Sage Journals: Discover world-class research

Abstract

Background:

Gut microbiota could affect the onset and development of vascular cognitive impairment (VCI) through modulating metabolic and immune pathways. However, the vascular mechanisms involved remain unclear.

Objective:

To investigate the gut microbiota associated with VCI and examine the mediating effects of regional cerebral blood flow (CBF) to explore potential therapeutic targets for VCI.

Methods:

This prospective study enrolled patients with VCI (n = 16) and healthy controls (n = 18) from the Chinese Imaging, Biomarkers, and Lifestyle study between January 1 and June 30, 2022. The gut microbiota composition and diversity were determined by 16 S ribosomal RNA gene sequencing. The association between gut microbiota and Montreal Cognitive Assessment (MoCA) scores was determined using Spearman’s correlation analysis. Regional CBF was calculated using pseudo-continuous arterial spin labeling. The mediating effects of regional CBF on the relationship between specific gut microbiota and cognition in VCI were investigated using mediation analysis.

Results:

Compared to healthy controls, patients with VCI had significantly greater abundance of Bifidobacterium, Veillonella, R uminococcus gnavus , Fusobacterium, and Erysipelatoclostridium and smaller abundance of Collinsella. The abundance of Ruminococcus gnavus was negatively associated with MoCA scores in patients with VCI, with the CBF in the left hypothalamus, right hypothalamus, and left amygdala accounting for 63.96%, 48.22%, and 36.51%, respectively, of this association after adjusting for confounders.

Conclusions:

Ruminococcus gnavus is associated with cognition in VCI, which is strongly mediated by CBF in the bilateral hypothalamus and left amygdala. These findings highlight the potential regulatory roles of nutrition and metabolism-related areas of the brain in VCI.

Keywords

Alzheimer’s disease cerebral blood flow cerebral small vessel diseases gut microbiota mediation analysis

INTRODUCTION

Vascular cognitive impairment (VCI) refers to a group of clinical syndromes in which at least one cognitive domain is damaged by clinical stroke or subclinical vascular injury [1]. With the changing lifestyles of the global population, vascular risk factors are becoming more prevalent in the rapidly growing aging population, and the incidence and disability rates of VCI are increasing [2]. VCI is the second most common type of dementia, accounting for approximately 15% –30% of cases [3]. It negatively affects quality of life and increases the burden on society and families [4, 5]. Despite its increasing prevalence worldwide, the diagnosis and management of VCI remain inconclusive [6]. Moreover, there is limited understanding of the underlying pathophysiology of VCI, although factors such as inflammation and endothelial dysfunction have been found to play a role [7]. Therefore, it is crucial to identify early diagnostic biomarkers and explore the potential mechanisms involved in the development of VCI to slow its progression.

The microbiota–gut–brain axis has recently received attention as an essential pathway in the development of VCI, suggesting that gut microbiota (GM) can affect brain cognition through neural, immune, endocrine, and metabolic pathways [8, 9]. Recently, a study reported that dysbiosis of several GM, including Enterobacteriaceae and Proteobacteria and their metabolites, could be linked to VCI and useful for distinguishing patients with VCI from normal controls [10]. However, the potential mechanisms behind the effects of GM on cognition, neuropsychiatric symptoms, and the development of VCI remain to be elucidated.

A recent review suggested that VCI was characterized by cerebral hemodynamic alterations ranging from decreased cerebral blood flow (CBF) to small vessel disease and involved a multifactorial process that led to demyelination and gliosis, including blood-brain barrier disruptions, hypoxia, hypoperfusion, oxidative stress, neuroinflammation, alterations in the neurovascular unit coupling, cerebral microbleeds, or superficial siderosis [11]. A retrospective study of 108 patients with ischemic cerebrovascular disease demonstrated that non-invasive 3D arterial spin labeling (ASL) imaging could be used to detect cerebral hemodynamics and predict prognosis in patients with VCI [12]. Another observational study using 3D pseudo-continuous ASL (pCASL) found that deficits in brain perfusion in the hippocampus, thalamus, and insula were related to the degree of cognitive impairment in patients with VCI [13]. However, evidence from recent research is limited, and findings have been inconsistent. Moreover, it remains unclear whether changes in GM affect cognition in VCI through the regulation of specific brain areas. Thus, this study aimed to investigate the relationship between specific GM and cognition in patients with VCI and to examine whether this association is mediated by regional CBF to explore the potential mechanisms of and intervention targets for VCI.

METHODS

Dataset and study design

This study enrolled consecutive patients with VCI and healthy controls (HCs) from the Chinese Imaging, Biomarkers, and Lifestyle (CIBL) Study between January 1 and June 30, 2022. The CIBL study has been described previously [14], which was approved by the Institutional Review Board of the Beijing Tiantan Hospital of Capital Medical University (KY-2021-028-01) and registered at Chictr.org.cn (ChiCTR2100049131). All participants or their relatives provided informed consent before enrollment.

