More than just a number: the gut microbiota and brain function across the extremes of life

ABSTRACT

Understanding the interrelationship between the gut microbiota and host physiology, although still in its relative infancy, has taken important steps forward over the past decade. In the context of brain disorders including those characterized by neurodevelopmental and neurodegenerative changes there have been important advances. However, initially research involved correlational analyses, had limited translational scope, and lacked functional assessments. Thus, largescale longitudinal clinical investigations that assess causation and underlying mechanisms via in depth analysis methods are needed. In neurodegeneration research, strong causal evidence now links the gut microbiome to Alzheimer's (AD), and Parkinson's Disease (PD), as supported by human-to-animal transplantation studies. Longitudinal interventions are being conducted in AD, PD, amyotrophic lateral sclerosis, Huntington's disease, and multiple sclerosis. Neurodevelopmental research has also seen a boon in microbiome-related clinical research including in autism, Attention-deficit/hyperactivity disorder, and schizophrenia, which is confirming prior animal model work regarding the key time-windows in the gut microbiome important for infant cognition. While recent research advances represent important progress, fundamental knowledge gaps and obstacles remain. Knowing how and why the gut microbiome changes at the extremes of life will develop our mechanistic understanding and help build the evidence base as we strive toward counteracting microbial missteps with precision therapeutic interventions.

1. Introduction

Over the course of the past decade, gut microbiome research has continued to provide a greater understanding of the important connections between the gut and the brain. This bidirectional communication network, now conceptually reframed as the microbiota-gut-brain axis,¹ has further benefited from continuing advances in next-generation sequencing technologies,² and the associated development of superior analysis via more advanced bioinformatic pipelines.^3,4 The initial focus in the field was based on characterizing the gut bacteria that were present, be it enriched or depleted (assessing diversity and composition),⁵ how they differed between different populations/regions,^6,7 across aging,⁸ or differences in gut bacterial composition between healthy individuals and in various disease states.^9–12 While this research broadly implicated the relevance of gut bacteria for different disease states or across the lifespan, from early life to older adulthood, this overreliance on compositional and correlational observations has been a common critique of the field.^13–15 These compositional assessments have now been complemented with causal and mechanistic research. For example, shotgun metagenomics is now favored over 16S rRNA sequencing, sampling of metabolic data over solely microbial composition is more common, as are larger sample sizes, and mechanisms of host microbe dialogue have been interrogated via interventional, longitudinal, and translational research to address the question of causality of the gut microbiome in neurodevelopment, healthy aging, and neurodegeneration (Figure 1).

Figure 1.

Descriptions of various modalities to assess causality in gut microbiome research, including (a) FMT in humans and animal research,^16–19 (b) antibiotics to ‘knock-out’ the gut microbiome and (c) using specific pathogen/germ free animals.²⁰ Various interventions modulate the gut microbiome, including (d) diet and/or fermented foods,²¹ and exercise,²² or (e) probiotic, prebiotic (“a substrate that is selectively utilized by host microorganisms conferring a health benefit”²³) or synbiotics (“a mixture comprising live microorganisms and substrate(s) selectively utilized by host microorganisms that confers a health benefit on the host”.²⁴) New and emerging analysis techniques include (f) assessing circadian rhythms of the gut microbiome^25,26 (g) along with field guides for gut microbiome analysis ^3,4 and methods for assessing functionality of the microbiota-gut-brain axis,²⁷ and reporting guidelines²⁸ to create consistency across the gut microbiota field. NB: current evidence supports the use of FMTs for treating C.diff infections, while for any other disease/disorder it is considered an investigational procedure.⁴⁵ Strict guidelines should be adhered to when conducting FMT in clinical contexts as there are some safety concerns and considerations, with some severe adverse events having occurred previously^46,47

These advances have been facilitated by guidelines for consistent and transparent reporting of gut microbiome results and preclinical studies (e.g. STORMS checklist and GRAFT guidelines),^18,28 open access platforms and methods of analyzing metagenomic sequences,²⁹ and accessible guides providing a framework for how to statistically assess the gut microbiome.^3,4 Altogether, these enable researchers to better understand, better plan, and better implement informative research about the gut microbiome. Furthermore, updates to statistical and bioinformatic approaches that focus on functionality,^27,30 have allowed the field to move past ‘who is there’ (composition), and facilitate answering the questions of ‘what are they doing?’ or ‘why are they important?’. Regarding experimental studies, the use of interventions in the form of pre- or probiotics,³¹ fecal microbiota transplants (FMT),¹⁹ or lifestyle factors which are known to modulate the gut microbiome such as diet,³² or exercise,³³ which are also implicated in healthy aging and neurodegeneration,³⁴ have enabled the inference of causal connections. Finally, longitudinal research measuring how the gut microbiota fluctuates over time at an individual level have begun to help answer the ‘chicken-and-egg’ problem of whether the gut microbiome is influencing disease/development, or if disease/development affects the gut microbiome. While the genetic risk(s) for various neurodevelopmental disorders and neurodegenerative diseases will not be focused on presently, it is important to contextualize the potential role of the microbiota-gut-brain axis in relation to genetic risk, and to consider how genetic and environmental risks, in concert with the microbiota-gut-brain axis, may modulate disease states and severity.

In relation to early life, the microbiome has seen much research scrutiny in many neurodevelopmental disorders such as autism spectrum disorder (ASD), attention-deficit hyperactivity disorder (ADHD), and schizophrenia (as a neurodevelopmental mental disorder in the broader definition). Altered maternal gut microbiomes, or disruptions to the infant gut microbiome, have both been linked to an increased risk of developmental disorders.^35,36 While this does not explain why those affected from these disorders have a different microbial profile, some interventional research is beginning to close this research gap. Understanding the gut microbiome changes across neurodevelopment, and how they may be causally related to the etiology of these disorders is yet to be fully elucidated.

A greater understanding of causality and the role of the gut microbiome has also begun to be elucidated in terms of healthy aging and neurodegenerative research. Primarily, this has related to healthy/unhealthy aging involving the investigations of chronological and biological aging, and a large focus on Alzheimer's,^37,38 and Parkinson's disease research.^10,39,40 Less research has been conducted on additional neurodegenerative conditions, e.g., Multiple Sclerosis (MS), Huntington's disease (HD), and Amyotrophic lateral Sclerosis (ALS), however these fields have been expanding in recent years.

In this review, we highlight how the neurodevelopment and neurodegeneration fields have progressed in the last decade focusing both on advances in basic research and outline the clinical evidence linking the gut microbiome to healthy/typical development and aging, along with links to neurodevelopmental disorders and neurodegenerative diseases. The environmental risks, and shared mechanisms of action across neurodevelopmental and neurodegenerative conditions (Figure 3), will be addressed.

Info box 1: Technological, statistical and experimental techniques used in the basic science, pre-clinical and clinical studies of the microbiome. Rather than only looking at the gut microbiome and ‘who is there’, more recent papers have begun to implement ‘multi-omics’ where multiple methods are employed to better understand the mechanisms, such as by Morton et al. who investigated metagenomic and metabolomic data across various studies.41 Multi-omics involves the integration of multiple dataset types including: – Metagenomics: Assessment of the structure and functions of the bacterial genes that are present. May include inferred functionality of 16S rRNA sequencing via PICRUSt2,42 or direct assessment via whole genome shotgun sequencing 29, and the use of gut metabolic or gut brain modules (GMMs and GBMs).27,30,43 – Metatranscriptomics: Focuses on what genes are expressed by the entire microbial community and allows for investigation of not just the functional potential, but the active functional profile of a microbial community. As an example, the bioBakery 3 platform allows investigation of not just the metagenome, but also the metatranscriptome.29 – Metaproteomics: An umbrella term covering various approaches of studying all the proteins in microbial communities and microbiomes from environmental sources. Metaproteomics goes past Metatranscriptomics by giving information about whether transcribed genes result in protein translation. See Li et al. for more details and examples.44 – Metabolomics: The study of small molecules, i.e., metabolites, within cells, biofluids, tissues or organisms, that are produced via bacteria. The metabolites and their interactions within a biological system/host is termed the metabolome. Common methods for assessing the metabolome include mass-spectrometry or nuclear magnetic resonance (NMR) spectroscopy.4

Figure 3.

Recent research has underscored the relevance of the microbiota-gut-brain axis across the extremes of life, from neurodevelopmental disorders through to neurodegenerative disorders. Various environmental factors are involved in these relationships, with shared mechanisms of action of the gut microbiome to both neurodevelopment and neurodegeneration. SCFA = Short chain fatty acid, GI = gastrointestinal, BBB = blood brain barrier.

2. Early-life and the gut microbiota

2.1. Neurodevelopment

Across the duration of pregnancy, the fetal brain is continuously developing to soon face the challenges of postnatal living and beyond. The very early folding of the neural tube, the precursor of the central nervous system, aids in aligning correct neuronal patterning and differentiation and is essential for life.⁴⁸ If this is not fully sealed neural tube defects will form, leading to neurodevelopmental consequences for the fetus,⁴⁸ including spina bifida occulta, meningocele, or anencephaly.⁴⁹ Disruption to development in later pregnancy also leads to complications. While most growth and drastic morphological change occurs within the first 20-weeks, specific fine tuning, including neuronal migration, myelination, and pruning, occurs later in pregnancy.^50–52 Complications during these processes may be associated with cognitive and behavioral neurodevelopmental complications, including ASD and schizophrenia.^53,54 These neuronal and behavioral changes occur in parallel with significant remodeling of the maternal gut microbiome across each trimester.^55–57 Over the last decade the involvement of the microbiome has been investigated in relation to neurodevelopment, particularly regarding ASD, schizophrenia, and ADHD, and a newer body of research now strongly supports earlier indications that the microbiome and neurodevelopmental trajectories are intricately linked.³⁵

During this period, important evidence has emerged linking the microbiome and neurodevelopment (See Table 1 for a summary of several key studies and the characteristics of each). It is important to note however that there are numerous approaches to microbiome testing, with techniques changing and improving overtime. The time of sample collection or the behavioral and milestone assessments may also differ between studies, and this should be considered when comparing various research outputs. Alterations seen in the gut microbiome during specific time windows have been linked with contemporaneous and later time point assessments of infant temperament, with correlations noted between increased relative abundances of Bifidobacteria and Streptococcus, with an overall more positive emotionality and an increased abundance of Akkermansia linked to more social behavior.^88,89 Conversely, microbial changes have also been associated with more negative behavioral outcomes, including impaired stress coping. Increased abundances of Bifidobacteria and Parabacteroides has been linked to increased negative reactivity, along with reduced abundance of Prevotella.⁹⁰ Interestingly, decreased abundances of Prevotella has also been implicated in adverse behavioral outcomes, notably internalizing problems.⁹¹ Separately, decreased relative abundances of Bifidobacteria, Lactobacillus, and Haemophilus have been linked with increased fear bias, which can be representative of emotionally directed attention.⁹² Cognition has also been linked to differential microbial profiles in infancy.⁹³

Table 1.

Summary of research studies and key characteristics of each research study in relation to early-life/neurodevelopment and the gut microbiome.

