Examination of shared gut microbiome signatures in aging and Parkinson’s disease

Parkinson’s disease (PD) is a common age-related neurodegenerative disorder whose prevalence is increasing rapidly in tandem with global population aging. The number of diagnosed PD patients has doubled over the past 3 decades and is expected to double again by year 2040 (Dorsey et al., 2023). Pathologically, PD is characterized by the loss of dopaminergic neurons (Langley et al., 2020) in the substantia nigra pars compacta (SNpc) of the midbrain that results in classical motor symptoms such as bradykinesia, rigidity, and resting tremors, which are behavioral markers for clinical diagnosis (Postuma et al., 2015; Moustafa et al., 2016). However, patients typically experience a prodromal period that is characterized by non-motor symptoms such as depression, constipation, REM sleep disorder, and olfactory loss (Haehner et al., 2011; Postuma et al., 2015; Liu et al., 2017) before the overt onset of movement deficits. The disease progression and symptom severity are clinically tracked using Hoehn and Yahr (H&Y) staging (Zhao et al., 2010) or the Unified Parkinson’s Disease Rating Scale (UPDRS) including the Movement Disorder Society (MDS) UPDRS (Movement Disorder Society Task Force on Rating Scales for Parkinson’s Disease, 2003; Goetz et al., 2008), with increasing score indicating worsening PD disability for both assessments (Holden et al., 2018).

The histopathological hallmark of PD is the accumulation of α-synuclein, a neuronal protein, into aggregates called Lewy bodies in the SNpc (Mackenzie, 2001). In 2003, the German pathologist Heiko Braak postulated that sporadic PD could develop from the spreading of these α-synuclein aggregates starting either from the nasal cavity or the gut before they infiltrate the brain (Rietdijk et al., 2017; Borghammer and Van Den Berge, 2019; Van Den Berge and Ulusoy, 2022). Supporting Braak’s hypothesis, there is now accumulating evidence suggesting an important role of the gut-brain axis in PD risk and progression (Tan et al., 2022). Indeed, studies have demonstrated the spreading of α-synuclein pathology from the gut to the brain via the vagus nerve connecting the two organs (Kim S. et al., 2019). Additionally, gut microbiota from PD patients worsens physical impairments when colonized into a PD mouse model (Sampson et al., 2016). Thus, the gut microbiome could play a role in PD pathogenesis.

Generally, PD risk factors include genetic predisposition (e.g., Pink1, Parkin, LRRK2 mutations) (Billingsley et al., 2018), head trauma (Bower et al., 2003) and exposure to environmental neurotoxicants such as pesticides (Kamel et al., 2007). However, exposure to such environmental factors are rather selective to certain populations, and genetics only account for 3–5% of all PD patients (Alafifi et al., 2020). The remaining 95% of PD cases are sporadic in nature, where aging remains the biggest contributing risk factor (Collier et al., 2011). Loss of neurons is common with aging; however, dopaminergic neurons in the SNpc are observed to be preferentially vulnerable to degeneration at a rate much higher than other neurons in the brain (Reeve et al., 2014). One study with 750 elderly (non-PD) participants showed that at least one-third of the study population displayed mild to severe neuronal loss in the SNpc, and 10% exhibited characteristic Lewy body pathology post-mortem (Buchman et al., 2012). Such vulnerability can be attributed to the accumulation of oxidative stress, DNA damage, dysfunctional mitochondria and protein aggregates that are a result of deteriorating cellular maintenance that comes with aging (Reeve et al., 2014). It also emphasizes the intimate relationship between aging and PD.

