Effect Modification by Obesity on Nonsurgical Periodontal Treatment

Abstract

Introduction:

Obesity is associated with increased periodontal disease prevalence and incidence. This retrospective cohort study examined whether body mass index (BMI) is an effect modifier of periodontal treatment outcomes in patients attending an urban dental school clinic.

Methods:

Data were extracted from electronic health records of 344 patients at a large urban dental school clinic who had at least 1 tooth with a probing pocket depth (PD) ≥5 mm at baseline and who subsequently received nonsurgical periodontal treatment. BMI was computed from self-reported weight and height and categorized as obese (≥30 kg/m²), overweight (25–29.9 kg/m²), or healthy (18–24.9 kg/m²). The primary treatment outcome of interest was defined as having no teeth with PD ≥5 mm in a quadrant on follow-up after nonsurgical periodontal therapy. That outcome was considered to represent treatment success in this study. Secondary outcomes included changes in mean PD and clinical attachment loss (CAL). Analyses included 879 treated quadrants among 344 patients (185 males, 159 females; mean age 49 ± 12 y at baseline; mean posttreatment follow-up of 6 ± 2 mo). Clinical outcomes in patients who were overweight or obese were compared to healthy-weight patients using generalized linear models for binary or continuous outcomes, accounting for clustering within patients. Covariates were age, gender, tobacco use, history of diabetes, insurance type, and number of baseline sites ≥5 mm.

Results:

Obesity was associated with a significantly lower likelihood of successful nonsurgical treatment (odds ratio = 0.47; 95% confidence interval, 0.25–0.88) than healthy weight. Being overweight was not associated with treatment success. Posttreatment reductions in the percentage of sites with pockets ≥5 mm and CAL ≥5 mm were greater in patients with healthy weight as compared to those either overweight or obese. However, posttreatment changes in mean PD and CAL did not differ among the BMI groups.

Conclusions:

Obesity adversely modifies the effectiveness of nonsurgical periodontal treatment among dental school clinic patients.

Knowledge Transfer Statement:

The results of this study may be used by dental providers to better understand and manage periodontal therapy in patients with obesity. Furthermore, patients will be better informed about their therapeutic options and outcome success.

Introduction

Nearly one-third of US adults are categorized as overweight and more than 40% as obese (Fryar et al. 2020) according to the US Centers for Disease Control and Prevention (CDC 2024) criteria. Periodontitis remains one of the most prevalent chronic diseases in the United States, affecting nearly half of adults (Eke et al. 2018), with a higher prevalence in persons who are obese (Kim et al. 2022). Adipose tissue is metabolically active, secreting cytokines and hormones that regulate the body’s response to infection. In obesity, levels of proinflammatory mediators and markers of oxidative stress are elevated, whereas proresolution mediators and antioxidants are decreased (Pamuk and Kantarci 2022). This imbalance, which leads to a prolonged inflammatory state, tissue destruction, and impaired wound healing, likely affects how well a patient responds to periodontal disease treatment. As a large segment of the population is affected by both conditions, it is important to know if obesity influences the success of periodontal treatment. However, to date, there is no consensus regarding the effect of obesity on changes in pocket depth and clinical attachment after nonsurgical treatment (Akram et al. 2016; Gerber et al. 2016).