Animal model studies
Study	Animal + Strain	Sex	Disease/Model	Intervention/ Exposure	Outcomes
Bhagavata Srinivasan et al.58	Sprague-Dawley rats	Female (mothers) + offspring (male/female)	Maternal obesity/western diet and exercise	6-weeks of diet then 10-days of exercise with mating occurring after 10 days of exercise - Control diet + sedentary (CS) - Control diet + Exercise (CE) - Obesogenic diet + sedentary (OS) - Obesogenic diet + exercise (OE)	- Maternal exercise (CE) decreased bacterial alpha-diversity of offspring - Maternal obesogenic diet decreased alpha-diversity irrespective of exercise - Clostridium sensu stricto, Clostridium_XlVa, Bacteroides, Blautia and Lactobacillus were enriched in obesogenic sedentary group compared to control sedentary group
Leyrolle et al.59	CD1 mice	Female (mothers) + offspring (males only)	Maternal immune activation (MIA) and maternal omega-3 deficiency	Upon males introduced to mating cage, Omega-3 deficient (DEF), and omega-3 sufficient (SUFF) diet was given for 48 hours. At embryonic day 17 pregnant females given saline (control), or LPS injection (to induce MIA)	Low maternal Omega-3 intake worsens maternal immune activation-induced gut dysfunction. MIA DEF mice at adulthood displayed learning deficits At phyla, family, and genus levels diet affected gut bacterial relative abundance
Sanguinetti et al.60	B6129SF2/J mice	Female (mothers) + offspring (male/female)	Maternal high-fat diet	Normal diet-fed mothers (NDm) and high-fat diet-fed mothers (HFDm). Mating occurred after 3-months of diet. Diet continued through gestation and lactation.	In offspring at 1- and 6-months gut microbiota compositional differences were present between diet groups
Leclercq et al.61	BALB/c mice	Female (mothers) + offspring (male/female)	Prenatal and early postnatal antibiotic exposure	One week before delivery pregnant females were treated with either drinking water (control), penicillin V (AB) or penicillin V and Lactobacillus rhamnosus JB-1 (AB/JB1), until pup weaning (postnatal day 21)	Penicillin had lasting effects on gut microbiota, increased cytokine expression in frontal cortex, modified blood–brain barrier integrity and altered behavior, irrespective of sex
Champagne-Jorgensen et al.,62	BALB/c mice	Female (mothers) + offspring (male/female)	Prenatal antibiotic exposure	Low-dose penicillin given in water from gestational day 15 to birth (approx. 6 days), control group received water no penicillin in water	Prenatal penicillin exposure induced anxiety-like behaviors for female mice, and abnormal social behavior for males Altered microbiota composition present for both sexes
O’Connor et al.63	C57BL/6 mice	Female (mothers) + offspring (male/female)	Postnatal antibiotic exposure	One day following birth, weaning mothers received either: Vehicle, Penicillin V, or antibiotic cocktail via drinking water for 7-days	Early life exposure to maternal antibiotics altered anxiety, social, and cognitive behaviors. Antibiotic cocktail broadly caused greater effects than single antibiotic
Vuong et al.64	C57Bl/6 J mice	Female (mothers) + offspring (male/female) reared as SPF or GF	Maternal microbiome influence on prenatal neurodevelopment	Four groups: animals reared without microbes (GF), animals with depleted microbiome (Antibiotic [ABx] treated from before conception to E14.5), conventionally colonized (SPF) animals (via SPF bedding at E14.5 till gestation and offspring had SPF bedding until behavioral testing), and colonization of ABX dams with Clostridia-dominant spore-forming (Sp) bacteria.	Embryos from ABX and GF dams had reduced expression of genes related to axonogenesis, impaired outgrowth of thalamic axons Colonization of microbiome depleted dams (Sp) prevented abnormal fetal brain gene expression and axonogenesis.
Lynch et al.65	NIH Swiss mice	Female (mothers) + offspring (male/female)	Antibiotic exposure during either; postnatal (P2-9), pre-weaning (P12-18) or post-weaning (P21-27)	Litters randomized to receive antibiotic cocktail, or 0.9% saline, for either the early postnatal (P2-9), pre-weaning (P12-18), or post-weaning (P21-27) periods.	The early-life antibiotic exposure elicited sex and time-dependent physiological effects. Antibiotic exposure, particularly in weaning period, elicited cecal microbiota disruption into adolescence. Microglial morphology was also altered by early-life antibiotic exposure
Kayyal et alo.66	BALB/c mice	Female (mothers) + offspring (male/female)	Postnatal antibiotic exposure and concurrent probiotic administration	Intervention started at postnatal day 14. Three groups: antibiotic group, antibiotic + L. rhamnosus JB1 and control (fed PBS only)	Male mice exposed to antibiobics developed long-term social behavior changes. All changes seen in brain, behavior and immune cells that were associated with antibiotic exposure, were absent in the antibiotic + L. rhamnosus JB1 group.
Gur et al.67	C57/Bl6 mice	Female (mothers) + offspring (female only [males reported in separate study])	Maternal prenatal stress	Pregnant females randomly assigned to stressed or non-stressed control group. Stressed group had restraint stress between embryonic day 10–16 for 2 h between 09:00 and 12:00. Offspring behavioral testing occurred between P60-P70.	Prenatal stress associated with fecal microbiota changes in pregnant females. BDNF was decreased from prenatal stress in utero and in adulthood. Anxiety and cognitive changes accompany microbial, IL-1β, and BDNF alterations
Gur et al.68	C57/Bl6 mice	Female (mothers) + offspring (male only [females reported in separate study])	Maternal prenatal stress	Pregnant females randomly assigned to stressed or non-stressed control group. Stressed group had restraint stress between embryonic day 10–16 for 2 h between 09:00 and 12:00. Offspring behavioral testing occurred between P60-P70.	Males exposed to prenatal stress had reduced social behavior in adulthood, neuroinflammation was present, along with decreased serotonin metabolism in adulthood. Males exposed to prenatal stress had reduced commensal microbes (i.e., Bacteroides and Parabacteroides) compared to non-prenatal stressed group
Nicolas et al.69	Sprague-Dawley rats	Offspring (female only)	Postnatal stress (Maternal separation)	Two groups: maternal separation from postnatal day 2–12 for 3 h/day, or non-separated group	Maternal separation altered LPS-induced inflammatory response in plasma of juvenile rats, enhanced LPS-induced increase in IL-1β in the ventral hippocampus and maternal separation attenuated an LPS-induced increase
Collins et al.70	Sprague Dawley rats	Offspring (male only)	Maternal separation/postnatal stress + milk fat globule membrane administration	At postnatal day 0, litters randomly assigned to maternally separated (separated at postnatal day 2 to postnatal day 12 for 3 h/day) or non-separated groups. Non-separated and separated groups split into control diets or diets + milk fat globule membrane	Early life stress alters changes in bejhaviour and visceral sensation. Early life stress-induced visceral pain was reversed by milk fat globule membrane
O’Mahony et al.71	Sprague Dawley rats	Offspring (male only)	Maternal separation/postnatal stress + milk fat globule membrane administration and prebiotic (GOS and PDX)	At postnatal day 0, litters randomly assigned to maternally separated (separated at postnatal day 2 to postnatal day 12 for 3 h/day) or non-separated groups. Non-separated and separated groups split into control diets, diets + milk fat globule membrane, diet + prebiotic, or diet + milk fat globule membrane and prebiotic at postnatal day 21.	MS-induced visceral hypersensitivity was ameliorated by MFGM and to greater extent with the combination of MFGM and prebiotic blend. Spatial learning and memory were improved by prebiotics, MFGM and with them combined. The combination of milk fat globule membrane and prebiotic reduced the long-term impact of maternal separation on a marker of myelination in the prefrontal cortex.
McVey Neufeld et al.72	Sprague-Dawley rats	Offspring (male only)	Maternal separation/postnatal stress + prebiotics and Lactobacillus rhamnosus GG (probiotic)	At postnatal day 0, litters randomly assigned to maternally separated (separated at postnatal day 2 to postnatal day 12 for 3 h/day) or non-separated groups. At postnatal day 21 animals randomized into control diet + Lactobacillus rhamnosus GG, control + water (no probiotic) or diet with prebiotics + Lactobacillus rhamnosus GG, or diet with prebiotic and water (no probiotic)	Diet containing prebiotic and probiotic attenuated early-life maternal separation effects on anxiety-like behavior and hippocampal-dependent learning.
Morais et al.73	NIH Swiss mice	Offspring (male only)	C-section, co-housing and probiotic/prebiotic administration.	Two groups: pups born via vaginal birth and a group born via C-section. A third control group to control for the effects of fostering was included (i.e., cross-fostered vaginal delivery). C-section born mice were either exposed to probiotic (Bifodobacterium breve M16V), prebiotic (shot chain GOS and long chain FOS) starting from birth and throughout experiment. Control vaginally born mice were given water.	C-section altered Bifidobacterium spp. Abundance in early-life C-section born mice had behavioral deficits through lifespan, and co-housing these mice with vaginally born mice corrected social deficits. The probiotic or prebiotic mixture also improved behvaiour in C-section born mice.
Sharon et al.74	C57BL/6 J mice	Male and female mice undergoing FMT	Gut microbiome transplant from ASD (or neurotypical) individuals to mice	Mice received FMT from either neurotypical donors, or ASD donors. FMT delivered via single oral gavage	FMT from ASD donors elicited impaired social communication and interaction behaviors in mice. Microbiome profiles of mice who had FMT, showed that specific bacterial taxa and metabolites may modulate ASD-like behaviors.
Sgritta et al.75	C57BL/6 J (Germ free), Shank3B+/– mice, DAT-cre mice, and BTBR mice	Male mice used for mouse models only	ASD mouse model, and treatment with	PBS (vehicle) or L.reuteri were added to drinking water of mouse models and treated water was consumed ad libitum for treatment period	L. reuteri treatment rescues social deficits in several ASD mouse models (including in germ free mice) L. reuteri also reverses social deficits via the vagus nerve
Dervola et al.76	Spontaneously hypertensive rats and Wistar Kyoto rats	Male and female rats used	Rat model of ADHD +	During pregnancy rats given n-3 PUFA enriched feed and offspring continued this diet. Control groups received control feed (n-6/n-3 of 7:1).	In males, PUFA supplemented feed enhanced reinforcement-controlled attention, while reducing hyperactivity and impulsiveness. Opposite or no effects seen in females. Enhanced dopamine and serotonin turnover rates observed in the male mice, whereas females exhibited no/limited change.
Zhu et al.77	C57BL/6 J mice. Fecal samples collected from 11 patients with schizophrenia, and 10 healthy controls	Male only mice use	FMT from drug-free schizophrenia patients to mice	Mice underwent oral gavage of antibiotics, and at 48-hours after last gavage, were randomized to receive FMT via oral gavage from either patients or controls, 9 times over 3 weeks	FMT from schizophrenia patients into mice caused psychomotor hyperactivity, and impaired learning and memory. 60 bacterial species were different between control FMT mice and schizophrenia FMT mice.
Clinical Studies
Slykerman et al.78	342 of the initially eligible 474 children completed cognitive testing at 11-years of age	Sex of children/participants not reported	Antibiotic exposure in early life	Maternal probiotic treatment during pregnancy: either Bifidobacterium animalis HN019, Lactobacillus rhamnosus HN001 or placebo	Children who received antibiotics within first 6-months of life had significantly lower overall cognitive and verbal comprehension and increased risk of problems with executive function, anxiety and attention-deficit hyperactivity disorder (after adjusting for mode of delivery, probiotic treatment, breastfeeding and income).
Aatsinki et al.79	Participants were mothers and 2.5-month-old infants with fecal microbiota composition data from the FinnBrain Birth Cohort study (N = 446)	Infant participants (male = 204, female n = 195)	Maternal prenatal psychological distress	Maternal prenatal psychological distress measured 3 times during pregnancy + hair cortisol measured at gestational week 24.	Positive associations were present between maternal chronic prenatal psychological distress and bacteria from the Proteobacteria phylum. Chronic prenatal psychological distress also negatively associated with Akkermansia. Prenatal psychological distress and hair cortisol did not correlate with infant fecal microbiota diversity.
Korpela et al.80	Mothers who had scheduled/planned C-Section deliveries were enrolled in study (	Infant participants (Female = 5, Male = 2)	Maternal FMT for C-section delivered infants	FMT group had maternal FMT given alongside mothers breastmilk (5 ml total) at first feeding. Two observational cohorts were used for control samples (collected at same time points as for the FMT-treated infants [HELMi and Jorvi cohorts]	Mode of delivery determines fecal microbiota development of infants The microbiota development of the FMT-treated C-Section born infants differed to the untreated C-section infants and showed significant similarity to vaginally born infants.
Zhou et al.81	Mothers who had scheduled/planned C-Section deliveries were enrolled in study (N = 68)	Infant participants (Female = 26, Male = 42)	Maternal vaginal microbiota transfer (via skin) for C-section delivered infants	Infants randomized to control group (exposed to sterile gauze) or vaginal microbiota transfer group (exposed to gauze with vaginal fluids)	Vaginal microbiota transfer associated with improved neurodevelopment in C-section born infants. Vaginal microbiota transfer treated infants and altered gut microbiota compared to controls, potentially this transfer could restore C-section born infants microbiome to resemble vaginally born infants.
Wilson et al.82	Mothers who had scheduled/planned C-Section deliveries (N = 25), or vaginal birth (control group, N = 22)	Infant participants (Female = 25, Male = 22)	Maternal vaginal microbiota transfer (via oral administration) for C-section delivered infants	At birth, infants randomized to receive 3 ml solution of maternal vaginal microbes (CS-seeded, n = 12) or sterile water (CS-placebo, n = 13). Vaginally born were the controls (VB, n = 22)	No observed differences in gut microbiome composition or functional potential between CS-seeded and CS-placebo infants at 1 month or 3 months of age
Shaaban et al.83	ASD children (N = 30) + Controls (N = 30)	Male = 19, Female = 11	Children with ASD administered probiotics	Probiotic contained Lactobacillus acidophilus, Lactobacillus rhamnosus and Bifidobacteria longum. 5 g of probiotic powder dissolved in water, given once daily for 3-months	Compared to baseline, following probiotic supplementation severity of autism symptoms was improved as were gastrointestinal symptoms.
Kang et al.84	ASD children (N = 18), neurotypical children (N = 20)	ASD children (Male = 16, Female = 2), neurotypical (Male = 18, Female = 2)	FMT given to ASD children	2-week antibiotic treatment, then FMT given at a ‘high dose’, plus daily, lower maintenance FMT doses for 7–8 weeks.	GI symptoms were reduced by 80% following treatment and persisted for 8-weeks after treatment. Behavioral ASD symptoms also improved after 8-weeks treatment ended. Overall bacterial diversity and abundance of Bifidobacterium, Prevotella, and Desulfovibrio increased following FMT
Kumperscak et al.85	ADHD Children & adolescents (N = 32)	Placebo group (Female = 3, Male = 11), probiotic group (Female = 6, Male = 12)	ADHD children & adolescents given probiotic	Probiotic group (N = 18) received Lactobacillus rhamnosus GG ATCC53103, and control group (N = 14) received placebo capsule. Capsules were taken once daily for 3-months	Self-report quality of life scores significantly improved in probiotic group, but not control. Results from psychometric tests and levels of inflammatory cytokines were ambiguous.
Skott et al.86	Children (N = 68), and Adults (N = 114) with ADHD diagnosis but no ASD diagnosis.	Children (Male = 50, Female = 18), Adults (Male = 33, Female = 81)	Effects of a synbiotic on patients with ADHD.	Synbiotic group (Children = 42, Adults = 57), placebo (Children = 26, Adults = 57). Treatment given once daily for 9-weeks. Synbiotic contained Pediococcus pentosaceus, Lactobacillus casei ssp paracasei, Lactobacillus plantarum 2362, and 2.5 g each of fermentable fibers betaglucan, inulin, pectin and resistant starch	Synbiotic and placebo group had equal improvements in ADHD symptoms, and neither group had significant changes to functioning or sub-diagnostic autism symptoms. However, Synbiotic treatment reduced sub-diagnostic autism symptoms in the domain restricted, repetitive and stereotyped behaviors in children, and improved emotion regulation in the domain of goal-directed behavior in adults.
Wang et al.87	ADHD children (N = 30)	Children (Male = 24, female = 6)	Effect of Bifidobacterium bifidum on ADHD	Probiotic group only (no control group), consumed 2-probiotic sachets daily for 8 weeks.	Across8-week intervention, patients ADHD symptoms improved. LEfSe analysis revealed that Firmicutes significantly decreased while Proteobacteria significantly increased during the 8-week treatment period.

