Sage Journals: Discover world-class research

Abstract

Introduction

The gut microbiota modulates dementia pathogenesis through immune interactions. Using Mendelian randomization, we investigate immune mediated mechanisms linking microbial dysbiosis to four dementia subtypes (Alzheimer's disease, Frontotemporal dementia, Vascular dementia, Parkinson's disease dementia . Our study tests whether gut microbiome effects on dementia are transmitted via immunoregulatory pathways.

Methods

Genome wide association studies data included gut microbiota, 731 immune traits, and dementia cohorts (Alzheimer's disease, Frontotemporal dementia, Vascular dementia, Parkinson's disease dementia). Two step Mendelian randomization with Inverse Variance Weighted analyses assessed mediation effects, controlled by F-statistics >10 and Steiger filtering. Sensitivity analyses addressed pleiotropy.

Results

A total of 37 gut microbiome species demonstrated potential causal effects relationships with four types of dementia, and 137 immune cell subsets exhibited potential causal effects associations with these four dementia subtypes. In the Two step Mendelian randomization analysis, CD45RA + CD28− CD8+ T cells, CD19 on IgD− CD38dim B cells, and BAFF-R on CD20− B cells were shown to exert mediating effects between class/order/family.Deltaproteobacteria and Alzheimer's disease. CD4+ CD8+ T cells were found to exert a mediating effect between genus.Roseburia and Parkinson's disease dementia . CD20− CD38− B cells, CD19 on CD20− B cells, and IgD on unswitched memory B cells were found to exert a mediating effect between class/order/family.Coriobacteriales,genus.Lactococcus and Vascular dementia.

Conclusion

This Mendelian randomization study revealed that certain immune cells serve as mediators in the pathway by which the gut microbiome contributes to the onset of dementia.

Keywords

Alzheimer's disease frontotemporal dementia vascular dementia Parkinson's disease dementia two step mendelian randomization immune cells gut microbiota

Introduction

Dementia is a syndrome characterized by cognitive and memory impairments. The subtypes of dementia include Alzheimer's disease (AD), frontotemporal dementia (FTD), and vascular dementia (VD).¹ Additionally, patients with Parkinson's disease (PD) may experience a full spectrum of cognitive deficits, ranging from subjective cognitive decline and mild cognitive impairment to Parkinson's disease dementia (PDD).²

The infiltration of both central and peripheral immune cells has been implicated in the pathogenesis of neurodegenerative diseases such as AD and PD.³ Conversely, certain immune cell populations, notably regulatory T cells (Tregs), have been shown to confer neuroprotection in models such as MPTP.⁴ The mechanisms by which peripheral immune cells influence central nervous system disease remain under investigation. One hypothesis posits that these cells release inflammatory cytokines, where dysregulation may lead to the development of neurological disorders.⁵ Alternatively, peripheral immune cells may exert effects via the choroid plexus, facilitating communication with the central nervous system.⁶ A cohort study indicated that elevated peripheral blood T cell levels are a risk factor for the development of dementia, whereas an increase in lymphocyte populations appears to confer a protective effect against this condition.⁷

The human gastrointestinal tract hosts trillions of microorganisms known collectively as the gut microbiota. These microbes play critical roles in regulating various physiological aspects of the host, particularly in the maturation and functionality of the immune system. Existing evidence suggests that the gut microbiota can influence the development of neurological diseases by affecting immune cells, positioning immune modulation by the microbiota as a crucial pathway for coordinating the microbiota‒gut‒brain axis.⁸ Both the gut microbiota and immune cells may influence the development of dementia, potentially through interactions between these two components. We hypothesize that immune cells could serve as mediators in the pathway from the gut microbiota to dementia. To investigate this possibility, we propose conducting a mediation Mendelian randomization (MR) analysis.

MR is an analytical method that uses germline genetic variants as instrumental variables (IVs) to assess potential causal relationships between an exposure (e.g., gut microbiota) and an outcome (e.g., Parkinson's disease). This approach mitigates confounding and reverse causality inherent in observational studies by emulating the design principle of a randomized controlled trial.⁹ In practice, genome wide significant single nucleotide polymorphisms (SNPs) associated with the exposure (typically p < 1 × 10⁻⁵) are selected as candidate IVs. Independent IVs are then obtained through linkage disequilibrium (LD) clumping (r² < 0.001 within a 10,000 kb window). These genetic instruments effectively stratify the population into groups based on allele carriage. The causal effect is estimated by comparing outcome differences between these genetically defined groups, often using ratio estimators.¹⁰

This investigation implemented a two step Mendelian randomization framework to 1) elucidate etiological pathways between gut microbial taxa, immunophenotypes, and four dementia subtypes (AD/FTD/VD/PDD), and 2) quantify the mediation proportion of immune cell profiles in microbiota and dementia relationships through causal mediation analysis.

Materials and methods

Study design

Our analytical pipeline comprised four stages: Microbiota and Dementia Causality: MR identified 37 dementia associated microbial features using inverse variance weighted regression. Immune and Dementia Links: Exposure wide MR across 137 immunophenotypes revealed causal associations with four dementia subtypes. Microbiota and Immune Interplay: MR analysis mapped 37 candidate taxa to 137 immunophenotypes. Mediation Verification: A causal mediation framework quantified immune cell contributions in microbiota and dementia relationships. Study design shown in Figure 1.

