1008 D3000 new imprint Word template Vol 13, No 1 (2025) ISSN 2167-8677 (online) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu Potential of Salivary IL-36g, IL-38, and RANKL in Differentiating Perio- dontal Health from Periodontitis Iqbal Hussain Ali, Suzan Ali Salman College of Den*stry, University of Baghdad, Iraq Abstract Objec5ve: To evaluate the diagnosFc potenFal of combined salivary biomarkers RANKL, IL- 36γ, and IL-38 in disFnguishing individuals with generalized unstable periodonFFs from those with a healthy periodonFum, by analyzing their concentraFons and correlaFon with clinical periodontal parameters. Material and Methods: A total of 120 parFcipants were included in this case-control study, comprising 90 subjects with generalized unstable periodonFFs and 30 periodontally healthy controls. Clinical periodontal measurements, PI, PPD, CAL, and BOP, were recorded. Before data collecFon, inter- and intra-examiner calibraFon were performed on five subjects each, using kappa and intraclass correlaFon coefficients to ensure measure- ment reliability (>70% for BOP and >90% for PPD and CAL). UnsFmulated saliva samples were collected, preserved in anFprotease soluFon, and stored on ice, then frozen for laboratory analysis. Salivary concentraFons of RANKL, IL-36γ, and IL-38 were measured using ELISA kits. Data analysis included comparison between groups, correlaFon with clinical parameters, and diagnosFc evaluaFon through sensiFvity, specificity, and AUC analysis. Data were analyzed using SPSS v29 and GraphPad Prism v9. Normality was assessed using the Shapiro-Wilk test. Parametric tests (ANOVA, t-test) and non-parametric tests (Kruskal-Wallis, Mann-Whitney) were applied as appropriate. CorrelaFons were evaluated using Spearman’s test. DiagnosFc accuracy was assessed through ROC curve analysis and AUC values, with staFsFcal signifi- cance set at p < 0.05. Results: RANKL and IL-36γ salivary levels in periodonFFs were signifi- cantly increased compared to healthy controls, while IL-38 levels were significantly reduced (p < 0.001). RANKL peaked in stage III, while IL-36γ was highest in stage IV. Conversely, IL-38 was consistently lower in both disease stages. Only IL-36γ with CAL correlaFon was a signifi- cant posiFve, while other clinical correlaFons were not staFsFcally significant different. ROC analysis demonstrated excellent diagnosFc accuracy for all three biomarkers in differenFat- ing periodontal health from disease, with AUC values of 0.944 (RANKL), 0.982 (IL-36γ), and 0.960 (IL-38), along with high sensiFvity and specificity. Conclusion: Salivary lev- els of RANKL, IL-36γ, and IL-38 demon- strated strong potenFal as non-invasive biomarkers for disFnguishing general- ized unstable periodonFFs from perio- dontal health. Their diagnosFc accuracy supports their uFlity in early detecFon and monitoring, although they showed limited ability to differenFate between periodonFFs severity stages. Open Access Cita%on: Ali IH, et al. (2025) Poten%al of Salivary IL-36g, IL-38, and RANKL in Differen%a%ng Periodontal Health from Periodon%%s. Den%stry 3000. 1:a001 doi:10.5195/d3000.2025.1008 Received: July 31, 2025 Accepted: August 9, 2025 Published: September 9, 2025 Copyright: ©2025 Ali IH, et al. This is an open access ar- %cle licensed under a Crea%ve Commons AUribu%on Work 4.0 United States License. Email: Iqbal.Ali2205@codental.uobaghdad.edu.iq Introduc)on Periodontitis is one of the most common worldwide oral diseases [1]. Between 2011 and 2020, periodontitis in dentate adults was estimated to be around 62%, and severe periodontitis 23.6%. These results show an unusually high prevalence of periodontitis compared to the previous estimates from 1990 to 2010 [2]. Periodontal disease is a chronic, multifacto- rial inFlammatory condition that affects the supporting structures of the teeth. It is pri- marily triggered by microbial plaque bioFilm and modiFied by host immune-inFlammatory responses, environmental, and genetic factors [3]. If left untreated, it may lead to progressive tissue destruction, alveolar bone loss, and ultimately, tooth loss [4]. Generalized unstable periodontitis represents an advanced stage of periodontal disease characterized by rapid progression and tissue breakdown. Early detection and accurate diagnosis are critical to managing PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 2 this condition effectively and preventing ir- reversible damage [5]. Conventional perio- dontal diagnostics are mainly based on clini- cal parameters such as probing pocket depth (PPD), clinical attachment level (CAL), bleed- ing on probing (BOP), and plaque index (PI). While these indicators provide useful infor- mation, they often reFlect past tissue damage and do not accurately capture current dis- ease activity or risk of progression 6. Recent advances in molecular diagnos- tics and salivary proteomics have opened new possibilities for non-invasive and real- time periodontal disease assessment. Saliva is increasingly recognized as a diagnostic Fluid due to its accessibility, non-invasive collection method, and the presence of bi- omarkers that reFlect both systemic and oral health conditions [7,8]. Among these biomarkers, Receptor Acti- vator of Nuclear Factor Kappa-Β Ligand (RANKL) has been identiFied as a key mole- cule in bone metabolism and periodontal de- struction [9]. Interleukin-36γ (IL-36γ) is a pro-inFlammatory cytokine involved in mu- cosal immunity and inFlammation [10], while Interleukin-38 (IL-38) plays an anti-inFlam- matory role by regulating immune responses and potentially protecting tissues from ex- cessive damage [11]. Therefore, the simultaneous evaluation of these salivary biomarkers may offer a more accurate and early diagnostic approach for periodontal disease compared to tradi- tional methods. We aim to analyze the diag- nostic potential of salivary RANKL, IL-36γ, and IL-38 in differentiating healthy perio- dontium from generalized unstable perio- dontitis. It further seeks to assess the corre- lation with the clinical periodontal parame- ters, as well as determine their diagnostic ac- curacy. Material and Methods Study settings The multicenter study, a case-control obser- vational study, was conducted at the Perio- dontics Departments of the College of Den- tistry, University of Kirkuk, the Kirkuk Spe- cialized Dental Center, and AL-Haweeja Health Center, Kirkuk, Iraq, from November 2023 to October 2024. A total of 120 participants were recruited for the study, comprising 61 females (50.83%) and 59 males (49.17%), with a female-to- male ratio of approximately 1.03:1. The par- ticipants were categorized into two groups: healthy periodontium (n=30) and unstable generalized periodontitis (n=90) based on the latest periodontal disease and condition classification. Sample Size Calculation Based on previously reported mean and standard deviation (SD) values of salivary RANKL concentrations, one of the primary biomarkers assessed in this study, the sam- ple size was calculated. Accordingly, the RANKL concentration in healthy individuals was estimated at 73.6 pg/mL, compared to 128.9 pg/mL in individuals with periodonti- tis. These values were used to assess the re- quired sample size using ChatGPT-based sta- tistical tools. The minimum calculated sam- ple size for the periodontitis group was 80 subjects. To account for potential attrition, this number was increased to 90. Similarly, the control group was initially estimated to require 20 participants and was subse- quently increased to 30 to minimize the risk of attrition bias. Study Population Consistent with the 2017 classiFication and the eligibility criteria of this study, subjects were grouped as: 1. Healthy periodontium group: the diagnostic criteria for periodontal health on an intact periodontium included: BOP <10%, PPD ≤3 mm, and no evidence of alveolar bone loss clinically. The unstable generalized periodontitis group was deFined by one or more of the fol- lowing criteria: sites with PPD ≥ 5 mm or 4 mm with positive BOP, BOP >10%, and in- volving ≥30% of teeth. Interdental Clinical Attachment Loss (CAL) detectable at ≥2 non- adjacent teeth, or buccal or oral CAL ≥3 mm with PPD >3 mm, present at ≥2 teeth. Eligibility Criteria Participants were systemically healthy indi- viduals (aged ≥18 years) diagnosed with ei- ther a healthy periodontium or unstable gen- eralized periodontitis. All were, had at least 20 natural teeth (excluding third molars), and had not taken any medications in the three months before enrolment. Exclusion criteria included systemic dis- eases (e.g. rheumatoid arthritis, cardiovas- cular disease, diabetes), smoking or alcohol use, pregnancy or lactation, presence of oral risk factors (e.g. carious or perio-endo le- sions), use of orthodontic appliances, prior periodontal treatment within six months, or medication use (e.g. antibiotics, corticoster- oids, biologics, or anti-inFlammatories) within the last three months. Ethical Approval The research methodology employed in this study adhered to the Declaration of Hel- sinki and its subsequent amendments con- cerning human subject research, as well as the STARD 2015 (Standards for Reporting of Diagnostic Accuracy Studies). The study was approved by the Ethics Committee of the University of Baghdad's College of Dentistry (reference number: 892, Project number: 892624, Date: 11-1-2024). Reliability Analysis To ensure measurement reliability, calibra- tion sessions were conducted before data collection. Inter- and intra-examiner calibra- tion were performed on Five subjects each, with a minimum of two hours between re- peated assessments. PPD, BOP, and CAL were recorded at buccal/labial sites. Agree- ment for BOP was assessed using the kappa statistic (>70% inter-examiner, >75% intra- examiner), while PPD and CAL reliability