Biomarkers in biological fluids in adults with periodontitis and/or obesity: A meta-analysis

Abstract

Obesity and periodontal diseases have been investigated to be interconnected, but the molecular mechanism underlying this association is still not clear. The aim of this systematic review is to assess the association of serum, salivary and gingival crevicular fluid (GCF) inflammatory markers (IMs), obesity, and periodontitis. Studies that evaluated IM of adults according to obesity status (O) and periodontitis status (P) (O+P+; O-P+; O+P-) were screened on several electronic databases and grey literature up until February 2019. Risk of bias assessment and level of evidence were evaluated through Fowkes and Fulton scale and Grading of Recommendations Assessment, Development and Evaluation (GRADE). Meta-analyses were grouped according to the biological matrix studied (serum/GCF) and groups (O+P+ vs. O−P+/O+P+ vs. O+P−). Out of the 832 studies screened, 21 were considered in qualitative synthesis and 15 in quantitative synthesis (meta-analysis). Although included studies showed mostly “no” or “minor” problems during the quality assessment, GRADE assessment indicated very low to moderate level of evidence based on the question answered. O+P+ adults exhibited significantly higher serum levels of C-reactive protein (CRP), interleukin 6 (IL-6), leptin, and tumor necrosis factor-α (TNF-alpha) and higher resistin GCF levels than O−P+. O+P+ adults showed significantly higher serum levels of IL-6 and leptin and lower adiponectin serum levels than O+P−. Only qualitative information could be obtained of the IM vaspin, omentin-1, chemerin, IL-10, progranulin, MCP-4, IL-1β, and interferon-γ (IFN-γ). Obesity and periodontitis, together or separately, are associated with altered serum and GCF levels of CRP, IL-6, leptin, TNF-alpha, adiponectin, and resistin. It was not possible to evaluate the association between obesity and periodontitis at salivary levels. The role of recently investigated biomarkers as vaspin, omentin-1, chemerin, IL-10, progranulin, MCP-4, IL-1β, and IFN-γ, which can be key points underlying the association between obesity and periodontitis, remains to be further investigated.

INTRODUCTION

The obesity global epidemy is a serious public health concern, because obesity is a risk factor to increased morbidity and mortality in cardiovascular diseases (CVDs), diabetes, cancers, and other chronic diseases.[1] Gingivitis and periodontitis are infectious inflammatory diseases whose pathogenesis may be affected by environmental factors and systemic disorders, such as diabetes, smoking, and obesity.[2]

The relationship between obesity and periodontal diseases has been investigated in previous clinical studies and systematic reviews.[3,4] Overall evidence has demonstrated positive consistent associations between parameters of obesity and clinical parameters of periodontitis – enough to postulate a pattern of increased risk of periodontitis in individuals with overweight or obesity.[4] However, the underlying biological mechanisms linking both diseases are not fully understood so far.[3,4]

Obesity negatively affects the immune response by increasing the susceptibility to infections.[5] Adipose tissue has emerged as an active participant in the regulation of several pathologic processes by means of the release of several cytokines that are involved in inflammatory processes.[6] Based on these concepts, some mechanisms have been proposed to explain the association of obesity and periodontal diseases. Obesity was associated with increased levels and proportions of periodontal pathogens in subgingival biofilm of patients with periodontitis and periodontal health.[7,8] Furthermore, increased body mass index (BMI) was related to altered levels of inflammatory mediators in gingival crevicular fluid (GCF) and serum of patients with periodontitis and periodontal health, whereas periodontitis per se seems to affect the circulatory levels of some adipose tissue-derived mediators.[9,10,11]

A previous systematic review investigated the cytokine profile in the GCF of adults and adolescents with periodontitis with and without obesity. It was concluded that the periodontal inflammation seems to have a greater influence than obesity on the levels of biomarkers in GCF.[12] However, a detailed analysis of the results from studies comparing the levels of inflammatory markers (IMs) in different biological fluids, such as GCF, serum, and saliva, in adults with obesity and periodontitis (O+P+), without obesity and with periodontitis (O−P+), and with obesity and without periodontitis (O+P−) has not been compiled so far. This distinction is important to better understand the underlying mechanisms between obesity and periodontitis. Therefore, the aim of this systematic review is to answer the following question: “Do adults with only obesity or only periodontitis or with both conditions differ in relation to the levels of IM in serum, saliva, or GCF?” The current investigation would provide an overview of the current status of the evidence on the impact of obesity and periodontitis on the IM underlying the pathogenesis of both diseases.

MATERIALS AND METHODS

The present systematic review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis checklist.[13]

Protocol and registration

This systematic review is registered at PROSPERO (CRD42018101943).

Study design

The general review question was divided in two subquestions to better clarify which IM is affected by either or both conditions simultaneously:

Subquestion 1: Is obesity associated with altered levels of IMs in adults with periodontitis?

Evaluation of papers that compared the levels of IM between O+P+ and O−P+ adults.

Subquestion 2: Is periodontitis associated with altered levels of IMs in adults with obesity?

Evaluation of papers that compared the levels of IM between O+P+ and O+P− adults.

Eligibility criteria

Inclusion criteria

1. Observational cross-sectional and longitudinal studies in adults that compared the levels of IM between O+P+ and O−P+ (Question 1) and between O+P+ and O+P− (Question 2) in serum, saliva, and/or GCF. Only baseline data of the longitudinal interventional studies were assessed and included

2. Studies that clearly presented a definition of periodontitis and assessed clinical parameters such as marginal bleeding, bleeding on probing, suppuration, probing depth, and clinical attachment level, based on the standards of the International World Workshop for Classification of Periodontal Disease and Conditions[18]

3. Studies that clearly defined obesity ass BMI ≥ 30 and <40 kg/m2, based on anthropometric measurements, including weight (kg), height (m), by means of body mass index (BMI), according to World Health Organization.[10]

Exclusion criteria

1. Studies in children or adolescents

2. Studies that included subjects with known systemic diseases (diabetes mellitus, cancer, hypertension, and hypothyroidism, among others) and recent history of antibiotic and anti-inflammatory therapy or periodontal treatment and studies that included pregnant or lactating women and smokers

3. In vitro and animal studies

4. Reviews, case reports, letters, and personal opinions

5. Studies without at least two of the three investigated groups (O+P+, O−P+, or O+P−).

Information sources

Literature search was performed up to February 6, 2019. Electronic databases consulted were the following: the Cochrane, Lilacs, PubMed (including MedLine), Scopus, and gray literature (OpenGrey, Google Scholar, and ProQuest).

Search strategy

Full strategy search according to each database is shown in Appendix 1. No date and language restrictions were applied. EndNote Web (Clarivate Analytics, USA) was used to identify and eliminate duplicate studies.

Study selection

Titles and abstracts of potential articles were screened independently by two reviewers (RSB and GSZ). Lists of selected studies were compared, and when necessary, a third reviewer was consulted (PMD) to make the final decision. After the initial search and selection phase, assessment of the full articles was performed. The reviewers also performed a hand-search in the reference list of all included studies. Potential full copies of articles to be included in this review were thereafter independently inspected by the same reviewers. No language restriction was applied.

Data collection

One author (RSB) collected the necessary information from the selected articles using predefined data collection worksheets. The second author (GSZ) cross-checked all obtained information and confirmed its accuracy. When needed, the authors of the included studies were contacted.

Risk of bias in individual studies

Quality assessment and risk of bias control were performed according to the guidelines described by Fowkes and Fulton.[14] It allows the classification of cross-sectional, cohort, controlled trial, and case–control studies. Questions were applied by two reviewers (RSB and GSZ). When evaluating criteria for each question, the importance of failures or missing information in relation to their expected effect on the results was scored as “major” (++) problem, “minor” (+) problem, “no problem” (0), or not applicable “NA.”

Summary measures

The levels of IM in serum, saliva, and GCF were considered the main outcomes. Comparable outcomes were standardized to the same measurement unit previously to meta-analysis procedures, when necessary.

