Estimation of the state of the microbiome in the elderly with impairments of carbohydrate and lipid exchange by the method of microbiome-associated exposomics Translated title: Оценка состояния микробиома у лиц пожилого возраста с нарушениями углеводного и липидного обмена методом микробиом-ассоциированной экспосомики

Abstract

We performed the biochemical analysis and studied the chemical composition of blood samples in 163 people 45-90 years old with type 2 diabetes mellitus and dyslipidemia. We examined the concentrations of the following compounds in the blood samples: fatty acids, aldehydes, styrenes of microbial origin, as well as the levels of glucose, triglycerides, cholesterol, and lipids of low and high density. The chemical composition of blood was determined by gas chromatography-mass spectrometry. The concentrations of fatty acids, aldehydes (including octadecenaldehyde (18a)), and styrenes were used to calculate the total molar concentration of small molecules originating from microbes (SMOM), of hydroxy acids - derivatives of the hydroxyl residue of Lipid A, and of bacterial endotoxin (3OH-FA), as well as grouped total concentrations of chemical compounds of microbial origin, which determine the representation of the main four phylotypes of the human microbiome: Actinobacteria, Bacteroidetes, Proteobacteria, and Firmicutes. Over the course of this study, we obtained data that show the increase in the total concentration of chemical compounds, the concentration of octadecenaldehyde and the concentration of 3OH-FA in patients with carbohydrate metabolism disorders due to diabetes mellitus type 2. We observed a decrease in the representation of Bacteroidetes in patients with carbohydrate metabolism disorders and a decrease in the representation of Proteobacteria and Firmicutes in patients with carbohydrate and lipid metabolism disorders, as well as an increase in the representation of Actinobacteria in patients with lipid metabolism disorders, including patients with combined lipid and carbohydrate metabolism disorders. There was a direct correlation between the representation of Bacteroidetes and the level of triglycerides in patients with type 2 diabetes mellitus as well as an inverse correlation between the representation of Firmicutes and glucose levels in these patients and in control group.

We did not reveal statistically significant changes in the concentrations of microbial markers nor statistically significant correlations between the biochemical blood parameters and the representation of microbiome phylotypes in the blood of patients with lipid metabolism disorders.

Translated abstract

Проведены биохимический анализ крови и исследование химического состава образцов крови 163 пациентов в возрасте 45–90 лет с сахарным диабетом 2-го типа и дислипидемиями. В крови определяли концентрации жирных кислот, альдегидов, стиролов, имеющих микробное происхождение, а также уровни глюкозы, триглицеридов, холестерина, липидов низкой и высокой плотности. Химический состав крови определяли методом газовой хромато-масс-спектрометрии. По концентрациям жирных кислот, альдегидов, стиролов рассчитывали суммарную молярную концентрацию малых молекул микробного происхождения (Small molecules originating from microbes, SMOM), концентрацию октадеценового альдегида (18a), суммарную концентрацию гидроксикислот, производных гидроксильного остатка липида А, бактериального эндотоксина (3OH-FA) и сгруппированные суммарные концентрации химических соединений микробного происхождения, определяющих представительство основных четырех филотипов микробиома человека: Actinobacteria, Bacteroidetes, Proteobacteria, Firmicutes.

В результате исследования были получены данные об увеличении суммарной концентрации химических соединений, концентрации октадеценового альдегида и концентрации 3OH-FA у пациентов при нарушении углеводного обмена по типу сахарного диабета 2-го типа. Снижение представленности Bacteroidetes отмечали при нарушении углеводного обмена и снижение представленности Proteobacteria и Firmicutes – при нарушении углеводного и липидного обменов, а также увеличение представленности Actinobacteria – при нарушении липидного обмена, в том числе сочетанного с нарушением углеводного обмена. У пациентов с сахарным диабетом 2-го типа и в контрольной группе отмечена обратная корреляция между представленностью Firmicutes и уровнем глюкозы в крови. В группе пациентов с сахарным диабетом 2-го типа отмечена прямая корреляция представленности Bacteroidetes и уровня триглицеридов в крови.

При нарушениях липидного обмена не выявлено ни статистически значимых изменений концентраций микробных маркеров в крови пациентов, ни статистически значимых корреляционных связей биохимических параметров крови и представленности филотипов микробиома.

