A literature search was conducted to identify all existing articles, up to August 2021, that were related to the role of ketogenic diet on the gut microbiota in the context of neurodegeneration. The search terms “Ketogenic diet,” “High fat diet,” “Gut microbiome,” “Gut microbiota,” “Neurodegeneration” and other synonym terms to the previous terms were used in three electronic databases; PubMed, Ovid MEDLINE, and SCOPUS. A title, abstract and keyword search was performed using the search terms. The Boolean operator “AND” was used to link the terms together on all databases. Articles were first screened through their titles and abstracts before proceeding with full text screening of relevant articles.

The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guideline was employed during the literature selection (Moher et al., 2015). The following inclusion criteria were used during the selection process; (1) peer-reviewed original research articles investigating the interplay between ketogenic diet, gut microbiome/microbiota and neurodegeneration and (2) English articles with full text available. The exclusion criteria were; (1) non-original research articles labeled as editorials, symposiums, conference papers, commentaries, book chapters, case reports, systematic reviews, and reviews, (2) duplicated articles, (3) articles that were not in the English language, and (4) articles that did not investigate ketogenic diet effects on gut microbiota in the context of neurodegeneration.

Different tools were employed to evaluate the quality of the selected relevant articles. The Quality Assessment Tool for Quantitative Studies by the Effective Public Health Practice Project (EPHPP) (Project, 1998) was used to assess the clinical studies. The Systematic Review Center for Laboratory Animal Experimentation Risk of Bias (SYRCLE RoB tool) was used to assess the quality of the preclinical animal studies. The literature search, selection and quality analysis of the selected articles were performed by two independent researchers.

There were no significant differences seen between the baseline alpha and beta diversity of the gut microbiota among the MCI participants compared to cognitively normal participants (Nagpal et al., 2019, 2020). On the other hand, MS participants depicted an overall significant decline in bacterial diversity by 36% and a bacterial mean total concentration by 24% compared to the healthy subject (Swidsinski et al., 2017). In addition, the data collected by Swidsinski et al. connotated that patients with multiple sclerosis have an underlying colonic dysbiosis due to the diminished biodiversity and concentration of the essential bacterial groups, such as Faecalibacterirum prausnitzii (Fprau) (Swidsinski et al., 2017). Fprau has been categorized as an anti-inflammatory microbe due to its capacity to generate butyrate (Bhargava and Mowry, 2014; Cantarel et al., 2015), which may play a central role in sustaining balance and maintaining the optimal ratio of regulatory T cells (T-reg) and T helper 17 cells (Th17) (Bhargava and Mowry, 2014; Cantarel et al., 2015). Hence, disruption to this balance may, in turn, induce inflammation (Bhargava and Mowry, 2014; Cantarel et al., 2015), thereby possibly leading to aggravation of the neurodegeneration condition.

Taxonomical evaluation of the gut microbiota illuminated a predominance of phyla Proteobacteria and Firmicutes amongst patients with underlying ND. There was a significant elevation in the phyla Proteobacteria in fecal samples of subjects with MCI in which the bacteria positively correlated with an AD-biomarker, Aß42/Aß40 ratio (Nagpal et al., 2019). Furthermore, family Enterobacteriaceae (phylum Proteobacteria) was also significantly elevated among MCI subjects and correlated positively with AD-biomarker tau-p181 and tau p-181/ Aß42 ratio (Nagpal et al., 2019). In Alzheimer's disease, high abundance of Enterobacteriaceae was also associated with worsening of disease progression and decline in cognitive functioning. Notably, patients that harbored a high proportion of Proteobacteria had significantly low Mini Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) (Liu et al., 2019).

Similarly, phyla Firmicutes was also elevated amongst those with MCI at baseline and positively correlated with AD-biomarker tau p-181 (Nagpal et al., 2019). Patients with MS were found to have an elevated abundance trend of rRNA bacteria Cvir1414 (Clostridium viridae group), Ehal (Eubacterium hallii), Ecyl387 (Eubacterium cylindroides), Lab158 (Lactobacillus, Enterococcus sp.), and Rfla729 (Ruminococcus albus) upon colonic microbiota analysis via fluorescence in-site hybridization method (FISH), whereby all these bacterial groups belong to the phylum Firmicutes, (Swidsinski et al., 2017) but these elevation was statistically insignificant. In accordance with this, the Sprague Dawley rats also demonstrated a relatively higher abundance of order Clostridales which belongs to the phylum Firmicutes in the AD group (Park et al., 2020).