The effect of obesity on the outcome of nonsurgical periodontal treatment has been investigated in numerous studies (Zuza et al. 2011; Al-Zahrani and Alghamdi 2012; Altay et al. 2013; Michalowicz et al. 2014; Suvan et al. 2014; Bouaziz et al. 2015; Goncalves et al. 2015; Balli et al. 2016; Duzagac et al. 2016; Peralta et al. 2020; Suvan et al. 2020; Zuza et al. 2020). Obesity was associated with smaller posttreatment improvements in mean pocket depth, clinical attachment loss, or percent sites with deep pockets compared to not being obese (Suvan et al. 2014; Bouaziz et al. 2015; Goncalves et al. 2015; Suvan et al. 2020). Two studies reported greater improvement in patients who were obese (Michalowicz et al. 2014; Duzagac et al. 2016) while others found no significant differences in these outcomes attributable to obesity (Zuza et al. 2011; Al-Zahrani and Alghamdi 2012; Altay et al. 2013; Balli et al. 2016; Peralta et al. 2020; Zuza et al. 2020). In most studies, patients with moderate to severe periodontal disease were treated with full-mouth scaling and root planing, oral hygiene instruction, and periodontal maintenance visits over a period of several months. Additional therapies included chlorhexidine irrigation or mouthwash (Altay et al. 2013; Michalowicz et al. 2014; Bouaziz et al. 2015; Peralta et al. 2020). Key measures of the response to treatment included pre- to posttreatment changes in pocket depth, clinical attachment loss, and bleeding on probing. Body mass index alone was used to define obesity and nonobese status in several studies (Al-Zahrani and Alghamdi 2012; Michalowicz et al. 2014; Suvan et al. 2014; Bouaziz et al. 2015; Suvan et al. 2020), whereas others incorporated waist circumference or waist/height ratio to account for the pattern of fat distribution. Uncertainty regarding the true impact of obesity on nonsurgical periodontitis treatment is also compounded by the very selective patient samples and treatment regimens in these studies. For example, individuals with mild periodontitis, prior periodontal treatment, use of certain medications, comorbidities, systemic disease such as diabetes, and habits such as smoking were frequently excluded, resulting in patient samples that may not be representative of those in private practice or dental clinics for whom the treatment is intended.

The purpose of this study was to determine whether clinical periodontal outcomes of nonsurgical periodontal therapy differ among patients who are obese, overweight, or healthy weight after accounting for baseline disease severity and other confounders using electronic health record data from an urban dental school clinic.

Methods

Study Design and Data Source

This retrospective cohort study used data obtained from electronic dental records of 344 patients who had an initial periodontal evaluation and underwent nonsurgical periodontal treatment between 2010 and 2019 at Boston University Henry M. Goldman School of Dental Medicine (GSDM). A data repository comprising deidentified clinical dental data (SALUD; Two Ten Health) was established by the GSDM Center for Clinical Research in 2018. As this study uses only deidentified data, it was deemed not human subjects research by the Boston University Medical Campus Institutional Review Board. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines (von Elm et al. 2007).

Periodontal Measurements

Predoctoral and postdoctoral dental students performed periodontal examinations and treatment under the supervision of faculty. Probing depth (PD) and recession were measured at 6 sites per tooth using Williams or University of North Carolina (UNC) probes. Clinical attachment loss (CAL) at each site was computed. American Dental Association (2021) Current Dental Terminology Codes (CDT) were assigned at each periodontal visit. Nonsurgical treatment procedure codes for scaling and root planing (D4341, D4342) were recorded by quadrant, performed on either 1 to 3 teeth per quadrant or 4+ teeth per quadrant.

Obesity Status

Weight and height were self-reported on patients’ medical history forms administered at baseline. Values of weight and height reported in Imperial units were converted to kilograms and meters, respectively. Values of height that were <1.22 m (48 in.) or >2.12 m (84 in.) and of weight <40 kg (90 lbs.) or >205 kg (450 lbs.) were considered unusable. Values of height that were missing or unusable at baseline were extrapolated from height reported at the nearest subsequent examination when possible, under the assumption that adult height does not change substantially over a short period of time, but no substitutions were made for missing or unusable baseline weights. Body mass index (BMI, kg/m²) was calculated and categorized according to CDC criteria (Centers for Disease Control and Prevention 2022): obese (≥30 kg/m²), overweight (25–29.9 kg/m²), or healthy weight (18.5–24.9 kg/m²).

Other Data

Self-reported data on age, gender, tobacco use, dental insurance type, and history of diabetes were obtained from the electronic records. Dental plaque was noted on each surface but was recorded in the clinical database as present or absent in a quadrant. However, recording of plaque was available for only 35% of quadrants.

Inclusion and Exclusion Criteria

Records were available for 5,535 patients who had an initial periodontal evaluation between July 1, 2010, and June 30, 2017, and underwent nonsurgical periodontal treatment in at least 1 quadrant within 6 mo of their initial periodontal examination (reference clinic population). Further inclusion criteria for our final sample include the following: 1) had a usable BMI value within 6 mo of their initial periodontal examination, 2) were followed for at least 2 y, 3) had at least 1 follow-up periodontal examination within 6 mo after nonsurgical periodontal treatment, and 4) had PD and CAL data on at least 14 teeth (to ensure no study participants had only partial-mouth assessments). Figure 1 illustrates the selection process for our final analytic sample (n = 344).