The inclusion criteria for patients with VCI were as follows: 1) patients were 60–85 years old; 2) patients met the 2018 standardized diagnosis of VCI guidelines [15]; 3) objective cognitive impairment was identified by Montreal Cognitive Assessment (MoCA) scores: illiteracy≤13 points; 1–6 years of education≤19 points; at least 7 years of education≤24 points; 4) patients had evidence of vascular brain injury (moderate to severe vascular brain injury on MRI defined as WMH Fazekas score > 1 and/or lacunar infarcts and/or intracerebral microhemorrhages, or mild vascular brain injury defined as WMH Fazekas score 1); 5) vascular injury was clearly a dominant factor affecting patients’ cognitive impairment. Excluded patients 1) had immediate and/or delayed cognitive decline that began within 6 months after a stroke and that does not reverse; 2) stroke within the previous 6 months and displayed moderate or severe stenosis in large vessels by MR angiography; 3) cognitive impairment caused by other neurodegenerative diseases including Alzheimer’s disease (AD), frontotemporal dementia, dementia with Lewy bodies, etc., intracranial tumors, or traumatic brain injury; 4) had history of a mental disorder in accordance with the Diagnostic and Statistical Manual of Mental Disorders-5 [16]; 5) taken antibiotics or probiotics, or other drugs that may affect gut microbiota, within one month before examination; and 6) an inability to cooperate to complete the data collection.

Healthy older participants from the CIBL cohort were recruited from the community and served as the HC group. These healthy individuals had no history of major medical, neurological, or psychiatric illnesses, and no apparent abnormal findings or cognitive impairments were identified by MRI or neuropsychological testing, and did not take antibiotics or probiotics, or other drugs that may affect gut microbiota, within one month before examination. A flowchart of the participant selection process is shown in Fig. 1.

Fig. 1

A flowchart of the inclusion and exclusion processes. VCI, vascular cognitive impairment; HC, healthy control; AD, Alzheimer’s dementia; FTD, frontotemporal dementia; DLB, dementia with Lewy bodies; pCASL, pseudo-continuous arterial spin labeling; RNA, ribose nucleic acid; DNA, deoxyribonucleic acid.

Data collection and neuropsychological assessments

Information on the patient’s age, sex, years of education, body mass index (BMI), diagnoses, and medical history, including hypertension, diabetes, coronary heart disease, smoking, drinking, and family history of dementia, were collected. Global cognitive function was evaluated using the MoCA, the severity of depression and anxiety were assessed using the Hamilton Depression Rating Scale (HAMD) and Hamilton Anxiety Rating Scale (HAMA), and the nutritional status and dietary habits were assessed using Mini Nutritional Assessment (MNA) and Dietary Diversity Score (DDS), respectively.

Fecal sample collection and 16 S ribosomal RNA (rRNA) sequencing

Fecal samples were collected on the morning of admission. Fresh fecal samples were immediately provided to doctors or nurses and collected in sterile tubes. These samples were stored at –80°C prior to microbial analysis.

16S rRNA gene sequencing was performed using an Illumina MiSeq PE300 platform (Shanghai Majorbio Bio-Pharm Technology Co., Ltd., Shanghai, China). Bacterial genomic DNA was extracted using an E.Z.N.A. Soil DNA Kit (Omega BioTek, Norcross, GA, USA). The V3–V4 hypervariable regions of the bacterial 16S rRNA genes were amplified using the forward primer 338F (5^′-ACTCCTACGGGAGGCAGCAG-3^′) and the reverse primer 806R (5^′-GGACTACVVGGGTATCTAATC-3^′) in a thermocycler polymerase chain reaction (PCR) system. All PCR amplicons were mixed and sequenced using an ABI GeneAmp 9700 PCR thermocycler (Applied Biosystems, Waltham, MA, USA). To study the constituent diversity of the species in the samples, the effective sequences of all samples were clustered into operational taxonomic units with 97% similarity. Shannon’s diversity index, the Chao1 richness estimate, the abundance-based coverage estimator index, and Simpson’s Diversity Index were used to assess diversity. Partial least squares discriminant analysis and principal coordinate analysis based on the weighted UniFrac algorithm were used for data analysis.

MRI acquisition and processing

Subjects were instructed not to use alcohol, caffeine, and nicotine for at least six hours prior to MRI measurement. Image acquisition was performed on a 3-T magnetic resonance imaging scanner (SIGNA Premier; GE Healthcare, Chicago, IL, USA) with a 48-channel head coil. High-resolution 3D T1-weighted scans were performed using the magnetization-prepared rapid gradient echo sequence: repetition time = 7.3 ms, echo time = 3.0 ms; flip angle = 12°, field of view = 256 × 256 mm², acquisition matrix = 256 × 256, slice thickness = 1.0 mm, slice number = 176, and scan time = 4 min 56 s. ASL was performed using a 3D pCASL sequence: axial acquisition, repetition time = 4849 ms, echo time = 10.6 ms, field of view = 220×220 mm, acquisition matrix = 512×512, slice thickness = 4 mm, slice number = 36, and post-labeling delay = 2025 ms. Brain MRI examination, including T2-weighted images, fluid-attenuated inversion recovery, pCASL, diffusion weighted imaging, and susceptibility-weighted imaging were performed for all the participants in this study. The imaging results were independently interpreted by a neurologist and a radiologist to achieve consistency, if their interpretations were inconsistent, they will subsequently be determined by a higher-level expert.