SPF = specific pathogen free, GF = germ free, PBS = Phosphate Buffered saline, GOS = galacto oligosaccharide, FOS = fructo oligosaccharide, PDX = Polydextrose, FMT = Fecal microbiota transplant, ASD = Autism Spectrum Disorder, ADHD = Attention-deficit/Hyperactivity Disorder, FMT = Fecal Microbiota Transfer, GI = Gastrointestinal, PUFA = Polyunsaturated fatty acids

Specific microbiome windows under consideration have expanded to include the maternal microbiome in preconception, pregnancy, at birth, in the early postnatal period and the infant's microbiome in the context of how these windows shape the long-term health and well-being of the infant (See Figure 2). Some major factors influencing the microbiome will be discussed in the following sections, including maternal diet, antibiotic use, and breastfeeding and birth mode are also known to shape neurodevelopment, mediated by the gut microbiome. Subsequently, the gut microbiome and specific neurodevelopmental conditions will then be discussed.

Figure 2.

Opportunities for the microbiome to affect development of an infant include both the maternal and potentially paternal microbiome in pre-conception, during pregnancy via indirect effects of maternal microbiome metabolites, birth mode which determines initial microbial colonization, along with the postnatal and infant periods that may be affected by environmental factors including breastfeeding vs formula feeding, antibiotic usage, diet and introduction of solid foods.

2.2. The maternal microbiome: From preconception to birth

The presence of microbes in the womb has long been disputed, however current research indicates that the placenta and the amniotic fluid is most likely sterile.^94,95 Since there is no direct transfer from the maternal microbiome in utero, all initial colonization occurs during the birth process and postnatal life. Despite this, it is widely accepted that the maternal microbiome plays a key role in neurodevelopment which will be discussed in detail. It has been suggested that the fetus is influenced indirectly by microbial metabolites produced by the mother.⁹⁶ These metabolites may cross the blood-placenta-barrier,⁶⁴ and then influence the immune response of the fetus.⁹⁷ The role that maternal metabolites have in neurodevelopment has been demonstrated in both animal,⁶⁴ and human studies.⁹⁸ The emergence of the paternal gut microbiome helps to further our understanding of potential mechanisms. A study examining the disrupted gut microbiome of fathers revealed that the offspring were at a greater risk for growth restriction and mortality,⁹⁹ and recent in mice suggests depletion of the paternal microbiome can induce intergenerational effects on offspring behavior and physiology.¹⁰⁰ Authors hypothesize that the gut microbiome may interact with male germ cells, whether directly or indirectly, and disruptions to the gut-germline axis may negatively affect the placenta. It is likely these changes are occurring due to epigenetic alterations during spermatogenesis.¹⁰¹ Between this, and the constant presence of the maternal microbiome, the gut health of both parents is key for shaping the long-term health of the fetus. Additional factors shaping the maternal and infant microbiome are discussed below.

2.2.1. Diet during preconception and pregnancy

Research links elevated homocysteine levels with poor psychomotor development,¹⁰² and folic acid, which aids in closure of the neural tube, to beneficial verbal and cognitive development.^103,104 Additionally, overconsumption of highly processed foods is associated with lower infant developmental scores.¹⁰⁵ Animal models have also demonstrated the potential for nutrient deficiencies, such as polyunsaturated fatty acids, to amplify the effects of stressors such as maternal immune activation on locomotion and memory in affected offspring.⁵⁹ However, when studies examine the effects of consumption of dietary micro- and macronutrients during pregnancy on neurodevelopment, the results are mixed with a systematic review finding limited associations.¹⁰⁶

One of the key factors shaping gut microbiome composition is diet.^107,108 Recent evidence evaluates in greater detail the links between the overall state of the maternal microbiome and the long-term health of the offspring. An overall anti-inflammatory diet in the preconception period seems to lower the risk of impairments in numerous neurodevelopmental parameters, such as motor skills, communication, and problem-solving skills.¹⁰⁹ Diets rich in fat and sugar have also been linked to poor cognitive development, while disrupting the gut microbiome of the offspring of rodent obesity models.^58,60 In humans, pre-pregnancy obesity was associated with alterations in the relative abundance of Lactobacillus and Bifidobacteria genera, both strongly associated with inflammation and has subsequently been shown to increase their risk of developing neurodevelopmental disorders.^110,111 Despite growing interest, the role of the microbiome in preconception diet has yet to be fully elucidated as most observational dietary studies do not yet routinely examine the maternal microbiome. However, future evidence from studies using animal models may give insight and direction toward a better understanding of the underlying mechanisms.

2.2.2. The impact of antibiotic use during pregnancy and early life

Antibiotics are prescribed to 20.8% of pregnant women, as they are vital for healthcare and their use cannot be avoided in severe cases of bacterial infections.¹¹² Initial studies using animal models in this field often used high doses of antibiotics, sometimes as a cocktail, over longer durations that spanned critical time windows.³⁵ Recently, there has been a shift in how these studies are performed, with more targeted microbial depletions at specific time points during gestation taking preference. Rodent studies have indicated that lower doses of a single antibiotic, in this case penicillin, during late pregnancy and early postnatal life, produces lasting effects in both sexes on gut microbiota, including increased cytokine expression in frontal cortex, modification of blood–brain barrier integrity and altered behavior in the offspring.⁶¹ Prenatal exposure to low-dose penicillin only during the last week of gestation has a sex-dependent effect on neurodevelopmental outcomes, with female mice less anxious while males displayed abnormal social behaviors.⁶² Comparisons between early life exposure to maternal antibiotics led to persistent alterations in anxiety, sociability, and cognitive behaviors. Animal models also note that the profile of behavioral effects in terms of anxiety, sociability, and cognitive behaviors due to a broad-spectrum antibiotic cocktail were greater than in animals treated with just penicillin, with the exception of greater deficits in social recognition.⁶³ Mechanistic insights have also accrued and following exposure to an antibiotic cocktail mixture, offspring displayed altered tactile sensitivity which was rescued by concurrent administration of specific beneficial microbial metabolites.⁶⁴ Little work has been done however to examine any potential differences between absorbable and nonabsorbable antibiotics. Most animal models now use various combinations of nonabsorbable antibiotics whose antimicrobial effects are confined to the gut.^113–115 The direct and indirect effects that absorbable antibiotics may have on neurodevelopment is yet to be fully elucidated, and it may prove that the different antibiotic drug classes along with different mechanisms of action (e.g. neurotoxic effects of metranidozole)¹¹⁶ may produce differential effects on both the offspring microbiome as well as many neurodevelopmental parameters, depending on the doses used.

In relation to human observational studies, prenatal antibiotic exposure significantly increases the risk of developing ADHD and ASD.^117–119 Regarding the increased odds of developing ADHD, studies report varying levels of antibiotic exposure are required for neurodevelopmental consequences to occur, with one study finding an association after one course and another study requiring three courses.^117,118 Maternal antifungal exposure has also been associated with an increased the risk of developing ADHD, however this has only been noted in males, hinting at a strong sex difference of prenatal exposures, which aligns with the current literature on sex differences in the prevalence of ADHD.¹¹⁸ While currently only evaluated in animal models, the supplementation of microbial metabolites or certain bacterial strains may be a beneficial future intervention option.

Antibiotics are often given prophylactically to infants in the neonatal intensive care unit ¹²⁰ and overuse is common,¹²¹ with earlier exposure during the first year of life a greater risk factor for subsequent difficulties.¹²² Antibiotic exposure in early life is known to negatively impact the infant's gut microbiome, which is of lower diversity and more unstable, and thus more susceptible to being influenced by external factors compared to that of an adult.¹²³ Early antibiotic use can delay the maturation of the infants microbiome and reduces the diversity of species present,¹²⁴ negatively impacting the microbiota-gut-brain axis, and influencing brain function and behavior.^1,125,126

Infants who received a course of antibiotics within their first 6 months have been found to have a higher risk of developing behavioral and emotional problems, as well as having deficits in language development.⁷⁸ The result from this study strongly supports the accumulating evidence that the microbiome is involved in neurodevelopment. Interestingly, the developmental outcomes are not as severe in infants that received antibiotic treatment aged 6–12 months which have improved cognitive outcomes.⁷⁸ Infants receiving their first course of antibiotics aged 12–24 months were shown to have enhanced cognition in comparison to both previous timepoints.⁷⁸ This suggests that a critical microbiome-gut-brain axis window may potentially exist. However, it is important to note that this does not mean that antibiotic exposure in later life will have no neurodevelopmental consequences. Research previously performed by the same group has noted that following antibiotic exposure within the first 12-months of life negatively impacts overall intelligence and reading ability.¹²⁷ Additionally, higher levels of ADHD-like behavior and depressive-like symptoms were noted in the peri-adolescent period, with an increased risk of ADHD being found if the exposure to antibiotics was within the first two years.^117,127 However, the risk of children developing ASD was found to be quite mixed, with some studies finding an increase in ASD risk following antibiotic usage in early life,^117,128 whereas others found no association.¹²⁹ In contrast to other findings, it has also been reported that oral vancomycin treatment can transiently alleviate ASD symptoms.¹³⁰ Despite this, it is important to note the small sample size of this study, containing only 10 participants, and the improvement significantly waned at the follow-up visit after the initial treatment.