Figure 1.

Study design. Research flow chart of the study. Exposures means the phenotypes used as the potential causal factors in MR analysis. Mediator means the variable between exposure and outcome. Outcome means the phenotype used as the affected variable in MR analysis. IVs represents instrument variables which can be screened by different p value levels. GM: gut microbiota. MR: Mendelian randomization. GWAS: genome wide association study.

Data sources

Summary level gut microbiota data were obtained from the MiBioGen consortium (https://mibiogen.gcc.rug.nl), comprising 18,340 participants across 24 cohorts, where 78% of the study population self reported European ancestry.¹¹ The MiBioGen consortium systematically compiled and evaluated genome wide genotype information in conjunction with 16S fecal microbiome datasets from subjects. In the context of genetic loci identification associated with relative abundance, only those taxa present in over 10% of the samples were taken into account. This rigorous selection process led to the inclusion of 211 taxa, which encompassed 131 genera, 35 families, 20 orders, 16 classes, and 9 phyla.¹² Meanwhile, genetic associations with AD were derived from a meta analysis of genome wide association studies (GWASs) conducted on individuals of European descent. The data for this analysis were provided by the International Genomics of Alzheimer's Project (IGAP), with a sample size of 21,982 cases and 41,944 controls.¹³ The R10 version of the FinnGen consortium served as the source of data for VD, FTD, and PDD.¹⁴ The disease diagnosis criteria are available at https://risteys.finngen.fi/, and the relevant disease diagnostic criteria are classified according to the International Classification of Diseases, 10th Revision (ICD-10). With the data set encompassing 2717 VD cases and 39,3024 controls, 129 FTD cases and 39,2463 controls, as well as 589 PDD cases and 18,400 controls, the characteristics of the respective GWAS data sources are outlined in Table S1.

Data extraction

The MR approach utilizes SNPs that are robustly linked to the exposure as IVs to assess the causal impact of an exposure on an outcome. For IVs to be deemed appropriate, three assumptions must hold true: (1) they must exhibit a strong association with the exposure of interest, (2) they must be independent of any confounding influences, and (3) they should not be directly associated with the outcome variable except through the exposure.¹⁵

To satisfy criterion 1, we implemented a stringent genome wide significance threshold (p < 1 × 10-5) and mandated that the F statistic be ≥ 10. This rigorous approach ensured the inclusion of strong IVs in our analysis. It is crucial to avoid weak IVs, as their utilization may introduce bias into MR results.¹⁶ To ensure the independence of the SNPs, linkage disequilibrium (correlation coefficient r2 < 0.001) was removed within a genetic distance range of 10,000 kb.¹⁷

MR and mediation analysis

We employed a two-step MR framework. First, univariable MR analyses were conducted to identify GM taxa and immune cell traits exhibiting significant causal effects on dementia. Analyses utilized the IVW method for exposures with multiple IVs. Only GM and immune cell pairs demonstrating significant associations with dementia (IVW p < 0.05), without heterogeneity or pleiotropy, were considered for mediation analysis.¹⁸

Within this mediation framework, GM acts as the exposure, immune cells as the mediator, and dementia as the outcome. To ensure biological plausibility, GM and immune cell pairs were required to satisfy specific directionality criteria based on the sign of their effects: if the total effect of a GM on dementia (βₑₒ) was positive, the effect of the GM on the immune cell (βₑₘ) and the effect of the immune cell on dementia (βₘₒ) had to share the same direction (βₑₘ × βₘₒ > 0); if negative, their effects had to be in opposite directions (βₑₘ × βₘₒ < 0). For qualifying pairs, the mediated proportion was calculated as (βₑₘ × βₘₒ) / βₑₒ.^19,20

Sensitivity analyses

Sensitivity analyses were performed using four MR methods—MR‒Egger, weighted median, simple mode, and weighted mode—that make different assumptions about pleiotropy. Cochran's Q statistic, calculated via both IVW and MR‒Egger regression, was used to assess heterogeneity (p > 0.05 indicated no significant heterogeneity). The MR‒Egger intercept test was applied to evaluate pleiotropy, with p > 0.05 suggesting no significant pleiotropic effects. All MR analyses were conducted using specific R packages (version 4.3.2), including “TwoSampleMR”, “tidyverse”, “ggplot2”.

