Periodontal disease–related nonalcoholic fatty liver disease and nonalcoholic steatohepatitis: An emerging concept of oral‐liver axis

Abstract Periodontal disease, a chronic inflammatory disease of the periodontal tissues, is not only a major cause of tooth loss, but it is also known to exacerbate/be associated with various metabolic disorders, such as obesity, diabetes, dyslipidemia, and cardiovascular disease. Recently, growing evidence has suggested that periodontal disease has adverse effects on the pathophysiology of liver disease. In particular, nonalcoholic fatty liver disease, a hepatic manifestation of metabolic syndrome, has been associated with periodontal disease. Nonalcoholic fatty liver disease is characterized by hepatic fat deposition in the absence of a habitual drinking history, viral infections, or autoimmune diseases. A subset of nonalcoholic fatty liver diseases can develop into more severe and progressive forms, namely nonalcoholic steatohepatitis. The latter can lead to cirrhosis and hepatocellular carcinoma, which are end‐stage liver diseases. Extensive research has provided plausible mechanisms to explain how periodontal disease can negatively affect nonalcoholic fatty liver disease and nonalcoholic steatohepatitis, namely via hematogenous or enteral routes. During periodontitis, the liver is under constant exposure to various pathogenic factors that diffuse systemically from the oral cavity, such as bacteria and their by‐products, inflammatory cytokines, and reactive oxygen species, and these can be involved in disease promotion of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. Also, gut microbiome dysbiosis induced by enteral translocation of periodontopathic bacteria may impair gut wall barrier function and promote the transfer of hepatotoxins and enterobacteria to the liver through the enterohepatic circulation. Moreover, in a population with metabolic syndrome, the interaction between periodontitis and systemic conditions related to insulin resistance further strengthens the association with nonalcoholic fatty liver disease. However, most of the pathologic links between periodontitis and nonalcoholic fatty liver disease in humans are provided by epidemiologic observational studies, with the causal relationship not yet being established. Several systematic and meta‐analysis studies also show conflicting results. In addition, the effect of periodontal treatment on nonalcoholic fatty liver disease has hardly been studied. Despite these limitations, the global burden of periodontal disease combined with the recent nonalcoholic fatty liver disease epidemic has important clinical and public health implications. Emerging evidence suggests an association between periodontal disease and liver diseases, and thus we propose the term periodontal disease–related nonalcoholic fatty liver disease or periodontal disease–related nonalcoholic steatohepatitis. Continued efforts in this area will pave the way for new diagnostic and therapeutic approaches based on a periodontologic viewpoint to address this life‐threatening liver disease.


| BACKG ROU N D
Periodontal disease is a common chronic inflammatory and infectious disease that is caused by an oral biofilm-mediated microbial dysbiosis that is predominantly comprised of anaerobic gram-negative bacteria, namely periodontopathic bacteria. 1,2 These biofilms are a continually renewing storehouse of lipopolysaccharide and other microbial molecules that are derived from the resident gram-negative bacteria. Biofilm components have ready access to the periodontal tissues and host circulation. Microbial challenges also initiate and perpetuate host immune responses in the periodontal tissues, resulting in production of high levels of inflammatory mediators and tissue-destructive enzymes. These responses, in turn, lead to periodontal tissue destruction and tooth loss. 3 The products from inflamed periodontal tissues also enter the circulation and enhance susceptibility to systemic diseases via several pathways. 1 In the field of research related to periodontal medicine, few papers to date have addressed the relationship between periodontal disease and the organs of the digestive system. Meanwhile, the relationship between periodontal disease and liver disease has received growing attention in recent years. The liver is the largest organ in the digestive system, and it plays an important role in maintaining the health of living organisms. 4 During the process of digestion, nutrients in food are absorbed through the numerous fine capillaries of the intestinal wall and they are carried into the veins. 5 These veins merge into larger veins and ultimately enter the liver through the portal vein. The liver removes bacteria and other foreign matter from the blood that enters through the portal vein, and it further breaks down many nutrients that have been absorbed by the intestine. 4 Blood rich in nutrients then recirculates for use throughout the body. Liver diseases occur due to various causes, including infectious diseases, pharmaceutical use, toxins, ischemia, and autoimmune diseases.
Many liver diseases cause liver cell damage, necrosis, and subsequent development of hepatic dysfunction, which leads to symptoms due to both the liver disease itself (eg, jaundice caused by acute hepatitis) and complications of the liver disease (eg, acute gastrointestinal bleeding as a result of liver cirrhosis and portal hypertension). Liver diseases such as hepatitis (which starts with a fatty liver caused by excessive alcohol consumption) and viral hepatitis are well known. However, in recent years, hepatitis and liver cirrhosis caused by fatty liver in the absence of alcohol consumption or in the presence of low alcohol consumption and without a viral infection have also been identified and are attracting attention. 6,7 insulin resistance and metabolic syndrome because many cases of nonalcoholic fatty liver disease arise from conditions such as obesity, diabetes, dyslipidemia, and hypertension. [7][8][9][10] Nonalcoholic fatty liver disease has a high worldwide prevalence of approximately 25%, and this is expected to increase in the future due to the increasing number of obese people who have metabolic syndrome. 11,12 Furthermore, nonalcoholic fatty liver disease is classified into nonalcoholic fatty liver, which has limited pathologic progression, and nonalcoholic steatohepatitis which has a more severe progressive nature. 13,14 Nonalcoholic fatty liver is a disease with a favorable prognosis, whereas nonalcoholic steatohepatitis can have fatal consequences with the gradual progression of inflammation and fibrosis transitioning into end-stage liver disease, such as cirrhosis and hepatocellular carcinoma. Therefore, appropriate strategic interventions for the prevention and early treatment of nonalcoholic steatohepatitis are required. 6,7 However, since the terms nonalcoholic fatty liver disease and nonalcoholic steatohepatitis do not reflect the cause of the disease and encompass numerous clinical conditions, there has been a movement in recent years to further subdivide the disease and develop new nomenclature to change the name of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis to "metabolic fatty liver disease" and "metabolic steatohepatitis." 6,7 Recently, there has also been a lively debate over the possible development of a periodontal disease-related nonalcoholic fatty liver disease and nonalcoholic steatohepatitis, which is the main theme of this chapter. Research related to periodontal disease and nonalcoholic fatty liver disease has gradually changed over time.
Between the 1990s and the early 2000s, a bidirectional association between poor oral hygiene with the presence of periodontal disease and chronic hepatitis and cirrhosis was suggested. [15][16][17][18] Later in the 2000s, the possible involvement of systemic inflammation and oxidative stress derived from periodontitis in the development of nonalcoholic fatty liver disease emerged from in vitro-based basic research. 19,20 Then, in the early 2010s, the possible involvement of Porphyromonas gingivalis, a common periodontopathic bacteria, in the development of nonalcoholic fatty liver disease was reported and continues to be discussed to this day. [21][22][23] Related to this, the concept of a gut-liver axis and gut dysbiosis was further proposed as another potential route linking the oral cavity and the liver. 24,25 Since the late 2010s, systematic reviews and meta-analyses [26][27][28] have continued to report on these associations based on growing evidence from epidemiologic studies 21, and on additional evaluation from in vivo research. [50][51][52][53][54][55][56][57][58][59] Moreover, as a next step, clinical studies with therapeutic intervention are expected to verify the effect of periodontal treatment on nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. 60,61 The relationship between periodontal disease and nonalcoholic fatty liver disease has been discussed from in vitro, in vivo, and epidemiologic perspectives, although no review has ever discussed these in a systematic manner, which is the aim of the current review.
In this review, we provide updates based on current evidence on the pathogenesis, clinical data, and treatment of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis involved with periodontal disease. After providing an explanation of the epidemiology and etiology of nonalcoholic fatty liver disease, the present status of the association between nonalcoholic fatty liver disease and periodontal disease will be presented. We will also explain the interrelationship of metabolic disorders and periodontal disease with nonalcoholic fatty liver disease and will organize the research evidence into the two pathways that link periodontal disease with liver disease, through the hematogenous and enteral routes. Furthermore, specific examples of periodontal disease-derived risk factors that play an important role in nonalcoholic fatty liver disease and nonalcoholic steatohepatitis will be discussed. Lastly, the possibility of periodontal treatment and the future outlook of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis research will be outlined.
As previously mentioned, separately classifying insulin resistance-associated nonalcoholic fatty liver disease and nonalcoholic steatohepatitis 6,7 from that which is associated with periodontal disease may be a development that emerges in the near future. This would support therapeutic intervention based on a periodontal approach, which may enable early treatment of this life-threatening liver disease.