Animal studies also confirmed the results found from observational human studies, demonstrating brain and behavioral changes following early life antibiotic administration.^65,66,131 Recent work has aimed to identify a specific temporal window in which microbial depletion is most critical.⁶⁵ For example, a more targeted depletion of the gut microbiome following broad-spectrum antibiotic treatment has unveiled varying microglia morphologies and gene expression in rodents, dependent upon time of treatment. This work highlights differential effects on cellular morphology and genetic expression dependent upon timing of the exposure, and this may be the key in relation to future neurodevelopmental outcomes. While this study only notes very subtle behavioral alterations, it provides insight into potential mechanisms.

2.2.3. The impact of prenatal and postnatal stressors on the microbiome and offspring brain

Maternal prenatal stress, whether it be psychological or experiencing a life-changing event, can lead to neurodevelopmental consequences for the fetus. It is possible that an altered gut microbiome may promote these outcomes, as stress has been shown to negatively impact the microbiome, although more work is required to establish definitive links.⁶⁷ In animal models, prenatal stress was shown to lead to increased anxiety and alterations to levels of brain derived neurotrophic factor in the placenta and the amygdala in adulthood.⁶⁷ The authors reported that the affected offspring's microbiome was disrupted compared to controls, with an altered relative abundance of Firmicutes, Bacteroidetes, and Rikenellaceae.⁶⁷ Disrupted social behavior is often reported, accompanied by a profile of gut-brain axis alterations including altered microbial composition, neuroinflammation, and tryptophan processing/metabolism, [^68,132 p. 2]. Indeed, the interface between the gut microbiome and the social brain is seen across species.¹³³

Shifts in infant microbial composition following prenatal stress are also seen in humans. This includes an increased abundance in many possibly inflammatory genera from the Proteobacteria group, such as Haemophillus and Campylobacter, following chronic maternal prenatal stress, as well as Finegoldia and Veillonella from the Firmicutes group.⁷⁹ While there is no doubt that prenatal stress impacts cognition and motor development, the association with the maternal microbiome is currently in need for further elaboration.¹³⁴

The relationship between postnatal stressors, such as maternal separation, and neurodevelopment has also been explored, albeit mostly performed in animal studies. Maternal separation has long been used as a model of early life stress, adverse exposures, and importantly, GBA dysfunction.¹³⁵ The isolation of the pups from postnatal day 2 until 12 has not only been shown to disrupt the serotonergic system, alter microglia, and lead to a chronic state of inflammation via polyunsaturated fatty acids,^69,136,137 but has also been shown to alter behavior in rodents,^70,138 such as increased anxiety levels and reduced learning and memory. Since the initial observation that this is also associated with microbiome disruption,¹³⁹ it has subsequently been demonstrated that different aspects of the emergent phenotype are microbiota dependent (anxiety-, depression-like behavior) and independent (HPA axis dysfunction).¹³⁸ Interventions targeting the microbiota that rescue or ameliorate the phenotype, with positive results are emerging. Prebiotic supplementation following early life stress has been shown to regulate the HPA axis, along with milk fat globule membrane (MFGM) which has been shown to reduce visceral sensitivity, with the compounds also improving cognition.^70,71 Anxiety phenotypes, stemming from early life stress has also been rescued following supplementation with prebiotic and Lactobacillus rhamnosus GG.⁷²

In relation to human studies, little work has been done to examine the effect that this specific stressor may have on neurodevelopmental outcomes. What is known however, is that separation of mother and child does negatively impact breastfeeding, which has been shown to impact long-term neurodevelopmental outcomes.^323,140

2.3. Early shaping of the infant microbiome

2.3.1. How preterm birth and birth mode impacts long-term neurodevelopment

Preterm birth, defined as delivery before 37 weeks’ gestation, affects just under 15 million infants annually and this premature delivery can lead to severe neurological consequences.¹⁴¹ Major brain development occurs in the final trimester, and delivery during this critical window may lead to brain injury and other neurodevelopmental consequences.^142,143 The gut microbiome of these infants has previously been shown to have unique microbial profiles.¹⁴⁴ Using 16S sequencing, recent work has shown overgrowth of Klebsiella in the gastrointestinal tract of these premature infants, an increased abundance of Bifidobacteria and Finegoldia species were also noted. Interestingly, these changes were seen in those with severe brain injuries and those without were found to have elevated levels of Streptococcus and Enterococcus.¹⁴³ Despite the many microbial changes, Klebsiella pneumoniae was found to be significantly association with brain injury, with the authors noting that Klebsiella overgrowth is a key for the dysregulation of the infant's microbiome-immune-brain axis, perhaps aggravating the brain injury.^143,145

The newborn's microbiome is heavily influenced by mode of delivery. If born vaginally, the infant is colonized with bacteria present in the mother's vaginal canal, including Lactobacillus which is beneficial and predominant in healthy women, and certain fecal bacteria, such as Bacteroides and Escherichia.⁸¹ In contrast, a cesarean section (C-section) delivery results in the infant colonized with bacteria including Staphylococcus and Klebsiella.⁸¹ Despite C-sections being a vital and often lifesaving healthcare innovation, they may be overutilized in high-to-middle income countries,¹⁴⁶ and they may have neurodevelopmental risks. Research has examined the long-term effect of C-section on cognition and C-sections are linked to an increased risk of intellectual disability, ADHD, and ASD.^147,322 Additionally, evidence shows, at least for ASD, that this association may be due to familial confounding by genetic and/or environmental factors.¹⁴⁹ Deficits in gross and fine motor development, along with language skills have also been noted by 5 months old.¹⁵⁰ Problem-solving has also been found to be delayed at 3 years following C-section delivery.¹⁵¹ Interestingly, the risk of certain disorders, such as ASD, seems to be greater if the C-section is elective.¹⁴⁸ It is possible that the underlying reasons for the C-section, along with the altered microbial colonization, may play a role in the neurodevelopmental deficits, and preclinical research highlights a role for the gut microbiome in the enduring behavioral effects induced by a c-section model.¹⁵² Given the gut microbiome alterations following C-section do not endure, there remains an urgent need to demarcate the time windows when interventions should be applied. The behavioral alterations that are seen in the human population has also been demonstrated in animal models. C-section has been shown to negatively affect behavior across the lifespan of mice, noting higher levels of anxiety and altered sociability.⁷³ Interestingly, these neurobehavioural alterations were partially rescued following co-housing with mice born conventionally, allowing a natural transfer of the microbiome. Similar positive changes were also noted following administration of Bifidobacterium breve, highlighting the potential impact that a probiotic intervention may have against these known C-section deficits. It has also been noted that this model results in increased body weight.¹⁵³ Interestingly, prenatal and peripartum exposure to antibiotics interface with C-section delivery to modify microbiome profiles in human infants,¹⁵⁴ a feature which needs to be built into more translationally faithful versions of this animal model.

Research has attempted to ameliorate the effects of a C-section birth on microbial colonization, by examining the impact of ‘seeding’ the neonate with maternal vaginal secretions. Seeding aims to recolonize the neonate's gut to match a vaginally born infant, which will have the microbes of the maternal gut and vagina, as opposed to the microbes of the maternal skin.¹⁵⁵ Recolonization may alleviate the negative developmental outcomes that can be associated with C-section.⁸¹ While not widely used, preliminary results have found that vaginal seeding of C-section exposed infants leads to neonate intestinal recolonization which partially matches a vaginally born infant.^81,156 Recolonization also helped to regulate levels of important metabolites and metabolic functions, including an increase in carbohydrate, amino acid, and energy metabolism and increased levels of L-lactic acid, homovanillic acid, and N-acetyl-D-glucosamine, which may potentially lead to better outcomes in later life.^81,156 Individuals receiving the vaginal microbial transplant scored higher in neurodevelopmental questionnaires than those who received a saline control.⁸¹ However, additional research indicates no positive effect of seeding.⁸² Separate research has shown that maternal FMT displays promising results in normalizing gut bacterial composition in C-section compared to vaginal births in mice.⁸⁰ While findings suggest the possibility of intervention at this early stage, to prevent development of certain disorders, there is currently limited evidence supporting this technique given that most studies are too small and exploratory in nature to form an actionable evidence base at present. Others have highlighted the potential dangers of seeding the infant with a suboptimal vaginal microbiome and/or one containing harmful pathogens, including group B Streptopcoccus, and various sexually transmitted infections, i.e., Neisseria gonorrhoeae or Chlamydia trachomatis, with screening of donor samples an important consideration in future guidelines.¹⁵⁷

2.3.2. The neurodevelopmental benefits of breastmilk

Breastfeeding has been recommended by the World Health Organization for a minimum of 6 months, but should ideally remain part of the diet until the infant reaches 2-years.^158–160 Through breastfeeding, the infant receives various nutrients including necessary proteins (lactoferrins and caseins), prebiotics (human milk oligosaccharides), polyunsaturated fats, and immunoglobulins.¹⁶¹ These nutritional compounds positively influence the developing microbiome of the infant and can increase the abundance of beneficial microbes, including Bifidobacteria, and can decrease abundance of other bacteria including Bacteroides.¹⁶² Interestingly, infants fed breast milk exclusively have a less diverse and more immature microbial composition than those receiving infant formula as their diet.¹⁶³ However, in this context, low diversity is perhaps more beneficial for the infant and recent evidence supports the idea that breastfeeding supports an appropriately paced assembly of the gut microbiota to control the acquisition of microbial species and functions.¹⁶⁴ Feeding type is known to alter gene expression and this ultimately enhances the transcription of genes that are associated with various immune and metabolic processes.¹⁶⁵ In comparison, the microbiome of formula fed infants is more adult-like, and this may not be ideal in early-life.¹⁶⁵ Breastfeeding for a prolonged period is strongly associated with optimal neurodevelopment, including a reduced risk of ADHD, with a similar protective effect found with ASD and schizophrenia.^166–170 Breastfeeding is also associated with greater IQ scores and enhanced cognition.^171,172 Magnetic resonance imaging studies show white matter microstructural development is increased in areas associated with language, vision, and higher order thinking and cognition in longer durations of breastfeeding, or when comparing breastfeeding to formula fed infants.¹⁷³ The role that the gut microbiome has in the neurodevelopmental benefits of breastmilk is still to be fully elucidated, but one potential component, MFGM, has shown great potential in improving certain neurodevelopmental outcomes following maternal separation in rats.^70,71 This includes spatial learning and memory, along with reducing visceral sensitivity stemming from the early life stress.^70,71 In the same study, prebiotics also displayed potential in improving learning and memory along with helping the HPA axis return to baseline levels following restraint stress, with the combination of prebiotics and MFGM proving to have some greater effects.⁷¹

2.4. Neurodevelopmental disorders and the gut microbiome

While the subsequent sections will focus on the role of the microbiota-gut-brain axis in neurodevelopment disorders, it is important to highlight the genetic component to neurodevelopmental disorders. Autism, ADHD and schizophrenia all have a genetic component increasing risk of developing the disorder,^174–176 with ASD genetic risk not only inherited but also involving de novo copy number variations,¹⁷⁷ while ADHD and schizophrenia may mostly involve heritable genetic risk.^178–180 Despite the literature highlighting the genetic component to these disorders, how this relates to environmental factors, such as breast feeding, and to the gut microbiome is not yet understood. Additionally, the genetic risk factors only explain a smaller part of the disease risk (despite high heritability in twin and family studies) and the gut microbiome could be one of the further contributing factors in developing a disorder. To develop a more whole understanding of the risk factors contributing to these disorders, researchers need to investigate gene-environment-development interaction.

