Association between the systemic immune inflammation index and periodontitis: a cross-sectional study

Our data comes from the NHANES database, which contains studies to examine the health and nutritional well-being of adults and children in the United States. The survey is conducted annually on a representative sample of around 5,000 people in US. These folks are distributed throughout the country, of which 15 counties are visited annually. During the NHANES interview, the demographic information, socioeconomic status, dietary habits, and health-related concerns of participants were collected. NHANES launched a six-year national periodontitis surveillance initiative in 2009. This project involved a FMPE study on individuals aged 30 or older, including assessment of the extent of gingival recession and measurement of their pocket depths. The National Center for Health Statistics' Ethics Review Board approved the study, which was conducted with the explicit written consent of all participants. Overview of our study design in Fig. 1.

Our study enrolled a total of 10,366 participants, spanning from 2009 to 2014. The following participants were excluded from our study: (1) incomplete data of platelets, neutrophils, and lymphocytes (N = 5219). (2) incomplete data of periodontitis (N = 13851). (3) incomplete data of covariates (N = 1032). Finally, we were able to enroll a sizeable sample of American adults with periodontitis (N = 10366), which was analyzed as a representative population of the entire nation.

The classical method of diagnosing periodontitis involves assessing clinical parameters such as the plaque index, periodontal pocket probing, bleeding on probing, and the clinical attachment levels [23]. Mild periodontitis was defined as having ≥ 2 interproximal sites with attachment loss (AL) ≥ 3 mm and ≥ 2 interproximal sites with probing depth (PD) ≥ 4 mm (not on the same tooth), or one site with PD ≥ 5 mm. Moderate periodontitis was defined as having ≥ 2 interproximal sites with AL ≥ 4 mm (not on the same tooth), or ≥ 2 interproximal sites with PD ≥ 5 mm (not on the same tooth). Severe periodontitis was defined as having ≥ 2 interproximal sites with AL ≥ 6 mm (not on the same tooth) and ≥ 1 interproximal site with PD ≥ 5 mm (Table 1). Additionally, for mild, moderate, and severe periodontitis, we describe extent and distribution as localized (< 30% of teeth involved) and generalized (≥ 30% of teeth involved) periodontitis [24].

Individual cell counts (neutrophils, lymphocytes, and platelets) from the NHANES were used to calculate the SII for the study population. The SII was calculated using a formula described in previous research papers (SII = platelet count × neutrophil count / lymphocyte count) [16]. The methods used to derive complete blood count parameters are based on the Beckman Coulter method of counting and sizing, and combination with an automatic diluting and mixing device for sample processing, and a single-beam photometer for hemoglobinometry. For more information, refer to Chapter 7 of the NHANES Laboratory/Medical Technologists Procedures Manual. We determined the optimal cutoff value of the SII level by using the receiver operating characteristics (ROC) curve.

We have selected the following variables as probable covariates in our study, as there are some extraneous factors that may impact the outcomes. Demographic data include gender (male or female), age (30–50 years or ≥ 50 years), race (Non-Hispanic White, Non-Hispanic Black, Mexican American, other Hispanic, or other races), education level (less than high school, high school, or above high school), and income to poverty ratio (PIR < 1.3 or PIR ≥ 1.3). The questionnaire data include alcohol consumption (1–10 drinks/month, 10–20 drinks/month, 20 + drinks/month, or non-drinker), smoking status (former smoker, current smoker, or never smoker), coronary heart disease (yes or no), myocardial infarction (yes or no), stroke (yes or no), cancer (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and diabetes (yes or no). Examination data include the body mass index (BMI < 25, 25 to < 30, ≥ 30).

To compare the SII subgroups, a survey-weighted logistic regression was used for continuous variables (mean ± SD) and a survey-weighted Chi-square test was used for categorical variables (n count). The correlation between SII and periodontitis was investigated through univariate and multivariate logistic regression analyses. In order to incorporate covariate adjustments, three models were developed. In Model 1, no adjustments were made for any covariates. In Model 2, adjustments were made for gender, age, race, education level, PIR, BMI, alcohol consumption, and smoking status. In Model 3, we made adjustments for systemic diseases that had previously been reported to be related to periodontitis, including coronary heart disease, myocardial infarction, stroke, cancer, hypertension, hypercholesterolemia, and diabetes [6‐12]. A comprehensive stratified logistic regression analysis was performed to identify the variables that influence the correlation among participants aged 50 or older. A sensitivity analysis was further conducted using multiple imputation approaches. Missing data was imputed through the use of a multiple imputation chain-equation [25].