| Anatomic features and physiologic role of the liver
The liver is a prominent organ in terms of its metabolism, synthesis, and detoxification functions. It also plays an important role in regulating blood glucose and lipids, and it has the potential to regenerate even after tissue damage. 4 The central function of the liver in homeostasis and the inflammatory response is made possible by its unique anatomic location; and it is the largest parenchymal organ, receiving a dual blood supply from systemic circulation and the gastrointestinal tract. 5 The liver receives 80% of its blood supply via the intestinal portal vein, which is rich in bacterial products, environmental toxins, and food antigens. The remaining 20% is derived from the hepatic artery, which is a feeding vessel branching from the abdominal aorta. The blood from the two circulatory systems joins at the hepatic hilum and then spreads throughout the liver via a capillary network called sinusoids. In other words, the liver is the hemodynamic confluence of the human body, and the large amount of blood that continuously flows into the liver through the sinusoids allows for a diverse composition of intrahepatic cell populations comprised of the metabolically active hepatocytes, nonparenchymal hepatocytes, and various immune cells. 62 In particular, liver function depends on its strong innate immune system to provide effective and rapid protection against potentially toxic substances without causing a harmful immune response. 4,5 This role includes intrahepatic enrichment of innate immune cells (Kupffer cells, hepatic stellate cells, natural killer, natural killer T, and T cells, etc), immunologic elimination of microorganisms, and removal of waste molecules. 63 Such complex communication between intrahepatic immune cells and hepatocytes is primarily mediated by cytokines, which activate effector functions of immune cells and hepatocytic intracellular signaling pathways controlling cell homeostasis. Kupffer cells and liver-infiltrating monocyte-derived macrophages are major sources of cytokines, such as tumor necrosis factor alpha and interleukin (IL)-6. Moreover, the biosynthesis of numerous soluble pathogen-recognition receptors and complement components plays an important role in controlling systemic innate immunity. 5 However, the liver is susceptible to metabolic and endocrine disorders due to the action of drugs, microorganisms, and environmental factors, and this imbalance can lead to pathologic consequences. 62 Given its regenerative capacity, the liver can overcome severe damage in many circumstances, but chronic damage progressively promotes a homeostatic imbalance, resulting in various chronic liver diseases, such as steatosis, hepatitis, fibrosis, cirrhosis, and hepatocellular carcinoma.

| Disease definition, prevalence, and epidemiology of nonalcoholic fatty liver disease/ nonalcoholic steatohepatitis
Nonalcoholic fatty liver disease, which affects both children and adults, is currently the most prevalent chronic liver disease worldwide. 64 Nonalcoholic fatty liver disease is defined as cases showing the presence of hepatic steatosis (greater than 5% of hepatocytes are fatty) but lacking common causes of secondary hepatic fat accumulation, such as excessive alcohol consumption, chronic viral hepatitis, autoimmune hepatitis, long-term use of steatosis-inducing medications, or congenital hepatic disorders. 6,9,65,66 The majority of nonalcoholic fatty liver diseases are nonalcoholic fatty liver (simple steatosis) with good prognosis (Figure 1), but a subgroup of about 20%-30% of these patients can develop into more severe and progressives forms of liver disease, namely nonalcoholic steatohepatitis. 9 Nonalcoholic steatohepatitis is characterized by histologic findings, including, in addition to lipid deposition, inflammatory cell infiltration, ballooning degeneration of hepatocytes, and fibrosis, and it is extremely difficult to distinguish between simple fatty liver and nonalcoholic steatohepatitis using noninvasive examination, such as blood biomarkers and ultrasonography. 67 Therefore, the gold standard for diagnosing nonalcoholic steatohepatitis remains a liver biopsy and exclusion of secondary causes. 68 Nonalcoholic steatohepatitis, also known as the liver phenotype of metabolic syndrome, is strongly associated with severe metabolic complications, such as obesity and diabetes mellitus. 8 Moreover, a portion of nonalcoholic steatohepatitis patients have been reported to progress to cirrhosis and hepatocellular carcinoma, which are end-stage liver diseases. 13,14 The prevalence of nonalcoholic fatty liver disease has been estimated to range between 20% and 50%, depending on the study population and diagnostic methods used, and it continues to increase worldwide as the number of obese individuals grows. [69][70][71] F I G U R E 1 Histologic features and prevalence of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH). A, Healthy liver normally contains some fat, but if more than 5% of hepatocytes are fatty, then it is diagnosed as fatty liver or steatosis. The spectrum of nonalcoholic fatty liver disease ranges from nonalcoholic fatty liver (NAFL: simple steatosis) to nonalcoholic steatohepatitis, which can ultimately progress to end-stage liver disease. In addition to the fatty deposition in the liver, nonalcoholic steatohepatitis is characterized by inflammation, hepatocellular damage, and cell death with or without fibrosis. Furthermore, nonalcoholic steatohepatitis can lead to scarring fibrosis and eventually progress to cirrhosis, hepatic insufficiency, and hepatocellular carcinoma. B, Global prevalence of nonalcoholic fatty liver disease was estimated to be 25% on average in the wide range from 17% to 50% according to the data from Estes et al 11 and Younossi et al. 12 Approximately 20% of nonalcoholic fatty liver disease cases would be classified as nonalcoholic steatohepatitis, which represents 3%-5% of the overall adult population. The worldwide prevalence of nonalcoholic fatty liver disease spectrum and subsequent cirrhosis have been projected to increase greatly by 2030 A meta-analysis study by Younossi et al 12 revealed that the global prevalence of nonalcoholic fatty liver disease is 25.24%, and is highest in the Middle East and South America, followed by Asia, North America, Europe, and Africa. It has been reported that the annual incidence of nonalcoholic fatty liver disease ranged between 20 and 50 cases per 1000 people in different countries. 6 Moreover, the overall mortality rate of patients with nonalcoholic fatty liver disease has increased significantly in recent years due to cardiovascular events and liver-related disorders, wherein the rate of nonalcoholic steatohepatitis patients is higher than that of patients with simple steatosis. [72][73][74] These surprising facts strongly indicate that nonalcoholic fatty liver disease and nonalcoholic steatohepatitis are at the center of the new pandemic of chronic liver disease, thus mediating a significant clinical and economic burden. 75,76