Two studies depicted that a ketogenic diet had resulted in a decrease in bacterial species diversity. For example, the rats within the AD-ketogenic diet group depicted a significant decline in bacterial species and a decrease in the Shannon Index (diversity) (Park et al., 2020). The finding by Park et al. aligns with an earlier study by Ma et al. in which introduction of a ketogenic diet regime to healthy mice resulted in decreased species diversity (Ma et al., 2018). A similar pattern of enrichment of gut bacterial α-diversity was observed in mice treated with etomoxir, as well as in mice with genetically modified lower CPT1 activity known as SOD1^Cpt1a/Cpt1a. However, when these mice were assigned a 60% high fat diet, they showcased a drop in the α-diversity. Hence, it is believed that the downregulation of CPT1 in ND models by ketogenic diet may decrease the gut microbiota diversity, which may attenuate disease progression. Furthermore, a loss of α-diversity was also seen after the commencement of high-fat diet (Bruce-Keller et al., 2015; Higarza et al., 2019), where a diminished proportion of Firmicutes, and an augmented proportion of Bacteroidetes and Proteobacteria was observed in the high-fat diet group (Higarza et al., 2019). Additionally, a decrease in family Ruminococcaceae, Lactobacillaceae, genus Lactobacillus, Clostridium cluster IV, Bifidobacterium, and an increase in family Peptostreptococcaceae, Clostridiaceae, Enterobacteriaceae, Desulfovibrionaceae, Bacteroidaceae, genus Bacteroides, Blautia, Rombustia, Bilophila, Escheria-Shigella were all identified in the high-fat diet group.

Similarly, another study showed an increase in phylum Bacteroidetes, genera Bacteroides, Dorea, Roseburia, Bilophila and a fall in size of phyla Actinobacteria, Cyanobacteria, Tenericutes, genera Bifidobacterium, Odoribacter, Dehalobacterium, Turibacter, after high-fat diet (Bruce-Keller et al., 2015; Wu et al., 2020). However, no specific pattern of changes in gut microbiota corresponding to the duration of dietary interventions were reported in these studies. In addition, differential abundance testing revealed a significant drop in harmful bacterial genera Rikenellaceae, Akkermansiaceae, Gastranaerophilales, and ⋎-proteobacteria in the group treated with etomoxir. Remarkably, class ⋎-proteobacteria and family Rikenellaceae have been widely held as pro-inflammatory bacteria, with the latter being linked to disruption of intestinal epithelial integrity and exacerbation of colonic inflammation via toll-like receptor (TLR)4 signaling pathway (Liu et al., 2019).

