Deviation from a healthy microbiome is commonly described as dysbiosis ^[1][66]. Dysbiosis can be defined as an altered composition and diversity of the gut microbiome that is associated with ill-health and disease. More specifically, dysbiosis disrupts immune suppressing host–microbe interactions and promotes inflammation, pathogenicity, and disease. With prolonged dysbiosis, there is an overgrowth of pathogenic bacteria with virulence factors, decreased regulatory or beneficial bacteria, and loss of bacterial diversity ^[2][72], which leads to dysfunction in the immune response and the promotion of a proinflammatory environment ^[1][3][65,66].

With dysbiosis, the homeostatic bacterial metabolic functions are altered, which can lead to a defective mucosal barrier ^[3][65]. It has been proposed that an increase in mucolytic bacteria such as Ruminococcus gnavus ^[4][73] and reduction in antimicrobial peptides such as Paneth cell-derived alpha-defensins, compromise the integrity of the mucosal barrier and are contributing factors in conditions such as inflammatory bowel disease (IBD) ^[5][74]. These factors also increase the risk of microbial overgrowth and bacterial translocation, which can further activate the host immune response via modulation of proinflammatory signalling pathways ^[3][65].

Pathobionts, or bacteria that display pathogenic properties opportunistically, can express virulence factors ^[6][75]. One example is the release of endotoxins, which act as ligands that bind cell surface receptors on innate immune cells, initiating cell activation ^[7][76]. Activated immune cells then upregulate their expression of pattern recognition receptors, which recognise microbe, pathogen, and danger-associated molecular patterns released by microbes and inflammatory cells ^[1][66]. In conjunction with reduced immune suppression from commensal bacteria, pathobionts lead to dysregulation in T cell differentiation that promotes the release of proinflammatory cytokines ^[8][77]. The resulting inflammatory environment induces host tissue damage and predisposes to diseases including IBD, irritable bowel syndrome, colorectal cancer, and extraintestinal diseases such as obesity and diabetes ^[9][10][11][78,79,80]. Table 1 outlines examples of several pathobionts and their molecular mediators suggested to contribute to disease.

A number of indices have been proposed to measure and assess dysbiosis. These include large-scale bacterial marker profiling, relevant taxa-based methods, neighborhood classification, random forest prediction, and combined alpha-beta diversity ^[21][22][23][89,90,91]. However, the most common method to measure and assess dysbiosis is to measure diversity. Alpha diversity describes the amount of unique taxa (richness) and their distribution (evenness) within a microbial community. Beta diversity is used to assess differences in microbial community composition between individuals, such as between those with a disease state and healthy controls ^[21][89]. Overall, multiple indices suggest that a single measure fails to address the broad concepts within dysbiosis. Furthermore, these measures cannot distinguish whether dysbiosis is a cause or consequence of disease ^[24][25][26][59,92,93].

A prominent inflammatory disease is IBD, which is a chronic inflammatory condition of the GI tract with episodes of relapse and remission ^[3][65]; it is predominantly classified into Crohn’s disease (CD) and ulcerative colitis (UC). CD can affect the entire GI tract with transmural involvement, while UC is characterised by a more superficial inflammation that is confined to the large intestine ^[27][94]. The prevalence of IBD has increased in Western countries, with a rising incidence in newly industrialised countries ^[28][71]. Up to 25% of patients present with symptoms during adolescence and young adulthood ^[29][95]. Children typically develop a more aggressive disease course ^[30][96], with detrimental effects on growth, development, and psychosocial wellbeing ^[28][71]. The pathogenesis of IBD is multifactorial, with interactions between host genetics, a dysregulated immune response, environmental exposures, and importantly, changes in the gut microbiome contributing to disease ^[3][65].

Different factors can alter the intestinal microbiome ^[31][70]. Diet, including feeding from birth, is a notable environmental factor that profoundly impacts microbiome composition. For instance, the microbiome of formula-fed infants is characterised by decreased diversity and bacterial richness, which is associated with inflammation ^[32][97]. Bolte et al. ^[33][98] also found that the Westernised diet, consisting of lower dietary fibre and higher consumption of animal fat, salt, and sugar ^[34][99], induces microbiome characteristics that correspond with intestinal inflammation, while plant-based diets were associated with SCFA-producers and lower abundance of pathogenic bacteria. These observations are also supported by Zheng et al. ^[35][100] who found correlations between the abundance of several bacterial species and a proinflammatory diet (defined as a diet containing dietary components known to be associated with inflammation). Two further studies have also found associations between IBD and the consumption of high-sugar foods ^[36][37][101,102].