Figure 1.

Flowchart of study population eligibility.

Statistical Analysis

Nonsurgical periodontal treatment was the exposure. The outcomes of interest were changes following treatment in mean PD and mean CAL in a quadrant and treatment success. For the purposes of these analyses, treatment success was defined a priori as having no teeth in a quadrant with pocket depth ≥5 mm after treatment in that quadrant. The 5-mm threshold was chosen because it is consistent with guidelines and optimal goals of active periodontal therapy in the management of periodontal disease (Sanz et al. 2015). We hypothesized that BMI modified the effect of nonsurgical periodontal treatment on treatment success.

Analysis of variance and the χ² test were used to examine associations between nonsurgical periodontal treatment outcomes (i.e., changes in PD and CAL following treatment), BMI, and covariates. Because treatment was specific to a quadrant rather than the whole mouth, and individual patients could have multiple quadrants with different exposure status, we used multivariable general linear models that account for the within-subject correlation of outcomes (binary outcome for treatment success and continuous outcomes for changes in PD and CAL). These models are appropriate for correlated data such as tooth quadrants that cluster within an individual. Only treated quadrants (n = 662) were used in the analysis of treatment success. Age, gender, insurance type, history of diabetes, and tobacco use were selected a priori for inclusion in the final models based on our prior analyses of this data repository (Kaye et al. 2023; Singhal et al. 2023). Additional covariates were included if bivariate analyses were statistically significant at P < 0.05. We estimated the odds ratios (OR) and 95% confidence intervals (CIs) of success of nonsurgical periodontal treatment among patients who were overweight or obese relative to healthy weight. The OR estimates were adjusted for age, gender, tobacco use (yes/no), history of diabetes (yes/no), dental insurance type (none, Medicaid, private, both Medicaid and private), and initial number of sites with pocket depth ≥5 mm.

Length of follow-up between nonsurgical treatment and reassessment of probing depth and CAL ranged from 2 wk to 6 mo (mean ± SD, 3.2 ± 1.6 mo). We also performed a sensitivity analysis of treatment success after excluding patients with follow-up times less than 1 mo. That analysis used 589 quadrants among 303 patients.

Models estimating changes in mean PD and CAL among the 3 BMI groups were controlled for age, gender, tobacco use, history of diabetes, dental insurance type, and number of sites probed. Statistical significance was set at P < 0.05. The number of quadrants available for analysis of mean change in pocket depth and attachment loss was 879 and included quadrants where no teeth exceeded the probing depth threshold of 5 mm at the initial examination. Analyses were repeated after the addition of plaque to the models, which was available for only 35% of quadrants. All analyses were performed with SAS v9.4 (SAS Institute).

Results

Table 1 presents characteristics of the 344 adult patients (185 men and 159 women) included in these analyses. Overall, the mean age was 49 ± 12 y at the initial periodontal examination and did not differ among the BMI categories. Patients who were obese were significantly more likely to be female, have a history of diabetes, and use Medicaid dental insurance than if they were nonobese. Twenty-eight percent of patients (n = 95) had 1 quadrant treated, 24% (n = 83) had 2 treated, 13% (n = 46) had 3 treated, and 35% (n = 120) had all 4 quadrants treated. The study population was similar to our reference population of periodontal clinic patients in terms of age, number of teeth present, tobacco use, dental insurance type, and mean probing depth and CAL. However, the study population had a higher percentage of males and patients with diabetes.

Table 1.

Characteristics of the Reference Population and Study Sample at the Initial Periodontal Examination by BMI Status.

