Skip to main content
Journal of Periodontal & Implant Science logoLink to Journal of Periodontal & Implant Science
. 2026 Apr 14;56(5):389–403. doi: 10.5051/jpis.2504980249

Identification of gingival junctional epithelium–associated genes reflected in periodontal clinical parameters

Misaki Matsuzaki 1, Yoshito Ochiai 1, Takumi Morita 1, Kazuma Igarashi 1, Masatomo Wada 1, Yumi Saito 1,2, Arisa Yamaguchi 1,2, Mizuho Yamazaki-Takai 1,2, Hideki Takai 1,2, Yorimasa Ogata 1,2, Yohei Nakayama 1,2,✉
PMCID: PMC13745167  PMID: 42252877

Abstract

Purpose

The expression of gingival junctional epithelium (JE)–associated genes, including amelotin (AMTN), odontogenic ameloblast-associated protein (ODAM), and follicular dendritic cell-secreted protein (FDCSP), has been reported to increase during gingival inflammation. However, the associations between the expression levels of JE-associated genes and periodontal clinical parameters remain unclear. This study aimed to investigate this association.

Methods

The study included patients with chronic periodontitis who were undergoing supportive periodontal therapy (SPT). Saliva samples were collected and analyzed using quantitative real-time polymerase chain reaction. Statistical analyses were performed to determine the associations between the RNA levels of JE-associated genes and periodontal clinical parameters.

Results

In total, 101 patients were included in this study. After excluding undetectable values for JE-associated gene messenger RNA levels, Spearman rank correlation analysis revealed weak positive correlations between AMTN expression and the bleeding on probing (BOP) rate (n=46, rs=0.4152), periodontal inflamed surface area (rs=0.3912), and the proportion of sites with probing depth of 4–5 mm (rs=0.3270). No significant correlations were observed between ODAM (n=79) or FDCSP (n=84) expression and periodontal clinical parameters. Multiple regression analysis using these clinical parameters as response variables and JE-associated genes as explanatory variables revealed a positive association between AMTN expression and BOP rate. In a second regression model with the reverse structure, AMTN expression was also associated with significant differences in BOP rate. Receiver operating characteristic curve analysis further indicated a significant relationship between AMTN expression and a 9% BOP rate threshold.

Conclusions

Despite the limited sample size resulting from undetectable JE-associated gene expression levels, salivary AMTN expression may have potential as a biomarker for the ancillary evaluation of periodontal inflammatory activity during SPT. Further studies with larger sample sizes are needed to confirm these findings across diverse populations.

Keywords: Biomarkers, Junctional epithelium, Saliva

Graphical Abstract

graphic file with name jpis-56-389-abf001.webp

INTRODUCTION

Chronic periodontitis is an inflammatory disease of the tooth-supporting tissues that results from complex interactions among oral bacteria, host factors, lifestyle habits, and environmental influences. This condition can lead to gingival swelling, bleeding, tooth mobility, alveolar bone resorption, and ultimately tooth loss [1]. The gingival junctional epithelium (JE) is a specialized epithelial tissue that attaches to the enamel surface through hemidesmosomes, thereby forming a barrier between the oral cavity and the periodontal tissues [2,3,4,5]. Consequently, the structural integrity of the JE is critical for maintaining periodontal tissue health. Located at the base of the gingival sulcus, the JE serves as an essential defense against microbial invasion of the tooth-supporting tissues.

The diagnosis of periodontal disease typically involves evaluation of clinical parameters, including probing depth (PD), clinical attachment level, and bleeding on probing (BOP), along with radiographic assessment of alveolar bone loss [6]. However, these diagnostic indicators are invasive and can vary depending on the examiner and clinical conditions.

Saliva is a noninvasive biological fluid that can be collected easily and painlessly. Both saliva and gingival crevicular fluid contain local and systemic markers associated with the pathophysiology of periodontal disease, making them promising sources of biomarkers for periodontitis and systemic diseases [7,8]. For example, salivary inflammatory cytokines such as interleukin-1 beta (IL1β) have been reported to correlate with several periodontal clinical parameters [9]. These findings support the potential utility of saliva as a noninvasive biomarker source reflecting periodontal disease activity.

Amelotin (AMTN) [10], odontogenic ameloblast-associated protein (ODAM) [11], and follicular dendritic cell-secreted protein (FDC-SP) [12] are expressed in the JE. AMTN plays key roles in bone and enamel mineralization [13] and is considered a component of the basal lamina that mediates epithelial cell attachment to the tooth surface [14]. AMTN and ODAM are differentially distributed within the basal lamina and may interact with each other to form a structural component of this matrix [15]. Their expression has been reported to increase in response to inflammation in gingival epithelial cells and in the JE of mice [16,17,18,19]. ODAM levels in gingival crevicular fluid have been found to reflect periodontal clinical parameters [20]. However, the relationships of AMTN and FDCSP gene expression in saliva with periodontal clinical parameters remain unclear, despite their roles as structural components of the JE.

Therefore, this study aimed to investigate the association between the messenger RNA (mRNA) expression levels of JE-associated genes in saliva and periodontal clinical parameters and to evaluate the potential usefulness of these genes as supplementary biomarkers for periodontitis during supportive periodontal therapy (SPT).

MATERIALS AND METHODS

Participants

This study included patients who had undergone periodontal therapy and had been receiving SPT for more than 2 years at Nihon University School of Dentistry, Matsudo Hospital. The study was approved by the Ethics Committee of Nihon University Matsudo Dental School (approval No. EC 20-015). Informed consent was obtained from all patients after they received a thorough explanation of the study objectives and procedures.

Examination items

Periodontal examinations for patients undergoing SPT included measurements of tooth mobility, PD (6-point method), BOP, and plaque control record data. These examinations were performed by dentists specializing in periodontology. Saliva samples were collected during SPT visits. The questionnaire obtained self-reported information on sex, age, systemic medical history, medication status, and smoking history.

Saliva collection

Participants were instructed not to eat or drink for 1 hour before saliva collection. Approximately 2 mL of unstimulated saliva was collected using a dedicated saliva collection device (SALIVA COLLECTION AID and Cryogenic Vials; Salimetrics, Carlsbad, CA, USA). After collection, saliva aliquots were immediately frozen and stored at −80°C. Before the subsequent experiments, the aliquoted saliva tubes were thawed at room temperature (RT).