Synthesis of the results

Results were separated according to the type of IM, type of biological fluids, and subquestions (subquestion 1/subquestion 2). Random effects meta-analysis was performed using Comprehensive Meta-Analysis V3 software (Biostat, USA). Statistical heterogeneity was calculated by inconsistency indexes (I²) and a value >90% was considered an indicator of substantial heterogeneity between studies.[15] The significance level was set at 5.0%.

Risk of bias across studies

The level of evidence (certainty) was assessed employing the Grading of Recommendations Assessment, Development and Evaluation (GRADE) for results included in meta-analyses. Risk of bias, inconsistency, directness, and imprecision were factors considered.

RESULTS

Study selection

From an initial amount of 832 articles, 571 were considered for abstract reading after eliminating duplicates and 5 papers were identified through hand-search. A total of 44 articles were screened thereafter for full-text review. Of the 44 articles evaluated, 23 studies did not fulfill the inclusion criteria and were later excluded [reasons for exclusion in Appendix 2]. Hence, 21 studies were included in the qualitative synthesis. All 21 included studies answered the subquestion 1 Table 1, whereas only 10 papers answered the subquestion 2 [Table 2]. Of the 21 studies that answered subquestion 1, 15 were included in meta-analysis, whereas of the 10 studies that answered subquestion 2, only 7 were included in meta-analysis. Concerning the quantitative synthesis, only studies including serum and GCF showed comparable results to be included in meta-analysis for both questions. Data about salivary levels of IM were not comparable between studies (e.g., different IM analyzed) and therefore salivary levels of IM were not included in meta-analysis. Flow diagram is shown in Figure 1.

Table 1.

Main characteristics and outcomes of included studies that answered the Subquestion 1 (comparison of serum, saliva, and/or gingival crevicular fluid levels of inflammatory markers between subjects with obesity and periodontitis O+P+ and subjects without obesity with periodontitis O−P+)