| Etiology and pathophysiology of nonalcoholic fatty liver disease/nonalcoholic steatohepatitis
The pathogenesis of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis involves multiple factors and processes, such as altered energy metabolism, an altered host immune system, enterobacteria, and genetic predisposition. Until now, the mechanism of its onset and progression has been explained from the perspective of a "two-hit theory" proposed by Day and James. 77,78 According to this theory, the first hit involves a sedentary lifestyle, high-fat diet, obesity, and insulin resistance, which enhance hepatic lipid accumulation and induce a fatty liver, thereby making the liver susceptible to further negative stimuli. Subsequently, it has been presumed that various hepatocyte-damaging factors, such as proinflammatory cytokines, gut microbiota-derived components, oxidative stress, and lipid peroxide, act as the second hit, leading to necrotic inflammation and fibrosis in the fatty liver. However, a twohit theory alone is not sufficient to explain all of the molecular and metabolic alterations occurring in nonalcoholic fatty liver disease, and in some cases it is necessary to assume that inflammation precedes the hepatic steatosis. 79,80 Therefore, the current widely accepted theory is that of a "multiple parallel hits hypothesis." This theory explains that there is an interaction between genetic and environmental factors, as well as changes in crosstalk between different organs, including adipose tissue, the intestine, the pancreas, and the liver. Together, this suggests that a more widespread and simultaneous metabolic dysfunction is involved in the process of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. 81

| Evaluation and diagnosis of nonalcoholic fatty liver disease/nonalcoholic steatohepatitis
The methods for evaluating nonalcoholic fatty liver disease and nonalcoholic steatohepatitis vary from study to study.
Representative methods 82 used in the literature will be discussed in this section.

| Pathologic diagnosis
A liver biopsy is the gold standard in the diagnosis of nonalcoholic steatohepatitis ( Figure 1 Also, Kleiner et al 85 scored liver tissue findings based on the degree of steatosis (score 0 to 3), the degree of lobular inflammation (score 0 to 3), and the frequency of hepatocyte ballooning (score 0 to 2), with total scores of 5 or more for nonalcoholic steatohepatitis, 2 or less for non-nonalcoholic steatohepatitis as definition of nonalcoholic fatty liver disease, and 3-4 for borderline cases; the total score is known as the nonalcoholic fatty liver disease activity score. In addition, they defined the stage of fibrosis using a score from 0 to 4, which is evaluated separately from the nonalcoholic fatty liver disease activity score.

| Abdominal sonography and computed tomography
Abdominal sonography (ultrasound) has a high detection capability in the presence or absence of moderate or high levels of fat deposits and, therefore, is useful in the diagnosis of nonalcoholic fatty liver disease. 82,86 However, it is difficult to assess the degree of inflammation and fibrosis. 87,88 It also cannot be used to differentiate between nonalcoholic fatty liver disease and early nonalcoholic steatohepatitis. 89,90 With this method, a fatty liver diagnosis was defined as a bright liver, increased liver echotexture compared with the kidneys, vascular blurring, and deep attenuation of the liver.
Abdominal computed tomography is also useful in the diagnosis of nonalcoholic fatty liver disease, and the liver-to-spleen ratio can be used to estimate the amount of fat deposition. 82,91 However, inflammation and fibrosis are difficult to determine by computed tomography, which cannot be used to identify nonalcoholic steatohepatitis. 92

| Blood biomarkers
Serum alanine aminotransferase, aspartate aminotransferase, gamma-glutamyl transpeptidase, platelets, albumin, triglyceride, cholinesterase, fasting plasma insulin, homeostasis model of assessment of insulin resistance, and other markers have been used as indicators of liver conditions. 82 Although there are no established biomarkers to detect nonalcoholic steatohepatitis, alanine aminotransferase may be a useful screening method for nonalcoholic fatty liver disease. 93 However, there is no consensus cutoff value for alanine aminotransferase, and it varies from 40 to 75 IU/L depending on the studies. [94][95][96] Alanine aminotransferase is also not a good indicator of the severity of the disease. In contrast, the ratio of aspartate aminotransferase to alanine aminotransferase is considered to be an indicator of fibrosis progression, and cutoff values of 1.0 for nonalcoholic steatohepatitis and 0.8 for nonalcoholic fatty liver disease are recommended.

| Formula scoring system
Several scoring systems that use a special formula have been developed for the diagnosis and prediction of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. The nonalcoholic fatty liver disease fibrosis score 97 is a system used to predict cases of advanced fibrosis. The formula for nonalcoholic fatty liver disease fibrosis score includes age, body mass index, impaired fasting glycemia or diabetes, the aspartate aminotransferase to alanine aminotransferase ratio, platelets, and albumin. The fatty liver index 98 has been developed to predict the onset of nonalcoholic fatty liver disease, and the formula consists of triglyceride, body mass index, gamma-glutamyl transpeptidase, and waist circumference. The fatty liver index was further modified for US citizens by taking ethnic differences into consideration. 99 The hepatic steatosis index 100 is a system for simplifying the nonalcoholic fatty liver disease evaluation, and the formula consists of alanine aminotransferase to aspartate aminotransferase ratio, body mass index, gender, and diabetes.

| EPIDEMIOLOG IC REL ATIONS HIP B E T WEEN PERI ODONTAL D IS E A S E AND NONALCOHOLIC FAT T Y LIVER DIS E A S E IN HUMANS
The relationship between periodontitis and liver disease has been previously discussed and is based on a growing number of epidemiologic studies. Between the 1990s and the early 2000s, Movin, 101 Novacek et al, 15  Recently, there has been a focus on the effects of periodontal disease on liver abnormalities, especially on nonalcoholic fatty liver disease. Thus, a literature search was conducted to answer the following question: Does periodontal disease affect the development or progression of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis? To answer that question, the following terms were searched in PubMed/MEDLINE: (periodontitis OR periodontal) AND (hepatic OR liver OR steatosis OR non-alcoholic fatty liver disease OR nonalcoholic fatty liver disease OR fatty liver OR NAFLD). Furthermore, filters for "Humans," "English," and "Adults: 19 years" were used. As a result, we found 154 articles.
We excluded studies on viral hepatitis and liver transplantation, case reports, animal studies, and studies with different objectives.
We also added six articles using a hand search. Consequently, 13 cross-sectional studies, two case-control studies, and three cohort studies were included (Table 1). These will be discussed in the following.