2.4.1. Autism Spectrum Disorder (ASD)

As described above numerous factors influence the microbiome in early life and may increase the risk of neurodevelopmental disorders. People with ASD often suffer from gastrointestinal symptoms, which may be due to a disruption of the microbiota-gut-brain axis.^181,182 A gut bacterial disruption could relate to the significant relative bacterial abundance differences in 27 genera, including Akkermansia, Bifidobacterium, Clostridium, and Ruminoccocus, when comparing ASD to controls.^181,183 Gut microbiome differences are also observed in animal models of ASD, including greater relative abundance of Firmicutes and Clostridia, along with lower Bacteroidetes abundance.^181,184 Recently, 16S sequencing has identified a signal unique to ASD patients that has been shown to be driven by 591 microbes.⁴¹ This, coupled with 138 microbial encoding genes linked to ASD through shotgun sequencing, may prove to be a beneficial aid in diagnoses practices.⁴¹

Of note, common ASD symptoms, i.e., repetitive behaviors and sensory preferences, are linked to certain food choices, and typically a restrictive diet leads to a less diverse microbial composition.¹⁸⁵ While this suggests that diet drives the microbiota alterations that have been reported, we should not discount the ongoing role of the gut microbiome in ASD symptom expression, or that it might represent a tractable target in this context.

In experimental research, FMTs from donors with ASD to mice transferred both the patient's microbiome, and an ASD-like phenotype to the mice (i.e., decreased communication and locomotion, and increased repetitive behaviors).⁷⁴ The shift in microbiome composition correlated with reduced social behaviors and repetitive behaviors, and Eisenbergiella tayi (02b40_Lachnospiraceae) correlated with increased repetition and social defects.⁷⁴ Overall, this study shows the potential causal role that the gut microbiome has in the pathophysiology of ASD, or at least of specific symptom profiles, and indicates that targeting the gut microbiome may be an effective therapeutic option for ASD. Yet, other work has deduced other potential mechanisms, such as the vagus nerve, to modulate certain ASD symptoms.⁷⁵ This has been seen following Lactobacillus reuteri administration in the Shank3B-/- mouse model, noting a negative impact on social interaction following vagotomy, showing an indirect mechanism for beneficial microbes to alter the symptom severity.⁷⁵ Probiotic interventions, using species such as Bifidobacterium longum, Lactobacillus rhamnosus, and Lactobacillus acidophilus, helped restore the microbiome and even reduced the severity of gastrointestinal and autistic symptoms in children with ASD.^83,183 FMT interventions have also been tested, which involve antibiotic consumption, fasting and bowel cleansing, and a particular refinement of FMT from neurotypical donors along with stomach acid suppressants, which resulted in children with ASD reporting a reductions in gastrointestinal symptoms, enhanced microbial richness and plasma metabolites, and alleviation of autistic symptoms.^84,186,187 These positive changes occurred after the initial microbiota transfer therapy and persisted for 2-years following cessation of the transfer, with some symptoms even improving further after the end of the study.¹⁸⁶ Across the timeframe of the study, an increase in Prevotella was noted, along with significant decreases in the abundance of certain microbes that had low uncertainty.⁴¹ Abundances of Anaerobutyricum and Butyricimonas, SCFA producers, remained stable across the study which may be potentially crucial in modulating symptoms.⁴¹ The beneficial effects of probiotics and microbial transplants highlight the intricate relationship between the gut microbiome and ASD, as well as the potential for it to act as a therapeutic target.

2.4.2. Attention-deficit/hyperactivity disorder (ADHD)

The gut microbiome has also been implemented as a factor that may be influencing ADHD. A significant amount of our understanding comes from human studies. When compared to those without ADHD, studies have found significant alterations in the composition of the gut, yet findings are not consistent.^188–191 Differences between patients and controls include an elevation of Odoribacter splanchnicus, Bifidobacterium and Bacteroidaceae abundances, and lower abundance of Faecalibacterium prausnitzii.^191–193 Greater Actinobacteria abundance is also associated with reduced ADHD symptom severity. ¹⁹¹ Given that the prior studies indicate relationships between gut bacteria and ADHD, studies have investigated the use of probiotics in ADHD. Lactobacillus rhamnosus GG, when given during pregnancy and early life, can significantly reduce the overall risk of ADHD and improve outcomes such as overall social and emotional functioning in children.^85,189,194 Beneficial impacts on behavior in children with ADHD were also observed with Bifidobacterium bifidum consumption, and a multi-strain probiotic, which alleviated symptoms of inattention and impulsivity.¹⁸⁹ However, no effect was seen with a synbiotic consisting of a Lactobacillus based probiotic and 3 different fermentable fibers on overall ADHD symptoms.⁸⁶ The synbiotic seemed to have a greater effect on those with elevated levels of soluble vascular cell adhesion molecule-1 before the beginning of the trial, with these adults displaying improved emotional regulation. Apart from probiotic interventions, some research has examined the effect of FMT on ADHD, where material from a neurotypical donor led to increased abundance of, the anti-inflammatory and beneficial, Faecalibacterium prausnitzii, reduced abundance in Bifidobacterium longum, and a reduction in ADHD symptoms.^87,195 While this result is promising, the research is a case-study of one participant. To generate more conclusive results extensive work needs be carried out to fully understand if targeting the microbiome via probiotics or FMT is a useful intervention in ADHD. Furthermore, the use of animal models has been greatly underutilized in studying this topic. Enriching rats diet with omega 3 polyunsaturated fatty acids was found to ameliorate symptoms of ADHD in males,⁷⁶ particularly through decreasing impulsive behaviors and improving attention, but not much more has come from animal studies.

2.4.3. Schizophrenia

As with ASD and ADHD, schizophrenia is a mental health disorder counted among the neurodevelopmental disorders, although occurring later in life compared to ASD and ADHD, and is becoming more convincingly associated with the gut microbiome.^196–198 Patients with schizophrenia display a different gut microbial composition to those without schizophrenia, including lower abundance of Clostridium, Faecalibacterium and Ruminococcus, with greater abundances of Lactobacillus, Collinsella, and Anaerococcus.^199,200 At the genus level, those with schizophrenia had greater α and β diversity, and had greater abundance of anaerobic bacteria, along with various species that typically reside in the oral microbiome.²⁰¹ While limited animal work has been performed tying the microbiome and schizophrenia together, one study demonstrated severe behavioral changes in mice that had received FMT from patients with schizophrenia.⁷⁷ These changes include impaired learning and memory and psychomotor activity and highlights the severe disruption of the patient's microbiome and provides insight into how the microbial alterations seen in the gut may be affecting the gut-brain axis and behavior. While probiotic treatments have improved certain GI symptoms, potentially due to the immunomodulatory effects of the probiotic bacteria, no definitive change in schizophrenia symptom profiles were shown.^202,203 Of note, distinct microbial profiles are associated with different symptoms of schizophrenia, i.e., positive symptoms including hallucinations and delusions, or negative symptoms such as anhedonia, avolition, and issues with intrapersonal relationships.^191,204 Positive symptom phenotypes are positively correlated with Lactobacillus, and negative symptoms are positively correlated with Lachnospiraceae and Ruminococcaceae.²⁰⁵ The severity of symptoms can also be correlated with numerous bacterial species in the gut, and furthermore, the microbial profile is known to shift and change following a first episode of psychosis,¹⁹⁹ p. 202.²⁰⁶ Given the literature highlights differential microbial profiles between patients based on their symptomology, future research should consider different phenotypes and trajectories of schizophrenia when assessing and implementing gut-based therapeutics. The potential for modulating the gut microbiome is not as clear with schizophrenia as it is for ASD and ADHD, yet the hope remains that careful microbial targeting, and regular psychotropic medication, may improve the outcomes and lives of those with schizophrenia.

3. Aging and the gut microbiota

Several hallmarks of aging have recently been proposed that relate to the microbiota-gut-brain axis; chronic inflammation and compositional/functional gut microbiome alterations.²⁰⁷ Aging can be both ‘healthy’, where age-related declines occur at a normal pace or are delayed, or can be ‘unhealthy’, when declines are accelerated.²⁰⁸ Gut bacterial differences have been observed across the age range from newborns to centenarians.⁸ In long living individuals (i.e., nonagenarians and centenarians), gut bacterial composition differs to younger age ranges,^209,210 and compositional and functional differences are also present in these populations depending on health status.³²¹ Additional research assessing 21,000 fecal samples determined that a loss of the ‘core’ microbiome and an increase in disease associated taxa were most correlated to unhealthy aging.²¹¹ In animal models, it has been shown that germ free mice can live longer than specific pathogen free mice, despite the fact that germ free animals have several altered body systems and behaviors, indicating the need for a gut microbiome for ‘normal’ aging and immune/nervous system function.²⁰ However, there is little current research that has thoroughly investigated the differences in physiology of aging between germ free mice and those with a gut microbiome, and the absence of pathological infections also needs to be considered as a factor.²⁰

Longitudinal assessments of cohorts across aging would help assess the link between gut bacteria and aging. However, existing research in humans is limited. A published protocol paper aimed to track older adults over a 4-year period,²¹² and investigations of centenarians’ gut bacteria over 1.5 years has occurred.²¹³ Compared to older adults, centenarians had a ‘youth-associated’ gut microbiota profile, including less potentially pathogenic bacteria and a greater abundance of Bacteroidetes. Centenarians also had greater within-individual similarity over time, compared to between-individuals, and individuals with the most within-individual similarity had greater Shannon's diversity, and Pielou's evenness index, indicating that greater evenness and diversity contribute to gut microbial stability in centenarians.

An important consideration when investigating aging, is chronological age (time since birth) vs. biological age [refers to biological and physiological (I.e., DNA), or lifestyle (I.e., nutrition and physical activity) related factors affecting longevity and functional capacity].^214,215 For research on determining biological vs chronological age see Rutledge et al.²¹⁶ Biological age relates to frailty, a decrease in function and physiological reserves that coincide with an increased risk of morbidity and mortality.²¹⁷ Assessment of biological age is important for gut microbiome research as frailty, or unhealthy aging may not relate to chronological age, and the associations between the hallmarks of biological age and gut microbiota may be more informative. Assessment of frailty and chronologic age in relation to the gut, oral and skin microbiota showed associations between the skin microbiota and frailty were present for several taxa at several skin microbiota sites.²¹⁸ The skin microbiota was not associated to chronological age, indicating the importance of assessing frailty, i.e., biological age. However, authors did not report associations between frailty and the oral or gut microbiota, and it is not explicitly stated if assessment of gut bacteria and frailty were conducted for both older adult groups, or separately for their community and nursing facility dwelling groups. Of note, nursing home dwellers were significantly frailer than community dwellers, which is to be expected, however location, not just frailty, would affect the observed skin microbiota associations and the reported associations between skin microbiota and frailty, cannot be disentangled from location. Other research shows associations between gut bacteria diversity and frailty of nursing home residents, with trends for greater composition of butyrate producers in individuals with lower frailty, and greater LPS biosynthesis in frailer individuals.²¹⁹

Several FMT trials in mice have helped elucidate causality of the gut microbiome in relation to aging. When FMT was conducted from young donor (3–4 months) mice to older recipients (19–20 months) the aging-associated differences in brain and peripheral immunity, and the hippocampal metabolome and transcriptome, were reversed.²²⁰ The reversal of age-related brain differences in mice from FMT has been replicated, with declining cognition rescued from FMT of young-to-older mice.²²¹ Additionally, intestinal microbiota were exchanged between young (3-months) and old (18-months) mice (i.e., young to old, and old to young FMTs).²²² When the aged microbiota was received by young mice, there was increased intestinal barrier permeability, greater inflammation (measured via serum LPS-binding protein and IL-6), and more markers of systemic and tissue markers of ‘inflammageing’. In the opposite direction, older mice in receipt of a young donor microbiota, the age-associated elevations of inflammation were reversed (i.e., reduction in serum TNF and LPS-binding protein). Understanding the role of aging on the gut microbiome is important in the context of neurodegenerative diseases as advancing age is the biggest risk factor for Alzheimer's and Parkinson's disease,^223–225 is a major risk factor for ALS,^226,227 and while disease onset occurs much earlier in terms of age (years) in HD and MS, it is still important to understand how aging impacts disease onset and progression.^228,229

3.1. Cognitive decline, MCI, and Alzheimer's disease

Some cognitive decline normally occurs with advancing age. When the rate of cognitive decline increases it may be considered mild-cognitive impairment (MCI), a common precursor to dementia.^230,231 Several observation and intervention studies have investigated cognition and the gut microbiome in humans, as covered previously.²³² Much of the literature focuses on the gut microbiome and cognitive impairment or dementia, predominately AD. Additionally, the potential role of the oral microbiome in AD is emerging.²³³ Purported mechanisms whereby the gut microbiota may relate to AD and cognitive decline is covered previously,^11,234 but to summarize, an altered gut microbiota may contribute to AD via inflammatory mediated pathways, and via microbial metabolites affecting cognition too. A systematic review of observational research comparing the gut microbiota of healthy controls to AD and/or MCI groups found the AD/MCI groups had lower diversity than controls, and differences in bacterial abundance at the phyla level, family level, and genus level (greater Proteobacteria abundance; lower Clostridiaceae abundance; greater Bifidobacterium abundance in AD/MCI, respectively).³⁷ Additionally, a meta-analysis found that probiotics were effective in improving cognition in people with AD or MCI.²³⁵ Another meta-analysis, solely focusing on AD, included three studies, covered in the prior review,²³⁵ but showed no beneficial effect of probiotic supplementation on cognition.²³⁶ The disparate findings may be due to the small number of trials, but may reflect that probiotic consumption is most effective earlier on in cognitive decline, prior to substantial cognitive decline occurring, such as in AD.