Studied have identified the gut microbiome as one factor that could be associated with the onset and progression of PD. The known relationship between the gut microbiome and PD is complex; studies suggesting that this relationship is casual (i.e., gut microbiome accelerates PD development) and studies proposing that the relationship is consequential both exist in current literature. As alluded to earlier, Sampson et al. (2016) published a seminal study highlighting the evidence for a causal link; here gut microbiota collected from PD patients sufficiently aggravated motor deficits when colonized into a PD mouse model (Sampson et al., 2016). Additional studies support this causal relationship (Ning et al., 2022; Zhu et al., 2022; Chen et al., 2025). On the other hand, different studies show that altered gut microbiome is a consequence of PD; medications for PD such as L-Dopa (Maini Rekdal et al., 2019; Zhong et al., 2023) and prodromal PD symptoms such as constipation (Yang et al., 2022) have been shown to change the gut microbiome. Various studies advocating for the consequential relationship show that manipulating the gut microbiome has minimal downstream effect on PD development (Scheperjans et al., 2024; De Sciscio et al., 2025). It should be noted that despite the exact nature of the relationship between gut microbiome and PD (casual or consequential) remaining uncertain, the association between the gut microbiome and PD has been shown to be robust. Recognizing this, metagenomic sequencing studies related to PD gut microbiome have been conducted across various ethnic groups and/or countries (Table 1). Although many of these studies have addressed confounding factors, such as dietary pattern, geography, medicine-use or comorbidities, the confounding effect of age is of particular relevance given its a major risk factor for PD and as well as a key determinant of gut microbiome composition (Hindle, 2010). Given this, we seek to elucidate a working model of the gut microbiome’s relationship that considers both aging and PD. Additionally, since dietary variations across different ethnic groups can affect the composition of gut microbiome (Leeming et al., 2019), we also examined the presence of unique subsets of PD gut microbiota that are linked with ethnicity.

We examined studies that employed 16S rRNA sequencing given the larger number of microbiome studies in PD and aging (Janda and Abbott, 2007). To get insights into recent developments in the field, papers on 16S rRNA sequencing of fecal samples from PD patients were gathered via PubMed in a publishing window of 2017–2022. The search term “parkinson’s + gut microbiome” was used. This yielded 689 results. Only original articles and meta-analyses that reported differentially abundant gut microbiome in human samples were selected. Review papers were examined for the original research articles that were cited. A total of 35 papers were examined for this review (Table 1; Bedarf et al., 2017; Hill-Burns et al., 2017; Hopfner et al., 2017; Li et al., 2017; Petrov et al., 2017; Heintz-Buschart et al., 2018; Lin et al., 2018; Qian et al., 2018; Aho et al., 2019; Barichella et al., 2019; Bedarf et al., 2019; Jin et al., 2019; Li C. et al., 2019; Li F. et al., 2019; Lin et al., 2019; Pietrucci et al., 2019; Weis et al., 2019; Baldini et al., 2020; Cirstea et al., 2020; Elfil et al., 2020; Nishiwaki et al., 2020; Ren et al., 2020; Wallen et al., 2020; Zhang et al., 2020; Li P. et al., 2021; Romano et al., 2021; Shen et al., 2021; Tan et al., 2021; Boertien et al., 2022; Jo et al., 2022; Li et al., 2022; Lubomski et al., 2022a; Romo-Vaquero et al., 2022; Toh et al., 2022; Zhang K. et al., 2022). It is interesting to note that the number of papers rose rapidly from just 48 papers in 2017 to over 200 papers per year in the last 3 years; this reinforces how PD research focus has increasingly embraced the gut-brain axis. A similar search process was done for aging-related studies using the search term “aging + gut microbiome” or “centenarian + gut microbiome.” Only original articles that studied gut microbiome differences between healthy young and healthy centenarians were selected, and those confounded with other experimental variables such as medication or supplement treatments were omitted. A total of 11 aging-related papers were examined for this review (Table 2; Maffei et al., 2017; Kim B. et al., 2019; Tuikhar et al., 2019; Wang et al., 2019; Wu et al., 2019; Badal et al., 2020; Palmas et al., 2022; Sepp et al., 2022; Wang J. et al., 2022; Wu J. et al., 2022; Wu L. et al., 2022). Additionally, we reviewed literature from the past 5 years (2021–2025) on the key PD-associated metabolite butyrate. A search using the terms “Parkinson’s” and “butyrate” yielded 67 publications, suggesting increasing interest in this metabolite.

