Association between Obesity and Periodontitis in US Adults: NHANES 2011–2014

Obesity is increasing worldwide and is becoming a significant social problem [1]. Obesity has a negative impact on an individual's physical and mental health, leading to a poor quality of life [2]. The World Health Organization (WHO) reports that obesity can lead to chronic diseases such as diabetes, cardiovascular disease, musculoskeletal disorders, and cancer [3]. Coronavirus patients who are obese have higher risks and worse outcomes than those who are not obese [4].

In the world, half of all adults suffer from periapical periodontitis, a disease with a high prevalence worldwide [5]. Periodontitis is highly prevalent in adults over 30 years in the USA, according to the National Health and Nutrition Examination Survey (NHANES) (2009–2014) [6]. Its social burden is increasing globally, warranting global changes in public health policy [7]. Therefore, more research on its risk factors is needed to prevent it.

Obesity is also a major risk factor for increased susceptibility to periodontal disease and may have a significant population-attributable risk. Obese adults have a higher prevalence of periodontitis, according to Suvan et al. [8]. It has also been found that obesity is a risk factor for periodontitis in numerous studies [9‒11]. In spite of numerous studies examining obesity and periodontitis, most of them are self-constructed databases, and fewer data are available from the US National Health and Nutrition Examination Survey. The findings, however, remain controversial. Some scholars [12‒14] have showed that body mass index (BMI) and waist circumference (WC) are associated with periodontitis; some scholars [15, 16] have reported that BMI is not associated with periodontitis but WC is associated with periodontitis; some scholars [17] have demonstrated that periodontitis is not associated with BMI and WC. Therefore, to better understand the relationship between obesity and periodontitis in the USA, more research is needed.

This cross-sectional study used NHANES data from 2011 to 2014 and was performed by the US Centers for Disease Control and Prevention. By using a stratified multistage probability survey, the NHANES project evaluated the health and nutrition status of noninstitutionalized Americans [18]. Through home visits, screenings, and laboratory testing conducted by mobile examination centers (MECs), the NHANES collects demographic and health information. Before participating in the NHANES, all participants completed written informed consent forms approved by the National Center for Health Statistics (NCHS) Ethics Review Committee. The secondary analysis did not require additional institutional review board approval. Individuals over 30 years old who had completed an interview participated in our study. We excluded pregnant women or individuals with missing data on BMI, WC, and periodontitis; a statistical estimate of the missing value of a covariate was used in place of the missing data. This study has been reported according to the STROBE statement.

The NHANES 2011–2014 “Oral Health – Periodontal Exam” protocol included measuring 6 sites on each tooth for a maximum of up to 28 teeth. Two sets of clinical periodontal measurements were included: clinical attachment loss (AL) and probing depth (PD) [19]. The researchers applied the Centers for Disease Control and Prevention/the American Academy of Periodontology classification/case definition to define “periodontitis” in the present study. It is defined as severe periodontitis in the algorithm as the presence of 2 or more interproximal sites with ≥6 mm clinical AL (not on the same tooth) and 1 or more interproximal site (s) with ≥5 mm PD. Severe periodontitis was not diagnosed in participants with moderate periodontitis, and they needed to have 2 or more interproximal sites with ≥4 mm clinical AL (not on the same tooth) or have 2 or more interproximal sites with ≥5 mm PD (not on the same tooth). Respondents categorized as having mild periodontitis did not belong to the preceding two groups and had ≥2 interproximal sites with ≥3 mm clinical AL and ≥2 interproximal sites with ≥4 mm PD (not on the same tooth) or 1 site with ≥5 mm PD. Participants who did not fall into any of these periodontitis categories were considered to be negative for periodontitis. The classification of “periodontitis” includes all patients who are classified as having mild, moderate, or severe periodontitis.

