The gut ecosystem encompasses various bacterial, viral, fungal, and protozoal species colonizing mucosal linings. The genetic density of the microbial genome outnumbers the human genome, with more than 3,000,000 microbial genes (1, 2). The integral role of the gut microbiome in metabolism, immune modulation, and vitamin production is governed by human cells' symbiotic relationship (3). Disturbances in balanced microbial structure, known as dysbiosis, result in increased pathogenic bacteria, and decreased commensal bacteria. For instance, in a comparison of oral and gut microbiomes, it has been shown that healthy aging individuals have an increased abundance of Akkermansia and Erysipelotrichaceae UCG-003 in their gut microbiome and, higher alpha diversity in their oral microbiomes; in that study, Streptococcus spp. was the only observed genus to have a significant reduction in its abundance in the oral and gut microbiomes of the healthy aging group (4).

Trimethylamine N-oxide (TMAO) is a key gut metabolite, derived from our diet, that has been shown to play a role in the development of various diseases. Mechanistic involvement of TMAO in the association between gut microbiome and stroke can provide potential therapeutic targets. The vast implications and effects of the gut microbiome in multi-disease processes make it difficult for research to be conducted to distinguish a single disease process. As the interdisciplinary topic of the gut microbiome and stroke is relatively novel, current investigations primarily focused on the microbial abundance in disease states are limited in providing functional or mechanistic information that may be vital in altering clinical outcomes.

Stroke, the fifth leading cause of death, affects more than 795,000 individuals in the United States every year and contributes to long-term disability, leading to both cognitive and ambulatory impairments (5). Stroke is broadly classified as ischemic and hemorrhagic. Risk factors for ishemic stroke include hypertension, diabetes, heart disease, hypercholesterolemia, smoking, and age. Ischemic stroke is classified into subtypes using the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification, including large artery atherosclerosis, cardioembolic, small vessel occlusion, other determined cause, or undetermined cause (6). There is also evidence that polygenic risk scores can augment stroke subtyping (7).

An emerging role of the gut microbiome in stroke patients provides deeper insights regarding potential avenues to improve outcomes at a personalized level. Our systematic review discusses current clinical evidence on the human gut microbiome's involvement in ischemic stroke at different stages of disease progression, recovery, and management. We further examine the role of ischemic stroke's etiological classification as a potential modulator. Gaining a deeper understanding of stroke modulators can guide the development of targeted strategies based on an ischemic event's etiological features. Knowledge about stroke-associated gut pathogenic bacteria (8) along with the development of machine learning algorithms, capable of recognizing and analyzing microbial parameters, can enhance diagnostic procedures, allow early identification of stroke risks, and facilitate precise interventions for improved outcomes. In the following sections, we aimed to answer the following questions: (1) What are the key findings in current literature that suggests the influence of the human gut microbiome on at-risk and current stroke patients? (2) What characteristics of gut metabolite trimethylamine N-oxide make it a suitable target for stroke prevention and intervention? We conclude with a brief discussion and perspectives for future directions with a focus on actionable biomarkers for targeted therapeutic interventions.

In phase I, we developed a comprehensive search strategy combining the three major themes of stroke, gut microbiome, and TMAO. The search string for stroke and gut microbiome were: (“Stroke”[Mesh] OR “Stroke, Lacunar”[Mesh] OR “Stroke Rehabilitation”[Mesh] OR “Infarction, Posterior Cerebral Artery”[Mesh] OR “Brain Stem Infarctions”[Mesh] OR “Infarction, Middle Cerebral Artery”[Mesh] OR “Infarction, Anterior Cerebral Artery”[Mesh] OR “Myocardial Infarction”[Mesh] OR “Anterior spinal artery stroke” [Supplementary Concept]) AND (“Microbiota”[Mesh] OR “Gastrointestinal Microbiome”[Mesh]) OR (stroke and microbiome). The search string for TMAO was: (“trimethyloxamine” [Supplementary Concept]) AND (“Diet”[Mesh] OR “Medication Therapy Management”[Mesh] OR Genetics). We included all published studies on these topics regardless of intervention target, approach, setting, and study design. We focused on human subject studies. PubMed database was searched and studies available between January 1, 1990, and March 22, 2020, were included. After performing the search, we manually screened the titles and abstracts for clinical applicability of gut microbiome in pre-stroke, stroke, and post-stroke stages.