Main article text

INTRODUCTION

The increase in the incidence of diabetes mellitus is contributing to the search for new opportunities in the prevention and treatment of this chronic systemic disease at the level of the macroorganism-microbiota relationship [1]. The development of systems biology technologies, such as genomics, proteomics, and metabolomics (OMICS technologies), to study the microbiome and metabolome of a macroorganism expands the possibilities of predictive diagnostics and the integral assessment of the state of the macroorganism-microbiome system based on the qualitative and quantitative composition of the microbiota and its impact on the metabolic processes in the human body [2, 3].

Human blood is saturated with microbial metabolites that are in dynamic equilibrium with intracellular metabolites produced by human cells [4]. Blood also contains chemical compounds introduced from the external environment with food, water, inhaled air, and substances produced by the microbiota, which are products of the lysis of microorganisms. All of these chemical compounds comprise the human metabolome that is divided into the metabolome of its own cells and exposome – something that is introduced from the external environment and affects the metabolic processes of the macroorganism. The study of the effect of the exposome on human metabolism using a systematic approach and methods of multivariate statistics has developed into one of the new, independent technologies - exposomics. The core of the exposome consists of the microorganisms’ metabolites and the lysis products of microorganisms [5]. The lipid composition of microorganisms is highly specific, which makes it possible to identify bacteria by their chemical composition [6]. Determination of the concentration of certain molecules in the blood makes it possible to study the microbiome-associated exposome: concentrations of the chemical compounds of microbial origin carry information about the changes in the ratio of microorganisms in the microbiome, permeability of the mucous membrane of the gastrointestinal tract (GIT), and intensity of the inflammatory processes and other processes of interaction between the microbiome and macroorganism [6].

The goal of this study was to carry out a comprehensive assessment of the state of the microbiome in elderly people with lipid and carbohydrate metabolism disorders using the criteria of microbiome-associated exposomics. To achieve this goal, we had to meet the following objectives: 1) determine the specificity of the interaction between the microbiome and the macroorganism by the integral indicators of endotoxemia as well as by the ratio of chemical compounds that characterize the activity of the main phylotypes of the human microbiome, and to assess the structure of the microbiome in patients with lipid and carbohydrate metabolism disorders; 2) to establish the relationship between the biochemical parameters of blood and the indicators of microbiome-associated exposomics in patients with lipid and carbohydrate metabolism disorders.

MATERIALS AND METHODS

Patients

The single blood samples taken from outpatients in the consulting and diagnostic center of the Russian Clinical and Research Center of Gerontology at the Pirogov Russian National Research Medical University were studied in this project. The study was conducted within the framework of the scientific program “Active longevity” of the Russian Clinical and Research Center of Gerontology. All of the patients signed an informed consent to participate in the study. In the course of this project, we examined 163 blood samples from patients with an average age of 68.1±1.7 years, of which 48 (29%) were females and 115 (71%) were males. Accounting for the differences in age, gender, and body weight in the experimental and control group did not lead to any statistically significant distribution of the study results. All of the patients were on sulfonylurea and biguanide drugs for the hypoglycemic therapy. The patients in a critical condition as well as those on insulin therapy were excluded from the study. The exclusion criteria also included the presence of a severe intestinal infection. Blood samples were divided into 4 groups depending on the type of metabolic disorder (Table 1).

Table 1.

Designation and size of the groups of patients

Group designation	Group size, number of people	Description of the patients’ disorders (Patients’ diagnoses)
DM2+DLP	67	Combination of dyslipidemia and type 2 diabetes mellitus. Patients with combined disorders of carbohydrate and lipid metabolism.
DLP	39	Dyslipidemia. Patients with lipid metabolism disorders.
DM2	39	Type 2 diabetes mellitus. Patients with carbohydrate metabolism disorders.
C	18	Control group. Patients without disorders of carbohydrate and lipid metabolism.

DM2 – type 2 diabetes mellitus, DLP – dyslipidemia, C – control

The diagnoses were made by an endocrinologist based on the results of a clinical and anamnestic examination and biochemical blood tests. Gender differences were not taken into account.