BMI and WC were used as indicators of obesity. BMI, as a measure of overall adiposity, was calculated by dividing weight (in kilograms) by height (in meters) squared and was divided into four categories: underweight (<18.5 kg/m²), normal (18.5–24.9 kg/m²), overweight (25–29.9 kg/m²), and obese (≥30 kg/m²). WC, as a measure of visceral adiposity, was divided into two categories, with a cutoff point of >102 cm for men and >88 cm for women. The cutoff points for BMI and WC are based on guidance from the National Heart, Lung, and Blood Institute (NHLBI) of the USA and the WHO [20, 21].

A variety of potential covariates were assessed according to the literature [12, 22, 23]. The selected variables were age, sex, race/ethnicity, education level, marital status, alcohol status, time since the last dental visit, physical activity, sleep time, family income, smoking status, how many days use dental floss, days used mouthwash for dental problem, arthritis, gout, congestive heart failure, coronary heart disease, angina pectoris, stroke, thyroid disease, diabetes, hypertension, energy consumption, fat consumption, cholesterol consumption, protein consumption, sugar consumption, and dietary fiber consumption. Race/ethnicity was classified as non-Hispanic White, non-Hispanic Black, Mexican American, or other races. Educational level was divided into less than 9 years, 9–12 years, and more than 12 years. Marital status was classified as either married, living with a partner, or living alone. Alcohol status was created from a question: “Had at least 12 alcohol drinks a year?” If the answer is “drinking at least 12 drinks of alcohol per year,” then the alcohol status is considered “Yes”; if the answer is “drinking less than 12 drinks of alcohol per year,” then the alcohol status is considered “No.” The time since the last dental checkup was categorized into three groups (≤1 year, 2–3 years, and >3 years). Physical activity was classified as sedentary, mild (walk or use a bicycle for at least 10 min continuously within the last 7 days, resulting in only light sweating or a light increase in breathing or heart rate), moderate (at least 10 min of movement within the last 7 days, resulting in only moderate sweating or a moderate increase in breathing or heart rate), and vigorous (at least 10 min of activity within the last 7 days, resulting in profuse sweating or an increase in breathing or heart rate). Sleep time was generated by a question: “How much sleep do you usually get on weekday or weeknight?” And it was divided into three categories (<7 h, 7–9 h, and ≥9 h). According to a US government report, family income was divided into three categories based on the poverty income ratio (PIR): low (PIR ≤1.3), medium (PIR >1.3–3.5), and high (PIR >3.5). According to preceding literature definitions, based on their answers to questions about whether they have smoked at least 100 cigarettes in their lifetime and whether they were currently smoking, smoking status was categorized as never smokers, current smokers, and former smokers. The number of days using dental floss was prompted by a question: “Other than brushing teeth, how many days in the past 7 days did you use dental floss or another device to clean between your teeth?” The number of days mouthwash was used to address dental problems was generated by the question: “In addition to brushing with a toothbrush, how many days in the past 7 days did you use mouthwash or other dental rinse products to treat dental disease or dental problems?” The determination of previous diseases (hypertension, diabetes, stroke, coronary heart arthritis, gout, congestive heart failure, angina pectoris, and thyroid disease) was based on the inquiry in the questionnaire of whether the doctor had been informed of the condition in the past. A dietary recall interview preceded an interview at MEC to obtain participants’ 24 h nutritional information, including energy consumption, fat consumption, cholesterol consumption, protein consumption, sugar consumption, and dietary fiber consumption.

This is a secondary examination of publicly accessible datasets. Categorical variables were represented by proportions (%), while continuous variables were described by the mean (standard deviation [SD]) or median (interquartile range [IQR]), depending on the situation. In order to compare differences among groups, one-way analyses of variance (ANOVA) (normal distribution), Kruskal-Wallis test (skewed distribution), and χ² test (categorical variables) were conducted. Univariate and multivariate logistic regression analyses were performed to evaluate the OR and 95% confidence intervals (95% CIs) of the relationship between BMI, WC, and periodontitis. Model I did not adjust any covariates. Model II was adjusted for age, sex, and race/ethnicity. Model III was adjusted for model II and education level, marital status, alcohol status, time since last dental visit, physical activity, sleep time, family income, smoking status, how many days used dental floss, and days used mouthwash for dental problem. Model IV was fully adjusted, including model III and arthritis, gout, congestive heart failure, coronary heart disease, angina pectoris, stroke, thyroid disease, diabetes, hypertension, energy consumption, fat consumption, cholesterol consumption, protein consumption, sugar consumption, and dietary fiber consumption.

