38 and that metabolites and gut
microbes may also be upstream or downstream of specific pathways, which
are currently under exploration and require further validation in the
future.
DISCUSSION
Our study revealed a novel mechanism of LVH development by intestinal
flora and metabolomics. The intestinal flora composition and metabolism
were significantly altered in LVH when compared with the hypertension
and control groups. LVH and control participants, as well as LVH and
hypertension participants, could be accurately distinguished by
specific bacteria and metabolites. We found a relationship between gut
microbiota and LVMI. There was a sex‐specific correlation between
metabolites and LVMI. In addition, we also observed different
correlations between metabolites and the flora in different groups.
Furthermore, BA and lipid metabolism pathways were significantly
changed during LVH development.
We observed an increase in the abundance of Chao and Abundance‐Based
Coverage Estimator in the LVH group, although the difference did not
reach statistical significance, which may indicate the tendency of
multiple harmful bacteria overgrowth in patients with LVH. β diversity
of gut microbiota in the 3 groups did not exhibit any significant
differences, which may suggest that the diversity and structure of gut
microbiota have limited impact on the occurrence and development of
LVH. It is not the composition of the whole gut microbiota but may be
certain taxa that change and have effects on LVH. We found differential
microbial taxa by LDA effective size analysis in the 3 groups. The
second possible reason is that our sample size is not large enough to
lead to insufficient statistical power. The third possible reason is
that diet and the use of antihypertensive drugs changed the composition
of the gut microbiome, resulting in no significant difference among the
3 groups. The shift of dominant bacteria in the gut microbiota further
characterized the imbalance of LVH‐specific gut microbial environment.
In the present study, bacteria such as Firmicutes and Actinobacteria
increased in both the LVH and hypertension groups, while Bacteroidetes
decreased. Contrary to our findings, Tsai et al[147] ^43 found that low
levels of the phylum Firmicutes were associated with an increased risk
of LVH. The possible reason is that their study population was
diabetic, unlike our study population, which excluded people with
diabetes. As shown by our study, an increased F/B ratio is represented
in LVH. The specific taxa with the highest LDA score in LVH belongs to
the Firmicutes phylum, and in controls and hypertension belong to the
Bacteroidetes phylum. For example, at the genus level, the
lactate‐producing bacteria Streptococcus and Turicibacter, both of
which belong to the Firmicutes phylum, were in higher quantities in the
LVH. On the contrary, acetate‐producing bacteria Prevotella, which
belong to the Bacteroidetes phylum, was found to be highly accumulated
in the control group. These changes in these genera may be the major
contributors to the increased F/B ratio in the LVH. A higher ratio of
F/B ratio has been widely considered a signature of gut dysbiosis and
has been related to diabetes, hypertension, and cardiovascular
disease.[148] ^44 , [149]^45 In our study, gut microbiota such as
Streptococcus, Romboutsia, Lactobacillus, Rothia, Granulicella,
Marmoricola, Enorma, Leuconostoc, Peptoniphilus, Faecalitalea,
Turicibacter, Desulfovibrio, Uruburuella, Ruminococcus torques group
and Erysipelotrichaceae UCG‐003 were typically enriched in the patients
with LVH. In contrast, healthy participants had higher levels of
Acetitomaculum, Fournierella, Odoribacter, Roseburia, Prevotella_2,
Christensenellaceae R‐7 group, Eubacterium ruminantium group,
GCA_900 066 575, Lachnospira, Lachnospiraceae NK4A136 group, and
Lachnospiraceae UCG‐010. Streptococcus, a genus of gram‐positive
pathological bacteria, has been shown to be associated with meningitis,
pneumonia, endocarditis,[150] ^46 bowel disease,[151] ^47
cirrhosis,[152] ^48 type 1 diabetes,[153] ^49 and hypertension.[154]
^50 Roseburia is one of the major short‐chain fatty acids producer in
human colon. Short‐chain fatty acids play a role in maintaining the
intestinal epithelial barrier, reducing intestinal permeability,
circulating lipopolysaccharide, and systemic inflammation.[155] ^51 A
recent study found that several strains of Turicibacter bacteria from
the mammalian gut microbiota modulate host bile and lipid
compositions.[156] ^52 Some studies have found that an increased
abundance of Turicibacter is associated with a high‐fat diet,[157] ^53
, [158]^54 but other studies have shown opposite relationships.[159]
^55 , [160]^56 The first reason may be that the phenotype of the
Turicibacter strain is different, and the host may experience different
lipid outcomes depending on its own specific Turicibacter strain.
Another reason may be due to host genetic and sex changes that lead to
differences in lipid responses between hosts. A third possible reason
is that the effectiveness of Turicibacterium colonization is influenced
by the biogeographic organization of individual microbiota.[161] ^52 An
improved Streptococcus abundance was also found in simple fat people
and type 1 diabetes.[162] ^49 , [163]^57 It is worth noting that
certain gut microbiota play paradoxical roles like Lactobacillus.