were conFirmed by intraclass correlation co- efFicients exceeding 90%. Salivary Collection Procedure Unstimulated whole saliva was collected be- fore clinical examination using the passive drooling method [12]. Participants refrained from food, drink, and oral hygiene for at least two hours and rinsed their mouths with wa- ter before sampling [13]. Saliva was pas- sively drooled into sterile tubes over 2–4 minutes [14]. Samples were coded, treated with a protease inhibitor, and stored on ice. They were cen- trifuged at 1000 rpm for 10 minutes, and 500 μL of the supernatant was transferred to la- beled Eppendorf tubes and stored at –20°C until ELISA analysis [15]. Periodontal Examination Periodontal status was determined accord- ing to the previously established criteria, af- ter saliva collection. A calibrated examiner performed full-mouth periodontal charting (PI, BOP, PPD, and CAL) using a UNC-15 per- iodontal probe (Medesy, Italy). The charting had begun at the distal surface of the upper right second/last molar and proceeded mesially, then lingually to cover all tooth surfaces. BOP, PPD, and CAL were assessed simultaneously. The probing force was standardized (20–25 g) as per cal- ibration sessions (16). Consequently, perio- dontitis staging and grading followed the 2017 classiFication guidelines, based on the site with the highest CAL Laboratory Analysis of Salivary Bi- omarkers After data collection was completed, frozen saliva samples were transported to the la- boratory, thawed at room temperature, and 100µm of the supernatant was collected for analyzing protein biomarkers. Salivary lev- els of IL-36γ, IL-38, and RANKL were meas- ured using the commercially available ELISA kits. PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 3 IL-36γ: Cloud-Clone Corp. ELISA kit, product no. SEL621Hu (Cloud-Clone Corp., Wuhan, China). -Sensitivity: <6.5pg/mL and detection range 15.6-1000 pg/mL. IL-38: Cloud-Clone Corp., Wuhan, China ELISA Kit, Product No. SEQ458Hu (Cloud- Clone Corp., Wuhan, China). -Sensitivity: 46.88pg/mL and the detection Range: 78.13-5000pg/mL. RANKL: Human RANKL ELISA Kit from Elab- science. Their Product No is E-EL-H5813 (headquartered in Houston, USA). -Sensitivity: 9.38 pg/mL, and the detection range is 15.63-1000 pg/mL. Each 96-well plate was pre-coated with a monoclonal antibody speciFic to the target analyte, employing the sandwich ELISA tech- nique. All procedures followed the manufac- turer’s protocol. Statistical Analysis For continuous variables, central tendency and dispersion were quantiFied using mean values and standard deviation (SD). The Shapiro-Wilks test was applied to evaluate the normality of data distribution. Inter- group comparisons were performed using an independent sample t-test, Mann-Whit- ney U test, and chi-square test. Within each group, Spearman correlation analysis was conducted to examine relationships between variables. Positive and negative predictive values were determined through contin- gency table analysis. Statistical signiFicance was deFined as p<0.05. The sensitivity, speciFicity, and optimal cut- off points for individual biomarkers, as well as their ratios, were evaluated using receiver operating characteristic (ROC) curve analy- sis and the area under the curve (AUC). For this purpose, the concentrations of each bi- omarker were dichotomized, assigning “0” to the healthy group and “1” to the periodonti- tis group. All statistical analyses were per- formed using GraphPad Prism software (ver- sion 9.0). A p-value of less than 0.05 was con- sidered statistically signiFicant. Results Out of the initially assessed 600 individuals, 120 participants met the inclusion criteria and were divided into two groups: 90 with generalized unstable periodontitis and 30 periodontally healthy controls. The demo- graphic analysis revealed signiFicant differ- ences in age and gender between the groups (p < 0.001), with the periodontitis group having a higher mean age and a greater pro- portion of males (Table 1). Clinical Parameters All clinical periodontal parameters (PI, BOP, PPD, and CAL) were signiFicantly higher in the periodontitis group compared to the con- trol group (p < 0.001). Stage III and IV perio- dontitis showed progressively increased se- verity, with stage IV presenting the highest mean values in BOP and CAL (Figure 2). Salivary Biomarker Levels RANKL levels were signiFicantly higher in the periodontitis group (261.61 ± 96.27 pg/ml) than in controls (125.62 ± 53.92 pg/ml), with the highest levels observed in stage III (Table 2). Similarly, IL-36γ levels were also signiFicantly elevated in periodontitis (340.77 ± 90.61 pg/ml) compared to con- trols (138.18 ± 50.68 pg/ml), while peaking in stage IV (Table 3). Conversely, IL-38 lev- els, in contrast, were signiFicantly lower in the periodontitis group (63.52 ± 21.24 pg/ml) than in controls (142.80 ± 28.53 pg/ml) (Table 4). Correlations of Salivary Biomarkers with the Periodontal Parameters IL-36γ showed a signiFicant positive correla- tion with CAL (r = 0.095, p = 0.024), whereas statistically non-signiFicant correlations were found between RANKL or IL-38 and clinical parameters (Table 5). Additionally, within stage III, IL-36γ showed signiFicant negative correlations with PI and BOP. In stage IV, BOP was signiFicantly and positively correlated with RANKL (Table 6). Inter-Biomarker Correlations A strong positive correlation was observed between RANKL and IL-36γ in both control (r = 0.671) and periodontitis groups (r = 0.560, p < 0.001). On the other hand, IL-38 was negatively correlated with both of them in the control group, while a signiFicant posi- tive correlation was observed in the perio- dontitis group (Table 7). Diagnostic Performance ROC analysis demonstrated the excellent di- agnostic performance of the three bi- omarkers in distinguishing between perio- dontal health and unstable generalized peri- odontitis. There were the AUCs: RANKL (0.944), IL-36γ (0.982), and IL-38 (0.960). On the Flap side, sensitivity ranged from 94.59% to 98.88%, and speciFicity ranged from 93.10% to 96.55%. However, the three biomarkers failed to signiFicantly distinguish between stage III and IV (AUC < 0.60). Discussion The present study evaluated the diagnostic potential of salivary IL-36γ, IL-38, and RANKL in distinguishing periodontal health from generalized unstable periodontitis us- ing ROC curves, sensitivity, and specificity concerning clinical periodontal parameters. The findings revealed high sensitivity and specificity for these biomarkers in differenti- ating health from disease, although their ability to distinguish between different peri- odontitis stages remains limited. Despite advancements in oral health care, periodontitis remains highly prevalent [17]. The Global Burden of Disease 2021 Study re- ported over 1 billion cases worldwide, with severe forms expected to rise by over 44% by 2050 [18]. This persistent burden is at- tributed to a reliance on restorative rather than preventive care, underscoring the ur- gent need for point-of-care diagnostics and personalized prevention strategies [19]. Periodontitis typically begins as gingivitis caused by biofilm accumulation, leading to dysbiosis, a shift toward pathogenic bacteria, and triggering a host immune-inflammatory response [20,21]. Current diagnostic meth- ods, including probing and radiographs, re- flect past damage and lack sensitivity to on- going inflammation or microbial changes [22]. Their subjectivity and variability high- light the need for more objective tools [23], prompting the inclusion of novel diagnostic domains in recent periodontal classifications [22]. A forward-looking approach aims to detect periodontitis before clinical signs emerge, utilizing biomarkers as predictive and prog- nostic indicators. They can identify disease presence, progression, or treatment re- sponse [24]. Salivary biomarkers like inter- leukins and RANKL have shown promise as non-invasive diagnostic tools. RANKL demonstrates excellent sensitivity and high specificity for distinguishing healthy from diseased states [25,26]. Similarly, IL-36γ, part of the IL-1 cytokine family, plays a key role in periodontal inflammation and bone loss, making it a significant emerging marker in chronic periodontal disease [27,28]. On the other hand, IL-38 can act as an inhib- itor of IL-36 pro-cytokines (IL-36α, ΙL-36β, and IL-36γ), as it partially blocks IL-36R, which is activated via these pro-cytokines and inhibited by IL-36Ra [29-31]. Therefore, the IL-38/IL-36R axis suggests the potential therapeutic benefit of IL-38 in inflammatory autoimmune diseases, primarily through its anti-inflammatory role in the development and resolution of these diseases [32]. Regarding periodontal disease, contrasting findings on IL-38 levels in different studies highlight its complex role in periodontal in- flammation. While increased GCF and saliva levels in one study suggest its potential as a marker for active disease [33], the decreased salivary levels in another study indicate that PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 4 further research is needed to fully under- stand its diagnostic utility [34]. Recently, studies demonstrated that individ- uals with periodontitis exhibited signifi- cantly lower salivary IL-38 levels compared to healthy individuals, suggesting a potential inverse relationship between IL-38 levels and the severity of periodontal disease [33,34]. Conversely, in another study, inter- leukin-36γ and IL-38 have been shown to have elevated salivary levels in periodontitis compared to healthy periodontium or gingi- vitis. Their salivary levels correlate with clin- ical periodontal parameters, suggesting a po- tential role in diagnosing periodontitis and assessing disease activity [33]. On the flap side, saliva contains