Calculation of periodontal data

Periodontal examination data, including tooth mobility, PD, BOP, and plaque control record data, were extracted from electronic medical records. From these data, the following variables were calculated: stage, grade, lesion extent, number of remaining teeth, mean PD (mm), proportion of sites with PD of 4–5 mm (%), proportion of sites with PD of ≥6 mm (%), BOP rate (%), periodontal epithelial surface area (PESA, mm2), periodontal inflamed surface area (PISA, mm2), percentage of mobile teeth (%), number of missing molars, and the scored Eichner classification. PISA and PESA were calculated using the Japanese Society of Periodontology JSP-Chart Ver. 4. The Eichner classification was converted from the records into numerical values as described in a previous study [21]. Stage and grade were assigned according to the 2017 World Workshop classification [22]. Lesion extent was defined as generalized when ≥30% of sites were affected (PD ≥4 mm) and localized when <30% of sites were affected.

RNA extraction and real-time polymerase chain reaction (PCR)

Salivary mRNA levels of JE-associated genes (AMTN, FDCSP, ODAM, laminin β3 [Lamβ3], and cytokeratin 13 [CK13]) and inflammatory genes (IL1β and tumor necrosis factor α [TNFα]) were quantified for the subsequent statistical analyses. Total RNA was extracted from the collected saliva using the Total RNA Life Puls Kit (Norgen Biotek Corp., Thorold, ON, Canada). RNA purity was assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). For RNA extraction, 400 μL of collected saliva was used, and DNase I treatment was performed according to the manufacturer’s instructions. Complementary DNA (cDNA) was synthesized using the ExScript RT Reagent Kit (Takara Bio Inc., Shiga, Japan) by reverse transcription at 37°C for 15 minutes. The cDNA was stored at −20°C. Quantitative real-time PCR was performed using primer sets (Table 1) and SYBR Premix Ex Taq on the TP800 Thermal Cycler Dice Real-Time System (Takara Bio Inc.). Cartridge-purified oligonucleotides were purchased from Sigma-Aldrich (St. Louis, MO, USA). The primers used have been shown to enable precise detection of JE-associated genes in gingival epithelial cells [23,24]. The PCR reaction consisted of 45 cycles, with denaturation at 95°C for 5 seconds and annealing/extension at 60°C for 30 seconds. mRNA levels were analyzed relative to glyceraldehyde 3-phosphate dehydrogenase. These data were used to investigate correlations between the mRNA expression levels of JE-associated genes and clinical parameters. Post-PCR melting curve analysis confirmed the specificity of single-target amplification. Relative gene-expression levels were calculated using the ΔΔCt method.

Table 1. Primer list for real-time polymerase chain reaction.

Description Length Sequence (5'–3')
JE-associated genes
AMTN
For. 25 GTTGAATGTACAACAGCAACTGCAC
Rev. 23 TTCCATCCTGGACATCTGGATTA
FDCSP
For. 20 GCCAGTCACTTGCCATTTCT
Rev. 20 GGGCAGATTCAGGTATTGGA
ODAM
For. 21 ATGGTCCTACCCTGGGAACAA
Rev. 21 ACAGCAATTTCAGGAGCTGTG
Lamβ3
For. 20 CCAAGCCTGAGACCTACTGC
Rev. 19 GAATCTCCTGTCCAGGTCC
CK13
For. 21 GACCGCCACCATTGAAAACAA
Rev. 22 TCCAGGTCAGTCTTAGACAGAG
Inflammatory cytokine
IL1β
For. 23 ATGATGGCTTATTACAGTGGCAA
Rev. 20 TGTGATGCGGTTTAGCTGAG
TNFα
For. 22 CCTCTCTCTAATCAGCCCTCTG
Rev. 21 GAGGACCTGGGAGTAGATGAG
Reference gene
GAPDH
For. 20 GCACCGTCAAGGCTGAGAAC
Rev. 20 ATGGTGGTGAAGACGCCAGT

JE: junctional epithelium, AMTN: amelotin, For.: forward, Rev.: reverse, FDCSP: follicular dendritic cell secreted protein, ODAM: odontogenic ameloblast-associated protein, Lamβ3: laminin β3, CK13: cytokeratin 13, IL1β: interleukin 1β, TNFα: tumor necrosis factor-alpha, GAPDH: glyceraldehyde 3-phosphate dehydrogenase.

Measurement of adenosine triphosphate (ATP) activity

To evaluate tissue and cell destruction, ATP activity was measured using a Lucifer HS kit (Kikkoman Biochemifa, Tokyo, Japan). A 100-μL saliva sample was used for this analysis. Following addition of the ATP elimination reagent (ATP-degrading enzyme) and mixing, the sample was transferred to a Lumitube and incubated at RT for 30 minutes. After the reaction, ATP extraction reagent was added to the Lumitube, and 20 seconds later, luminescent reagent was added and mixed. The luminescent emission (in relative light units) was measured using a Lumitester (Kikkoman Biochemifa) within 10 seconds after mixing. These variables were used in the subsequent statistical analyses.

Analysis of the correlation between gingival JE-associated genes and clinical parameters

The following statistical analyses were conducted using BellCurve for Excel (Social Survey Research Information Co., Ltd., Tokyo, Japan). Correlations between clinical parameters and gingival JE-associated genes were analyzed statistically. A total of 15 clinical parameters, 2 smoking-related environmental factors, 5 gingival JE-associated genes, 2 inflammatory cytokines, and ATP activity were included in the analyses. Correlations were assessed using Spearman rank correlation coefficients.

Among the variables displaying statistically significant correlations, clinical parameters associated with gingival JE-associated genes were further evaluated using multiple regression analysis by calculating the standardized partial regression coefficients and coefficients of determination (R2). When high collinearity was identified by the software, one variable was retained and the other was excluded from the multiple regression analysis. The Durbin–Watson statistic was examined to assess whether autocorrelation was present in the error terms, with values of approximately 2.0 considered acceptable. In addition, the mean absolute error was calculated to evaluate the accuracy of each multiple linear regression model. Multiple regression analyses were performed using gingival JE-associated genes as response variables and clinical parameters as explanatory variables, and a second model with the reverse structure was also analyzed to explore directionality. For items that showed significant differences in the multiple regression analyses, expression levels of gingival JE-associated genes were compared between groups divided according to the mean values of the clinical parameters, and the results were presented as box plots. Significant differences were assessed using the Welch t-test based on statistical normality and homogeneity of variance. Differences in the distribution of categorical variables between the high- and low-gene expression groups were evaluated using χ2 tests.