Source	Country and place of recruitment	Study design	O+P + n	O−P + n	Mean age or range	Periodontitis definition	Type of sample, inflammatory markers evaluated, method of assay, and kit brand employed	Main outcomes	Observations
1. Al-Zahrani and Alghamdi 2012[16]	Kingdom of Saudi Arabia, University clinic	Prospective interventional	20	20	43.7±8	≥3 mm of CAL at ≥30 of the sites	Serum CRP levels: ELISA Kit: Information not available	Serum CRP level was significantly higher in the O+P + group, compared to the O−P + group (0.96 vs. 0.60; P=0.001)	Baseline values included in meta-analysis
2. Balli et al., 2016[17]	Turkey, University clinic	Prospective interventional	19	19	O+P+: 42 O−P+: 40	Standards of the international world workshop for classification of periodontal disease and conditions[18]	GCF Vaspin and Omentin-1: ELISA, Hangzhou Eastbiopharm (China); TNF-α: ELISA, Boster Biological Technology (USA)	O+P + subjects showed higher levels of vaspin and TNF-α (P<0.008), and lower levels of omentin-1 levels (P<0.008), compared to O−P + subjects	Not included in meta-analysis (just Spearman’s rank correlation values of inflammatory markers are shown)
3. Balli et al., 2013[19]	Turkey, University clinic	Prospective; Interventional	20	20	O+P+: 40.56±4.11 O−P+: 39.67±3.87	Standards of the international world workshop for classification of periodontal disease and conditions[18]	GCF Chemerin: ELISA, Hangzhou Eastbiopharm (China);IL-6: ELISA, Boster Biological Technology (USA)	O+P + group showed higher GCF levels of chemerin and IL-6, compared to O−P + group (P<0.05 and P<0.008 respectively)	Not included in meta-analysis (just Spearman’s rank correlation values of inflammatory markers are shown)
4. Boyapati et al., 2018[20]	India, University clinic	Cross-sectional	25	25	O+P+: 54.92±6.82 O−P+: 43.24±8.96	Standards of the international world workshop for classification of periodontal disease and conditions[18]	Serum Leptin: ELISA Adiponectin: ELISA Kit: information not available	Leptin serum levels were statistically higher at O+P + group when compared to O−P + group (26.33 vs. 1.92), while serum adiponectin levels were statistically lower at O+P + group (11.92 vs. 16.96) (P<0.001)	Included in meta-analysis
5. Duzagac et al., 2016[21]	Turkey, University clinic	Prospective, interventional	15	15	O+P+: 40.66±7.98 O−P+: 41.06±7.10	PD ≥4 mm in ≥30% of periodontal sites; BOP in ≥50% of sites; Interproximal CAL >2mm in ≥20% of sites; radiographic evidence of bone loss;	Serum and GCF Adiponectin: ELISA, Human Adiponectin ELISA kit (Adipogen, Switzerland); IL-6, IL-10, CRP and TNF alpha: ELISA (Invitrogen, USA)	GCF levels of adiponectin, IL-6, IL-10 and TNF-α did not differ between O+P + and O−P + groups (8.54 vs. 10.07 mg/L; 8.28 vs. 7.27 pg/mL; 6.53 vs. 7.15 pg/mL; 13.84 vs. 13.09 pg/mL respectively; P>0.05) Serum levels of adiponectin, IL-10 and CRP did not differ between O+P + and O−P + groups (6.39 vs. 8.34mg/L; 6.47 vs. 3.22pg/mL; 0.63 vs. 0.27 mg/L respectively, P>0.05). Serum levels of IL-6, and TNF-α were higher for O+P + group than for O−P + group (3.03 vs. 1.59 pg/mL; 45.52 vs. 4.53 pg/mL respectively, P<0.05)	Baseline inflammatory markers values included in meta-analysis; with exception of IL-10 (only paper to study this biomarker).
6. Gonçalves et al., 2015[22]	Brazil, University clinic	Prospective, interventional	18	21	O+P+: 48.59±5.9 O−P+: 48.4±9.5	>30% of the sites with concomitant PD and CAL ≥4 mm and a minimum of six teeth distributed in the different quadrants presenting at least one site with PD and CAL ≥5 mm and BOP	Serum Adiponectin and leptin: ELISA, Quantikine, R and D Systems, Minneapolis, USA	Leptin serum levels were higher in O+P + group than in O−P + group (441.8 vs. 292.7, P=0.04, pg/mL × 102) while adiponectin levels did not differ between groups (52.5 vs. 44.2 ng/mL × 102; P>0.05)	Just baseline inflammatory markers values included in meta-analysis
7. Guruprasad and Pradeep 2018[23]	India, University clinic	Cross-sectional	10	10	O+P+: 31.40±6.04 O−P+: 31.80±4.96	GI ≥1 mm, PD ≥5 mm, CAL ≥3 mm and radiographic evidence of bone loss at >30% of sites	GCF and serum IL-34: ELISA, RAND D Systems, Minneapolis, USA	GCF levels of IL-34 were not statistically different between O+P + and O−P + groups (969.44 vs. 899.20 pg/mL) (P>0.001). Serum levels also have not shown statistically difference between O+P + and O−P + groups (774.02 vs. 641.76 pg/mL) (P>0.001)	Not included in meta-analysis (only study to include this biomarker)
8. Jentsch et al., 2017[24]	Germany, University clinic	Cross-sectional	25	25	O+P+: 50.3±12.3 O−P+: 46.7±14.7	Standards of the international world workshop for classification of periodontal disease and conditions[22]	GCF, saliva, and serum Acylated and total ghrelin: ELISA (Merck Millipore, Darmstadt, Germany) Chemerin and IL-1β: ELISA (R and D Systems Europe Ltd., Abingdon, UK)	At GCF, there was no statistically difference among acylated ghrelin levels between O+P + and O−P + patients, as well as chemerin levels. Total ghrelin and IL-1B levels were higher in O−P + patients, when compared to O−P + patients (P<0.05) Salivary levels of acylated ghrelin and chemerin have not statistically differed among O+P + and O−P + groups, while total ghrelin and IL-1B were higher in O−P + groups when compared to O+P + groups (P<0.05) Serum levels of IL-1B, total and acylated ghrelin have not differed among O+P + and O−P + groups, while serum chemerin levels were higher in O+P + group when compared to O−P + group (P=0.29)	Not included in meta-analysis (only study to include these biomarkers)
9. Kanoriya et al., 2017[25]	India, University clinic	Cross-sectional	20	20	O+P+: 34.75±6.01 O−P+: 35.10±6.24	Signs of gingival inflammation clinically - PD ≥5 mm, GI >1, CAL ≥3 mm	Serum and GCF Retinol-binding protein 4 (RBP4): Human RBP4 ELISA kit, Ray biotech Inc (USA) Leptin: Human Leptin Elisa kit, Ray biotech Inc (USA)	Both serum and GCF levels of RBP4 and leptin were higher for O+P + than for O−P+ (27.3 vs. 19.75 ng/mL; 375.55 vs. 272.70 pg/mL respectively in serum (P<0.05) and in GCF 23.50 vs. 16.75 ng/mL; 132.00 vs. 33.00 pg/mL (P<0.05) respectively	Included in meta-analyses (leptin). RBP-4 was not included because it was the only paper to evaluate this biomarker
10. Kose et al., 2015[26]	Turkey, University clinic	Cross-sectional	22	22	O+P+: 36.90±4.54 O−P+: 36.63±4.33 (mean±sd)	Two or more tooth sites with PD ≥4 mm or CAL of 4 mm that bleed on probing	Serum and saliva IL-6, TNF-α: ELISA IL-6: Human IL-6 ELISA kit (EK0410) TNF-α: Human TNF-α ELISA kit (EK0525)	Serum and salivary levels of TNF-α did not differ between groups (75.15 vs. 71.40; 50.11 vs. 47.21 pg/mL respectively, P>0.05). Serum and salivary IL-6 levels were higher for O+P + subjects than for O−P + subjects (32.15 vs. 25.05; 22.62 vs. 17.41 pg/mL respectively, P<0.05)	Included in meta-analysis (serum biomarkers). Salivary adipokines were not included because it was the only paper to evaluate those biomarkers in saliva
11. Mendoza-Azpur et al., 2015[27]	Peru, University clinic and clinical practices	Cross-sectional	24	19	O+P+: 43.5±7.3 O−P+: 41.4±6.5	BOP in at least four teeth with a PD >4 mm	Serum Adiponectin, leptin: ELISA TNF-α: fully automated chemiluminescence immunoassay Kit: Information not available	O+P + group showed higher serum levels of leptin and adiponectin compared to O−P+ (13.5 vs. 9.9 pg/mL; 13.6 vs. 10.5 ug/mL respectively P<0.05), while TNF-α levels did not differ between groups (4.6 vs. 4.9 pg/mL, P>0.05)	Included in meta-analysis
12. Patel and Raju 2014[28]	Turkey, Not specified	Cross-sectional	30	30	23-53	Presence of clinical inflammation, GI >1, PD ≥5 mm and CALs ≥3 mm with radiographic evidence of bone loss	GCF and serum Resistin: ELISA, Quantikine human Resistin Immunoassay, RanD Systems, USA	Serum and GCF resistin levels were higher for O+P + group than for O−P + group (16.66 vs. 12.34, P<0.01, and 9.99 vs. 7.47 ng/Ul, P<0.05, respectively)	Included in meta-analysis.
13. Pradeep et al., 2012[29]	Turkey, University clinic	Cross-sectional	10	10	O+P+: 36.87±3.32 O−P+: 35.22±3.11	Clinical signs of gingival inflammation, PD ≥5 mm, CAL ≥3 mm, with radiographic evidence of bone loss and GI >1	GCF and serum Progranulin: ELISA, Adipogen International, South Korea hs-CRP: RANDOX immunoturbidimetrical analyzer	Serum and GCF levels of progranulin (237.6 vs. 182.2; 231.5 vs. 176.6 ng/mL) and CRP (6.49 vs. 3.89 mg/L, 1.14 vs. 0.93) were higher for O+P + group compared to O−P + group (P<0.05)	Included in meta-analysis (serum and GCF values of CRP) Progranulin values were not included in meta-analysis because it was the only paper evaluating this biomarker
14. Pradeep et al., 2013[11]	Turkey, University clinic	Cross-sectional	10	10	O+P+: 31.60±3.81 O−P+: 32.80±4.76	Clinical signs of gingival inflammation, PD ≥5 mm, CAL ≥3 mm, with radiographic evidence of bone loss and GI>1	GCF and serum MCP-4: ELISA, Raybiotech, USA hs-CRP: RANDOX immunoturbidimetrical analyzer	Serum and GCF levels of MCP-4 (274.3 vs. 175.10 pg/mL and 53.40 vs. 34.30 ug/mL respectively) and CRP (6.07 vs. 3.95mg/L and 1.17 vs. 0.99, P<0.05) were higher for O+P + group than for O−P + group (P<0.05)	Included in meta-analysis (serum and GCF values of CRP). MCP-4 values were not included in meta-analysis because it was the only paper evaluating this biomarker
15. Pradeep et al., 2016[30]	Turkey, University clinic	Cross-sectional	10	10	O+P+: 33.80±3.96 O−P+: 35.10±3.98	Clinical signs of gingival inflammation, PD ≥5 mm, CAL ≥3 mm, with radiographic evidence of bone loss and GI >1	GCF Vaspin: ELISA, ABO Swiss, China	GCF concentration of vaspin was higher for O+P + than for O−P+ (1.84 vs. 1.35 P<0.01)	Not included in meta-analysis because it was the only article assessing vaspin with available data for meta-analysis
16. Pradeep et al., 2016[31]	India, University clinic	Cross-sectional	10	10	O+P+: 32.50±5.83 O−P+: 30.90±5.76	PD ≥5 mm, GI >1, CAL ≥3 mm	GCF and tear fluid (not evaluated in this systematic review) Lipocalin-2: ELISA	GCF lipocalin levels were higher in O+P + group than in O−P + group (106.51 vs. 84.32 ug/L) (P<0.05)	Not included in meta-analysis because it was the only article assessing this biomarker
17. Suresh et al., 2016[32]	India, University clinic	Cross-sectional	25	25	O+P+: 38.56±5.378 O−P+: 35.96±5.806	Clinical attachment loss of 3-5 mm at more than 30% of sites PI, GI and CAL	GCF Resistin: ELISA, Raybio Human Resistin ELISA kit	GCF resistin levels were higher for O+P + group than for O−P + group (15.4 vs. 12.74 ng/mL, P<0.01)	Included in meta-analysis
18. Suresh et al., 2018[33]	India, Not specified	Prospective, interventional	30	30	O+P+: 35.67±4.080 O−P+: 37.17±4.764	CAL of ≥3 mm in >30% of sites	GCF and serum Resistin: ELISA	GCF and serum levels of resistin were higher in O+P + patients when compared to O−P + patients (15.055 vs. 12.833 ng/mL), as well as serum levels of resistin (25.832 vs. 19.065 ng/mL) (P<0.05)	Included in meta-analysis
19. Varghese et al., 2018[34]	India, University clinic	Prospective, interventional	100	100	O+P+: 37.3 O−P+: 38.4	Patients with CAL of equal to or more than 3 mm in more than one-third of sites	GCF and serum Resistin: ELISA	GCF and serum levels of resistin were higher in O+P + group, when compared to O−P + group (16.02 vs. 13.10 ng/mL and 26.82 vs. 20.34 ng/mL) (P<0.05)	Not included in meta-analysis (standard deviation not available)
20. Zimmermann et al., 2013[10]	Brazil, University clinic	Cross sectional	20	20	O+P+: 51.5±7.6 O−P+: 47.8±7.7	Presence of ≥30% of the sites with PD and CAL ≥4 mm and ≥noncontiguous teeth with ≥1 site with PD and concomitant CAL ≥5 mm	GCF and serum Resistin, adiponectin, leptin, TNF-α, and IL-6: ELISA, Quantikine, R and D Systems, Minneapolis	Serum concentration of resistin, adiponectin, leptin, TNF-α and IL-6 did not differ between groups (2.3 vs. 2.5 ng/mL × 5; 44.00 vs. 44.2 ng/mL × 100; 426.8 vs. 294.7 pg/mL × 100; 3.3 vs. 3.2 pg/mL; 3.4 vs. 2.6 pg/mL, P>0.05). GCF levels of resistin, adiponectin, leptin and IL-6 did not differ between groups (1.52 vs. 2.04 ng/uL; 1.70 vs. 1.21 ng/uL; 0.71 vs. 0.67 pg/uL; 0.26 vs. 0.59pg/uL, P>0.05), while GCF concentration of TNF-α was higher for O+P + group than for O−P+ (0.51 vs. 0.12 pg/uL) (P<0.05)	Included in meta-analysis (resistin, adiponectin, TNF-α, leptin and IL-6 serum and GCF levels)
21. Zuza et al., 2010[35]	Brazil; University clinic	Prospective, interventional	27	25	O+P+: 45.1±8.6 O−P+: 42.9±7.7	Presence of ≥2 teeth with PD ≥5 mm[22]	Serum IL-1β, IL-6, TNF-α and IFN-γ: ELISA, Bioscience (USA)	Serum levels of IL-1β, IL-6 and TNF-α were higher for the O+P + group compared to O−P+ (2.44 vs. 1.62 pg/mL; 1.84 vs. 1.14 pg/mL; 22.29 vs. 17.42 pg/mL respectively P≤0.05). IFN-γ levels did not differ between groups (0.15 vs. 0.14) pg/mL (P>0.05)	Baseline values included in meta-analysis (IL-6 and TNF-α serum values). IL-1β and IFN-γ serum values were not included because it was the only paper in this systematic review to employ those markers