a variety of biomarkers, including proteins, DNA, and RNA, which can be used to diagnose multiple diseases, such as cancer, cardiovascular dis- eases, and oral diseases [35,36]. Along with this, due to its elements that reflect the activ- ity of all periodontal sites, saliva content re- flects a consensus ‘whole mouth’ inflamma- tory status rather than at active disease sites, as with GCF analysis [37]. Besides, saliva col- lection, especially in children and large pop- ulations, is a non-invasive, easy-to-handle, and suitable method for mass screening [35,38]. Moreover, quantifying cytokines is particu- larly crucial for establishing threshold levels of the selected biomarkers, which are essen- tial for developing a chairside clinical tool. Therefore, the ELISA technique was chosen for quantifying biomarker concentrations due to its user-friendliness, high sensitivity, and specificity, which are attributed to the unique interaction between the antibody and antigen [39-41]. Alongside its excellent reproducibility and quantitative detection ability for pro-inflammatory and anti-in- flammatory cytokines [37,39,41,42]. Not only that, but ELISA can also be adapted for various applications, including the detection of cancer biomarkers and cytokines, and is compatible with different sample types such as blood, serum, and plasma [42-44]. Additionally, ELISA, by detecting specific bi- omarkers, is extensively used for early dis- ease diagnosis, including Alzheimer's and cancer [39,43,45]. Therefore, the technique employed in both clinical diagnostics and re- search laboratories is for its reliability and adaptability [44,46]. Demographic Distribution and Group Characteristics A sample size of 120 participants was calcu- lated, comprising a control group with healthy periodontium (n = 30) and a case group with generalized unstable periodonti- tis (n = 90). The unequal distribution of par- ticipants is attributed to several factors related to study design and participants' characteristics. Many studies use specific matching criteria to form control groups, such as age, sex, and socioeconomic status, which can limit the number of eligible control participants. For instance, in a study on periodontitis and rheumatoid arthritis, controls were matched 1:1 with periodontitis participants based on these criteria, which can inherently lead to unequal distribution if the pool of eligible controls is smaller [47]. Similarly, in our research, the control group criteria involved systemically and periodon- tally healthy people, aged >18 years, who had a BOP <10%, with more than 20 teeth in their mouth, and were free from current or previous periodontitis. Ordinarily, these cri- teria resulted in a lower percentage of peo- ple in the health center’s community com- pared with the periodontitis group criteria. Furthermore, the prevalence of periodontitis in some populations can lead to a higher number of participants with the condition compared to those without. Likewise, by us- ing the 2017 classification, periodontitis was found in 36.5% of the sampled Iraqi popula- tion, with severe cases (stages III and IV) be- ing the most common, accounting for 77.3% of periodontitis cases [48]. On the other hand, there were significant de- mographic differences between the groups, including gender and age (Table 1). The con- trol group consisted predominantly of fe- males (72.4%) and younger individuals (26.93 ± 4.14 years), while the periodontitis group was older (42.79 ± 12.39 years) with a higher percentage of males (56.2%). These differences are consistent with the well-es- tablished understanding that periodontitis is more prevalent in older individuals [49-52], and may be more pronounced in males [50,53-55]. Alongside other factors such as smoking and low socioeconomic status, male was identified as a significant risk factor for periodontitis [51,53]. Although the smokers were excluded from the current study, the socioeconomic status was not, and the gen- der selection was random. What is more, the distribution of periodonti- tis stages (Table 2) in the periodontitis group showed that the majority of participants were in advanced stages (Stage III, 39%, and Stage IV, 31.4%), which is significant in the context of understanding the severity of the condition in this cohort. The relatively low proportion of patients in Stage I (2.1%) and Stage II (4.2%) may suggest a study bias to- wards individuals with more severe forms of periodontitis, potentially due to the inclu- sion criteria favoring those with more no- ticeable clinical symptoms, such as involving only periodontitis patients with the general- ized extent and unstable status. Otherwise, this distribution is consistent with the various studies, which indicate that a significant portion of individuals are in ad- vanced stages of the disease. In a Norwegian population, Stage III and IV periodontitis were observed in 17.6% of the study popula- tion, with severe forms primarily occurring after 60 years of age [56]. Similarly, in a ru- ral Chinese population, more than half of the individuals had Stage III/IV periodontitis, with a rapid progression noted [57]. Like- wise, in a Sámis population in Northern Nor- way, 20.1% were in Stage III/IV [58]. Be- sides, the prevalence of severe periodontitis increases with age [57]. As well as, individu- als with Stage III/IV periodontitis reported more impairment in quality of life and mas- ticatory function [59,60], which led to a higher percentage of those patients in healthcare centers. In the context of comparisons between the study groups, clinical periodontal parame- ters (PL, BOP, PPD, and CAL) showed signifi- cant differences between the healthy control and periodontitis groups. Both PLI and BOP were significantly higher in the periodontitis group, especially in stages III and IV, com- pared to the control group (Figure 2, A, B). These findings support the expected link be- tween clinical inflammation and the severity of periodontitis, as higher PLI and BOP scores are associated with increased sever- ity. These parameters reflect the inflamma- tory condition of the gingival tissue and are used to evaluate the progression of perio- dontal disease [61-63]. The PPD and CAL also showed significant dif- ferences between stages III and IV (Figure 2, C, D), highlighting the increased severity of periodontal destruction in more advanced stages of the disease. Various factors, includ- ing host responses, influence this progres- sion [64], oxidative stress [65,66], microbial factors [67], functional impairments [60], and systemic conditions [68,69]. Studies in- dicate that higher CAL values correlate with more severe periodontal disease [61,62,70,71]. Likely, increased PPD is con- sistently associated with higher disease se- verity [61-63,72]. Salivary Biomarkers in Periodontal Health and Disease Salivary biomarkers, specifically RANKL, IL- 36γ, and IL-38, were examined to assess their potential diagnostic biomarkers, differ- entiating between periodontal health and generalized unstable periodontitis. RANKL In the context of periodontitis, literature has demonstrated that RANKL is a key mediator in bone resorption and a hallmark of perio- dontal disease [73,74], primarily through its PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 5 expression by activated T and B lymphocytes [75,76], osteocytes, and macrophages [77], and its regulation by immune responses and inflammatory cytokines [75]. Furthermore, the increased RANKL expression is at- tributed to the periodontopathic bacteria, such as Porphyromonas gingivalis, which are known to exacerbate periodontal disease [78]. In our study, RANKL demonstrated an AUC of 0.941, with a sensitivity of 97.22% and specificity of 93.10%, suggesting a highly ef- fective capacity of RANKL in differentiating healthy periodontium from generalized un- stable periodontitis (Table 9) (Figure 3-A). Consistent with the previous study, RANKL exhibits perfect sensitivity (1.00) and high specificity (0.92) in determining these condi- tions, making it a reliable biomarker for early detection and management of perio- dontal diseases [79]. Moreover, RANKL levels were significantly higher in the periodontitis group compared to the control group, with stage III showing a slightly higher mean level than stage IV (Ta- ble 3). This elevation is consistent with avail- able literature, which demonstrated that RANKL levels, both in saliva and gingival tis- sues, are significantly higher in patients with periodontitis compared to healthy controls has been observed in various forms of perio- dontitis [78,80-85], indicating its potential as a marker for periodontal disease [79,81,82]. Furthermore, in our study, RANKL levels cor- relate with PI, and PPD positively but non- significantly (Table 6). This is consistent with various studies, where RANKL levels are elevated in periodontal disease, charac- terized by the correlation between RANKL levels and these clinical parameters is often not statistically significant, indicating varia- bility in individual responses or other influ- encing factors [73,81,86]. Treatments such as antimicrobial photody- namic therapy and the use of resveratrol- containing mouthwash have been shown to reduce RANKL levels and improve periodon- tal parameters. However, the changes in RANKL levels do not always correlate di- rectly with changes in PI and PPD [87,88]. Despite that, these factors were considered excluded criteria in this study. Therefore, we suggested that while RANKL is an important factor in periodontal disease, its levels alone may not be a reliable predictor of clinical se- verity as measured by PI and PPD. Other fac- tors, including individual variability and ad- ditional inflammatory mediators, likely play significant roles. In addition, the relationship