| Studies using biomarkers as an indicator of liver abnormalities
Early studies of the association between periodontal disease and liver abnormalities using blood biomarkers were conducted mainly in Japan. Saito et al 29 studied the association between periodontitis and liver status in 172 women with an average age of 40.9 years who attended a health promotion program. The results showed that age-adjusted regression coefficients of serum aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, gamma-glutamyl transpeptidase, cholinesterase, high-density lipoprotein cholesterol, fasting blood glucose, blood cell count, total protein, and urea were significantly associated with the severity of periodontitis. The levels of aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transpeptidase, lactate dehydrogenase, alkaline phosphatase, and cholinesterase in serum were significantly higher in patients with periodontitis than in nonperiodontitis patients. A linear multiple regression analysis was performed using data from these blood tests as independent variables, adjusted for age, smoking history, and oral hygiene; the results showed that serum aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transpeptidase, cholinesterase, lactate dehydrogenase, and high-density lipoprotein cholesterol (inversely proportional) were significantly correlated with the severity of periodontitis. Logistic regression analysis showed significant odds ratios for serum alanine aminotransferase, aspartate aminotransferase to alanine aminotransferase ratio, and cholinesterase for periodontitis (probing pocket depth of 4 mm and over) incidence with or without adjustment for body mass index, age, smoking history, oral hygiene, and/or body fat percentage.
In another region of Japan, Furuta et al 30 conducted a cross-sectional study of the relationship between periodontal disease and liver abnormalities in 2225 students that were 18 to 19 years of age. In male subjects, normal serum alanine aminotransferase levels (less than 20 IU/L) were observed in 95.8% of nonperiodontitis patients and in 4.2% of periodontitis patients, whereas abnormal levels were found in 87.4% of nonperiodontitis patients and in 12.6% of periodontitis patients. These differences between normal and abnormal levels were statistically significant. When using logistic regression analysis, males were significantly more likely to have periodontitis if their serum alanine aminotransferase was high (greater than or equal to 41 IU/L) than if it was low (adjusted odds ratio of 2.3). However, no significant relationship was found in females. These results differ from those of Saito et al, who found an association between periodontitis and liver abnormalities in females.
In addition, Ahmad et al 32 investigated the association between hepatic abnormality, metabolic syndrome, and periodontal status in 5477 employees of a manufacturing company in Japan.
They found that the mean probing pocket depth in the low alcohol consumption group with higher alanine aminotransferase and metabolic syndrome was significantly higher than the mean probing pocket depth in the normal alanine aminotransferase without metabolic syndrome group. However, no difference was found in females, which is partly consistent with the results of Furuta et al. The data obtained in this study were limited to measurements on radiographs and biomarkers in blood samples, which may have prevented the authors from finding a relationship.
3.1.2 | Studies using imaging and/or scoring systems to diagnosis nonalcoholic fatty liver disease The studies described so far have primarily used serum biomarkers as indicators of abnormalities in the liver, and in most cases, multivariate analyses have been performed with periodontal parameters as the dependent variable. However, the direction of research in cross-sectional studies has now focused on using periodontal disease parameters as the independent variable and liver disease parameters as the dependent variable. Accordingly, in addition to serum biomarkers, other diagnostic methods have been used as parameters of liver disease.
Alazawi et al 37 investigated the association between periodontitis and nonalcoholic fatty liver disease in two groups: a population-based study in the United States and a patient-based study in the UK. Data from the United States National Health and Nutrition Examination Survey III were used for the population-based study.
Periodontitis was defined as the presence of two or more sites with probing pocket depth of 3 mm or sites of 5 mm or more. Nonalcoholic fatty liver disease was defined using the nonalcoholic fatty liver disease fibrosis score. Although nonalcoholic fatty liver disease was significantly correlated with several periodontal parameters, only the mean probing pocket depth remained significant after adjustment for confounding factors. Furthermore, the percentage of subjects with a clinical attachment level of 3 mm or more were 7.5% in the low nonalcoholic fatty liver disease fibrosis score group and 14.7% in the moderate or higher nonalcoholic fatty liver disease fibrosis score group, and this difference was statistically significant. Similarly, the mean clinical attachment level was significantly higher in the group with moderate or higher nonalcoholic fatty liver disease fibrosis score. The patient-based study in the UK included 69 patients with a mean age of 49.2 years. Periodontitis was defined as the presence of a site with probing pocket depth 3.5-5.5 mm in more than two sextants or probing pocket depth greater than 5.5 mm. Nonalcoholic     18.1% in subjects with 0% of sites with probing pocket depth of 4 mm or more, 26.6% in the less than 30% group, and 39.2% in the 30% or more group. Periodontitis and nonalcoholic fatty liver disease were correlated with the level of serum C-reactive protein, but there was no significant association with weighted genetic C-reactive protein scores. Furthermore, when C-reactive protein was less than 1 mg/L, the adjusted prevalence odds ratio for nonalcoholic fatty liver disease for the 30% or more sites was 2.39, while the ratio for C-reactive protein 1-3 mg/L and greater than 3 mg/L was 0.97 and 1.12, respectively. In other words, there was a significant association between periodontitis and nonalcoholic fatty liver disease in subjects with low levels of C-reactive protein, but no relationship was found at higher levels of C-reactive protein. Based on these findings, the authors concluded that serum C-reactive protein may be a modifier of the relationship between periodontitis and nonalcoholic fatty liver disease. This finding may explain some of the variations in the relationship between periodontitis and nonalcoholic fatty liver disease.

| Case-control studies
Yoneda et al 21

| Cohort studies
A population-based cohort study was performed using the Study of Subjects were divided into 0%, less than 30%, and 30% or more of sites with 3 mm or more clinical attachment level or 4 mm or more probing pocket depth at baseline. The liver conditions after more than 5 years (median 7.7 years) were investigated by sonography and serum alanine aminotransferase. Relative to subjects without a clinical attachment level of 3 mm or more, the nonalcoholic fatty liver disease incidence was elevated in participants with both less than 30% and 30% or more of sites affected. The adjusted incidence rate ratio for nonalcoholic fatty liver disease was statistically significant at 1.28 for less than 30% of sites and 1.60 for 30% or more of sites affected, respectively. Similarly, the incidence difference was 5.49 for less than 30% of sites and 11.11 for 30% or more of sites affected with a statistically significant difference. On the other hand, no such dose-response relationship was observed for the probing pocket depth of 4 mm or more. In addition, in patients showing a clinical attachment level of 3 mm or more, the unadjusted incidence rate ratio for 1 mm or more of attachment loss during the observation period was 1.78, and that for 2 mm or more was 2.32, with statistical significance, but it did not reach the level of significance when adjusted. Thus, the authors of this study suggested that a history of periodontitis may be a risk factor for nonalcoholic fatty liver disease.
Helenius-Hietala et al 45  were performed annually. In addition, blood aspartate aminotransferase and alanine aminotransferase were measured. The number of sites with a probing pocket depth of 6 mm or more or a clinical attachment level of 6 mm or more at baseline was the independent variable, and the increase or decrease in aspartate aminotransferase or alanine aminotransferase over 8 years was the dependent variable, and these relationships were analyzed using logistic regression analysis, which was adjusted for confounding factors.
The relationships were also analyzed for individuals that smoked and consumed alcohol. Analysis showed that increased alanine aminotransferase was significantly correlated with periodontal parameters, with an adjusted odds ratio of 1.10 for a probing pocket depth of 6 mm or more and 1.03 for a clinical attachment level of 6 mm or more. However, there was no correlation with aspartate aminotransferase. In subjects with smoking habits but not drinking habits, alanine aminotransferase correlated significantly with probing pocket depth of 6 mm or more (adjusted odds ratio 1.20) and clinical attachment level of 6 mm or more (adjusted odds ratio 1.04).