Based on the results of the multiple regression analyses, stratified descriptive statistics were calculated for JE-associated genes and clinical parameters. Groups were stratified according to the expression levels of JE-associated genes and the mean values of the clinical parameters that showed significant differences in the multiple regression analyses. The BOP rate was stratified using ≤9% (low risk in SPT risk assessment) and ≥25% (high risk), as reported in a previous study [24]. All analyses were conducted using samples without zero mRNA values for all genes as measured by quantitative PCR.

Furthermore, for items that showed significant bidirectional differences in the multiple regression analyses, the discriminatory ability of salivary JE-associated gene expression levels for clinical indicators was evaluated using receiver operating characteristic (ROC) curves.

RESULTS

Periodontal examination and patient profiles

A total of 112 patients were enrolled in this study. Periodontal examinations and saliva collection were performed. Of the 112 patients, 11 were excluded due to insufficient saliva volume for real-time PCR or because they had fewer than 10 remaining teeth. Consequently, 101 patients were included in the final analysis. The number of samples with undetectable mRNA levels was 55 for AMTN, 22 for ODAM, 17 for FDCSP, and 18 for Lamβ3, and these values were recorded as zero. Periodontal examination findings showed that the mean PD was 2.5±0.3 mm, the proportion of sites with PD of 4–5 mm was 3.0%±4.5%, the proportion of sites with PD of ≥6 mm was 1.1%±2.0%, the BOP rate was 12.0%±8.9%, the PISA was 156.9±121.9 mm2, and the PESA was 1,079.9±296.2 mm2. The mean number of remaining teeth was 22.5±4.8, the mean number of missing molars was 4.7±3.6, and the mean Eichner classification score was 4.1±2.1. For smoking history, nonsmokers were coded as 0, former smokers as 1, and current smokers as 2. The number of patients in each group was 61, 32, and 8, respectively. The Brinkman index was 123.9±239.2 for the overall cohort (Table 2).

Table 2. Profiling of subjects.

Subjects All Values (n=101)
Sex Male 44
Female 57
Age Male 71.7±9.6
52–88
Female 68.3±10.4
44–87
Examination item Number of teeth 22.5±4.8
10–31
Mean of PD (mm) 2.5±0.3
2–3.6
Rate of PD 4–5 mm (%) 3.0±4.5
0–25.4
Rate of PD ≥6 mm (%) 1.1±2.0
0–11.8
BOP rate (%) 12.0±8.9
0–41.7
PISA (mm2) 156.9±121.9
0–648.4
PESA (mm2) 1,079.9±296.2
439.9–1,742.4
Number of missing molars 4.7±3.6
0–15
Eichner classification 4.1±2.1
1–9
Smoking (0/1/2) 61/32/8
Br Index 123.9±239.2
0–1,125

Data shown as mean ± standard deviation or minimum–maximum.

PD: probing depth, BOP: bleeding on probing, PISA: periodontal inflamed surface area, PESA: periodontal epithelial surface area, Br Index: Brinkman index.

Correlation analysis

Correlations among 15 clinical parameters, 2 smoking-related environmental factors, 5 JE-associated genes, 2 inflammatory cytokines, and ATP activity were investigated using Spearman rank correlation coefficients. In the analysis excluding zero mRNA values, AMTN (n=46) showed a weak positive correlation with the proportion of sites with PD of 4–5 mm. In contrast, ODAM (n=79) and FDCSP (n=84) displayed no significant correlations with the clinical parameters. These findings suggest that AMTN expression is associated with clinical parameters related to periodontal inflammation.

In the analysis including zero mRNA values, weak positive correlations were observed between AMTN and BOP rate, PISA, and stage. FDCSP showed a weak positive correlation with PISA and weak negative correlations with the number of missing molars and the Eichner classification. IL1β displayed weak positive correlations with the proportion of sites with PD of ≥6 mm, BOP rate, PISA, and smoking history. In contrast, ODAM, Lamβ3, CK13, and TNFα were correlated with either 1 or no clinical parameters. In particular, correlations between ODAM and the clinical parameters were minimal and not statistically significant.

Chi-square tests were performed for stage and AMTN, which showed weak positive correlations in the Spearman correlation analysis, and for smoking history and IL1β, because all of these variables were categorical. The results indicated an association between stage classification and AMTN (Figure 1).

Figure 1. Correlations among 15 clinical parameters, 2 environmental factors, 5 junctional epithelium-associated genes, 2 inflammatory cytokines, and ATP activity were analyzed using Spearman rank correlation coefficients and are presented as a heatmap. The total sample size was 101. For the analysis excluding zero messenger RNA values, the sample size for each gene is also shown (AMTN, n=46; ODAM, n=79; and FDCSP, n=84).

Figure 1

ATP: adenosine triphosphate, AMTN: amelotin, ODAM: odontogenic ameloblast-associated protein, FDCSP: follicular dendritic cell-secreted protein, Lamβ3: laminin β3, CK13: cytokeratin 13, TNFα: tumor necrosis factor α, IL1β: interleukin-1 beta, PD: probing depth, PISA: periodontal inflamed surface area, PESA: periodontal epithelial surface area, Br Index: Brinkman index.

a)P<0.01; b)P<0.05.

Multiple regression analysis

Multiple regression analysis was performed using clinical parameters that showed significant correlations in the Spearman correlation analysis as explanatory variables and JE-associated genes as response variables.

In the analysis excluding zero mRNA values, AMTN expression was significantly explained by the BOP rate (R2=0.3268, standardized regression coefficient=0.3268). In the reverse model, the BOP rate was also significantly explained by AMTN (R2=0.3268, standardized regression coefficient=0.3268) (Table 3). No significant associations were observed between the other JE-associated genes and clinical parameters in the analysis excluding zero mRNA values (data not shown). The Durbin–Watson values calculated to assess autocorrelation in each multiple regression analysis were approximately 2.0. The mean absolute error was calculated as the absolute difference between the predicted and measured values.

Table 3. Multiple regression analysis of junctional epithelium-associated genes and clinical parameters in zero values-excluding data.

Dependent variable Explanatory variable Standardized β R R2 DW MAE
AMTN BOP rate (%) 0.3268a) 0.3268 0.1068a) 1.8927 0.3
Constant term -
BOP rate (%) AMTN 0.3268a) 0.3268 0.1068a) 1.6816 7.5
Constant term -

ODAM and IL1β were initially considered based on correlation coefficients (|r|>0.15) but were excluded from the final regression model due to a lack of statistical significance (P>0.05).