Results

Potential genetic causality and correlations between the GM and dementia

Univariate MR analysis revealed significant associations between dementia and the GM, as determined by the IVW method (Figure 2). The results from the weighted median, MR‒Egger, simple mode, and weighted mode methods are provided in the supplementary material (Table S2). The SNP conditions are provided in Table S3, the polygenic results are listed in Table S4, and the heterogeneity is detailed in Table S5. The bacterial genus Oscillospira (OR: 1.322, 95% CI: 1.114–1.568) represents the most significant risk factor for AD, whereas the genus Ruminococcus1 (OR: 0.769, 95% CI: 0.643–0.919) is the most notable protective factor. For VD, the bacterial family Porphyromonadaceae (OR: 1.611, 95% CI: 1.083–2.398) is identified as the most significant risk factor, whereas the genus Ruminococcus1 (OR: 0.7, 95% CI: 0.517–0.949) again emerges as the most prominent protective factor. The genus Phascolarctobacterium (OR: 5.082, 95% CI: 1.417–18.228) is the most significant risk factor for FTD, with the class Betaproteobacteria (OR: 0.225, 95% CI: 0.051–0.985) providing the most substantial protective effect. Additionally, the genus Ruminiclostridium9 (OR: 2.429, 95% CI: 1.091–5.41) is recognized as the most significant risk factor for PDD, whereas the order Victivallales (OR: 0.367, 95% CI: 0.219–0.617) and the class Lentisphaeria (OR: 0.367, 95% CI: 0.219–0.617) are highlighted as the most notable protective factors. Our analysis of the ORs for the impact of the GM on different types of dementia revealed that, in terms of the magnitude of direct effects estimated by univariable MR, certain gut microbial taxa showed larger effect sizes on FTD compared to other dementia subtypes, whereas the observed direct effects on AD were relatively modest. It is important to note that this refers solely to the strength of initial statistical associations and does not imply established biological pathways, which were further explored in the subsequent mediation analysis; for example, the class Betaproteobacteria displays protective roles in both AD and FTD, whereas the genus Ruminococcus1 is protective in both AD and VD. Interestingly, the genus Roseburia acts as a protective factor in FTD but is identified as a risk factor in PDD. These findings suggest that although there may be shared mechanisms underlying the role of the gut microbiota in dementia, distinct differences exist across the various types of dementia.

Figure 2.

Potential genetic causality and correlation between GM and dementia. Note: nSNP, number of selected IVs; or, odds ratio; pval, MR analysis p value;Alzheimer's disease (AD), frontotemporal dementia (FTD), vascular dementia (VaD), and Parkinson's disease dementia (PDD).

Potential genetic causality and correlations between immune cells and dementia

Univariate MR analysis revealed significant associations between dementia and immune cells, as determined by the IVW method (Figure 3). The results from the weighted median, MR‒Egger, simple mode, and weighted mode methods are provided in the supplementary material (Table S6). The SNP conditions are provided in Table S7, the polygenic results are in Table S8, and the heterogeneity is detailed in Table S9. Our results indicate that 43 immune cell subsets can influence the onset of AD, which is the highest number compared with other types of dementia. When the ORs of immune cell effects on different dementia types were compared, we found that immune cells had a more significant effect on FTD than on other forms of dementia. Furthermore, we observed that certain immune cells can affect the onset of multiple dementia types. For example, CD64 on monocytes is associated with the occurrence of both PDD and VD. Similarly, CD8 on CD28+ CD45RA− CD8+ T cells serves as a protective factor in both AD and FTD. HLA-DR expression on CD14+ CD16− monocytes and on CD14 + monocytes is considered a risk factor for both AD and FTD.

Figure 3.

Potential genetic causality and correlation between immune cell and dementia. Note: nSNP, number of selected IVs; or, odds ratio; pval, MR analysis p value;Alzheimer's disease (AD), frontotemporal dementia (FTD), vascular dementia (VaD), and Parkinson's disease dementia (PDD).

Interestingly, some immune cell markers yield contrasting effects across different dementia types. For example, CCR2 on monocytes, CD8 on effector memory CD8+ T cells, and the absolute count of IgD + CD24− B cells are considered risk factors in AD but are protective in FTD. Additionally, PDL-1 on monocytes has been identified as a risk factor in VD and a protective factor in FTD. The percentage of plasmacytoid dendritic cells (% dendritic cells) is deemed a risk factor in FTD and plays a protective role in AD. These findings underscore the complex and context dependent roles of immune cells in the pathophysiology of various dementia types.

Potential genetic causality and correlations between immune cells and the GM

We conducted MR analysis on the GM and immune cells identified as positive in the previous two steps. We propose that the following criteria must be met for the mediation effect to be considered valid: first, the IVW p values of the MR results for both the gut microbiota and immune cells must be less than 0.05; second, the directions of the indirect and direct effects must be consistent; and finally, the 95% confidence interval for the mediation effect must be entirely above or below zero. Table 1 and Figure 4 present the results of the mediation effects established for the gut microbiota and corresponding immune cells linked to different types of dementia. The results from the weighted median, MR‒Egger, simple mode, and weighted mode methods are provided in the supplementary material (Table S10). The SNP conditions are provided in Table S11, the polygenic results are listed in Table S12, and the heterogeneity is detailed in Table S13. Our results indicate that AD is the type of dementia most influenced by the gut microbiota through immune cells. In contrast, FTD does not yield positive results due to the opposing directions of the direct and indirect effects or because the 95% confidence interval for the indirect effect crosses zero. Furthermore, we observed that the same gut microbiota can influence the onset of dementia through different immune cells. For example, bacteria from the order, family, or class Coriobacteriales can affect the occurrence of VD through the CD20− CD38− B-cell percentage and IgD on unswitched memory B cells. Similarly, the class Deltaproteobacteria promotes AD through the CD45RA + CD28− CD8+ T-cell percentage, CD19 on IgD− CD38dim B cells, and BAFF-R on CD20− B cells. similarly, the same immune cell type can be influenced by different gut microbiota to drive disease onset; for example, the percentage of CD45RA + CD28− CD8+ T cells can be influenced by the order/family Desulfovibrionales and class Deltaproteobacteria, thereby promoting the occurrence of AD.

Figure 4.

Different GM communities influence the onset of various types of dementia through distinct immune cell mechanisms. Note: Alzheimer's disease (AD), frontotemporal dementia (FTD), vascular dementia (VaD), and Parkinson's disease dementia (PDD).