CRP – C reactive protein; GCF – Gingival crevicular fluid; TNF-α – Tumor necrosis factor; IL-34 – Interleukin-34I; IL-6 – Interleukin-6; IL-10 – Interleukin-10; IL-1β – Interleukin-1β; RBP-4 – Retinol-binding protein-4; PD – Probing depth; CAL – Clinical attachment level; BOP – Bleeding on probing; GI – Gingival index; hs-CRP – High sensitive C reactive protein; MCP-4 – Active Monocyte Chemotactic Protein-4; PI – Periodontal index; IFN-γ – Interferon-γ

Table 2.

Main characteristics and outcomes of included studies that answered the Subquestion 2 (comparison of serum, saliva and/or gingival crevicular fluid levels of adipokines between subjects with obesity and periodontitis O+P+ and subjects with obesity without periodontitis O+P− patients)

Source	Country and place of recruitment	Study design	O+P + n	O+P − n	Mean age or Range	Sample evaluated Inflammatory markers	Main outcomes	Observations
1. Guruprasad and Pradeep 2018[23]	India, University clinic	Cross- sectional	10	10	O+P+: 31.40±6.04 O+P−: 32.00±4.40	Serum and GCF IL-34	Serum and GCF levels were significantly higher in O+P + patients, when compared to O+P − patients (774.02 vs. 184.36 pg/mL and 969.44 vs. 446.18 pg/mL) (P<0.001)	Not included in meta-analysis (only study to evaluate this biomarker)
2. Jentsch et al., 2017[24]	Germany, University clinic	Cross- sectional	15	15	O+P+: 50.3±12.3 O+P−: 38.9±13.4	Serum, saliva, and GCF Total and acylated ghrelin, interleukin-1β, and chemerin	GCF levels of acylated ghrelin have not differed among O+P + and O+P − groups. O+P + group showed higher levels of total ghrelin, as well as of chemerin and lower levels of IL-1B, compared to O−P + group Salivary levels of chemerin, total and acylated ghrelin have not differed among O+P + and O−P + groups, while it was higher for O+P + group, when compared to O−P + group Serum levels of all evaluated biomarkers have not differed among O+P + and O+P − groups (P<0.05)	Not included in meta-analysis (only study to evaluate these biomarkers)
3. Kanoriya et al., 2017[25]	India, University clinic	Cross- sectional	20	15	O+P+: 34.75±6.01 O+P−: 35.33±6.70	Serum and GCF RBP-4 and leptin	Serum and GCF levels of RBP-4 were higher in O+P + than in O+P − patients (27.30 vs. 12.20 ng/mL; 23.50 vs. 9.00 ng/mL), as well as serum leptin (375.55 vs. 178.00 pg/mL) (P<0.05). GCF leptin was higher for O+P − when compared to O+P+ (330.93 vs. 132.00 pg/mL)	Included in meta-analysis
4. Kose et al., 2015[26]	Turkey, University clinic	Cross- sectional	22	22	O+P+: 36.90±4.54 O+P−: 34.63±4.40	Serum and saliva IL-6, TNF-α	Serum and salivary IL-6 levels were higher in O+P + than in O+P − patients (32.14 vs. 26.11 pg/mL and 22.62 vs. 17.91, P<0.05), while TNF-α levels did not differ between groups (75.15 vs. 73.01 pg/mL and 22.62 vs. 17.91 pg/mL) (P>0.05)	Data about serum biomarkers was included in meta-analysis. Salivary levels were not included because it was the only study to evaluate this biological source
5. Mendoza-Azpur et al., 2015[27]	Peru, University clinic	Cross- sectional	24	21	O+P+: 43.5±7.3 O+P−: 42.9±7.8	Serum Adiponectin, leptin, TNF-α	Serum adiponectin (13.6 vs. 17.5 ug/mL, P<0.05) and leptin (13.5 vs. 16.4 pg/mL) levels were lower in O+P + patients than in O+P − individual. Serum TNF-α did not differ between groups (4.6 vs. 5.1 pg/mL, P>0.05)	Included in meta-analysis
6. Pradeep et al., 2012[29]	Turkey, University clinic	Cross- sectional	10	10	O+P+: 36.87±3.32 O+P−: 35.2±3.54	Serum and GCF Progranulin and hs-CRP	Serum and GCF concentrations of progranulin (237.6 vs. 205.9 ng/mL and 231.5 vs. 197.8 ng/mL) and CRP (6.49 vs. 4.03 mg/L and 1.14 vs. 1.00, P<0.05) were higher for O+P + than for O+P − individuals (P<0.05)	Included in meta-analysis (serum and GCF values CRP) Progranulin values were not included in meta-analysis
7. Pradeep et al., 2013[11]	Turkey, University clinic	Cross- sectional	10	10	O+P+: 31.60±3.81 O+P−: 33.20±3.65	Serum and GCF MCP-4 and hs-CRP.	Serum and GCF MCP-4 levels were higher for O+P + group than for O+P − group (274.3 vs. 241.70pg/mL and 53.40 vs. 19.60pg/uL). Serum CRP concentration was higher for O+P + group than for O+P − group (6.07 vs. 3.95 mg/L), P<0.05). GCF concentration of CGP did not differ between groups (1.17 vs. 1.08 mg/L, P>0.05).	Included in meta-analysis (serum and GCF values of CRP) MCP-4 values were not included in meta-analysis (only study to include this biomarker)
8. Pradeep et al., 2016[31]	India, University clinic	Cross sectional	10	10	O+P+: 32.50±5.83 O+P−: 31±5.03	GCF Lipocalin-2	GCF levels of lipocalin-2 were higher for O+P + patients when compared to O+P − group (106.51 vs. 80.9 ug/L) (P<0.05)	Not included in meta-analysis (only study to include this biomarker)
9. Suresh et al., 2016[32]	India, University clinic		25	25	O+P+: 38.56±5.378 O+P−: 31.40±6.325	GCF Resistin	GCF resistin levels were higher for O+P + compared to O+P − patients (15.14 vs. 9.06 ng/mL, P<0.001)	Included in meta-analysis
10. Zimmermann et al., 2013[10]	Brazil, University clinic	Cross sectional	20	18	O+P+: 51.5±7.6 O+P−: 43.2±7.4	GCF and serum Resistin, adiponectin, leptin, TNF-α, and IL-6	Serum resistin levels were higher (2.3 vs. 1.1 ng/mL ×5, (P<0.05) while serum adiponectin and TNF-α levels were lower for O+P + than for O+P − subjects (44.0 vs. 81.6 ng/mL × 100 and 3.3 vs 6.5 pg/mL, P<0.05). Serum leptin and IL-6 levels did not differ between groups (426.8 vs. 479.6 pg/mL × 100 and 3.4 vs. 1.1 pg/mL, P>0.05) GCF resistin and leptin levels were lower for O+P + than for O+P − individuals (1.52 vs. 2.83 ng/uL and 0.71 vs. 4.11 pg/uL, P<0.05). GCF adiponectin, TNF-α and IL-6 did not differ between groups (1.70 vs. 3.40 ng/uL, 0.51 vs. 0.59pg/uL and 0.26 vs. 0.26 pg/uL, P>0.05)	Included in meta-analysis

GCF – Gingival crevicular fluid; IL-34 – Interleukin-34; TNF-α – Tumor necrosis factor; IL-6 – Interleukin-6; IL-1β – Interleukin-1β; RBP-4 – Retinol-binding protein-4; hs-CRP: high sensitive C reactive protein; MCP-4 – Active Monocyte Chemotactic Protein-4

Figure 1.