R: adjusted multiple correlation coefficient, R2: coefficient of determination, DW: Durbin–Watson, MAE: mean absolute error, AMTN: amelotin, BOP: bleeding on probing, ODAM: odontogenic ameloblast-associated protein, IL1β: interleukin 1β.

a)P<0.05.

In the analysis including zero mRNA values, AMTN was significantly associated with IL1β, BOP rate, and stage (R2=0.2291). PISA was excluded as an explanatory variable because of its high collinearity with the BOP rate. The variables with R2≥0.1 were Lamβ3, which was significantly associated with ATP-activated luciferase and lesion extent (R2=0.1168), and IL1β, which was significantly associated with PISA and smoking history (R2=0.1197). BOP rate was excluded as an explanatory variable because of its high collinearity with PISA. The variable with R2<0.1 was FDCSP, which was significantly associated with PISA (R2=0.0872) (Table 4). The number of missing molars was excluded as an explanatory variable because of its high collinearity with the Eichner classification. In another model with the reverse structure of the multiple regression analysis, using JE-associated genes as the explanatory variables and clinical parameters as the response variables, the clinical parameters with R2≥0.1 were BOP rate for AMTN (R2=0.1082) and PISA for FDCSP and IL1β (R2=0.1082). Given that PISA showed high collinearity with the BOP rate, it was excluded from the multiple regression analysis (Table 5).

Table 4. Multiple regression analysis of junctional epithelium-associated genes and clinical parameters.

Dependent variable Independent variable Standardized β R R2 DW MAE
AMTN IL1β 0.2257b) 0.4786 0.2291a) 2.021 0.2
BOP rate (%) 0.2209b)
Stage 0.2587a)
Constant term -
FDCSP PISA (mm2) 0.2234b) 0.2953 0.0872b) 2.142 7.9
Eichner classification −0.1484
Constant term -
Lamβ3 ATP luciferase activity 0.2419b) 0.3418 0.1168a) 1.576 12.4
Extent of disease 0.2350b)
Constant term -
IL1β PISA (mm2) 0.2324a) 0.3460 0.1197a) 1.844 21.6
Smoking 0.2313b)
Constant term -

Variables such as 'Rate of PD ≥6 mm (%)' were initially considered but excluded from the final models due to lack of statistical significance (P>0.05).

R: adjusted multiple correlation coefficient, R2: coefficient of determination, DW: Durbin–Watson, MAE: mean absolute error, AMTN: amelotin, IL1β: interleukin 1β, BOP: bleeding on probing, FDCSP: follicular dendritic cell secreted protein, PISA: periodontal inflamed surface area, Lamβ3: laminin β3, ATP: adenosine triphosphate.

a)P<0.01; b)P<0.05.

Table 5. Multiple regression analysis of clinical parameters and junctional epithelium-associated genes.

Dependent variable Independent variable Standardized β R R2 DW MAE
BOP rate (%) AMTN 0.3122a) 0.3122 0.0975a) 1.446 6.4
Constant term -
PISA (mm2) FDCSP 0.2211b) 0.3290 0.1082a) 1.840 86.2
IL1β 0.2079b)
Constant term -

IL1β was initially considered for the BOP rate model based on correlation coefficients (|r|>0.15) but was excluded from the final model due to a lack of statistical significance (P>0.05).

R: adjusted multiple correlation coefficient, R2: coefficient of determination, DW: Durbin–Watson, MAE: mean absolute error, BOP: bleeding on probing, AMTN: amelotin, PISA: periodontal inflamed surface area, FDCSP: follicular dendritic cell secreted protein, IL1β: interleukin 1β.

a)P<0.01; b)P<0.05.

Descriptive statistics and intergroup analysis of JE-associated genes and clinical parameters

For the stratified descriptive analysis of JE-associated genes and clinical parameters, items were selected based on the multiple regression results. Comparisons were made by stratifying the data into 2 groups according to JE-associated gene expression and the mean values of the clinical parameters. In addition, tests of homogeneity of variance were performed to assess variability within strata, followed by between-group comparisons.

In the analysis excluding zero mRNA values, AMTN expression was significantly higher in the group with a BOP rate of ≥9% (P=0.0300), and the BOP rate was significantly higher in the group with AMTN mRNA levels of ≥0.34 (P=0.0111) (Figure 2, Table 6).

Figure 2. To assess intergroup variability, tests for equality of variance were performed, followed by between-group comparisons visualized using box plots. The mean AMTN value (0.34) was used as the reference, and standard BOP rate values (9% or 25%) used in SPT assessment were applied. Significant relationships were observed for AMTN with BOP rate and stage, as well as for interleukin-1 beta with periodontal inflamed surface area.

Figure 2

AMTN: amelotin, BOP: bleeding on probing, SPT: supportive periodontal therapy.

a)P<0.05.

Table 6. Stratified descriptive statistics between JE-associated genes and clinical parameters in zero values-excluding data.

JE-associated genes/clinical parameters Stratified descriptive statistics Testing of differences of population mean
Multiple regression analysis explanatory variables Stratified standard of response variables Numbers Mean ± SD of explanatory variables Correlation ratio (η2) Hypothesis testing for the homogeneity of variances Methods P value Statistical power
AMTN BOP rate <0.34 36 12.01±9.53 0.1379a) P=0.8211 t-test P=0.0111a) 0.7372
≥0.34 10 21.11±9.85
BOP rate AMTN <9 20 0.131±0.197 0.1026a) P=1.0000 t-test P=0.0300a) 0.5924
≥9 26 0.492±0.696
AMTN <25 40 0.330±0.589 0.0006 P=0.3848 t-test P=0.8724 0.0529
≥25 6 0.370±0.397

JE: junctional epithelium, SD: standard deviation, AMTN: amelotin, BOP: bleeding on probing.

a)P<0.05.

In the analysis including zero mRNA values, AMTN expression was significantly higher (≥0.15) in the groups with a higher BOP rate and more severe stage (P=0.0085 and P<0.001, respectively). In contrast, IL1β showed only a weak association and did not reach statistical significance. For FDCSP, expression tended to be higher in the high-PISA group, but the difference was not significant (P=0.0679). Lamβ3 levels were higher in the group with higher ATP activity, although the difference was not significant. In addition, IL1β expression was significantly elevated in the high-PISA group (P=0.0056). However, no significant association was observed with smoking history (Table 7). In stratified analyses using clinical parameters as outcome variables and including zero mRNA values, AMTN expression was significantly higher in the group with a BOP rate of ≥9% (P=0.0228) and also differed significantly in the group with a BOP rate of ≥25% (P=0.0315). Furthermore, in the high-PISA group, expression of FDCSP (P=0.0358) and IL1β (P=0.0205) was significantly elevated (Table 8).