Table 1.

Mediation analysis results.

Exposure	Mediator	Outcome	Direct effect (β_EM)	Direct effect (β_MO)	Direct effect (β_EO)	Indirect effect (β_EM×β_MO)	95%CI	Proportion mediated (β_EM×β_MO*/β_EO)
order.Desulfovibrionales.id.3156	CD45RA + CD28- CD8+ T cell %T cell	AD	2.1398	0.0008	0.2384	0.0017	0.0003,0.0039	0.71%(95% CI :0.13%, 1.64%)
family.Desulfovibrionaceae.id.3169	CD45RA + CD28- CD8+ T cell %T cell	AD	2.1080	0.0008	0.2683	0.0017	0.0002,0.0041	0.63%(95% CI :0.07%, 1.53%)
class.Deltaproteobacteria.id.3087	CD45RA + CD28- CD8+ T cell %T cell	AD	1.9311	0.0008	0.2234	0.0015	0.0002,0.0037	0.67%(95% CI :0.09%, 1.66%)
class.Deltaproteobacteria.id.3087	CD19 on IgD- CD38dim B cell	AD	0.2491	0.0731	0.2234	0.0182	0.0004,0.0448	8.15%(95% CI :0.18%,20.05%)
class.Deltaproteobacteria.id.3087	BAFF-R on CD20- B cell	AD	−0.2853	−0.1096	0.2234	0.0313	0.0026,0.0717	14.01%(95% CI :1.16%, 32.09%)
genus.Roseburia.id.2012	CD4 + CD8+ T cell %leukocyte	PDD	0.3423	0.3174	0.6647	0.1086	0.0129,0.2430	16.34%(95% CI:1.95%, 36.56%)
class/order/family.Coriobacteriales.	CD20- CD38- B cell %B cell	VD	−0.3561	−0.0696	0.3477	0.0248	0.0026,0.0569	7.13%(95% CI:0.75%, 16.36%)
class/order/family.Coriobacteriales.	IgD on unswitched memory B cell	VD	0.3474	0.0756	0.3477	0.0263	0.0010,0.0634	7.56%(95% CI:0.29%, 18.23%)
genus.Lactococcus.id.1851	CD19 on CD20- B cell	VD	0.2159	−0.1224	−0.0510	−0.0264	−0.0035,-0.0594	51.76%(95% CI:6.86%, 100%)

Note: β_EM is the MR casual effect of exposure E on mediator M, β_MO is the MR casual effect of mediator M on outcome O, and β_EO is the ‘ total ‘ effect of exposure E on outcome O.95%CI, calculate the 95%confidence interval for the Indirect effect.

Discussion

Our research revealed that different gut microbiota influence the development of various types of dementia (i.e., AD, PDD, and VD) through distinct immune cell pathways. It is worth noting that our research findings suggest the potential of gut microbiota to influence the occurrence and development of dementia through its effects on immune cells.

Class Betaproteobacteria exhibited protective effects against both AD and FTD, while genus Ruminococcus1 was protective against both AD and VD. Conversely, genus Roseburia demonstrated opposing effects, serving as a protective factor in FTD but as a risk factor in PDD. These patterns suggest that while certain microbial taxa may exert broad neuroimmunomodulatory effects, others have context-dependent roles that vary across dementia subtypes.

The relationship between the gut microbiota and AD has been confirmed in several studies. There is a gut-brain axis between the gastrointestinal tract and the central nervous system, with frequent bidirectional communication. Dysbiosis of the gut microbiota may contribute to the development of AD, and modulating the gut microbiota may be helpful for the prevention of AD.²¹ However, the mechanism by which the gut microbiota causes AD through immune cells has not been elucidated. In the present study, we found that Desulfovibrionales can contribute to the onset of AD by influencing CD45RA + CD28− CD8+ T cells. Order.Desulfovibrionales, most species of which are anaerobic microorganisms that can promote the conversion of sulfate to sulfide, primarily generating hydrogen sulfide.²² Research by Dizbrou et al. has demonstrated that the levels of hydrogen sulfide and its metabolites in the plasma of AD patients are significantly elevated compared with those in controls, suggesting their potential as biomarkers for the diagnosis and intervention of AD.²³ Additionally, an animal study revealed that sulfate reducing bacteria can induce T-cell proliferation, activate inflammatory responses, and contribute to the development of colitis.²⁴ These findings suggest that Desulfovibrionales may promote T-cell proliferation through hydrogen sulfide, thereby contributing to the onset of AD.²⁵ Interestingly, in the present study, the class Deltaproteobacteria may similarly affect the occurrence of AD through their impact on CD45RA + CD28− CD8+ T cells. This correlation may be attributed to the fact that the class Deltaproteobacteria encompasses various bacteria, including sulfate reducing bacteria.²⁶

In the literature on the promotion of AD development by CD45RA + CD8+ T cells, a considerable number of relevant studies exist; for example, CD8+ CD45RA + T cells accumulate in elderly individuals, exhibiting characteristics such as high cytotoxicity, low proliferation, and sensitivity to apoptosis, which are associated with excessive inflammation and various chronic inflammatory states.²⁷ CD8+ CD45RA + T cells are also capable of increasing the expression of the terminal differentiation marker CD57, which has been utilized to identify senescent T cells.²⁸ CD8+ T cells may participate in the pathogenesis of AD by releasing cytokines such as IFN-γ, TNF-α, and IL-17, which contribute to neuronal damage.²⁹ These cytokines increase the permeability of the blood‒brain barrier and facilitate the migration of T cells into the central nervous system (CNS) parenchyma. CD8+ CD45RA + T cells can also trigger apoptosis.³⁰ These findings align with our results. However, whether Deltaproteobacteria influences CD8+ CD45RA + T cells through hydrogen sulfide, thereby contributing to the onset of AD, requires further investigation.

