Academic Editor: Arthur J. Chu
Disorders such as inflammatory bowel disease (IBD) and celiac disease (CeD) result in intestinal hyperpermeability or ‘leaky’ gut. The increased permeability of the intestinal barrier allows microbial metabolites, toxins, and pathogens to infiltrate the bloodstream and extraintestinal tissues, causing systemic inflammation. Despite differences in aetiology and pathophysiology, IBD and CeD share several extraintestinal manifestations such as neuroinflammation, neurological and psychiatric manifestations, and sensorineural hearing loss (SNHL). This narrative review focuses on the association between intestinal hyperpermeability with the brain and inner ear diseases. We postulate that the microbial metabolites and pathogens released from the gut increase the permeability of natural barriers, such as the blood-brain barrier (BBB) and blood-labyrinth barrier (BLB). The barrier breakdown allows the spreading of inflammatory processes to the brain and inner ear, leading to disease.
If an inflammatory response in the gut does not naturally resolve, it may lead to a state of chronic inflammation. This development may result in pathologies such as inflammatory bowel disease (IBD) and celiac disease (CeD) [1]. Evidence shows that these conditions are associated with a pathological shift in gut bacteria [2, 3, 4].
An imbalance of intestinal flora results in gut dysbiosis, bringing about changes to the permeability of the intestinal barrier (IB) [5, 6]. Consequently, pathogens can then infiltrate the circulation, enabling them to spread to other organ systems, thus resulting in secondary extraintestinal infections, which are often life-threatening [5]. Indeed, both IBD and CeD have been linked with extraintestinal manifestations (EIMs), including neuroinflammatory diseases and sensorineural hearing loss (SNHL) [7, 8, 9]. However, despite reported associations between diet and hearing loss, the current literature does not recognise IBD-induced gut dysbiosis as an aetiology of SNHL. Instead, it advocates that SNHL in IBD has an autoimmune background [10].
IBD is an umbrella term used to label two disorders that involve chronic inflammation of the gastrointestinal tract (GIT): Crohn’s disease (CD) and ulcerative colitis (UC) [11]. Interestingly, IBD and sub-clinical manifestations of the inflammatory gut disease have been associated with gut dysbiosis and significantly increased levels of bacterial plasma components such as lipopolysaccharide (LPS) [12, 13, 14]. These findings provide evidence for IB hyperpermeability. In both IBD and CeD, bacterial metabolites leak from the intestinal lumen into the bloodstream, where they can potentially infiltrate the brain, producing local neuroinflammatory processes [15, 16] and other neurological conditions, including the so-called “celiac brain” [15]. Similarly, we previously postulated that, in gut dysbiosis caused by a high-fat diet (HFD), systemic immune responses might enhance the permeability of the blood-labyrinth barrier (BLB), thus causing cochlear inflammation following the infiltration of inflammatory cells and cytokines and the deposition of immune complexes [17]. However, IBD and CeD do not rely on HFD to trigger increased IB permeability. Instead, these disorders are multifactorial “leaky gut” diseases resulting from immune system malfunctioning and autoimmunity [18, 19, 20, 21, 22].
This review entails the existence of a gut-inner ear axis that links IBD and CeD with SNHL. This concept is analogous to the gut-brain axis, linking inflammatory gut diseases with brain disorders.
A broad literature search spanning from 1998 to 2021 was conducted using PubMed, Google Scholar, and Embase medical databases. References from the relevant papers were used. Following Boolean search logic, the main keywords included: “inflammatory bowel disease” OR “Crohn’s disease” OR “ulcerative colitis” OR “celiac disease” OR “coeliac disease” OR “gut dysbiosis”) AND (“hearing loss” OR “sensorineural hearing loss” OR cochlea OR “blood-labyrinth barrier” OR “blood-brain barrier” OR inflammation OR lipopolysaccharides). The search results were subsequently examined according to their relevance to this review. Only publications in the English language were included.
Since first described by Samuel Wilks in 1859, IBD has been associated with altering the gut microbiome [23, 24]. Even though the aetiology of IBD is not yet fully understood, it has been tightly linked with IB hyperpermeability (“leaky gut”) [25], similar to CeD [2, 26]. Consequently, these conditions are often referred to as “leaky gut disorders” [27, 28, 29].
A compromised IB can result in systemic pathological processes, such as increased oxidative stress (OS) [30], inflammation [31], and decreased insulin sensitivity, affecting various organs and tissues [32]. Interestingly, it appears that CeD represents a risk factor for IBD and vice versa [33]. Pinto-Sanchez et al. [34] found a 9-fold increase in the risk of developing IBD in CeD patients compared with the control population.
However, CeD and IBD are very different in their aetiology and pathophysiology. Nevertheless, they share similarities, such as gut dysbiosis, increased IB permeability, and inflammatory responses. In addition, both diseases present with similar EIMs, including neuroinflammation, neurodegeneration, and hearing loss. Based on the similitudes between CeD and IBD, it is prudent to assume their EIMs might have comparable pathophysiology. Here, we discuss the links between these gut diseases and the potential mechanisms of their EIMs.
The intestinal wall coating is a single epithelial layer that connects the host
to the external environment, known as the IB. Intestinal epithelial cells (IECs)
are responsible for maintaining a functional barrier. The lining of intestinal
epithelia exhibits distinctive intercellular connections known as tight junctions
(TJ), desmosomes, and adherent junctions. These connections have a role in the
selective permeability of gut lining by permitting the passage of nutrients and
fluid absorption while preventing the displacement of microbial metabolites and
antigens from the gut [35, 36]. TJs function as an active structural barrier in
the paracellular space [36]. TJ protein composition includes claudins, occludins,
and junctional adhesion molecules (JAM) [35]. The cytokines that regulate the
immune system, such as tumour necrosis factor-
The mammalian immune system comprises two integrated subsystems: the innate and
adaptive (Fig. 1). The innate immune system is the first to respond to pathogens
and is evolutionarily older. If the pathogenic challenge persists, the adaptive
immune system will engage with the pathogen with specificity and memory [40, 41, 42].