Every paper recorded is given a unique key, P_x (where x is a numeric), for PD in Table 1, or AG_x, for aging in Table 2. These unique keys serve as headers for each list of microbiomes reported in Supplementary Tables 1, 2 for mapping of the data to the specific papers. The microbes’ identities were collated exactly as they were reported in the respective papers. The microbiomes reported were sorted and recorded according to whether there was an increase (Supplementary Table 1) or decrease (Supplementary Table 2) in abundance for each of these papers. The intersect (overlapping microbiome trends between PD and aging) and complement (microbiome trends unique to PD or aging) were then examined. Differences in gut microbiome across ethnicity (based on Race column in Table 1) and PD symptom severity (green for mild PD, pink for moderate PD in Table 1) were also studied. PD symptom severity was classified as mild for UPDRS III scores lower or equal to 32, moderate for scores between 33 and 58 and severe for scores above or equal to 59 (Martínez-Martín et al., 2015). Amongst the papers that have information on UPDRS III scores, 13 were mild and 4 were moderate. In addition, a Jaccard Index, represented as J(A,B) = | A ∩ B| /| A ∪ B|, was calculated to represent the similarity between PD and aging datasets. The Jaccard Distance (1-Jaccard Index) was also calculated to represent the dissimilarity between ethnicity datasets. All analyses were done using R (v1.4.1106).

A total of 35 papers on PD gut microbiome and 11 papers on aging-related gut microbiome from the 5-year period between 2017 and 2022 were examined. Of the 35 PD papers, one analyzed appendix samples (Li P. et al., 2021) while the rest of the 34 used fecal samples. The 35 PD papers consist of one review, five meta-analyses, and 29 research articles; they cover patients from ages 52–88, with sample sizes from 10 up to 200 subjects, spanning across 12 different countries that capture both Asian and Western populations. Among the papers that provided a numerical score specifically for UPDRS III PD symptom severity grading, 13 were mild and 3 were moderate.

Similarly, the 11 aging-related papers analyzed fecal samples. These papers consist of one review and 10 articles covering young, elderly, and centenarian groups, with sample sizes of 9 subjects up to 158 subjects, spanning across 7 countries that capture both Asian and Western populations. All the aging studies only examined elderly and centenarians who are healthy and free from medical interventions to prevent confounding effects of other diseases and drugs. Many studies also attempt to compare the young and old from similar geographical proximity (e.g., from the same village) or from the same household to reduce confounding effects from dietary differences.

In the last 5 years (2021–2025), more population studies have emerged identifying decreased butyrate as a biomarker for PD (Yan et al., 2022; Liu et al., 2024; Zhao et al., 2024; Rust et al., 2025). In addition to constipation, studies suggest butyrate plays an early role in other prodromal PD non-motor symptoms such as REM sleep disorder and depression; butyrate-producing bacteria such as Lachnospira, Butyricicoccus, and [Eubacterium]_ventriosum_group where found to be decreased in REM sleep disorder (Huang et al., 2023), and likewise butyrate-producing Roseburia and Romboutsia were observed to be reduced in depression (Xie et al., 2022). Even within the PD group, lower abundance of butyrate producing Butyricimonas synergistica, was associated with worse non-motor symptoms (Nuzum et al., 2023). Other studies have suggested that the reduction of fecal butyrate correlated with clinical severity of PD (Chen et al., 2022), such as worse postural instability-gait disorder scores (Tan et al., 2021). Interestingly, in a human A53T α-synuclein transgenic mouse model, enteric α-synuclein expression was shown to decrease fecal butyrate levels (Pellegrini et al., 2022). Furthermore, in vitro fermentation experiments using fecal samples from PD patients and age-matched healthy controls demonstrated that, although butyrate production can be stimulated in PD patients, overall butyrate production rate remain significantly lower (Baert et al., 2021). Taken together, these findings suggest that PD-initiating factors such as α-synuclein may contribute to an early reduction in butyrate levels, which are subsequently maintained at low levels throughout disease progression due to impaired production in PD patients. This highlights butyrate as a promising biomarker for PD.

A natural question that follows is whether changing the gut microbiome and more specifically supplementation of butyrate can then improve PD symptoms. In a recent study, fecal microbiota transplantation (FMT) from healthy human individuals to 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced PD mouse models was able to significantly improve motor function, with the therapeutic effects associated with increased levels of butyrate (Ni et al., 2025). Importantly, supplementation of butyrate alone (in the form of sodium butyrate) was sufficient to reduce motor deficits in an α-synuclein pre-formed fibrils (PFF) mouse model (Kakoty et al., 2021), improve non-motor symptoms like anxiety in 6-hydroxydopamine (6-OHDA) mice (Avagliano et al., 2025), and normalize sleep architecture in MPTP mice (Duan et al., 2025).