Antibiotic exposure is an additional environmental factor that can induce a loss of health-promoting bacteria and reduced expression of antibacterial agents and immunoglobulins, resulting in the potential to increase susceptibility to infections ^[38][55]. Antibiotic use corresponds with an increase in inflammatory cytokines, alteration of insulin sensitivity, and modulation of the metabolism of SCFAs and bile acids ^[39][103]. A cohort study conducted by Hviid et al. ^[40][104] reported associations between antibiotic exposure during childhood and the development of IBD. Additionally, maternal antibiotic use during pregnancy and infantile antibiotic exposure increases the risk of very early onset IBD ^[26][41][93,105]. Antibiotic use in infants small for gestational age showed an increase in pathogenic species associated with a ‘dysbiotic’ microbiome ^[42][43][13,106]. In addition to antibiotic exposure in premature and small-for-gestational-age infants, other early life factors including Caesarean section delivery and formula feeding increase the risks of asthma, atopy, obesity, type 2 diabetes mellitus, necrotising enterocolitis, and sepsis ^{[42][44][45][46][47][48][49][50]}[5,6,13,36,37,107,108,109].

Significant efforts have been directed at identifying the pathogenic microbiome characteristics of inflammatory disease and in particular IBD ^{[51][52][53][54][55][56][57]}[110,111,112,113,114,115,116]. Most studies have identified a decrease in alpha diversity in individuals with CD compared to UC and healthy controls ^{[52][54][55][56][57]}[111,113,114,115,116]. However, a systematic review by Pittayanon et al. ^[51][110] reported that the most common findings of a decrease in the species Faecalibacterium prausnitzii and an increase in Enterobacteriaceae, were characteristic of IBD. F. prausnitzii is a SCFA-producer with metabolomic studies showing reduced SCFAs in the GI tract of individuals with IBD ^[53][112].

Furthermore, SCFA-producers, such as F. prausnitzii and Roseburia hominis, reduce proinflammatory cytokines including interleukin (IL)-12 and interferon-gamma and increase anti-inflammatory cytokines including IL-10 ^[24][59], implying that reduced SCFA-producing bacteria may be a marker of the pathogenic microbiome. Enterobacteriaceae, such as E. coli and Shigella, are facultative anaerobes and are enriched in inflammation at the expense of obligate anaerobes, which may be due to oxidative stress during inflammation ^[22][90]. These bacteria produce lipopolysaccharides (LPS), a pathogen-associated molecular pattern that activates Toll-like receptor signalling and a consequent inflammatory cascade ^[12][81].

As an illustration of this, a study of GF mice found that the mice developed colitis when colonised by a highly endotoxic intestinal microbiome containing a high proportion of Enterobacteriaceae (including E. coli) and a low proportion of Bacteroidetes ^[58][117]. Pathogenic strains of E. coli, including adherent-invasive E. coli (AIEC), have also been reported to be associated with IBD ^[59][118]. AIEC can induce the expression of cell adhesion molecules and they possess virulence factors such as type 1 pili and long polar fimbriae, which can facilitate colonisation in the intestinal mucosa ^[13][82]. Although E. coli can be considered part of the normal GI microbiome, a pathogenic and disproportionate abundance of E. coli may also constitute a pathogenic microbiome.

A caveat is that most studies on the microbiome in IBD are cross-sectional or longitudinal over short periods such as one year. Therefore, it is difficult to determine whether the previously described microbiome characteristics are a cause or consequence of disease. Furthermore, the GI microbiome is not homogenous, with reported differences in the microbiome between faecal and biopsy samples ^[12][81]. In addition, IBD is a heterogenous disease and findings for individuals with mild disease may not be valid for individuals with moderate or severe disease ^[52][111].