Biochemical analysis of the patients’ blood

The levels of the following biochemical parameters were determined in the patient’s blood samples: concentration of glucose, triglycerides (TG), total cholesterol (TC), the low-density lipoprotein cholesterol (LDL-C), and the high-density lipoprotein cholesterol (HDL-C). Venous blood collected from patients after fasting 12 h (not earlier than 12 h after the last meal) was centrifuged for 10 min at 3000 rpm at 15°C. The resulting blood plasma was an alyzed on an automatic biochemical analyzer “Advia 1800 Siemens Healthcare Diagnostics” (USA – Germany) according to the manufacturer’s instructions.

Determination of the concentration of small molecules of microbial origin in the blood

Methylsilyl derivatives of fatty acids, aldehydes, and sterols were separated by chromatography on 50 m long high-performance capillary columns with a methyl silicone stationary phase using a GC-MS instrument with electron ionization (70 eV).

The selected compounds had a carbon chain 10 to 26 atoms long and were divided into 5 different groups. Saturated aldehydes were analyzed in the homologous series: from tetradecanal (14a) to 11-octadecenaldehyde (18:1d11a); hydroxy acids from 2-hydroxylauric (2h12) to 2-hydroxyhexacosanoic (2h26); saturated fatty acids from decanoic (10:0) to eicosenoic (20:0) acid; unsaturated fatty acids from isomyristic (i14) to cis-vaccenic (18:1d11) acid; sterols: campesterol, β-sitosterol and ergosterol [7].

The combined pool of selected reference points was determined by summing up all the concentrations of the analyzed chemical compounds. The representation of Actinobacteria, Bacteroidetes, Proteobacteria, and Firmicutes phylotypes was determined by summing up the numerical values of the representation rates of all the microorganisms belonging to the corresponding phylotypes [8, 9].

The microbial marker of bacterial plasmalogen was determined by the concentration of octadecenaldehyde (18:1a), which is a product of the cell wall acid methanolysis of microorganisms that are representatives of the indigenous normobiota and belong to Bifidobacterium spp. and Eubacterium spp. genera [10].

The microbial marker of bacterial endotoxin in the blood samples was determined as the sum of the concentrations of hydroxy acids: 3h12, 2h12, hi13, 3h13, 3h14, 2h14, 2hi15, 3hi15, h16, 3hi17, h18, h15, which characterize the amount of lipopolysaccharides (LPS) in the cell wall of gram-negative bacteria: Acinetobacter, Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Bacteroides hypermegas, Fusobacterium, Haemophilus, Sphingomonas, Flavobacterium, Porphyromonas, Prevotella, Bacteroides fragilis, Helicobacter pylori [11].

Statistical methods

The variation statistics methods used in this project included the nonparametric Mann-Whitney U-test. Differences were considered statistically significant at p<0.05. Pearson’s goodness-of-fit test (χ²) was used to evaluate the adequacy of the obtained results to the theoretical hypothesis. Calculations were carried out using Microsoft Office Excel and Statistica 8.0 software packages.

RESULTS

An integral assessment of the disorders of carbohydrate (DM2) and lipid (DLP) metabolism as well as their combination (DM2 + DLP) in patients was conducted based on the general changes in the macroorganism-microbiota system. The determination of the concentration of octadecenaldehyde that is a product of the cell wall acid methanolysis of Bifidobacterium spp. and Eubacterium spp.- the main source of bacterial plasmalogen, of the total concentration of hydroxy acids - derivatives of the hydroxy residue of lipid A of bacterial endotoxin (3OH-fatty acids - 3OH-FA) as well as of the total concentration of small molecules of microbial origin (SMOM) in the blood were used for the integral assessment of the structure of the human microbiome. The comparison results are shown in Fig. 1.

Fig. 1.

Integral indicators of the state of the human microbiome in cases of carbohydrate and lipid metabolism disorders. The data are represented as the median with interquartile range M [Q1, Q3]. (**) means a statistically significant difference compared to the C group (p<0.01) according to the Mann-Whitney U-test.