Several studies suggest that a key mechanism underlying the beneficial properties of butyrate is through re-shaping the gut microbiota and reducing colonic and intestinal barrier disruption (Avagliano et al., 2022; Xu et al., 2022; Guo et al., 2023; Zhang Y. et al., 2023) across various neurotoxin-induced mouse models of PD involving rotenone (Zhang Y. et al., 2023), 6-OHDA (Avagliano et al., 2022) or MPTP (Xu et al., 2022; Guo et al., 2023). These studies also reported another key mechanism of reduced gut and brain inflammation. Supporting this, sodium butyrate was shown to be able to suppress MPP+ activation of BV2 microglia cells and reduce the production of nitrite and pro-inflammatory cytokines (Xu et al., 2024). In a separate study, butyrate attenuated 6-OHDA-induced dopaminergic neuronal injury via inhibiting microglia activation and neuroinflammatory factors production (Xu et al., 2025). Other pathways suggested includes PGC1α-autophagy activation to reduce rotenone-induced α-synuclein accumulation and aggregation (Zhang Y. et al., 2022), inhibition of pro-inflammatory pathways involving JAK2/STAT3 signaling (Ji et al., 2023), NF-κB and MAPK signaling in the SNpc (Hou et al., 2022), and protection against Fe and Mn toxicities (Tizabi et al., 2023). Additionally, there are also suggestions to supplement L-Dopa treatment with butyrate as HNK-butyrate esters have been shown to inhibit E. faecalis growth in the gut, thus increasing the amount of unmetabolized L-Dopa that can reach the brain for therapeutic effects (Cheng et al., 2025).

While butyrate shows therapeutic promise, its efficient delivery remains an important point of consideration. As with other SCFAs, butyrate is easily absorbed in the human small intestine (specifically the jejunum) which limits butyrate bioavailability in the lower gastrointestinal tract such as the colon (Schmitt et al., 1976; Hodgkinson et al., 2023). There are currently several ways of butyrate delivery that aim to overcome this issue. The first technique is to provide butyrate as a triglyceride. Butyrate triglyceride is composed of 3 butyric acid molecules connected to a glycerol backbone; importantly, this conformation prevents premature absorption of butyrate in the small intestine (Bedford and Gong, 2018). Based on this principle, a retrospective clinical study examining existing medical records of PD patients showed that combination probiotics supplementation of butyrate triglyceride (302.86 mg) with Crocus sativus L. (30 mg) and vitamin D3 (100 mcg) was sufficient to improve UPDRS III scores (Alexoudi et al., 2023). Secondly, a similar technique of butyrate delivery involves the use of microencapsulated sodium butyrate; this colonic-release butyrate capsules have been shown to be effective in alleviating symptoms of irritable bowel syndrome (IBS) (Banasiewicz et al., 2013), ulcerative colitis (UC) (Vernero et al., 2020), and symptomatic uncomplicated diverticular disease (SUDD) (Tursi et al., 2025). Thirdly, butyrate levels could be improved by administering prebiotics such as resistant starch. Prebiotic resistant starch is able to pass through the small intestine intact and reach the colon where it can then aid butyrate production (through microbial fermentation) (Dobranowski and Stintzi, 2021; Hodgkinson et al., 2023). Recently, a clinical trial aimed at altering fecal SCFAs used an 8-week resistant starch prebiotic intervention to significantly increase fecal butyrate concentrations (Becker et al., 2022) in PD patients. Lastly, FMT could be used to deliver healthy butyrate-producing bacteria directly to the colon. A 2017 randomized controlled trial showed that UC patients who responded clinically to FMT treatment had more butyrate-producing bacteria post-FMT (Fuentes et al., 2017). In the context of PD however, FMT has been less successful in improving clinical outcomes in patients (Scheperjans et al., 2024). Nevertheless, FMT may perform better in the case of prodromal PD. By the time a PD patient is clinically diagnosed more than 50% of nigrostriatal dopaminergic neurons have already been lost (DeMaagd and Philip, 2015); this could potentially limit how effective FMT could be at this clinical stage. Therefore, the timing of FMT intervention may play an important role in overall success. All in all, butyrate shows promise as a biomarker and a feasible therapeutic target for PD.