In phase 2, we updated our search on May 1, 2020, and reviewed recent animal studies covering the underlying mechanisms. The following data were extracted from the included articles: study design, study setting, sample size, and limitations. Two review authors (ViS and VaS) extracted the data from eligible studies and the extracted data were reviewed for accuracy and relevance. The principal summary measure was the number of current article subtypes on the gut microbial abundance in stroke disease states. Studies with multiple study designs were counted as two separate studies for accurate synthesis. Data were imported into a table format for visualization. Missing data were not applicable as the study design data was apparent in the eligible studies. To better characterize the current literature on gut-microbiome, the range of sample sizes of the grouped study design was also calculated. Given the complexity and the non-comprehensive nature of the current literature, studies on gut microbiota and stroke were grouped based on clinical features such as stroke risk factors, stroke onset, and post-stroke recovery. Studies on TMAO were grouped based on the source, genetics, nutrition, comorbidities, medication, and mechanisms.

The search protocol regarding the gut microbiome and stroke resulted in 222 articles, from which 68 human studies were eligible for preliminary analysis, shown in Supplementary Tables 1, 2. Table 1 outlines the different study types, the number of articles in each study type, and the range of sample size. Five out of the 68 human subject articles had 2 study designs. Our preliminary analysis showed that the majority of articles in our present study were cohort studies (n = 27) and cross-sectional studies (n = 25); only one study was a clinical trial. The study design with the highest maximum sample size included cohort and metagenomic (human) studies.

Among the preliminary 68 human subject studies, 24 articles were selected for further in-depth analysis and inclusion in our present study based on their pertinent findings for the gut microbiome and stroke. As human studies were limited in reporting underlying mechanisms, we included nine additional animal studies in our review. Our search was expanded to include 40 articles that focused on the effects of genetics, nutrition, comorbidities, medications, and mechanisms related to TMAO. In total, we included 73 articles in this review, as depicted in Figure 1.

Our results showed that human gut microbial communities can be involved in different stroke stages, from facilitating the progression of stroke in the presence of its risk factors, to the recurrence of stroke. We identified altered gut microbes in pre-stroke, stroke onset, and post-stroke phases, which may serve as potential clinical biomarkers. The current understanding of the gut microbiome revolves around the large artery atherosclerosis subtype, probably due to its high prevalence compared to other etiologies. However, the microbiome's vast effects can play a role in the pathogenesis of all stroke etiologies.

Our findings also demonstrate that TMAO is a powerful metabolite that plays an essential role in the association between microbiome and stroke and may serve as an actionable target. As there is new evidence that polygenic risk scores can be used to augment stroke subtypes (7), microbiome composition can also be an additional layer of information that can be used to better inform care providers when planning individualized care paths for their patients to improve outcome. Taken together, the gut microbiome plays an important role in stroke and can transform how stroke and its subtypes are diagnosed and treated.

Gut dysbiosis and gut metabolite TMAO, involved in different stages of stroke, may contribute to stroke development and associated outcomes. These findings demonstrate the importance of maintaining a healthy gut by preserving the normal balance of commensal and pathogenic bacteria and optimal TMAO levels to reduce the risk of stroke incidence. Implementation of personalized solutions focused on restoring microbial communities and lowering TMAO through the diet may improve the quality of life for individuals living with morbid conditions. Dietary interventions can change gut microbial composition from a disease state to a homeostatic state.

Prebiotics are dietary fiber-derived, non-digestible carbohydrates that selectively stimulate the growth or activity of certain colon-resident bacteria (80). For instance, prebiotics can enhance the growth of anti-inflammatory bacteria Lactobacillus and Bifidobacterium (81). Examples of prebiotics include inulin, human milk oligosaccharides, fructo-oligosaccharides, galacto-oligosaccharides (82). Sources of prebiotics include soybeans, raw oats, unrefined wheat, and unrefined barley (83). Diets rich in prebiotics can shift the microbiome composition toward increased commensal bacteria, thereby promoting host health.

While prebiotics enhances intrinsic microbial communities, probiotics are exogenous live microorganisms, taken orally, that confer health and benefit to the host via modulation of intestinal microflora and inhibition of pathogen colonization (84). Probiotics can restore the proper balance of gut microbiota, exert metabolic effects, and enhance immunomodulatory pathways (85). Probiotics belonging to the genera Lactobacillus, Bifidobacterium, Lactococcus, Streptococcus, and Enterococcus (85), are found in fermented foods such as cheese, kefir, bushera, salami, kimchi, sauerkraut, olives (86). Incorporating prebiotics, probiotics, or both (known as synbiotic) into a personalized diet may attenuate the development of pathogen-dominant microbiome-associated diseases, especially in patients with bacterial translocation or inflammation. For instance, the probiotic species Bifidobacterium animalis can potentially aid in the reversal of bacteremia and lower inflammation in type 2 diabetes, a common risk factor for stroke (87).

The gut microbiome exhibits inter-individual variability in age, sex, ethnicity, and disease states (88); therefore, similar dietary recommendations given to different individuals may not effectively improve their dysbiosis. Care pathways could be designed for individual variations when incorporating personalized nutrition therapies with specific probiotic strains or prebiotic consumption.

