Saito et al, 2006, Japan Cross-sectional N = 172 (all participants were females, mean age 40.9 y) No Periodontitis subjects having at least one sextant with deepest periodontal probing depth ≥4 mm (code 1 or 2) Eight designated molars and two incisors with periodontal probing depth recorded

Liver steatosis, hepatic condition

Percentage body fat

Elevated levels of aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transpeptidase, lactate dehydrogenase, alkaline phosphatase, cholinesterase

aspartate aminotransferase: ≥32 IU/L

alanine aminotransferase: ≥32 IU/L

aspartate aminotransferase-alanine aminotransferase ratio: ≤1

cholinesterase: ≥1.23 (Δ pH)

Age-adjusted regression analysis

Logistic regression models

Periodontitis as the dependent variables

Regression coefficient:

Aspartate aminotransferase 0.98

Alanine aminotransferase 0.56

Cholinesterase 40

Adjusted odds ratio (95% confidence interval) for periodontitis:

Aspartate aminotransferase 4.88 (1.18-20.21)

Alanine aminotransferase 6.79 (1.27-36.36)

Aspartate aminotransferase-alanine aminotransferase ratio 2.34 (0.99-5.5)

Cholinesterase 3.82 (1.33-10.96)

Yes Hepatic steatosis is associated with periodontitis in Japanese women 29 Furuta et al, 2010, Japan Cross-sectional N = 2225 (1264 males and 961 females, aged 18-19 y) Yes Presence of ≥1 teeth with periodontal probing depth ≥4 mm Randomly selected quadrants, one maxillary and one mandibular with periodontal probing depth and percentage bleeding on probing recorded

Alanine aminotransferase normal: ≤20 IU/L

Alanine aminotransferase subclinical: 21-40 IU/L

Alanine aminotransferase abnormal: ≥41 IU/L

Logistic regression analysis

Periodontitis as the dependent variables

Adjusted odds ratio (95% confidence interval) for periodontitis:

Males, alanine aminotransferase 2.3 (1.0-5.2)

Females, alanine aminotransferase 1.0 (0.1-9.3)

Yes (for males) Elevated alanine aminotransferase is a potential risk indicator for periodontitis among healthy young males 30 Yoneda et al, 2012, Japan Case-control study

N = 210 (150 nonalcoholic fatty liver disease, 102 nonalcoholic steatohepatitis, and 48 nonalcoholic fatty liver) patients, 64 males and 86 females, mean age of 54.6 y

N = 60 healthy subjects, 29 males and 31 females, mean age of 52.9 y

No Detection of Porphyromonas gingivalis, Treponema denticola, Prevotella intermedia, Tannerella forsythia, Aggregatibacter actinomycetemcomitans, and Campylobacter rectus by polymerase chain reaction technique Not mentioned

Histopathologic findings (liver biopsy)

Steatosis and necroinflammatory activity (criteria of Matteoni et al)

P. gingivalis–positive rate (%)

Multiple regression analysis: liver disease as the dependent variables

Rate of various fimbriae A types on nonalcoholic fatty liver disease patients (%)

P. gingivalis (+): non-nonalcoholic fatty liver disease, 21.7%; nonalcoholic fatty liver disease, 35.4%; nonalcoholic steatohepatitis, 52.0%

Adjusted odds ratio (95% confidence interval) for nonalcoholic fatty liver disease: 2.62 (1.00-6.83)

94.3% of P. gingivalis–positive specimens were invasive fimbriae A genotypes

Yes (not for nonalcoholic fatty liver to control) P. gingivalis infection was noted at a significantly high frequency in nonalcoholic fatty liver disease and nonalcoholic steatohepatitis patients 21 Case-series N = 10 No Patients had a periodontal probing depth of >5 mm in at least four teeth Not mentioned Abnormal levels of aspartate aminotransferase and alanine aminotransferase Level of aspartate aminotransferase and alanine aminotransferase decreased Yes Ahmad et al, 2015, Japan Cross-sectional N = 5477 (4207 males, mean age of 45.4 years and 1270 females, mean age 45.9 y) Yes, interexaminer Not mentioned Mesio-buccal and mid-buccal sites for all teeth, except for third molars, with periodontal probing depth and clinical attachment level Alanine aminotransferase ≥40 IU/L

Multiple regression models

Periodontitis as the dependent variables

Mean and standard deviation of periodontal probing depth (mm) in low alcohol consumption group in males