Current studies tend to examine aging and Parkinson’s disease separately. In this review, we have examined gut microbiome changes that relate to both aging and PD. We have also examined the gut microbiome differences that are related to ethnicity and identified several microbes that were ethnically specific: Alistipes and Odoribacter for Asians, Lactobacillus and Roseburia for Western populations. Importantly, we observed that the gut microbiome seemed to converge by exerting its effects through butyrate. Notably, the major producers of butyrate—Faecalibacterium and Ruminococcus—decreased in abundance across both Asian and Western populations as well as with age. One possible reason for this observation could be the beneficial roles played by butyrate as a histone deacetylation inhibitor and an important contributor to intestinal barrier and gut permeability. Our findings are reaffirmed by a recently published metagenomics by Wallen et al. (2022) on the largest PD cohort of microbiome data, which determined that the abundance of bacteria such as Blautia, Faecalibacterium, Fusicatenibacter, Roseburia and Ruminococcus were decreased in PD, while Bifidobacterium and Lactobacillus were increased in PD (Wallen et al., 2022). Additionally, we observed bidirectional changes for Prevotella that was likewise pointed out by Wallen et al and resolved to be generally upregulated in their meta-analysis; the study also discussed the observed increases in Akkermansia in PD that was not detected in the metagenomics (Wallen et al., 2022). Many of these microbes along with butyrate are altered with age suggesting that these may be early targets for preventive measures against PD.

In recent years, several papers have stratified their PD cohorts to examine the effect of medication as a confounder on gut microbiome (Hill-Burns et al., 2017; Palacios et al., 2021; Lubomski et al., 2022b; Gorecka-Mazur et al., 2024). Only patients who were on Levodopa-Carbidopa and Deep Brain Stimulation (DBS) (Palacios et al., 2021; Lubomski et al., 2022b) reflected changes in gut microbiome but not those on COMT inhibitors (Hill-Burns et al., 2017). Interestingly, DBS and Levodopa affect different microbiomes, with the common downregulated microbes being Hespellia and the common upregulated microbes being Prevotella and Bacillus (Lubomski et al., 2022b). The sensitivity of Prevotella to PD treatments may partly explain the bidirectional changes in this genus reported across studies. Furthermore, levodopa-associated microbiome changes emerged only at 6 months (Lubomski et al., 2022b) and not at the earlier 3-month time point (Palacios et al., 2021), suggesting a delayed or cumulative treatment effect. While Roseburia, a key butyrate-producing genus, was upregulated following prolonged levodopa exposure, most taxa that increased across treatment intervals (including Prevotella, Bacillus, Methanobrevibacter, and Veillonella) are not direct butyrate producers and are more commonly associated with acetate, succinate, propionate production, or methanogenesis. Importantly, these medication-associated microbiome changes were distinct from the core microbial signatures identified as PD- or aging-related, underscoring the need to account for treatment effects when interpreting disease-associated gut microbiome alterations.

This study has several limitations. Firstly, only two major demographic populations (Asian and Western) were included in the analysis which limits how well the findings extrapolate on a global scale, especially since the gut microbiome is sensitive to factors such as dietary pattern and geographical location (Leeming et al., 2019). This limitation is due to the lack of literature studying certain demographics, particularly from developing nations which may lack the required resources, and highlights the need for more thorough demographic representation in the field. Secondly, this review primarily reports on 16S rRNA sequencing data. While this approach allows quantification of bacterial composition, it fails to capture any information on the other microorganisms that make up the gut microbiome (e.g., fungi and viruses) (Bars-Cortina et al., 2024). Additionally, 16S rRNA data does not account for functional activity (Durazzi et al., 2021). Although more costly, newer techniques such as metagenomic sequencing address these limitations while offering better taxonomic resolution (Kuczynski et al., 2012; Durazzi et al., 2021). Thirdly, none of the studies examined in this review reported on the gut microbiome of severely symptomatic PD patients (UPDRS III above 58) who may present different gut microbial signatures than mildly (UPDRS III below 33) and moderately symptomatic PD (UPDRS III between 33 and 58) (Martínez-Martín et al., 2015); therefore, although we show that gut microbes that are associated with both aging and PD did not appear to correlate with mild and moderately symptomatic PD, caution should be used when extrapolating these findings to more symptomatic PD patients.