| Systematic review and meta-analyses
One systematic review and two meta-analyses have been published on the relationship between periodontitis and nonalcoholic fatty liver disease ( Table 2). Alakhali et al 26 discussed 12 articles (N = 53384), and all but one of them found a significant correlation between periodontal or bacteriologic parameters and nonalcoholic fatty liver disease. The quality of the papers included was also assessed based on the Strengthening of Reporting of Observational Studies in Epidemiology guidelines, with four papers scoring 7, the highest points possible, four scoring 6, and the others 4-5, which can be considered good. However, the authors did not perform any statistical analysis, such as a meta-analysis, due to heterogeneity and inconsistency among the studies included.
In a review by Wijarnpreech et al, 28 five papers that met their inclusion criteria were selected. The unadjusted odds ratio for periodontitis with a probing pocket depth of 3.5-4 mm or more was sta-

| Summary of epidemiologic studies
Most evidence on the association between periodontitis and nonalcoholic fatty liver disease has been from cross-sectional studies.
Although significant associations have been found in most studies, results have varied, likely due to differences in age, gender, 29,30 and ethnicity. 39 In some cases, the significance of the association may have disappeared after adjusting for confounding factors, and a more detailed analysis of the factors and their synergistic effects on the association is necessary. Although cross-sectional studies alone do not reveal a causal relationship, three cohort studies 36,44,45 suggested that periodontitis is a potential risk factor for nonalcoholic fatty liver disease.
In addition, it has been suggested that P. gingivalis is involved in the progression of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. 21 Early publications 29,30,32 primarily used blood samples to assess nonalcoholic fatty liver disease, but their accuracy was limited, especially in assessing the severity of disease. Using a biopsy is the most useful method for the diagnosis of nonalcoholic steatohepatitis, but it is not practical for repeated assessments because it is an invasive test. Thus, imaging and scoring systems offer advantages to these other methods.
The relationship between periodontal disease and nonalcoholic fatty liver disease has been obtained primarily from observational studies. Although one study 21 suggested that nonsurgical periodontal treatment reduced P. gingivalis levels and improved liver health, the effect of periodontal treatment on liver disease is still largely unknown. Future randomized controlled trials on this topic will be needed to validate this claim.

| REL ATI ON S HIP B E T WEEN PERIODONTAL DIS E A S E , NONALCOHOLIC FAT T Y LIVER DIS E A S E/NONALCOHOLIC S TE ATOHEPATITIS , AND ME TABOLIC SYNDROME
Metabolic syndrome is a critical risk factor for both periodontal disease and nonalcoholic fatty liver disease. These diseases mediate a bidirectional three-way relationship, centered on insulin resistance associated with obesity and diabetes ( Figure 2).

| Role of metabolic syndrome and insulin resistance in the pathophysiology of nonalcoholic fatty liver disease/nonalcoholic steatohepatitis
Nonalcoholic fatty liver disease is considered the hepatic manifestation of metabolic syndrome because it is closely associated with obesity, insulin resistance, hypertension, and dyslipidemia. 9,10 TA B L E 2 Summary of the systematic review and meta-analyses It is important to highlight that obesity characterized by excess adipose tissue due to an increase in the number and volume of adipocytes is strongly associated with the development of nonalcoholic fatty liver disease, since it causes fat accumulation in the liver through insulin resistance. Adipose tissue is a multifunctional organ that regulates energy consumption, insulin sensitivity, and inflammatory processes via various inflammatory mediators. 104,105 In obese people, excess adipose tissue mediates several negative effects, such as increased macrophage infiltration, disruption of adipocytokine production (IL-1β, IL-6, tumor necrosis factor alpha, leptin, resistin, visfatin, adiponectin, plasminogen activator inhibitor-1, etc), and subsequent defective insulin secretion. 106 Hyperinsulinemia promotes further obesity because insulin is an anabolic hormone that promotes glucose uptake and fat storage.
In addition, increased blood levels of proinflammatory adipokines produced in inflamed adipose tissue cause insulin resistance, which is accompanied by low-grade systemic inflammation, resulting in increased hepatic influx and accumulation of fatty acids. [107][108][109][110] However, reduction of serum adiponectin, which has an anti-inflammatory effect, may also induce hepatic fat accumulation, inflammation, and insulin resistance. 106,111 Many studies have shown that the inflammation occurs as a consequence of obesity, and it may cause insulin resistance and other disturbances of energy homeostasis.
In fact, both excessive body mass index and visceral obesity are recognized as risk factors for nonalcoholic fatty liver disease, and nearly two-thirds of patients with obesity and type 2 diabetes have hepatic steatosis. 112,113 In patients with nonalcoholic fatty liver disease, the presence of multiple components of metabolic syndrome are associated with more severe liver disease and are more likely to progress to nonalcoholic steatohepatitis and cirrhosis. 114 specimens, and this is thought to be mediated by inflammatory cytokines. Furthermore, the presence of metabolic syndrome among nonalcoholic fatty liver disease patients is associated with an increased risk for fibrosis in nonalcoholic steatohepatitis and the risk for eventual liver failure. 117,118 The liver is not simply a passive participant, since hepatic steatosis has systemic consequences as it worsens metabolic syndrome. 119 For example, nonalcoholic fatty liver disease itself has been reported to enhance insulin resistance, predict the emergence of metabolic complications, and increase the risk for cardiovascular events. 73,120 Intracellular lipid content in the liver also decreases insulin clearance, causing the hyperinsulinemia, which is a sign of prediabetes. 121 In other words, even if nonalcoholic steatohepatitis does not directly lead to end-stage liver disease, it may have a significant impact by promoting extrahepatic complications in individuals with metabolic syndrome. 122