Elevated alanine aminotransferase (−), metabolic syndrome (−): 2.09 ± 0.36

Elevated alanine aminotransferase (+), metabolic syndrome (−): 2.12 ± 0.35

Elevated alanine aminotransferase (−), metabolic syndrome (−): 2.18 ± 0.41

Elevated alanine aminotransferase (−), metabolic syndrome (+): 2.21 ± 0.36

Yes Significant association of liver abnormalities and metabolic syndrome with periodontal condition in males with low alcohol consumption 32 Wiener et al, 2016, USA Cross-sectional N = 5758 (50.1% females, 41.9% 30-44 y, 28.7% 45-54 y, 29.4% 55-69 y) No Following American Academy of Periodontology/Centers for Disease Control and Prevention definition: mild periodontitis, moderate periodontitis, severe periodontitis Not mentioned Alanine aminotransferase ≥40 IU/L

Logistic regression analysis

Periodontitis as a dependent variable

Adjusted odds ratio: 1.17 (0.85-1.60) No (yes for unadjusted odds ratio) Positive but attenuated association of periodontitis and alanine aminotransferase failed to reach significance when other known, strong factors of periodontitis were included in the analysis 34 Akinkugbe et al, 2017, Pomerania Cohort study N = 2623 (41% males and 59% females, mean age of 46 y) No Proportion of sites with clinical attachment level ≥4 mm or periodontal probing depth ≥3 mm (0%, <30%, ≥30%) Mesio-buccal, mid-buccal, disto-buccal, and mid-lingual site for all teeth except for third molars in two quadrants, with periodontal probing depth and clinical attachment level

Abdominal sonography

Serum alanine aminotransferase >0.57 μmol/system of units (34.2 IU/L) for men, >0.4 μmol/system of units (24 IU/L) for women

Median of 7.7 y incidence

Weighted Poisson regression estimated

Median of 7.7 y incidence

Incidence rate difference with multiple imputation

Liver disease as a dependent variable

Adjusted incidence rate relative to no site of clinical attachment level of 3 mm:

 <30%: 1.28 (0.84-1.95)

 ≥30%: 1.60 (1.05-2.43)

Adjusted incidence rate difference relative to no site of clinical attachment level of 3 mm:

 <30%: 5.49 (−2.53-13.5)

 ≥30%: 11.9 (4.09-19.6)

Adjusted incidence rate relative to no site of periodontal probing depth of 4 mm:

 <30%: 1.53 (1.00-2.35)

 ≥30%: 0.77 (0.44-1.33)

Adjusted incidence rate difference relative to no site of periodontal probing depth of 4 mm:

 <30%: 14.6 (8.87-20.4)

 ≥30%: −6.34 (−13.7-1.02)

Yes History of periodontitis as an independent risk factor contributing to nonalcoholic fatty liver disease incidence in a population-based sample 36 Akinkugbe et al, 2017, Pomerania Cross-sectional study N = 2481 (55% females, mean age of 47 y) No Proportion of sites with periodontal probing depth ≥3 mm (0%, <30%, ≥30%) 4 sites per tooth on 2 quadrants Liver ultrasonography: increase in liver echogenicity

Logistic regression analysis

Stratified according to the median value (1.98) for the C-reactive protein–specific weighted genetic score (wGSCRP) and for low (<1 mg), intermediate (1-3 mg) and high (>3 mg) levels of serum C-reactive protein

Liver disease as a dependent variable

Adjusted prevalence odds ratio (95% confidence interval):

Subjects with wGSCRP ≤ 1.98

 <30% sites of periodontal probing depth ≥4 mm: 1.08 (0.75-1.57)

 ≥30% sites of periodontal probing depth≧4 mm: 1.14 (0.72-1.80)

 Subjects with wGSCRP > 1.98

 <30% sites of periodontal probing depth ≥4 mm: 1.33 (0.94-1.89)

 ≥30% sites of periodontal probing depth≧4 mm: 1.65 (1.07-2.55)

 Subjects with serum C-reactive protein <1 mg

 <30% sites of periodontal probing depth ≥4 mm: 1.62 (1.00-2.61)

 ≥30% sites of periodontal probing depth ≥4 mm: 2.39 (1.32-4.31)

 Subjects with serum C-reactive protein 1-3 mg:

 <30% sites of periodontal probing depth ≥4 mm: 1.37 (0.90-2.08)

 ≥30% sites of periodontal probing depth ≥4 mm: 0.97 (0.57-1.66)

 Subjects with serum C-reactive protein >3 mg

 <30% sites of periodontal probing depth ≥4 mm: 0.70 (0.45-1.10)

 ≥30% sites of periodontal probing depth ≥4 mm: 1.12 (0.65-1.93)