Flow diagram of literature search and selection criteria

Study characteristics

Study characteristics are summarized in Tables 1 and 2.

Risk of bias within studies

Table 3 presents the results of the risk of bias assessment in the individual studies.

Table 3.

Results of individual studies according to Fowkes and Fulton checklist

Author	Study design appropriate to objetives?		Study sample representative?		Control group acceptable?		Quality of measurements and outcomes?		Completeness?		Distorting influences?
Al-Zahrani and Alghamdi 2012[16]	Prevalence	NA	Source	0	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	0	Size	0	Matching	+	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Balli et al., 2016[17]	Prevalence	NA	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	++	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	0	Size	0	Matching	0	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Balli et al., 2013[19]	Prevalence	NA	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	++	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	0	Size	0	Matching	0	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Boyapati et al., 2018[20]	Prevalence	NA	Source	0	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	+	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	0	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	++
Duzagac et al., 2016[21]	Prevalence	NA	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	0	Size	0	Matching	0	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	0	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Gonçalves et al., 2015[22]	Prevalence	NA	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	0	Size	0	Matching	0	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Guruprasad and Pradeep 2018[23]	Prevalence	0	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Jentsch et al., 2017[24]	Prevalence	0	Source	+	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	++	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Kanoriya et al., 2017[25]	Prevalence	0	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	0	Blindness	0	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Kose et al., 2015[26]	Prevalence	NA	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	++	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	0	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Mendoza-Azpur et al., 2015[27]	Prevalence	0	Source	0	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	+	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Patel, Raju, 2014[28]	Prevalence	0	Source	+	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Pradeep et al., 2012[29]	Prevalence	0	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	++	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Pradeep et al., 2013[11]	Prevalence	0	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	++	Matching	0	Blindness	0	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Pradeep et al., 2016[30]	Prevalence	0	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Pradeep et al., 2016[31]	Prevalence	0	Source	0	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	++	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	0	Blindness	0	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Suresh et al., 2016[32]	Prevalence	0	Source	0	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	++	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	++	Matching	++	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	++
Suresh et al., 2018[33]	Prevalence	NA	Source	+	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	++	Drop outs	NA	Contamination	NA
	Treatment	0	Size	++	Matching	+	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	++
Varghese et al., 2018[34]	Prevalence	NA	Source	+	Definition	0	Validity	+	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	++	Drop outs	NA	Contamination	NA
	Treatment	0	Size	++	Matching	+	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	++
Zimmermann et al., 2013[10]	Prevalence	0	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	NA	Size	0	Matching	++	Blindness	+	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0
Zuza et al., 2010[35]	Prevalence	NA	Source	0	Definition	0	Validity	0	Compliance	NA	Extraneous treatment	NA
	Prognosis	NA	Method	0	Source	0	Reproducibility	0	Drop outs	NA	Contamination	NA
	Treatment	0	Size	0	Matching	0	Blindness	++	Deaths	NA	Changes over time	NA
	Cause	NA	Criteria	0	Characteristics	0	Quality control	0	Missing data	NA	Confounding factors	0
											Distorting reduced by analysis	0

++: Major problem, +: Minor problem, 0: No problem; NA: Not applicable

Results of individual studies

Qualitative results for individual studies are presented in Tables 1 and 2.

Meta-analysis results

Subquestion 1 meta-analysis results

Serum C-reactive protein, interleukin-6, leptin, tumor necrosis factor-α, adiponectin, and resistin in O+P+ and O−P+ groups

While serum adiponectin and resistin levels did not differ significantly between O+P+ and O−P+ groups (P > 0.05), serum C-reactive protein (CRP), interleukin 6 (IL-6), leptin, and tumor necrosis factor-α (TNF-alpha) levels were higher in the O+P+ than in the O−P+ group (P< 0.05) [Figures 2-4].

Figure 2.

Quantitative results of serum adiponectin and C-reactive protein (O+P+ vs. O−P+)

Figure 4.

Quantitative results of serum resistin and tumor necrosis factor-α (O+P+ vs. O−P+)

Figure 3.

Quantitative results of serum interleukin-6 and leptin (O+P+ vs. O−P+)

Gingival crevicular fluid C-reactive protein, tumor necrosis factor-α, adiponectin, resistin, interleukin-6, and leptin in O+P+ and O−P+ - groups

The GCF levels of most of the IM did not differ between O+P+ and O−P+ groups, including adiponectin, CRP, leptin, TNF-α, and IL-6 (P > 0.05). GCF levels of resistin were higher in the O+P+ than in the O−P+ group (P< 0.05) [Figures 5-7].

Figure 5.

Quantitative results of gingival crevicular fluid adiponectin and C-reactive protein (O+P+ vs. O−P+)

Figure 7.

Quantitative results of gingival crevicular fluid tumor necrosis factor-α and IL-6 (O+P+ vs. O−P+)

Figure 6.

Quantitative results of gingival crevicular fluid leptin and resistin (O+P+ vs. O−P+)

Subquestion 2 meta-analysis results

Serum C-reactive protein, interleukin-6, tumor necrosis factor-α, leptin, and adiponectin in O+P+ and O+P− groups

Serum TNF-α and CRP did not differ between O+P+ and O+P− groups (P > 0.05). Serum IL-6 and leptin levels were higher, whereas serum adiponectin levels were lower in the O+P+ when compared to the O+P− group (P< 0.05) [Figures 8-10].

Figure 8.

Quantitative results of serum adiponectin and C-reactive protein (O+P+ vs. O+P−)

Figure 10.

Quantitative results of serum tumor necrosis factor-α and leptin (O+P+ vs. O+P−)

Figure 9.

Quantitative results of serum IL-6 and leptin (O+P+ vs. O+P−)

Gingival crevicular fluid C-reactive protein, resistin, and leptin in O+P+ and O+P− groups

GCF levels of CRP, leptin, and resistin did not differ between O+P+ and O+P− groups (P > 0.05) [Figures 11 and 12].

Figure 11.

Quantitative results of gingival crevicular fluid C-reactive protein and leptin (O+P+ vs. O+P−)

Figure 12.

Quantitative results of gingival crevicular fluid resistin (O+P+ vs. O+P−)

Confidence in cumulative evidence

GRADE results for subquestion 1 (serum and GCF samples) are presented in [Tables 4 and 5], respectively, whereas for subquestion 2 (serum and GCF samples) are presented in [Tables 6 and 7], respectively, which were included as Supporting Material. Overall quality of evidence varied between “very low” to “moderate.”

Table 4.