By comparing healthy individuals to preclinical AD (determined via investigating tau and/or Aβ plaques through PET imaging), but lacking AD symptoms, Ferreiro et al., found distinct gut microbiome abundance profiles associated with either preclinical AD or healthy controls.²³⁷ The differences in the gut microbiome composition correlated with Beta amyloid and tau pathological biomarkers, but not with biomarkers of neurodegeneration. Additionally, microbial pathways of L-arginine, L-ornithine, and 4-aminobutanoate degradation were associated with preclinical AD status via regression models, with all three pathways sharing succinate as a product. Overall, this shows an altered gut microbiome may contribute to initial AD pathophysiology, but might be less relevant to ongoing AD progression, and highlights the potential for an altered SCFA metabolism state to be present in preclinical AD. Prior research shows lower SCFAs levels in MCI compared to controls, with AD patients compared to MCI having lower levels of SCFAs.²³⁸ Supporting this, research shows reduced levels of putatively butyrogenic bacterial genera in AD compared to healthy controls.²³⁹ Apart from butyrate, succinate is likely immunomodulatory,^240,241 and inflammation via the gut is associated with AD pathophysiology.²⁴² Secondly, succinate is a precursor to propionate, and excess propionate has been associated with AD in humans.²⁴³ Despite research indicating the likely involvement of gut microbiota in AD pathophysiology via multiple routes,²⁴² particularly early in the disease progression, research is currently not at the point where a gut microbiota biomarker is available to be used as an early-detection biomarker.

FMT research has investigated cognition and the gut microbiome, and shows that a ‘young’ microbiota helps improve cognition when transplanted into older mice.²²⁰ Mechanistically, FMT from aged to young recipient mice decreased adult neurogenesis and novelty-induced neuronal activation of the hippocampus.²⁴⁴ By interrogating vagus-nerve related neuronal activity the authors showed that mice receiving the aged microbiota FMT had reduced neuronal activity in the ascending vagus-nerve output brain structure. In an AD animal model, near daily FMTs across 4-months (wild-type donors to AD model) rescued cognitive deficits, and reduced formation of amyloid β plaques and neurofibrillary tangles.²⁴⁵ Increasing the translatability of FMT research, Grabrucker et al.,²⁴⁶ transplanted human fecal samples from healthy donors and people with AD into animals. Authors displayed that behavioral AD symptoms, specifically cognitive/memory deficits (which correlated with donors Clinical Dementia Rating scale), were transferred to previously healthy animals via gut microbiota, confirming a causal and ongoing role of gut microbiota in AD symptoms. Therefore, while human research showed the gut microbiome to be related to early AD pathophysiology,²³⁷ human-to-animal FMT studies indicate the prolonged role of the gut microbiota throughout AD. Grabrucker et al.,²⁴⁶ also substantiated prior research, with FMT from AD patients to animals impairing behaviors dependent on adult hippocampal neurogenesis, and the severity of impairments in mice were correlated with cognitive scores of the AD donors.

A case study of FMT to treat clostridium difficile infection, in a person with Alzheimer's, showed improvements in cognition (measured via mini-mental state examination[MMSE] and Montreal Cognitive Assessment[MoCA]) one-month following FMT.²⁴⁷ The same group assessed 5 people with AD and clostridium difficile infection over 3-months, reporting no serious adverse events from the FMT and all patients having improved cognition (measured via MMSE, MoCA and Clinical Dementia Rating Scale Sum of Boxes).²⁴⁸ Whether cognitive improvements are related to the FMT affecting AD pathophysiology, or simply treating the infection, is not clear. Apart from FMT, probiotics have been investigated as a therapeutic option across several animal models for AD, which has been summarized previously.²⁴⁹ Future research assessing FMT in AD specifically, should use cognitive tests that address specific cognitive domains, and that are more sensitive to changes to cognition (I.e., N-back, spatial working memory etc.), rather than cognitive screening forms including the MoCA or MMSE.

3.2. Additional neurodegenerative diseases

Alzheimer's, discussed above, and PD, ALS, MS, and HD described below (with key gut microbiome research studies and study characteristics described in Table 2), all have a genetic/heritable component. In AD the presence of the apolipoprotein E gene-e4 allele is a contributing risk factor ²⁶⁰ with additional gene mutations implicated in AD onset.^223,261 A genetic component to PD onset and severity/symptoms is also present,^262,263 which similar to the genetic risk in AD, mostly relates to early-onset cases with reduced relevance to late onset cases. For ALS,²⁶⁴ and MS,²⁶⁵ genetic/familial risk factors exist, although genetics alone explain only a small proportion of the risk of disease development.²⁶⁶ Comparatively HD is considered a solely genetic disease, with children of those who carry the increased number of ‘CAG’ triplets on chromosome 4 having a 50% chance of inheriting the gene mutation and developing HD,²⁶⁷ with the number of ‘CAG’ repeats associating with age of HD onset (along with additional genetic factors explaining variance).²⁶⁸ Despite the genetic component to HD, environmental factors may exist that alter disease progression or age of onset.²⁶⁹ Given the genetic and environmental factors underpinning these neurodegenerative diseases, future neurodegenerative disease research should assess the genetic, environmental and microbiome factors. Doing so will be important to understand the interaction between genetics and environmental factors and to uncover where the microbiota-gut-brain axis fits into the web of risk factors.

Table 2.

Summary of research studies and key characteristics of each research study in relation to later-life, neurodegenerative diseases and the gut microbiome.

Animal model studies
Study	Animal + Strain	Sex	Disease/Model	Intervention	Outcomes
Armstrong,220	C57BL/6 J mice	Male only used	Aging related behavioral deficits in mice countered with ‘young’ FMT	Aged mice randomized to receive FMT from young or aged mice. FMT delivered via oral gavage once a day for 3-days, then twice weekly for 8-weeks	FMT from young donor (3–4 months) mice to older recipients (19–20 months) reversed the aging-associated differences in brain and peripheral immunity, and the hippocampal metabolome and transcriptome.
Hou et al.221	C57BL/6 J mice	Male only used	Age-related cognitive decline model	Antibiotic treatment preceded FMT, with FMT performed twice, 72 h apart. Mice (15–16 months old) received FMT from young (8-week-old) donor mice,	Cognitive deficits from aging-related gut microbiome remodeling were rescued by FMT from young donors.
Reeve et al.222	C57BL/6 J	Male only used	Effects of aging in mice and FMT from ‘young’ donors	Young (3 months), old (18 months), or aged (24 months) male mice were randomized to experimental cages to receive either FMT or PBS (control). 3-day delivery of broad-spectrum antibiotic given via oral gavage with palatable antibiotics via drinking water. FMT delivered twice 72 h apart	Older/aged and young microbiome profile was successfully transferred via FMT to young and older mice respectively. Aged donor FMT to young mice accelerated CNS related inflammation. Such effects were reversed in older mice by young donor FMT.
Kim et al.244	RjOrl: SWISS mice	Male only	Effects of ‘young’ and ‘aged’ gut microbiome (from humans and mice) FMT on young mice	Following antibiotics (1-week) young mice received FMT from either young mice donors, or aged mice donors. Separately, young mice received FMT from old or young humans.	FMT from aged, but not young, human/animal donors into young mice triggers profound hippocampal alterations, including astrogliosis, decreased adult neurogenesis, decreased novelty-induced neuronal activation, and impairment in hippocampus-dependent memory
Ağagündüz245	WT mice and ADLPAPT transgenic mice	Female only mice	AD mouse model (ADLPAPT transgenic mouse model) and FMT	FMT from WT mice given orally to ADLPAPT mice for 16-weeks, and given to ABX-pretreated ADLPAPT mice for 4 weeks	Gut microbiota composition differed between the WT and ADLPAPT mice. ADLPAPT mice also displayed loss of epithelial barrier integrity and intestinal and systemic chronic inflammation. FMT from WT mice ameliorated formation of amyloid β plaques and neurofibrillary tangles and cognitive impairment.
Fratiglioni et al.246	41 healthy controls, 54 AD patients. Adult Sprague-Dawley rats	25 female controls, 31 female AD patients. Male rats used only.	FMT from AD patients and age-matched healthy controls into young adult rats	Rats administered ABX-cocktail for 7-days via drinking water. 72 h later FMT given via oral gavage of donor microbiota daily for 3-days. N = 16 rats received control FMT, N = 16 rats received AD FMT	AD symptoms were transferred to rats via FMT from human AD patient donors. Behavioral Impairments in rats receiving FMT from AD patients related to adult hippocampal neurogenesis, and severity of these impairments correlated with clinical cognitive scores of AD donors.
Ingre et al.250	C57BL/6 J mice	Male mice	Rotenone induced PD mice model	FMT from control mice given to rotenone induced PD-like mice for 2-weeks, once daily. 15 rotenone/PD mice received FMT while 15 mice did not.	Rotenone-induced mice developed gut microbiota dysbiosis, GI impairment and poor behavior performances. Akkermansia and Desufvibrio bacteria were increased also. In FMT treated mice, FMT restored the gut microbiome community, ameliorating GI dysfunction and motor deficits, and reduced systemtic inflammation levels.
Waubant et al.251	C57BL/6 mice	Male mice	MPTP induced PD mice model + FMT treatment	N = 15 control mice, N = 15 MPTP + PBS group, and N = 15 MPTP then FMT treated mice. MPTP injection for 5-days, then FMT treatment (from control mice) for 7-days via gavage.	Murine model of PD exhibited gut microbial dysbiosis. FMT from control mice to PD mice improves motor function, and ncreased striatal DA and 5-HT content of PD mice.
Olsson et al.252	C57BL/6 J mice Human PD patients and human controls as stool donors for FMT	Male mice Sex of human donors not reported	MPTP-induced PD mouse model and FMT treatment	MPTP given once daily for 5-days + PBS (N = 12), control mice received saline + PBS in place of any FMT treatment (N = 12). FMT given via gavage daily for 10-days for the PD FMT mouse group (N = 12), and the control FMT mouse group (N = 12)	MPTP-mice had elevated Desulfvibrio, mice receiving FMT from PD donors had elevated Akkermansia. FMT from PD donors to MPTP treated mice significantly aggravated motor impairments, dopaminergic neurodegeneration and colonic inflammation. FMT from healthy controls improved these factors in MPTP treated mice.
Curran et al.148	SOD1G93A ALS model mice and age-matched wild-type mice	Male and female mice used	ALS mice model, treated with butyrate or antibiotics	SOD1G93A/ALS model mice randomized to either non-treatment group (N = 6), or sodium butyrate treatment (N = 10). Sodium butyrate given in 2% concentration in filtered drinking water. SOD1G93A/ALS model mice randomized to either non-treatment group (N = 8), or ABX treatment (N = 11). ABX given via filtered drinking water.	Butyrate or ABX treated ALS mice had longer latency to fall (i.e., improved performance) in rotarod test, reduced SOD1G93A aggregation and enhanced enteric neuromuscular function.
Myers et al.253	G93A transgenic mice and age matched wild type mice	Not disclosed	G93A transgenic ALS mouse model, treated with butyrate	G93A mice randomized into two groups. Butyrate-treated group received sodium butyrate 2% concentration in filtered drinking water. Control group received filtered drinking water (no sodium butyrate)	Butyrate treatment of ALS model mice restored intestinal microbiota homeostasis, improved GI integrity. Abnormal Paneth cells were significantly decreased in ALS model mice following butyrate treatment
Lee et al.254	C57BL/6 mice	Female mice	Experimental autoimmune encephalomyelitis (mouse model of MS] + FMT treatment	Immunized mice randomized into a FMT or no-FMT group. FMT group received FMT via oral gavage daily for 42 days.	FMT may rectify altered gut microbiota in the MS-model mice. FMT also reduced microglia and astrocyte activation and conferred protection on the blood-brain-barrier
Mo et al.255	Hemizygous R6/1 transgenic mice crossed with F1 (CBA x C57Bl/6) mice to have male wild-type (WT) and R6/1 (HD) littermates	Male mice only	Transgenic mouse model of HD	No intervention, observation of HD mouse model across time and compared to WT mice	HD mouse model had gut dysbiosis at 12-weeks, not earlier, Gut microbiome volatility was increased in HD mouse model, pre-symptomatic stage. Increase in butanoate metabolism pathway in HD mice model
Fasano et al.256	R6/1 transgenic male mice (CBA x C57Bl/6 strain) crossed with females (CBA x C57Bl/6) to generate wild-type (WT) and R6/1 (HD) littermates	Male mice only	Transgenic mouse model of HD and fiber diet	WT and R6/1 mice randomized to control diet (5% fiber), zero-fiber diet or high-fiber (10% resistant starch fiber) from 6–20 weeks.	High-fiber reduced pro-inflammatory bacteria in gut microbiome, and treated affective/cognitive deficits in HD mice. Functional analysis revealed high-fiber diet decreased potentially pathogenic functional bacterial pathways in HD mice.
Hawkes et al.257	Male R6/1 hemizygous mice were crossed with female CBAxC57Bl/6 F1 mice to generate male and female WT and R6/1 (HD) littermates	Male and female mice	Transgenic mouse model of HD, and FMT	Mice randomized to control (vehicle/vehicle), ATB-only (ATB/vehicle) or ATB/FMT group. FMT group received oral administration of WT donor ceca content for 3-days	FMT from WT donor to HD mice positively alters cognitive outcomes (particularly in females). In HD male mice, FMT engraftment was inefficient.
Clinical Studies
Study	Participants	Sex	Disease/Exposure	Intervention	Outcomes
Billingsley et al.248	5 AD patients, + stool donors (healthy controls)	3 females, 2 males	AD + Clostridioides difficile infection then undergoing FMT	AD patients receiving FMT first underwent bowel preparations as indicated by their condition and doctor's instructions. FMT delivered to intestinal tract during colonoscopy by gastroenterologist, following FMT guidelines	Improvements in MMSE and MoCA were observed pre-post FMT. Gut microbiota composition changes post-FMT, with increase in Bacteroidaceae and a decrease in Enterococcaceae.
O’Neill et al.258	PD patients and healthy donors	N = 3 females, N = 9 males	PD + FMT	PD patients in FMT group (N = 8) received 60 g of donor fecal material lyophilized into powder and in capsule, administered twice a week for 12-weeks. Placebo group (N = 4) received placebo capsules at same rate.	Some non-severe upper GI symptoms occurred in FMT group. Baseline Beta-diversity similar between groups, but FMT group beta-diversity significantly increased at 6 and 13-weeks. Objective motor behaviors/symptoms showed transient improvements, with subjective symptom improvements reported compared to baseline in FMT group.
La Reau et al.259	PD patients and healthy donors	N = 11 male, N = 4 female	PD + FMT	FMT delivery occurred within 1-hour of fecal sample collection and was done via colonic route (N = 10) or nasointestinal route (N = 5). FMT occurred once.	Scores for PSQI, HAMD, HAMA, PDQ-39, NMSQ and UPDRS-III significantly decreased (i.e., improved) after FMT treatment. 5 cases had adverse events, including diarrhea (N = 2), abdominal pain (N = 2) and flatulence (N = 1)