Characteristic	Periodontal Clinic Reference Population	Study Sample That Met Inclusion Criteria
		Total Study Sample	Healthy BMI (18.5–24.9 kg/m2)	Overweight (25.0–29.9 kg/m2)	Obese(≥30 kg/m2)
Number of participants	5,535	344	110	117	117
Age, y	46 ± 14	49 ± 13	47.3 ± 13.0	50.1 ± 13.3	49.5 ± 10.4
Height, m	1.69 ± 0.11 (n = 5,140)	1.68 ± 0.1	1.70 ±0.1	1.69 ± 0.1	1.66 ± 0.11
Weight, kg	78.5 ± 18.8 (n = 5,156)	81.1 ± 19.6	65.8 ± 9.0	78.3 ± 9.8	98.3 ± 20.9
Number of teeth	27 ± 8	26 ± 4	27 ± 4	26 ± 4	26 ± 4
Male, %	48	54	57	67	38
Female, %	52	46	43	33	62
Diabetic, %	11	19	10	16	32
Tobacco use, %	17	16	16	19	14
Insurance type, %
None	26	28	25	31	27
Medicaid	48	46	46	38	50
Medicaid & private	4	8	4	15	4
Private	22	19	26	16	18
Number of quadrants analyzed	13,963	879	280	301	298
No. of teeth/quadrant	7 ± 1	7 ± 1	7 ± 1	7 ± 1	7 ± 1
Mean probing depth (mm)/quadrant	3.0 ± 0.7	3.0 ± 0.8	2.9 ± 0.9	3.0 ± 0.7	2.9 ± 0.7
Mean attachment loss (mm)/quadrant	3.3 ± 0.9	3.2 ± 1.0	3.2 ± 1.1	3.2 ± 0.9	3.2 ± 0.8
Sites with probing depth ≥5 mm/quadrant, %	14	13	13	13	15

Values are presented as mean ± SD unless otherwise indicated.

BMI, body mass index.

As shown in Table 2, obesity was associated with significantly lower odds of a successful outcome after a nonsurgical procedure (OR = 0.47; 95% CI, 0.25–0.88) relative to healthy weight. Among those who were overweight, the odds of success of nonsurgical treatment (OR = 0.86; 95% CI, 0.48–1.52) were not significantly different from those who were healthy weight. Results did not change when the analysis excluded patients who were followed less than 1 mo (obese: OR = 0.49; 95% CI, 0.26–0.94; overweight: OR = 0.86; 95% CI, 0.48–1.55). The inclusion of plaque in the model further decreased the odds of treatment success among both the obese (OR = 0.39; 95% CI, 0.13–1.17) and overweight groups (OR = 0.42; 95% CI, 0.17–1.02), but neither decrease was statistically significant.

Table 2.

Multivariable Adjusted Probability of Success of Nonsurgical Periodontal Treatment.

	Success of Nonsurgical Treatment (n = 662 Quadrants)
Characteristic	Odds Ratio	95% CI
BMI
Obese (≥30 kg/m2)	0.47*	0.25–0.88
Overweight (25.0–29.9 kg/m2)	0.86	0.48–1.52
Normal (18.5–24.9 kg/m2)	1.00 (reference)
Number of sites ≥5 mm at initial visit	0.84*	0.79–0.9
Age	1.00	0.98 –1.02
Female gender	1.85*	1.12–3.04
Insurance
None	0.49*	0.25–0.98
Medicaid	1.05	0.56–1.94
Medicaid and private	0.91	0.29–2.89
Private	1.00 (reference)
Tobacco use	0.94	0.49–1.83
History of diabetes	0.82	0.4–1.66

Success is defined as no teeth with ≥5-mm pocket depth in quadrant after nonsurgical treatment.

CI, confidence interval; BMI, body mass index.

^*Statistically significant, P < 0.05.

Changes in mean pocket depth and attachment loss following nonsurgical treatment tended to be less favorable among patients who were overweight or obese than in healthy-weight patients, but average changes were slight and did not differ significantly among BMI groups (Table 3). Results were similar when the analysis was restricted to quadrants in which 1 or more teeth had pocket depth or CAL of 5 mm or greater (data not shown).

Table 3.

Multivariable Adjusted Millimeter Reduction in Pocket Depth and Clinical Attachment Loss in All Quadrants, Relative to Healthy Weight, after Nonsurgical Periodontal Treatment.