Yes (for interaction for serum C-reactive protein levels) Periodontitis was positively associated with higher prevalence odds of nonalcoholic fatty liver disease and this relationship was modified by serum C-reactive protein levels 35 Widita et al, 2017, Japan Cohort study N = 265 (133 males and 132 females, mean age of 72.5 y) Yes, interexaminer Periodontal probing depth ≥6 mm and clinical attachment level ≥6 mm Six sites around each tooth Elevation of aspartate aminotransferase, alanine aminotransferase, and aspartate aminotransferase/alanine aminotransferase ratio in 8 y

Logistic regression analysis

Liver disease as a dependent variable

Stratified according to smoking status and alcohol drinking habits

Adjusted odds ratio (95% confidence interval) for fatty liver index

Aspartate aminotransferase as a dependent variable

Periodontal probing depth ≥6 mm: 1.10 (0.99-1.22)

Clinical attachment level ≥6 mm: 1.02 (0.99-1.05)

Alanine aminotransferase as a dependent variable

Periodontal probing depth ≥6 mm: 1.10 (1.00-1.21)

Clinical attachment level ≥6 mm: 1.03 (1.00-1.06)

Smokers

 periodontal probing depth ≥6 mm: 1.20 (1.00-1.26)

Clinical attachment level ≥6 mm: 1.04 (1.00-1.07)

Yes for alanine aminotransferase levels (significant interaction of alanine aminotransferase with smoking status) The elevation of alanine aminotransferase levels might be associated with clinical periodontal parameters among with clinical periodontal parameters among non-institutionalized Japanese elderly, and this association was modified by smoking status 44 Alzawi et al, 2017, USA and UK Cross-sectional (population-based and patient-based) study

Population-based study in USA

N = 8172 (3796 males and 4376 females, 20-74 y)

Population-based study

No

Population based

2 sites with periodontal probing depth ≥3 mm from different sextans or serum immunoglobulin G antibodies against 19 bacterial species in 8153 participants aged ≥40 y

Not mentioned

Population-based study

Presence of steatosis on gallbladder ultrasonography

Nonalcoholic fatty liver disease fibrosis score

Population-based study

Logistic regression analysis

Liver disease as a dependent variable

Population based study

Unadjusted odds ratio (95% confidence interval)

Bleeding on probing (%): 1.10 (1.04-1.07)

Periodontal probing depth ≥4 mm (%): 1.06 (1.01-1.10)

Mean periodontal probing depth: 1.11 (1.05-1.18)

Clinical attachment level >3 mm (%) : 1.13 (1.06-1.20)

Mean clinical attachment level: 1.12 (1.04-1.21)

Adjusted for demographic socioeconomic and behavioral factors

Bleeding on probing (%): 1.07 (1.00-1.17)

Mean periodontal probing depth: 1.11 (1.05-1.08)

Adjusted for demographic socioeconomic factor, behavioral factors, and cholesterol

Mean periodontal probing depth: 1.08 (1.00-1.17)

Odds ratio (95% confidence interval)

Antibodies of Selenomonas noxia: 1.13

Antibodies of Streptococcus oralis: 1.14

Yes (for unadjusted model and some adjusted models) Complementary evidence from an epidemiologic survey and a clinical study show that nonalcoholic fatty liver disease is associated with periodontitis and the association is stronger with significant liver fibrosis 37

Patient-based study in UK

N = 69 (periodontitis patients: mean age of 49.2 y, no periodontitis patients: mean age of 50.6 y)

Patient-based study

No

Patient-based study

Basic periodontal examination code 3 (periodontal probing depth: 3.5-5.5 mm) in 2 or more sextant or 4 (periodontal probing depth >5.5 mm) in any sextant

Patient-based study

Not mentioned

Patient-based study

Kleiner criteria (liver biopsy)

Patient-based study

Spearman test

Odds ratio, relative risk

Periodontitis as a dependent variable

Patient-based study

Liver stiffness (kPa): periodontitis: 15.3, no periodontitis: 8.9

Number of periodontitis patients: 11/38 in nonalcoholic steatohepatitis, 1/31 in nonalcoholic fatty liver

 Odds ratio (95% confidence interval)

Nonalcoholic steatohepatitis to nonalcoholic fatty liver disease: 12.2 (1.48-101.0)

 Relative risk (95% confidence interval)

Nonalcoholic steatohepatitis and diabetes: 1.54 (1.04-2.28)

Nonalcoholic steatohepatitis without diabetes: 1.14 (0.95-1.38)

Yes Komazaki et al, 2017, Japan Cross-sectional N = 52 with nonalcoholic fatty liver disease, mean age of 55 y No Antibody titers against A. actinomycetemcomitans, F. nucleatum, P. gingivalis