Table 7. Stratified descriptive statistics between JE-associated genes and clinical parameters.

JE-associated genes Stratified descriptive statistics Testing of differences of population mean
Multiple regression analysis explanatory variables Stratified standard of response variables Numbers Mean ± SD of explanatory variables Correlation ratio (η2) Hypothesis testing for the homogeneity of variances Methods P value Statistical power
AMTN BOP rate <0.15 79 10.45±7.35 0.1173a) a) Welch’s t-test P=0.0085a) 0.7830
≥0.15 22 17.78±11.43
IL1β <0.15 79 14.04±25.38 0.0724a) a) Welch’s t-test P=0.1031 0.3696
≥0.15 22 41.76±25.38
Stage <0.15 79 1.81±0.92 0.1520a) b) Welch’s t-test P<0.001a) 0.9997
≥0.15 22 2.68±0.57
FDCSP PISA <7.48 75 140.81±103.62 0.0506b) a) Welch’s t-test P=0.0679 0.4495
≥7.48 26 203.21±157.08
Lamβ3 ATP luciferase activity <10.1 82 215,973.8±283,385.8 0.1058a) a) Welch’s t-test P=0.0618 0.4698
≥10.1 19 592,181.4±815,014.3
Extent of disease <10.1 82 0.01±0.11 0.0129 a) Welch’s t-test P=0.4629 0.1100
≥10.1 19 0.05±0.23
IL1β PISA <20.1 79 134.25±102.15 0.1249a) b) Welch’s t-test P=0.0056a) 0.8278
≥20.1 22 238.13±152.27
Smoking <20.1 79 0.42±0.61 0.0291 P=0.3209 t-test P=0.1157 0.3951
≥20.1 22 0.68±0.72

JE: junctional epithelium, SD: standard deviation, AMTN: amelotin, BOP: bleeding on probing, IL1β: interleukin 1β, FDCSP: follicular dendritic cell secreted protein, PISA: periodontal inflamed surface area, Lamβ3: laminin β3, ATP: adenosine triphosphate.

a)P<0.01; b)P<0.05.

Table 8. Stratified descriptive statistics between clinical parameters and junctional epithelium-associated genes.

Clinical parameters Stratified descriptive statistics Testing of differences of population mean
Multiple regression analysis explanatory variables Stratified standard of response variables Numbers Mean ± SD of explanatory variables Correlation ratio (η2) Hypothesis testing for the homogeneity of variances Methods P value Statistical power
BOP rate AMTN <9 46 0.057±0.144 0.0450b) a) Welch’s t-test P=0.0228b) 0.6327
≥9 55 0.233±0.535
<25 93 0.142±0.417 0.0079 P=0.8467 t-test P=0.0315b) 0.5797
≥25 8 0.278±0.377
PISA FDCSP <156.9 59 4.98±7.91 0.0537b) a) Welch’s t-test P=0.03585b) 0.5491
≥156.9 42 11.00±17.16
IL1β <156.9 59 10.57±21.24 0.0702a) a) Welch’s t-test P=0.0205b) 0.6510
≥156.9 42 33.43±59.57

SD: standard deviation, BOP: bleeding on probing, AMTN: amelotin, PISA: periodontal inflamed surface area, FDCSP: follicular dendritic cell secreted protein, IL1β: interleukin 1β.

a)P<0.01; b)P<0.05.

These findings indicate that AMTN expression is strongly associated with BOP rate and stage, whereas IL1β and FDCSP are associated with PISA. In addition, the expression of JE-associated genes appears to reflect clinical inflammatory markers.

ROC curve analysis

ROC curve analysis was performed to evaluate whether expression levels of salivary JE-associated genes could discriminate clinical parameters, focusing on items that showed significant differences even when the response and explanatory variables were reversed in the multiple regression analyses. The discriminatory ability of AMTN for the BOP rate was assessed using ROC curves and the area under the curve (AUC). Zero values for AMTN mRNA expression were excluded from this analysis. Clinical cutoff values of 9% and 25% were used during SPT risk assessment [25]. When 9% was used as the cutoff value, the AUC was 0.6904, the 95% confidence interval (CI) was 0.5351–0.8456, the P value was 0.0162, the cutoff value was 0.2110, the sensitivity was 0.5769, and the false positive rate was 0.2000. When 25% was used as the cutoff value, the AUC was 0.6542, the 95% CI was 0.4179–0.8905, the P value was 0.2010, the cutoff value was 0.2818, the sensitivity was 0.6667, and the false positive rate was 0.2500. A statistically significant result was observed when 9% was used as the reference value (Figure 3).

Figure 3. Receiver operating characteristic curve analysis was performed to evaluate whether salivary AMTN expression levels could distinguish between low-risk (≤9%) and high-risk (>25%) groups in SPT risk assessment using the BOP rate as an indicator. When 9% was used as the cutoff value, the AUC was 0.6904, the 95% CI was 0.5351–0.8456, the P value was 0.0162, the cutoff value was 0.2110, the sensitivity was 0.5769, and the false positive rate was 0.2000. When 25% was used as the cutoff value, the AUC was 0.6542, the 95% CI was 0.4179–0.8905, the P value was 0.2010, the cutoff value was 0.2818, the sensitivity was 0.6667, and the false positive rate was 0.2500.

Figure 3

AMTN: amelotin, SPT: supportive periodontal therapy, BOP: bleeding on probing, AUC: area under the curve, CI: confidence interval, TPF: true positive fraction, FPF: false positive fraction.

DISCUSSION

Recently, salivary biomarkers, including inflammatory cytokines and enzymes that reflect periodontal tissue destruction and inflammatory activity, have attracted attention as point-of-care tests [25,26]. For example, aspartate aminotransferase, hemoglobin, and lactate dehydrogenase have been recognized as candidate biomarkers. In practice, these biomarkers have been used as adjunctive measures for assessing periodontal conditions. However, no studies have attempted to evaluate periodontal disease by assessing structural components of the gingival JE in saliva. Therefore, this study analyzed gingival JE-associated genes.

The JE is a critical epithelial structure that attaches to the tooth surface [2,3] and serves a central role in defense against bacteria and external stimuli within the gingival crevice. Accordingly, maintenance of its structure and function is essential for periodontal tissue health. JE destruction is associated with gingivitis, periodontitis, and alveolar bone loss [27]. Thus, JE status is an important indicator for evaluating the onset and progression of periodontitis.