between clinical parameters such as BOP and CAL with RANKL levels has been explored (Table 6), indicating a negative, non-significant correlation between BOP and CAL with RANKL levels. Likewise, various studies indi- cate a negative correlation between BOP and CAL with RANKL levels, which is also not sta- tistically significant. These suggest that changes in BOP and CAL do not strongly pre- dict changes in RANKL levels in the context of periodontitis [86,88,89]. Alternatively, other studies demonstrate that the higher RANKL levels correlate posi- tively with clinical indicators of periodontal diseases, such as (PPD) and (CAL), suggest- ing that RANKL could act as a diagnostic marker for periodontal disease [78,81,84,85]. As well, detecting RANKL in saliva makes it a non-invasive biomarker for diagnosing periodontal disease, with high sensitivity and specificity in distinguishing between periodontal health and disease [79,81]. Additionally, in comparison with stages III and IV of periodontitis (Table 3), RANKL lev- els are significantly higher than in healthy controls, with stage III showing slightly higher levels than stage IV. Consistent with previous literature, it suggests a nuanced role of RANKL in different stages of perio- dontitis [26]. In other words, suggests that RANKL could be used as a biomarker for the severity of periodontitis, differentiating be- tween different stages. Although in this study, RANKL lacks discrimination between stage III periodontitis and stage IV periodon- titis (Table 9) (Figure 3), it could be because both stages represent severe forms of the disease, or a larger sample size is needed for that purpose. IL-36γ As RANKL, the current study showed IL-36γ with an even higher AUC of 0.979, with iden- tical sensitivity (97.22%) and slightly higher specificity (96.55%), reinforcing its robust role as a diagnostic marker for generalized unstable periodontitis (Table 9) (Figure 3). In another study, a combination of IL-1β with IL-36γ is particularly effective in distin- guishing periodontitis from periodontal health, suggesting its utility as part of a bi- omarker panel for early detection and moni- toring of periodontal disease. Despite its high sensitivity in detecting periodontitis, IL- 36γ exhibits a relatively low specificity when identifying individuals without periodontitis [34]. Recently, studies have consistently shown that IL-36γ levels are significantly elevated in patients with periodontitis compared to healthy individuals, particularly pronounced in more advanced stages of the periodontitis, where there is a greater degree of tissue de- struction and bone loss, as stages III and IV [27,33,90]. Our finding that IL-36γ levels fol- lowed a similar pattern as RANKL, with elevated levels in the periodontitis group, particularly in stages III and IV (Table 4), re- flects its role in the inflammatory response during periodontitis, consistent with the previous studies, given its involvement in pro-inflammatory signaling pathways. As an inflammatory response indicator, IL- 36γ enhances the expression of other pro-in- flammatory cytokines (IL-1β, IL-6, and TNF- α). While it is a bone loss biomarker associ- ated with periodontitis, it increases the RANKL/OPG ratio [27,91]. Additionally, it promotes neutrophil chemotaxis, a crucial component of the immune response at the oral barrier, which is mediated through fi- broblasts and is a key factor in the progres- sion of periodontitis [91]. Moreover, IL-36γ enhances the MAPK and TLR4 signaling pathways associated with chronic periodontitis; so that IL-36γ not only indicates disease presence but actively par- ticipates in its progression [91,92]. There- fore, targeting IL-36γ and its pathways could offer new therapeutic avenues for managing periodontitis. Hence, inhibiting IL-36γ activ- ity decreased neutrophil infiltration and bone resorption in experimental models, suggesting its potential as a therapeutic tar- get [90,91]. In harmonious, when examining the correla- tion between clinical periodontal parame- ters and salivary biomarkers (Table 6, Table 7), it was observed that CAL was positively correlated with IL-36γ in the periodontitis group, which suggests a link between the in- flammatory response (as indicated by IL-36γ levels) and clinical attachment loss, a key marker of periodontal destruction. Previ- ously, a study has indicated a positive corre- lation between salivary IL-36γ levels and clinical periodontal parameters, such as PPD and CAL. This implies that periodontal dis- ease severity increases IL-36γ levels, reflect- ing the inflammatory state of the periodon- tium [33]. Other clinical parameters, such as PLI, BOP, and PPD, did not show significant correla- tions with the biomarkers. This may indicate that these markers are more directly associ- ated with the chronicity and severity of peri- odontal destruction rather than with early inflammatory responses. Studies have identified a positive correlation between the concentrations of IL-36γ in oral fluids and certain clinical periodontal pa- rameters, including probing depth and clini- cal attachment loss. This evidence highlights the potential of these biomarkers in as- sessing the activity and severity of periodon- titis [33]. Conversely, in our study, BOP, the unique predictive test used by periodontists for routinely assessing the stability or pro- gression of periodontitis [93], showed nega- tive and non-significant correlations with IL- PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 6 36γ salivary biomarkers (Table 6). In addi- tion, IL-36γ is significantly elevated in perio- dontitis patients, correlated with the RANKL/OPG ratio, suggesting IL-36γ's in- volvement in perpetuating gingival inflam- mation and alveolar bone resorption, mak- ing it a potential therapeutic [27]. IL-38 In this study, IL-38 showed an AUC of 0.973, with perfect sensitivity (100%) and a speci- ficity of 93.10% (Table 9). These findings suggest that IL-38 may have a superior abil- ity to detect generalized unstable periodon- titis, further supporting its role as a diagnos- tic biomarker, consistent with S. Kc et al. [94]. To our knowledge, this is the first study to evaluate the diagnostic potential of IL-38. The contrasting findings on IL-38 levels in different studies highlight its complex role in periodontal inflammation. The increased levels of GCF and saliva in one study and of GCF in another one suggested its potential as a marker for active disease [33,95]. While the decreased salivary levels in another study [34] align with our findings, IL-38 was significantly lower in the periodontitis group compared to the healthy controls, and its lev- els were lower in stages III and IV (Table 5). This finding suggests that IL-38 might play an inhibitory role in periodontal inflamma- tion, as its reduced expression could be a re- sponse to chronic inflammation in periodon- tal tissues. The general trend supports the use of multi- ple biomarkers for better diagnostic preci- sion. Combining multiple biomarkers, such as IL-1β, IL-6, and MMP-8, improved diag- nostic accuracy for periodontitis. While RANKL, IL-36γ, and IL-38 are not specifically highlighted in these combinations, the gen- eral trend supports the use of multiple bi- omarkers for better diagnostic precision [94,96,97]. Aligning with previous studies, their gingival crevicular fluid, saliva, and serum levels cor- relate with disease severity and activity, providing a non-invasive means to assess periodontal health [26,33]. These findings support the potential of RANKL, IL-36y, and IL-38 as biomarkers for the early detection and monitoring of periodontal disease. The sensitivity and specificity values reflect that these biomarkers can accurately identify pa- tients with periodontitis and potentially as- sist in differentiating it from periodontal health across various disease stages. There- fore, combining IL-36γ, IL-38, and RANKL could serve as a comprehensive diagnostic biomarker panel for periodontal disease. The determination of diagnostic thresholds for biomarkers is crucial for distinguishing between periodontal health and various stages of periodontitis [94,96,98]. Recent research highlights several promising bi- omarkers and their potential diagnostic ap- plications. In clinical diagnostics [98,99], de- termining effective cut-off points for bi- omarkers is crucial, as they help patient stratification based on disease risk or treat- ment response [100,101]. The method selection for determining cut-off points should align with the biomarker's in- tended use, whether for diagnosis, progno- sis, or screening. Different applications may require different statistical approaches [102]. While ROC curve analysis is widely used, it is a common method for determining cut-off points, particularly for diagnostic purposes. It involves selecting a point that maximizes sensitivity and specificity, often using the Youden index [103,104]. How- ever, ROC curves are not ideal for prognostic biomarkers as they do not account for time- to-event data [102]. The proposed cut-off concentrations for dif- ferentiating periodontal health from perio- dontitis are 143.8 pg/ml, 196.9 pg/ml, and 67.05 pg/ml of RANKL, IL-36y, and IL-38, re- spectively (Table 9) (Figure 3-A). Additionally, proposed cut-off points for the biomarkers provide valuable insights into their diagnostic thresholds for identifying periodontal health and different stages of periodontitis. The cut-off points for the bi- omarkers RANKL, IL-36y, and IL-38 were 143.8 pg/ml, 196.9 pg/ml, and 66.69 pg/ml, respectively (Table 9) (Figure 3-B), suggest- ing the potential to discriminate between periodontal health and periodontitis stage III. On the other hand, the cut-off points for these biomarkers were 161.3 pg/ml, 256.8 pg/ml, and 111.9 