Lactobacillus, in our study, was overrepresented in both the LVH and
hypertension groups. A study from a Brazilian cohort, consistent with
our results, showed that in patients with hypertension, the abundance
of Lactobacillus increased.[164] ^58 In addition, an increase in
Lactobacillus was also observed in ST‐segment–elevation myocardial
infarction.[165] ^59 However, some other studies reported the
beneficial role of Lactobacillus consumption in BP regulation.[166] ^60
, [167]^61 , [168]^62 , [169]^63 The genera of Enorma, Faecalitalea,
Turicibacter, Ruminococcus torques group, and Erysipelotrichaceae
UCG‐003 were positively correlated with LVMI in our study. Two models
were adjusted when we examined the association between microbiota and
LVMI using linear models. Considering that antihypertensive drugs,
lipid profile, fasting blood glucose, and body mass index may have an
effect on the gut microbiome, we included them as covariates in the
adjustment model. An association between Enorma and other diseases was
also reported previously, such as atrial fibrillation,[170] ^64
hypertension,[171] ^61 and spinal muscular atrophy.[172] ^65
Erysipelotrichaceae UCG‐003 have been reported to be associated with
inflammatory bowel diseases.[173] ^66
Metabolic profiles of both plasma and fecal samples analyzed from
patients with LVH demonstrated significant alterations, which could be
derived from altered bacterial functions in patients with LVH. Most
importantly, our study uncovered that the expression of differential
metabolite pathways are different across groups. In particular, the
pathway related to BA metabolism was highly expressed in the LVH group
compared with the hypertension group. This was accompanied by changes
in the plasma and fecal metabolites of BAs, and BAs were significantly
correlated with LVMI. Metabolites from plasma and feces, such as BAs,
can act as a signaling molecule involved in glycolipid metabolism by
activating the nuclear receptor farnesoid X receptor. BAs plays a role
in myocardial hypertrophy, cardiomyocyte apoptosis and cardiac
hemodynamic abnormalities.[174] ^67 In the present study, conjuncted
BAs such as glycocholic acid were down‐expressed and free BAs including
DCA, hyodeoxycholic acid, CDCA, ursodeoxycholic acid, and lithocholic
acid were up‐expressed in patients with LVH. A study using
Langendorff‐perfused isolated heart demonstrated the role of BAs on the
heart, and the results showed that CDCA and DCA could exert positive
inotropic effects by increasing cytoplasmic Ca^2+ levels in
cardiomyocytes and negative isotropic effects by inhibiting sinus node
activity in rat hearts within specific concentration ranges.[175] ^68
TMAO is an important metabolite of intestinal flora, which is
associated with the incidence of hypertension, diabetes, obesity, and
atherosclerosis.[176] ^28 The present study showed that choline and
TMAO were overexpressed in LVH. It is widely known that Firmicutes and
Proteobacteria are producers of trimethylamine and TMAO.[177] ^69 ,
[178]^70 In addition, trimethylamine and TMAO levels were associated
with an increased F/B ratio.[179] ^71 , [180]^72 This provides new
ideas for alleviating disease progression by downregulating the
choline–trimethylamine –TMAO production pathway in gut microbiota. A
meta‐analysis showed that high levels of TMAO were associated with
major adverse cardiovascular events compared with low levels (risk
ratio [RR] =1.62).[181] ^29 There is a dose‐dependent relationship
between TMAO and the risk of death from cardiovascular disease. An
animal study found cardiac remodeling in rats after TMAO intervention,
with increases in left ventricular posterior wall thickness and left
ventricular anterior wall dimensions at both diastole and systole.[182]
^73 TMAO can induce the expression of inflammatory factors and has
obvious proinflammatory effects. The inflammatory response associated
with the development of hypertension is an important factor in the
occurrence of target organ damage, such as LVH. Furthermore, a
decreased level of LA in patients with LVH was observed in our study.
LA and α‐LA have a protective effect against high BP and helps reduce
the risk of cardiovascular disease. A meta‐analysis showed that dietary
LA intake is inversely associated with coronary heart disease risk in a
dose–response manner.[183] ^74
We acknowledge that the present study has several limitations. Our
sample size is relatively small. In the future, further studies with a
larger sample size are warranted to truly understand the association
between LVH and gut microbiota and metabolites. The present study also
lacks mechanistic studies in animal models to determine the causality
and role of gut microbiota on the development of LVH. Furthermore,
information about exercise and diet, which may also have an impact on
the gut microbiota, was not collected and corrected for in our study.
Despite the limitations, our study identified an association between
gut microbiota and its metabolites and LVH in clinical trials. Lack of
early intervention leads to irreversible myocardial hypertrophy and
interstitial fibrosis, resulting in impaired cardiac function and
worsening long‐term outcomes. Our exploration is of great significance
to provide new therapeutic ideas for other therapeutic options besides
drug therapy in clinics, such as intestinal flora transplantation and
metabolite supplementation. Our study can provide ideas for early
intervention targets to prevent patients with hypertension from
developing LVH and to provide means of reversal for patients who have
already developed LVH.
Sources of Funding
This work was supported by the Clinical Research Special Fund of Wu
Jieping Medical Foundation (320.6750.19089‐90).
Disclosures
None.
Supporting information
Tables S1–S10
[184]JAH3-13-e034230-s001.pdf^ (2.5MB, pdf)
Figures S1–S15
[185]JAH3-13-e034230-s002.xlsx^ (46.5MB, xlsx)
Acknowledgments