PCASL Data processing was performed using CereFlow software (Anying Technology [Beijing] Co., Ltd., China), with the following steps: 1) CBF was calculated from the ASL proton density images and perfusion-weighted images using the standard simple compartment model with the assumption that the arterial transit time was equivalent to the post-labeling delay; 2) the M0 image (obtained with pCASL acquisition within the same image space as the CBF) was matched with the anatomical T1-weighted image; 3) the T1-weighted images were normalized to the Montreal Neurological Institute template; 4) the forward transformation matrix derived from T1-weighted images was used to warp the CBF image into the Montreal Neurological Institute space; 5) the regional CBF was extracted using the automated anatomical labeling atlas. Six brain areas involved with eating, metabolism, and inflammation were analyzed, including the bilateral hypothalamus, amygdala, and insula [17 –19].

Statistical analysis

All statistical analyses were performed using SPSS statistical software (v26.0; IBM Corp., Armonk, NY, USA). Categorical variables are presented as the total number (n) and percentage (%) per group, and the chi-squared or Fisher’s exact test was used to assess statistical differences. The mean and standard deviation were calculated for continuous variables with normal distributions, whereas the median and interquartile range were used for continuous variables with non-normal distributions. Similarly, the independent t-tests were used for normally distributed data, whereas the Mann–Whitney U test was used for non-normally distributed data. The Wilcoxon Rank Sum test was used at the genus level to determine changes in the abundance of GM in both groups. The false discovery rate approach was used to adjust for multiple tests. Spearman’s correlation analysis was performed to examine the association between GM abundance and MoCA scores in patients with VCI.

Mediation models were developed using Model 4 in PROCESS macro (v3.5) with a 95% confidence interval and 5,000 bootstrap samples to analyze the effects of regional CBF on the association between the abundance of Ruminococcus gnavus and MoCA scores [20]. The total effect (TE) was decomposed into two components: the natural direct effect (NDE) and the natural indirect effect (NIE). The NDE represented the effect of Ruminococcus gnavus on MoCA scores that were independent of CBF. The NIE represented the effects of Ruminococcus gnavus on MoCA scores that could be explained by changes in regional CBF. The mediating effect was considered significant if the 95% confidence interval (CI) was not zero. The percentage of the mediated effect was calculated using the following formula: (NIE/TE)×100. Under this framework, three multivariate linear regression models were fitted: one model for MoCA scores (outcome) on Ruminococcus gnavus (exposure) and all study confounders, one model for regional CBF (mediator) conditional on the abundance of Ruminococcus gnavus (exposure) and all study confounders, and one model for MoCA scores (outcome) conditional on the abundance of Ruminococcus gnavus, regional CBF, and all study confounders. Confounders included age and the CBF of the whole brain, which was found in the univariate analysis. Iterative reweighted least squares algorithm was used to evaluate sensitivity and robustness of the model confounders. The threshold for statistical significance was set at a two-tailed p < 0.05.

RESULTS

Baseline characteristics of clinical and regional CBF in both groups

The demographic and neuropsychological characteristics of the 16 patients with VCI and the 18 HCs are shown in Table 1. Patients with VCI were older (p = 0.026) and had lower MoCA scores (p < 0.001) than HCs. There were no significant differences between the two groups in terms of sex, years of education, BMI, hypertension, diabetes, coronary heart disease, family history of dementia, HAMD scores, HAMA scores, MNA scores, or DDS scores. (all p > 0.05).

The regional CBF values of the six brain areas, namely the bilateral hypothalamus, amygdala, and insula, and the whole brain were significantly lower in the VCI group than in the HC group (all p < 0.001; Table 1).

Table 1

Baseline characteristics of clinical and regional CBF in both groups

	HC (n = 18)	VCI (n = 16)	t/χ²/z	p
Age (y)	65.11±5.13	69.75±6.44	–2.33	0.026
Sex (% Male)	5 (27.78)	7 (43.75)	0.95	0.331
Education (y)	13.22±4.70	11.75±3.57	1.61	0.254
BMI (kg/m²)	23.59±3.08	24.41±3.12	–0.77	0.450
Hypertension (%)	7 (38.89)	10 (62.50)	1.89	0.169
Diabetes (%)	3 (16.67)	5 (31.25)	—	0.317^a
Coronary heart disease (%)	5 (27.78)	5 (31.25)	0.05	0.824
Smoking (%)	3 (16.67)	6 (37.50)	—	0.169 ^a
Drinking (%)	6 (33.33)	7 (43.75)	0.39	0.533
Family history of dementia (%)	6 (33.33)	3 (18.75)	—	0.336 ^a
MoCA (scores)	15.50 (12.00, 20.00)	25.00 (24.00, 28.00)	–4.99	<0.001
HAMD (scores)	7.00 (3.00, 10.00)	7.50 (4.75, 11.25)	–0.67	0.502
HAMA (scores)	8.00 (3.00, 11.00)	4.50 (2.00, 9.00)	–0.78	0.435
MNA (scores)	26.00 (24.4,28.3)	24.75 (20.3, 27.0)	–1.81	0.089
DDS (scores)	9.00 (8.00,9.00)	9.00 (9.00, 9.00)	–1.27	0.205
CBF of left hypothalamus (mL/100 g/min)	48.34±6.80	32.77±9.07	5.70	<0.001
CBF of right hypothalamus (mL/100 g/min)	47.92±6.41	32.32±7.76	6.42	<0.001
CBF of left amygdala (mL/100 g/min)	43.29±6.52	28.71±8.82	5.52	<0.001
CBF of right amygdala (mL/100 g/min)	40.99±5.30	28.10±6.85	6.18	<0.001
CBF of left insula (mL/100 g/min)	48.80±7.68	32.65±8.04	5.89	<0.001
CBF of right insula (mL/100 g/min)	51.92±8.85	35.93±6.67	5.82	<0.001
CBF of whole brain (mL/100 g/min)	44.47±5.75	41.85±4.70	1.44	0.040