pg/ml, respectively. Sug- gesting a potential to discriminate between periodontal health and periodontitis stage IV, as illustrated in (Table 9) (Figure 3-C). These cut-off concentrations further solidify the role of these biomarkers as effective di- agnostic tools for identifying early and mod- erate stages of periodontitis. However, these biomarkers may need to be supplemented with other diagnostic approaches, such as clinical examination or advanced imaging techniques, to improve accuracy in distin- guishing the most severe stages of periodon- titis [105,106]. Despite their potential strong diagnostic per- formance in distinguishing between healthy periodontium and periodontitis, the bi- omarkers demonstrated much lower diag- nostic accuracy when comparing stages III and IV of periodontitis. The AUC values be- tween stages III and IV for all three bi- omarkers were low, with values 0.532 pg/ml, 0.580 pg/ml, and 0.540 pg/ml, for three biomarkers RANKL, IL-36y, and IL-38, respectively (Table 9) (Figure 3-D) indicat- ing that these biomarkers are ineffective in discriminating between these two advanced stages of periodontitis as AUC below 0.7 is typically regarded as poor or ineffective [106-108]. This lack of discrimination between stages of periodontitis may be because both represent severe forms of the disease, where the in- flammatory and pathological processes may involve overlapping immunological path- ways. This makes further distinctions more difficult using these biomarkers alone [109,110], or biomarker levels may reach saturation, reducing their effectiveness in distinguishing between different stages of periodontal disease [109,111]. Limitations and Future Directions The findings of this study suggest the poten- tial utility of salivary biomarkers in diagnos- ing periodontitis, although several limita- tions exist. The unequal distribution of par- ticipants between the control and periodon- titis groups could introduce bias, particularly in the biomarker comparison across differ- ent periodontitis stages. Additionally, they were insufficient to distinguish between dif- ferent severity stages of periodontitis (stage III vs. stage IV). Therefore, the relatively small sample size in the control group may limit the generalizability of these findings. Furthermore, interpreting large datasets from salivary diagnostics can be complex, re- quiring advanced analytical techniques [35]. While many potential biomarkers have been identified, further research is needed to validate their effectiveness and reliability in clinical settings [96,100]. Future studies with larger, more balanced cohorts and lon- gitudinal designs are recommended for in- vestigating the role of these biomarkers in distinguishing between healthy and diseased states. Moreover, the integration of multiple salivary biomarkers may enhance early diag- nosis and monitoring of periodontal disease in a non-invasive and patient-friendly man- ner. Therefore, salivary RANKL, IL-36γ, and IL-38 represent promising tools for future diagnostic protocols in personalized perio- dontal care. Conclusion Salivary RANKL and IL-36γ levels were ele- vated, while IL-38 was reduced in periodon- titis, indicating their roles in periodontitis pathogenesis. ROC analysis revealed RANKL had the highest diagnostic accuracy. These Findings support the potential use of RANKL, IL-36γ, and IL-38 as non-invasive bi- omarkers for early detection and monitoring of periodontal disease. Conflict of Interest No conFlicts to declare. PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 7 References 1. Chapple IL, Mealey BL, Van Dyke TE, Bartold PM, Dommisch H, Eickholz P, et al. Periodontal health and gingival diseases and conditions on an intact and a reduced periodontium: Consensus report of workgroup 1 of the 2017 World Workshop on the ClassiPication of Periodontal and Peri-Implant Dis- eases and Conditions. Journal of Periodontology. 2018; 89: S74-S84. 2. Trindade, D., Carvalho, R., Machado, V., Cham- brone, L., Mendes, J. J., Botelho, J. (2023). Preva- lence of periodontitis in dentate people between 2011 and 2020: A systematic review and meta- analysis of epidemiological studies. Journal of clin- ical periodontology, 50(5), 604–626. 3. Kassebaum NJ, Bernabé E, Dahiya M, Bhandari B, Murray CJ, Marcenes W. Global burden of severe periodontitis in 1990-2010: a systematic review and meta-regression. Journal of dental research. 2014 Nov; 93(11): 1045-53. 4. GBD 2017 Oral Disorders Collaborators, et al. Global, regional, and national levels and trends in burden of oral conditions from 1990 to 2017: a systematic analysis for the global burden of dis- ease 2017 study. Journal of dental research, 2020, 99. 4: 362-373. 5. Tonetti, Maurizio S., et al. Impact of the global burden of periodontal diseases on health, nutri- tion and wellbeing of mankind: A call for global ac- tion. Journal of clinical periodontology, 2017, 44. 5: 456-462. 6. Graziani, Filippo; Tsakos, Georgios. Patient- based outcomes and quality of life. Periodontology 2000, 2020, 83. 1: 277-294. 7. Tsuchida, Satoshi, et al. Dark matter signals on a laser interferometer. Physical Review D, 2020, 101. 2: 023005. 8. Liccardo, Daniela, et al. Periodontal disease: a risk factor for diabetes and cardiovascular disease. International journal of molecular sciences, 2019, 20. 6: 1414. 9. Asher, Wesley B., et al. GPCR-mediated β-arres- tin activation deconvoluted with single-molecule precision. Cell, 2022, 185. 10: 1661-1675. e16. 10. Papapanou PN, Sanz M, et al. Periodontitis: Consensus report of Workgroup 2 of the 2017 World Workshop on the ClassiPication of Perio- dontal and Pre-Implant Diseases and Conditions. J Clin Periodontol. 2018; 45(Suppl 20): S162–S170. 11. Cohen, Jérémie F., et al. STARD 2015 guidelines for reporting diagnostic accuracy studies: expla- nation and elaboration. BMJ open, 2016, 6. 11: e012799. 12. Sungkar, S., Chismirina, S., Nasution, A., Imaduddin, H. (2020). The Effect of Cheese and Milk on Buffering Capacity of Saliva in Children 10- 12 Years. Journal of Biomimetics, Biomaterials and Biomedical Engineering, 48, 105 - 110. 13. Grzesiak-Gasek, I., Dębska-Łasut, K. (2025). Oc- currence of dental caries, the prevalence of hy- giene and dietary habits, and the inPluence of the workout session on the selected parameters of sa- liva in youth swimmers. Dental and medical prob- lems. 14. Yau, J., Fong, T., Wan, A., Ho, R. (2022). Compar- ison of passive drool and cotton-based collection methods for salivary C-reactive protein measure- ment. American Journal of Human Biology, 34. 15. Lagerlöf F. Caries-Protective Factors in Saliva. Advances in Dental Research. 1994;8(2):229-238. doi:10.1177/08959374940080021601 16. Ainamo, J., Bay, I. (1975). Problems and pro- posals for recording gingivitis and plaque. Interna- tional dental journal, 25 4, 229-35. 17. M. X. Chen, Y. J. Zhong, Q. Q. Dong, H. M. Wong, Y. F. Wen. Global, Regional, and National Burden of Severe Periodontitis, 1990–2019: An Analysis of the Global Burden of Disease Study 2019, Journal of Clinical Periodontology 48, no. 9 (2021): 1165– 1188. 18. G. G. Nascimento, S. Alves-Costa, M. Romandini. Burden of Severe Periodontitis and Edentulism in 2021, With Projections up to 2050: The Global Burden of Disease 2021 Study, Journal of Perio- dontal Research 59, no. 5 (2024): 823–867. 19. Dong, A., Proctor, G., Zaric, S. (2025). Diagnos- tic Accuracy of Microbiome-Derived Biomarkers in Periodontitis: Systematic Review and Meta- Analysis. Journal of periodontal research. 20. Murakami S, Mealey BL, Mariotti A, Chapple ILC. Dental plaque-induced gingival conditions. J Clin Periodontol. 2018; 45(Suppl 20): S17-S27. doi: 10. 1111/jcpe. 12937). 21. Chapple, I., Abdulkareem, A., Al-Taweel, F., Gul, S., Al-Sharqi, A., Sha, A. (2023). Current concepts in the pathogenesis of periodontitis: from symbiosis to dysbiosis. Journal of Oral Microbiology, 15. https: //doi. org/10. 1080/20002297. 2023. 2197779. 22. Tonetti, Maurizio S.; Greenwell, Henry; Kor- man, Kenneth S. Staging and grading of periodon- titis: Framework and proposal of a new classiPica- tion and case dePinition. Journal of periodontology, 2018, 89: S159-S172. 23. J. Liukkonen, U. K. Gürsoy, E. Könönen, et al. Salivary Biomarkers in Association with Periodon- tal Parameters and the Periodontitis Risk Haplo- type, Innate Immunity 24, no. 7 (2018): 439–447. 24. CaPiero, C., Spagnuolo, G., Marenzi, G., Martus- celli, R., Colamaio, M., Leuci, S. (2021). Predictive Periodontitis: The Most Promising Salivary Bi- omarkers for Early Diagnosis of Periodontitis. Journal of Clinical Medicine, 10. 25. Akram, H., Salman, S., Najah, A., Saliem, S., Fad- hil, R., Mahmood, A. (2024). Diagnostic ability of salivary TNF-α and RANKL to differentiate perio- dontitis from periodontal health (case-control study). Journal of Emergency Medicine, Trauma and Acute Care. 26. Relvas, M., Silvestre, R., Gonçalves, M., Cabral, C., Mendes-Frias, A., Monteiro, L., Da Costa, A. (2023). Analysis of Salivary Levels of IL-1β, IL17A, OPG and RANK-L in Periodontitis Using the 2017 ClassiPication of Periodontal Diseases—An Explor- atory Observational Study. Journal of Clinical Med- icine, 12 27. Cloitre, Alexandra, et al. IL-36γ is a pivotal in- Plammatory player in periodontitis-associated bone loss. ScientiPic reports, 2019, 9. 1: 19257. 28. Papathanasiou, E., Conti, P., Carinci, F., Lauri- tano, D., Theoharides, T. (2020). IL-1 Superfamily Members and Periodontal Diseases. Journal of Dental Research, 99, 1425 - 1434. 29. Boutet, M. A.; Bart, G.; Penhoat, M.; Amiaud, J.; Brulin, B.; Charrier, C.; Vigne, S. Distinct expression of interleukin (IL)-36 alpha, beta and gamma, their antagonist IL-36Ra and IL-38 in psoriasis, rheu- matoid arthritis and Crohn’s disease. Clin. Exp. Im- munol. 2016, 1841, 59–73. 