There is a pronounced increase in beneficial bacterial genera Lachnospiraceae and Odoribacter in the SOD1^Cpt1a/Cpt1a mice group, as compared to their respective control counterparts. Odoribacter, has been known to generate lower levels of the pro-inflammatory cytokine, tumor necrosis factor-alpha (TNF-α) (Trabjerg et al., 2021). This increase in beneficial bacterial composition was also seen with the genus Akkermansia, where after ketogenic diet intervention, the genus Akkermansia was elevated in MCI subjects (Nagpal et al., 2019) and even in healthy mice (Ma et al., 2018). It was also shown that in a high-fat diet animal model, the cognitive decay was halted, and spatial memory was improved by increasing the kynurenic acid production with the help of Akkermansia (Wu et al., 2020). Akkermansia was also reported to relieve systemic inflammation by decreasing the concentration of pro-inflammatory markers such as IL-10 and increasing anti-inflammatory markers (Wu et al., 2020). Furthermore, Akkenmansia was also depicted to restore gastrointestinal ecosystem by increasing the concentration of genus Bifidobacterium, neuroprotective bacteria, and decreasing the neurotoxic genera Bilophila and Bacteroidetes (Wu et al., 2020). In addition, Akkermansia may also produce beneficial SCFA, predominantly acetate and propionate, which was shown to be vital for gut and neuronal communication and homeostasis (Wu et al., 2020). Taken together, this evidence suggest that ketogenic diet not only reduced harmful bacterial composition, but also promoted the flourishment of beneficial bacterial species such as Akkermansia. In addition, harmful bacteria such as genus Lachnobacterium (of phylum Firmicutes) was found to be downregulated in patients with MCI after ketogenic diet intervention (Nagpal et al., 2019). Similarly, prominent down-regulation of pro-inflammatory bacteria genera Desulfovibrio and Turicibacter were also observed in mice on ketogenic diet (Ma et al., 2018). In fact, the concentration of Desulfovibrio was undetectable, whereas Turicibacter was two times lower in concentration compared to the control mice. Likewise, positive gut microbiota modulation was seen with ketogenic diet intervention, whereby the bacterial composition in patients with MS, was restored and almost coincided with the healthy control subjects after the diet (Swidsinski et al., 2017).

Mice that were on the ketogenic diet demonstrated enhanced abundance of the beneficial bacteria genus Lactobacilli by 3.2 times (Ma et al., 2018). The increase in Lactobacilli and the decline in Desulfovibrio by ketogenic diet provided a shift within the internal gut environment to an anti-inflammatory state (Solas et al., 2017; Leigh and Morris, 2020). In support, the administration of probiotics containing Lactobacillus demonstrated restoration of cognitive impairment, whereas feeding of chow containing Desulfovibrio was shown to perpetuate the cognitive decline (Leigh and Morris, 2020), thus suggesting the positive impact of ketogenic diet on these two bacterial genera and possibly on neurodegeneration as well. The data collected by Swidsinski et al. connotated that patients with MS have an underlying colonic dysbiosis due to the diminished biodiversity and concentration of the essential bacterial groups, such as Faecalibacterirum prausnitzii (Fprau) (Swidsinski et al., 2017). Fprau has been categorized as an anti-inflammatory microbe due to its capacity to generate butyrate (Bhargava and Mowry, 2014; Cantarel et al., 2015) which may play a central role in sustaining balance and maintaining the optimal ratio of regulatory T cells (T-reg) and T helper 17 cells (Th17). Hence, disruption to this balance may, in turn, induce inflammation, thereby possibly leading to aggravation of the neurodegeneration condition (Bhargava and Mowry, 2014; Cantarel et al., 2015).

Ketogenic diet has the ability to diminish the oxidative stress in neurodegenerative diseases, particularly MS, by increasing the levels of glutathione and catathione, which are enzymes that may exert antioxidant properties and counteract the increasing levels of reactive oxygen and nitrogen species, respectively (Tobore, 2021). Mice on the ketogenic diet intervention have also demonstrated enhanced global cerebral perfusion, whereby regional blood flow to the ventromedial hypothalamus (VMH) was significantly increased by 11.82% (Ma et al., 2018). In addition, Aβ clearance was also augmented in the mice under ketogenic diet (Ma et al., 2018). From a molecular perspective, the ketogenic diet reduced the production of mTOR protein by dampening its gene expression by 29.9%. Inhibition of mTOR protein expression may aid in delaying cognitive deterioration, as hyperactivation of this protein has been associated with memory impairment (Jahrling and Laberge, 2015). Besides that, endothelial nitric oxide synthase (eNOS) levels were also found to be significantly boosted by 111.5% increase in the ketogenic diet mice (Ma et al., 2018). eNOS has been seen as a potent vasodilator, and hence increased circulating levels of eNOS by ketogenic diet may promote cerebral perfusion especially to the hypothalamus (Liao et al., 2021). A trial comprising of eNOS deficient mice models depicted pathological signs suggestive of the neurodegenerative process such as loss of white matter mass leading to cortical atrophy, along with decreased concentration of oligodendrocytes, resulting in lack of myelin sheaths (Liao et al., 2021). Hence, this elucidated the vital function of eNOS in maintaining a healthy CNS environment. In addition, the expression of P-glycoprotein, an essential protein that may aid Aβ clearance from the brain, was also significantly elevated by 50% in ketogenic diet mice and its activity was promoted by 185.38% (Ma et al., 2018). The study by Park et al. further elucidated that integration of ketogenic diet as an intermittent diet schedule have proven to be more favorable, where a decline in Firmicutes and an augmentation in Akt phosphorylation were observed (Park et al., 2020). Upstream signaling of the Akt molecule causes the activation of a cascade of molecular processes, which ultimately may lead to dampening of the AD pathogenesis, such as decreased Aβ deposition and formation of neurofibrillary tangles (Long et al., 2021).