GRADE working group grades of evidence for serum inflammatory markers levels subjects with obesity and periodontitis O+P + and subjects without obesity with periodontitis O−P+ (Subquestion 1)

Outcome	Certainty assessment
	Studies (n)	N (O+P+/O−P+)	Risk of bias	Inconsistency	Indirectness	Impression	Other	Grade quality
Adiponectin	5	100/112	Not seriousa	Seriousb	Not serious	Seriousc	Not serious	⨁⨁◯◯ LOW
CRP	4	55/55	Not seriousa	Very seriousb	Not serious	Not serious	Not serious	⨁⨁◯◯ LOW
IL-6	4	84/83	Seriousa	Very seriousb	Not serious	Not serious	Not serious	⨁◯◯◯ VERY LOW
Leptin	5	102/116	Not seriousa	Very seriousb	Not serious	Not serious	Not serious	⨁⨁◯◯ LOW
Resistin	2	50/50	Seriousa	Very seriousb	Not serious	Seriousc	Not serious	⨁◯◯◯ VERY LOW
TNF-α	6	221/168	Seriousa	Very seriousb	Not serious	Not serious	Not serious	⨁◯◯◯ VERY LOW

Legend: ^aRisk of bias evaluated according to Fowkes and Fulton employed parameters [Table 3]; ^bInconsistency evaluated mostly through heterogeneity results (I²) (See Quantitative results); ^c=Imprecision analyzed through confidence interval examination (95%) (See Quantitative results). CRP – C reactive protein; IL-6 – Interleukin-6; TNF-α – Tumor necrosis factor

Table 5.

GRADE working group grades of evidence for gingival crevicular fluid inflammatory markers levels of subjects with obesity and periodontitis O+P+ and subjects without obesity with periodontitis O−P+ (Subquestion 1)

Outcome	Certainty assessment
	Studies (n)	N (O+P+/O−P+)	Risk of bias	Inconsistency	Indirectness	Impression	Other	Grade quality
Adiponectin	2	35/35	Seriousa	Not serious	Not serious	Seriousc	Not serious	⨁⨁◯◯ low
CRP	2	20/20	Seriousa	Not serious	Not serious	Seriousc	Not serious	⨁⨁◯◯ low
IL-6	2	35/35	Seriousa	Seriousa	Not serious	Seriousc	Not serious	⨁◯◯◯ very low
Leptin	2	40/40	Seriousa	Very seriousb	Not serious	Not serious	Not serious	⨁◯◯◯ very low
Resistin	3	75/75	Seriousa	Very seriousb	Not serious	Not serious	Not serious	⨁◯◯◯ very low
TNF-α	2	35/35	Serious	Not serious	Not serious	Seriousc	Not serious	⨁⨁◯◯ low

Legend: ^aRisk of bias evaluated according to Fowkes and Fulton employed parameters [Table 3]; bInconsistency evaluated mostly through heterogeneity results (I²) (See Quantitative results); ^cImprecision analyzed through confidence interval examination (95%) (See Quantitative results). CRP – C reactive protein; IL-6 – Interleukin-6; TNF-α – Tumor necrosis factor

Table 6.

GRADE working group grades of evidence for serum inflammatory markers levels of subjects with obesity and periodontitis O+P+ and subjects with obesity without periodontitis O+P− patients (Subquestion 2)

Outcome	Certainty assessment
	Studies (n)	N (O+P+/O+P−)	Risk of bias	Inconsistency	Indirectness	Impression	Other	Grade quality
Adiponectin	2	39/39	Seriousa	Very seriousb	Not serious	Seriousc	Not serious	⨁◯◯◯ Very low
CRP	2	20/20	Seriousa	Very seriousb	Not serious	Serious	Not serious	⨁◯◯◯ Very low
IL-6	2	42/40	Seriousa	Not seriousb	Not serious	Not seriousc	Not serious	⨁⨁⨁◯ Moderate
Leptin	3	59/54	Not seriousa	Very seriousb	Not serious	Not seriousc	Not serious	⨁⨁◯◯ Low
TNF-α	4	157/84	Seriousa	Not serious	Not serious	Seriousc	Not serious	⨁⨁◯◯ Low

^aRisk of bias evaluated according to Fowkes and Fulton employed parameters [Figure 2]; ^bInconsistency evaluated mostly through heterogeneity results (I²) (See Quantitative results); ^c=Imprecision analyzed through confidence interval examination (95%) (See Quantitative results). CRP – C reactive protein; IL-6 – Interleukin-6; TNF-α – Tumor necrosis factor

Table 7.

GRADE working group grades of evidence for gingival crevicular fluid adipokine levels of subjects with obesity and periodontitis O+P+ and subjects with obesity without periodontitis O+P − patients (Subquestion 2)

Outcome	Certainty assessment
	Studies (n)	N (O+P+/O+P−)	Risk of bias	Inconsistency	Indirectness	Impression	Other	Grade quality
CRP	2	20/20	Seriousa	Not serious	Not serious	Seriousc	Not serious	⨁⨁◯◯ low
Leptin	2	40/33	Seriousa	Very seriousb	Not serious	Seriousc	Not serious	⨁◯◯◯ very low
Resistin	2	45/43	Seriousa	Very seriousb	Not serious	Seriousc	Not serious	⨁◯◯◯ very low

^aRisk of bias evaluated according to Fowkes and Fulton employed parameters [Table 3]; ^bInconsistency evaluated mostly through heterogeneity results (I²) (see quantitative results); ^cImprecision analyzed through confidence interval examination (95%) (see quantitative results). CRP – C reactive protein

DISCUSSION

Although the underlying mechanisms on the link between obesity and periodontitis remain to be completely elucidated, it is suggested that a critical role of inflammatory mediators including adipokines, cytokines, and chemokines can be facilitated by increased adipose and inflamed periodontal tissues.[5,6] The current systematic review compiled data of 21 studies evaluating the serum, GCF, and salivary levels of IM in subjects presenting either obesity or periodontitis or both simultaneously. Comparisons were made in obese subjects without periodontitis and nonobese subjects with periodontitis. Classical IM, including CRP, IL-6, and TNF-α, were the most investigated biomarkers, whereas several newly discovered IM (e.g., IL-34, omentin, vaspin, and chemerin) were poorly explored, precluding a more thorough evaluation by means of meta-analysis. Serum and GCF were the most studied biological matrices, whereas only two studies analyzed the IM in saliva.[24,26] Probably, the low quantity of studies evaluating salivary levels of biomarkers underlying the association among periodontitis and obesity is justified because saliva is a contaminated and less trustworthy source. Overall qualitative findings and meta-analysis demonstrated significant higher circulatory and local levels of pro-IMs in subjects with both obesity and periodontitis (O+P+ group), when compared to subjects presenting periodontitis only or obesity only. Together, these findings suggest that both conditions may act as cofactors of the inflammatory process, favoring a pro-inflammatory profile.

The subquestion 1 Table 1 focused on the comparison of IM levels between adults affected by periodontitis differing by the status of obesity. According to meta-analysis, four IMs with pro-inflammatory biological activities (CRP, IL-6, TNF-α, and leptin) were significantly elevated in serum of O+P+ compared to O−P+ adults. It supports the classical concept that obesity may induce a shift toward a systemic pro-inflammatory profile, resulting in the development of a chronic, low-grade inflammatory state, which can contribute to the pathogenesis of several comorbidities and complications.[36] In GCF, the levels of resistin were significant higher in O+P+ than in O−P+ subjects. Resistin is an adipocyte- and monocyte-derived cytokine associated with insulin resistance, which plays important roles in amplifying the inflammatory state related to several diseases, including metabolic syndrome, Type 2 diabetes mellitus, and CVDs.[37] Previous investigations have shown higher levels of resistin in patients with periodontitis when compared to periodontally healthy controls, suggesting resistin as a possible marker of periodontal diseases.[28,38] Herein, we noticed that periodontal resistin levels are even more increased in the presence of obesity, being a possible pathway of exacerbation of periodontal breakdown in those patients. It seems that obesity yields negative impact at systemic and periodontal levels by increasing the levels of some pro-inflammatory IMs.