Furthermore, potential modifications of the relationship between BMI, WC, and periodontitis were assessed, including the following variables between BMI and periodontitis: sex, age (30–44, 45–59, 60–80 years), race/ethnicity, education level (<9, 9–12, ≥12 years), marital status (married, living with a partner, living alone), family income (low, medium, high), sleep time(<7, 7–9, ≥9 h), mouthwash (<3, 3–4, 5–7 days), floss (<3, 3–4, 5–7 days), smoking status (never, current, former), alcohol status (no, yes), diabetes (no, yes), hypertension (no, yes); including the following variables between WC and periodontitis: sex, age (30–44, 45–59, 68–80 years), race/ethnicity, marital status (married, living with a partner, living alone), family income (low, medium, high), sleep time (<7, 7–9, ≥9 h), mouthwash (<3, 3–4, 5–7 days), floss (<3, 3–4, 5–7 days), last time (<1, 1–3, > 3 years), smoking status (never, current, former), alcohol status (no, yes), diabetes (no, yes), hypertension (no, yes); then performed an interaction test. Tests for effect modification for those of subgroup indicators were followed by the likelihood ration test. To assess the robustness of our results, we compared the baseline characteristics of missing and nonmissing populations, and the interactive effect of age and categorized BMI on periodontitis was also analyzed.

Because the sample size was determined solely on the data provided, no a priori statistical power estimates were performed. All analyses were performed using the statistical software packages R 3.3.2 (http://www.R-project.org, The R Foundation, Shanghai, China) and Free Statistics software version 1.7 [24]. A descriptive study was conducted on all participants. By conducting a two-tailed test, it is declared that a p value <0.05 is significant.

In total, 19,931 participants completed the interviews. We excluded pregnant women (n = 40), those with missing data on periodontitis (n = 12,969), those with missing data on BMI (n = 47), and those without WC (n = 213). Missing covariate data were replaced with statistical estimates of missing values, and mean, median, or mode were used as missing values for continuous variables, while the missing values for categorical variables were treated as a separate group, named 9. Ultimately, this cross-sectional study included 6,662 participants from NHANES 2011–2014 in the analysis. Detailed inclusion and exclusion processes can be found in Figure 1.

The basic characteristics of participants before covariate missing value replacement, after excluding covariate missing values, and with missing values of covariates are displayed in an online supplementary material (Table S1) (for all online suppl. material, see https://doi.org/10.1159/000534751). Table 1 shows the baseline characteristics of all individuals with and without periodontitis. Study participants had a mean age of 51.8 (±14.1) years, average BMI was 29.1 (±6.6) kg/m², average WC was 99.7 (±15.6) cm, and 3,295 (49.5%) of them were male. Among patients with periodontitis, 58.2% of the participants were males, 60.7% were married or living with a partner, more than 44% of the participants had at least 12 years of education. Among the 6,662 participants, 83.1% of them do not have diabetes, while among the 6,662 participants, 47.7% of them do not smoke, and nearly, half of the subjects had a dental visit for less than 1 years.

After conducting a multivariate analysis, it was found that BMI and WC have a significant association with the incidence of periodontitis. For every 1-unit increase in BMI, there is a 1% increase in periodontitis incidence, and for every 1-unit increase in WC, there is also a 1% increase in periodontitis incidence. After stratification by BMI at 18.5, 25, and 30 kg/m², periodontitis incidence during the follow-up period increased (p > 0.05) in the higher BMI subgroup compared with the subgroup with BMI <18.5 kg/m² (p for trend = 0.035). After stratification by WC, periodontitis incidence increased (p = 0.01) in the obese subgroup compared with the normal subgroup (Table 2).