In addition, our research indicates that the class Deltaproteobacteria may promote the onset of AD by increasing the level of CD19 on IgD− CD38dim B cells or by reducing the expression of BAFF-R on CD20− B cells. According to our findings, BAFF-R on CD20− B cells appears to be associated with a decreased incidence of AD. However, it remains unclear which specific bacteria within the class Deltaproteobacteria or whether all bacteria in this class influence these two immune cell subsets. Research on CD20− B cells is limited; however, CD20+ B cells are known to be involved in the development of multiple sclerosis (MS), and therapeutic antibodies targeting CD20+ B cells, such as ocrelizumab, the first humanized anti CD20 monoclonal antibody approved for the treatment of relapsing–remitting MS (RRMS) and primary progressive MS (PPMS),³¹ have been developed. This antibody depletes both immature and mature B cells in the circulation without affecting CD20− plasma cells. Additionally, B-cell-activating factor (BAFF) can inhibit CD20-mediated apoptosis.³² Given the relationship between CD20+ B cells and CD20− B cells, as well as the role of BAFF in inhibiting CD20-mediated apoptosis, further investigations are warranted to elucidate whether BAFF-R on CD20− B cells may confer protection against AD. While the precise role of CD20⁻ B cells in AD remains unclear, an intriguing parallel may be drawn from the established role of CD20⁺ B cells in MS, where they contribute to pathogenesis. This observation is hypothesis generating, suggesting that CD20⁻ B cells could potentially fulfill a complementary or regulatory function in our context. However, this extrapolation remains speculative and merits direct experimental investigation in future studies to delineate the specific phenotype and mechanistic role of CD20⁻ B cells in AD

Similarly, research on the role of CD19 on IgD− CD38dim B cells in AD is limited. CD19+ B cells are recognized as risk factors for the onset and progression of neuromyelitis optica spectrum disorder (NMOSD), and several therapeutic agents targeting CD19+ B cells, such as inebilizumab, have been extensively utilized as therapeutic interventions.³³ Although existing studies have not demonstrated a direct relationship between CD20 + and CD19+ B cells and AD, some data support the involvement of the adaptive immune system, including B cells, in the pathogenesis of AD. Certain HLA alleles, including HLA-DRB1*15:01, have been associated with an increased risk of AD, and a reduction in peripheral B-cell subpopulation levels has been observed in some patients with AD, which may be related to genetic alterations in these cells.³⁴ Therefore, further investigations into how CD20 + and CD19+ B cells may contribute to the pathogenesis of AD are warranted.

A recent study delineates that human induced pluripotent stem cell (hiPSC) derived dopaminergic neurons necessitate exposure to alpha-synuclein and immunostimulants to form Lewy body like inclusions. Moreover, lysosomal dysfunction instigated by immunostimulation appears to be a pivotal factor in promoting the pathology of Lewy body formation, underscoring its critical role in the neurodegenerative process.³⁵ Our data indicate that Roseburia may influence the occurrence of PDD through its effects on CD4+ CD8+ T cells. There appears to be no direct evidence supporting the notion that Roseburia promotes the number of CD4+ CD8+ T cells. However, an animal study demonstrated that Roseburia can increase the number of CD4+ CD25+ Foxp3+ T cells, which are involved in the regulation of inflammation.³⁶ These findings suggest that Roseburia may have regulatory effects on various T-cell subsets. Notably, T cells expressing CD4 + and CD8 + are considered markers of aging.³⁷ Additionally, a case‒control study revealed elevated levels of CD4+ CD8+ T cells in patients with PD.³⁸ This finding aligns with our results, which indicate that Roseburia may contribute to the development of PDD through the increase in CD4+ CD8+ T-cell levels.

Interestingly, our results indicate that Coriobacteriales may contribute to the development of VD and that this effect may be achieved by regulating CD20− CD38− B cells and IgD on unswitched memory B cells. In a case‒control study comparing an obese group with cognitive impairment and an obese group with normal cognition, Coriobacteriales was negatively correlated with obesity related cognitive scores,³⁹ although direct evidence linking this bacterium to VD remains lacking. Moreover, an experimental study demonstrated that probiotic supplementation can decrease the abundance of Coriobacteriales and reduce the proportion of CD4+ T cells.⁴⁰

Our results indicate that CD20− CD38− B cells serve as protective factors for VD and, as previously mentioned, that CD20− B cells act as protective factors in AD. These findings suggest that CD20− B cells may play a significant role in the pathogenesis of dementia. Additionally, CD38+ B cells are associated with multiple myeloma, for which the anti-CD38 monoclonal antibody daratumumab has been approved for treatment.⁴¹ This raises the question of whether there may be shared pathogenic mechanisms between certain hematological diseases, such as multiple myeloma and dementia. Notably, some case reports have linked intravascular large B-cell lymphoma with rapidly progressive dementia.^42,43

In addition, IgD on unswitched memory B cells has been found to activate and promote the production of proinflammatory cytokines during SARS-CoV-2 infection.⁴⁴ Additionally, a reduction in the levels of IgD on unswitched memory B cells has been observed in patients with rheumatoid arthritis treated with rituximab.⁴⁵ These findings suggest that IgD on unswitched memory B cells plays a significant role in immune damage, which may also be one of the mechanisms contributing to the development of VD.