Data were evaluated using the χ² or Mann–Whitney U test. ^ap values were evaluated using Fisher exact tests. HC, healthy control; VCI, vascular cognitive impairment; BMI, body mass index; MoCA, Montreal cognitive assessment; HAMD, Hamilton depression rating scale; HAMA, Hamilton anxiety rating scale; CBF, cerebral blood flow; MNA, Mini Nutritional Assessment; DDS, dietary diversity scores.

GM diversity and composition

The β-diversity of GM was higher in patients with VCI than in HCs in the partial least squares discriminant analysis model (Fig. 2B), whereas there was no significant difference between the two groups in the principal coordinate analysis model (Fig. 2A). There was no significant difference in α-diversity between the two groups (all p > 0.05, Fig. 2C–F). At the genus level, the abundance of Bifidobacterium (p = 0.023), Veillonella (p = 0.022), Ruminococcus gnavus (p = 0.002), Fusobacterium (p = 0.026), and Erysipelatoclostridium (p = 0.021) was greater in patients with VCI than in HCs, whereas the abundance of Collinsella (p = 0.002) was smaller in patients with VCI than in HCs (Fig. 2G, H). There was no significant difference in the abundance of GM at the phylum level between the two groups (all p > 0.05).

Fig. 2

The diversity and composition of GM in both groups. The β-diversity of GM on the PCoA model (A) and the PLS-DA model (B). No difference in α-diversity between two groups, identified by the Ace index (C), Chao1 index (D), Shannon index (E), and Simpson index (F). The composition of GM in two groups (G). Genus-level gut microbial taxa comparisons showed there were significant difference in the relative abundance of Bifidobacterium, Veillonella, Ruminococcus gnavus, Fusobacterium, Erysipelatoclostridium, and Collinsella between two groups by using Wilcoxon rank sum test (H). VCI, vascular cognitive impairment; HC, healthy control; GM, gut microbiota; PCoA, principal coordinate analysis; PLS-DA, Partial least squares discriminant analysis; Ace, Abundance-based coverage estimator.

Spearman’s correlation analysis between the above-mentioned six genera of GM and MoCA scores demonstrated that the abundance of Ruminococcus gnavus was significantly negatively associated with MoCA scores in patients with VCI (r = –0.793, p = 0.009) and was not related to MoCA scores in HCs (p > 0.05; Fig. 3A). There was no significant association between the abundance of the other five GM and MoCA scores (all p > 0.05, Fig. 3B–F).

Fig. 3

Spearman correlation analysis between the abundance of six genera in the GM and MoCA scores in VCI and HC. Six genera include Ruminococcus gnavus (A), Veillonella (B), Fusobacterium (C), Erysipelatoclostridium (D), Bifidobacterium (E), Collinsella (F).VCI, vascular cognitive impairment; HC, healthy control; MoCA, Montreal Cognitive Assessment; r, Correlation Coefficient; ***p < 0.001.

Mediating effects of regional CBF on the association between GM and cognition

Table 2 shows the mediation effect of the regional CBF on the association of the abundance of Ruminococcus gnavus with MoCA scores. The abundance of Ruminococcus gnavus was negatively associated with MoCA scores (β= –3.405, p = 0.002). The regional CBF of the left hypothalamus (β= 380, p = 0.002), right hypothalamus (β= 0.418, p = 0.004), and left amygdala (β= 0.309, p = 0.006) were significantly positively associated with MoCA scores in VCI, whereas there was no significant association between the regional CBF of the right amygdala or bilateral insula and MoCA scores after adjusting for age and the CBF of the whole brain (all p > 0.05). The proportion of the association between the abundance of Ruminococcus gnavus and MoCA scores mediated by the CBF of the left hypothalamus was 63.96%, the right hypothalamus was 48.22%, and the left amygdala was 36.51% in patients with VCI (Fig. 4).