Ultrasonography: increase in echoes in the liver

Abdominal computed tomography: liver-spleen ratio, fat area

Spearman test

Correlation coefficient ρ

Anti–A. actinomycetemcomitans immunoglobulin G to total fat area: 0.38

Anti–F. nucleatum immunoglobulin G to total fat area: 0.31

Anti–A. actinomycetemcomitans immunoglobulin G to visceral fat area: 0.37

Yes Infection of A. actinomycetemcomitans affects nonalcoholic fatty liver disease by altering the gut microbiota and glucose metabolism 38 Nakahara et al, 2018, Japan Case control study

nonalcoholic fatty liver disease patients

N = 200 (106 males and 94 females, mean age of 51.5 y)

Non-nonalcoholic fatty liver disease patients

N=? (data has not been provided)

No Serum immunoglobulin G antibody titers against P. gingivalis fimbriae A type 1, 2, and 4

Liver biopsy: criteria of Matteoni, Brunt, and Kleiner

Abdominal computed tomography: visceral fat area

Logistic regression analysis

Liver disease as a dependent variable

Univariate odds ratio (95% confidence interval)

Type 1: 1.81 (0.99-3.32)

Type 2: 1.49 (0.83-2.67)

Type 4: 2.17 (1.12-3.99)

Yes (for type 4) P. gingivalis infection is an important risk factor for pathologic progression in nonalcoholic fatty liver disease 42 Iwasaki et al, 2018, Japan Cross-sectional study N = 1226 (772 males and 454 females, mean age of 50 y) Yes, interexaminer One or more teeth with ≥4 mm periodontal probing depth

Mesio-bucal, mid-buccal, disto-buccal

Mesio-lingual, mid-lingual, disto-lingual per tooth (data of subject teeth has not been provided)

Ultrasonography in the absence of other case of chronic liver disease

Bright liver, increased liver echotexture with kidneys, a vascular blurring, and deep attenuation of the liver

Logistic regression analysis

Liver disease as a dependent variable

Nonalcoholic fatty liver disease prevalence rate (%) significantly increased according to the severity of periodontal disease

 Odds ratio (95% confidence interval)

For all: 1.88 (1.18-2.99)

Males: 1.62 (0.95-2.78)

Females: 2.97 (1.11-7.98)

Yes (for females) There appears to be a positive association between ultrasound-diagnosed nonalcoholic fatty liver disease and having periodontal probing depth ≥4 mm 40 Kuroki et al, 2018, Japan Cross-sectional study N = 110 (66 males and 44 females, mean age of 73.3 y) Yes, interexaminer Not mentioned Mesial and distal sites of alveolar bone loss (percentage of distance between cementoenamel junction to alveolar crest and cementoenamel junction-apex) for all remaining teeth, including third molars on panoramic radiography

Aspartate aminotransferase >30 IU/L

Aspartate aminotransferase >42 IU/L for males

Aspartate aminotransferase >23 IU/L for females

Gamma-glutamyl transpeptidase >32 IU/L for females

Logistic regression analysis

Liver abnormalities as a dependent variable

Adjusted odds ratio

Aspartate aminotransferase: 1.43 (0.46-4.48)

Alanine aminotransferase: 1.24 (0.37-4.18)

Gamma-glutamyl transpeptidase: 0.95 (0.03-1.16)

No There was no significant association between the elevation of serum live enzyme levels and alveolar bone loss in Japanese adults 41 Akinkugbe et al, 2018, USA (Hispanic and Latino) Cross-sectional study N = 11 914 (45.1 males and 54.9% females, mean age of 40.4 y) No Percentage of sites (none, <30%, ≥30%) affected by clinical attachment level ≥3 mm or periodontal probing depth ≥4 mm Not mentioned

Nonalcoholic fatty liver disease

Alanine aminotransferase >40 IU/L for males

Alanine aminotransferase >31 IU/L or aspartate aminotransferase >37 IU/L for females

Fatty liver index score ≥60%

Prevalence odds ratio

Liver disease as a dependent variable

Adjusted prevalence odds ratio

 Clinical attachment level ≥3 mm

 <30%: 1.03 (0.87-1.21)

 ≥30%: 0.91 (0.70-1.18)

 Periodontal probing depth ≥4 mm

 <30%: 1.03 (0.88-1.20)

 ≥30%: 1.00 (0.72-1.38)

No Previously reported associations between periodontitis and nonalcoholic fatty liver disease were not replicated in a diverse group of Hispanic/Latino men and woman 39 Shin, 2019, South Korea Cross-sectional study N = 4061 (1476 males and 2585 females, >19 y) No Presence of periodontal pockets (community periodontal index score 3-4) 10 index teeth: the first and second molars, the upper right incisor, and the lower left incisor