The innate immune system in the gut includes several physiological barriers that
protect the body from the insurgence of pathogens. These barriers include mucus,
TJs, IECs, antimicrobial enzymes, pattern recognition receptors, transforming
growth factor-
Overview of the mammalian immune system.
In the first 2–3 years of life, the microbial landscape of the gut rapidly
changes, by the end reaching similar functionality to the one seen in adults
[55]. During this period of weaning from breastmilk to solid food, the changing
microbiota leads to “weaning rejection” of the immune system. At this point,
several studies have observed shifts in the global gene expression in the
intestines, including genes encoding defensins (a major family of host defence
peptides expressed predominantly in neutrophils and epithelial cells), chemokine
receptors, and mucins [56]. At the weaning stage, commensal microbiota induces
gene expression of the pro-inflammatory cytokines (TNF-
Antibiotic administration in early childhood may impact the microbiome for life via so-called pathological imprinting [56, 58]. In South Korea, children up to two years of age receive an average of 3.4 courses of antibiotics per year [59]. In New Zealand, clinical studies demonstrated that 94% of children received at least one course of antibiotics by the age of five, with an average of eight courses by the same age [60]. Several studies have identified multiple early childhood factors that are directly associated with the risk of developing IBD later in life, such as mode of delivery [61], feeding type [62], childhood hygiene [63], and antibiotic use [64]. Antibiotics in infancy also increase the chance of developing obesity [65] and CeD [66], both linked with gut dysbiosis.
IBD denotes diseases that involve chronic inflammation within the GIT, with two apparent phenotypes [67]. In the first phenotype (CD), the inflammation can develop in any part of the GIT, often characterised by patchy and transmural damage through the IB. The second phenotype (UC) involves confluent inflammation in the colonic mucosa [68]. IB impairment leads to a leaky gut in both cases [11]. IBD is considered by many as having an autoimmune origin [69]; however, some authors have suggested a novel, autoinflammatory background [70].
IBD’s aetiology is multifactorial, complex, and still not fully understood. It combines genetic, environmental, and microbial factors, which influence the immune system; however, none of these factors can cause disease alone [68]. The genetic factors constitute only a relatively small proportion of IBD cases [68, 71]. Therefore, if the host carries genetic risk variants, they must be exposed to environmental or microbial challenges to develop IBD [68]. Rapidly increasing IBD incidence in the modern world indicates the significance of diet, lifestyle, and a changing environment [72].
The increased incidence of IBD parallels the “westernisation” of countries [73, 74]. IBD affects around 3.1 million people in the USA alone, with cases increasing worldwide. Between 2003 and 2013, the number of new IBD cases in NZ increased by an average of 8.1% per year [75]. The incidence of IBD in NZ among adults and children is considered very high [76, 77, 78], with NZ and Australia among the top five countries for incidence of CD [78]. In Asia, the incidence rate for UC has risen by 60% since 1988 and 70% for CD [79]. The number of new IBD cases in South Korea is one of the highest globally and rapidly increases [80].
Another inflammatory gut disorder that results in a “leaky gut” is celiac disease (CeD). It is an autoimmune condition that results from an immune response to gluten in genetically predisposed adults and children [81]. CeD affects approximately 1% of the global population [82]. Over the past 50 years, CeD prevalence has increased 4-fold [83].
The severity of the disease might be influenced by genetic and environmental factors and immune imbalance [84]. Like IBD, genetic background on its own is insufficient to develop the disorder; gluten is a crucial contributor [85]. However, gut dysbiosis may trigger pathogenic pathways leading to CeD progression [81]. In response to the build-up of gluten fragments (such as gliadin) in the intestines, the adaptive and innate immune responses lead to villous atrophy, crypt hyperplasia, and IB hyperpermeability [86].
Several established factors can alter homeostasis between gut microbiota and the immune system; however, the multifactorial aetiology of IBD is still not fully understood [40].
IBD and CeD have common immunological, genetic, and environmental factors
contributing to their manifestations. Genome-wide association studies (GWAS) have
shown that CeD and CD share genetic risk loci [87]. Moreover, CeD and IBD share
elements of aetiology such as increased IB permeability [88], compromised
regulatory T cell (Treg) function [89, 90], upregulation of pro-inflammatory
cytokines ( IL-13, IL-17, IL-21 and IFN-
In a healthy individual, commensal microbiota, IECs, and immune cells function in concinnity. When a soluble antigen enters the GIT, local immunity is suppressed due to immune tolerance. However, many factors can disturb this homeostasis between pro-, and anti-inflammatory mediators, favouring pro-inflammatory responses in susceptible individuals. Prolonged activation of the innate immune system leads to a state of unresolved, chronic inflammation, such as IBD. IB hyperpermeability in IBD allows harmful molecules (such as LPS, CpG motifs, pathogens, luminal antigens) to infiltrate the bloodstream, causing the release of pro-inflammatory cytokines, which can alter homeostasis in distant organs [59, 72, 95]. In CD, increased IB permeability has been observed prior to clinical relapse, suggesting its role in disease exacerbation [96, 97]. However, it is still unclear if the immune activation in IBD results from gut dysbiosis or loss of immune tolerance in the gut’s immune system [40].