Table 2

Mediation effect of the regional CBF on the association of the abundance of Ruminococcus gnavus with MoCA scores

Mediators	Pathway	Effect	SE	95% CI	Mediating
Total effect	Ruminococcus gnavus → MoCA	–3.405	0.853	–5.263∼ –1.546
Hypothalamus_L	Ruminococcus gnavus → Hypothalamus_L → MoCA				63.96%
Direct effect		–1.227	0.778	–2.940∼ 0.485
Indirect effect		–2.178	1.206	–5.136∼ –0.065
Hypothalamus_R	Ruminococcus gnavus →Hypothalamus_R → MoCA				48.22%
Direct effect		–1.710	0.753	–3.367∼ –0.053
Indirect effect		–1.642	1.710	–6.555∼ –0.603
Amygdala_L	Ruminococcus gnavus → Amygdala_L → MoCA				36.51%
Direct effect		–2.162	0.728	–3.764∼ –0.559
Indirect effect		–1.243	1.635	–6.401∼ –0.313
Amygdala_R	Ruminococcus gnavus → Amygdala_R →MoCA				—
Direct effect		–2.327	0.831	–4.156∼ –0.499
Indirect effect		–1.025	1.274	–3.395∼ 0.119
Insula_L	Ruminococcus gnavus→Insula_L→MoCA				—
Direct effect		–3.163	0.943	–5.265∼ –1.060
Indirect effect		–0.205	1.151	–3.105∼ 1.432
Insula_R	Ruminococcus gnavus→ Insula_R →MoCA				—
Direct effect		–3.352	0.934	–5.589∼ –1.476
Indirect effect		0.181	0.730	–1.903∼ 1.803

MoCA, Montreal cognitive assessment; SE, standard error; CI, confidence interval.

Fig. 4

Mediation analyses of regional CBF on the association between the abundance of Ruminococcus gnavus with MoCA scores in patients with VCI after adjusting confounders. Confounders include age, CBF of whole brain. CBF, cerebral blood flow; MoCA, Montreal Cognitive Assessment; PM, percentage mediated.

In the sensitive analysis, the correlation between the CBF of the left hypothalamus and MoCA scores (coefficient = –0.380) and between the abundance of Ruminococcus gnavus and CBF of the left hypothalamus (coefficient = –5.731) remain unchanged with different parameter values. The correlation between the CBF of the right hypothalamus and MoCA scores (coefficient = –0.426 to –0.411) and between the abundance of Ruminococcus gnavus and CBF of the left hypothalamus (coefficient = –5.731) were not significantly different within the full sample analysis results. The correlation between the CBF of the left amygdala and MoCA scores (coefficient = –0.399 to –0.410) and between the abundance of Ruminococcus gnavus and CBF of the left hypothalamus (coefficient = –4.025) also were stable. These findings suggested that the model was robust and stable.

DISCUSSION

This study found that the CBF of the bilateral hypothalamus and the left amygdala were significant mediators in the relationship between the abundance of Ruminococcus gnavus and MoCA scores in patients with VCI, which revealed potential metabolic mechanisms and therapeutic targets for this disease. First, we found that the abundance of Ruminococcus gnavus in the GM was negatively correlated with cognitive function in patients with VCI. Previous evidence demonstrated that Ruminococcus gnavus, a prevalent member of the human GM, was related to autism, depression, and other neurological disorders [21, 22]. Despite the lack of direct evidence connecting Ruminococcus gnavus and VCI, some studies have described changes in this bacterium in AD-related dementia. For instance, an observational study showed that a higher abundance of Ruminococcus gnavus was related to higher plasma levels of neurofilament light chain, regarded as a biomarker for cognitive impairment in patients with AD-related dementia [23]. Park et al. found that the abundance of Ruminococcus gnavus was higher in transgenic APP/PS1 model mice than in wild-type mice [24]. An animal study showed that metabolites of Ruminococcus gnavus, such as isoamylamine, accelerated the expression of genes related to aging and promoted the apoptosis of microglial cells, resulting in a decline in cognitive function [25]. Additionally, it has been established that Ruminococcus gnavus influences metabolic and neuroinflammatory processes [26, 27]. Thus, we speculated that the increased abundance of Ruminococcus gnavus could lower MoCA scores in patients with VCI via the regulation of metabolic and neuroinflammatory pathways.