When taking into account subquestion 2 Table 2, the focus was the impact of periodontitis on the levels of IMs in subjects with or without obesity. This subject was less explored than Question 1, given the more limited number of papers included on this topic. Of the five serum IMs assessed in meta-analysis, serum levels of the pro-IMs IL-6 and leptin were significantly higher, whereas adiponectin, an anti-inflammatory adipokine,[39] was lower in O+P+ than in O+P− subjects. Those findings support the notion that periodontitis might intensify the systemic pro-inflammatory state of subjects with obesity, further increasing the risk systemic diseases.[40] Noteworthy, it was shown that the levels of IL-6 and leptin seem to be significantly higher in serum of subjects with periodontitis and obesity when compared to those with obesity or periodontitis alone. IL-6 is a cytokine enrolled in chronic and acute inflammatory states, as well as in maturation of B-cells and in vascular damage. It is also associated with increased susceptibility of diseases such as diabetes mellitus and rheumatoid arthritis.[41,42] Leptin plays a dual role as a hormone and as a cytokine, affecting endocrine functions, bone metabolism, energy homeostasis, and of inflammatory responses.[43] Therefore, it appears that the interaction of obesity and periodontitis increases the levels of some IM to a greater extent than when these conditions act separately, suggesting that obese patients with periodontitis may be at an even increased risk for systemic complication related to systemic inflammatory burden.

The current systematic review indicated from “very low” to “moderate” certainty levels of evidence through GRADE evaluation.[15,44,45] This is attributed to confounders that were also evaluated through Fowkes and Fulton checklist, such as, mismatching of age and sex, different brands of kits employed, and lack of blindness during biological fluid collection or laboratory analysis. Another confounder factor that should be mentioned is the lack of standardization on the time of body fluid collection, because some biomarkers might be influenced by circadian variation in several conditions.[46]

It is important to highlight that just cross-sectional data were analyzed, which provides only evidence of association on the impact of obesity and/or periodontitis on IM levels. Longitudinal data could not be included in the present analysis because it is unethical to expose individuals to obesity and/or periodontitis. Therefore, randomized clinical trials evaluating this relationship focus on the evaluation of periodontal treatment impact on IM levels of individuals with obesity,[22,35] which was not analyzed by this systematic review. Cohort studies were not found.

Another important consideration is that only qualitative analysis could be performed for some biomarkers (i.e., vaspin, omentin-1, chemerin, IL-10, progranulin, MCP-4, IL-1β, and interferon-γ [IFN-γ]) that have been pointed as key adipokines related to obesity and/or to the pathogenesis of periodontal diseases.[47] The actual role of these most recently identified markers on the association between obesity and periodontitis needs to be further evaluated.

Scientific evidence indicates that obesity alters the serum levels of IMs such as CRP, IL-6, leptin, TNF-alpha, adiponectin, and resistin in subjects with periodontitis, whereas periodontitis alters the levels of IMs in subjects with obesity, both favoring a pro-inflammatory profile. In individuals having both conditions, the systemic inflammatory profile appears to be increased. The current evidence also indicates that resistin concentration is increased in the GCF periodontal sites of obese subjects with periodontitis than in those of nonobese subjects. Additional studies are necessary to estimate the impact of obesity and/or periodontitis on salivary IM levels. The identified certainty levels ranged from “very low” to “moderate,” implying that future-related research might significantly change the direction and strength of the identified differences.

Thus, the following recommendations are listed for future research:

1. Inclusion of a greater number of patients, because the longitudinal studies cannot be performed

2. Inclusion of a greater number of biomarkers, which can be key points on the association among periodontitis and obesity, such as vaspin, omentin-1, chemerin, IL-10, progranulin, MCP-4, IL-1β, and IFN-γ, in such a way that the inflammatory cascade of both diseases can be extensively understood

3. Standardization of body fluid collection time

4. Employment of international recognized classification for both conditions, that is, obesity and periodontitis.

CONCLUSION

Obesity alters the serum levels of specific IMs in subjects with periodontitis, while periodontitis alters the levels of IMs in subjects with obesity, both favoring a pro-inflammatory profile. In individuals having both conditions the systemic inflammatory profile appears to be increased. The identified certainty levels ranged from “very low” to “moderate” and future research might change the direction and strength of identified differences.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Appendix 1