The prevalence of genetic factors in IBD is relatively low, making environmental factors a key player. Interestingly, the incidence of IBD is higher in urban than rural areas [98, 99], and there is a clear association between diet and incidence [73, 100, 101, 102, 103, 104]. Factors that impact gut microbiota, such as early exposure to animals, having many siblings [105, 106], natural mode of delivery [61], and breastfeeding [107], effectively decrease the chances of developing IBD in contrast to the mode of delivery via C-section [61, 74], excessive paediatric hygiene [63] and early-life antibiotic therapies [64, 108, 109]. Furthermore, treatments that typically affect microbiota, such as faecal diversion and antibiotic therapy, are often used for IBD management [110, 111, 112, 113, 114, 115]. However, an extensive nested case-control analysis of the population-based University of Manitoba Inflammatory Bowel Disease Epidemiologic Database has demonstrated that antibiotic use may be a risk factor for developing IBD, as a significant number of patients included in this study had been prescribed antibiotics 2–5 years prior to being diagnosed with IBD [108]. Moreover, a random-effects meta-analysis demonstrated that antibiotic use increases the risk of CD development [116]. Interestingly, the lesions characteristic for IBD coincide with higher concentrations of commensal bacteria [117, 118], and IBD patients present higher yields of antibodies against commensal bacteria than healthy individuals [119].
Over the years, the increase in IBD incidence has been associated with the
increased consumption of the western diet, which is higher in fats and additives,
but with a decreased amount of fruits, vegetables, and fibre [59, 72, 120, 121, 122].
Even though regional differences of the western diet have been observed [72, 123, 124], dietary factors associated with that diet commonly induce gut dysbiosis in
obesity and metabolic syndrome [122, 125, 126, 127, 128]. Moreover, many pro-inflammatory
cytokines common for CeD and IBD (e.g., TNF-
The other established aspect of IBD aetiology is a failure of the immune
regulatory control in the active phases of the disease [132]. The increased
population of T-cells and increased expression of pro-inflammatory cytokines,
including TNF-
Interestingly, GF animals tend to have impaired Th17 cell development, decreased IL-17 cytokine production in the colon [137], and impaired Treg cells [138], suggesting an essential role for microbiota in IBD immunopathology. GF animals also have an altered mucus layer [139], further implicating gut microbiota in IBD pathophysiology. While healthy subjects can generally tolerate autologous microbiome, in some cases, the breakdown of this symbiosis is associated with chronic intestinal inflammation [40, 140, 141]. It was proposed that gut dysbiosis negatively impacts the interaction between the immune response and microbiome, leading to an overactivation of the immune system [40].
In comparison, dietary gluten in CeD stimulates innate and adaptive immune
systems in a susceptible individual, increasing the production of IL-15, which
plays a major role in developing inflammatory and protective immune responses to
microbial invaders and parasites. IL-15 also causes epithelial cell death in the
gut and increases IB permeability, enabling gluten peptides to infiltrate lamina
propria [142, 143]. At the same time, transglutaminase type 2 stimulates
deamination of gluten-derived peptides producing epitopes that bind to
HLA-DQ2/DQ8 heterodimers on antigen-presenting cells, thus provoking a T-cell
response [144, 145]. Moreover, gliadin (a prolamine component of gluten)
accelerates the dissembling of intercellular junctional proteins via epidermal
growth factor receptor (EGFR) pathway activation [3, 146, 147]. As a consequence
of uncontrolled antigen trafficking from the lumen through the IB,
immunoregulatory deficits are further escalated. It was suggested that the onset
of inflammation with a secondary production of pro-inflammatory cytokines
(TNF-
Interestingly, CeD and IBD share genetic backgrounds associated with the innate immune response against pathogens and pro-inflammatory activation [133]. However, the Th17 response pathways and autophagy (natural cell degradation that removes unnecessary or dysfunctional components through a lysosome-dependent regulated mechanism) are only involved in IBD. Autophagy also plays a role in recognising and eliminating pathogens [133, 150, 151].
It is now well established that CeD and IBD need external, environmental stimuli to activate the immune system. Known environmental triggers include diet [152, 153], smoking, microorganisms, hygiene, early antibiotic exposure, and urban living, to mention a few [154]. Not surprisingly, the initial trigger is different for each disorder [133].
The bacterial species Bacteroides and Firmicutes make up 90% of eubiotic human gut microbiota [155], where along with other phyla, they orchestrate pro-and anti-inflammatory responses [156]. For example, in a healthy microbiome, Bacteroides engage in the recruitment of cytotoxic T cells to target immune cells (microbial antigen-loaded, antigen-presenting cells) that can trigger IBD, thus preventing IBD development [157]. Gut dysbiosis in CD is characterised by a decrease in Bacteroides and Firmicutes and an increase in Gammaproteobacteria and Actinobacteria, which leads to disease progression [71]. Interestingly, enterotoxigenic Bacteroides fragilis has been strongly associated with IBD and colorectal cancer [158, 159, 160]. In addition, 33% of IBD patients have an increased quantity of E.coli in the gut [161]. These strains of bacteria can cross the mucosal barrier, disturbing the epithelial lining [4, 71], thus allowing bacterial metabolites and pathogens to translocate into the systemic circulation [4, 25, 39, 156, 162]. Gram-negative bacterial metabolites such as LPS can also induce colitis and local inflammation in the intestines [163].