All statistical analyses were conducted using R (version 4.2.2). A p-value of less than 0.05 was considered significant for two-tailed tests.

A total of 30,468 individuals were involved in the national periodontitis surveillance project conducted by NHANES from 2009 to 2014. According to NHANES rules, participants who were 30 years or older and had at least one tooth (excluding third molars) and did not fulfill any of the health exclusion criteria were eligible for the periodontal evaluation. Our study eventually included 10,366 participants after excluding 5129 participants who lacked complete data on their platelet count, neutrophil count, and lymphocyte count, along with 13,851 participants who lacked complete data on periodontitis and 1032 participants who lacked complete data on covariates (Fig. 2). The study population consisted of 5126 males (49.45% of the total) and 5240 females (50.55%). The average age of the study population was 51.00 years. The weighted study population represents adults aged 30 or older throughout the United States. Periodontitis was diagnosed in 3,984 participants, accounting for 38.43% of the total number of participants. Among them, there were 428 participants with mild periodontitis (including 268 with localized and 160 with generalized periodontitis), 2861 participants with moderate periodontitis (including 1,554 with localized and 1307 with generalized periodontitis), and 695 participants with severe periodontitis (all with generalized periodontitis).We found that there was no statistical difference in SII variation between mild, moderate, and severe periodontitis (P > 0.05, Additional file 1: Table S1), or between localized and generalized periodontitis (P > 0.05, Additional file 2: Table S2). The median of SII in mild, moderate, and severe periodontitis was significantly higher than that of participants without periodontitis (P < 0.01, Additional file 1: Table S1). Therefore, the periodontal status is divided into two groups: those with mild, moderate, or severe periodontitis (classified as "yes") and those without (classified as "no"). The optimal cutoff for SII level was 978 × 10⁹/L (AUC = 0.635, 95% CI: 0.624—0.645, P < 0.01, Additional file 6: Figure S1). Furthermore, it has been noted that individuals with high SII levels often have neutrophilia, lymphopenia, or thrombocytosis [26]. We found that participants with high SII had higher neutrophil counts, lower lymphocyte counts, and higher platelet counts (P < 0.01, Additional file 3: Table S3).

The clinical characteristics of the participants were analyzed according to their periodontal status, as presented in Table 2. The results demonstrate a statistically significant disparity in gender, age, race, education level, PIR, alcohol consumption, smoking status, high SII scores, coronary heart disease, myocardial infarction, hypertension, and diabetes (all P < 0.05).

The results in Table 3 indicated that the risk of suffering from periodontitis is significantly higher among participants who are aged 50 or older, smokers, are non-Hispanic Black, Mexican American, other Hispanic, or other races, have coronary heart disease, have myocardial infarction, have hypertension, have diabetes, and have high SII scores (OR > 1, P < 0.05). However, the risk of periodontitis is reduced in participants who are female, have a higher level of education, and have a PIR of 1.3 or above (OR < 1, P < 0.05).

The findings display in Table 4 revealed a significant difference between SII and periodontitis in three weighted logistic regression models (OR > 1, P < 0.05). Participants with a SII greater than 978 × 10⁹/L are more likely to develop periodontitis. The close correlation between age and periodontitis is widely recognized, and Table 4 also offers additional insight after breaking down the ages of the participants. Logistic regression analysis showed that there was no evidence in Models 1–3 that a SII above 978 × 10⁹/L is associated with a higher risk of periodontitis in the 30–50 years old age group (Model 1: OR = 1.301, 95% CI 0.849, 1.993, P = 0.215; Model 2: OR = 1.266, 95% CI 0.768, 2.086, P = 0.337; Model 3: OR = 1.213, 95% CI 0.722, 2.039, P = 0.445). Finally, a significant association between a high SII and periodontitis was observed in the subgroup of participants who were aged 50 or older after adjusting for all potential confounding factors (Model 3: OR = 1.409, 95% CI 1.037, 1.915, P = 0.022).

Subsequently, a stratified logistic regression analysis was conducted on participants aged 50 or older, and it was found that there were no factors that could affect the relationship between SII and periodontitis in this age group (Additional file 4: Table S4). In order to ensure the stability of our findings, we employed multiple imputation methods to impute any missing data and conducted a sensitivity analysis. The distribution of characters in the baseline was shown in Additional file 5: Table S5. The sensitivity analysis results revealed that the model, which was adjusted for all potential variables among the participants aged 50 or older, generated consistent results to the aforementioned conclusion (OR = 1.382, 95% CI 1.019, 1.876, P = 0.029, Table 5).