As shown in Fig. 1, the statistically significant changes in the concentrations of octadecenaldehyde 18a and in the total concentration of 3OH-FA hydroxy acids as well as in the total SMOM concentration were characteristic only for the DM2 and DM2+DLP groups. Therefore, in cases with the carbohydrate metabolism disorders, the total concentration of SMOM increased 2 times while in the cases with a combination of disorders in carbohydrate and lipid metabolism, SMOM increased 1.5 times. In cases of carbohydrate metabolism disorder, the concentration of compound 18a increased 3 times; when carbohydrate and lipid metabolism disorders were combined, it increased 2 times. In cases of the carbohydrate metabolism disorders, the concentration of 3OH-FA increased 2 times, while in the cases with a combination of carbohydrate and lipid metabolism disorders-1.5 times.

The results of the biopsy and morphological studies showed that the duodenal mucosa has a characteristic inflammatory profile in patients with diabetes mellitus. The mucous membrane is abundantly infiltrated with macrophages and has all the signs of immune activation. The increased infiltration of the mucosa by bacteriophages may be the reason for the decrease in its barrier function [12]. Therefore, it is precisely in the cases of diabetes mellitus that a sharp increase in the concentration of SMOM in the blood was observed in our study. The combination of diabetes mellitus and dyslipidemia tends to reduce the concentration of SMOM in blood, which indicates a decrease in mucosal permeability due to oppositely directed processes occurring in carbohydrate and lipid metabolism disorders.

The representation of the main 4 phylotypes of the human microbiome was determined based on the ratio of the markers’ concentrations of microorganisms that belong to certain phylotypes. Assignments of the corresponding markers to the microorganisms of a certain phylotype were performed based on the previously published data [10, 11, 12] (Table 2).

Table 2.

Microbial markers of phylotypes of microorganisms

Designation	Chemical compound
Actinobacterium
i18	Isooctadecanoic acid
i17a	Isoheptadecanal
14a	Tetradecanal
i14a	Isomyristyl aldehyde
i15	Isopentadecanoic acid
10Me15	10-Methylpentadecanoic acid
i16	Isopalmitic acid
10Me16	10-Methylhexadecanoic acid
a17	Anteisoheptadecanic acid
10Me17	10-Methylheptadecanoic acid
16:1d9t	trans-9-hexadecenoic acid
10Me18	10-Methyloctadecanoic acid
i14	Isomyristic acid
14:1d11	11-Tetradecenoic acid
Bacteroides
3h13	3-Hydroxytridecanoic acid
2hi15	2-Hydroxyisopentadecanoic acid
3h16	3-Hydroxypalmitic acid
3hi17	3-Hydroxyisoheptadecanoic acid
3h15	3-Hydroxypentadecanoic acid
Furmikuts
i16a	Isopalmitaldehyde
18:1d11a	11-octadecenaldehyde
18:1d9a	9-octadecenaldehyde
i15a	Isopentadecanaldehyde
a15a	Anteisopentadecanaldehyde
10h18	10-Hydroxystearic acid
a13	Anteisotridecanoic acid
i12	Isolauric acid
a15	Anteisopentadecanoic acid
a19	Anteisononadecanoic acid
19cyc	Cyclononedecanoic acid
20:1d11	11-Eicosenoic acid
14:1d9	9-Tetradecenoic acid
15:1d9	9-Pentadecenoic acid
16:1d7	7-Hexadecenoic acid
18:1d11	Cis-vaccenic acid
16:1d11	11-Hexadecenoic acid
14:1d7	7-Tetradecenoic acid
Coprostanol	Coprostanol
Proteobacterium
3h12	3-Hydroxylauric acid
2h12	2-Hydroxylauric acid
3hi13	3-Hydroxyisotridecanoic acid
3h14	3-Hydroxymyristic acid
2h14	2-Hydroxymyristic acid
3hi15	3-Hydroxyisopentadecanoic acid
3h18	3-Hydroxystearic acid
3hi20	3-Hydroxyisoeicosanoic acid
17cyc	Cycloheptadecanoic acid
i17:1d9	9-Isoheptadecenoic acid
i16:1d9	9-isohexadecenoic acid

To level the influence of permeability of the intestinal wall, the activity of the immune system, etc., we used the relative concentrations (representation) of the components. The representation was calculated as the fraction of the presence of each component in the total concentration of all the components in the blood. The representation of phylotype was calculated as the sum of all relative concentrations of the cell wall components of bacteria that belong to the corresponding phylotype. The resulting phylotype distribution differs significantly from similar indicators obtained by sequencing or other methods since the pool of measured concentrations of cell wall components is limited by the number of unique compounds. The assessment of the informativity of this approach was also one of the objectives of this study. The results of representation of phylotypes in the human exposome are shown in Fig. 2.