Inflammatory cytokines play a key role in initiating and maintaining inflammatory immune responses by promoting the production of secondary mediators that contribute to connective tissue destruction and osteoclast-mediated bone resorption [28]. Among these cytokines, IL1β is involved in the pathogenesis of periodontal disease [29,30] and has been reported to correlate with the BOP rate and periodontal pocket depth [31]. Although inflammatory cytokines induce JE-associated gene expression in human gingival fibroblasts and epithelial cells [16,17,18,19,32,33], no study has yet reported JE-associated gene expression in saliva. Therefore, we investigated the association between salivary JE-associated genes and clinical parameters.

In this study, AMTN expression showed weak positive correlations with stage, IL1β, BOP rate, and PISA. Additionally, multiple regression analysis demonstrated a significant association, with an R2 value of 0.2291, which exceeded those observed for the other clinical parameters discussed below. Group comparisons further showed that AMTN expression was significantly higher in the groups with a higher BOP rate and more advanced stage. These findings suggest that salivary AMTN expression is correlated with periodontal inflammation. AMTN localizes to the inner basal lamina of the gingival epithelium, whereas ODAM and FDC-SP are distributed throughout the JE [34]. Therefore, quantitative detection of AMTN expression in saliva may be more difficult than detection of ODAM and FDC-SP during inflammation. At the same time, AMTN detection may be more sensitive for mild inflammation because the inner basal lamina is disrupted during the initial inflammatory response. Collectively, these findings suggest that AMTN may serve as a biomarker of gingival inflammatory activity, as reflected by BOP rate and PISA, in patients receiving SPT.

The correlation analysis revealed a weak positive correlation between FDCSP and PISA, along with weak negative correlations between FDCSP and the number of missing molars (rs =−0.1979) and the Eichner classification (rs =−0.1999) (Table 3). FDC-SP is a secreted protein expressed by follicular dendritic cells, the periodontal ligament, and the JE [13]. JE atrophy has been reported to occur when mechanical stimulation is reduced [35]. This atrophy leads to a decline in barrier function, thereby facilitating bacterial invasion and subsequent inflammation [36]. The loss of occlusal support has also been reported to reduce mechanical stimulation to the JE, potentially suppressing FDCSP expression.

IL1β showed weak positive correlations with the proportion of sites with PD of ≥6 mm, the BOP rate, PISA, and smoking history. Multiple regression analysis identified PISA and smoking history as significant explanatory variables. Group comparisons revealed a significant difference for PISA, but not for smoking history. Because PISA represents the area of inflammation caused by periodontal disease, these findings suggest that IL1β may serve as a biomarker of periodontitis progression. Salivary IL1β expression appears to be useful for diagnosing periodontitis. However, previous studies have shown lower diagnostic accuracy in smokers [36]. Smoking exerts a clear confounding effect on salivary biomarkers. Therefore, smoking should be considered when salivary biomarkers are used for periodontal diagnosis [37], and further studies in nonsmokers may be warranted.

Previous research has indicated that ODAM protein levels in gingival crevicular fluid reflect periodontal parameters [20]. However, the present study found no meaningful correlation between ODAM expression and periodontal parameters (Figure 1). This discrepancy may be attributable to differences in sample type (saliva vs crevicular fluid) and inflammatory status (SPT vs initial periodontal therapy). Saliva may reflect full-mouth conditions, such as BOP rate, PISA, and PESA. However, for the evaluation of inflammatory activity, salivary detection is inferior to crevicular fluid analysis, particularly in patients receiving SPT. In the previous study, gingival crevicular fluid was collected from 8 patients with periodontitis, among whom at least 18% (and up to 68.5%) of sites had a PD of ≥4 mm, and 60% (and up to 88.7%) of sites were BOP-positive [20]. In contrast, in the present study, the mean proportion of BOP-positive sites was 12.0% ± 8.9% (range, 0%–41.7%) (Table 2).

In the multiple regression analysis, a moderate interrelationship was observed between the BOP rate and AMTN expression. Therefore, AMTN was associated with the BOP rate and may be a strong candidate molecular marker of local inflammatory activity. ROC curve analysis was performed to evaluate the discriminatory ability of salivary JE-associated gene expression levels for clinical parameters. AMTN showed moderate discriminatory ability when a BOP cutoff value of 9%, but not 25%, was used (Figure 3). A BOP cutoff of 25% has been used as a critical value for identifying moderate or high risk in SPT assessment [24]. These findings raise the possibility that AMTN detection may be useful for identifying high inflammatory risk during SPT assessment.

In addition, a weak interrelationship was observed between FDCSP expression and PISA in the multiple regression analysis. FDCSP expression in the JE has been reported to fluctuate in response to changes in mechanical stimulation, thereby contributing to the maintenance of barrier function [35]. In contrast, PISA is a clinical indicator that quantifies the inflamed area in periodontal tissues and reflects the extent of inflammatory activity. Research has indicated that PISA is useful for quantifying the inflammatory burden [38,39]. Based on previous studies, PISA cutoff values of 100 mm2 and 130 mm2 were used in the ROC analysis. No significant difference was observed for FDCSP at either PISA cutoff value. In contrast, IL1β showed significant differences at both cutoff values. These findings indicate that direct evaluation of inflammation through IL1β detection may reflect PISA more closely than detection of JE components, although the influence of smoking history should be considered when applying salivary IL1β as a biomarker for periodontitis.

This study focused on patients in the SPT phase and included many patients with relatively mild inflammation. Therefore, highly sensitive detection of JE-associated gene expression was challenging. Further studies are needed to clarify how AMTN and FDCSP expression levels change in states of more severe inflammation during initial and active periodontal therapy. In addition, the presence of undetectable mRNA values for JE-associated genes, which contributed to the small sample size in some analyses, reduced the reliability of the statistical findings, although analyses excluding zero mRNA values were also performed. Future studies should include protein-level measurements to clarify the clinical significance of these genes and to determine whether AMTN protein expression reflects periodontal inflammation as a supportive biomarker before and after initial periodontal therapy.

This study statistically evaluated the associations between salivary JE-associated gene expression levels and periodontal clinical parameters. Among the JE-associated genes examined, AMTN showed the strongest association with the BOP rate. These findings suggest that JE-associated genes may be useful as supplementary indicators of periodontitis. However, due to the moderate discriminatory performance observed, the results should be validated in larger cohort studies incorporating protein-level detection.