Some studies have suggested that ketogenic diet exerts a negative effect on the gut microbiota. The conflicting results indicate that ketogenic diet as a therapy for gut dysbiosis in neurodegeneration may warrant further investigations, particularly in a clinical setting, where influences of environmental factors and inter-species interactions in the intestinal flora may govern the therapeutic effects of ketogenic diet. A surge in abundance of family Streptococcaceae and Desulfovibrionaceae, which have been known for their pro-inflammatory and endotoxin-producing properties, as well as a decrease in the family Lachnospiraceae have been reported after a high-fat diet regime suggesting a inflammatory mediating effect of high-fat diet (Li et al., 2021; Trabjerg et al., 2021). Additionally, signs of gut dysbiosis have been observed in the AD rat group after ketogenic intervention, where there was an increase in the order Enterobacteriales (phylum Proteobacteria) (Park et al., 2020). Furthermore, the ketogenic dietary pattern also resulted in a further increase in Enterobacteriaceae from baseline among MCI subjects (Nagpal et al., 2019). Likewise, diminutions of the family Bifidobacteriacea (of phylum Actinobacteria) upon ketogenic diet in MCI patients also depicted an unfavorable outcome of this diet pattern in the neurodegeneration context (Nagpal et al., 2019). In addition, within the MS population, a decline in levels of Akkermansia was observed after six months of ketogenic intervention (Swidsinski et al., 2017), unlike the positive increase seen in MCI subjects (Nagpal et al., 2019).

Similarly, in an ALS transgenic animal model receiving high fat diet, worsening of neurological score, earlier onset of tremor, weaker grip strength, reduced overall activity level, and lesser right turns in Y-maze test were reported further exemplifying the diet's possible deleterious consequences in ALS (Trabjerg et al., 2021). Correspondingly, from the pathological aspect, a heightened state of neuroinflammation was also shown by the rise in astrocyte reactivity marker, microglial marker (Cx3cr1 and Tmem119), TLR 2&4, and the decrease in ZO-1 and claudin 5, which are indicative of compromised neurovascular integrity, in subjects receiving the high fat diet (Bruce-Keller et al., 2015; Reilly et al., 2020; Zeltser et al., 2020). High fat diet was also found to render intestinal tight junction more permeable (lower jejunum and colonic occluding, increased DAO), thereby inducing a state of systemic inflammation, exemplified by increased colonic NOS, phosphorylation of p65 subunit of NFKD, plasma endotoxin and the pro-inflammatory IL-1β (Bruce-Keller et al., 2015; Trabjerg et al., 2021). In fact, Wu et al. also showed that a high-fat diet promoted memory decay and depression in mice, suggesting that a high-fat diet may indeed be more harmful than beneficial (Wu et al., 2020). Besides that, the contrasting behavioral and pathological effects seen between ketogenic diet intervention and high fat diet intervention may be related to the state of carbohydrate deprivation, which is higher in ketogenic diet than high fat diet. Thus, based on the evidence accumulated in this systematic review, ketogenic diet may be more beneficial toward neurodegenerative diseases than high fat diet, especially when integrated as an intermittent diet schedule, but the duration of the diet should be monitored closely. There is a substantial amount of evidence that proposes the positive molecular modulation of ketogenic diet in a healthy environment, but whether these positive modulations may persist in a neurodegeneration model may warrant further investigations.