Fig. 2.

The abundance of phylotypes in the microbiome exposome of the elderly patients with different combinations of carbohydrate and lipid metabolism disorders.The abundance of phylotypes in the graphs is represented as M [Q1, Q3]. (*) means a statistically significant difference compared to the C group according to the Mann-Whitney U-test (p<0.05), (**) - p<0.01.

Fig. 2 shows that, in the cases with carbohydrate metabolism disorders, the representation of Bacteroidetes decreases (Fig. 2B); in the cases of lipid metabolism disorders, the representation of Actinobacteria increases (Fig. 2A). With carbohydrate metabolism disorders, lipid metabolism disorders as well as with their combination, the representation of Firmicutes and Proteobacteria decreases (Fig. 2C, D).

The results of the correlation analysis of the concentrations of biochemical parameters in cases of carbohydrate and lipid metabolism disorders with the representation of the phylotypes of microorganisms is shown in Table 3.

Table 3.

The results of the correlation analysis of the concentrations of biochemical parameters in the cases with carbohydrate and lipid metabolism disorders with the representation of phylotypes of microorganisms

Factor	TC	TG	LDL-C	HDL-C	Glu-cose	TC	TG	LDL-C	HDL-C	Glu-cose
	Control					Dyslipidemia
Actinobacteria	0.21	-0.42	0.29	-0.02	0.55	0.13	0.15	0.10	0.01	-0.15
Bacteroidetes	0.09	-0.46	0.14	0.15	0.64	0.02	0.13	0.15	-0.23	0.26
Proteobacteria	-0.14	0.37	-0.15	-0.22	-0.62	0.02	0.07	-0.04	0.02	0.30
Firmicutes	0.04	0.23	0.03	-0.23	-0.91*	-0.19	0.06	-0.16	-0.22	0.35
	Type 2 diabetes mellitus					Combination of type 2 diabetes mellitus and dyslipidemia
Actinobacteria	-0.26	0.10	-0.28	-0.16	0.02	0.07	0.12	-0.04	0.22	0.18
Bacteroidetes	0.03	0.45*	-0.20	0.11	-0.29	0.08	-0.11	0.00	-0.10	-0.31
Proteobacteria	0.18	-0.13	0.29	0.04	0.05	-0.06	-0.08	0.06	-0.23	-0.06
Firmicutes	0.14	0.04	-0.24	-0.07	-0.39*	0.03	-0.04	0.07	-0.10	-0.20

*

means the statistical significance of the Pearson correlation coefficient, p<0.05.

According to the data shown in Table 3, there is a direct correlation between the level of triglycerides and the representation of Bacteroidetes in DM2. There is a strong inverse correlation between the glucose levels and representation of Firmicutes in the control group. A less pronounced inverse correlation is noticed between the same parameters in the DM2 group. This may be due to the heterogeneity of the Firmicutes phylotype, which includes microorganisms with different metabolic pathways that enable the rapid switching of microbial digestion from one substrate to another. In dyslipidemia, including its combination with diabetes mellitus, correlations were not found.

DISCUSSION

Using the method of microbiome-associated exposomics, we showed a reduction in the representation of all phylotypes except Actinobacteria in cases with carbohydrate metabolism disorders. In most studies of the microbiome in cases with diabetes mellitus that were performed using the sequencing of bacterial DNA isolated from blood samples, a decrease in the representation of Bacteroidetes and an increase in the representation of Firmicutes were noted [12, 13, 14, 15, 16, 17]. The results obtained by molecular genetic methods and microbiome-associated exposomics differ since these methods are based on the analysis of different types of samples - blood and stool samples. An assessment of the ratio of microorganisms in the stool samples indicates the number of microorganisms in the intestine. The ratio of the number of microorganisms, measured by the concentrations of SMOM in the blood, corresponds to the ratio of colonies of microorganisms lysed as a result of interaction with the immunocompetent cells of macroorganism. It is also important to note that the determination of the phylotype ratio by the sequencing of microbiota DNA isolated from the stool samples is a more accurate method of characterizing the microbiome than microbiome-associated exposomics. Many multidirectional biochemical processes have an impact on the results obtained by microbiome-associated exposomics. Therefore, these results indicate the ratio of SMOM from different phylotypes in the bloodstream rather than the ratio of phylotypes in the microbiome.