Furthermore, our research indicates that Lactococcus plays a protective role in vascular dementia by increasing the level of CD19 on CD20− B cells. Animal studies have demonstrated that Lactococcus can protect against neurodegeneration induced by beta amyloid protein in Caenorhabditis elegans.⁴⁶ Furthermore, supplementation with Lactococcus lactis has been shown to confer protective effects on patients with dementia. However, there is currently a lack of direct evidence linking Lactococcus to vascular dementia. Currently, there is a scarcity of direct evidence linking CD19+ CD20− B cells to VD. However, intriguingly, these findings suggest that B cells lacking CD20 expression may serve as protective factors against VD. These findings indicate that CD20− B cells might play a significant role in the prevention of dementia.

Regrettably, our study did not conclude that the gut microbiota influences the occurrence of FTD through its effects on immune cells. Although positive MR results indicated associations between the gut microbiota and FTD, as well as between immune cells and FTD, further results connecting these pathways are lacking.

Some of the proportion of the total effect mediated by immune cells, as estimated in our study, was quantitatively small. The potential biological relevance of these modest estimates warrants consideration. First, MR estimates reflect the lifelong effect of genetic predisposition, and even small effect sizes can point to mechanistically meaningful pathways. Second, the exposure in question gut microbiota composition and immune cell abundance is chronic and subject to long-term cumulative biological interaction. Small per individual genetic effects, when sustained over a lifetime and across populations, may translate into meaningful public health relevance. Thus, while the direct clinical translation of these specific effect sizes is limited, our findings serve a primary purpose of identifying robust, genetically supported candidate pathways (e.g., specific GM taxa and immune traits) that merit priority in future experimental and longitudinal research to elucidate their precise biological roles and potential for therapeutic targeting.

Limitation

Our study has several limitations. First, our conclusions are derived solely from SNPs with a significance threshold of p < 1e-5. Second, we defined our final results based on a 95% confidence interval for mediation effects that did not cross zero, aiming to establish the most reliable conclusions; however, this may have led to false negative results. For example, in the relationship where the gut microbiota influences FTD via immune cells, the triangular mediation effect was valid, but we did not adopt it as a final result because its confidence interval crossed zero. Third, the mediating effects we observed—such as the percentages of CD45RA + CD28− CD8+ T cells, CD4 + CD8+ T cells, and CD20− CD38− B cells—reflect the proportions of these cells within different leukocyte populations. Although an increase in these cell types can increase their proportions, a corresponding decrease in total leukocytes can also lead to an increase in these proportions; our discussion focused primarily on increases cell numbers. Fourth, the Finnish database used ICD-10 F02.3 for PDD diagnosis and that potential misdiagnosis of Lewy body dementia as PDD cannot be ruled out. And, MR provides potential evidence supporting causality between exposure and outcome, not definitive proof like RCTs. Also，an important limitation of this study concerns the interpretation of findings for FTD. The FTD sample size (n = 129 cases) was substantially smaller than those for other dementia outcomes. Consequently, the statistical power to detect significant mediation pathways for FTD was severely limited. Meanwhile, in the exploratory univariable MR analyses used to screen candidate microbiota and immune cell traits for inclusion in the mediation framework, we applied a nominal significance threshold (p < 0.05) without strict multiple testing correction (e.g., Bonferroni or FDR). This approach was adopted to maximize the identification of potential pathways for subsequent mediation testing, acknowledging that it may increase the risk of false positive associations. Therefore, the identified associations should be interpreted as hypothesis-generating and require validation in independent cohorts and functional studies.

In conclusion

In this study, we discovered that the gut microbiota can influence the development of various types of dementia through its effects on immune cells, identifying specific microbial strains and immune cell subsets involved in these processes. We also observed that the same microbial strains can affect different immune cells, sometimes with opposing effects. Furthermore, we identified commonalities in the mechanisms underlying different types of dementia caused by different immune cells, suggesting the potential existence of shared mechanisms in immune cell-driven dementia subtypes. In conclusion, we comprehensively explored the potential causal relationships between the gut microbiota, immune cells, and dementia, identifying several mediation pathways that offer insights for future research.

Supplemental Material

sj-docx-1-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-docx-1-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-2-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-2-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-3-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-3-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-4-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-4-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-5-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-5-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-6-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-6-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-7-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-7-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-8-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-8-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-9-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-9-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-10-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-10-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-11-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-11-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Supplemental Material

sj-csv-12-ini-10.1177_17534259261426829 - Supplemental material for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study

Supplemental material, sj-csv-12-ini-10.1177_17534259261426829 for Immune cells play a mediating role in the relationship between the gut microbiota and dementia: A Mendelian randomization study by Jianzhun Chen, Liuhui Zhu, Jie Liu, Jieyu Chen, Chunyu Liang, Chenxi Liu and Fang Wang, Yongyun Zhu, Xinglong Yang in Innate Immunity

Footnotes

Acknowledgements section

We want to acknowledge the participants and investigators of the FinnGen study, MiBioGen consortium and the International Genomics of Alzheimer's Project (IGAP)

ORCID iD

Xinglong Yang

Ethical considerations

The data used in this study were all from public databases and did not involve ethics approval and consent to participate.