To evaluate the potential impact of BMI and WC on the relationship with periodontitis, a sensitivity analysis was conducted, including a stratified analysis in various subgroups. The results of this analysis are presented in Figures 2 and 3. The study found that there was a significant association between BMI and WC with periodontitis in female subjects but not in male subjects. The stratified analysis revealed a highly consistent pattern. Regardless of the subgroup, effect size of BMI and WC on periodontitis was stable. The interaction analysis revealed that age (p for interaction = 0.032) and diabetes (p for interaction = 0.013) had a significant interactive effect on the relationship between BMI and periodontitis, last time (p for interaction = 0.042) played an interactive role in the association between WC and periodontitis. Considering multiple testing, a p value of less than 0.05 for the interaction diabetes may not be clinically significant between BMI and periodontitis. Similarly, last time may not be clinically significant between WC and periodontitis. Table 3 showed the relationship between different BMI groups and periodontitis in different age groups. BMI ≥30 kg/m² was associated with periodontal disease prevalence in individuals aged 30–44 years, which remained statistically significant after adjusting for all covariates (p < 0.001) but not in other age groups. A dose-response relationship between BMI and periodontitis in different age groups is presented in the online supplementary material (online suppl. Fig. S1).

Our findings showed that the risk for periodontitis increased with BMI, and WC was similarly correlated among individuals aged 30–80 years. In female subjects, BMI and WC showed a significant association with periodontitis; however, no such association was found in male subjects. While no significant association was observed between BMI and periodontitis when BMI was categorized into four groups, individuals with high WC (>102 cm in men; >88 cm in women) had 1.18 times higher odds of developing periodontal disease compared to those with normal WC. Young individuals with a BMI of 30 kg/m² or higher had a 33% higher prevalence of periodontal disease compared to their normal-weight counterparts. In order to effectively analyze the association between obesity and periodontitis, researchers need to consider the number of levels involved in the analysis.

Multiple studies have demonstrated a direct link between obesity and periodontitis. For instance, Khader et al. [13] and Sarlati et al. [25] surveyed BMI and WC, while Kim et al. [26], Jae et al. [27], and Dalla Vecchia et al. [28] assessed BMI. Additionally, Saito et al. [29] evaluated BMI, WHR, and body fat in Japan. All these studies indicated a positive correlation between obesity and periodontitis. However, the biological mechanism of this association was not well understood. According to a review [30], adipose tissue secreted a bioactive substance called “diopocytokine” that could potentially harm the periodontal tissue; plasminogen activator inhibitor-1, expressed in visceral fat, could induce visceral blood clotting and increase the risk of ischemic vascular disease, which might also reduce blood flow to the periodontium in obese individuals, promoting the development of periodontal disease. Another mechanism involved oral bacteria. Nakai et al. [31] suggested that obese individuals might have different oral bacteria than normal-weight individuals, which could contribute to the development of periodontitis. In vitro studies have shown that stimulation of adipocytes with C-reactive protein and inflammatory cytokines promotes adipocyte differentiation and increases the production of mature proteases such as matrix metalloproteinases. Furthermore, adipocytes secreted proinflammatory cytokines, such as TNF-α and IL-6, leading to systemic inflammation [32]. TNF-a promoted periodontitis by stimulating osteoclast formation, inducing alveolar bone destruction and connective tissue destruction. TNF-a was considered to be a contributor of early stage periodontitis in the obese population [33]. Moreover, another possible mechanism is the ingestion of periodontal pathogens; some scholars have found one pathogen, also found in the periodontal biofilm, Selenomonas noxia, which, at levels 41.05% of the total salivary bacteria, could identify obese individuals with a high sensitivity and specificity. S. noxia belongs to the phylum Firmicutes, whose relative proportion of the gut microbiota has been reported to be elevated in obese individuals [34, 35].