| Bidirectional relationship between periodontal disease and nonalcoholic fatty liver disease
Given the close connection between nonalcoholic fatty liver disease and metabolic syndrome and the fact that periodontal disease is bidirectionally associated with metabolic syndrome, it is important to consider periodontal disease in the pathology of nonalcoholic fatty liver disease. 55 Numerous epidemiologic studies have shown that periodontal disease can exacerbate various metabolic disorders, such as diabetes, obesity, dyslipidemia, and chronic kidney disease. [123][124][125] Periodontitis-related systemic inflammation may contribute to insulin resistance through elevated blood levels of adipocytokines, such as tumor necrosis factor alpha, IL-6, and leptin, which inhibit the insulin receptor and its downstream signaling. [126][127][128][129] The presence of both obesity and periodontal disease significantly increases the risk for diabetes because of the exacerbated insulin resistance due to periodontitis, which further increases glucose and insulin levels in blood. 130 Insulin resistance also promotes dyslipidemia through increased circulating free fatty acids in blood. Furthermore, periodontal disease is directly and indirectly involved in cardiovascular disease owing to its exacerbation of systemic inflammation and metabolic syndrome. 131,132 Proinflammatory mediators and periodontopathic bacteria and their products may damage endothelial cells and promote atherogenesis and thrombus formation, thereby increasing the risk for cardiovascular disease. [133][134][135] Further, intervention studies have reported that periodontal treatment improves insulin resistance, blood glucose levels, lipid profiles, and endothelial function in patients with periodontitis. [136][137][138][139][140] F I G U R E 2 Bidirectional three-way relationship among metabolic syndrome, nonalcoholic fatty liver disease (NAFLD)/nonalcoholic steatohepatitis (NASH), and periodontal disease, centering on insulin resistance. The black arrows indicate an established link by accumulated evidence. Red arrows indicate possible link that still has unproven causality. Blue arrow indicates indeterminate link because of no or little evidence It is well known that diabetes and obesity negatively impact the progression of periodontal disease. 141,142 Poor glycemic control in diabetic patients has been correlated with increased risk for periodontal attachment loss and tooth loss compared with nondiabetic subjects. 143,144 Through the formation of advanced glycation end products and a glucose-rich environment, diabetes can accelerate the inflammatory process and inhibit wound healing in the periodontal tissues, thereby promoting tissue destruction by periodontitis. 145 Therefore, the latest classification of periodontal disease includes diabetes as a critical element in determining the grade of periodontitis, and its importance as a risk factor for the progression of periodontal disease is emphasized. 146 In terms of obese patients, they have approximately twice the risk for periodontal disease and their condition may negatively affect the responsiveness to periodontal treatment compared with normal weight subjects. 129,147,148 The mechanism by which obesity exacerbates periodontitis is still unclear, but increased adipokines in the gingival crevicular fluid, decreased periodontal immune response, and impaired gingival microcirculation have been proposed. 145

| Association between periodontal disease and nonalcoholic fatty liver disease/ nonalcoholic steatohepatitis with a focus on metabolic syndrome
As noted for the various metabolic disorders mentioned previously, periodontal disease can affect nonalcoholic fatty liver disease and nonalcoholic steatohepatitis via disturbances in energy homeosta- sis. An updated meta-analysis using four cross-sectional studies and one retrospective cohort study showed that the association between periodontitis and nonalcoholic fatty liver disease was no longer significant when adjusting for insulin resistance and various metabolic parameters, suggesting that those metabolic conditions (and not periodontitis itself) are predisposing factors for nonalcoholic fatty liver disease. 28 However, animal studies have shown that periodontal inflammation and infection by periodontal pathogens can cause mild fatty liver and hepatitis, even in healthy animals without metabolic disease. 24,149 For example, studies using a ligature-induced periodontitis rodent model have reported an altered hepatic glycolipid metabolism through increased blood levels of inflammatory cytokines, total cholesterol, triglycerides, and oxidative stress. 50 However, these studies, because of their cross-sectional nature, do not support a causal relationship, and the mechanisms involved have not been fully examined. Thus, to the best of our knowledge, there is currently limited evidence that liver disease, at least nonalcoholic fatty liver disease and nonalcoholic steatohepatitis, affects periodontal disease.

| P OTENTIAL DUAL PATHWAYS LINKING PERI ODONTAL D IS E A S E AND NONALCOHOLIC FAT T Y LIVER DIS E A S E/ NONALCOHOLIC S TE ATOHEPATITIS
Although the mechanism by which harmful factors are transported from diseased periodontal tissue to the liver is unclear, the following two routes have been proposed based on the unique anatomical characteristics of the liver (Figure 3).

| Periodontal microulceration, general circulation, and hepatic arterial system
One possible route connecting periodontal disease and nonalcoholic fatty liver disease/nonalcoholic steatohepatitis is the hematogenous physical diffusion of immunogenic factors and oral pathogenic bacteria from the periodontal tissues. The mechanism linking periodontal disease to systemic disease has long been explained by the concept of microulceration in the periodontal pocket. 1,132,157,158 The gingival epithelium in a healthy periodontium normally covers the connective tissue, including blood and lymph vessels, and acts as a barrier to obstruct noxious biofilm components. 159,160 However, in diseased tissues, increases in permeability and microulceration of the gingival epithelium readily allow invasion of noxious substances and microorganisms into the circulation via the periodontal tissues. 157,161 In addition, inflammation-induced capillary structural changes, vasodilation, and perturbed blood flow may enhance the diffusion of pathogenic factors. 162,163 The hematogenous diffusion is known to be further enhanced by mechanical perturbation of the gingival tissues. Studies have revealed that oral mechanical injuries caused by daily dental activity (eg, brushing, flossing, chewing), periodontal procedures (eg, scaling and root planing, probing), and other dental procedures (eg, orthodontics, tooth extraction) cause a bacteremia. [164][165][166] Patients with periodontal disease show a further increase in serum/circulating bacteria and lipopolysaccharide derived from these oral injuries compared with individuals with healthy periodontal tissue. 167,168 Specific periodontal pathogens and other oral bacteria have been detected in areas distant from the oral cavity, including atherosclerotic plaques, joint cavities, the brain, and the liver, suggesting their association with various systemic diseases. 22,134,169,170 Furthermore, periodontal host cells activated by immune interactions with biofilm bacteria enhance the release of reactive oxygen species and inflammatory cytokines, such as IL-1β, IL-6, and tumor necrosis factor alpha. [171][172][173] It has been reported that these pro-inflammatory cytokines and oxidative stress molecules are elevated in patients with periodontitis, not only in gingival crevicular fluid and gingival tissue but also in serum. 128,129,174 Therefore, potential liver damage derived from periodontal disease may be delivered to the liver in a hematogenous manner and it may promote the progression of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. The various substances transferred into the blood via the capillaries of the periodontal tissue first pass through the left and right jugular veins, then they join the superior vena cava and then flow into the heart. After entering the pulmonary circulation for gas exchange, they are pumped from the heart through the aorta and then diffuse throughout the F I G U R E 3 Dual possible pathway for the link between periodontal disease and nonalcoholic fatty liver disease (NAFLD)/nonalcoholic steatohepatitis (NASH). A, One possible mechanism is hematogenous systemic diffusion of bacteria, endotoxin, and inflammatory mediators through microulceration in the periodontal pocket. A proper hepatic artery, which is branched from the abdominal aorta, is presumed the main transportation route from systemic circulation to liver. B, Another mechanism is gut microbial dysbiosis induced by the transport of oral bacteria through the gastrointestinal tract. The oral bacteria-mediated gut dysbiosis can cause impairment of the gut barrier function and immune modulation, further leading to hepatic exposure to bacteremia, endotoxemia, and bacterial metabolite through the enterohepatic circulation by portal vein system. IL-1β, interleukin 1 beta; IL-6, interleukin 6; TNF-α, tumor necrosis factor alpha; FFA, free fatty acids; SCFA; short-chain fatty acids; EtOH, ethanol body by the systemic circulation. Regarding the liver, the proper hepatic artery, the potentially vegetative blood vessel of the liver branching from the abdominal aorta, can be presumed the main transportation route.
Indeed, epidemiologic studies have shown that C-reactive protein, which is synthesized in hepatocytes and activated by proinflammatory cytokines, including tumor necrosis factor alpha and IL-6, is a modifying factor of periodontitis and nonalcoholic fatty liver disease. 35,40 Serum C-reactive protein levels are also known to increase with the severity of periodontal disease. In addition, animal studies have reported that increased serum levels of proinflammatory cytokines and oxidative stress markers, as well as C-reactive protein, may contribute to nonalcoholic fatty liver disease progression after inducing periodontitis. 20,149,151 Our previous study also showed that adding P. gingivalis to ligature-induced periodontitis in rats exacerbated nonalcoholic fatty liver disease, which was accompanied by increased serum lipopolysaccharide activity and C-reactive protein. 53 These data suggest that periodontally derived circulating inflammatory molecules play a critical role in the pathogenesis of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis.
As for periodontopathic bacteria, Furusho et al 22 reported that P. gingivalis was detected by immunochemical staining in 52.5% of liver biopsy specimens from nonalcoholic steatohepatitis patients.
The P. gingivalis-positive liver cases showed a significantly higher fibrosis score than the P. gingivalis-negative cases. Furthermore, Ishikawa et al 23 found that orally administered SNAP26b-tagged P. gingivalis in mice was detected in the liver tissue, and the translocation of P. gingivalis from the oral cavity to the liver was further promoted by hyperglycemia. Interestingly, P. gingivalis is known to have the ability to invade and survive inside immune cells, such as macrophages and dendritic cells, 175,176 suggesting that periodontopathic bacteria may hijack circulating leukocytes to serve as Trojan horses for dissemination of infection from the oral cavity to the liver. 177