	Pocket Depth		Clinical Attachment Loss
Characteristic	Estimate (mm)	95% CI	Estimate (mm)	95% CI
BMI
Obese (≥30 kg/m2)	–0.05	–0.23 to 0.13	0.00	–0.19 to 0.18
Overweight (25.0–29.9 kg/m2)	–0.02	–0.21 to 0.17	0.02	–0.17 to 0.20
Healthy (18.5–24.9 kg/m2)	0.00 (reference)		0.00 (reference)
N sites measured	0.00	0.00 to 0.01	0.01*	0.00 to 0.02
Age	0.01	0.00 to 0.01	0.00	0.00 to 0.01
Female gender	0.07	–0.08 to 0.21	0.03	–0.11 to 0.18
Insurance
None	0.00	–0.16 to 0.17	0.04	–0.15 to 0.22
Medicaid	–0.06	–0.22 to 0.10	–0.09	–0.26 to 0.08
Medicaid and private	0.20	–0.14 to 0.54	0.13	–0.19 to 0.46
Private	0.00 (reference)		0.00 (reference)
Tobacco use	–0.18	–0.43 to 0.07	–0.08	–0.30 to 0.14
History of diabetes	–0.03	–0.23 to 0.17	0.07	–0.13 to 0.27

N = 840 quadrants. Negative estimates indicate the magnitude of reduction was smaller (worse outcome) than the reference group; positive estimates indicate the reduction was greater (better outcome).

^*Statistically significant, P < 0.05.

Treatment did result in a shift in the distributions of sites with shallow and deep pockets. The percentages of sites with pocket depths up to 4 mm increased after treatment as compared to pretreatment, whereas at successively deeper pocket levels, the percentages of sites up to those thresholds declined after treatment. These trends differed by BMI status (Fig. 2A). Among patients with healthy weight, there was a steeper decline in percent sites at thresholds above 4 mm (up to 4%) compared to those who were overweight or obese (0% to 1%). A similar trend was seen for clinical attachment loss (Fig. 2B).

Figure 2.

Treatment resulted in a shift in the distributions of sites with shallow and deep pockets - the % sites with pocket depths up to 4 mm increased after treatment while the % sites with deeper pockets declined after treatment. Trends differed by BMI status. Similar results were seen for clinical attachment loss. Change in cumulative percentage of sites below pocket depth thresholds (A) or clinical attachment loss thresholds (B) after nonsurgical periodontal treatment.

Discussion

The results of this study indicate that obesity modifies the effect of nonsurgical periodontal treatment on the success of treatment among patients in a dental school clinic. Patients who were obese had only half the chance of a successful resolution of deep periodontal pockets after nonsurgical periodontal treatment compared to those of healthy weight despite having a similar extent of deep pockets at baseline. Obesity is characterized by chronic systemic inflammation, which is mediated by hormones and cytokines secreted mainly by visceral, or abdominal, fat tissue. Levels of the anti-inflammatory hormone adiponectin tend to be lower in persons who are obese than healthy weight, while leptin, which induces proinflammatory cytokines such as TNF-α and IL-6, is positively correlated with the amount of body fat, but it is not clear if these mediators play a role in the differential response to therapy. Offenbacher et al. (2009) found that nonsurgical periodontal treatment reduced the odds of an elevated serum C-reactive protein, a marker of inflammation, only in patients without obesity. Reductions in C-reactive protein and some cytokine levels were found in response to treatment regardless of obesity status (Zuza et al. 2011; Al-Zahrani and Alghamdi 2012; Altay et al. 2013; Balli et al. 2016; Toy et al. 2023). Other investigators reported no effect of periodontal treatment on the hormones leptin and adiponectin (Goncalves et al. 2015). More research is needed to clarify the mechanisms driving the poorer response to periodontal treatment in obesity.

Our results regarding outcomes of nonsurgical treatment in patients who are obese are consistent with those of several prior intervention studies (Suvan et al. 2014; Bouaziz et al. 2015; Goncalves et al. 2015; Suvan et al. 2020) despite differences in methodology and patient population characteristics. The previous studies evaluated only full-mouth nonsurgical periodontal treatment or debridement, whereas most of our patients did not require full-mouth treatment. Follow-up time in our analysis was determined by the treating clinician’s assessment of progress and included only clinical measurements taken during the time of active treatment. We evaluated treatment responses in patients who were overweight separately from healthy weight and obese. In addition, we had attempted to make our findings relevant to treatment outcomes expected in a clinical practice with a diverse patient population. Therefore, our population included patients who smoked, had systemic diseases, had undergone prior periodontal treatment, and used anti-inflammatory medications. While we controlled for the characteristics that were most strongly related to periodontal treatment outcomes (e.g., diabetes and smoking) in our analyses, there may be additional unmeasured confounders that introduce bias into our findings.