Consent for publication

Written informed consent for publication of this paper was obtained from the First Affiliated Hospital of Kunming Medical University, Kunming, and all authors.

Contributions

Xinglong Yang offered funding and provided overall supervision and manuscript review. Jianzhun Chen handled the main research design, data analysis and manuscript writing. Liuhui Zhu gave experimental support and helped with data analysis. Jie Liu assisted in data analysis and literature review. Jieyu Chen participated in literature review and research coordination. Chunyu Liang offered technical support and aided in data management. Chenxi Liu, Yongyun Zhu and Fang Wang managed the data and ensured quality control. All authors had participated in paper writing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Applied Basic Research Foundation of Yunnan Province [Grant Numbers 202301AS070045, 202101AY070001-115]; National Natural Science Foundation of China [Grant number 81960242]; The Major Science and Technology Special Project of Yunnan Province [Grant number 202102AA100061];535 Talent Project of First Affiliated Hospital of Kunming Medical University[2023535D18]，The Innovative Team of Yunnan Province (202305AS350019）.

National Natural Science Foundation of China, The Innovative Team of Yunnan Province, The Major Science and Technology Special Project of Yunnan Province, the Applied Basic Research Foundation of Yunnan Province, 535 Talent Project of First Affiliated Hospital of Kunming Medical University, (grant number 81960242, 202305AS350019, 202102AA100061, 202301AS070045, 202101AY070001-115, 2023535D18).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

Supplemental material

Supplemental material for this article is available online.

References

Maclin

JMA

Wang

Xiao

. Biomarkers for the diagnosis of Alzheimer's disease, dementia Lewy body, frontotemporal dementia and vascular dementia. Gen Psychiatr 2019; 32: e100054.

Aarsland

Batzu

Halliday

, et al. Parkinson disease-associated cognitive impairment. Nat Rev Dis Primers 2021; 7: 47.

Berriat

Lobsiger

Boillée

. The contribution of the peripheral immune system to neurodegeneration. Nat Neurosci 2023; 26: 942–954.

Berriat

Lobsiger

Boillée

. The contribution of the peripheral immune system to neurodegeneration. Nat Neurosci 2023; 26: 942–954.

Liu

Tan

Zhang

, et al. Neuroimmune mechanisms underlying Alzheimer's disease: insights into central and peripheral immune cell crosstalk. Ageing Res Rev 2023; 84: 101831.

Ito

Komai

Mise-Omata

, et al. Brain regulatory T cells suppress astrogliosis and potentiate neurological recovery. Nature 2019; 565: 246–250.

Zhang

Wang

Chen

, et al. Peripheral immunity is associated with the risk of incident dementia. Mol Psychiatry 2022; 27: 1956–1962.

Fung

Olson

Hsiao

. Interactions between the microbiota, immune and nervous systems in health and disease. Nat Neurosci 2017; 20: 145–155.

Lawlor

Harbord

Sterne

, et al. Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med 2008; 27: 1133–1163.

10.

Bowden

Holmes

. Meta-analysis and Mendelian randomization: a review. Res Synth Methods 2019; 10: 486–496.

11.

Kurilshikov

Medina-Gomez

Bacigalupe

, et al. Large-scale association analyses identify host factors influencing human gut microbiome composition. Nat Genet 2021; 53: 156–165.

12.

Wang

Dai

Hou

, et al. Dissecting causal relationships between gut Microbiota, blood metabolites, and stroke: a Mendelian randomization study. J Stroke 2023; 25: 350–360.

13.

Kunkle

Grenier-Boley

Sims

, et al. Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates aβ, tau, immunity and lipid processing. Nat Genet 2019; 51: 414–430.

14.

Kurki

Karjalainen

Palta

, et al. Finngen provides genetic insights from a well-phenotyped isolated population. Nature 2023; 613: 508–518.

15.

Davies

Holmes

Davey Smith

. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. Br Med J 2018; 362: k601.

16.

Palmer

Lawlor

Harbord

, et al. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res 2012; 21: 223–242.

17.

Burgess

Small

Thompson

. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res 2017; 26: 2333–2355.

18.

Chen

Liang

, et al. Liver enzyme and risk of vascular dementia: a univariable and multivariable Mendelian randomization of European descent. Neuroprotection 2024; 2: 310–317.

19.

Carter

Sanderson

Hammerton

, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol 2021; 36: 465–478.

20.

Chen

Peters

Prins

, et al. Systematic Mendelian randomization using the human plasma proteome to discover potential therapeutic targets for stroke. Nat Commun 2022; 13: 6143.

21.

Zhang

Gao

Kwok

, et al. Gut microbiome-targeted therapies for Alzheimer's disease. Gut Microbes 2023; 15: 2271613.

22.

Kushkevych

Hýžová

Vítězová

, et al. Microscopic methods for identification of sulfate-reducing bacteria from various habitats. Int J Mol Sci 2021; 22: 4007.

23.