30. Vigne, S.; Palmer, G.; Lamacchia, C.; Martin, P.; Talabot-Ayer, D.; Rodriguez, E.; Ronchi, F.; Sal- lusto, F.; Dinh, H.; Sims, J. E.; et al. CS16-5. IL-36R Ligands are Potent Regulators of Dendritic and T Cells. Cytokine 2011, 56, 105–106. 31. Mutamba, S.; Allison, A.; Mahida, Y.; Barrow, P.; Foster, N. Expression of IL-1Rrp2 by human mye- lomonocytic cells is unique to DCs and facilitates DC maturation by IL-1F8 and IL-1F9. Eur. J. Immu- nol. 2012, 42, 607–617. 32. Xie, L., Huang, Z., Li, H., Liu, X., Zheng, S., Su, W. (2019). IL-38: A New Player in InPlammatory Au- toimmune Disorders. Biomolecules, 9. 33. Korkmaz, H., Hatipoğlu, M., Kayar, N. (2023). Interleukin-38: A crucial player in periodontitis. Oral diseases. 34. Gürsel, M., Bayat, T., Ndar, N., Tak, A., Karadu- man, B. (2024). Investigating the diagnostic poten- tial of IL-1β, IL-10, and IL-36γ in gingival crevicu- lar Pluid in patients with different periodontal con- ditions. Biomarkers, 29, 118 - 126. 35. Khurshid, Z., Warsi, I., Moin, S., Slowey, P., Latif, M., Zohaib, S., Zafar, M., Zafar, M. (2020). Biochem- ical analysis of oral Pluids for disease detection. Advances in clinical chemistry, 100, 205-253. https: //doi. org/10. 1016/bs. acc. 2020. 04. 005. 36. Li, F., Yoshizawa, J., Kim, K., Kanjanapangka, J., Grogan, T., Wang, X., Elashoff, D., Ishikawa, S., Chia, D., Liao, W., Akin, D., Yan, X., Lee, M., Choi, R., Kim, S., Kang, S., Bae, J., Sohn, T., Lee, J., Choi, M., Min, B., Lee, J., Kim, J., Kim, Y., Kim, S., Wong, D. (2018). Dis- covery and Validation of Salivary Extracellular RNA Biomarkers for Noninvasive Detection of Gas- tric Cancer. Clinical chemistry, 64 10, 1513-1521. https: //doi. org/10. 1373/clinchem. 2018. 290569. 37. Serdar, C., Guney, Z., Balci, N., Altingoz, S., Ser- dar, M., Kurgan, Ş. (2025). Salivary trimethylamine N-oxide: a novel non-invasive marker for perio- dontal inPlammation. Clinical Oral Investigations, 29. https: //doi. org/10. 1007/s00784-025- 06295-4. 38. Papagerakis, P., Zheng, L., Kim, D., Said, R., Eh- lert, A., Chung, K., Papagerakis, S. (2019). Saliva and Gingival Crevicular Fluid (GCF) Collection for Biomarker Screening. Methods in Molecular Biol- ogy, 1922, 549-562. https: //doi. org/10. 1007/978-1-4939-9012-2_41. 39. Pérez-Ruiz, E., Decrop, D., Ven, K., Tripodi, L., Leirs, K., Rosseels, J., Van De Wouwer, M., Geukens, N., De Vos, A., Vanmechelen, E., Winderickx, J., Lammertyn, J., Spasic, D. (2018). Digital ELISA for the quantiPication of attomolar concentrations of Alzheimer's disease biomarker protein Tau in bio- logical samples. Analytica chimica acta, 1015, 74- 81. https: //doi. org/10. 1016/j. aca. 2018. 02. 011. 40. Cohen, L., Cui, N., Cai, Y., Garden, P., Li, X., Weitz, D., Walt, D. (2020). Single Molecule Protein Detec- tion with Attomolar Sensitivity Using Droplet Dig- ital Enzyme-Linked Immunosorbent Assay. ACS nano. https: //doi. org/10. 1021/acsnano. 0c02378. 41. Byun, J., Lee, K., Park, Y., Song, D., Min, Y., Kim, D. (2025). Revisiting ELISA with in situ ampliPica- tion of biomarkers to boost its sensitivity. Sensors and Actuators B: Chemical. https: //doi. org/10. 1016/j. snb. 2024. 136780. 42. Arya, S., Estrela, P. (2018). Recent Advances in Enhancement Strategies for Electrochemical ELISA-Based Immunoassays for Cancer Biomarker Detection. Sensors (Basel, Switzerland), 18. https: //doi. org/10. 3390/s18072010. 43. Japp, N., Souchek, J., Sasson, A., Hollingsworth, M., Batra, S., Junker, W. (2021). Tumor Biomarker In-Solution QuantiPication, Standard Production, PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 8 and Multiplex Detection. Journal of Immunology Research, 2021. https: //doi. org/10. 1155/2021/9942605. 44. Murdock, R., Shen, L., GrifPin, D., Kelley-Lough- nane, N., Papautsky, I., Hagen, J. (2013). Optimiza- tion of a paper-based ELISA for a human perfor- mance biomarker. Analytical chemistry, 85 23, 11634-42. https: //doi. org/10. 1021/ac403040a. 45. Shi, J., Barman, S., Cheng, S., Zeng, Y. (2024). Metal-organic framework-interfaced ELISA probe enables ultrasensitive detection of extracellular vesicle biomarkers. Journal of materials chemis- try. B. https: //doi. org/10. 1039/d4tb00585f. 46. Maley, A., Garden, P., Walt, D. (2020). Simpli- Pied Digital Enzyme-Linked Immunosorbent Assay Using Tyramide Signal AmpliPication and Fibrin Hydrogels. ACS sensors. https: //doi. org/10. 1021/acssensors. 0c01661. 47. Choi, Y., Lee, K. (2021). Periodontitis as a Risk Factor for Rheumatoid Arthritis: a Matched-Co- hort Study. International Dental Journal, 71, 516 - 521. https: //doi. org/10. 1016/j. identj. 2021. 01. 006. 48. Imran, N., Abdulbaqi, H., Milward, M. (2024). The prevalence of periodontitis in an Iraqi popula- tion using the 2017 classiPication. Journal of Bagh- dad College of Dentistry. https: //doi. org/10. 26477/jbcd. v36i2. 3668. 49. Ju, X., Harford, J., Luzzi, L., Jamieson, L. (2021). Prevalence, extent, and severity of periodontitis among Australian older adults: Comparison of two generations. Journal of Periodontology, 93, 1387 - 1400. https: //doi. org/10. 1002/JPER. 21-0458. 50. Renvert, S., Persson, R., Persson, R. (2013). Tooth loss and periodontitis in older individuals: results from the Swedish National Study on Aging and Care. Journal of periodontology, 84 8, 1134- 44. https: //doi. org/10. 1902/jop. 2012. 120378. 51. Eke, P., Borgnakke, W., Genco, R. (2020). Re- cent epidemiologic trends in periodontitis in the USA. Periodontology 2000, 82 1, 257-267. https: //doi. org/10. 1111/prd. 12323. 52. Eke, P., Wei, L., Borgnakke, W., Thornton-Ev- ans, G., Zhang, X., Lu, H., Mcguire, L., Genco, R. (2016). Periodontitis prevalence in adults ≥ 65 years of age, in the USA. Periodontology 2000, 72 1, 76-95. https: //doi. org/10. 1111/prd. 12145. 53. Bengtsson, V., Persson, G., Berglund, J., Renvert, S. (2021). Periodontitis related to cardiovascular events and mortality: a long-time longitudinal study. Clinical Oral Investigations, 25, 4085 - 4095. https: //doi. org/10. 1007/s00784-020-03739-x. 54. Wulandari, P., Widkaja, D., Nasution, A., Syah- putra, A., Gabrina, G. (2022). Association between age, gender and education level with the severity of periodontitis in pre-elderly and elderly patients. Dental Journal (Majalah Kedokteran Gigi). https: //doi. org/10. 20473/j. djmkg. v55. i1. p16-20. 55. Groenewegen, H., Bierman, W., Delli, K., Dijks- tra, P., Nesse, W., Vissink, A., Spijkervet, F. (2019). Severe periodontitis is more common in HIV- in- fected patients. The Journal of infection, 78 3, 171- 177. https: //doi. org/10. 1016/j. jinf. 2018. 11. 008. 56. Stødle, I., Verket, A., Høvik, H., Sen, A., Koldsland, O. (2021). Prevalence of periodontitis based on the 2017 classiPication in a Norwegian population. The HUNT Study. Journal of clinical periodontology. https: //doi. org/10. 1111/jcpe. 13507. 57. Iao, S., Pei, X., Ouyang, X., Liu, J., Liu, W., Cao, C. (2020). Natural progression of periodontal diseases in Chinese villagers based on the 2018 classiPication. Journal of periodontology. https: //doi. org/10. 1002/JPER. 20-0199. 58. Bongo, A., Brustad, M., Oscarson, N., Jönsson, B. (2020). Periodontal health in an indigenous Sámi population in Northern Norway: a cross-sectional study. BMC Oral Health, 20. https: //doi. org/10. 1186/s12903-020-01098-3. 59. Goergen, J., Albandar, J., Oppermann, R., Rösing, C., Susin, C., Haas, A. (2021). Periodontitis Stage and Grade are associated with poor oral health re- lated quality of life: Pindings from the Porto Alegre Cohort Study. Journal of clinical periodontology. https: //doi. org/10. 1111/jcpe. 13527. 60. Uy, S., Deng, K., Fok, C., Fok, M., Pelekos, G., Tonetti, M. (2021). Food Intake, Masticatory Func- tion, Tooth Mobility, Loss of Posterior Support and Diminished Quality of Life Are Associated with More Advanced Periodontitis Stage Diagnosis. Journal of clinical periodontology. https: //doi. org/10. 1111/jcpe. 13588. 61. González-Febles, J., Rodrıǵuez-Lozano, B., Sánchez-Piedra, C., Garnier-Rodrıǵuez, J., Busta- bad, S., Hernandez-Gonzalez, M., González-Dávila, E., Sanz, M., Dıáz-González, F. (2020). Association between periodontitis and anti-citrullinated pro- tein antibodies in rheumatoid arthritis patients: a cross-sectional study. Arthritis Research & Ther- apy, 22. https: //doi. org/10. 1186/s13075-020- 2121-6. 62. Rodrıǵuez-Lozano, B., González-Febles, J., Gar- nier-Rodrıǵuez, J., Dadlani, S., Bustabad-Reyes, S., Sanz, M., Sánchez-Alonso, F., Sánchez-Piedra, C., González-Dávila, E., Dıáz-González, F. (2019). As- sociation between severity of periodontitis and clinical activity in rheumatoid arthritis patients: a case–control study. Arthritis Research & Therapy, 21. https: //doi. org/10. 1186/s13075-019-1808- z. 63. Pham, T. (2018). The association between per- iodontal disease severity and metabolic syndrome in Vietnamese patients. International Journal of Dental Hygiene, 16, 484–491. https: //doi. org/10. 1111/idh. 12350. 64. Zhao, X., Liu, W., Wu, Z., He, X., Tang, Y., He, Q., Lin, C., Chen, Y., Luo, G., Yu, T., Wang, X. (2024). Hepatocyte growth factor is protective in early stage but bone-destructive in late stage of experi- mental periodontitis. Journal of periodontal re- search. https: //doi. org/10. 1111/jre. 13237. 65. Zamora-Perez, A., Ortiz-Garcıá, Y., Lazalde-Ra- mos, B., Guerrero-Velázquez, C., Gómez-Meda, B., Ramıŕez-Aguilar, M., Zúñiga-González, G. (2015). Increased micronuclei and nuclear abnormalities in buccal mucosa and oxidative damage in saliva from patients with chronic and aggressive perio- dontal diseases. Journal of Periodontal Research, 50 1, 28-36. https: //doi. org/10. 1111/jre. 12175. 66. Sczepanik, F., Grossi, M., Casati, M., Goldberg, M., Glogauer, M., Fine, N., Tenenbaum, H. (2020). Periodontitis is an inPlammatory disease of oxida- tive stress: We should treat it that way. Periodon- tology 2000, 84 1, 45-68. https: //doi. org/10. 