Gut dysbiosis in CeD is characterised by an increase of Gram-negative and a decrease of Gram-positive bacteria [2]. The increased presence of Bacteroides fragilis has been found in celiac patients and was associated with IB hyperpermeability and CeD pathogenesis [164]. Intriguingly, celiac patients presenting with EIM and those with typical GIT symptoms have different microbiome landscapes [165]. Wacklin et al. [165] demonstrated that CeD patients with only GIT symptoms had lower microbial diversity dominated by Proteobacteria compared to those with EIMs. The latter had a high abundance of Firmicutes [165].
CeD facilitates barrier breaches in the immune-privileged organs (brain, cochlea); however, the mechanism is unknown [16, 166, 167]. CeD can also increase the risk of developing sepsis due to the increased mucosal permeability and altered composition of the intestinal glycocalyx (the layer of gut epithelial cells considered the primary site for adhesion of commensal bacteria) [167].
The EIMs of IBD are widespread and include the brain and the inner ear [168, 169] (Fig. 2). The IBD-related EIMs occur in approximately 50% of patients [170]. The therapy of IBD and associated EIMs is primarily concerned with the dietary modifications and systemic anti-bacterial and immunosuppressive agents (antibiotics, sulfasalazine, corticosteroids, azathioprine, and dapsone) [171, 172]. IBD likely results in low-grade systemic inflammatory responses, which can spill over to the extraintestinal organs. Although the cause of EIMs is unknown, they have been linked to IB hyperpermeability [25]. Both IBD [173] and CeD [174] have been associated with gut dysbiosis and significantly increased levels of bacterial plasma components such as LPS [12, 13, 14]. Remarkably, after eliminating gluten from the diet in celiac patients, LPS levels decrease [174], and the IB regains its integrity. Therefore, it can be postulated that, as the pathogens infiltrate the circulation, they breach the natural barriers of immune-privileged organs, allowing pathogens in and causing localised inflammatory reactions.
Extraintestinal manifestations of gut dysbiosis. Gut dysbiosis contributes to intestinal tissue damage, leading to intestinal barrier hyperpermeability (leaky gut). The pathogens and their metabolites infiltrate the bloodstream, leading to a spread of inflammation, which can reach remote organs and thus cause extraintestinal manifestations.
Despite the general hypothesis that EIMs of IBD are related to immune reactions [175, 176, 177, 178, 179], their pathogenesis is not fully understood. There is increasing evidence that IBD results from a malfunction of the immune system and autoimmunity. Being diagnosed with IBD also increases the likelihood of developing other autoimmune diseases [180]. Due to shared epitopes, the “leaky intestine” of the damaged GIT mucosa may trigger immune responses at various extraintestinal sites [175, 178]. Resultant from IB hyperpermeability, commensal bacteria metabolites and pathogens in the bloodstream may trigger autoimmune reactions due to the similarity between the bacterial and host epitopes [169, 175, 178].
Not all IBD and CeD patients develop EIMs. Genetic factors play a significant role in presenting IBD EIMs [181], with a concordance rate of 70% of parent-child pairs and 84% of sibling pairs [182]. Studies investigating the relationships between EIMs and major histocompatibility complex loci have shown that UC and CD do not share HLA genotypes [181, 183, 184]. CD patients are more likely to express HLA-A2, HLA-DR1, and HLA-DQw5 genotypes, whereas UC patients tend to express HLA-DR103, -B27, and -B58 genotypes [181, 183]. It appears that the specific HLA genotypes are related to different EIMs [181, 184, 185]. For example, HLA-DR3 is associated with the increased risk of primary sclerosing cholangitis in UC, HLA-B27 and HLA-B58 are associated with EIMs related to skin and eyes, and HLA-B27 to ankylosing spondylitis in 90% of IBD patients [183, 184, 186].
EIMs of CeD and IBD include neuroinflammation and psychiatric disorders [15, 105, 187, 188, 189]. A substantial proportion of adult CeD patients develop neuroinflammatory and neurological conditions [15, 188, 189, 190]. Neurological deficits have been reported in 22.5% of adults and up to 24.5% of children in clinical studies [191, 192, 193]. In IBD, patients carry an increased risk of developing neurodegenerative diseases such as Parkinson’s disease (PD) or Alzheimer’s disease (AD) (PD: adjusted hazard ratio [HR], 1.56; 95% confidence interval [CI], 1.24–1.97; AD: adjusted HR, 1.14; 95% CI, 1.05–1.25) [194]. Younger IBD patients are more likely to develop PD than their healthy counterparts [194]. Furthermore, the incidence of dementia is significantly increased in IBD patients when compared to age-matched controls (5.5% vs 1.4%; HR, 2.54) and occurs earlier in life (76.24 years old on average, compared with 83.45 among controls) [195]. In addition, Elsehety and Bertorini found neurological or psychiatric EIMs in 84 of 253 patients with CD (frequency 33.2%) [196].
Neuropsychiatric manifestations of IBD and CeD can be stress-related due to the
difficulty living with these diseases [40, 197, 198]. Nevertheless, opinions
regarding the association between mental health, stress, and IBD are conflicting
[40]. The alteration of the gut-brain axis compromises relationships between gut
microbiota, gut-associated lymphoid tissues, neuroendocrine network, and
neuro-cognitive functions [16]. Gut dysbiosis could cause a breach of the
blood-brain barrier (BBB) via hormonal secretion, small molecules such as LPS
[199], vascular endothelial growth factors [200] and free radicals [201]. The
metabolic cofactors (e.g., homocysteine [202] and nicotinamide adenine
dinucleotide [203]) and inflammatory mechanisms [16, 204] have also been
postulated. For example, Matisz et al. [205] suggested that chronic
inflammation due to gut dysbiosis remodels anterior cingulate cortex physiology,
resulting in the inaccurate judgment of danger. This remodelling was induced by
chronic stimulation of the threat-coping system by endocrine signalling and
anxiety [205]. LPS overproduction and release from the leaky gut likely plays a
crucial role in developing neurodegenerative diseases [16, 206, 207]. In PD,
LPS-CD14 complexes interact with toll-like receptor TLR4, initiating signalling
events involving mitogen-activated protein kinases (MAPK) and transcription
factors such as NF-
It appears that CeD, IBD, and mental disorders (autism, schizophrenia) share mechanisms involving microbial-derived metabolites that can cause neuroinflammation and damage in different brain regions [16, 215, 216, 217]. The process starts with gut dysbiosis and leads to neurodegeneration and cognitive deficits via inflammatory pathways [16, 214, 215, 216, 217, 218]. It has been established that gut dysbiosis induced by ageing, diet, obesity, alcohol abuse, and antibiotics could underpin the dysfunctional gut-brain axis [189], leading to inflammatory processes in the brain. Early antibiotic use is also a risk factor in developing autism and other neurological conditions [219].