Fatty liver index score >60%

Hepatic steatosis index >36

Chi-square test

Generalized linear model

Liver disease as a dependent variable

Prevalence (%) of nonalcoholic fatty liver disease for women

1) In fatty liver index ≥60 subjects 1-1) No periodontal pockets: 4.6 1-2) Periodontal pockets: 13.2 2) In hepatic steatosis index ≥36 subjects 2-1) No periodontal pockets: 15.9 2-2) Periodontal pockets: 29.5 3) Adjusted odds ratio (95% confidence interval) for fatty liver index for women 3-1) Mild periodontitis: 1.51 (0.78-2.91) 3-2) Severe periodontitis: 2.05 (1.20-3.52) 4) Adjusted odds ratio (95% confidence interval) for HIS for women 4-1) Mild periodontitis: 1.89 (1.13-3.16) 4-2) Severe periodontitis: 1.40 (0.88-2.24) Yes Significant association between the presence of periodontal pockets measured by community periodontal index and nonalcoholic fatty liver disease in the Korean population 49 Weintraub et al, 2019, USA Cross-sectional N = 5421 (47.9% males and 52.1% females, 21-71 y) No

Moderate periodontitis ≥2 teeth with clinical attachment level ≥4 mm or periodontal probing depth ≥5 mm at interproximal

Severe periodontitis ≥2 teeth with clinical attachment level ≥6 mm and ≥1 tooth with periodontal probing depth ≥5 mm at interproximal

Not mentioned

Nonalcoholic fatty liver disease

Ultrasonography: moderate to severe hepatic steatosis

Nonalcoholic fatty liver disease fibrosis score ≥−1.455

Fatty liver index ≥30

US fatty liver index ≥30

Logistic regression analysis

Liver disease as a dependent variable

Odds ratio (95% confidence interval)

 Nonalcoholic fatty liver disease assessed by Ultrasonography :1.54 (1.06-2.24)

 Nonalcoholic fatty liver disease fibrosis score : 3.10 (2.31-4.17)

 Fatty liver index: 1.61 (1.13-2.28)

 US fatty liver index: 2.21 (1.74-2.98)

Yes Nonalcoholic fatty liver disease was significantly associated with tooth loss, moderate to severe periodontitis, and for some nonalcoholic fatty liver disease measures, untreated caries, after adjusting for several key sociodemographic factors 46 Helenius-Hietala et al, 2019, Finland Cohort study N = 6165 (45.3% males and 54.7% females, mean age of 49.5 y) Yes

At least one tooth with a periodontal pocket at least 4 mm deep;

Mild to moderate periodontitis: 1-4 teeth with ≥4 mm deep periodontal pockets

Advanced periodontitis: ≥5 teeth with ≥4 mm deep periodontal pockets

Each tooth excluding wisdom teeth on four surfaces

Nonalcoholic fatty liver disease (for baseline):

Fatty liver index >60 with alcohol use <30 g/d for men or <20 g/d for women

13-y incidence as follows:

 First hospitalization owing to liver disease

 Liver-related death

 Diagnosis of primary liver cancer

Cox model; hazard ratio

Severe liver event as a dependent variable

Adjusted hazard ratio (95% confidence interval)

 Mild periodontitis: 2.24 (0.98-4.84)

 Advanced periodontitis: 3.29 (1.53-7.05)

Adjusted hazard ratio (95% confidence interval) in baseline nonalcoholic fatty liver disease patient

 Mild periodontitis: 3.23 (0.62-16.8)

 Advanced periodontitis: 6.94 (1.43-33.6)

Yes (for advanced periodontitis) Epidemiologic link independent of multiple confounders beween periodontitis and incident severe liver disease were found 45 Kim et al, 2020, South Korea Cross-sectional study N = 4272; 1113 with periodontitis (51.7% males and 48.3% females), mean age of 53.1 y, and 3159 of nonperiodontitis (38.9% males and 61.1% females), mean age of 41.2 y No Community periodontal index score 3 and 4 10 index teeth: the first and second molars, the upper right incisor, and the lower left incisor Fatty liver index divided by quartile Logistic regression analysis

Adjusted odds ratio (95% confidence interval)

 2nd quartile of fatty liver index: 1.29 (0.97-1.71)

 3rd quartile of fatty liver index: 1.43 (1.06-1.93)

 4th quartile of fatty liver index: 1.63 (1.24-2.16)

Yes Fatty liver index may be associated with periodontitis prevalence, especially in subjects with diabetes 48