We also found that the association between the increased abundance of Ruminococcus gnavus and decreased MoCA scores were strongly mediated by the decreased CBF in the bilateral hypothalamus and left amygdala in patients with VCI. As part of the hypothalamic–pituitary–adrenal axis (HPA), the hypothalamus plays important roles in neuroendocrine, cognitive, and metabolic functions [28]. HPA axis dysfunction can elevate proinflammatory cytokines and chemokines, reduce synaptic plasticity, and persistently activate microglia, which can further cause amyloid beta plaque accumulation and Tau hyperphosphorylation, typical pathological markers of AD [29]. A previous study showed that the increased abundance of R uminococcus gnavus is related to intestinal and central nervous system inflammatory response activation and oxidative stress, which could decrease CBF in regions regulating eating, such as the hypothalamus [30]. Moreover, and importantly, dysfunction of the HPA axis also leads to GM dysbiosis, forming a vicious cycle of GM dysbiosis, disrupted HPA axis regulation, and cognitive impairment [31]. In addition, reduced amygdala CBF was a mediator of the association between Ruminococcus gnavus dysbiosis and cognitive impairment in VCI. Similar to the present findings, previous studies showed that the basolateral amygdala is directly related to short and long-term memory associated with significant experiences [32, 33], and learning and episodic memory are regulated by the central amygdala and its output to the hypothalamus [34]. Previous evidence demonstrated that reduced amygdala activity could lead to memory deficits via metabolic dysregulation and, particularly, insulin resistance [35]. Insulin resistance can cause deficits in neuroplasticity, pathological amyloid-beta deposition, increased tau protein phosphorylation, oxidative stress and neuroinflammation, leading to cognitive decline [36]. In an observational study, a greater abundance of Ruminococcus gnavus was found in people with insulin resistance, which may be associated with chronic inflammation and damage to the intestinal and blood–brain barriers [37]. Therefore, we concluded that the cognitive decline associated with a higher abundance of Ruminococcus gnavus and amygdala hypoperfusion could be related to both insulin resistance and other metabolic syndromes.

Considering the role of GM and metabolism-related brain regions on cognition, dietary nutrition could provide a novel therapeutic route for the improvement of cognition in patients with VCI. A randomized controlled trial identified that the Mediterranean diet could generate consistent metabolome changes, including lower plasma and urinary carnitine levels and decreased levels of potentially proinflammatory Ruminococcus gnavus. In addition, previous studies showed that the Mediterranean diet could significantly increase CBF in the left inferior frontal cortex, parahippocampal gyri, posterior cingulate, precuneus, and hippocampus regions, thereby improving cognitive function [39, 40].

The present study had some limitations. First, this was a single-center study with a relatively small number of patients; therefore, our findings may not be generalizable to other settings due to inherent selection bias. Second, we did not perform further metabolism-related laboratory indicators and other multiomics research to validate the relationship among regional CBF, cognitive function, and GM.

Conclusions

This study demonstrated that the abundance of Ruminococcus gnavus was significantly negatively associated with cognitive function in patients with VCI, and this association was strongly mediated by CBF in the bilateral hypothalamus and left amygdala. These findings suggested that Ruminococcus gnavus could be considered an important biomarker for early screening and intervention in VCI. Moreover, increased CBF of the three metabolism-related brain regions more directly reflected the activation pathway of the microbiota–gut–brain axis and highlighted that dietary nutrition interventions targeting Ruminococcus gnavus and the bilateral hypothalamus and left amygdala could provide an opportunity for improving cognition in patients with VCI. Prospective longitudinal studies are needed to evaluate the diagnostic value of Ruminococcus gnavus in the early detection of VCI in addition to translational studies to elucidate the potential metabolic mechanisms of the association among Ruminococcus gnavus, regional CBF, and cognition in patients with VCI.

Footnotes

ACKNOWLEDGMENTS

We would like to thank AnImage (Beijing) Technology Co., Ltd for pCASL consultation.

FUNDING

This work was supported by the National Key Research and Development Program of China (2021YFC2500100 and 2021YFC2500103) and the National Natural Science Foundation of China (82071187 and 81870821).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author.

References

Rundek

, Tolea

, Ariko

, Fagerli

, Camargo

(2022) Vascular cognitive impairment (VCI). Neurotherapeutics 19, 68–88.

Malik

, Georgakis

, Neitzel

, Rannikmae

, Ewers

, Seshadri

, Sudlow

CLM

, Dichgans

(2021) Midlife vascular risk factors and risk of incident dementia: Longitudinal cohort and Mendelian randomization analyses in the UK Biobank. Alzheimers Dement 17, 1422–1431.

Wolters

, Ikram

(2019) Epidemiology of vascular dementia. Arterioscler Thromb Vasc Biol 39, 1542–1549.

Moon

, Han

, Bae

, Suh

, Kim

, Kwak

, Kim

, Moon

, Park

, Ryu

, Youn

, Lee

, Jhoo

, Kim

(2021) Disease burdens of Alzheimer’s disease, vascular dementia, and mild cognitive impairment. J Am Med Dir Assoc 22,2093–2099.e2093.

Jhang

, Wang

, Chang

, Wu

(2021) Characteristics predicting a high caregiver burden in patients with vascular cognitive impairment: Using the apriori algorithm to delineate the caring scenario. Risk Manag Healthc Policy 14, 1335–1351.

Biesbroek

, Biessels

(2023) Diagnosing vascular cognitive impairment: Current challenges and future perspectives. Int J Stroke 18, 36–43.

Oveisgharan

, Dawe

, Yu

, Kapasi

, Arfanakis

, Hachinski

, Schneider

, Bennett

(2022) Frequency and underlying pathology of pure vascular cognitive impairment. JAMA Neurol 79, 1277–1286.

Kaur

, LaForce

, Mallela

, Saha

, Buffa

, Li

, Sangwan

, Rothenberg

, Zhu

(2023) Exploratory transcriptomic profiling reveals the role of gut microbiota in vascular dementia. Int J Mol Sci 24, 8091.