A further consideration of the contribution of the microbiome to disease is age. As previously described, in early childhood the gut microbiome is generally considered to be similar in composition and diversity to that of adults ^[60][119]. However, some studies report differences in the microbial composition between younger children and adolescents, suggesting continual development of the microbiome with age ^[32][61][62][62,97,120]. Furthermore, studies of the microbiome of treatment-naïve children with IBD suggests that the microbiome is characterised by reduced abundances of certain bacterial species rather than an increase in pathogens ^[22][63][64][63,90,121]. This highlights that in children at least, the pathogenic microbiome may be characterised by a loss of physiological functions.

However, a systematic review of gut microbiome profiles of children with IBD, commonly reported decreased alpha diversity and beta diversity that differed from healthy controls ^[28][71]. Overall, the IBD microbiome in children showed increased Enterococcus and decreased Anaerostipes, Blautia, Coprococcus, Faecalibacterium, Roseburia, Ruminococcus, and Lachnospira ^[28][71]. As these findings are also similar to adult IBD, it was proposed that the conditions for adult-onset IBD may be established during childhood ^[28][71].

Associations between the microbiome and disease severity or outcomes have also been investigated. Olbjørn et al. ^[62][120] reported that patients with higher abundance of Proteobacteria were more likely to require aggressive treatment and surgery. Proteobacteria abundance also correlated with increased CD complications and an absence of mucosal healing ^[62][120]. It is of interest that E. coli is classified within the Proteobacteria phylum. However, no firm conclusions can be reached whether specific species within the Proteobacteria phylum, or whether a common characteristic shared by the phylum, is contributing to these findings ^[62][120].

Paediatric studies have also investigated associations between dysbiosis, disease activity and treatment response ^{[23][28][62][63][64][65][66]}[63,71,91,120,121,122,123]. Overall, dysbiosis positively correlated with disease activity in children ^[28][66][71,123]. Of interest was that hydrogen sulfide (H2S)-producers, such as Fusobacterium, Prevotella, and Streptococcus, were increased with inflammation ^[28][71]. H2S can damage the intestinal epithelium and influence the microbiome, mucous, and biofilm interactions ^[28][71]. Therefore, H2S and H2S-producers should also be a consideration of the pathogenic microbiome.

Contrary to these findings is that de Meij et al. ^[64][121] found no association between microbiome profile and disease activity. Further studies have also reported that with treatment, bacterial profiles and dysbiosis remain ^[23][62][91,120]. However, these studies also highlighted that consideration should also be given to the type of treatment, and/or treatment success. Kolho et al. ^[65][122] reported that microbial composition and diversity increased in treatment responders, following six weeks of anti-tumour necrosis factor-α therapy. Specific changes associated with treatment response have also been identified with Eubacterium and Bifidobacterium associated with a favorable response to medication ^[65][67][122,124]. A further study of the oral microbiome also found associations between bacterial species and treatment response ^[68][125]. These findings support the idea that analysis of the microbiome may potentially assist in the assessment and management of IBD.

A further factor that requires consideration is that the microbiome itself may be contributing to dysbiosis. The essence of this hypothesis is that symbiont bacteria can potentially evolve into pathobiont bacteria under specific selective pressures ^[69][126], which is feasible due to the rapid reproductive cycle of prokaryotes. The potential for a rapidly evolving microbiome is illustrated by the example that there are approximately the same number of reproductive events occurring in the average human microbiome in 5 days, as in the entire human population over the last 66 million years ^[70][127].

In recognition of these factors, the concept of microbiome engineering has emerged recently ^[71][128]. These include primary preventive strategies such as education on the consequences of Caesarean section delivery, conservative use of antibiotics during pregnancy, administration of antibiotics after cord clamping to limit foetal exposure and the adoption of the WHO Baby Friendly Hospital Initiative ^[72][9]. Current secondary prevention methods include pre- and probiotics supplementation of the mother during pregnancy and the neonate after birth, faecal microbiome transplantation and phage therapy ^[69][126]. At present, the literature surrounding neonatal microbiome engineering is relatively unexplored but poses a pivotal opportunity to promote robust microbiome development and prevent long-term pathologies associated with a disrupted microbiome.