1111/prd. 12342. 67. How, K., Song, K., Chan, K. (2016). Porphy- romonas gingivalis: An Overview of Periodonto- pathic Pathogen below the Gum Line. Frontiers in Microbiology, 7. https: //doi. org/10. 3389/fmicb. 2016. 00053. 68. Sato, K., Yamazaki, K., Kato, T., Nakanishi, Y., Tsuzuno, T., Yokoji-Takeuchi, M., Yamada-Hara, M., Miura, N., Okuda, S., Ohno, H., Yamazaki, K. (2021). Obesity-Related Gut Microbiota Aggra- vates Alveolar Bone Destruction in Experimental Periodontitis through Elevation of Uric Acid. mBio, 12. https: //doi. org/10. 1128/mBio. 00771-21. 69. Sonnenschein, S., Meyle, J. (2015). Local in- Plammatory reactions in patients with diabetes and periodontitis. Periodontology 2000, 69 1, 221- 54. https: //doi. org/10. 1111/prd. 12089. 70. Durand, R., Roufegarinejad, A., Chandad, F., Rompré, P., Voyer, R., Michalowicz, B., Emami, E. (2019). Dental caries is positively associated with periodontal disease severity. Clinical Oral Investi- gations, 23, 3811 - 3819. https: //doi. org/10. 1007/s00784-019-02810-6. 71. Habibagahi, Z., Khorshidi, H., Hekmati, S. (2016). Periodontal Health Status among Patients with Behçet's Disease. ScientiPica, 2016. https: //doi. org/10. 1155/2016/7506041. 72. Kaneko, C., Kobayashi, T., Ito, S., Sugita, N., Mu- rasawa, A., Ishikawa, H., Tabeta, K. (2021). Associ- ation among periodontitis severity, anti-agalacto- syl immunoglobulin G titer, and the disease activ- ity of rheumatoid arthritis. Journal of Periodontal Research. https: //doi. org/10. 1111/jre. 12867. 73. Ibraheem, W., Fageeh, H., Preethanath, R., Al- zahrani, F., Al-Zawawi, A., Divakar, D., Al-Kheraif, A. (2020). Comparison of RANKL and osteopro- tegerin levels in the gingival crevicular Pluid of young cigarette- and waterpipe-smokers and indi- viduals using electronic nicotine delivery systems. Archives of oral biology, 115, 104714. https: //doi. org/10. 1016/j. archoralbio. 2020. 104714. 74. Chen, B., Wu, W., Sun, W., Zhang, Q., Yan, F., Xiao, Y. (2014). RANKL Expression in Periodontal Disease: Where Does RANKL Come From? BioMed Research International, 2014. https: //doi. org/10. 1155/2014/731039. 75. Taubman, M., Valverde, P., Han, X., Kawai, T. (2005). Immune Response: The Key to Bone Re- sorption in Periodontal Disease. Journal of perio- dontology, 76 Suppl 11S, 2033-2041. https: //doi. org/10. 1902/jop. 2005. 76. 11-S. 2033. 76. Kawai, T., Matsuyama, T., Hosokawa, Y., Makihira, S., Seki, M., Karimbux, N., Gonçalves, R., Valverde, P., Dibart, S., Li, Y., Miranda, L., Ernst, C., Izumi, Y., Taubman, M. (2006). B and T lympho- cytes are the primary sources of RANKL in the bone resorptive lesion of periodontal disease. The American journal of pathology, 169 3, 987-98. https: //doi. org/10. 2353/AJPATH. 2006. 060180. 77. Kittaka, M., Yoshimoto, T., Levitan, M., Urata, R., Choi, R., Teno, Y., Xie, Y., Kitase, Y., Prideaux, M., Dallas, S., Robling, A., Ueki, Y. (2023). Osteocyte RANKL Drives Bone Resorption in Mouse Liga- ture-Induced Periodontitis. Journal of Bone and Mineral Research, 38, 1521 - 1540. https: //doi. org/10. 1002/jbmr. 4897. 78. Wara-Aswapati, N., Surarit, R., Chayasadom, A., Boch, J., Pitiphat, W. (2007). RANKL upregulation associated with periodontitis and Porphyromonas gingivalis. Journal of periodontology, 78 6, 1062-9. https: //doi. org/10. 1902/JOP. 2007. 060398. 79. Abdullameer, M., Abdulkareem, A. (2023). Di- agnostic potential of salivary interleukin-17, RANKL, and OPG to differentiate between perio- dontal health and disease and discriminate stable and unstable periodontitis: A case-control study. Health Science Reports, 6. 80. Teodorescu, A., Martu, I., Teslaru, S., Kappen- berg-Niţescu, D., Goriuc, A., Luchian, I., Mârțu, M., Solomon, S., Martu, S. (2019). Assessment of PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 9 Salivary Levels of RANKL and OPG in Aggressive versus Chronic Periodontitis. Journal of Immunol- ogy Research, 2019. 81. Asif, S., Ahmad, B., Hamza, S., Taib, H., Kassim, N., Zainuddin, S. (2021). Investigation of Salivary RANKL and OPG Levels in Periodontitis Patients at Hospital Universiti Sains Malaysia. European Jour- nal of Dentistry, 16, 173 - 178. https: //doi. org/10. 1055/s-0041-1731930. 82. Ochanji, A., Matu, N., Mulli, T. (2017). Associa- tion of salivary RANKL and osteoprotegerin levels with periodontal health. Clinical and Experimental Dental Research, 3, 45 - 50. https: //doi. org/10. 1002/cre2. 49. 83. Crotti, T., Smith, M., Hirsch, R., Soukoulis, S., Weedon, H., Capone, M., Ahern, M., Haynes, D. (2003). Receptor activator NF kappaB ligand (RANKL) and osteoprotegerin (OPG) protein ex- pression in periodontitis. Journal of periodontal research, 38 4, 380-7. https: //doi. org/10. 1034/J. 1600-0765. 2003. 00615. X. 84. Bostancı, N., Ilgenli, T., Emingil, G., Afacan, B., Han, B., Töz, H., Atilla, G., Hughes, F., Belibasakis, G. (2007). Gingival crevicular Pluid levels of RANKL and OPG in periodontal diseases: implications of their relative ratio. Journal of clinical periodontol- ogy, 34 5, 370-6. https: //doi. org/10. 1111/J. 1600-051X. 2007. 01061. X. 85. Al-Ghurabi BH, Mohssen SM. Salivary Level of RANKL and OPG in Chronic Periodontitis. J Bagh Coll Dent. 2015; 27(1): 189–94. https: //doi. org/10. 12816/0015286). 86. Krasavcevic, A., Nikolic, N., Mijailović, I., C�arkić, J., Milinković, I., Jankovic, S., Aleksić, Z., Milašin, J. (2020). Impact of Notch signalling molecules and bone resorption regulators on clinical parameters in periodontitis. Journal of periodontal research. https: //doi. org/10. 1111/jre. 12801. 87. Rahman, S., Gv, G., Mehta, D. (2020). A clinico- microbiological and biochemical study evaluating the adjunctive use of antimicrobial photodynamic therapy and local drug delivery of 1. 2 % simvas- tatin gel compared to scaling and root planning alone. Photodiagnosis and photodynamic therapy, 102017. https: //doi. org/10. 1016/j. pdpdt. 2020. 102017. 88. Mohammed, S., Akram, H. (2024). Evaluating the EfPicacy of Resveratrol-Containing Mouthwash as an Adjunct Treatment for Periodontitis: A Ran- domized Clinical Trial. European journal of dentis- try. https: //doi. org/10. 1055/s-0044-1788686. 89. Karcı, B., O� ncü, E., Doğan, M. (2021). The Effect of Different Dental Implant Surface Characteristics on Bone Immunologic Biomarkers and Microbio- logic Parameters: A Randomized Clinical Study. The International journal of periodontics & restor- ative dentistry, 41 4, 589-597. 90. Liu, J., Meng, H., Mao, Y., Zhong, L., Pan, W., Chen, Q. (2024). IL-36 Regulates Neutrophil Chem- otaxis and Bone Loss at the Oral Barrier. Journal of Dental Research, 103, 442 - 451. 91. Babaloo, A., Shirmohammadi, A., San- doghchian, S., Kamalzadeh, A., Ghasemi, S. (2018). Evaluation of the effect of IL-36γ expression on chronic periodontitis by enhancing the MAPK and TLR4 signaling pathways: A basic research. Jour- nal of Dental Research, Dental Clinics, Dental Pro- spects, 12, 159 - 165. 92. Babaloo, A., Shirmohammadi, A., Ghasemi, S., Sandoghchian, S., Parvan, S., Fathi, N. (2019). Ef- fect of IL-36γ expression on chronic periodontitis with an increase in NF-KB signaling pathway. Jour- nal of Advanced Periodontology & Implant Dentis- try, 11, 28 - 33. 93. Lang, N. P.; Joss, A.; Orsanic, T.; Gusberti, F. A.; Siegrist, B. E. Bleeding on probing. A predictor for the progression of periodontal disease? J. Clin. Per- iodontol. 1986, 13, 590–596. 94. Kc, S., Wang, X., Gallagher, J. (2019). Diagnostic sensitivity and speciPicity of host-derived salivary biomarkers in periodontal disease amongst adults: Systematic review. Journal of Clinical Peri- odontology. 95. Das, N. (2024). Assessment of IL-38 And IL-40 Levels in Gingival Crevicular Fluid in Gingivitis and Stage III Grade C Periodontitis: A Comparative Clinical Study. South Eastern European Journal of Public Health. 96. Blanco-Pintos, T., Regueira-Iglesias, A., Seijo- Porto, I., Balsa-Castro, C., Castelo-Baz, P., Nibali, L., Tomás, I. (2023). Accuracy of periodontitis diag- nosis obtained using multiple molecular bi- omarkers in oral Pluids: A systematic review and meta-analysis. Journal of clinical periodontology. 97. Ebersole, J. L.; Nagarajan, R.; Akers, D.; Miller, C. S. Targeted salivary biomarkers for discrimina- tion of periodontal health and disease(s). Front. Cell Infect. Microbiol. 2015, 5, 62. 98. Arias-Bujanda, N., Regueira-Iglesias, A., Balsa- Castro, C., Nibali, L., Donos, N., Tomás, I. (2019). Accuracy of single molecular biomarkers in saliva for the diagnosis of periodontitis: a systematic re- view and meta-analysis. Journal of clinical perio- dontology. 99. De Lima, C., Acevedo, A., Grisi, D., Taba, M., Guerra, E., De Luca Canto, G. (2016). Host-derived salivary biomarkers in diagnosing periodontal dis- ease: systematic review and meta-analysis. Jour- nal of clinical periodontology, 43 6, 492-502. https: //doi. org/10. 1111/jcpe. 12538. 100. Denkert, C., Loibl, S., Müller, B., Eidtmann, H., Schmitt, W., Eiermann, W., Gerber, B., Tesch, H., Hilfrich, J., Huober, J., Fehm, T., Barinoff, J., Jackisch, C., Prinzler, J., Rüdiger, T., Erbstößer, E., Blohmer, J., Budczies, J., Mehta, K., Minckwitz, G. (2013). Ki67 levels as predictive and prognostic parame- ters in pretherapeutic breast cancer core biopsies: a translational investigation in the neoadjuvant GeparTrio trial. Annals of oncology: ofPicial journal of the European Society for Medical Oncology, 24 11, 2786-93. https: //doi. org/10. 1093/an- nonc/mdt350. 101. Polley, M., Dignam, J. (2021). Statistical Con- siderations in the Evaluation of Continuous Bi- omarkers. The Journal of Nuclear Medicine, 62, 605 - 611. 102. Duarte, P. (2021). Cut Point IdentiPication of Continuous Biomarkers: A Challenge That Goes Beyond Statistical Aspects. The Journal of Nuclear Medicine, 62, 1833 - 1833. https: //doi. org/10. 