In addition to diets focused on restoring gut microbiota as a whole system, personalized nutrition therapies could specifically target and lower TMAO levels, which are often linked to numerous disease states. Studies have reported that dietary precursors can predict TMAO levels (89) and precursor phosphatidylcholine can heighten TMAO levels (74). Increased precursor choline or betaine are further associated with increased risk for major adverse cardiac events (90). These findings suggest that decreased consumption of TMAO precursors may lower TMAO levels and may improve its associated conditions. TMAO precursors are found in red meat and egg yolk. Starting a dietary approach to stop hypertension (DASH) (91), vegetarian diet, or Mediterranean diet may reduce TMAO, especially for patients with high plasma TMAO levels (92). As a long-term diet is associated with both the metabolome and microbiome (93), persistent dietary changes, focused on improving gut health, may reduce stroke risk and promote secondary prevention. Personalized nutrition strategies account for the personalized microbiota response and personalized host response to a specific diet or nutrients.

Machine learning-based personalized dietary interventions, developed based on the patient's microbiome composition, clinical parameters, genetics, lifestyle patterns, and personal goals, can be designed to impact the microbiome, metabolome, and gut physiology (94–97). Patient TMAO levels can be integrated into the model to tailor a diet focused on lowering TMAO and mitigating the risk of developing related conditions. Quantification of various biomarkers such as choline, carnitine, betaine, trimethylamine, and TMAO in the clinical setting can be a convenient and effective way to identify actionable risk factors for stroke recurrence or poor outcome and facilitate the development of technologies with quantitative features for targeted interventions. A well-designed and validated model can analyze patient profile, including patient's altered gut microbes, increased TMAO levels, systemic inflammatory markers, medication history, and traditional stroke risk factors obtained from the electronic health record, and estimate the risk of recurrence or poor outcome. Such models can also integrate other non-clinical variables such as nutrition patterns for a more precise and personalized assessment and recommendations (95, 98).

The gut microbiome contributes to ischemic stroke progression and includes large vessel subtypes and small vessel pathophysiology. Our findings in the present study are considered preliminary. Lack of standardization between articles in an interdisciplinary field limits applicability to the clinical setting. Factors such as lifestyle, physiologic traits, alcohol consumption, and bowel movements are confounding variables that should be considered in future studies. Vujkovic-Cvijin and colleagues provide a list of important confounding factors to consider when designing such studies on this topic (99). One of the limitations in our present study is the inclusion of articles with small sample sizes, which warrant additional large-scale investigations. Reports with large sample sizes included genomic databases while articles with small sample sizes included investigations performed in hospital or university settings. Further, the diversity and composition of gut microbiota lead to complex network analysis, and articles in our present study primarily examined increases or decreases in microbiota composition. Studies included here did not incorporate the concept of functional redundancy-induced stability of gut microbiota due to the multiple functioning capacity of different microbes. Future studies need to examine functional changes rather than compositional changes of the gut microbes to demonstrate how the microbes induce changes within the human body. Current biochemical knowledge on microbiome's effects on stroke was extrapolated from animal studies that were similar but not specific to gut influences on stroke. Further, our review yields a low number of articles on hemorrhagic stroke compared to ischemic stroke; the latter could be in part since hemorrhagic stroke is associated with higher morbidity, mortality, and a lower prevalence than ischemic stroke. Additional investigations are needed to explore the gut microbiome alterations that occur during a hemorrhagic stroke. Some of the articles included in our present study were specific to particular ethnic populations, such as Caucasians and Asians. The latter may provide insight into how genetics or environment (nutrition and physical activity) may differentially impact gut microbiomes across ethnicities, which is important for better patient representation and addressing health disparity, and the development of personalized therapies. Limitations in our review process include selecting published articles in English and using only one database (Pubmed).

In conclusion, our findings support the involvement of the gut microbial landscape and key metabolite TMAO in stroke risk factors, occurrence and recurrence of ischemic stroke, and functional outcomes after stroke. Future human subject studies are needed to elucidate specific functions and underlying mechanisms of gut microbes that play a role in the pathophysiology of stroke. Translating this knowledge into clinical practice can aid care providers in delivering personalized and preventative care to stroke patients efficiently.

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

VA and RZ: conception of the work. ViS, VaS, VA, and AK: data collection. VA, RZ, ViS, VaS, AK, DC, SS, JL, and DW: data analysis and interpretation. ViS, VaS, and VA: drafting the article. ViS, VaS, DW, JL, SS, DC, RZ, and VA: critical revision of the article. All authors approved the final version of the article.

Unrelated to this study, VA had financial research support from the National Institute of Health (NIH) grant No. R56HL116832 sub-awarded to Geisinger during the study period. RZ had financial research support from Bucknell University Initiative Program, Roche—Genentech Biotechnology Company, the Geisinger Health Plan Quality fund, and receives institutional support from Geisinger Health System during the study period. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.