Accordingly, the expression of the peroxisome proliferator-activated
receptor-gamma (PPAR
Micro-RNAs are small RNA molecules (~20–23 nucleotide long), which regulate gene expression post-transcription by binding to homologous sequences on the 3’ untranslated regions (UTRs; homologous base pairings between miRNA seed nucleotides 2 to 7 and the 3’ UTR) [227]. Micro-RNAs control most cellular processes such as cell proliferation, differentiation, apoptosis, cell signalling, immune and inflammatory responses [227]. Biopsies of intestinal tissue of celiac patients showed downregulation of miRNAs: miR-192-5p, miR-31-5p, miR-338-3p, and miR-197 and upregulation of chemokine C-X-C motif ligand 2 (CXCL2) and nucleotide oligomerisation domain-2 (NOD2) at mRNA and protein expression levels [229]. These miRNAs play a significant role in innate immune responses. For example, miR-192-5p, a critical player in intestinal homeostasis, is downregulated in UC [230] and CeD [229, 231]. Moreover, miR-449a, responsible for negative regulation of Notch receptor 1 (Notch1) and krüppel like factor 4 (KLF4) that regulate goblet cell proliferation and differentiation, are upregulated in pediatric CeD patients, thus explaining the reduced amount of goblet cells in CeD [232]. Another study demonstrated elevated miR-204 in rhesus macaques with CeD [226]. This miRNA directly targets the intestinal TJ protein claudin-1, reducing its expression [227]. The loss of TJ proteins can compromise the BBB, leading to the translocation of intestinal LPS to the brain, activation of microglia, and neuroinflammation [227].
EIMs of IBD and CeD include SNHL (Tables 1 and 2) [9, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246] . Between 40–60% of pediatric CeD patients demonstrate at least unilateral SNHL [234, 236, 237]. Therefore, it was suggested that children presenting with idiopathic hearing loss should also be checked for CeD [236]. Correspondingly, if the child is diagnosed with CeD, the audiometric examination is warranted to identify early hearing deficits [237]. However, the pathophysiology of CeD-induced SNHL is still enigmatic. Several hypotheses have been put forward, including vasculitis, malnutrition, labyrinth infiltration by activated lymphocytes, anti-neuronal antibodies, and deposition of immune complexes [245]. The endolymphatic sac contains and recirculates IL-2-producing immunocompetent cells, which regulate immune responses [245]. IL-2 activation in endothelial cells of the spiral modiolar vein stimulates intercellular adhesion molecule-1 (ICAM-1) to attract more leukocytes to the target tissue and initiate immune and inflammatory reactions [247, 248, 245]. The autoimmune aetiology of IBD-induced SNHL was also proposed [10], even though the presentation of SNHL in IBD patients is more consistent with chronic inflammation, similar to the neuroinflammatory processes in the brain.
Year | Authors | Title | Methodology | Results | Conclusion |
2005 | Akbayir et al. [233] | Sensorineural hearing loss in patients with inflammatory bowel disease: a subclinical extraintestinal manifestation | Clinical study involving 39 patients with IBD (21 Crohn’s disease, 18 ulcerative colitis) and 25 healthy age- and sex-matched controls. To assess auditory function, otoscopy, tympanometry, and pure tone audiometry were carried out. | o ll patients and control subjects had normal otoscopy findings, and tympanometry was unremarkable, excluding middle ear disease and conductive hearing loss. | “(…) it was demonstrated that a subclinical SNHL may be associated with UC and somewhat with CD, affecting mainly the high frequencies. In light of this finding, it may be advisable to investigate labyrinth functions as well as other extraintestinal manifestations in patients with IBD.” |
o The average hearing thresholds were raised significantly in the IBD group at higher frequencies (2, 4, and 8 kHz). | |||||
o There is a significant threshold increase for the UC group at frequencies 2, 4, and 8 kHz and for the CD group only at 4 kHz. | |||||
o A trend of SNHL worsening with the patient age and extent of ulcerative colitis was observed. | |||||
o No significant correlation between SNHL and sex, involvement site in GI tract, medication history for IBD, and coexistence of other EIMs. | |||||
2009 | Karmody et al. [240] | Sensorineural hearing loss in patients with inflammatory bowel disease | A clinical study was conducted over 11 years. Medical and audiometric documentation of 38 patients with a diagnosis of IBD (ulcerative colitis and Crohn’s disease) was reviewed. | o Of 38 patients with a history of IBD, 58% (n = 22) recorded SNHL. | “(…) this study demonstrates a correlation between SNHL and IBD, but a larger controlled investigation is needed. If IBD is an autoimmune disorder, the inner ear could be affected by the underlying systemic immune dysfunction. Unravelling the pathophysiology of IBD should explain the mechanism of its association with dysfunction of the inner ear.” |
o 19 patients with SNHL had no other identifiable aetiology for their inner ear dysfunction. | |||||
o 14 patients with SNHL had been diagnosed with UC, and 5 had CD. | |||||
o 16 had bilateral SNHL, and 3 patients had unilateral SNHL. | |||||
o 70% developed hearing loss before the age of 50 years. | |||||
o Only one SNHL patient had a lasting response to medical treatment. | |||||
2014 | Wengrover et al. [241] | Hearing loss in patients with inflammatory bowel disease | A prospective blinded comparative study was conducted over 3 years. | o 21% (n = 16) of the IBD patients complained of hearing loss since the first IBD diagnosis; 13% had hearing deficits. | “Sensorineural hearing loss may be another EIM of IBD. It is found in 30% of IBD patients and in up to 43% of patients with other EIMs. Early hearing evaluation should be recommended to IBD patients who have other EIMs.” |
o Audiometric examination showed hearing loss (mild to severe) in 23 (30%) of the IBD patients, matched with 3 (10%) of the controls. | |||||