2967/jnumed. 121. 262905. 103. Bantis, L., Nakas, C., Reiser, B. (2013). Con- struction of conPidence regions in the ROC space after the estimation of the optimal Youden index- based cut-off point. Biometrics, 70. https: //doi. org/10. 1111/biom. 12107. 104. U� nal, I. (2017). DePining an Optimal Cut-Point Value in ROC Analysis: An Alternative Approach. Computational and Mathematical Methods in Medicine, 2017. https: //doi. org/10. 1155/2017/3762651. 105. Sexton, W., Lin, Y., Kryscio, R., Dawson, D., Ebersole, J., Miller, C. (2011). Salivary biomarkers of periodontal disease in response to treatment. Journal of clinical periodontology, 38 5, 434-41. https: //doi. org/10. 1111/j. 1600-051X. 2011. 01706. x. 106. Zucchelli, A., Vetrano, D., Grande, G., Calde- rón-Larrañaga, A., Fratiglioni, L., Marengoni, A., Rizzuto, D. (2019). Comparing the prognostic value of geriatric health indicators: a population- based study. BMC Medicine, 17. https: //doi. org/10. 1186/s12916-019-1418-2. 107. Viasus, D., Del Rio-Pertuz, G., Simonetti, A., Garcıá-Vidal, C., Acosta-Reyes, J., Garavito, A., Car- ratalá, J. (2016). Biomarkers for predicting short- term mortality in community-acquired pneumo- nia: A systematic review and meta-analysis. The Journal of infection, 72 3, 273-82. https: //doi. org/10. 1016/j. jinf. 2016. 01. 002. 108. Cong, S., T., Di, X., Tian, C., Zhao, M., Wang, K. (2020). Diagnostic value of neutrophil CD64, pro- calcitonin, and interleukin-6 in sepsis: a meta- analysis. BMC Infectious Diseases, 21. https: //doi. org/10. 1186/s12879-021-06064-0. 109. Relvas, M., Mendes-Frias, A., Gonçalves, M., Salazar, F., López-Jarana, P., Silvestre, R., Da Costa, A. (2024). Salivary IL-1β, IL-6, and IL-10 Are Key Biomarkers of Periodontitis Severity. Interna- tional Journal of Molecular Sciences, 25. https: //doi. org/10. 3390/ijms25158401. 110. Teles, F., Chandrasekaran, G., Martin, L., Patel, M., Kallan, M., Furquim, C., Hamza, T., Cucchiara, A., Kantarcı, A., Urquhart, O., Sugai, J., Giannobile, W. (2024). Salivary and serum inPlammatory bi- omarkers during periodontitis progression and af- ter treatment. Journal of Clinical Periodontology, 51, 1619 - 1631. https: //doi. org/10. 1111/jcpe. 14048. 111. Kinney, J., Morelli, T., Oh, M., Braun, T., Ram- seier, C., Sugai, J., Giannobile, W. (2014). Crevicular Pluid biomarkers and periodontal disease progres- sion. Journal of clinical periodontology, 41 2, 113- 120. 112. Safari S, Baratloo A, ElPil M, Negida A. Evi- dence-Based Emergency Medicine; Part 5 Receiver Operating Curve and Area under the Curve. Emerg (Tehran). 2016 Spring;4(2):111-3. PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 10 Figure 1. Flowchart of the study. Table 1. Demographic variables of the study population. Total (N=120) Control (n=30) Periodontitis (90) p-value Sex Male N (%) 59(49.2%) 9(27.6%) 50(56.2%) <0.001* Female N (%) 61(50.8%) 21 (72.4) 40(43.8%) Mean ±SD 26.93± 4.14 42.79±12.39 <0.001** Minimum 21 18 Maximum 38 75 M: male, F: female; N: number; *Comparison done by chi-square; level of significance ≤ 0.05; Sig: significant. ** Independent t-test, Sig: signif- icant; SD: standard deviation; Min: minimum; Max: maximum. Table 2. Periodontitis stages distribution. Periodontitis Stage I Stage II Stage III Stage IV N 2 5 46 37 % 2.2% 5.5% 51.1% 41.1% PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 11 A B C D Figure 2. Comparisons of clinical periodontal parameters in periodontal health and disease. A) The PLI means between control Vs periodontitis and control Vs stage III, IV. B) The BOP means between control Vs periodontitis and control Vs stage III, IV. C) The PPD mean between stage III Vs IV. D) The CAL mean between stage III Vs IV. Contro l Peri odontiti s Stag e I II Stag e I V 0 50 100 150 M ea n of P LI ✱ ✱ ✱ ns 29.68 96.47 95.52 97.31 Contro l Peri odontiti s Stag e I II Stag e I V 0 20 40 60 80 100 M ea n of B O P ✱ ✱ ✱ ✱ 6.62 81.73 75.26 91.81 Stage III Stage IV 0 2 4 6 Me an of P PD ✱ 4.39 4.93 Stag e I II Stag e I V 0 2 4 6 M ea n of C AL ✱ 3.31 5.09 PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 12 Table 3. RANKL (pg\ml) among the study groups. Biomarker Groups Mean ±SD Min Max Comparison p-value RANKL (pg\ml) Control 125.62 53.92 87.54 353.08 Periodontitis 261.61 96.27 126.95 914.55 C VS P* <0.001 Stage III 263.93 118.01 126.95 914.55 C VS PS** <0.001 Stage IV 256.91 72.03 160.64 506.13 Abbreviations: C: control, P: periodontitis, PS: periodontitis stage, S-III, IV: stage III, IV *comparison using Mann-Whitney test **; Comparison using Kruskal-Wallis test; level of significance ≤ 0.; SD: Standard deviation; Sig: Signif- icant Table 4. IL-36γ (pg\ml) among groups. Biomarker Groups Mean ±SD Min Max Comparison p-value IL-36γ (pg\ml) Control 138.18 50.68 33.11 318.11 Periodontitis 340.77 90.61 169.56 753.86 C VS P * <0.001 Stage III 329.72 82.91 169.56 658.18 C VS PS** <0.001 Stage IV 359.53 98.61 254.34 753.86 Abbreviations: C: control, P: periodontitis, PS: periodontitis stage, S-III, IV: stage III, IV *comparison using Mann-Whitney test **; Comparison using Kruskal-Wallis test; level of significance ≤ 0.; SD: Standard deviation; Sig: Signif- icant. Table 5. IL-38 (pg\ml) among groups. Biomarker Groups Mean ±SD Max Min Comparison p-value IL-38 (pg\ml) Control 142.80 28.53 55.92 175.84 Periodontitis 63.52 21.24 20.80 151.76 C VS P* <0.001 Stage III 61.63 20.42 20.80 129.89 C VS PS ** <0.001 Stage IV 65.25 22.82 29.77 151.76 Abbreviations: C: control, P: periodontitis, PS: periodontitis stage, S-III, IV: stage III, IV. *comparison using Mann-Whitney test **; Comparison using Kruskal-Wallis test; level of significance ≤ 0.05; SD: Standard deviation; Sig: Sig- nificant PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 13 Table 6. Correlation between clinical periodontal variables (%PI, % BOP, PPD, and CAL) with salivary biomarkers (RANKL, IL-36γ, and IL-38) levels in the unstable generalized periodontitis group. Variables RANKL IL-36γ IL-38 R p-value R p-value R p-value PI 0.078 0.468 -0.110 0.304 0.137 0.2 BOP -0.023 0.828 -0.001 0.992 -0.121 0.257 PPD 0.052 0.631 0.178 0.24 0.036 0.739 CAL -0.080 0.46 0.095 0.024 -0.100 0.299 Abbreviations: RANKL, Receptor activator of nuclear factor kappa-Β ligand; IL-36γ, Interleukin 36γ; IL-38, Interleukin-38; BOP, bleeding on probing; PLI, plaque index; NS, non-significant; sig, significant; r, Spearman’s correlation coefficient Table 7. Correlation between clinical periodontal parameters with RANKL, IL-36, and IL-38 salivary levels in periodontitis stage III and stage IV. Periodontitis severity variables RANKL IL-36γ IL-38 R p-value R p-value R p-value Stage III PLI 0.199 0.191 -0.383 0.009 0.223 0.136 BOP 0.054 0.723 -0.154 0.024 -0.137 0.362 PPD -0.005 0.975 -0.100 0.507 -0.089 0.558 CAL -0.167 0.273 0.201 0.18 -0.203 0.176 Stage IV PLI -0.067 0.697 0.273 0.103 0.127 0.452 BOP 0.280 0.003 -0.131 0.441 -0.184 0.276 PPD 0.216 0.2 0.171 0.31 0.174 0.304 CAL 0.119 0.483 0.171 0.311 -0.014 0.933 Abbreviations: RANKL, Receptor activator of nuclear factor kappa-Β ligand; IL-36γ, Interleukin-36γ; IL-38, Interleukin-38; PPD, probing pocket depth; CAL, clinical attachment level; NS, non-significant; sig, significant; r, Spearman’s correlation coefficient PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 14 Table 8. Correlation between RANKL, IL-36γ, and IL-38 in control and periodontitis groups. Interaction RANKL IL 38 R p-value R p-value Control IL36γ 0.671 0.001 -0.476 0.012 Periodontitis IL 36γ 0.560 0.001 0.632 0.001 Control IL 38 -0.773 0.001 - - Periodontitis IL38 0.482 0.001 - - Abbreviations: RANKL, Receptor activator of nuclear factor kappa-Β ligand; IL-36γ, Interleukin 36γ; IL-38, Interleukin-38; sig, significant; r, Spearman’s correlation coefficient. A) B) C) D) Figure 3. ROC curves of RANKL, IL-36, and IL-38, (A) Control Vs. Periodontitis (B) Control Vs. Stage III, (C) Control Vs. Stage IV (D) Stage III Vs. Stage IV. 0 20 40 60 80 100 0 20 40 60 80 100 ROC curve 100% - Specificity% Se ns iti vi ty % RANKL IL-3 8 IL-36 Reference line 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 y 8 0 20 40 60 80 100 0 20 40 60 80 100 ROC curve 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 ROC curve 100% - Specificity% S en si tiv ity % RANKL IL-38 IL-36y Reference line 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 ROC curve 100% - Specificity% Se ns iti vi ty % RANKL IL-36y IL-38 Reference line 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 ROC curve 100% - Specificity% Se ns iti vi ty % RANKL IL-36 y L-38 Reference line PotenFal of Salivary IL-36, IL-38, and RANKL in DifferenFaFng Periodontal Health from PeriodonFFs Vol 13, No 1 (2025) DOI 10.5195/d3000.2025.1008 http://dentistry3000.pitt.edu 15 Table 9. Sensitivity, specificity, and cut-off points of RANKL, IL-36, and IL-38. Test Result Varia- ble(s) AUC P-value Optimal cutoff point %Sensitivity %Specificity Control VS Periodon- titis RANKL 0.944 Excellent <0.0001 143.8 98.88 93.10 IL-36 0.982 Excellent <0.0001 196.9 98.88 96.55 IL-38 0.960 Excellent <0.0001 67.05 65.17 96.55 Control VS Stage III RANKL 0.942 Excellent <0.0001 143.8 97.85 93.10 IL-36 0.980 Excellent <0.0001 196.9 97.83 96.55 IL-38 0.967 Excellent <0.0001 66.69 71.74 96.55 Control VS Stage IV RANKL 0.947 Excellent <0.0001 161.3 94.59 93.10 IL-36 0.986 Excellent <0.0001 256.8 97.30 96.55 IL-38 0.949 Excellent <0.0001 111.9 97.30 93.10 Stage III VS Stage IV RANKL 0.532 Fail 0.614 238.5 54.05 52.17 IL-36 0.580 Fail 0.209 320.7 56.76 54.35 IL-38 0.540 Fail 0.527 57.64 51.53 47.83 AUC: area under the curve, %: percentage. Based on a rough classiFication system, AUC can be interpreted as follows: 90 -100 = excellent; 80 - 90 = good; 70 - 80 = fair; 60 - 70 = poor; 50 - 60 = fail [112].