Footnotes

Funding: This study was supported in part by the Japan Society for the Promotion of Science (Chiyoda, Tokyo, Japan) KAKENHI Grants-in-Aid for Scientific Research (C) No. 24K12935 to Y.N.

Conflict of Interest: No potential conflict of interest relevant to this article was reported.

Author Contributions:
  • Conceptualization: Yohei Nakayama.
  • Data curation: Misaki Matsuzaki.
  • Formal analysis: Misaki Matsuzaki, Yohei Nakayama.
  • Investigation: Misaki Matsuzaki, Yoshito Ochiai, Takumi Morita, Kazuma Igarashi, Masatomo Wada, Yumi Saito, Arisa Yamaguchi, Mizuho Yamazaki-Takai, Hideki Takai, Yorimasa Ogata.
  • Methodology: Yohei Nakayama.
  • Project administration: Yohei Nakayama.
  • Resources: Yohei Nakayama.
  • Supervision: Yorimasa Ogata, Yohei Nakayama.
  • Writing - original draft: Misaki Matsuzaki.
  • Writing - review & editing: Yohei Nakayama.

References

  • 1.Keyes PH, Likins RC. Plaque formation, periodontal disease, and dental caries in Syrian hamsters. J Dent Res. 1946;25:166. [PubMed] [Google Scholar]
  • 2.Hou M, Liu S, Yan K, Sun Z, Li S. Downregulation of odontogenic ameloblast-associated protein in the progression of periodontal disease affects cell adhesion, proliferation, and migration. Arch Oral Biol. 2023;145:105588. doi: 10.1016/j.archoralbio.2022.105588. [DOI] [PubMed] [Google Scholar]
  • 3.Groeger S, Meyle J. Oral mucosal epithelial cells. Front Immunol. 2019;10:208. doi: 10.3389/fimmu.2019.00208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hormia M, Owaribe K, Virtanen I. The dento-epithelial junction: cell adhesion by type I hemidesmosomes in the absence of a true basal lamina. J Periodontol. 2001;72:788–797. doi: 10.1902/jop.2001.72.6.788. [DOI] [PubMed] [Google Scholar]
  • 5.Schroeder HE, Listgarten MA. The junctional epithelium: from strength to defense. J Dent Res. 2003;82:158–161. doi: 10.1177/154405910308200302. [DOI] [PubMed] [Google Scholar]
  • 6.Armitage GC. The complete periodontal examination. Periodontol 2000. 2004;34:22–33. doi: 10.1046/j.0906-6713.2002.003422.x. [DOI] [PubMed] [Google Scholar]
  • 7.Kaufman E, Lamster IB. Analysis of saliva for periodontal diagnosis--a review. J Clin Periodontol. 2000;27:453–465. doi: 10.1034/j.1600-051x.2000.027007453.x. [DOI] [PubMed] [Google Scholar]
  • 8.Ozmeric N. Advances in periodontal disease markers. Clin Chim Acta. 2004;343:1–16. doi: 10.1016/j.cccn.2004.01.022. [DOI] [PubMed] [Google Scholar]
  • 9.Relvas M, Mendes-Frias A, Gonçalves M, Salazar F, López-Jarana P, Silvestre R, et al. Salivary IL-1β, IL-6, and IL-10 are key biomarkers of periodontitis severity. Int J Mol Sci. 2024;25:8401. doi: 10.3390/ijms25158401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Iwasaki K, Bajenova E, Somogyi-Ganss E, Miller M, Nguyen V, Nourkeyhani H, et al. Amelotin-a novel secreted, ameloblast–specific protein. J Dent Res. 2005;84:1127–1132. doi: 10.1177/154405910508401207. [DOI] [PubMed] [Google Scholar]
  • 11.Moffatt P, Smith CE, St-Arnaud R, Nanci A. Characterization of Apin, a secreted protein highly expressed in tooth-associated epithelia. J Cell Biochem. 2008;103:941–956. doi: 10.1002/jcb.21465. [DOI] [PubMed] [Google Scholar]
  • 12.Iwai Y, Noda K, Yamazaki M, Kato A, Mezawa M, Takai H, et al. Tumor necrosis factor-α regulates human follicular dendritic cell-secreted protein gene transcription in gingival epithelial cells. Genes Cells. 2018;23:161–171. doi: 10.1111/gtc.12561. [DOI] [PubMed] [Google Scholar]
  • 13.Nakayama Y, Holcroft J, Ganss B. Enamel hypomineralization and structural defects in amelotin-deficient mice. J Dent Res. 2015;94:697–705. doi: 10.1177/0022034514566214. [DOI] [PubMed] [Google Scholar]
  • 14.Kawasaki K, Weiss KM. Mineralized tissue and vertebrate evolution: the secretory calcium-binding phosphoprotein gene cluster. Proc Natl Acad Sci U S A. 2003;100:4060–4065. doi: 10.1073/pnas.0638023100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Holcroft J, Ganss B. Identification of amelotin- and ODAM-interacting enamel matrix proteins using the yeast two-hybrid system. Eur J Oral Sci. 2011;119(Suppl 1):301–306. doi: 10.1111/j.1600-0722.2011.00870.x. [DOI] [PubMed] [Google Scholar]
  • 16.Nakayama Y, Kobayashi R, Matsui S, Matsumura H, Iwai Y, Noda K, et al. Localization and expression pattern of amelotin, odontogenic ameloblast-associated protein and follicular dendritic cell-secreted protein in the junctional epithelium of inflamed gingiva. Odontology. 2017;105:329–337. doi: 10.1007/s10266-016-0277-y. [DOI] [PubMed] [Google Scholar]
  • 17.Nakayama Y, Takai H, Matsui S, Matsumura H, Zhou L, Kato A, et al. Proinflammatory cytokines induce amelotin transcription in human gingival fibroblasts. J Oral Sci. 2014;56:261–268. doi: 10.2334/josnusd.56.261. [DOI] [PubMed] [Google Scholar]
  • 18.Jin Z, Yamaguchi A, Takai H, Nakayama Y, Ogata Y. Interleukin-6 regulates human ODAM gene expression in gingival epithelial cells. J Periodontal Implant Sci. 2025;55:349–364. doi: 10.5051/jpis.2402980149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Nakayama Y, Inoue E, Kato A, Iwai Y, Takai-Yamazaki M, Tsuruya Y, et al. Follicular dendritic cell-secreted protein gene expression is upregulated and spread in nifedipine-induced gingival overgrowth. Odontology. 2020;108:532–544. doi: 10.1007/s10266-020-00483-2. [DOI] [PubMed] [Google Scholar]