A total of 105 participants (76 patients and 29 controls), where 59 (77%) had CD, and 17 (23%) had UC. The mean age was 36 years; 51% were males, and 40% were hospitalised due to IBD exacerbation. | o SNHL constituted 93% of hearing deficits. Out of 46 patients whose EIM status was clearly documented, 20 (43%) had SNHL. | ||||
o IBD phenotype, hospitalisation, and disease type were not different between the groups. | |||||
2016 | Wengrower et al. [9] | Hearing loss in patients with inflammatory bowel disease | A prospective blinded comparative study was conducted over three years. 76 IBD patients and 29 controls underwent a complete otorhinolaryngological examination and audiometry test. | o Hearing loss (mild to severe) was found in 29 (38%) of the IBD patients, and 4 (14%) of the control group. | “Hearing loss may be another EIM of IBD. It is found in 38% of IBD patients and up to 52% of patients with other EIMs; hearing loss increases over the age of 40. Early hearing evaluation should be recommended to these high-risk IBD patients.” |
o Moderate to severe hearing loss was found in 7/33 (21%) in the EIM-positive group compared to 4/43 (9%) in the EIM-negative group. | |||||
o Out of 11 patients over 40 with other EIMs, all (100%) had hearing loss compared to 8/12 (66%) of patients over 40 without other EIMs. | |||||
2020 | Polat et al. [242] | Assessment of hearing function in children with inflammatory bowel disease | The clinical study involved 32 pediatric patients with IBD and 31 age-matched controls. Examinations involved detailed ENT examination, pure tone audiometry (PTA), high-frequency audiometry (HFA), signal-to-noise ratio (SNR) and distortion product (DP) otoacoustic emissions testing. | o No differences in age and gender and PTA thresholds at low frequencies between controls and children with IBD. | “(…) SNHL in pediatric patients with IBD was seen at the high frequencies. It could represent a potential early indicator of SNHL in this population. We recommend hearing function tests twice a year for early diagnosis. HFA and DPOAE can be used safely in this population for monitoring the hearing loss.” |
o The mean PTA responses at 1,000; 8,000; 10,000; 12,500; 16,000; SNR1400; SNR2000; SNR2800; and SNR4000Hz of the IBD group were significantly higher than those of the controls (p | |||||
2021 | Yozgat et al. [243] | Ulcerative colitis may be a risk factor for sensorineural hearing loss | The clinical study involved 53 patients with IBD and 20 matched controls within period of 4 months. Examinations involved tympanometry, otoscopy and audiometry. | o No significant difference in terms of gender and age between the IBD and control groups. | “SNHL has been detected in a significant number of UC patients. Also, the hearing function deteriorated significantly as the age of the patients and the duration of the disease increases. It should be recommended to evaluate UC patients over 40 years of age and with the long-term disease for SNHL.” |
o No significant difference in air and bone conduction in both ears in patients with CD. | |||||
o A significant difference in both air and bone conduction in ulcerative colitis (p = 0.0001 in the left ear, p = 0.004 in the right ear). | |||||
o SNHL was detected in 45.2% (n = 14) of UC patients and 13.6% (n = 3) of CD patients using audiometry. | |||||
o Three UC patients had moderate, one had moderate to severe, and one had profound hearing loss. |
Year | Authors | Title | Methodology | Results | Conclusion |
2011 | Hizli et al. [234] | Sensorineural hearing loss in pediatric celiac patients | A sample of 32 biopsies and serologically proven newly diagnosed pediatric celiac patients and matched healthy subjects (control group) were involved in this study. Pure-tone audiometry at frequencies 250–8000 Hz was performed in all subjects. Slight/mild SNHL was defined as a loss of sound detection within the 16–40 dB range. The mean age of the patient and control group was 11.9 and 11.3, respectively (p |
· SNHL was found in 40.6% (n = 13) celiac pediatric patients (6 unilateral and 7 bilateral), and 3.1% (n = 1) control group. | “(…) a higher prevalence of SNHL in pediatric celiac patients than in controls, suggesting an association between CeD and SNHL. The findings of this study suggest that hearing impairment should be investigated in newly diagnosed pediatric CeD patients. Further longitudinal investigations on a larger sample size will be necessary to confirm the present data and to search the immunological processes which could be the basis of the association between CD and SNHL.” |
2011 | Karabulut et al. [244] | Audiological findings in celiac disease | 41 pediatric celiac patients and 31 controls were included in the study. Both groups were evaluated with audiometry, tympanometry, transiently evoked otoacoustic emission (TEOAE), distortion product otoacoustic emission (DPOAE), and contralateral suppression of the TEOAE. | · The average PTA thresholds at 250 Hz of the CeD patients were significantly higher (p |
“(…)CeD seems to have an important impact on the auditory system and results in an elevation of the PTA thresholds at 250 Hz and a decrease in the amplitudes of DPOAE and linear TEOAE at 1 kHz in children.” |
· The signal-to-noise ratio (SNR) amplitudes in DPOAE testing and SNR with and without contralateral acoustic stimulus in TEOAE testing were significantly lower at 1 kHz in the CeD group than in the control group. | |||||
· There was no significant difference between the CeD and the control group regarding contralateral suppression amplitudes. | |||||
2012 | Solmaz et al. [236] | Celiac disease and sensorineural hearing loss in children | 25 pediatric patients with biopsy-proven celiac disease were diagnosed in the pediatric gastroenterology department, and 25 healthy control subjects were included in the study. All subjects underwent tympanometry and pure tone audiometry at frequencies 250–8000 Hz. | · Tympanometry showed normal peak compliance, gradient, peak pressure, ear canal volume, and acoustic reflexes in the patients and controls. | Sensorineural hearing loss (SNHL) and celiac disease (CeD) may be observed coincidentally. Children with clinical signs of hearing deficiency of unknown aetiology should be assessed for CeD. |