However, others have not found this correlation. Linden et al. [36] in Northern Ireland reported that there was no significant relation between BMI and high-threshold periodontitis among 60–70 years in western European men. In a study conducted by Ylo¨stalo et al. [37], it was found that there was no correlation between body weight and periodontal infection in 2,841 dentate nondiabetic individuals aged between 30 and 49 years. Torrungruang et al. [17] concluded that periodontitis was not associated with BMI and WC. The reason may be that the association between obesity and periodontal disease varies by region and race.

We found a significant association between obesity and periodontitis in young adults, which was consistent with the findings of Zahrani et al. [12]. Various potential mechanisms could explain the findings above. The first explanation could be the observed cohort effect, but our current research is cross sectional. Younger and older participants may have been exposed to different lifestyles. Dietary trend studies have already documented a shift in American teenagers’ dietary patterns towards less healthy food, and overweight young participants may have unhealthy dietary patterns that are deficient in essential nutrients and have excess sugar and fat content [12, 38], and the carbonated beverages also have an impact on the relationship between obesity and periodontal disease. For instance, glucose-containing carbonated soft drinks increase carbonyl stress burden, which may result in a decrease in antioxidant concentration in oral saliva and enhance oxidative nitration in association with the inflammatory reaction [39]. Such dietary patterns could increase the risk of periodontal disease. The second explanation could be due to the dilution of the effect of obesity in the older age groups. It may be that the influence of obesity on the periodontal condition of older participants was masked by the presence of stronger risk factors (such as age). Thus, more nonobese individuals would develop periodontal disease as they age. Finally, when faced with social shame related to obesity, young people face greater pressure than older people. Social psychological factors may contribute to the mechanism of periodontal disease, including physiological changes in blood supply and salivary flow; changes in immune responses to periodontal pathogens; or changes in health-related behaviors, such as smoking, dietary intake, and oral hygiene habits [40].

In this study, sub-analyses by gender showed that BMI and WC were significantly associated with periodontitis in obese females. These results are consistent with the findings of Dalla Vecchia et al. [28] and Hisayama et al. [41]. Possible mechanisms need to be further explored.

The main limitation of our study is its cross-sectional design. Therefore, the association between obesity and periodontitis does not imply a causal relationship. However, the study has several strengths including a large sample size, a significant association between obesity and periodontitis even after adjusting for many potential confounding factors, a multivariate logistic regression model, and its generalizability to the US population. However, cohort studies and intervention studies are needed to further understand the causal relationship between periodontitis and obesity, especially in young people, and to determine if obesity is a true risk factor for periodontal disease. If so, preventing and managing obesity may be an adjunct method to improving periodontal health. Although many confounding factors were adjusted for, residual confounding may still affect the association between periodontitis and obesity.

In summary, after adjusting for many variables, there is a significant association between BMI and WC with obesity, which is also significant in young adults (aged 30–44 years) with obesity (BMI ≥30 kg/m²). This has important implications for the study of obesity and periodontitis. Prospective cohort studies may help to determine the causal relationship between periodontitis and obesity.

Ethical approval was not required as this study was based on publicly available data. The NHANES was authorized by the National Center for Health Statistics (NCHS) Ethics Review Committee, approval number # 2011-17. All participants completed written informed consent forms before participation. The secondary analysis did not require additional Institutional Review Board approval.

The authors declare no conflict of interest.

This study was funded by the Sichuan Medical Association (Q21037) and Chengdu Medical College (21Z079).

Ling Liu conducted data analysis and wrote the manuscript. Linyu Xia conducted data collection and data interpretation. Ren guo Gong designed the study and reviewed the manuscript. Jing Xu conducted data analysis and reviewed the manuscript. Yu jie Gao conducted the data collection. Xiu hua Dong conducted data collection. All authors have reviewed and approved the manuscript for publication.

Ling Liu and Lin Yu Xia are co-first authors and contributed equally to this work.

All data in the article can be obtained from the NHANES database (https://www.cdc.gov/nchs/nhanes/index.htm). Further inquiries can be directed to the corresponding author.