We also showed the presence of correlations between the ratio of microbiome phylotypes and the main biochemical characteristics of blood. These correlations hint at the most significant indicators characterizing carbohydrate metabolism disorders. The existence of a correlation cannot be considered as proof of the direct cause-effect relationship of these parameters, but it indicates the direct involvement of the microbiota in the pathogenesis of DM2.

Studies of the interaction between the microbiome and macroorganism by microbiome-associated exposomics enable the evaluation of the overall result that includes the carbohydrate metabolism disorders, the changes in the main phylotypes ratios of the human microbiome, the role of inflammatory reactions, the permeability of the mucous membrane of various sections of the intestine and periodontium, and other multidirectional processes.

Integral criteria - the total concentration of SMOM, the concentration of octadecenaldehyde 18a and 3OHFA – can be used to characterize bacterial endotoxemia in patients with DM2 and other disorders.

We did not reveal statistically significant changes in the concentrations of microbial markers in the blood of patients with lipid metabolism disorders as well as statistically significant correlations between the biochemical blood parameters and the representation of microbiome phylotypes. In the blood of patients with carbohydrate metabolism disorders, we found statistically significant excess in SMOM concentrations – 2–3 times versus the corresponding parameters in healthy individuals. The highest (threefold) excess was observed in the blood level of octadecenaldehyde 18a.

The ratios of the main phylotypes of the microbiome is the reflection of many processes determined by micro-biome-associated exposomics and should be considered as the result of the interaction between the microbiome and the macroorganism. We have revealed the suppression of Bacteroidetes in patients with diabetes mellitus which fully corroborates with the data that describe the effect of this disorder on the state of the human microbiome [18]. A direct correlation was also found between the representation of Bacteroidetes and the level of triglycerides in cases with a carbohydrate metabolism disorders. According to our data, the development of dyslipidemia reduced the intensity of endotoxemia (total level of SMOM) that develops in DM2. Therefore, the level of triglycerides can be considered as the most significant factor of dyslipidemia that prevents the development of diabetes mellitus.

Footnotes

Conflict of interest: Authors have no commercial or financial interests.

Funding: The study was conducted within the framework of the scientific topics of the Russian Clinical and Research Center of Gerontology (Pirogov Russian National Research Medical University) and G. N. Gabrichevsky Research Institute for Epidemiology and Microbiology.

REFERENCES

1. IbAwad SF, Al-Mawali A, Al-Lawati JA, Morsi M, Critchley JA, Abu-Raddad LJ. Forecasting the type 2 diabetes mellitus epidemic and the role of key risk factors in Oman up to 2050: Mathematical modeling analyses. J Diabetes Investig 2021; 12(7), 1162-74. doi: 10.1111/jdi.13452.

2. Nicholson JK, Lindon JC. Systems biology: Metabonomics. Nature 2008; 455(7216), 1054-6 doi: 10.1038/4551054a.

3. Nicholson JK, Lindon JC, Holmes E. ‘Metabonomics’: understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data. Xenobiotica 1999; 29(11), 1181-9. doi: 10.1080/004982599238047.

4. Kolokolova T, Savel’ev O, Sergeyev N. Metabolic analysis of human biological fluids by 1H NMR spectroscopy. J Anal Chem 2011; 63, 104-20. doi: 10.1134/S1061934808020020.

5. Moon Y. Microbiome-Linked Crosstalk in the Gastrointestinal Exposome towards Host Health and Disease. Pediatr Gastroenterol Hepatol Nutr 2016; 19(4), 221-8. doi: 10.5223/pghn.2016.19.4.221.