· There was no air-bone gap in any of the participants. | |||||
· There was a statistically significant difference (p | |||||
2012 | Leggio et al. [245] | Coeliac disease and hearing loss: Preliminary data on a new possible association | Twenty-four adult celiac patients and 24 healthy subjects matched for gender, age, smoking and drinking habits were enrolled in the study. Among the celiac patients, 6 were newly diagnosed, and 18 patients were on a gluten-free diet for at least one year. | · A hearing loss was found in 47.1% (n = 10) of celiac patients and 9.1% (n = 2) controls. | “Despite the low number of subjects evaluated, the present study showed a higher prevalence of hearing loss in celiac patients than in healthy controls, suggesting an association between CeD and hearing loss. Immunological processes such as ear-specific and non-specific autoantibodies and vasculitis could be the basis of this association. Further longitudinal investigations on a larger sample size will be necessary to confirm the present data. ” |
· All celiac patients with hearing impairment developed SNHL. | |||||
· The prevalence of SNHL was not significantly different between untreated (33.3%) and treated (44.4%) coeliac patients. | |||||
2015 | Urganci et al. [237] | Sensorineural hearing loss in pediatric patients with celiac disease | Otoscopy, tympanometry and pure tone audiometry were performed in 44 pediatric patients with celiac disease and 20 matched controls. | · SNHL was detected in only 6.8% (n = 3) patients within 1−3 years after diagnosing CeD. | “(…) subclinical sensorineural hearing loss was demonstrated in adult patients with CeD; therefore, we recommend to perform audiometric examinations in pediatric patients for recognising hearing loss early during the course of the disease.” |
· None of the patients or controls had symptoms such as hearing loss, tinnitus or balance disturbance. | |||||
· All group members had normal otoscopy and tympanometry, excluding middle ear disease and conductive hearing loss. Pure tone audiometry showed no abnormality. | |||||
2015 | Sahin et al. [235] | Evaluation of hearing loss in pediatric celiac patients | The study included 110 pediatric patients with biopsy-confirmed celiac disease and 41 matched controls. The hearing was evaluated using tympanometry and pure tone audiometry (250–8000 Hz frequency). | · Audiometric bone conduction thresholds were significantly (p |
“These results indicate that subclinical hearing loss may be present in children with CeD, which could precede more serious hearing impairments at older ages and later stages of the disease. Hearing screenings should be recommended for children with CeD in order to prevent the potentially unfavourable effects of hearing loss on the emotional, behavioural, cognitive, and sensorimotor development of these patients.” |
· There were no significant differences in pure-tone averages for air conduction (p | |||||
· When the results for celiac patients were analysed according to the duration of disease ( |
|||||
2019 | Yazici et al. [246] | Does celiac disease cause autoimmune sensorineural hearing loss? | The prospective study included 103 adult celiac patients and 79 healthy controls between 2012 and 2018. Celiac patients were divided into two groups: remission or active, according to their gluten-free diet duration and serum levels of anti-tissue transglutaminase. They underwent pure-tone audiometry after detailed ear examination. | · Only 3.88% (n = 4) of celiac patients showed SNHL. | “In this study with a higher number of CeD patients when compared with the previous studies, it has been shown that CeD does not appear to cause autoimmune SNHL. In addition, patients in the remission of CeD did not show different PTA thresholds than the active cases.” |
· There was no statistically significant difference between the hearing levels of the celiac patients and the control group in air and bone conductions measurements. | |||||
· The PTA thresholds comparing the remission and active celiac patients did not differ in air and bone conduction frequencies. | |||||
2020 | Yaprak et al. [239] | Hearing evaluation with ABR in pediatric patients with celiac disease | 38 pediatric celiac patients were included in the study. The patients had confirmed diagnosis of Celiac disease through duodenal biopsies and transglutaminase antibody. The control group consisted of 18 children aged 3 to 17 years old who were all admitted to the pediatric gastroenterology department due to complaints of constipation and transglutaminase Ab. All children underwent Auditory-Brain-Stem-Evoked Responses (ABR). | · The results of the ABR examination did not show any difference between the patient group and control group as regards the latency of the waves I, III, V. | “The exact pathogenesis of neurological damage observed in CeD is still unknown. Humoral immune mechanisms are the most frequently attributed cause. Although no significant difference was found in ABR responses between the study group and healthy control group, there is a need for further research on this subject.” |
· No difference was observed between the two groups in the interpeak latencies of the ABR waves I–III, I–V and III–V. None of the patients was observed to have clinical hearing loss. |
The blood-brain barrier (BBB) is a critical anatomical and physiological structure protecting neural tissue. The BBB is formed by the blood vessels of the central nervous system (CNS). In the CNS, the blood vessels are not permissive, restricting the infiltration of pathogens into neural tissue [249, 250, 251]. The integrity of the BBB can be affected by multiple mechanisms, which, in turn, allow pathogens and inflammatory cells to infiltrate neural tissues [249, 252]. This can result in neuroinflammatory disorders, such as multiple sclerosis [253, 254], acute disseminated encephalomyelitis [255], or transverse myelitis [256]. Neuroinflammation can also be triggered by an injury, exposure to a neurotoxin, neurodegenerative disease, or ageing [257, 258, 259].