FMT = Faaecal microbiota transplant, CNS = central nervous system, AD = Alzheimer's Disease, WT = Wild type, ABX = antibiotics, MMSE = Mini Mental State examination, MoCA = Montreal Cognitive Assessment, PD = Parkinson's disease, MPTP = 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), PBS = Phosphate Buffered Saline, ALS = Amyotrophic Lateral Sclerosis, MS = Multiple Sclerosis, HD = Huntington's disease, PSQI = Pittsburgh Sleep Quality Index, HAMD = Hamilton Depression rating scale, HAMA = Hamilton Anxiety rating scale, PDQ-39 = Parkinson's disease Questionnaire, NMSQ = Non-motor symptom questionnaire and UPDRS-III = Unified Parkinson's Disease Rating Scale – III

3.2.1. Parkinson's disease (PD)

Early indications that the gut may be involved in PD including constipation and gastrointestinal issues (GI) preceding motor impairments,²⁷⁰ and the dual-hit hypothesis,²⁷¹ led to researchers investigating the role of the microbiota-gut-brain axis in PD. Since then, it has been shown that gut microbiota can metabolize PD medication,²⁷² and several reviews and meta-analyses have surmised that 1) the gut microbiome differs between healthy controls and people with PD (at a compositional and functional level) and 2) have proposed potential mechanisms underpinning the involvement of the gut microbiota in PD pathophysiology.^10,40

Observational and mechanistic studies have enabled some insight into the mechanisms of the microbiota-gut-brain axis in PD pathophysiology. Broadly, SCFA (particularly butyrate) producing bacteria appear to be reduced in PD compared to healthy controls, and a greater abundance of pro-inflammatory bacteria and an increase in bacteria that may degrade the gastrointestinal wall is observed in PD compared to controls.^10,40 However, most prior research uses 16S sequencing, limiting the functional inferences that can be drawn, reducing taxonomic resolution compared to shotgun sequencing, and meaning species or strain level analysis is not typically possible [strain-level analysis is also dependent on shotgun sequencing depth].^273,274 Wallen et al., ²⁷⁵ conducted shotgun sequencing on 490 PD, and 234 controls and showed depletion of butyrate producing bacteria in PD compared to controls, but did not find statistical significance for elevation of Akkermansia in PD compared to controls, as prior research had.^276,277 Lower levels of fecal SCFAs have also been observed in PD compared to controls.^278,279 However, in the plasma, PD patients had elevated levels of SCFAs.^278,280 Of note, it is difficult to interpret SCFA concentrations (fecal or plasma) as true reflections of microbial production, given the utilization of SCFAs by colonocytes, alterations in gastrointestinal absorption and processing by the liver.²⁸¹ Such research may indicate a breakdown in the typical use of SCFAs by colonocytes, subsequently altering SCFA levels, rather than SCFA concentrations being influenced via production alone. While Wallen et al. do not report PD symptoms or severity,²⁷⁵ other research indicates participants with more PD non-motor symptoms correspond with lower levels of the butyrogenic Butyricimonas synergistica.²⁸² However, Nuzum et al. ²⁸² did not observe a statistical difference of butyrogenic bacteria abundance between PD and control groups, as observed by Wallen et al.,²⁷⁵ The authors postulate this may be due to PD severity level differences, as the cohort in Nuzum et al., ¹⁰ had on average fewer years since PD diagnosis and less severe symptoms when compared to the studies included in a prior systematic review.¹⁰

Longitudinal research in PD and the gut microbiome has also now been conducted,^283–286 and research covering PD severity,²⁸⁷ has begun to shed light on causality. Cilia et al., ²⁸⁴ conducted a 3-year prospective study and found that lower abundance of Roseburia at baseline was associated with worsening motor (via Hoehn and Yahr staging) and non-motor symptoms (via non-motor symptom questionnaire) at the 3-year follow-up. Similarly, Lubomski et al., showed a underrepresentation of butyrate producing genera across 0, 6 and 12 month timepoints when comparing PD and control groups,²⁸⁵ however, these findings were not replicated in prior longitudinal research. Minato et al., showed lower counts of Bifidobacterium were associated with worsening PD symptoms,²⁸⁶ and Aho et al., did not find an association between bacterial genera and progressing disease severity.²⁸³ As, these papers spanned only a few years and all used 16S sequencing, longer timeframes, and the use of shotgun sequencing to assess functional differences with time or between stable vs. progressed PD phenotypes, is warranted. Huang et al. assessed early PD, REM sleep behavior disorder (RBD), first-degree relatives of RBD, and control participants.²⁸⁷ Investigating across these different groups aligns with the brain/body-first hypothesis of Parkinson's disease,^288,289 as isolated RBD was aligned with a body-first progression subtype,²⁸⁹ and allows for insights to be drawn on the early stages of Parkinson's disease development, potentially prior to onset of motor symptoms. Huang et al., showed that abundance of two different genera, Faecalibacterium and Butyricicoccus, were reduced with increasing α-synucleinopathy with the RBD group then the early PD group having the second-highest and highest amounts of α-synucleinopathy.²⁸⁷ While other research did not show iRBD groups had decreased SCFA producing bacteria, they did contain increased Akkermansia, which may relate to a potential route of PD pathophysiology through impaired intestinal barrier function.²⁹⁰ Indeed, both a lower abundance of SCFA producing bacteria, and a greater abundance of mucin degrading bacteria predicted increases in disease severity/progression among people with PD.²⁹¹ This led the authors to postulate that individuals with an altered gut microbiome, may progress through PD faster than those with no disturbances.

In addition to the involvement of SCFAs in PD pathophysiology, animal experiments indicate curli producing E.coli bacteria results in increased alpha-synuclein deposition in the brain and gut of animals, with the same animals having increased expression of markers of inflammation (i.e., TLR2, IL-6 and TNF).²⁹² Similar research with Desulfvibrio bacteria, harvested from individuals with PD and transplanted into C. elegans, showed increased amounts of, and larger, alpha-synuclein aggregates compared to the same bacteria taken from healthy individuals.²⁹³ FMT in animal models has also been utilized in PD research with daily FMTs (from control mice) administered over 2-weeks to a rotenone induced model of PD in mice, subsequently restoring gut microbiota, reducing colon LPS levels, and improving GI dysfunction and motor impairments observed prior to FMT.²⁵⁰ Similar findings were observed in a 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model, with FMT attenuating the MPTP induced gut microbiota changes.²⁵¹ Authors also completed FMT from PD to control mice, eliciting motor dysfunction and neurotransmitter loss.²⁵¹ Fecal samples taken from healthy people and those with PD, were used in FMT to control and PD model mice (MPTP model of PD).²⁵² Daily FMT (10-days) from people with PD to MPTP-mice worsened motor impairments, dopaminergic neurodegeneration, and colonic inflammation, while FMT from healthy people improved motor behavior, slowed dopaminergic neurodegeneration, and relieved colonic inflammation. These results indicate the causal role the gut microbiota has in PD pathophysiology. Preliminary FMT studies in people with PD have been conducted,^258,259 with research indicating that FMT is well-tolerated, with some mild-to-moderate adverse events reported (E.g., GI disturbances). Dupont et al., ²⁵⁸ administered the FMT orally (N = 12; placebo group = 12) showing temporary improvements to objectively assessed motor function. Xue et al., ²⁵⁹ administered the FMT via nasal-jejunal tube (N = 5) or via colonoscopy (N = 10), finding that the colonic route offered greater improvements to PD symptoms which persisted for longer than in the naso-intestinal route. Lastly, animal model research on α-synuclein and gut microbiota, indicated that α-synuclein pathology alters the ENS and gut microbiome, but those gut microbiota alterations can be modulated by exercise.²⁹⁴ In addition to FMT, exercise in PD should be investigated for its therapeutic benefits, and capacity to modulate the gut microbiota.