Further research is needed regarding the potential effect modification by overweight on periodontal treatment outcomes (Suvan et al. 2014). In our study, the odds of treatment success in the overweight group were not significantly different from the healthy-weight group. Suvan et al. (2014) likewise did not find differences in mean PD or CAL or sites with PD >4 mm between participants who were overweight and normal weight.

In contrast to several intervention studies (Michalowicz et al. 2009; Suvan et al. 2014; Goncalves et al. 2015), we found that neither obesity nor overweight was significantly associated with changes in mean pocket depth and attachment loss. Means levels obscure large differences that occur in only a few sites, and the variability in measurement among a large number of examiners in our study may also have affected these results. More research is also needed to determine if obesity influences the clinical outcomes after surgical procedures as well as nonsurgical treatment.

Our study used real-world clinical data from existing electronic health records, an approach that can complement evidence from randomized clinical trials (RCTs) (Bartlett et al. 2019). In contrast to RCTs, which have robust internal validity, studies using real-world clinical treatment data can yield results with strong external validity and may better reflect the clinical environments and patient populations in which treatment is typically performed (Bartlett et al. 2019). This is especially true for populations that are often excluded from RCTs, such as older adults and those with comorbid conditions (He et al. 2020).

We encountered several limitations inherent in the use of electronic health data that are not specifically collected for research purposes. Chief among them were the reliance on self-reported weight and height to calculate BMI, as well as the use of BMI itself as a surrogate for adiposity-related metabolic disturbances. Underestimation of weight and overestimation of height are common when individuals self-report (Connor Gorber et al. 2007) and may have resulted in misclassification bias of our effect modifier. However, such misclassification, to the extent it exists, would most likely bias our results to the null. That is, the bias would have attenuated the true strength of the association we found. Although BMI is widely used because it is easy to assess and predicts greater risk of many chronic diseases, it does not distinguish fat from muscle and bone or describe the distribution of body fat. Indices of abdominal obesity such as waist circumference, waist/height ratio, and sagittal abdominal diameter may be stronger indicators of risk of obesity-linked health problems than BMI but are less likely to be documented in dental clinic records. BMI was computed at baseline examination and used as a time-fixed variable given the relatively short follow-up time, although we acknowledge that it is possible to have a time-varying BMI.

Other limitations of this study include a high frequency of missing values for bleeding on probing, plaque measurements, and race/ethnicity. We also had no data on calibration of the pocket depth and attachment loss measures among the many examiners.

Finally, the patients included in this current study were a small subset of a larger cohort of periodontal clinic patients on which we previously reported the use of periodontal treatment resources by obesity status (Kaye et al. 2023; Singhal et al. 2023). To increase the likelihood that any changes in clinical measures resulted directly from the treatment procedure, we limited this current study sample to those who had initial PD and CAL measurements, BMI values, and follow-up periodontal measurements all within a short time frame. This strategy excluded a large number of patients who received nonsurgical periodontal treatment and follow-up examinations on a less stringent schedule. However, the similarities between the study sample and the larger reference population of patients attending the periodontal clinic with regard to clinical periodontal measures suggest our results can be generalized to the larger patient population. Although our patient population was diverse, which is consistent with the school’s focus on serving disadvantaged communities, the results still are only applicable to patients who seek and undergo periodontal treatment.

In conclusion, patients with obesity attending a dental school clinic were less likely to have complete resolution of deep pockets following nonsurgical periodontal treatment after accounting for differences in baseline disease severity and other confounders. Consideration of a patient’s body mass index may help both patients and providers to better understand and manage the course of periodontal therapy.

Author Contributions

E. Kaye, contributed to conception and design, data analysis and interpretation, drafted and critically revised the manuscript; R. McDonough, contributed to conception and design, data acquisition, analysis, and interpretation, critically revised manuscript; A. Singhal, R.I. Garcia, contributed to conception, data interpretation, critically revised the manuscript; M. Jurasic, contributed to conception and design, data acquisition, analysis, and interpretation, critically revised the manuscript. All authors gave final approval and agreed to be accountable for all aspects of the work.