  • 20.Lee HK, Kim SJ, Kim YH, Ko Y, Ji S, Park JC. Odontogenic ameloblast-associated protein (ODAM) in gingival crevicular fluid for site-specific diagnostic value of periodontitis: a pilot study. BMC Oral Health. 2018;18:148. doi: 10.1186/s12903-018-0609-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Tabe S, Nakayama Y, Kobayashi R, Oyama K, Kitano D, Ogihara J, et al. Association between dietary habit and clinical parameters in patients with chronic periodontitis undergoing supportive periodontal therapy. Nutrients. 2022;14:4993. doi: 10.3390/nu14234993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Tonetti MS, Greenwell H, Kornman KS. Staging and grading of periodontitis: Framework and proposal of a new classification and case definition. J Periodontol. 2018;89(Suppl 1):S159–S172. doi: 10.1002/JPER.18-0006. [DOI] [PubMed] [Google Scholar]
  • 23.Nakayama Y, Igarashi K, Jin Z, Yamaguchi A, Ganss B, Ogata Y. Conditioned medium from cultured cementoblasts upregulates amelotin gene expression via the SOCS3 signaling pathway. J Periodontal Implant Sci. 2025;55:255–272. doi: 10.5051/jpis.2403080154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Lang NP, Tonetti MS. Periodontal risk assessment (PRA) for patients in supportive periodontal therapy (SPT) Oral Health Prev Dent. 2003;1:7–16. [PubMed] [Google Scholar]
  • 25.Uzar E, Gurbuz N, Akpınar Z, Ergin M, Koşkan Ö, Ay ZY. The potential of serum and salivary interleukin-1β and matrix metalloproteinase-8 as determinants of periodontitis grading. BMC Oral Health. 2025;25:1139. doi: 10.1186/s12903-025-06529-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Relvas M, Silvestre R, Gonçalves M, Cabral C, Mendes-Frias A, Monteiro L, et al. Analysis of salivary levels of IL-1β, IL-17A, OPG and RANK-L in periodontitis using the 2017 classification of periodontal diseases—an exploratory observational study. J Clin Med. 2023;12:1003. doi: 10.3390/jcm12031003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Zhu S, Xiang C, Charlesworth O, Bennett S, Zhang S, Zhou M, et al. The versatile roles of odontogenic ameloblast-associated protein in odontogenesis, junctional epithelium regeneration and periodontal disease. Front Physiol. 2022;13:1003931. doi: 10.3389/fphys.2022.1003931. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Zhou M, Li S, Pathak JL. Pro-inflammatory cytokines and osteocytes. Curr Osteoporos Rep. 2019;17:97–104. doi: 10.1007/s11914-019-00507-z. [DOI] [PubMed] [Google Scholar]
  • 29.Rawlinson A, Dalati MH, Rahman S, Walsh TF, Fairclough AL. Interleukin-1 and IL-1 receptor antagonist in gingival crevicular fluid. J Clin Periodontol. 2000;27:738–743. doi: 10.1034/j.1600-051x.2000.027010738.x. [DOI] [PubMed] [Google Scholar]
  • 30.Faizuddin M, Bharathi SH, Rohini NV. Estimation of interleukin-1β levels in the gingival crevicular fluid in health and in inflammatory periodontal disease. J Periodontal Res. 2003;38:111–114. doi: 10.1034/j.1600-0765.2003.01649.x. [DOI] [PubMed] [Google Scholar]
  • 31.Cheng R, Wu Z, Li M, Shao M, Hu T. Interleukin-1β is a potential therapeutic target for periodontitis: a narrative review. Int J Oral Sci. 2020;12:2. doi: 10.1038/s41368-019-0068-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Nakayama Y, Takai H, Matsui S, Zhou L, Abiko Y, Ganss B, et al. Transcriptional regulation of amelotin gene by proinflammatory cytokines in gingival fibroblasts. Connect Tissue Res. 2014;55(Suppl 1):18–20. doi: 10.3109/03008207.2014.923848. [DOI] [PubMed] [Google Scholar]
  • 33.Tsuruya Y, Yamaguchi A, Yamazaki-Takai M, Zhenyu J, Takai H, Nakayama Y, et al. Interleukin-1β regulates odontogenic ameloblast-associated protein gene transcription in human gingival epithelial cells. Odontology. 2022;110:557–568. doi: 10.1007/s10266-022-00689-6. [DOI] [PubMed] [Google Scholar]
  • 34.Nishio C, Wazen R, Kuroda S, Moffatt P, Nanci A. Expression pattern of odontogenic ameloblast-associated and amelotin during formation and regeneration of the junctional epithelium. Eur Cell Mater. 2010;20:393–402. doi: 10.22203/ecm.v020a32. [DOI] [PubMed] [Google Scholar]
  • 35.Yuan X, Liu B, Cuevas P, Brunski J, Aellos F, Petersen J, et al. Linking the mechanics of chewing to biology of the junctional epithelium. J Dent Res. 2023;102:1252–1260. doi: 10.1177/00220345231185288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Arias-Bujanda N, Regueira-Iglesias A, Blanco-Pintos T, Alonso-Sampedro M, Relvas M, González-Peteiro MM, et al. Diagnostic accuracy of IL1β in saliva: the development of predictive models for estimating the probability of the occurrence of periodontitis in non-smokers and smokers. J Clin Periodontol. 2020;47:702–714. doi: 10.1111/jcpe.13285. [DOI] [PubMed] [Google Scholar]
  • 37.Lahdentausta L, Paju S, Mäntylä P, Buhlin K, Pietiäinen M, Tervahartiala T, et al. Smoking confounds the periodontal diagnostics using saliva biomarkers. J Periodontol. 2019;90:475–483. doi: 10.1002/JPER.18-0545. [DOI] [PubMed] [Google Scholar]
  • 38.Inoue Y, Hatanaka K, Yamamoto T, Hirata T, Minabe M, Yamamoto T, et al. Reference values of periodontal inflamed surface area as a clinical index determined by a multicenter retrospective observational study. J Jpn Soc Periodontol. 2019;61:159–167. [Google Scholar]
  • 39.Nesse W, Abbas F, van der Ploeg I, Spijkervet FK, Dijkstra PU, Vissink A. Periodontal inflamed surface area: quantifying inflammatory burden. J Clin Periodontol. 2008;35:668–673. doi: 10.1111/j.1600-051X.2008.01249.x. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Periodontal & Implant Science are provided here courtesy of Korean Academy of Periodontology

RESOURCES