Search Strategy from different databases

Database	Search
PubMed	“Periodontal Diseases”[Mesh: NoExp] OR “Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Chronic Periodontitis”[Mesh] OR “Periodontitis”[Mesh] OR “Periodontitis” OR “Periodontitides” AND “Body Weights and Measures”[Mesh] OR “Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip Ratio” OR “Waist-Hip Ratios” OR “Waist-to-Hip Ratio” OR “Waist-to-Hip Ratios” OR “Body Size”[Mesh] OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity”[Mesh: NoExp] OR “Obesity” OR “Obesities” OR “Obesity, Abdominal”[Mesh] OR “Obesity, Morbid”[Mesh] OR “obese” OR “obeses” OR “Overweight”[Mesh: NoExp] OR “Overweight” OR “Overweights” OR “Overnutrition”[Mesh] OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight”[Mesh] OR “Ideal Body Weight” OR “Ideal Body Weights”
	AND “Adipokines”[Mesh] OR “Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein”OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin”OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers”[MeSH Terms] OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines”[Mesh] OR “Cytokines”
Scopus	TITLE-ABS-KEY(“Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Periodontitis” OR “Periodontitides” ) AND TITLE-ABS-KEY(“Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip Ratio” OR “Waist-Hip Ratios” OR “Waist-to-Hip Ratio” OR “Waist-to-Hip Ratios” OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity” OR “Obesities” OR “obese” OR “obeses” OR “Overweight” OR “Overweights” OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight” OR “Ideal Body Weights” OR “NormalBody Weight” OR “Normal Body Weights” OR “Ideal Body Mass” ) AND TITLE-ABS-KEY(“Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein”OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin”OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines”)
Web of Science	“Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Periodontitis” OR “Periodontitides” AND “Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip Ratio” OR “Waist-Hip Ratios” OR “Waist-to-Hip Ratio” OR “Waist-to-Hip Ratios” OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity” OR “Obesities” OR “obese” OR “obeses” OR “Overweight” OR “Overweights” OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight” OR “Ideal Body Weights” OR “NormalBody Weight” OR “Normal Body Weights” OR “Ideal Body Mass” AND “Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein”OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin”OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines”
Cochrane	“Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Periodontitis” OR “Periodontitides” AND “Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip Ratio” OR “Waist-Hip Ratios” OR “Waist-to-Hip Ratio” OR “Waist-to-Hip Ratios” OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity” OR “Obesities” OR “obese” OR “obeses” OR “Overweight” OR “Overweights” OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight” OR “Ideal Body Weights” OR “NormalBody Weight” OR “Normal Body Weights” OR “Ideal Body Mass” AND “Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein”OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin”OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines”
ProQuest Dissertations and Theses	all(“Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Periodontitis” OR “Periodontitides” ) AND all(“Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip Ratio” OR “Waist-Hip Ratios” OR “Waist-to-Hip Ratio” OR “Waist-to-Hip Ratios” OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity” OR “Obesities” OR “obese” OR “obeses” OR “Overweight” OR “Overweights” OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight” OR “Ideal Body Weights” OR “Normal Body Weight” OR “Normal Body Weights” OR “Ideal Body Mass” ) AND all( “Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein” OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin” OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines” )
Open Grey	(“Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Periodontitis” OR “Periodontitides” ) AND (“Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip Ratio” OR “Waist-Hip Ratios” OR “Waist-to-Hip Ratio” OR “Waist-to-Hip Ratios” OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity” OR “Obesities” OR “obese” OR “obeses” OR “Overweight” OR “Overweights” OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight” OR “Ideal Body Weights” OR “Normal Body Weight” OR “Normal Body Weights” OR “Ideal Body Mass”) AND (“Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein” OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin” OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines”)
Lilacs and BBO	(tw:(“Periodontal Diseases” OR “Parodontosis” OR “Parodontoses” OR “Periodontitis” OR “Periodontitides” OR “DoençasPeriodontais” OR “Doença Periodontal” OR paradontose OR parodontose OR “Piorreia Alveolar” OR “EnfermedadesPeriodontales” OR “Enfermedad Periodontal” OR paradontosis OR parodontosis OR “Piorrea Alveolar”)) AND (tw:(“Body Measure” OR “Body Measures” OR “body weights” OR “body weight” OR “Body Mass Index” OR “Body Mass Indexes” OR “Quetelet Index” OR “Quetelets Index” OR “Quetelet’s Index” OR “Waist-Hip” OR “Waist-to-Hip” OR “Body Size” OR “Body Sizes” OR “Waist Circumference” OR “Waist Circumferences” OR “Obesity” OR “Obesities” OR “obese” OR “obeses” OR “Overweight” OR “Overweights” OR “Overnutrition” OR “Hypernutrition” OR “Ideal Body Weight” OR “Ideal Body Weights” OR “Normal Body Weight” OR “Normal Body Weights” OR “Ideal Body Mass” OR “Medida corporal” OR “medidascorporais” OR “medidacorpórea” OR “medidascorpóreas” OR “peso corporal” OR “pesos corporais” OR “peso corpóreo” OR “pesos corpóreos” OR “índice de massa corporal” OR “índice de massascorporais” OR “índice de quetelet” OR imc OR “Índice de Massa Corpórea” OR “cintura/quadril” OR “cintura-quadril” OR “tamanho corporal” OR “tamanhoscorporais” OR “circunferência abdominal” OR “circunferênciasabdominais” OR obesidad*or obeso* OR obesa* OR sobrepeso OR “sobre peso” OR sobrenutrição OR “sobrenutrição” OR hipernutrição OR “massa corporal” OR “massascorporais” OR “medidascorporales” OR “pesos corporales” OR “Índice de Masa Corporal” OR “cintura/cadera” OR “cintura-cadera” OR “Tamaño Corporal” OR “TamañosCorporales” OR “CircunferenciasAbdominales” OR hipernutrición OR sobrenutrición OR “Peso Corporales” OR “masa corporal” OR “masascorporales”)) AND (tw:(“Adipokines” OR “Adipokine” OR “Adipocytokines” OR “LEPTIN” OR “ADIPONECTIN” OR “Adipocyte Complement-Related Protein 30-kDa” OR “Adipocyte Complement Related Protein 30 kDa” OR “Adipose Most Abundant Gene Transcript 1” OR “apM-1 Protein” OR “apM 1 Protein” OR “ACRP30 Protein” OR “RESISTIN” OR “TUMOR NECROSIS FACTOR-ALPHA” OR “Tumor Necrosis Factor alpha” OR “Cachectin-Tumor Necrosis Factor” OR “Cachectin Tumor Necrosis Factor” OR “TNFalpha” OR “TNF-alpha” OR “Tumor Necrosis Factor” OR “Cachectin” OR “INTERLEUKIN-6” OR “Interleukin 6” OR “IL6” OR “COMPLEMENT FACTOR D” OR “ADIPSIN” OR “Factor D” OR “Leptin” OR “Obese Protein” OR “NicotinamidePhosphoribosyltransferase” OR “Visfatin” OR “Interleukin-8” OR “Interleukin 8” OR “IL8” OR “IL-8” OR “chemerin protein, human” OR “RBP4 protein, human” OR “Biomarkers” OR “Biomarker” OR “Biologic Markers” OR “Biologic Marker” OR “Biological Marker” OR “Biological Markers” OR “Immunologic marker” OR “Immunologic markers” OR “Cytokines” OR adipocina* OR adipocitocina* OR leptina OR “ProteínaBloqueadora da Sensação de Fome” OR “Proteína Ob” OR “ProteínaAntiobesidade” OR adiponectina* OR resistina OR “Fator de NecroseTumoralalfa” OR “NecroseTumoral” OR caquetina OR “TNF-alfa” OR “Interleucina-6” OR “IL-6” OR “Fator de Crescimento de Hibridoma” OR “Fator de Crescimento de Plasmocitoma” OR “Fator Estimulador de Hepatócito” OR “Proteína Indutora da Diferenciação Mieloide” OR “MGI-2” OR “Fator D do Complemento” OR adipsina OR “Ativador de C3 Convertase” OR “Ativador de Convertase C3” OR “C3PA Convertase”
	OR “Fator D” OR “Fator Properdina D” OR “Nicotinamida Fosforribosiltransferase” OR “Fator Otimizador de Colônia de Células B Precursoras” OR visfatina OR “Quimiocina CXCL8” OR “Fator Quimiotático” OR “IL-8” OR “Peptídeo Ativador” OR biomarcador* OR “Marcador Biológico” OR “Marcadores Biológicos” OR “Marcador Bioquímico” OR “Marcadores Bioquímicos” OR “Marcador Clínico” OR “Marcadores Clínicos” OR “Marcador Imunológico” OR “Marcadores Imunológicos” OR “Marcador de Laboratório” OR “Marcadores de Laboratório” OR “Marcador de Soro” OR “Marcadores de Soro” OR “Marcador Substituto” OR “Marcadores Substitutos” OR “Marcador Sérico” OR “Marcadores Séricos” OR “Marcador Viral” OR “Marcadores Virais” OR citocina* OR citoquina* OR adipoquina* OR adipocitoquina* OR “Proteína Ob” OR “Proteína Obesa” OR “Factor de Necrosis Tumoral alfa” OR “Necrosis Tumoral” OR caquectina OR “TNF-alfa” OR “Factor de Crecimiento de Hibridoma” OR “Factor de Crecimiento de Plasmacitoma” OR “Factor Estimulante de Hepatocito” OR “Proteína Mieloide Diferenciada-Inducida” OR “MGI-2” OR “Factor D” OR “Activador de C3 Convertasa” OR “Activador de laConvertasa de C3” OR “C3PA Convertasa” OR “Nicotinamida Fosforribosiltransferasa” OR “Nicotinamida Fosforibosiltransferasa” OR “Factor Amplificador de Colonias de Células pre-B” OR visfatina OR “Quimiocina CXCL8” OR “Factor Quimiotáctico” OR “PéptidoActivador” OR “Marcador Inmunológico” OR “Marcadores Inmunológicos” OR “Marcador Sustituto” OR “Marcadores Sustitutos” OR “Marcadores Virales”)) AND (instance:”regional”) AND ( db:(“LILACS” OR “BBO”))
Google scholar	(“Periodontal Diseases” OR “Periodontitis”) AND (“Body Measure” OR “body weight” OR “Body Mass Index” OR “Quetelet Index” OR “Waist-Hip” OR “Body Size” OR “Waist Circumference” OR “Obesity” OR “Overweight” OR “Overnutrition” OR “Hypernutrition”) AND (“Adipokines” OR “Adipocytokines” OR “Biomarkers” OR “Biologic Markers” OR “Immunologic marker” OR “Cytokines”)

Appendix 2

Appendix 2.

Excluded articles and reasons for exclusion (n=18)

Author, year	Reason for exclusion
Akman et al., 2012[48]	1
Akram et al., 2017[49]	2
Altay et al., 2013[50]	3
Al-Hamoudi et al., 2018[51]	1
Buduneli et al., 2014[52]	1
Bostrom et al., 2015[53]	1
D’Aiuto et al., 2005[54]	1
Dogan et al., 2015[55]	1
Eldin et al., 2013[56]	2
Fell et al., 2013[57]	2
Karthikeyan et al., 2007[58]	2
Karthikeyan et al., 2007[59]	2
Khanna et al., 2010[60]	2
Kim et al., 2016[61]	1
Lundin et al., 2009[16]	1
Martinez-Herrera et al., 2018[62]	3
Rangé et al., 2013[63]	3
Recker et al., 2015[64]	2
Saito et al., 2008[65]	3
Satpathy et al., 2015[66]	2
Saxlin et al., 2009[67]	1
Selvarajan et al., 2015[68]	1
Shimada et al., 2010[69]	1
Taşdemir et al., 2016[70]	3
Teles et al., 2012[71]	2
Thanakun et al., 2014[72]	3
Zhong et al., 2007[73]	2

Legend: Focus question not answered (n=12); Lack of control group as evaluated in this systematic review (O+P− or O−P+) (n=9); Includes nicotine dependent individuals or patients with diabetes or metabolic syndrome (n=5)