The leaky gut and resulting systemic inflammation can negatively impact the
integrity of the BBB. As a result, the BBB’s ability to selectively restrict the
passage of pathogens and neurotoxic agents to the brain is diminished.
Inflammation and hypoxia can be classified as primary culprits [16], often by
weakening the TJ of the BBB [16]. Han et al. [260] demonstrated that
dextran sodium sulfate (DSS)-induced colitis in mice could provoke systemic
inflammation leading to cortical brain inflammation via up-regulation of
inflammatory cytokines in the serum. Other studies demonstrated that colitis
induced by trinitrobenzene sulphonic acid (TNBS), which affects the IB
permeability, can also disturb the integrity of the BBB [252, 261]. In monkeys,
the alteration of the gut microbiome by antibiotics can also increase BBB
permeability [262] due to gut dysbiosis [260, 263]. The permeability of the BBB
can be increased by pro-inflammatory cytokines such as IL-1
Interestingly, germ-free (GF) mice show increased resistance to neuroinflammatory diseases [265]. However, Braniste et al. [204] reported an increase in the permeability of the BBB in GF mice. A possible explanation is that the normal gut microbiota also regulates the TJ structure and function within the BBB; thus, a lack of commensal bacteria may lead to an aberrant formation of TJs, such as occludin and claudin-5 [204].
In conclusion, there is strong evidence that “leaky gut” disorders can induce a variety of EIMs resulting from breached barriers of immune-privileged organs, such as the BBB.
Despite the anatomical differences between the IB, BBB, and BLB, these barriers also have several commonalities [263, 266]. Diseases associated with pathological alterations of gut microbiota (e.g., diabetes, obesity, IBD, CeD) also share an ability to affect the permeability of all three barriers. Therefore, we postulate that gut dysbiosis and leaky gut have a similar influence on both the BBB and BLB via a feedback loop driven by microbial solutes and the innate immune system.
One of the factors that consistently affects the permeability of the BLB is inflammation. Another potential mechanism that might affect the permeability of the BLB and the BBB in IBD and CeD is OS [11, 15, 267, 268, 269].
OS results from the overproduction of reactive oxygen species (ROS) [267, 270, 271]. Along with inflammation, it is considered one of the primary mechanisms in CeD. It was suggested that OS might predispose CeD patients to other autoimmune disorders [267]. Previous studies have shown that brain-derived microvascular endothelial cells of the BBB, when exposed to OS, express elevated matrix metallopeptidase 9 (MMP-9) activity that affects TJ protein occludin [264, 272]. Similarly, elevated OS markers such as inducible nitric oxide synthase (iNOS) show that vascular endothelial cells of the BLB are also prone to oxidative damage, triggering inflammatory pathways within the cochlea [273].
In mice, specific cytokines (IL-1, IL-6, and MIP-1
Leaky gut can cause systemic inflammation resulting in a breach of the blood-brain and blood-labyrinth barrier and local low-grade chronic inflammation in the brain and the cochlea. Abbreviations: BM, basement membrane; EC, endothelial cell; PVM/M, perivascular-resident macrophage-like melanocyte; TJ, tight junction; SV, stria vascularis; SL, spiral ligament; SG, spiral ganglion.
CpG motifs can also activate innate and adaptive immune responses via TLR9
receptor-mediated MAPK and NF-
Cytokines and chemokines increase the permeability of the BBB by stripping off its protective glycocalyx [269]. The glycocalyx is an equivalent of the superficial unstirred mucus layer in the IB and forms both the BBB and the BLB [289]. As a result, the endothelial cells are exposed to inflammatory mediators, allowing for their erosion [269].
“Leaky gut” disorders, such as IBD and CeD, lead to an increase in IB permeability. The compromised IB allows pathogens and microbial metabolites to infiltrate the circulatory system and spread to distant organs. The immune-privileged organs (brain, cochlea) are protected by barriers with a similar structure; thus, these barriers are more likely to be compromised by the same type of stimuli. EIMs of CeD and IBD include dysfunctions of the blood-brain and the blood-labyrinth barrier (Fig. 3). Based on the current literature, we postulate that the breach of these barriers causes neuroinflammation and inflammation-induced SNHL. We have coined the term gut-inner ear axis to describe the crosstalk between the gut and inner ear, analogous to the gut-brain axis.
DK designed the study and performed the literature search, and SMV provided advice. DK analysed the literature and drafted the review. Both authors contributed to editorial changes in the manuscript. Both authors read and approved the final manuscript.
Not applicable.
We thank Mr Corey Beran for proofreading this manuscript.
This study was funded by Eisdell Moore Centre (Auckland, New Zealand), grant number 3721994.
The authors declare no conflicts of interest.