Disbrow

Stokes

Ledbetter

, et al. Plasma hydrogen sulfide: a biomarker of Alzheimer's disease and related dementias. Alzheimers Dement 2021; 17: 1391–1402.

24.

Figliuolo

Dos Santos

Abalo

, et al. Sulfate-reducing bacteria stimulate gut immune responses and contribute to inflammation in experimental colitis. Life Sci 2017; 189: 29–38.

25.

Kida

Ichinose

. Hydrogen sulfide and neuroinflammation. Handb Exp Pharmacol 2015; 230: 181–189.

26.

Suzuki

Ueki

Amaishi

, et al. Desulfovibrio portus sp. nov., a novel sulfate-reducing bacterium in the class Deltaproteobacteria isolated from an estuarine sediment. J Gen Appl Microbiol 2009; 55: 125–133.

27.

Lanna

Gomes

Muller-Durovic

, et al. A sestrin-dependent erk-jnk-p38 MAPK activation complex inhibits immunity during aging. Nat Immunol 2017; 18: 354–363.

28.

Akbar

Henson

Lanna

. Senescence of T lymphocytes: implications for enhancing human immunity. Trends Immunol 2016; 37: 866–876.

29.

Villegas-Mendez

Greig

Shaw

, et al. IFN-γ-producing CD4+ T cells promote experimental cerebral malaria by modulating CD8+ T cell accumulation within the brain. J Immunol 2012; 189: 968–979.

30.

Medana

Gallimore

Oxenius

, et al. MHC Class I-restricted killing of neurons by virus-specific CD8+ T lymphocytes is effected through the fas/FasL, but not the perforin pathway. Eur J Immunol 2000; 30: 3623–3633.

31.

Mancinelli

Rossi

Capra

. Ocrelizumab for the treatment of multiple sclerosis: safety, efficacy, and pharmacology. Ther Clin Risk Manag 2021; 17: 765–776.

32.

Saito

Miyagawa

Onda

, et al. B-cell-activating factor inhibits CD20-mediated and B-cell receptor-mediated apoptosis in human B cells. Immunology 2008; 125: 570–590.

33.

Lee

DSW

Rojas

Gommerman

. B cell depletion therapies in autoimmune disease: advances and mechanistic insights. Nat Rev Drug Discov 2021; 20: 179–199.

34.

Sabatino

Pröbstel

Zamvil

. B cells in autoimmune and neurodegenerative central nervous system diseases. Nat Rev Neurosci 2019; 20: 728–774.

35.

Bayati

Ayoubi

Aguila

, et al. Modeling Parkinson's disease pathology in human dopaminergic neurons by sequential exposure to α-synuclein fibrils and proinflammatory cytokines. Nat Neurosci 2024; 27: 2401–2416.

36.

Zhu

Song

Shen

, et al. Roseburia intestinalis inhibits interleukin-17 excretion and promotes regulatory T cells differentiation in colitis. Mol Med Rep 2018; 17: 7567–7574.

37.

Williams-Gray

Wijeyekoon

Scott

, et al. Abnormalities of age-related T cell senescence in Parkinson's disease. J Neuroinflammation 2018; 15: 66.

38.

Hisanaga

Asagi

Itoyama

, et al. Increase in peripheral CD4 bright+ CD8 dull+ T cells in Parkinson disease. Arch Neurol 2001; 58: 1580–1583.

39.

Zhao

Huang

, et al. Association between erythrocyte membrane fatty acids and gut bacteria in obesity-related cognitive dysfunction. AMB Express 2023; 13: 48.

40.

Kim

Jung

Choi

, et al. Probiotic supplementation has sex-dependent effects on immune responses in association with the gut microbiota in community-dwelling older adults: a randomized, double-blind, placebo-controlled, multicenter trial. Nutr Res Pract 2023; 17: 883–898.

41.

Vockova

Svaton

Karolova

, et al. Anti-CD38 therapy with daratumumab for relapsed/refractory CD20-negative diffuse large B-cell lymphoma. Folia Biol (Praha) 2020; 66: 17–23.

42.

Kerbauy

Pasqualin

DDC

Smid

, et al. Diffuse large B-cell lymphoma of the central nervous system presenting as “lymphomatosis cerebri” and dementia in elderly man: case report and review of the literature. Medicine (Baltimore) 2019; 98: e14367.

43.

Lin

Huang

, et al. Intravascular large B-cell lymphoma presenting as rapidly progressive dementia and stroke: a case report. Medicine (Baltimore) 2021; 100: e27996.

44.

Castleman

Santos

Lesteberg

, et al. Activation and pro-inflammatory cytokine production by unswitched memory B cells during SARS-CoV-2 infection. Front Immunol 2023; 14: 1213344.

45.

Ramwadhdoebe

van Baarsen

LGM

Boumans

MJH

, et al. Effect of rituximab treatment on T and B cell subsets in lymph node biopsies of patients with rheumatoid arthritis. Rheumatology (Oxford) 2019; 58: 1075–1085.

46.

Komura

Aoki

Kotoura

, et al. Protective effect of Lactococcus laudensis and Pediococcus parvulus against neuropathy due to amyloid-beta in caenorhabditis elegans. Biomed Pharmacother 2022; 155: 113769.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.10 MB

0.26 MB

0.52 MB

0.75 MB

1.19 MB

0.10 MB

3.71 MB

19.69 MB

0.34 MB

0.72 MB

1.67 MB

0.01 MB