, Shao

, Li

, HuangFu

, Wang

, Liu

, Cai

, Zhao

(2018) Vascular cognitive impairment and the gut microbiota. J Alzheimers Dis 63, 1209–1222.

10.

Koszewicz

, Jaroch

, Brzecka

, Ejma

, Budrewicz

, Mikhaleva

, Muresanu

, Schield

, Somasundaram

, Kirkland

, Avila-Rodriguez

, Aliev

(2021) Dysbiosis is one of the risk factor for stroke and cognitive impairment and potential target for treatment. Pharmacol Res 164, 105277.

11.

Iadecola

, Duering

, Hachinski

, Joutel

, Pendlebury

, Schneider

, Dichgans

(2019) Vascular cognitive impairment and dementia: JACC Scientific Expert Panel. J Am Coll Cardiol 73, 3326–3344.

12.

Zhou

, He

, Wang

(2022) Application of 3D-ASL in hemodynamic analysis and prognosis evaluation of vascular cognitive impairment. Am J Transl Res 14, 7960–7968.

13.

Sun

, Cao

, Ding

, Wang

, Han

, Zhou

, Xu

, Zhang

, Xu

(2016) Cerebral blood flow alterations as assessed by 3D ASL in cognitive impairment in patients with subcortical vascular cognitive impairment: A marker for disease severity. Front Aging Neurosci 8, 211.

14.

Jiang

, Wang

, Liu

, Yao

, Sun

, Jiang

, Li

, Jiang

, Zhang

, Wang

, Zhang

, Jia

, Zou

, Xu

(2023) Spatiotemporal characteristics of regional brain perfusion associated with neuropsychiatric symptoms in patients with Alzheimer’s disease. J Alzheimers Dis 95, 981–993.

15.

Skrobot

, Black

, Chen

, DeCarli

, Erkinjuntti

, Ford

, Kalaria

, O’Brien

, Pantoni

, Pasquier

, Roman

, Wallin

, Sachdev

, Skoog

VICCCS group;

Ben-Shlomo

Passmore

, Love

, Kehoe

(2018) Progress toward standardized diagnosis of vascular cognitive impairment: Guidelines from the Vascular Impairment of Cognition Classification Consensus Study. Alzheimers Dement 14, 280–292.

16.

American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders (5th ed.). American Psychiatric Association, Washington, DC.

17.

Liu

, Xu

, Yi

, Tong

, Cai

(2022) The hypothalamus for whole-body physiology: From metabolism to aging. Protein Cell 13, 394–421.

18.

Ghaemi Kerahrodi

, Michal

(2020) The fear-defense system, emotions, and oxidative stress. Redox Biol 37, 101588.

19.

Koren

, Yifa

, Amer

, Krot

, Boshnak

, Ben-Shaanan

, Azulay-Debby

, Zalayat

, Avishai

, Hajjo

, Schiller

, Haykin

, Korin

, Farfara

, Hakim

, Kobiler

, Rosenblum

, Rolls

(2021) Insular cortex neurons encode and retrieve specific immune responses. Cell 184, 5902–5915.e5917.

20.

VanderWeele

(2016) Mediation analysis: A practitioner’s guide. Annu Rev Public Health 37, 17–32.

21.

Han

, Zhao

, Zhou

, Yang

, Xiong

, Lu

, Deng

, Luo

, Chen

, Qiu

, Shang

, Huang

, Mo

, Huang

, Liu

, Yang

, Zhai

, Ning

, Lin

, Huang

, Cheng

, Zhong

LLD

, Li

, Bian

, Fang

(2022) Altered metabolome and microbiome features provide clues in understanding irritable bowel syndrome and depression comorbidity. ISME J 16, 983–996.

22.

Coletto

, Latousakis

, Pontifex

, Crost

, Vaux

, Perez Santamarina

, Goldson

, Brion

, Hajihosseini

, Vauzour

, Savva

, Juge

(2022) The role of the mucin-glycan foraging Ruminococcus gnavus in the communication between the gut and the brain. Gut Microbes 14, 2073784.

23.

Marizzoni

, Mirabelli

, Mombelli

, Coppola

, Festari

, Lopizzo

, Luongo

, Mazzelli

, Naviglio

, Blouin

, Abramowicz

, Salvatore

, Pievani

, Cattaneo

, Frisoni

(2023) A peripheral signature of Alzheimer’s disease featuring microbiota-gut-brain axis markers. Alzheimers Res Ther 15, 101.

24.

Park

, Choi

, Lee

, Kwon

, Yang

, Jeong

, Kim

, Han

(2017) Metagenome analysis of bodily microbiota in a mouse model of Alzheimer disease using bacteria-derived membrane vesicles in blood. Exp Neurobiol 26, 369–379.

25.

Teng

, Mu

, Xu

, Zhang

, Sriwastva

, Liu

, Li

, Lei

, Sundaram

, Hu

, Zhang

, Park

, Hwang

, Rouchka

, Zhang

, Yan

, Merchant

, Zhang

(2022) Gut bacterial isoamylamine promotes age-related cognitive dysfunction by promoting microglial cell death. Cell Host Microbe 30,944–-960.e948.

26.