3.2.2. Amyotrophic Lateral Sclerosis (ALS)

Like PD, ALS patients often suffer from constipation. Whether constipation precedes ALS diagnosis, as it can in PD, is not currently known.²⁹⁵ Non-motor functions in ALS are not well understood, however, prior hypotheses have implicated the microbiota-gut-brain axis in ALS progression/onset.²⁹⁶ Animal models have shown correlation(s) between ALS and gut microbiota composition, with ALS mice having significant gut microbiota composition differences compared to controls.²⁹⁷ Researchers also took bacteria associated with ALS and mono-inoculated ALS model mice, showing that Parabacteroides distasonis and Ruminococcus torques were associated to increased disease severity.²⁹⁷ Authors also observed reductions in Akkermansia muciniphila aligning with ALS progression.²⁹⁷ Research using SOD1^G93A mice conducted longitudinally with and without butyrate treatment showed that without butyrate treatment gut bacterial differences in the SOD1^G93A mice compared to wild-type were observed, whereas less differences were observed in the butyrate-treated group.¹⁴⁸ Importantly, butyrate treatment lead to greater motor task performance, and prior research indicates butyrate treatment may delay ALS progression in mice,²⁵³ potentially highlighting that reduced butyrate production is involved with ALS progression.

Research in patients with ALS is sparse, with inconsistent findings. Several bacteria at the genus level were reportedly more abundant in ALS patients compared to controls ²⁹⁸ (including pro-inflammatory bacterial genera Escherichia and Shigella, and Akkermansia, which opposes animal-model work).²⁹⁷ Other research has found some differences at a genus level, including the genera Subdoligranulum [which contains the butyrate producer Subdoligranulum variabile ²⁹⁹] but no fecal SCFA differences were observed between cohabiting controls and people with ALS.³⁰⁰ Authors also reported people with ALS had lower cytokine levels (IL-15, IL-8, MCP-1 and VEGF-A) compared to controls. Separate research reveals no difference in diversity, composition or function (via PICRUSt1) between ALS and controls.³⁰¹ Additional research that has observed differences between control and people with ALS had small sample sizes (N < 6 per ALS group), limiting interpretability and comparability to larger studies.^302,303 Overall, a lower abundance of butyrate producers, and increased proinflammatory bacteria may be involved in ALS progression. Impairment of the intestinal, blood-spinal cord,³⁰⁴ and blood-brain barrier, have also been reported in ALS patients, and may relate to microbiota-gut-brain axis involvement in ALS pathophysiology.

3.2.3. Multiple sclerosis (MS)

A systematic review of case-control observational studies of MS and the microbiota-gut-brain axis identified ten studies (all using 16S sequencing), with none of the eight that investigated alpha diversity finding a difference between MS and control groups.³⁰⁵ At a genera and species level, ≥2 studies observed lower abundance of Prevotella, Faecalibacterium prausnitzii (with one study showing increased Faecalibacterium comparing MS to controls), Bacteroides coprophilus, Bacteroides fragilis, and a higher abundance of Methanobrevibacter and Akkermansia muciniphila in MS cases versus controls. While there are concordant differences observed, the evidence is not universally consistent but nevertheless indicates the microbiota-gut-brain axis involvement in MS. A different review also highlights the influence the gut microbiome has on immune function and the relevance of this in relation to MS pathophysiology.³⁰⁶

FMT studies in MS mice models indicate trends of FMTs to ameliorate some of the observed gut microbiome differences between MS mice and controls.²⁵⁴ FMT also led to reduced activation of microglia and astrocytes while conferring some protection on the BBB, myelin and axons in the MS mice models.²⁵⁴ These findings showcase the potential mechanisms of the microbiota-gut-brain axis influencing MS, and highlight the usefulness of FMT for MS. Several case studies of FMT in people with MS have indicated its potential safety and efficacy in addressing MS symptoms.^307,308 A pilot FMT trial in MS patients (N = 9) were given monthly FMTs for 6-months.³⁰⁹ Authors indicated that FMT was safe for this population, however this study was underpowered and stopped early (see paper for details), therefore potential benefits of FMT cannot be determined currently.

Several papers have assessed MS and the microbiota-gut-brain axis longitudinally.^310–312 Different findings were observed across all papers, with some broader community differences observed dependent on disease severity or whether it worsened or not,³¹¹ but no distinct bacterial communities were identified between MS and control groups, with some bacterial composition differences observed (lower Bifidobacterium longum, Clostridium leptum, and Faecalibacterium prausnitzii in MS compared to controls).³¹⁰ In terms of composition, findings also appear to be specific to disease severity or worsening status, as lower levels of Akkermansia were observed in disease progressors compared to non-progressors, and lower Akkermansia abundance correlated to worsening MS progression.³¹²

3.2.4. Huntington's disease (HD)

Evidence of the microbiota-gut-brain axis’ potential to affect HD pathophysiology included plasma metabolome differences of gut-derived metabolites between controls and premanifest HD individuals or early symptomatic HD individuals.³¹³ Research comparing HD mice models to wild-type controls showed gut bacteria compositional differences at the phylum level, and increased diversity in only male HD mice compared to controls.³¹⁴ In a follow-up study authors longitudinally assessed fecal microbiome data for wild-type and HD mice models via shotgun sequencing,²⁵⁵ and at phylum and family levels no gut bacterial differences were observed between groups, across any time point. However, between group differences were observed for gut microbiome function at week 12 (prior to motor symptom onset) with a greater relative abundance of sulfur metabolism, lysine degradation, glutathione metabolism and butanoate metabolism in the HD compared to wild-type mice. Greater butanoate metabolism did not correspond to greater plasma butyrate levels in the HD group (in fact authors note a non-significant [p = 0.11] decrease in the HD mice), nor other SCFAs. Authors discussed that pathogenic butyrate producers, such as Fusobacterium, may preferentially use lysine over pyruvate for butyrate production, which releases toxic byproducts, i.e., ammonia,³¹⁵ with this perhaps relating to the increased lysine degradation pathways they observed. The same lab also conducted a fiber intervention in HD mice,²⁵⁶ reporting several benefits of the high-fiber diet to HD mice, including improvements to affective and cognitive deficits and improved GI function. Within the HD mice groups exposed to the different diets (no fiber, control, and high fiber), there were no specific changes to the gut microbiome (study used 16S sequencing) observed at 14 or 20 weeks of age. Comparisons between wild-type and HD mice exposed to the high-fiber diet revealed gut microbiome abundance differences observed at 14-weeks only at the phyla and family level. Examples of these included Actinobacteriota and Proteobacteria being decreased in the HD mice in the high fiber group compared to the high fiber wild-type mice. At the family level Bacteroidaceae, Butyricicoccaceae, and Ruminococcaceae had increased abundance in the high fiber HD mice when compared to wild-type high fiber mice. Additionally, FMT from wild type to R6/1 HD model mice positively altered cognitive outcomes.²⁵⁷ This indicates the potential for FMT to effectively treat some of the symptoms associated with HD while also reinforcing the likely involvement of the gut microbiota within HD.

Comparisons of HD patients to controls revealed, that the HD group had lower diversity, while sex-specific differences at the phylum and family level, were only observed for males.³¹⁶ Several inferred functional pathways (via PICRUSt) were indicated to be involved in HD; Starch degradation V, Methylerythritol phosphate I and II, and NAD biosynthesis I were all greater (in terms of relative abundance) in HD group. Research investigating the gut microbiome and cytokine levels of HD patients and controls did not reveal any sex-specific effects observed previously, but found several genera (Intestinimonas, Bilophila, Lactobacillus, Oscillibacter, Gemmiger, and Dialister) had greater abundance in the HD compared to control group.³¹⁷ Additionally, Intestinimonas was positively correlated with total functional capacity scores, while Lactobacillus was negatively associated with MMSE score. Across 10 plasma inflammatory markers only IL-4 differed, being greater in the control compared to HD group. Associations between bacteria and cytokines were completed in the HD group only, with Intestinimonas positively correlated with plasma IL-4, and Porphyromonas positively correlated with IL-4, IL-10, and IL-13, while Bilophia, Oscillibacter and Gemmiger negatively associated to IL-6. A review that details much of the above discussed evidence for involvement of the microbiota-gut-brain axis in MS, and incorporates a deeper discussion of potential mechanisms can be found here.³¹⁸

4. Conclusions and Future directions

The microbiome-gut-brain axis research field has expanded horizons significantly since the initial focus on conducting compositional and correlational analyses. While such studies provided important impetus to the field, the past decade has answered the calls for more mechanistic studies and started to address the need to translate promising studies from animal models to humans. Collectively, these studies have further shown the importance of the gut microbiome to neurodevelopment and neurodegeneration.

These advances have seen a proliferation in clinical research in neurodevelopmental disorders including ASD, ADHD and schizophrenia. This research has largely resonated with the insights from earlier animal research, linking time-window specific microbiome configurations to infant temperament and cognition, and animal models work continues to document mechanistic insights, including the role of the vagus nerve.⁷⁵ It is anticipated that animal research will ultimately accrue the information necessary to pinpoint of targeted temporal microbial manipulations in neurodevelopmental disorders,⁶⁵ although there are still many pieces missing in this puzzle .

In neurodegeneration research, there is now strong causal evidence linking the gut microbiome to symptom expression in Alzheimer's,²⁴⁶ and PD ²⁵² using human-to-animal FMT. Such approaches could form the basis of new gut-brain animal models of these disorders to allow the assessment of possible microbiota interventions. A greater amount of clinical work investigating Alzheimer's, PD, ALS, HD, and MS has also been conducted, with longitudinal research becoming more prevalent. Some burning questions remain, including the interface between biological sex and the gut microbiome in these disorders, the role of host genetics, the gut virome's role in neurodegeneration including PD,³¹⁹ and whether the gut microbiome during prodromal stages could be predictive of future psychopathology and be amenable to prevention strategies.³²⁰ Moreover, the issue of psychiatric comorbidity and possible transdiagnostic microbial profiles also requires further resolution.

While the recent research advances represent important progress, there are still fundamental knowledge gaps and obstacles to overcome. For example, there is still no consensus on what a ‘healthy microbiome’ is, either during adulthood or at the extremes of life, and developing this understanding is paramount to exploiting the diagnostic and therapeutic potential of recent observations. Future neurodevelopment research should include more translational human intervention research, such as evaluating the potential of probiotics, prebiotics, postbiotics or FMT interventions. Understanding and overcoming the undesirable impacts of factors disrupting the gut microbiome such as maternal and infant antibiotic usage, and an increased focus on researching the potential benefits and safer application of ‘seeding’ an infant's microbiome via the mother's vaginal microbiome following C-section are warranted.⁸¹ While ALS, HD and MS research in the microbiota-gut-brain axis is expanding, a clearer focus on mechanism and causation is needed, akin to that demonstrated in in Alzheimer's and PD, and overall, the neurodegeneration field would benefit from longitudinal research assessing humans in prodromal phases prior to and after disease diagnosis, to define disease-relevant gut microbiota changes. In these complex disorders, the microbiota is increasingly recognized as a key pathophysiological component. Knowing how and why the gut microbiome changes at the extremes of life will further develop our mechanistic understanding and help build the evidence base as we strive toward counteracting microbial missteps with precision therapeutic interventions.

Funding Statement

APC Microbiome Ireland is a research centre funded by Science Foundation Ireland (SFI; (SFI/12/RC/2273_P2)). NN receives project funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 101034270.

Disclosure Statement

SKS has received speaker's and authors’ honoraria from Medice Arzneimittel Pütter GmbH Co KG, Janssen and Takeda. J.F.C. has spoken at conferences organized by Bromotech and has received research funding from Reckitt, Nutricia, Dupont/IFF, and Nestle. G.C. received honoraria from Janssen, Probi, and Apsen and research funding from Pharmavite, Fonterra, Reckitt, Tate and Lyle and Nestle and is or has been a paid consultant for Yakult, Zentiva and Heel Pharmaceuticals. This support did not influence or constrain this review. The other authors report there are no competing interests to declare.

Data availability statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.