6. Vlasov AA, Salikova SP, Grinevich VB, Bystrova OV, Osipov GA, Meshkova ME. Gut microbiota and systemic inflammation in patients with chronic heart failure. Kardiologiia 2020; 60(5), 74-82. (In Russian). doi: 10.18087/cardio.2020.5.n859.

7. Wild CP. The exposome: from concept to utility. Int J Epidemiol 2012; 41(1), 24-32. doi: 10.1093/ije/dyr236.

8. Gavini F, Cayuela C, Antoine J-M, Lecoq C, Lefebvre B, Membré J-M, Neut C. Differences in the Distribution of Bifidobacterial and Enterobacterial Species in Human Faecal Microflora of Three Different (Children, Adults, Elderly) Age Groups. Microbial Ecology in Health and Disease 2001; 13(1), 40-5. doi: 10.1080/089106001750071690.

9. He F, Ouwehand AC, Isolauri E, Hosoda M, Benno Y, Salminen S. Differences in Composition and Mucosal Adhesion of Bifidobacteria Isolated from Healthy Adults and Healthy Seniors. Curr Microbiol 2014; 43(5), 351-4. doi: 10.1007/s002840010315.

10. Pisanov RV, Shipko ES, Duvanova OV, Simakova DI. Identification of microorganisms using gas chromato-mass-spectrometry. Zhurnal mikrobiologii, èpidemiologii i immunobiologii (Journal of microbiology, epidemiology and immunobiology) 2020; 97(4), 356-62. (In Russian). doi: 10.36233/0372-9311-2020-97-4-8.

11. Bezrodny SL, Mardanly SG, Zatevalov AM, Tereshina EV, Mironov AY, Pomazanov VV. Assessment of the state of intestinal microbiocenosis based on bacterial endotoxin and plasmalogen in elderly persons with type 2 diabetes mellitus pathology. Klinicheskaya Laboratornaya Diagnostika (Russian Clinical Laboratory Diagnostics) 2021; 66(9), 565-70. (In Russian). doi: 10.51620/0869-2084-2021-66-9-565-570.

12. Tkach SM, Dorofeeva AA. The ratio of the main types of intestinal microbiota in patients with type 2 diabetes mellitus. Klinichna Endocrinologia ta Endocrinna Chirurgia (Clinical Endocrinology and Endocrine Surgery) 2018; 3(62), 7-14. (In Russian). doi: 10.24026/1818-1384.3(63).2018.142668.

13. Sheveleva SA, Kuvaeva IB, Efimochkina NR, Markova YuM, Prosyannikov MYu. Gut microbiome: from the reference of the norm to pathology. Voprosy pitaniia (Problems of Nutrition) 2020; 89(4), 35-51. (In Russian). doi: 10.24411/0042-8833-2020-10040.

14. Gorvitovskaia A, Holmes SP, Huse SM. Interpreting Prevotella and Bacteroides as biomarkers of diet and lifestyle. Microbiome 2016; 4, 15. doi: 10.1186/s40168-016-0160-7.

15. Matos J, Matos I, Calha M, Santos P, Duarte I, Cardoso Y, Faleiro ML. Insights from Bacteroides Species in Children with Type 1 Diabetes. Microorganisms 2021; 9(7), 1436. doi: 10.3390/microorganisms9071436.

16. Liu X, Cheng YW, Shao L, Sun SH, Wu J, Song QH, et al. Gut microbiota dysbiosis in Chinese children with type 1 diabetes mellitus: An observational study. World J Gastroenterol 2021; 27(19), 2394-414. doi: 10.3748/wjg.v27.i19.2394.

17. La-Ongkham O, Nakphaichit M, Nakayama J, Keawsompong S, Nitisinprasert S. Age-related changes in the gut microbiota and the core gut microbiome of healthy Thai humans. 3 Biotech 2020; 10(6), 276. doi: 10.1007/s13205-020-02265-7.

18. Wang J, Li W, Wang C, Wang L, He T, Hu H, et al. Enterotype Bacteroides Is Associated with a High Risk in Patients with Diabetes: A Pilot Study. J Diabetes Res 2020; 2020, 6047145. doi: 10.1155/2020/6047145.