Grahnemo

, Nethander

, Coward

, Gabrielsen

, Sree

, Billod

, Engstrand

, Abrahamsson

, Langhammer

, Hveem

, Ohlsson

(2022) Cross-sectional associations between the gut microbe Ruminococcus gnavus and features of the metabolic syndrome. Lancet Diabetes Endocrinol 10, 481–483.

27.

Henke

, Brown

, Cassilly

, Vlamakis

, Xavier

, Clardy

(2021) Capsular polysaccharide correlates with immune response to the human gut microbe Ruminococcus gnavus. Proc Natl Acad Sci U S A 118, e2007595118.

28.

Marx

, Lane

, Hockey

, Aslam

, Berk

, Walder

, Borsini

, Firth

, Pariante

, Berding

, Cryan

, Clarke

, Craig

, Su

, Mischoulon

, Gomez-Pinilla

, Foster

, Cani

, Thuret

, Staudacher

, Sanchez-Villegas

, Arshad

, Akbaraly

, O’Neil

, Segasby

, Jacka

(2021) Diet and depression: Exploring the biological mechanisms of action. Mol Psychiatry 26, 134–150.

29.

Tan

, Norhaizan

(2019) Effect of high-fat diets on oxidative stress, cellular inflammatory response and cognitive function. Nutrients 11, 2579.

30.

Bini

, Parikh

, Lacadie

, Hwang

, Shah

, Rosenberg

, Seo

, Lam

, Hamza

, De Aguiar

, Constable

, Sherwin

, Sinha

, Jastreboff

(2022) Stress-level glucocorticoids increase fasting hunger and decrease cerebral blood flow in regions regulating eating. Neuroimage Clin 36, 103202.

31.

Westfall

, Iqbal

, Sebastian

, Pasinetti

(2019) Gut microbiota mediated allostasis prevents stress-induced neuroinflammatory risk factors of Alzheimer’s disease. Prog Mol Biol Transl Sci 168, 147–181.

32.

Choi

, Kim

, Lee

, Kim

, Sung

, Choi

, Venkat

, Park

, Jung

, Kaang

(2021) Synaptic correlates of associative fear memory in the lateral amygdala. Neuron 109, 2717–2726.e3.

33.

Zheng

, Tu

, Li

, Hua

, Liu

, Pan

, Hu

, Zhang

(2021) Neuroinflammation induces anxiety- and depressive-like behavior by modulating neuronal plasticity in the basolateral amygdala. Brain Behav Immun 91, 505–518.

34.

Roesler

, Parent

, LaLumiere

, McIntyre

(2021) Amygdala-hippocampal interactions in synaptic plasticity and memory formation. Neurobiol Learn Mem 184, 107490.

35.

Song

(2023) Amygdala activity and amygdala-hippocampus connectivity: Metabolic diseases, dementia, and neuropsychiatric issues. Biomed Pharmacother 162, 114647.

36.

Sedzikowska

, Szablewski

(2021) Insulin and insulin resistance in Alzheimer’s disease. Int J Mol Sci 22, 9987.

37.

, Wang

, Chen

, Li

, Zhao

, Ding

, Zhang

(2023) Microbiome features differentiating unsupervised-stratification-based clusters of patients with abnormal glycometabolism. mBio 14, e0348722.

38.

Meslier

, Laiola

, Roager

, De Filippis

, Roume

, Quinquis

, Giacco

, Mennella

, Ferracane

, Pons

, Pasolli

, Rivellese

, Dragsted

, Vitaglione

, Ehrlich

, Ercolini

(2020) Mediterranean diet intervention in overweight and obese subjects lowers plasma cholesterol and causes changes in the gut microbiome and metabolome independently of energy intake. Gut 69, 1258–1268.

39.

Hoscheidt

, Sanderlin

, Baker

, Jung

, Lockhart

, Kellar

, Whitlow

, Hanson

, Friedman

, Register

, Leverenz

, Craft

(2022) Mediterranean and Western diet effects on Alzheimer’s disease biomarkers, cerebral perfusion, and cognition in mid-life: A randomized trial. Alzheimers Dement 18, 457–468.

40.

Stillman

, Jakicic

, Rogers

, Alfini

, Smith

, Watt

, Kang

, Erickson

(2021) Changes in cerebral perfusion following a 12-month exercise and diet intervention. Psychophysiology 58, e13589.

Mediation of Regional Cerebral Blood Flow in the Relationship between Specific Gut Microbiota and Cognition in Vascular Cognitive Impairment

Abstract

Background:

Objective:

Methods:

Results:

Conclusions:

Keywords

INTRODUCTION

METHODS

Dataset and study design

Data collection and neuropsychological assessments

Fecal sample collection and 16 S ribosomal RNA (rRNA) sequencing

MRI acquisition and processing

Statistical analysis

RESULTS

Baseline characteristics of clinical and regional CBF in both groups

GM diversity and composition

Mediating effects of regional CBF on the association between GM and cognition

DISCUSSION

Conclusions

Footnotes

ACKNOWLEDGMENTS

FUNDING

CONFLICT OF INTEREST

DATA AVAILABILITY

References