| Gut microbial dysbiosis and enterohepatic circulation
Another potential route of communication between the oral cavity and the liver is via transport of oral bacteria through the gas- It is widely known that gut microbiome dysbiosis is closely associated with nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. [186][187][188] All blood from the gut travels via the portal vein to reach the liver, which performs the metabolic, immunological, and detoxification processes before the blood reaches the systemic circulation. 5,186 Therefore, through the enterohepatic circulation, the liver is constantly exposed to bacterial components and metabolites absorbed from the gut, which can potentially affect the condition of the liver. In fact, it is known that in gut dysbiosis there is an increase in choline metabolism (which is essential for lipolysis), hepatotoxins (such as lipopolysaccharide and ethanol), and volatile organic compounds. [189][190][191][192] Furthermore, dysbiosis enhances intestinal permeability by impairing intercellular tight junctions in the gut wall and promotes the transfer of hepatotoxins and enterobacteria to the liver. 193,194 From the foregoing, it has been suggested that dysbiosis due

| P OTENTIAL MECHANIS MS BY WHI CH PERI ODONTAL D IS E A S E MAY IN CRE A S E THE RIS K OF NONALCOHOLIC FAT T Y LIVER DIS E A S E/NONALCOHOLIC S TE ATOHEPATITIS
The liver, which is located at a hemodynamic convergence point in the body, connects the hepatic arterial and portal systems, allowing a mixture of oxygenated blood and blood from the portal system.
Therefore, in a state of periodontitis, the liver is under constant exposure to various pathogenic factors that are diffused systemically from the oral cavity, such as bacteria and their components, inflammatory cytokines, and reactive oxygen species, and these can be involved in the disease promotion of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis (Figure 4).

| Periodontopathic bacteria
Data from over the last decade strongly suggest that P. gingivalis is involved in nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. P. gingivalis has many virulence factors (such as collagenase, aminopeptidase, and a trypsin-like enzyme) and other components (including lipopolysaccharide and fimbriae) that are also known to trigger intracellular signaling events. 195 Yoneda et al, 21  Our previous animal studies showed that a combination of P. gingivalis infection with ligature-induced periodontitis increased serum levels of alanine aminotransferase and lipopolysaccharide as well as hepatic fat deposition in rats with high-fat diet-induced obesity and insulin resistance. 53,55 However, the intervention with either P. gingivalis or ligature placement alone did not show similar changes.
Therefore, our results suggest that P. gingivalis or its products may enter the blood circulation via the inflamed periodontal tissues and thereby contribute to nonalcoholic steatohepatitis progression. As already mentioned (Section 5.1), it is known that P. gingivalis can diffuse from the oral cavity to the systemic circulation and reach the liver. 22,23 Some studies have clarified the molecular mechanism and ex- with healthy livers. 193,194,201 Moreover, the inhibition of lipopolysaccharide receptors was associated with significant protection against the development of nonalcoholic fatty liver disease in various animal models. 202,203 In patients with periodontal disease, the degree and frequency of endotoxemia is increased with the severity of the disease, 168,204 which contributes to systemic inflammation, including increased blood levels of C-reactive protein, IL-6, and tumor necrosis factor alpha. 205,206 During the development of oral dysbiosis, periodontopathic bacteria of the genera Fusobacterium, Porphyromonas, and Prevotella are predominant in the periodontal microenvironment, thereby promoting lipopolysaccharide production. 207 The endotoxemia can be explained not only by the translocation of lipopolysaccharide from inflamed periodontal tissue to the systemic circulation, but also by liver exposure to lipopolysaccharide via the portal system due to periodontitis-induced gut dysbiosis (see Section 5, Figure 3).
Although a trace amount of lipopolysaccharide (about 1.0 ng/mL) is usually present in the portal circulation even under normal physiologic conditions, the liver of a healthy subject is hardly responsive F I G U R E 4 Mechanisms through which periodontal disease increases the risk of nonalcoholic fatty liver disease (NAFLD)/nonalcoholic steatohepatitis (NASH). LPS, lipopolysaccharide; TNF-α, tumor necrosis factor alpha; IL-6, interleukin 6; ROS, reactive oxygen species; FFA, free fatty acids; TLR, toll-like receptor; PAR2, protease activated receptor 2; TGF-β, transforming growth factor beta to such low concentrations of this endotoxin. 208 In this regard, it is known that fatty liver increases hepatic macrophages (Kupffer cells) and enhances their susceptibility to the low-dose lipopolysaccharide. 209 Imajo et al 210 54 The results showed that most of the lipopolysaccharide spread through the circulation and accumulated markedly in the liver more than in other organs, including the kidney, brain, and spleen. It is noteworthy that this accumulation of P. gingivalis lipopolysaccharide was increased and maintained in the fatty liver for a longer period of time than in the healthy liver. Furthermore, in ongoing studies in our laboratory, we are finding that the high-fat diet may delay the metabolic clearance of P. gingivalis lipopolysaccharide from the liver.
This change in lipopolysaccharide kinetics in fatty liver may be due to the aforementioned increased Kupffer cells and upregulation of toll-like receptor 4 and toll-like receptor 2 pathways.

| Proinflammatory cytokines and adipokines
Adipokines, such as tumor necrosis factor alpha, IL-6, leptin, and adiponectin produced by adipose tissue, are closely involved with hepatic lipid deposition, inflammation, fibrosis, and carcinogenesis in nonalcoholic fatty liver disease. 215 In the enlarged adipose tissue of obese people, increased secretion of chemoattractant protein-1 causes an infiltration of inflammatory cells, primarily macrophages, which then secrete inflammatory cytokines and chemokines that disrupt the balance of adipokine production by adipocytes. 216 Adipokines affect not only chronic inflammation and insulin resistance in local adipose tissue, but also hepatic insulin sensitivity directly. 81 Periodontal disease is characterized by a low-grade systemic inflammatory state that increases blood levels of proinflammatory cytokines, including tumor necrosis factor alpha and IL-6, similar to obesity, suggesting a potential risk for nonalcoholic fatty liver disease in the bidirectional relationship between periodontal disease and obesity (see Section 4, Figure 2). In particular, tumor necrosis factor alpha plays a major role in hepatic insulin resistance, and it inactivates the insulin receptor substrate by serine phosphorylation through activation of a serine/threonine kinase, thus blocking the insulin receptor signaling cascade. 217 IL-6, which is upregulated by tumor necrosis factor alpha, is also associated with decreased insulin signaling and induction of fatty acid oxidation, as well as secretion of C-reactive protein by the liver. 145,218 Like adipocytes, cells within periodontal tissue can also secrete various adipokines. 145

| Oxidative stress
Oxidative stress is defined as a deleterious condition resulting from an imbalance between reactive oxygen species and antioxidant capacity. 230,231 Reactive oxygen species is a collective term that broadly describes a variety of molecules derived from oxygen molecules and free radicals: singlet oxygen, superoxide, hydrogen peroxide, hydroxyl, and nitric oxide. 232 Under physiologic conditions, these reactive oxygen species effects are rapidly eliminated by antioxidant defenses and repair enzymes in the body. 233 However, when excessive reactive oxygen species are produced, this causes nonspecific cell death and tissue injury through oxidative damage to deoxyribonucleic acid (DNA), fatty acids, and proteins due to their high reactivity.
In the development of periodontal disease, activated polymorphonuclear leukocytes produce a large amount of reactive oxygen species, which are involved in periodontal tissue destruction. 234,235 Oxidative stress is also one of the major mediators used to explain the mechanism connecting periodontitis and systemic diseases, because it is associated with various diseases, including periodontitis, obesity, and nonalcoholic fatty liver disease. 236 243,244 MicroRNAs may also play an important role in the pathogenesis of nonalcoholic fatty liver disease, and they have recently been explored as new molecular markers for the diagnosis and prognosis of fatty liver. 245 In addition, circulating microRNAs from some organs, such as adipose tissue, are known to act as metabolic regulators and alter specific gene expression in the liver. 246 Although the study of microRNAs in periodontology is still at an early stage, one study using a ligature-induced periodontitis mouse model reported that changes in blood micro-RNAs were consistent with hepatic apoptosis-related messenger RNA expression. 247

| OR AL AND G UT MI CROB I OME-TARG E TED PROB I OTI C THER APY IN MANAG EMENT OF NONALCOHOLIC FAT T Y LIVER D IS E A S E
Periodontal disease is currently considered to be the result of a harmful shift in the balance of the normally stable resident oral microbiota. 248 As mentioned earlier, gut dysbiosis induced by enteral translocation of periodontopathic bacteria may be involved in nonalcoholic fatty liver disease. One mechanism assumed to link the gut microbiome with nonalcoholic fatty liver disease is the disruption of the gut epithelial barrier, which may allow leakage of microbial products and metabolites into the portal circulation. Namely, changes in lipopolysaccharide and bacterial metabolites due to gut dysbiosis can induce intestinal inflammation and increase permeability, thereby promoting hepatic exposure to these components, which can directly cause nonalcoholic fatty liver disease and liver fibrosis. 249 Thus, there is increasing interest in the potential of the human oral and gut microbiome to serve as a target for prophylactic and therapeutic interventions in nonalcoholic fatty liver disease.
Diverse strategies for manipulating the gut microbiome in the management of nonalcoholic fatty liver disease have been proposed, including the use of antibiotics, probiotics, prebiotics, and symbiotics (a combination of probiotics and prebiotics). Probiotics are defined as live cultures of microorganisms that are beneficial to the human body. 250 Prebiotics, fermentable foods that contain dietary fiber, have an indirect effect on the human body by affecting the activity of probiotics. 251 Antibiotics exert beneficial effects on metabolic disorders by nonspecifically suppressing the microbiome, but they may be accompanied by harmful side effects and potential emergence of antibiotic-resistant bacterial strains. Therefore, recently, supplementation with probiotics and symbiotics in the treatment of nonalcoholic fatty liver disease has been favorably accepted due potential enhanced safety for humans and the environment. [252][253][254] Preclinical animal studies have shown that probiotics suppress the development of insulin resistance and hepatic inflammatory signaling and improve steatosis through regulation of the gut microbiota. [255][256][257][258] A recent systematic meta-analysis by Sharpton et al, 252 which consisted of 21 randomized clinical trials, revealed that the use of probiotics or symbiotics improved liver-specific markers of hepatic function (alanine aminotransferase), liver stiffness measurements, and liver steatosis in patients diagnosed with nonalcoholic fatty liver disease.
In the oral context, the application of probiotics in the treatment of gingivitis and periodontitis can improve microbiological outcomes in saliva and subgingival plaque with or without nonsurgical periodontal treatment, such as scaling and root planing. 259  Recently, our studies have reported that an antimicrobial peptide, nisin, which is produced primarily by Lactococcus species, has effectiveness in the context of periodontal disease. 259,260 Nisin, a type of bacteriocin, belongs to a group of cationic peptide antimicrobials collectively called Type A (I) lantibiotics. 261 Nisin and other lantibiotics have gathered a lot of attention in the food industry and the medical field because of their potent and broad-spectrum activity even at trace concentrations, low cytotoxicity at antibacterial concentrations, and low likelihood of promoting the development of bacterial resistance. [262][263][264][265] Interestingly, our data showed that in oral salivary-derived biofilms, nisin-producing Lactococcus lactis and nisin reduce the levels of bacterial pathogens while retaining oral commensal bacteria, such as Neisseria species. 260 The probiotic L. lactis and nisin also significantly inhibited the formation, structure, and viability of biofilms spiked with periodontopathic bacteria. We

CO N FLI C T O F I NTE R E S T
The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.